1500 words-statistical nutrition paper due 7hours
mark21
referencesresourcesdatanotesfortheorder.zipHenry BMR equation and SACN PAL (1)(1).pdf
Limitations to using factorial method(1).pdf
Underestimation of Daily Energy Expenditure With the Factorial Method: Implications for Anthropological Research
WILLIAM R. LEONARD,1,2* VICTORIA A. GALLOWAY,2,3 AND EVGUENI IVAKINE2 1Department of Anthropology, University of Florida, Gainesville, Florida 32611 2Department of Human Biology and Nutritional Sciences, University of Guelph, Guelph, Ontario N1G 2W1, Canada 3Department of Physiotherapy, University of Toronto, Toronto, Ontario M5T 1W5, Canada
KEY WORDS energetics; factorial method; heart-rate monitoring; adaptability; Siberia
ABSTRACT Under field conditions, total daily energy expenditure (TDEE) has generally been estimated using time allocation techniques (the factorial method). However, recent work suggests that the factorial method underesti- mates TDEE relative to newer, more accurate methods such as doubly labelled water (DLW) and heart-rate (HR) monitoring. This study compares estimates of TDEE obtained by the factorial and HR-monitoring methods for a sample of 61 indigenous (Evenki; 17 males, 44 females) and 32 nonindigenous (‘‘Russian’’; 10 males, 22 females) subjects from three communities in Central Siberia. Energy expenditures obtained from the two methods were signifi- cantly correlated (r 5 0.495; P , 0.0001), but the factorial method signifi- cantly underestimated TDEE relative to the HR-monitoring technique (8.95 6 2.73 vs. 8.25 6 1.34 MJ/d; P , 0.005).
Interpopulational analyses of data compiled from this and other studies indicate that the factorial method consistently underestimates TDEE relative to DLW and HR monitoring and that the magnitude of underestimation increases with expenditure levels. Indeed, among sedentary populations, factorial estimates of TDEE converge on those of the other methods, whereas at high activity levels the disparity is quite large. These results imply that the daily energy requirements of many subsistence-level populations have been underestimated, thus providing an overly favorable picture of energy balance. Moreover, it is likely that underrepresentation of TDEE is most problematic in rural societies of the developing world which tend to have high activity levels and great risk of malnutrition. Am J Phys Anthropol 103:443–454, 1997. r 1997 Wiley-Liss, Inc.
Energetic approaches have been widely used in biological anthropology to examine adaptive strategies of human populations (Ulijaszek, 1995). Information on energy expenditure is critical for assessing dietary adequacy and the likelihood undernutrition (James and Schofield, 1990). Additionally, estimates of energy costs for daily activities have been used to assess the efficiency of subsistence behavior among hunting-and-
gathering (Hawkes et al., 1982; Smith, 1981), agricultural (Thomas, 1973; Dufour, 1983), and pastoral (Galvin, 1985) societies. Yet,
Contract grant sponsor: Natural Sciences and Engineering Research Council of Canada; Contract grant number: OGP- 0116785.
*Correspondence to: William R. Leonard, Department of An- thropology, 1350 Turlington Hall, University of Florida, Gaines- ville, FL 32611. E-mail: [email protected]
Received 21 June 1996; accepted 19 May 1997.
AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 103:443–454 (1997)
r 1997 WILEY-LISS, INC.
while information on activity patterns and energy expenditure provides important in- sights into the ecology and health of human groups, such data are difficult to obtain on traditional ‘‘free-living’’ populations.
Total daily energy expenditure (TDEE) in anthropological settings has generally been estimated by the factorial method, in which the researcher or subject records the amount of time the subject spends in various activi- ties throughout the day. The activity times are then converted into energetic equiva- lents. Activity-specific energy costs can be determined by indirect calorimetry while in the field (e.g., Thomas, 1973; Dufour, 1983) or by using standard tables (e.g., Durnin and Passmore, 1967; James and Schofield, 1990; Passmore and Durnin, 1955). TDEE is then determined by summing the energy expended in each activity throughout the day. In 1985, this approach was recom- mended by the World Health Organization (WHO) as the preferred method for estimat- ing individual and populational energy re- quirements (FAO/WHO/UNU, 1985).
Recently, however, a number of studies have raised questions about the accuracy of the factorial method for estimating TDEE. For example, work by Roberts et al. (1991) and Haggarty et al. (1994) on American and British men found that the factorial method significantly underestimated TDEE relative to the doubly labelled water method, the technique generally regarded as the most accurate for measuring free-living energy costs. Similarly, studies by Leonard et al. (1995), Spurr et al. (1996), and Dufour et al. (1996) all showed the that the factorial method underrepresented TDEE relative to estimates obtained from daily heart-rate (HR) monitoring. Together, these studies suggest that the factorial method may sys- tematically underestimate TDEE when com- pared to other, more accurate methods.
The purpose of this research is to further evaluate the utility of the factorial method for estimating TDEE under field conditions. First, we compare estimates of TDEE ob- tained by the factorial and HR-monitoring methods for men and women of two different ethnic groups (indigenous Evenki and nonin- digenous Russians) living in rural communi- ties of Central Siberia. Next, interpopula-
tional analyses are presented using these and previously published data to examine patterns of bias associated with the factorial method. The implications of these results for our understanding of energy dynamics in subsistence-level human populations are then discussed.
METHODS Sample
This research was conducted among indig- enous (Evenki) and nonindigenous (Rus- sian) subjects living in three communities from the Baykit District of the Stony Tun- guska region of Central Siberia—Surinda, Poligus, and Baykit (see Fig. 1). Both Surinda and Poligus are relatively small villages, which have remained as the administrative centers for several Evenki reindeer-herding brigades (Leonard et al., 1994). As of 1994, the populations of Surinda and Poligus were 613 and 481, respectively. According to gov- ernmental records, 95% of the Surinda popu- lation was indigenous (largely Evenki), as compared to 48% in Poligus. In contrast, Baykit, the district capital, is a larger and more urbanized town. According to the 1994 census, the population of Baykit was 5,187, most of whom were nonindigenous (93%).
Anthropometric, dietary, activity recall, and HR data were collected on a sample of 61 Evenki (17 males, 44 females) and 32 Russian (10 males, 22 females) subjects. The male subjects ranged in age from 13–57 years, whereas the females ranged from 14–59. The mean ages of the Evenki and Russian subjects do not significantly differ in either sex. All data were collected in the community health posts from July to Sep- tember of 1995. The research protocol was approved by the Human Subjects Review Committee of the University of Guelph.
Anthropometry
Anthropometric dimensions included stat- ure (in centimeters), body weight (in kilo- grams), and skinfold measures (in millime- ters) at the triceps and subscapulum. Stature was measured to the nearest millimeter using a portable field anthropometer, and body mass was measured to the nearest 0.5 kg using a standing scale (Seca Corp., Colum-
444 W.R. LEONARD ET AL.
Fig. 1. Map showing the locations of the Stony Tunguska region within Central Siberia (A) and the three study communities—Surinda, Poligus, and Baykit (B). Modified from Katzmarzyk (1993).
445UNDERESTIMATION OF ENERGY EXPENDITURE
bia, MD). Skinfold measurements were taken with Lange callipers (Cambridge Scientific, Cambridge, MD) and were recorded to the nearest 0.5 mm. All measurements were taken by a single observer (W.R.L.) using the techniques described in Lohman et al. (1988).
Energy expenditure
The factorial and HR-monitoring methods were used to assess TDEE for the same day on each subject.
Factorial method. A 24 h activity recall was administered to each subject by E.I. and V.A.G. Each subject was asked about the times that he or she awoke and went to sleep and was then queried about activities dur- ing half-hour blocks of his or her waking day.
In coding the data, activities were divided into eight major categories as derived from James and Schofield (1990) and outlined in Table 1. Each category was assigned a physi- cal activity ratio (PAR) reflecting the energy cost as a multiple of basal metabolic rate (BMR). Basal requirements for each subject were calculated from body weight using age- and sex-specific regression equations com- piled in the WHO’s most recent protein and energy recommendations (FAO/WHO/UNU, 1985). The amount of time spent in activities of each category was then multiplied by the energetic cost of the activity level and summed to obtain an estimate of TDEE.
Heart-rate monitoring. Energy expendi- ture (EE) from HR monitoring was assessed
using the flex-HR technique of Spurr and colleagues (1988). With this method, indi- vidual HR vs. expenditure relationships are first established for each subject before hav- ing them wear an HR monitor for an entire active day. Heart rates were recorded with a Polar Vantage XL HR monitor (Polar Elec- tro, Stamford, CT). EE (kilojoules/minute) for basal, resting, and exercising conditions was determined using the Aerosport TEEM 100 Metabolic System (Aerosport Inc., Ann Arbor, MI). The oxygen (O2) and carbon dioxide (CO2) analyzers were calibrated with external air and tanks of compressed gas containing 16.02% O2 and 5.03% CO2 (Scott Specialty Gases, Troy, MI). Expired volumes were measured using pneumotachometers that were calibrated with a 3 liter syringe (Hans Rudolph, Inc., Kansas City, MO). Energy costs for the basal and resting condi- tions were measured using the low flow pneumotach heads, whereas exercising en- ergy costs were monitored with the medium flow head.
During the evening prior to the metabolic measurements, subjects came into the health post and were informed of the research protocol. Additionally, demographic and di- etary information were collected on each subject. Subjects were also given a chance to familiarize themselves with the equipment (e.g., HR monitors, mouthpieces, nose clips). After the briefing session, the subjects slept overnight in the health post.
Basal metabolic rates were measured shortly after the subjects arose the following morning. All subjects had been in a fasted state for at least 10 h when measured, and, when necessary, a portable space heater was used to heat the room to insure thermoneu- tral conditions. Basal energy costs were determined as the average of the final 8 min of measurement, after the subject had be- come accustomed to the apparatus, as evi- denced by stabilization of the HR, oxygen consumption (VO2), and respiratory quo- tient (RQ) values. Following the basal mea- surements, energy costs of three resting positions (lying, sitting, and standing) were measured for 3 min each. Resting metabolic rate (RMR) was determined as the average costs for those three positions.
TABLE 1. Activity categories used for estimating energy expenditure from time allocation data
Activity category PAR1 Selected examples
Sleep 1.0 1.2 Sitting eating, relaxing
Light 1.4 Visiting friends, watching TV, boiling water
1.6 Washing, dressing, washing dishes, sewing, walking with no load
Moderate 2.1 Household chores, cooking food, feeding animals
2.8 Washing clothes, picking fruit, sweeping floor
Heavy 3.8 Carrying firewood, carrying water
Very heavy 5.1 Running; walking with a load (20 kg)
1 PAR, physical activity ratio 5 [energy cost of the activity/basal metabolic rate].
446 W.R. LEONARD ET AL.
HR vs. EE relationships during exercise were assessed while subjects performed a graded stepping test (Fitness Canada, 1986). The steps consisted of a double riser, each 21 cm high, making for a total ascent of 42 cm. Subjects performed a 3 min bout of exercise before proceeding to the next level of the test, each successive level requiring a faster cadence. Each subject continued until at least three bouts had been completed. En- ergy costs were measured during the final minute of each exercise bout and HR re- corded at the 2.5 minute point, once it had stabilized.
An EE-HR relationship for each subject was plotted, and the least squared regres- sion line for the exercising points was calcu- lated. A flex-HR point was determined as the mean of the highest resting HR and the lowest exercising HR.
Once the EE-HR relationship for a subject had been determined, he/she was fitted with a HR monitor which was worn for an entire day. At the completion of the 24 h period, the subject’s HR data were directly downlowded to a portable computer. Small gaps in the HR profile during the active day were filled with the average of the HRs on either side of the gap. The recorded HRs were converted to energetic equivalents by comparing them to the subject’s EE-HR relationship. HRs above the flex point were converted using the least squared regression line through the exercise points. For HRs at or below the flex, RMR was assigned as the energy cost. Energy expenditure while sleeping was as- sumed to equal the measured basal meta-
bolic rate (after Goldberg et al., 1988). Total daily energy expenditure was then calcu- lated by summing the active, resting, and sleeping components of the day with the equation TDEE 5 AEE 1 REE 1 SEE, where TDEE is the total daily energy expen- diture, AEE is the active energy expenditure (for HRs . flex), REE is the resting energy expenditure (for HRs # flex), and SEE is the sleep energy expenditure.
Statistical methods
Mean differences between the HR and factorial estimates were compared using paired t-tests (one-tailed). Pearson product- moment correlations were used to assess the concordance between the two measures of energy expenditure. Agreement between the two methods was also measured using the statistical approach outlined by Bland and Altman (1986). All analyses were performed using SPSS (SPSS, Chicago, IL) version 4.0.
RESULTS
Table 2 compares the anthropometric char- acteristics of the Evenki and Russian samples. Not surprisingly, Russian men and women are significantly taller, heavier, and fatter than their Evenki counterparts. Rus- sian men are 13 cm taller and almost 20 kg heavier than Evenki men. Among women, the Russian-Evenki differences are over 7 cm in height and almost 14 kg in mass. Even when weight is adjusted for stature using the BMI, the Russians are significantly heavier than the Evenki. Likewise, the Rus- sians are also relatively fatter, as evidenced
TABLE 2. Anthropometric dimensions for male and female Evenki and Russian subjects of the Baykit District of Central Siberia1
Sample n Age (years) Weight (kg) Stature (cm) BMI (kg/m2) Sum skin2 (mm)
Males Evenki 17 31.5 6 11.4 53.9 6 8.0 159.7 6 8.5 21.1 6 2.8 14.2 6 3.1 Russian 10 35.9 6 13.5 72.4 6 14.8** 172.7 6 7.6*** 24.0 6 3.1* 23.5 6 6.2***
Total 27 33.1 6 12.1 60.7 6 14.1 164.5 6 10.3 22.2 6 3.2 17.6 6 6.3
Females Evenki 44 31.3 6 11.4 51.0 6 8.3 150.3 6 5.4 22.7 6 3.8 34.7 6 15.0 Russian 22 32.6 6 13.9 64.7 6 15.8*** 157.7 6 6.2*** 26.0 6 6.0* 45.4 6 20.7*
Total 66 31.7 6 12.2 55.6 6 13.0 152.8 6 6.6 23.8 6 4.9 38.3 6 17.7 1 Values expressed as mean 6 SD. 2 Sum of triceps and subscapular skinfold measures. * Evenki-Russian difference is significant at P , 0.05. ** Evenki-Russian difference is significant at P , 0.01. *** Evenki-Russian difference is significant at P , 0.001.
447UNDERESTIMATION OF ENERGY EXPENDITURE
by significantly higher sum of skinfold mea- sures. Comparisons to US normative data for BMI and skinfolds (from Frisancho, 1990) indicate that males of both ethnic groups tend to be relatively lighter and leaner than their female counterparts.
Comparison of the factorial and HR methods
Table 3 presents the metabolic param- eters for the Russian and Evenki subjects. Estimates of TDEE using the HR method are significantly higher than those derived by the factorial method (8.95 6 2.73 MJ (2,139 kcal) vs. 8.25 6 1.34 MJ (1972 kcal); P , 0.005). Daily expenditure levels among males are, on average, 10.33 MJ (2,470 kcal) for the HR method and 9.27 MJ (2,216 kcal) for the factorial method (P , 0.01). The dif- ferences between the methods are some- what smaller for females, as the HR method gave an average of 8.38 MJ (2,003 kcal) compared to 7.83 MJ (1,872 kcal) with the factorial approach (P , 0.05).
Some of the differences in TDEE between the two methods may reflect the differences between measured and predicted BMR. For the entire male sample, measured BMR is significantly higher than that predicted from the FAO/WHO/UNU (1985) norms (7.24 vs. 6.69 MJ; P , 0.05). Among the women, on the other hand, measured BMRs are compa- rable to the predicted values for the entire sample and significantly lower among the Russians. When the measured rather than the predicted BMRs are used in the factorial estimates, the HR estimates of TDEE re-
main significantly higher (8.95 6 2.73 vs. 8.38 6 2.11 MJ; P , 0.01).
Expressing daily energy expenditure as a multiple of basal requirements provides a measure of activity level (i.e., physical activ- ity level (PAL) 5 TDEE/BMR). The FAO/ WHO/UNU (1985) guidelines for energy re- quirements have suggested that a PAL of 1.4 is necessary for minimum maintenance ac- tivities and that those required for light, moderate, and heavy occupational activities are 1.55, 1.78, and 2.10, respectively, in men and 1.56, 1.64 and 1.82 in women. From the data presented in Table 3, we see that men and women of both ethnic groups appear to have relatively sedentary lifestyles. For all four groups, PAL values obtained from the HR method are higher than those of the factorial approach. These differences are statistically significant for Evenki women (PAL 5 1.59 vs. 1.42; P , 0.01) as well as for the entire female sample (PAL 1.57 vs. 1.42; P , 0.01) and the total sample (1.54 vs. 1.41; P , 0.01).
The relationship between the factorial and HR estimates of TDEE is presented in Figure 2. Estimates from the two methods are significantly correlated (r 5 0.495; P , 0.0001). The slope of the regression line, however, is significantly shallower than 1 (b 5 0.244 6 0.045), indicating that the fac- torial method systematically underestimates TDEE in this sample. A comparable pattern is seen when the HR estimates of TDEE are compared to the factorial estimates based on measured BMR. In this latter case the corre- lation between the HR and factorial esti-
TABLE 3. Measures of energy expenditure for Evenki and Russian men and women1
Sample
BMR (MJ/d) TDEE (MJ/d) PAL
Measured Estimated HR Factorial HR Factorial
Males Evenki 6.84 6 1.45 6.31 6 0.49 9.99 6 2.82 8.91 6 1.56* 1.48 6 0.37 1.41 6 0.22 Russian 7.92 6 1.34 7.32 6 0.87 10.92 6 2.32 9.89 6 1.21 1.39 6 0.25 1.36 6 0.15
Total 7.24 6 1.48 6.69 6 0.81* 10.33 6 2.65 9.27 6 1.50** 1.44 6 0.33 1.39 6 0.20
Females Evenki 5.39 6 1.03 5.30 6 0.40 8.45 6 2.54 7.51 6 0.88** 1.59 6 0.50 1.42 6 0.14** Russian 5.45 6 0.79 5.93 6 0.51** 8.26 6 2.70 8.48 6 0.99 1.53 6 0.52 1.43 6 0.11
Total 5.41 6 0.95 5.51 6 0.53 8.38 6 2.60 7.83 6 1.02* 1.57 6 0.51 1.42 6 0.13** 1 Values expressed as mean 6 SD. * Differences between pairs of energy expenditure measures are significant at P , 0.05. ** Differences between pairs of energy expenditure measures are significant at P , 0.01.
448 W.R. LEONARD ET AL.
mates is higher (r 5 0.588; P , 0.0001), while the slope of the regression is steeper but still significantly less than 1 (b 5 0.453 6 0.065).
Differences between the factorial and HR estimates are examined further in Figure 3, which compares the two techniques using the statistical method developed by Bland and Altman (1986). In this figure, the facto- rial estimates are those derived using esti- mated BMRs. For each subject, the differ- ence between the two estimates (factorial 2 HR) is plotted against their av- erage. The factorial estimates average 0.70 6 2.37 MJ (167 kcal) less than the HR esti- mates, an underestimation of about 8%. Additionally, the plot demonstrates that the magnitude of underestimation is signifi- cantly correlated with absolute expenditure levels (r 5 20.664; P , 0.001). This implies that the factorial method does not provide an unbiased estimate of TDEE relative to HR monitoring; rather, the degree of under- estimation increases with greater expendi-
ture levels. Indeed, for average TDEEs un- der 10 MJ/day, there is relatively high concordance between the two methods, with the estimates differing by only 0.5% (.04 MJ). Above 10 MJ, on the other hand, the two methods depart by 29% (3.77 MJ).
Carrying out the Bland-Altman analysis using the factorial estimates based on mea- sured BMR produces similar results. In this case, the factorial estimates are 0.57 6 2.27 MJ (136 kcal) lower than the HR estimates (,7% underestimation). As with the previ- ous analysis, the degree of underestimation is significantly correlated with absolute ex- penditures; however, the level of association is more modest (r 5 20.309; P , 0.01).
Interpopulational comparisons
Table 4 presents PAL estimates compiled from recent studies that have compared the factorial approach to other methods of indi- rect calorimetry (either DLW or HR monitor- ing) under free-living conditions. In addition to the information presented here, compara-
Fig. 2. Relationship between the factorial vs. HR estimates of energy expenditure. Estimates from the two methods are highly correlated (r 5 0.495; P , 0.0001); however, the slope of the regression is significantly shallower than identity (b 5 0.244 6 0.045; P , 0.001).
449UNDERESTIMATION OF ENERGY EXPENDITURE
tive data were available for samples of American and British men (from Roberts et al. (1991) and Haggarty et al. (1994), respec- tively), men and women from farming com-
munities of highland and coastal Ecuador (from Leonard et al., 1995), and wage- earning and non-wage-earning urban Colom- bia women (from Spurr et al., 1996). For all
Fig. 3. Difference in energy expenditure (factorial 2 HR) vs. mean energy expenditure for the HR and factorial methods. The factorial method underestimates TDEE by an average of 0.70 6 2.37 MJ, and the differences between the methods are significantly correlated with absolute expenditure levels (r 5 20.664; P , 0.001).
TABLE 4. Comparison of physical activity level (PAL) estimates using the factorial method vs. other methods of indirect calorimetry (DLW or HR monitoring) for selected populations
Group Sex n Comparison
method PAL11 PAL22 Reference
US adults M 14 DLW 1.56 1.98 Roberts et al. (1991) British adults M 17 DLW 1.79 1.96 Haggarty et al. (1994) Ecuador
Highland farmers M 11 HR monitoring 1.95 2.39 Leonard et al. (1995) F 11 HR monitoring 1.71 1.97
Coastal farmers M 5 HR monitoring 1.43 1.58 Leonard et al. (1995) F 5 HR monitoring 1.56 1.62
Colombia Urban adults F 29 HR monitoring 1.47 1.83 Spurr et al. (1996) Urban adults F 23 HR monitoring 1.50 1.90
Siberia Evenki villagers M 17 HR monitoring 1.41 1.48 Present study
F 44 HR monitoring 1.42 1.59 Russian villagers M 10 HR monitoring 1.36 1.39 Present study
F 22 HR monitoring 1.43 1.53 Mean 6 SD 1.55 6 0.18 1.77 6 0.29* 1 PAL1, physical activity level estimate using the factorial method. 2 PAL2, physical activity level estimate using the DLW or HR-monitoring method. * PAL estimates different at P , 0.001.
450 W.R. LEONARD ET AL.
12 sex-specific samples, the PALs obtained from the factorial method are lower than those derived from the comparator method (either DLW or HR monitoring). On average, the factorial PALs were significantly lower than the HR and DLW estimates (1.55 vs. 1.77; P , 0.001).
The data from Table 4 are presented graphically in Figure 4. This plot shows that despite the systematic pattern of underesti- mation with the factorial approach, the fac- torial estimates are highly correlated with those derived from the DLW and HR meth- ods (r 5 0.890; P , 0.0001). The slope of the regression line, however, is significantly less than 1 (b 5 0.557 6 0.090; P , 0.01), again suggesting that the degree of underestima- tion is greater at higher expenditure levels. This analysis suggests that at very low activity levels (PALs # 1.3), the factorial estimates converge with those of other meth- ods. At moderate to high activity levels (PALs $ 1.7), on the other hand, there is
substantial underestimation with the facto- rial approach.
DISCUSSION
Results of the present study are consis- tent with much recent work suggesting that the factorial method systematically underes- timates total daily energy expenditure in free-living populations (see Dufour et al., 1996; Haggarty et al., 1994; Leonard et al., 1995; Roberts et al., 1991; Spurr et al., 1996). There are a number of possible expla- nations for the apparent disparity between estimates obtained from the factorial ap- proach relative to other, more accurate meth- ods of assessing TDEE. One source of the bias may be the use of activity recalls rather than direct observations for determining the factorial estimates. However, this does not appear to be the entire explanation. Spurr et al. (1996) found substantial underestima- tion of TDEE even with factorial estimates that are based on continuous minute-by-
Fig. 4. Relationship between factorial and DLW or HR-monitoring estimates of physical activity level (PAL) for the 12 sex-specific samples presented in Table 4. The factorial estimates of PAL are highly correlated with those derived from the other methods of indirect calorimetry (r 5 0.890; P , 0.0001); however, the slope of the regression is significantly less than 1 (b 5 0.557 6 0.090; P , 0.01).
451UNDERESTIMATION OF ENERGY EXPENDITURE
minute observations. Interpopulational varia- tion in BMR is widely documented (e.g., Roberts, 1978; Henry and Rees, 1991; Rode and Shephard, 1995) and may contribute to underestimation of TDEE with the FAO/ WHO/UNU’s (1985) factorial method. Re- sults presented here suggest that differ- ences in TDEE between the two methods are partially attributable to differences in BMR. Using measured rather the predicted BMRs in the factorial estimates resulted in a 19% reduction in the average level of underesti- mation; however, these factorial estimates remained significantly lower than those ob- tained from the HR method. Additionally, in our previous study among Ecuadorian farm- ers (Leonard et al., 1995), we utilized pre- dicted BMRs for both our HR and factorial estimates and still found significant differ- ences between the methods. Thus, it ap- pears that variation in BMR cannot entirely explain the systematic underestimation of TDEE by the factorial method.
It has also been suggested that because much of the standard reference data for activity costs have been derived on Western populations (e.g., Durnin and Passmore, 1967; FAO/WHO/UNU, 1985; James and Schofield, 1990) these data cannot be effec- tively applied to populations of different social and ethnic backgrounds (Dufour et al., 1996; Durnin, 1990; Spurr et al., 1996). Similarly, Spurr et al. (1996) found that among Colombian women the FAO/WHO/ UNU’s (1985) values for several activities gave energy costs that were significantly lower than those measured by indirect calo- rimetry. In contrast, Katzmarzyk et al. (1996) found that in three subsistence-level popula- tions (Siberians and highland and coastal Ecuadorians), the energy costs of a submaxi- mal stepping exercise did not significantly deviate from those predicted from Canadian normative values. These authors suggested that the underestimation of TDEE with the factorial method in traditionally living popu- lations may largely reflect error in estimat- ing energy costs at rest (i.e., lying, sitting, standing) rather than during activity and exercise.
Regardless of what the source of bias is, the current results have enormous implica- tions for research in human population biol-
ogy and nutrition. These findings seem to confirm the suspicion of Durnin (1990) and others who have suggested that low expendi- tures reported for many free-living popula- tions are physiologically unlikely. In the anthropological literature, for example, ex- tremely low TDEE and PAL estimates have been reported for such populations as the !Kung San foragers of the Kalahari (Lee, 1979; Leslie et al., 1984), the Turkana pasto- ralists of Kenya (Galvin, 1985; Little and Gray, 1990), and the Quechua agropastoral- ists of Peru (Leonard, 1992; Thomas, 1973). In light of the present findings, it is probable that the daily energy requirements of these groups have been underestimated, thus pro- viding an overly favorable picture of popula- tion energy balance.
The problem of underestimating energy needs (and overestimating dietary adequacy) is likely to be most acute among highly active groups. This study, as well as those of Haggarty et al. (1994) and Leonard et al. (1995), demonstrated that the level of under- estimation of the factorial method increases with TDEE. It would therefore seem that the errors associated with the factorial method will be greatest when applied to rural developing-world populations, who tend to have higher daily activity levels than populations of the developed world (see Schulz and Scholler, 1994). This implies that the WHO’s recommended method for esti- mating TDEE is most likely to underrepre- sent the problem of energy stress among those populations that are at greatest risk for malnutrition (i.e., rural, developing- world agrarian societies).
Beyond these issues of daily energy re- quirements and energy balance, the present results also have implications for examining the efficiency of human subsistence regimes. For example, the application of optimal for- aging theory for understanding subsistence behavior among hunting and gathering populations has generally relied on stan- dard reference data such as Durnin and Passmore (1967) to determine the energetic costs and benefits of exploiting different resources (see Smith, 1981; Winterhalder, 1981). Potential bias in estimating the en- ergy costs of foraging tasks should therefore tend to skew calculations of the relative
452 W.R. LEONARD ET AL.
benefits of exploiting different potential re- sources, leading to errors in predictions about which resources are to be included in an ‘‘optimal diet.’’
Yet, despite the clear limitations in the factorial method, it does appear to effec- tively determine relative activity levels within and among different populations. The high correlations between energy expendi- ture estimates obtained by the factorial and the DLW or HR methods indicate that the factorial approach is able to rank relative daily expenditure levels across individuals and groups. Although the factorial method may not be able to quantify TDEE with a high degree of accuracy, it can effectively discriminate groups that are highly active from those that are moderately active and sedentary.
In conclusion, this study has shown that the factorial approach significantly and sys- tematically underestimates total daily en- ergy expenditure both within and between human populations relative to other meth- ods of indirect calorimetry used with free- living conditions. The degree of underestima- tion is related to absolute expenditure levels, such that relatively greater bias is seen at higher TDEEs. These results raise concern about the utility of the WHO’s recommended methodology for assessing human energy needs, since it may tend to mask the severity of energy stress evident in populations of the developing world. They also suggest that estimates of dietary adequacy reported for many traditional, subsistence-level popula- tions must be viewed with some caution. While the factorial method does effectively sort groups into relative activity levels, it underrepresents TDEE, especially at high expenditure levels.
ACKNOWLEDGMENTS
This study was conducted in collaboration with Drs. Ludmilla Osipova (Institute of Cytology and Genetics, Russian Academy of Science, Novosibirsk). We are grateful to all the subjects who participated in this study. Additionally, we thank Mark Smith (Aero- sport, Inc.) and Dr. Brian Wilson for techni- cal assistance with the TEEM 100 and Anne Keenleyside and Marina Kazakovtseva for assistance with data collection in the field.
Comments by Drs. Marcia Robertson and Catherine Ponter-Brick and two anonymous reviewers substantially improved the paper.
LITERATURE CITED
Bland JM, and Altman DG (1986) Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1:307–310.
Dufour DL (1983) Nutrition in the Northwest Amazon: Household dietary intake and time-energy expendi- ture. In RB Hames and WT Vickers (eds.): Adaptive Responses of Native Amazonians. New York: Aca- demic Press, pp. 329–355.
Dufour DL, Haught TA, and Spurr GB (1996) Estimat- ing daily energy expenditure using the factorial meth- od: A comparison of published values. Am. J. Hum. Biol. 8:114 (abstract).
Durnin JVGA (1990) Low energy expenditures in free- living populations. Eur. J. Clin. Nutr. 44(Suppl. 1):95– 102.
Durnin JVGA, and Passmore R (1967) Energy, Work and Leisure. London: Heineman Educational Books.
FAO/WHO/UNU (Food and Agriculture Organization/ World Health Organization/United Nations Univer- sity) (1985) Energy and Protein Requirements. WHO Technical Support Series, No. 724. Geneva: World Health Organization.
Fitness Canada (1986) Canadian Standardized Test of Fitness (Operations Manual). Ottawa: Government of Canada.
Frisancho AR (1990) Anthropometric Standards for the Assessment of Growth and Nutritional Status. Ann Arbor, MI: University of Michigan Press.
Galvin KA (1985) Food Procurement, Diet, Activities and Nutrition of Ngisonyoka Turkana Pastoralists in an Ecological and Social Context. Ph.D. dissertation, SUNY-Binghamton.
Goldberg GR, Prentice AM, Davies HL, Murgatroyd PR (1988) Overnight and basal metabolic rates in men and women. Eur. J. Clin. Nutr. 42:137–144.
Haggarty P, McNeill G, Abu Manneh MK, Davidson L, Milne E, Duncan G, and Ashton J. The influence of exercise on the energy requirements of adult males in the UK. Brit. J. Nutr. 72:79.
Hawkes K, Hill K, and O’Connell JF (1982) Why hunters gather: Optimal foraging and the Ache of Eastern Paraguay. Am. Ethnol. 9:379–398.
Henry CJK, and Rees DG (1991) New predictive equa- tions for the estimation of basal metabolic rate in tropical peoples. Eur. J. Clin. Nutr. 45:177–185.
James WPT, and Schofield EC (1990) Human Energy Requirements: A Manual for Planners and Nutrition- ists. Oxford: Oxford University Press.
Katzmarzyk PT (1993) The Adaptive Significance of Energy Expenditure among Indigenous Siberian Popu- lations. M.Sc. thesis, University of Guelph.
Katzmarzyk PT, Leonard WR, Stephen MA, Berti PR, and Ross AGP (1996) Differences between observed and predicted energy costs at rest and during exercise in three subsistence-level populations. Am. J. Phys. Anthropol. 99:537–545.
Lee RB (1979) The !Kung San: Men, women, and work in a foraging society. Cambridge: Cambridge University Press.
Leonard WR (1992) Age and sex differences in the impact of seasonal energy stress among Andean agri- culturalists. Hum. Ecol. 19:351–368.
Leonard WR, Katzmarzyk PT, Comuzzie AG, Crawford MH, and Sukernik RI (1994) Growth and nutritional status of the Evenki reindeer herders of Central Siberia. Am. J. Hum. Biol. 6:339–350.
453UNDERESTIMATION OF ENERGY EXPENDITURE
Leonard WR, Katzmarzyk PT, Stephen MA, and Ross AGP (1995) Comparison of the heart rate-monitoring and factorial methods: Assessment of energy expendi- ture in highland and coastal Ecuadoreans. Am. J. Clin. Nutr. 61:1146–1152.
Leslie PW, Bindon JR, and Baker PT (1984) Caloric requirements of human populations: A model. Hum. Ecol. 12:137–162.
Little MA, and Gray SJ (1990) Growth of young nomadic and settled Turkana children. Med. Anthropol. Q. 4:296–314.
Lohman TG, Roche AF, and Martorell R (eds.) (1988) Anthropometric Standardization Reference Manual, abridged ed. Windsor, Canada: Human Kinetics.
Passmore R, and Durnin JVGA (1955) Human energy expenditure. Physiol. Rev. 35:801–839.
Roberts DF (1978) Climate and Human Variability, 2nd ed. Menlo Park, CA: Cummings.
Roberts SB, Heyman MB, Evans WJ, Fuss P, Tsay R, and Young VR (1991) Dietary energy requirements of young adult men, determined by using the doubly labeled water method. Am. J. Clin. Nutr. 54:499–505.
Rode A, and Shephard RJ (1995) Basal metabolic rate of Inuit. Am. J. Hum. Biol. 7:723–729.
Schulz LO, and Schoeller DA (1994) A compilation of total daily energy expenditures and body weights in healthy adults. Am. J. Clin. Nutr. 60:676–681.
Smith EA (1981) The application of optimal foraging theory to the analysis of hunter-gatherer group size.
In B Winterhalder and EA Smith (eds.): Hunter- Gatherer Foraging Strategies: Ethnographic and Ar- cheological Analyses. Chicago: University of Chicago Press, pp. 36–65.
Spurr GB, Prentice AM, Murgatroyd PR, Goldberg GR, Reina JC, and Christman NT (1988) Energy expendi- ture from minute-by-minute heart-rate monitoring: Comparison with indirect calorimetry. Am. J. Clin. Nutr. 48:552–559.
Spurr GB, Dufour DL, and Reina JC (1996) Energy expenditure of urban Columbian women: A compari- son of patterns and total daily energy expenditure by the heart rate and factorial methods. Am. J. Clin. Nutr. 63:870–878.
Thomas RB (1973) Human adaptation to a high Andean energy flow system. Occasional Papers in Anthropol- ogy, No. 7. University Park, PA: Pennsylvania State University.
Ulijaszek SJ (1995) Human Energetics in Biological Anthropology. Cambridge: Cambridge University Press.
Winterhalder B (1981) Foraging strategies in the boreal forest: An analysis of Cree hunting and gathering. In B Winterhalder and EA Smith (eds.): Hunter-Gath- erer Foraging Strategies: Ethnographic and Archeo- logicalAnalyses. Chicago: University of Chicago Press, pp. 66–98.
454 W.R. LEONARD ET AL.
goran 2005 Public Health Nutrition(1).pdf
Estimating energy requirements: regression based prediction equations or multiples of resting metabolic rate
Michael I Goran* Institute for Prevention Research and Departments of Preventive Medicine and Physiology & Biophysics, Keck School of Medicine, University of Southern California, Los Angeles, CA, USA
Abstract
Energy requirements have traditionally been determined based on multiples of resting metabolic rate (RMR), known as Physical Activity Levels (PAL). With more data from doubly labelled water studies alternative approaches for estimating energy requirements have been suggested. Statistical analysis reveals that body weight explains more of the variance in total energy expenditure (TEE) than does RMR. The explanation for this phenomenon is that body weight contributes to the variance of both RMR and the other major determinant of TEE, i.e. physical activity related energy expenditure. Thus, in effect, the regression-based approach provides a more physiological appropriate model for TEE. Its major departure from tradition, difference from current adult proposals, and time taken for acceptance are the disadvantages of the regression-based approach.
Keywords Resting metabolic rate Physical activity level
Regression based approach
Introduction
Energy requirements have traditionally been determined
based onmultiples of resting metabolic rate (RMR), known
as Physical Activity Levels (PAL). This approach was
developed prior to the availability of doubly labelled
water, which provides a more direct estimate of energy
requirements as a measure of total energy expenditure
(TEE). As more doubly labelled water data becomes
available, alternative approaches for estimating energy
requirements have been suggested. The purpose of this
paper is to provide a comparison of advantages and
disadvantages of these two approaches for estimating
energy requirements (see Table 1).
Regression based approach
The regression based approach for estimating TEE has
been suggested in various other studies1,2, as well as, in
the analysis of data for infants3 and children4 for this
consultation in this series of papers. This approach uses
multiple regression techniques to develop prediction
models of TEE as a function of measured predictor
variables such as body weight, age, RMR, etc. The
regression approach is a major departure from the
convention approach of predicting energy requirements
as multiples of RMR.
There are several advantages of the regression based
approaches. The first major advantage is that this approach
is evidence based and has been developed based on a
pool of actual measures of TEE and potential predictor
variables. It is worthy to note that both Butte3 and Torun4
worked independently to review existing data from infants
and children, but yet derived a very similar approach for
estimating total energy requirements using a simple, yet
effective prediction model. Both prediction equations
were based on simple available measures (body weight)
and provided relatively accurate predictions of TEE.
Interestingly, the data sets that have been used to define
regression approaches for TEE are significantly larger than
the data sets used elsewhere in this series of papers to
derive prediction equations for RMR in children. Clearly
further studies and analysis will be required to expand the
data sets into more heterogeneous groups, as well as
perform cross-validations of the proposed prediction
equations on independent data sets.
The statistical analysis is revealing in that clearly body
weight explains more of the variance in TEE than does
RMR, as has been observed in other analysis where RMR
typically only explains ,50% of the variance in TEE5. The explanation for this phenomenon is that body weight
contributes to the variance of both RMR and the other
major determinant of TEE, i.e. physical activity related
energy expenditure. Thus, in effect, the regression based
approach provides a more physiological appropriate
model for TEE, as compared to the PAL approach which
assumes TEE is composed of multiples of RMR.
The disadvantages of the regression based approach are
that it is a major departure from tradition, is different from
current adult proposals, and will take time to be accepted.
In addition, current equations are based mainly on weight
(several also include height, age and gender), and may
q The Author 2005*Corresponding author: Email [email protected]
Public Health Nutrition: 8(7A), 1184–1186 DOI: 10.1079/PHN2005803
need to be modified with other variables to specify
different activity levels. Nevertheless, since both the
regression based approach and the PAL approach are
based mainly on weight, a direct comparison of the two
approaches can be easily obtained.
PAL approach (multiples of RMR)
The major advantage of the PAL approach is that it is a
traditional and commonly used approach. On the surface
this approach is simple to use but a closer analysis of this
approach reveals several limitations.
First the PAL approach is actually awkward to use in that
a two-step prediction is required that could potentially
compound prediction error. In the first step, RMR is
predicted using body weight based equations. The
equations proposed are based on a limited data set for
children and infants. In the second step a PAL is applied
and the predicted RMR is multiplied by that factor to obtain
an estimate of energy requirements. The PALs that are
frequently used for light, moderate, and heavy physical
activity are somewhat arbitrary, and were not developed
specifically for use in children. In general, the PAL
approach may not be statistically or mathematically
appropriate because it is a ratio (PAL ¼ TEE:RMR). In other words the PAL model assumes a linear relationship
between TEE and RMR that has a slope equivalent to PAL
and a non-zero intercept. Several studies, including a
meta-analysis6 reveal that the relationship between TEE
and RMR is linear but has a variable slope, and more
importantly, a variable and frequently non-zero intercept.
The non-zero intercept is important since under this
condition use of a constant PAL results in spurious
differences when subjects are compared across the range
of TEE levels (see Fig. 1). In an analysis of 574 measures of
TEE7, the PAL index was validated by showing that it was
uncorrelated with RMR, but it was not clear whether the
TEE vs. RMR relationship satisfied the non-zero intercept
requirement.
From a physiological perspective the assumption of
multiples of RMR is also an inaccurate model of TEE, since
TEE is composed of the sum of various components
Table 1 Summary of strengths and limitations of the regression based approach and the PAL ratio approach for estimating energy requirements
Strengths Limitations
Regression based Evidence based (i.e. based on measures of TEE) Major departure from tradition Several researchers have independently proposed this new approach
If used for children may not be consistent with adults Current equations mainly limited to weight based
Simple and practical to implement prediction and may need modification to specify activity Typically based on body weight which predicts more of the variance in TEE than RMR
Physiologically sound Statistically and mathematically sound
PAL ratio approach Traditional approach Need to first predict RMR then apply PAL factor Commonly used and widely accepted 2-stage prediction may compound prediction error
Clumsy to apply PALs are arbitrary and have not been specifically developed for children and infants
Different PALs are not needed for infants due to limited range of activity
PAL is highly variable, even within heterogeneous groups PAL ratio violates mathematical assumption of non-zero intercept
Not based on sound physiological model of TEE (should be sum of components, not multiples of one component)
Assumes physical activity energy cost is dependent on the same factors as for RMR
Abbreviations: PAL – Physical Activity Level; TEE – total energy expenditure; RMR – resting metabolic rate.
Fig. 1 Spurious ratios when comparing children at low, medium and high levels of total energy expenditure (TEE). Data from 231 children previously published2 showing a regression relationship between TEE and RMR. The three ‘hypothetical’ children are shown as large circles at low, medium and high levels of energy expenditure. All three children lie on the regression line and there- fore have a similar total relative to resting energy expenditure, however, the TEE:RMR ratios are widely different. The difference in the TEE:RMR ratio is a spurious observation because this ratio fails to take into account the non-zero intercept in the relationship between total and resting energy expenditure
Estimating energy requirements 1185
(RMR þ activity þ thermic effect of a meal) rather than multiples of one single component. Another physiological
limitation is that applying PALs to RMR assumes that
physical activity energy expenditure is dependent entirely
on RMR, which is not the case. Moreover, since RMR is
primarily weight dependent (or more specifically lean
mass dependent), the PAL approach assumes that physical
activity is primarily weight dependent. This may be true
for some physical activities but unlikely to be true for all
types and levels of physical activity.
Summary and conclusion
A summary of the strengths and limitations of the two
approaches is provided in Table 1. The regression-based
approach is recommended since it is evidence based,
easier to apply and more statistically and physiologically
appropriate. However, during the transition from the
traditional PAL approach, it is relatively straightforward to
make both approaches available so they can be compared.
References
1 Goran MI, Poehlman ET. Total energy expenditure and energy requirements in healthy elderly persons. Metabolism 1992; 41: 744–53.
2 Goran MI, Nagy TR, Gower BA, Mazariegos M, Solomons N, Hood V, Johnson R. Influence of sex, seasonality, ethnicity and geographic location on the components of total energy expenditure in young children: implications for energy requirements. American Journal of Clinical Nutrition 1998; 68(3): 675–82.
3 Butte N. Energy requirements of infants. Public Health Nutrition 2005; 8(7A): 953–67.
4 Torun B. Energy requirements of children and adolescents. Public Health Nutrition 2005; 8(7A): 968–93.
5 Goran MI. Variation in total energy expenditure in humans. Obesity Research 1995; 3: 59–66.
6 Carpenter WH, Poehlman ET, O’Connell M, Goran MI. Influence of body composition and resting metabolic rate on variation in total energy expenditure: a meta analysis. American Journal of Clinical Nutrition 1995; 61: 4–10.
7 Black AE, Coward WA, Cole TJ, Prentice AM. Human energy expenditure in affluent societies: an analysis of 574 doubly labelled water measurements. European Journal of Clinical Nutrition 1996; 50: 72–92.
MI Goran1186
measuring physical activity(1).pdf
A Practical Guide to Measuring Physical Activity
Louisa G. Sylvia, PhD, Assistant Professor of Psychology; The Massachusetts General Hospital; Bipolar Clinic & Research Program, 50 Staniford Street, Suite 580, Boston, MA 02114; (phone) 617-643-4804 (fax) 617-726-6768
Emily E. Bernstein, BS, Clinical Research Coordinator; The Massachusetts General Hospital; Bipolar Clinic & Research Program, 50 Staniford Street, Suite 580, Boston, MA 02114; (phone) 617-726-7591 (fax) 617-726-6768
Jane L. Hubbard, MS, RD, Dietician; Massachusetts General Hospital, Clinical Research Center, 55 Fruit St, Boston, MA 02114 (phone) (617) 724-2830; (617) 726-7563
Leigh Keating, MS, RD, and Dietician, Brigham & Women’s Hospital; Clinical Center for Investigation, 221 Longwood Avenue, Boston MA 02115 (phone) 617-732-7783; (fax) 617-732-7900
Ellen J. Anderson, MS, RD Bionutrition/MPC Director, Massachusetts General Hospital, Clinical Research Center, 55 Fruit St, Boston, MA 02114 (phone) (617) 724-2830; (617) 726-7563 Louisa G. Sylvia: [email protected]; Emily E. Bernstein: [email protected]; Jane L. Hubbard: [email protected]; Leigh Keating: [email protected]; Ellen J. Anderson: [email protected]
Keywords Physical Activity; Assessment; Research; Methodology; Exercise
Considerations for Measuring Physical Activity Research has demonstrated the benefits of physical activity (PA) and the negative consequences of sedentary behavior for physical and mental wellbeing [1–5]. Thus, PA has become increasingly prominent as an intervention tool; however, research is often hindered by the challenge of employing a valid, reliable measure that also adequately satisfies the research question or design [1, 4–7]. The doubly labeled water method (DLW) remains the gold standard for assessing total energy expenditure; however, it is not often used for research studies as it is expensive, has high subject burden, is time-intensive, and cannot capture qualitative data [8–9]. The aim of this commentary is to summarize the main methods of measuring PA as well as offer examples of their uses in research trials [10–12].
© 2013 Academy of Nutrition and Dietetics. Published by Elsevier Inc. All rights reserved.
Corresponding Author: Louisa Sylvia, Bipolar Clinic & Research Program, The Massachusetts General Hospital, 50 Staniford Street, Suite 580, Boston, MA 02114, [email protected].
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
NIH Public Access Author Manuscript J Acad Nutr Diet. Author manuscript; available in PMC 2015 February 01.
Published in final edited form as: J Acad Nutr Diet. 2014 February ; 114(2): 199–208. doi:10.1016/j.jand.2013.09.018.
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
Methods of Measuring PA Self-Report Questionnaires
These questionnaires are the most common method of PA assessment [13] and rely on participants’ recall ability. Questionnaires vary by what they measure (e.g., mode, duration, or frequency of PA), how data are reported (e.g., activity scores, time, calories), quality of the data (e.g., measures of intensity, differentiating between habitual and merely recent activities, inclusion of leisure and non-leisure activity), and how data are obtained (e.g., paper and pencil assessment, computerized questionnaire, interview) [11, 14]. Validation studies comparing self-report questionnaires to DLW are inconsistent [9]; however, their advantages include cost effectiveness, ease of administration, and accuracy in measuring intense activity [15–16], determining discrete categories of activity level (e.g., low, moderate, high)[16], ranking individuals or groups in their PA[17], providing details about the PA, and showing improvement across groups or individuals [14, 18–19]. Potential disadvantages are that self-report questionnaires are less robust in measuring light or moderate activity [14], assessing energy expenditure [18–19] and may be limited by the dependency on written language (i.e., questions) [20] and external factors (i.e., social desirability, complexity of the questionnaire, age, and seasonal variation) [21–25]. Self-report questionnaires are significantly more reliable at the group than the individual level [9, 17–19] as well as when the questionnaire is structured chronologically and with discrete periods [26].
In Table 1, we provide details on seven well-studied, commonly used self-report questionnaires: Modifiable Activity Questionnaire (MAQ) [27], Previous Week Modifiable Activity Questionnaire (PWMAQ) [28], Recent Physical Activity Questionnaire (RPAQ) [29], International Physical Activity Questionnaires (IPAQ) [3, 30], Previous Day Physical Activity Recall (PDPAR) [31], and 7-day Physical Activity Recall (PAR) [2, 32].
Self-Report Activity Diaries/Logs Self-report diaries require participants to record PA in real time which provides the most detailed data [11, 26] and can overcome some limitations of questionnaires (i.e., less susceptible to recall errors, social desirability bias, measurement bias) [26, 33]. To illustrate, Bouchard’s Physical Activity Record (BAR) [34] is a widely used diary in which participants report PA for each 15 minute interval over three days. Activities are rated on a scale of 1 to 9 (1 = sedentary activity, 9 = intense manual work or high intensity sports) to yield a total energy expenditure score; however, the diary is burdensome, particularly for individuals with cognitive dysfunction [30]. In addition, questionnaires not completed in real time could be subject to memory bias as well as participant reactivity, the phenomenon of behavior change due to awareness of being observed [35–37].
Direct Observation In direct observation, an independent observer monitors and records PA [38–39]. This method of assessment is often used when activity is restricted to a delineated space (e.g., a classroom) [39–41]. It is also a popular method for young children as they have difficulty recalling their PA [42]. This flexible method is valuable in gathering contextual information (e.g., preferred location, time, and clothing) and details of the PA (e.g., type, personalized variations to activities). Disadvantages include high cost of time and energy [30], potential reactivity [35–37], difficulty obtaining ethical approval [37], and the lack of objective measures of energy expenditure [37].
Devices: Accelerometers In recent decades, accelerometers have gained popularity given their accuracy, ability to capture large amounts of data, and ease of administration, particularly in large studies [9].
Sylvia et al. Page 2
J Acad Nutr Diet. Author manuscript; available in PMC 2015 February 01.
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
Accelerometers measure acceleration (counts) in real time and detect movement in up to three orthogonal planes (anteroposterior, mediolateral, and vertical) [30, 43]. These counts are then translated into a metric of interest, which can be biological (e.g. energy expenditure) or PA patterns (e.g. stationary) [44]. Devices can be worn in numerous places on the body, including waist, hip, and thigh. Table 2 summarizes commonly-used triaxial accelerometers.
As demonstrated in large studies, such as the ongoing National Health and Nutrition Examination Survey (NHANES) conducted by the National Center for Health Statistics (part of the Centers for Disease Control and Prevention), investigators can use accelerometer data to compute physical activity volume, rate, and time spent in different intensities of exercise, and can be used for broader characterizations such as achievement of public health guidelines and classification by physical activity levels [45]. New accelerometers demonstrate better validity, compared to DLW, than older models. For example, the TracmorD has improved validity over the Tritrac R3D [46–47]. Strengths of accelerometers include minute-by-minute on-line monitoring [12], capturing intensity level [48–50], feasibility with young children [51], accuracy with static and dynamic behaviors [14, 20, 52], and large memory capacities [53]. However, accelerometers are expensive [6] and require technical expertise, specialized hardware, software, and individual programming [6]. Accelerometers also lack a standard protocol for managing or reducing data [37], can induce a reactivity bias [30], and do not provide any contextual information. Additionally, some accelerometers are unable to differentiate body position (i.e., sitting, lying, standing) or walking intensity [37]. Notably, the relationship between accelerometer activity counts and energy expenditure depends on the count cut-point applied to the data; choosing different cut-points can differentially influence measurements of physical activity intensity [54].
Devices: Pedometers Pedometers measure number of steps taken with a horizontal, spring-suspended lever arm which is deflected when the subject’s hip accelerates vertically with a force beyond a chosen threshold. Pedometers correlate strongly with uniaxial accelerometers, and directly observed duration of activities [30, 55–57]. Their simplicity, relatively low cost, and ability to pick up short durations of PA (often missed by self-report measures) make these devices popular. Pedometer data also tend to be correlated with biological outcomes and predictors (e.g. age, BMI) [58]. Pedometers appear to yield the most accurate data for running and moderate walking, as these behaviors require forward vertical motion. Disadvantages of pedometers include inability to record PA involving horizontal motion occurring during periods of inactivity, leisure activity, or solely upper body movements [59–60]. Pedometer brands differ in the set vertical acceleration threshold needed to register a step, which necessarily yield varying PA sensitivity and thus different outputs [18]. Pedometers do not record intensity, frequency, or duration of PA [53, 61], have significantly less data storage capacity than accelerometers [53], and can also induce reactivity in subjects [30, 35, 62]. Pedometers work best for documenting relative changes in PA or ranking individuals [61]. Table 2 includes a summary of widely used pedometers.
Devices: Heart-Rate Monitors Heart rate [63] monitoring is a physiological indicator of PA and energy expenditure [64], providing real-time data on the frequency, duration, and intensity of PA in an unobtrusive (e.g., they can be worn as watches or on the chest), low-effort way for periods up to one month [12, 65–66]. HR monitors capture energy expenditure during activities not involving vertical trunk displacement that many accelerometers and pedometers miss [67] and are best suited to categorize subjects’ PA levels (i.e., highly active, somewhat active, sedentary) as opposed to the exact amount of PA. These devices tend to show discrepancies particularly at very high and low intensities [1, 53, 65–66, 68]. Discrepancies are due to HR and energy
Sylvia et al. Page 3
J Acad Nutr Diet. Author manuscript; available in PMC 2015 February 01.
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
expenditure not sharing a linear relationship at rest and low-intensity (as the PA is confounded by unrelated factors such as caffeine, stress, body position) or high intensity PA [69]. Age, body composition, muscle mass, gender, and fitness level also affect this linear relationship or reduce its accuracy [61].
Devices: Armbands In recent years, armband technology has been developed and validated using DLW [70] in an effort to address the limitations of other devices. Several versions of the armband exist (e.g., SenseWear, HealthWear, bodybugg) [71] and they use motion and heat-related sensors (i.e., heat flux, galvanic skin response, skin temperature, body temperature) to measure energy expenditure and monitor metabolic PA [71]. This dual measurement strategy (i.e., body temperature and motion) is more sensitive to assessing the energy expenditure associated with complex and non-ambulatory activities, such as walking while carrying a heavy load [72–73]. Thus, armbands have proven to be excellent devices for tasks of daily life (or low to moderate activity), but have not been ideal for higher intensity exercise [74]. Thus, researchers have developed exercise-specific algorithms to correct this error in armband technology [75]; however, it can still be a limitation especially if the type and duration of exercise are unknown.
Choosing a Measure of PA Four key features of a PA measure should be considered when choosing one for a research study: (1) quality of PA measured (e.g. activity type, intensity, frequency, duration), (2) objectivity of the data, subject burden (e.g. time and/or effort required to complete), (3) cost/ burden to administer, and (4) specific limitations, discussed above. To further assist in choosing a PA assessment, we considered the main factors of a study population (i.e., age, gender, body weight, co-morbid conditions) that may impact choosing a PA measure.
Age Age groups differ in regards to activity level (i.e., frequency and duration), PA type, cognition, and ability to focus or sustain attention. For example, children typically exhibit intense, but sporadic bursts of PA [39, 76]. Thus, self-report measures limited to total time of activities [76], accelerometers that assume a consistent intensity of exercise [61], and HR monitors which would record sustained elevated HR [61] are not ideal for younger study participants. Accelerometers (e.g. actigraph, activPal) [77], several self-report questionnaires (i.e., PDPAR, IPAQ, PAR, BAR), and direct observation have been validated for children. Armbands have been validated only at sedentary, low, and moderate levels of activity for children with child-specific algorithms applied to the data [78–79].
Adults are more likely to demonstrate consistently low, but steady PA (e.g., walking) and high sedentary activity at work [13], whereas the elderly often have physical restrictions that narrow their scope and type of PA [80]. Therefore, tools that do not accurately record walking may not be best for adult or elderly groups. Pedometers, which often fail to record slower, shuffling gates, will not adequately reflect the PA of older, frailer populations [46]. Thus, the PAR [81], PDPAR [82–83], IPAQ, and accelerometers (e.g. activPal, Tritrac) [84–86]
have been validated in elderly populations.
Furthermore, adults have demonstrated adequate recall ability for self-reported PA assessments [87], but children and the elderly have more difficulty with this type of assessment [10–11, 21, 88]. Self-report measures specifically for children and elderly include those that employ prompts, cued recall, or recognition rather than spontaneous generation [89], divide questionnaires or logs into discrete, logical time periods [90] or cover fewer days.
Sylvia et al. Page 4
J Acad Nutr Diet. Author manuscript; available in PMC 2015 February 01.
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
Gender Making gender-specific assumptions about exercise regimes can be difficult given the many confounding factors (e.g., weight, medical comorbidity, age) and that gender differences can vary by culture and/or country. However, studies in the United States and Canada have identified some trends for women versus men that could help elucidate why certain PA assessment tools have not proven valid or reliable for women. For example, women tend to walk and participate in light PA more so than men and men tend to partake in vigorous PA more so than women [32, 91–93]. As such, for women, some accelerometers (e.g. Tritrac accelerometer [94]), BAR [92], and some HR monitors (e.g. Polar S410 [67]) have failed validity tests given their limitations with light activity. Other accelerometers (e.g. activPal), HR monitors, and pedometers are likely to be more accurate. Similarly, self-report measures such as the RPAQ and MAQ may not be ideal for women as they do not account for many forms of light PA, while the PDPAR and PAR do.
Body Weight High body mass index (BMI) can reduce accuracy of devices, particularly pedometers [32]
armbands [75, 95], and HR monitors [61]. Additionally, studies have found under- [22] and overestimation [96] of self-report PA among obese participants as compared to non-obese respondents. Research has shown that obese individuals, in addition to having significantly higher BMIs, tend to be less active than the rest of the population [22, 97]. In general, recall of PA among inactive individuals is less accurate [98]. Furthermore, because of the relatively low engagement in PA, self-report measures like the MAQ and RPAQ that only encompass leisure activities and do not include unstructured daily activities, such as housework, are ill- advised [88, 93], whereas the PDPAR is recommended because it captures such domains and has been validated for an obese population [99]. Some accelerometers, like the activPal, but not all (e.g. Tritrac), have been validated for this clinical population [100–101]. The combination of high BMI and inactive lifestyle poses a unique challenge in identifying accurate methods of estimating one’s energy expenditure.
Psychiatric and Medical Co-morbidities When studying a population with severe mental illness, certain characteristics should be considered, such as low levels of leisure PA [20, 36, 102–104] and cognitive impairment, including shorter attention span, memory deficits, and errors in comprehension and reporting [24, 105–106]. Thus, it is suggested that measurements account for frequency, varying intensity and duration, and all possible contexts (e.g. structured exercise, housework) of the PA [6, 105, 107–108]. Self-reports like the MAQ and RPAQ that only account for structured leisure time may not be advantageous [88]; additionally, the PAR has demonstrated questionable validity in this population [105]. Cognitive impairments restrict the feasibility of self-reports [24], particularly complex or lengthy questionnaires [102], but the IPAQ has been validated for participants with severe mental illness [109–110]. Furthermore, specific psychiatric conditions are associated with varying frequency of PA; for example, lower PA can be associated with anxiety and depression, while greater PA can be associated with eating disorders and alcohol abuse [109, 111]. As noted above, different levels of habitual PA intensity merit different measures. In general, objective measures such as accelerometers or pedometers are suggested as the primary assessment tool for such populations [105].
For individuals with serious medical co-morbidities, ability to exercise is a key moderator of PA [6], making structured leisure activities and moderate to vigorous PA often difficult. As such, the MAQ and RPAQ are not ideal for individuals with a high medical burden [88], but the PAR questionnaire is recommended [112]. Furthermore, accelerometers can fail to distinguish between standing and sitting, a distinction that may be crucial for disabled
Sylvia et al. Page 5
J Acad Nutr Diet. Author manuscript; available in PMC 2015 February 01.
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
participants; the activPal specifically has been validated with elderly individuals with impaired walking [84]. Additionally, armband monitors have had inconsistent results for users with serious medical co-morbidities [113–115].
Conclusions PA is a multi-dimensional construct and thus, there is no measure that can assess all facets of PA. Thus, investigators should approach PA measure selection with a clear concept of the type of data they intend to collect. For many studies, combining multiple PA assessments is recommended [8, 116], however, it is possible multiple measures may not be necessary if an investigator is only interested in one facet of PA. Given that further research is needed to validate individual PA measures for different populations, it is difficult to determine an optimal PA assessment. Thus, investigators when selecting a PA measure need to pay close attention to each assessment’s strengths and limitations. We recommend consulting with a provider who has expertise in the area of PA assessment prior to choosing a measure, but hope that this commentary provides the knowledge base for investigators without this expertise to ask the questions that most pertain to their area of study.
Acknowledgments Funding Disclosure.
The project described was partially supported by Grant Number 1 UL1 RR025758-04, Harvard Clinical and Translational Science Center, from the National Center for Research Resources; Grant Number 8 UL1 TR000170-05, Harvard Clinical and Translational Science Center, from the National Center for Advancing Translational Science; and Grant Number 5K23MH091182-02, Nutrition/Weight Loss, Exercise, Wellness Treatment for Bipolar Disorder, from the National Institute of Mental Health.
References 1. Macfarlane DJ, Lee CC, Ho EY, et al. Convergent validity of six methods to assess physical activity
in daily life. J Appl Physiol. 2006; 101(5):1328–34. [PubMed: 16825525]
2. Blair SN, Haskell WL, Ho P, et al. Assessment of habitual physical activity by a seven-day recall in a community survey and controlled experiments. Am J Epidemiol. 1985; 122(5):794–804. [PubMed: 3876763]
3. Craig CL, Marshall AL, Sjostrom M, et al. International physical activity questionnaire: 12-country reliability and validity. Med Sci Sports Exerc. 2003; 35(8):1381–95. [PubMed: 12900694]
4. Lee IM, Hsieh CC, Paffenbarger RS Jr. Exercise intensity and longevity in men. The Harvard Alumni Health Study. JAMA. 1995; 273(15):1179–84. [PubMed: 7707624]
5. Reilly JJ, Penpraze V, Hislop J, et al. Objective measurement of physical activity and sedentary behaviour: review with new data. Arch Dis Child. 2008; 93(7):614–9. [PubMed: 18305072]
6. Dishman RK. The measurement conundrum in exercise adherence research. Med Sci Sports Exerc. 1994; 26(11):1382–90. [PubMed: 7837959]
7. Welk, G. Introduction to physical activity research. In: Welk, GJ., editor. Physical activity assessments for health-related research. Human Kinetics Publishers, Inc; Champaign: 2002. p. 4
8. Melanson EL Jr, Freedson PS. Physical activity assessment: a review of methods. Crit Rev Food Sci Nutr. 1996; 36(5):385–96. [PubMed: 8725670]
9. Westerterp KR. Assessment of physical activity: a critical appraisal. Eur J Appl Physiol. 2009; 105(6):823–8. [PubMed: 19205725]
10. Janz KF, Witt J, Mahoney LT. The stability of children’s physical activity as measured by accelerometry and self-report. Med Sci Sports Exerc. 1995; 27(9):1326–32. [PubMed: 8531633]
11. Sallis JF. Self-report measures of children’s physical activity. J Sch Health. 1991; 61(5):215–9. [PubMed: 1943046]
12. Welk GJ, Corbin CB. The validity of the Tritrac-R3D Activity Monitor for the assessment of physical activity in children. Res Q Exerc Sport. 1995; 66(3):202–9. [PubMed: 7481081]
Sylvia et al. Page 6
J Acad Nutr Diet. Author manuscript; available in PMC 2015 February 01.
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
13. Castillo-Retamal M, Hinckson EA. 2011. Work. 2011; 40(4):345–57. [PubMed: 22130052]
14. Jacobs DR Jr, Ainsworth BE, Hartman TJ, et al. A simultaneous evaluation of 10 commonly used physical activity questionnaires. Med Sci Sports Exerc. 1993; 25(1):81–91. [PubMed: 8423759]
15. Besson H, Brage S, Jakes RW, et al. Estimating physical activity energy expenditure, sedentary time, and physical activity intensity by self-report in adults. Am J Clin Nutr. 91(1):106–14. [PubMed: 19889820]
16. Ishikawa-Takata K, Tabata I, Sasaki S, et al. Physical activity level in healthy free-living Japanese estimated by doubly labelled water method and International Physical Activity Questionnaire. Eur J Clin Nutr. 2008; 62(7):885–91. [PubMed: 17522602]
17. Corder K, van Sluijs EM, Wright A, et al. Is it possible to assess free-living physical activity and energy expenditure in young people by self-report? Am J Clin Nutr. 2009; 89(3):862–70. [PubMed: 19144732]
18. Shephard RJ. How much physical activity is needed for good health? Int J Sports Med. 1999; 20(1):23–7. [PubMed: 10090457]
19. Shephard RJ. Limits to the measurement of habitual physical activity by questionnaires. Br J Sports Med. 2003; 37(3):197–206. discussion 206. [PubMed: 12782543]
20. Lassenius O, Akerlind I, Wiklund-Gustin L, et al. Self-reported health and physical activity among community mental healthcare users. J Psychiatr Ment Health Nurs. 2013
21. Baranowski T, Dworkin RJ, Cieslik CJ, et al. Reliability and validity of self report of aerobic activity: family health project. Research Quarterly. 1984; 55:309–317.
22. Klesges RC, Eck LH, Mellon MW, et al. The accuracy of self-reports of physical activity. Med Sci Sports Exerc. 1990; 22(5):690–7. [PubMed: 2233209]
23. Uitenbroek DG. Seasonal variation in leisure time physical activity. Med Sci Sports Exerc. 1993; 25(6):755–60. [PubMed: 8321115]
24. Durante R, Ainsworth BE. The recall of physical activity: using a cognitive model of the question- answering process. Med Sci Sports Exerc. 1996; 28(10):1282–91. [PubMed: 8897386]
25. Vanhees L, Lefevre J, Philippaerts R, et al. How to assess physical activity? How to assess physical fitness? Eur J Cardiovasc Prev Rehabil. 2005; 12(2):102–14. [PubMed: 15785295]
26. van der Ploeg HP, Merom D, Chau JY, et al. Advances in population surveillance for physical activity and sedentary behavior: reliability and validity of time use surveys. Am J Epidemiol. 2010; 172(10):1199–206. [PubMed: 20855469]
27. Kriska AM, Knowler WC, LaPorte RE, et al. Development of questionnaire to examine relationship of physical activity and diabetes in Pima Indians. Diabetes Care. 1990; 13(4):401–11. [PubMed: 2318100]
28. Gabriel KP, McClain JJ, Schmid KK, et al. Reliability and convergent validity of the past-week Modifiable Activity Questionnaire. Public Health Nutrition. 2010; 14(3):435–442. [PubMed: 20843404]
29. Besson H, Brage S, Jakes RW, et al. Estimating physical activity energy expenditure, sedentary time, and physical activity intensity by self-report in adults. Am J Clin Nutr. 2010; 91(1):106–14. [PubMed: 19889820]
30. Rachele JN, McPhail SM, Washington TL, et al. Practical physical activity measurement in youth: a review of contemporary approaches. World J Pediatr. 2012; 8(3):207–16. [PubMed: 22886192]
31. Weston AT, Petosa R, Pate RR. Validation of an instrument for measurement of physical activity in youth. Med Sci Sports Exerc. 1997; 29(1):138–43. [PubMed: 9000167]
32. Sallis JF, Haskell WL, Wood PD, et al. Physical activity assessment methodology in the Five-City Project. Am J Epidemiol. 1985; 121(1):91–106. [PubMed: 3964995]
33. Tudor-Locke C, van der Ploeg HP, Bowles HR, et al. Walking behaviours from the 1965–2003 American Heritage Time Use Study (AHTUS). Int J Behav Nutr Phys Act. 2007; 4:45. [PubMed: 17897472]
34. Bouchard C, Tremblay A, Leblanc C, et al. A method to assess energy expenditure in children and adults. Am J Clin Nutr. 1983; 37(3):461–7. [PubMed: 6829488]
35. Ling FC, Masters RS, McManus AM. Rehearsal and pedometer reactivity in children. J Clin Psychol. 2010; 67(3):261–6. [PubMed: 21254054]
Sylvia et al. Page 7
J Acad Nutr Diet. Author manuscript; available in PMC 2015 February 01.
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
36. Lindamer LA, McKibbin C, Norman GJ, et al. Assessment of physical activity in middle-aged and older adults with schizophrenia. Schizophr Res. 2008; 104(1–3):294–301. [PubMed: 18550338]
37. Hardy LL, Hills AP, Timperio A, et al. A hitchhiker’s guide to assessing sedentary behaviour among young people: Deciding what method to use. J Sci Med Sport. 2013
38. McKenzie TL, Marshall SJ, Sallis JF, et al. Leisure-time physical activity in school environments: an observational study using SOPLAY. Prev Med. 2000; 30(1):70–7. [PubMed: 10642462]
39. Sleap M, Warburton P. Physical activity levels of 5–11-year-old children in England: cumulative evidence from three direct observation studies. Int J Sports Med. 1996; 17(4):248–53. [PubMed: 8814504]
40. Sallis JF, McKenzie TL, Conway TL, et al. Environmental interventions for eating and physical activity: a randomized controlled trial in middle schools. Am J Prev Med. 2003; 24(3):209–17. [PubMed: 12657338]
41. McKenzie TL. Observational measures of children’s physical activity. J Sch Health. 1991; 61(5): 224–7. [PubMed: 1943048]
42. Anderssen N, Jacobs DR Jr, Aas H, et al. Do adolescents and parents report each other’s physical activity accurately? Scand J Med Sci Sports. 1995; 5(5):302–7. [PubMed: 8581574]
43. Chen KY, Bassett DR Jr. The technology of accelerometry-based activity monitors: current and future. Med Sci Sports Exerc. 2005; 37(11 Suppl):S490–500. [PubMed: 16294112]
44. Freedson P, Pober D, Janz KF. Calibration of accelerometer output for children. Med Sci Sports Exerc. 2005; 37(11 Suppl):S523–30. [PubMed: 16294115]
45. Tudor-Locke C, Brashear MM, Johnson WD, et al. Accelerometer profiles of physical activity and inactivity in normal weight, overweight, and obese U.S. men and women. Int J Behav Nutr Phys Act. 2010; 7:60. [PubMed: 20682057]
46. Plasqui G, Westerterp KR. Physical activity assessment with accelerometers: an evaluation against doubly labeled water. Obesity (Silver Spring). 2007; 15(10):2371–9. [PubMed: 17925461]
47. Vanhelst J, Mikulovic J, Bui-Xuan G, et al. Comparison of two ActiGraph accelerometer generations in the assessment of physical activity in free living conditions. BMC Res Notes. 2012; 5:187. [PubMed: 22534207]
48. Healy GN, Dunstan DW, Salmon J, et al. Objectively measured light-intensity physical activity is independently associated with 2-h plasma glucose. Diabetes Care. 2007; 30(6):1384–9. [PubMed: 17473059]
49. Healy GN, Wijndaele K, Dunstan DW, et al. Objectively measured sedentary time, physical activity, and metabolic risk: the Australian Diabetes, Obesity and Lifestyle Study (AusDiab). Diabetes Care. 2008; 31(2):369–71. [PubMed: 18000181]
50. Matthews CE, Chen KY, Freedson PS, et al. Amount of time spent in sedentary behaviors in the United States, 2003–2004. Am J Epidemiol. 2008; 167(7):875–81. [PubMed: 18303006]
51. Pate RR, Almeida MJ, McIver KL, et al. Validation and calibration of an accelerometer in preschool children. Obesity (Silver Spring). 2006; 14(11):2000–6. [PubMed: 17135617]
52. Godfrey A, Culhane KM, Lyons GM. Comparison of the performance of the activPAL Professional physical activity logger to a discrete accelerometer-based activity monitor. Med Eng Phys. 2007; 29(8):930–4. [PubMed: 17134934]
53. Freedson PS, Miller K. Objective monitoring of physical activity using motion sensors and heart rate. Res Q Exerc Sport. 2000; 71(2 Suppl):S21–9. [PubMed: 10925821]
54. Loprinzi PD, Lee H, Cardinal BJ, et al. The relationship of actigraph accelerometer cut-points for estimating physical activity with selected health outcomes: results from NHANES 2003–06. Res Q Exerc Sport. 2012; 83(3):422–30. [PubMed: 22978192]
55. McClain JJ, Tudor-Locke C. Objective monitoring of physical activity in children: considerations for instrument selection. J Sci Med Sport. 2009; 12(5):526–33. [PubMed: 19054715]
56. Bassett, J.; Strath, SJ. Use of pedometers to assess physical activity. In: Welk, GJ., editor. Physical activity assessment for health-related research. Human Kinetics; Champaign: 2002. p. 163-178.
57. Tudor-Locke C, Williams JE, Reis JP, et al. Utility of pedometers for assessing physical activity: convergent validity. Sports Med. 2002; 32(12):795–808. [PubMed: 12238942]
Sylvia et al. Page 8
J Acad Nutr Diet. Author manuscript; available in PMC 2015 February 01.
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
58. Cleland VJ, Schmidt MD, Salmon J, et al. Correlates of pedometer-measured and self-reported physical activity among young Australian adults. J Sci Med Sport. 2011; 14(6):496–503. [PubMed: 21622024]
59. Tudor-Locke C, Ainsworth BE, Thompson RW, et al. Comparison of pedometer and accelerometer measures of free-living physical activity. Med Sci Sports Exerc. 2002; 34(12):2045–51. [PubMed: 12471314]
60. Kilanowski CK, Consalvi AR, Epstein LH. Validation of an electronic pedometer for measurement of physical activity in children. Pediatric Exercise Science. 1999; 11:63–8.
61. Trost SG. Objective measurement of physical activity in youth: current issues, future directions. Exerc Sport Sci Rev. 2001; 29(1):32–6. [PubMed: 11210445]
62. McNamara E, Hudson Z, Taylor SJ. Measuring activity levels of young people: the validity of pedometers. Br Med Bull. 2010; 95:121–37. [PubMed: 20562207]
63. Schmidt MD, Blizzard CL, Venn AJ, et al. Practical considerations when using pedometers to assess physical activity in population studies: lessons from the Burnie Take Heart Study. Res Q Exerc Sport. 2007; 78(3):162–70. [PubMed: 17679489]
64. Janz, KF. Use of heart rate monitors to assess physical activity. In: Welk, GJ., editor. Physical Activity Assessments for Health-Related Research. Human Kinetics; Champaign: 2002. p. 143-161.
65. Sirard JR, Pate RR. Physical activity assessment in children and adolescents. Sports Med. 2001; 31(6):439–54. [PubMed: 11394563]
66. Pahkala K, Heinonen OJ, Lagstrom H, et al. Leisure-time physical activity of 13-year-old adolescents. Scand J Med Sci Sports. 2007; 17(4):324–30. [PubMed: 16903899]
67. Crouter SE, Albright C, Bassett DR Jr. Accuracy of polar S410 heart rate monitor to estimate energy cost of exercise. Med Sci Sports Exerc. 2004; 36(8):1433–9. [PubMed: 15292754]
68. Terbizan DJ, Dolezal BA, Albano C. Validity of seven commercially available heart rate monitors. Measurement in Physical Education and Exercise Science. 2002; 6(4):243–247.
69. Livingstone MB. Heart-rate monitoring: the answer for assessing energy expenditure and physical activity in population studies? Br J Nutr. 1997; 78(6):869–71. [PubMed: 9497439]
70. Mignault D, St-Onge M, Karelis AD, et al. Evaluation of the Portable HealthWear Armband: a device to measure total daily energy expenditure in free-living type 2 diabetic individuals. Diabetes Care. 2005; 28(1):225–7. [PubMed: 15616258]
71. Andre, D.; Pelletier, R.; Farringdon, J., et al. The Development of the SenseWear ® armband, a Revolutionary Energy Assessment Device to Assess Physical Activity and Lifestyle. BodyMedia, Inc; 2006.
72. Welk GJ, McClain JJ, Eisenmann JC, et al. Field validation of the MTI Actigraph and BodyMedia armband monitor using the IDEEA monitor. Obesity (Silver Spring). 2007; 15(4):918–28. [PubMed: 17426327]
73. Johannsen DL, Calabro MA, Stewart J, et al. Accuracy of armband monitors for measuring daily energy expenditure in healthy adults. Med Sci Sports Exerc. 42(11):2134–40. [PubMed: 20386334]
74. Drenowatz C, Eisenmann JC. Validation of the SenseWear Armband at high intensity exercise. Eur J Appl Physiol. 111(5):883–7. [PubMed: 20972880]
75. Jakicic JM, Marcus M, Gallagher KI, et al. Evaluation of the SenseWear Pro Armband to assess energy expenditure during exercise. Med Sci Sports Exerc. 2004; 36(5):897–904. [PubMed: 15126727]
76. Dale, D.; Welk, GJ.; Matthews, CE. Methods for assessing physical activity and challenges for research. In: Welk, GJ., editor. Physical Activity Assessments for Health-Related Research. Human Kinetics Publishers, Inc; Champaign: 2002. p. 19-34.
77. Martin A, McNeill M, Penpraze V, et al. Objective measurement of habitual sedentary behavior in pre-school children: comparison of activPAL With Actigraph monitors. Pediatr Exerc Sci. 2011; 23(4):468–76. [PubMed: 22109774]
78. Andreacci JL, Dixon CB, McConnell TR. Validation of SenseWear ® Armband to Assess Energy Expenditure in Children Ranging in Body Size. Med Sci Sports Exerc. 2006; 38(5):S255.
Sylvia et al. Page 9
J Acad Nutr Diet. Author manuscript; available in PMC 2015 February 01.
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
79. Calabro MA, Welk GJ, Eisenmann JC. Validation of the SenseWear Pro Armband algorithms in children. Med Sci Sports Exerc. 2009; 41(9):1714–20. [PubMed: 19657300]
80. Caspersen, CJ.; Merritt, RK.; Stephens, T. International physical activity patterns: a methodological perspective. In: Dishman, RK., editor. Advances in Exercise Adherence. Human Kinetics Publishers; Champaign: 1994. p. 73-110.
81. Dubbert PM, Vander Weg MW, Kirchner KA, et al. Evaluation of the 7-day physical activity recall in urban and rural men. Med Sci Sports Exerc. 2004; 36(9):1646–54. [PubMed: 15354050]
82. Westerterp KR, Saris WHM, Bloemberg BPM, et al. Validation of the Zutphen physical activity questionnaire for the elderly with double labeled water. Med Sci Sports Exerc. 1992; 34:68.
83. Schuit AJ, Schouten EG, Westerterp KR, et al. Validity of the Physical Activity Scale for the Elderly (PASE): according to energy expenditure assessed by the doubly labeled water method. J Clin Epidemiol. 1997; 50(5):541–6. [PubMed: 9180646]
84. Taraldsen K, Askim T, Sletvold O, et al. Evaluation of a body-worn sensor system to measure physical activity in older people with impaired function. Phys Ther. 2011; 91(2):277–85. [PubMed: 21212377]
85. Grant PM, Dall PM, Mitchell SL, et al. Activity-monitor accuracy in measuring step number and cadence in community-dwelling older adults. J Aging Phys Act. 2008; 16(2):201–14. [PubMed: 18483442]
86. Kochersberger G, McConnell E, Kuchibhatla MN, et al. The reliability, validity, and stability of a measure of physical activity in the elderly. Arch Phys Med Rehabil. 1996; 77(8):793–5. [PubMed: 8702373]
87. Blair SN, Dowda M, Pate RR, et al. Reliability of long-term recall of participation in physical activity by middle-aged men and women. Am J Epidemiol. 1991; 133(3):266–75. [PubMed: 2000844]
88. Pettee Gabriel K, McClain JJ, Schmid KK, et al. Reliability and convergent validity of the past- week Modifiable Activity Questionnaire. Public Health Nutr. 2011:1–8. [PubMed: 21211099]
89. Baranowski T, Domel SB. A cognitive model of children’s reporting of food intake. Am J Clin Nutr. 1994; 59(1 Suppl):212S–217S. [PubMed: 8279427]
90. Baranowski T. Validity and reliability of self report measures of physical activity: an information- processing perspective. Research Quarterly for Exercise & Sport. 1988; 59:314–327.
91. Hovell MF, Sallis JF, Hofstetter CR, et al. Identifying correlates of walking for exercise: an epidemiologic prerequisite for physical activity promotion. Prev Med. 1989; 18(6):856–66. [PubMed: 2626418]
92. Friedenreich CM, Courneya KS, Neilson HK, et al. Reliability and validity of the Past Year Total Physical Activity Questionnaire. Am J Epidemiol. 2006; 163(10):959–70. [PubMed: 16524954]
93. Tortolero SR, Masse LC, Fulton JE, et al. Assessing physical activity among minority women: focus group results. Womens Health Issues. 1999; 9(3):135–42. [PubMed: 10340019]
94. Campbell KL, Crocker PR, McKenzie DC. Field evaluation of energy expenditure in women using Tritrac accelerometers. Med Sci Sports Exerc. 2002; 34(10):1667–74. [PubMed: 12370570]
95. Predieri B, Bruzzi P, Lami F, et al. Accuracy of SenseWear Pro2 armband to predict resting energy expenditure in childhood obesity. Obesity (Silver Spring).
96. Lichtman SW, Pisarska K, Berman ER, et al. Discrepancy between self-reported and actual caloric intake and exercise in obese subjects. N Engl J Med. 1992; 327(27):1893–8. [PubMed: 1454084]
97. Rising R, Harper IT, Fontvielle AM, et al. Determinants of total daily energy expenditure: variability in physical activity. Am J Clin Nutr. 1994; 59(4):800–4. [PubMed: 8147322]
98. Harada ND, Chiu V, King AC, et al. An evaluation of three self-report physical activity instruments for older adults. Med Sci Sports Exerc. 2001; 33(6):962–70. [PubMed: 11404662]
99. Racette SB, Schoeller DA, Kushner RF. Comparison of heart rate and physical activity recall with doubly labeled water in obese women. Med Sci Sports Exerc. 1995; 27(1):126–33. [PubMed: 7898328]
100. Kozey-Keadle S, Libertine A, Lyden K, et al. Validation of wearable monitors for assessing sedentary behavior. Med Sci Sports Exerc. 2011; 43(8):1561–7. [PubMed: 21233777]
Sylvia et al. Page 10
J Acad Nutr Diet. Author manuscript; available in PMC 2015 February 01.
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
101. Jacobi D, Perrin AE, Grosman N, et al. Physical activity-related energy expenditure with the RT3 and TriTrac accelerometers in overweight adults. Obesity (Silver Spring). 2007; 15(4):950–6. [PubMed: 17426330]
102. Krupa T, McLean H, Eastabrook S, et al. Daily time use as a measure of community adjustment for persons served by assertive community treatment teams. Am J Occup Ther. 2003; 57(5):558– 65. [PubMed: 14527118]
103. Pack S. Poor physical health and mortality in patients with schizophrenia. Nurs Stand. 2009; 23(21):41–5. [PubMed: 19248449]
104. Nyboe L, Lund H. Low levels of physical activity in patients with severe mental illness. Nord J Psychiatry. 2012
105. Soundy A, Taylor A, Faulkner G, et al. Psychometric properties of the 7-Day Physical Activity Recall questionnaire in individuals with severe mental illness. Arch Psychiatr Nurs. 2007; 21(6): 309–16. [PubMed: 18037441]
106. Farnam CR, Zipple AM, Tyrrell W, et al. Health status risk factors of people with severe and persistent mental illness. J Psychosoc Nurs Ment Health Serv. 1999; 37(6):16–21. [PubMed: 10382153]
107. Vancampfort D, Correll CU, Probst M, et al. A review of physical activity correlates in patients with bipolar disorder. J Affect Disord. 2012
108. Sallis JF, Saelens BE. Assessment of physical activity by self-report: status, limitations, and future directions. Res Q Exerc Sport. 2000; 71(2 Suppl):S1–14. [PubMed: 10925819]
109. Park S, Cho MJ, Cho SJ, et al. Relationship between physical activity and mental health in a nationwide sample of Korean adults. Psychosomatics. 2011; 52(1):65–73. [PubMed: 21300197]
110. Vancampfort D, Probst M, Knapen J, et al. Associations between sedentary behaviour and metabolic parameters in patients with schizophrenia. Psychiatry Res. 2012
111. El Ghoch M, Calugi S, Pellegrini M, et al. Measured physical activity in anorexia nervosa: Features and treatment outcome. Int J Eat Disord. 2013
112. Johansen KL, Painter P, Kent-Braun JA, et al. Validation of questionnaires to estimate physical activity and functioning in end-stage renal disease. Kidney Int. 2001; 59(3):1121–7. [PubMed: 11231369]
113. Tierney M, Fraser A, Purtill H, et al. A study to determine the criterion validity of the sensewear armband as a measure of physical activity in people with rheumatoid arthritis. Arthritis Care Res (Hoboken).
114. Manns PJ, Haennel RG. SenseWear Armband and Stroke: Validity of Energy Expenditure and Step Count Measurement during Walking. Stroke Res Treat. 2012; 247165
115. Hill K, Dolmage TE, Woon L, et al. Measurement properties of the SenseWear armband in adults with chronic obstructive pulmonary disease. Thorax. 65(6):486–91. [PubMed: 20522844]
116. Haskell WL. Physical activity by self-report: a brief history and future issues. J Phys Act Health. 2012; 9(Suppl 1):S5–10. [PubMed: 22287448]
117. Hagstromer M, Bergman P, De Bourdeaudhuij I, et al. Concurrent validity of a modified version of the International Physical Activity Questionnaire (IPAQ-A) in European adolescents: The HELENA Study. Int J Obes (Lond). 2008; 32(Suppl 5):S42–8. [PubMed: 19011653]
118. Ottevaere C, Huybrechts I, De Bourdeaudhuij I, et al. Comparison of the IPAQ-A and actigraph in relation to VO2max among European adolescents: the HELENA study. J Sci Med Sport. 2011; 14(4):317–24. [PubMed: 21444243]
119. Grant PM, Ryan CG, Tigbe WW, et al. The validation of a novel activity monitor in the measurement of posture and motion during everyday activities. Br J Sports Med. 2006; 40(12): 992–7. [PubMed: 16980531]
120. Nichols JF, Morgan CG, Sarkin JA, et al. Validity, reliability, and calibration of the Tritrac accelerometer as a measure of physical activity. Med Sci Sports Exerc. 1999; 31(6):908–12. [PubMed: 10378921]
121. Bonomi AG, Plasqui G, Goris AH, et al. Estimation of free-living energy expenditure using a novel activity monitor designed to minimize obtrusiveness. Obesity (Silver Spring). 18(9):1845– 51. [PubMed: 20186133]
Sylvia et al. Page 11
J Acad Nutr Diet. Author manuscript; available in PMC 2015 February 01.
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
122. Crouter SE, Schneider PL, Karabulut M, et al. Validity of 10 electronic pedometers for measuring steps, distance, and energy cost. Med Sci Sports Exerc. 2003; 35(8):1455–60. [PubMed: 12900704]
123. Karabulut M, Crouter SE, Bassett DR Jr. Comparison of two waist-mounted and two ankle- mounted electronic pedometers. Eur J Appl Physiol. 2005; 95(4):335–43. [PubMed: 16132120]
124. Brage S, Brage N, Franks PW, et al. Reliability and validity of the combined heart rate and movement sensor Actiheart. Eur J Clin Nutr. 2005; 59(4):561–70. [PubMed: 15714212]
125. Crouter SE, Churilla JR, Bassett DR Jr. Accuracy of the Actiheart for the assessment of energy expenditure in adults. Eur J Clin Nutr. 2008; 62(6):704–11. [PubMed: 17440515]
Sylvia et al. Page 12
J Acad Nutr Diet. Author manuscript; available in PMC 2015 February 01.
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
Sylvia et al. Page 13
Ta bl
e 1
Su m
m ar
y of
s el
f- re
po rt
q ue
st io
nn ai
re s
to m
ea su
re p
hy si
ca l a
ct iv
ity
M ea
su re
P er
io d(
s) o
f In
te re
st C
at eg
or ie
s of
A ct
iv it
y In
cl ud
ed In
pu t
O ut
pu t
Sp ec
ia l N
ot es
M A
Q [2
7] L
if et
im e,
P as
t y ea
r Pa
st w
ee k
L ei
su re
O cc
up at
io n
T ra
ns po
rt
D ur
at io
n Fr
eq ue
nc y
N um
be r
of h
ou rs
( or
M E
T h
ou rs
) pe
r w
ee k
of PA
In cl
ud es
n o
m ea
su re
o f
in te
ns ity
P W
M A
Q [2
8] Pa
st w
ee k
L ei
su re
T el
ev is
io n
C om
pu te
r us
e D
is ab
ili ty
-r el
at ed
in ac
tiv ity
D ur
at io
n Fr
eq ue
nc y
N um
be r
of h
ou rs
( or
M E
T h
ou rs
) pe
r w
ee k
of PA
M od
if ie
d ve
rs io
n of
th e
M A
Q In
cl ud
es n
o m
ea su
re o
f in
te ns
ity
R P
A Q
[2 9]
Pa st
4 w
ee ks
L ei
su re
O cc
up at
io n
T ra
ns po
rt H
om e
D ur
at io
n Fr
eq ue
nc y
T ot
al e
ne rg
y ex
pe nd
itu re
PA e
ne rg
y ex
pe nd
itu re
In cl
ud es
n o
m ea
su re
o f
in te
ns ity
IP A
Q -S
[3 , 3
0] H
ab itu
al o
r pa
st w
ee k
V ig
or ou
s PA
M od
er at
e PA
W al
ki ng
Si tti
ng
D ur
at io
n Fr
eq ue
nc y
T ot
al P
A s
co re
s fo
r ea
ch c
at eg
or y
D es
ig ne
d to
b e
ea si
ly a
da pt
ed in
m an
y la
ng ua
ge s
an d
co un
tr ie
s
IP A
Q -L
[3 , 3
0] H
ab itu
al o
r pa
st w
ee k
L ei
su re
O cc
up at
io n
T ra
ns po
rt H
om e
Y ar
d &
g ar
de n
Si tti
ng
D ur
at io
n Fr
eq ue
nc y
T ot
al P
A s
co re
s fo
r ea
ch c
at eg
or y
V er
si on
s ex
is t f
or s
pe ci
fi c
po pu
la tio
ns (
e. g.
yo ut
h, e
ld er
ly , a
nd f
or ei
gn la
ng ua
ge sp
ea ke
rs [1
17 –1
18 ] )
P D
P A
R [3
1] Pa
st d
ay , 3
o r
7 da
ys 3:
00 –1
1: 00
pm 30
m in
ut e
in te
rv al
s
E at
in g
Sl ee
pi ng
/b at
hi ng
T ra
ns po
rt W
or k/
sc ho
ol Sp
ar e
tim e
Pl ay
/r ec
re at
io n
E xe
rc is
e/ w
or ko
ut
Pr im
ar y
ac tiv
ity p
er in
te rv
al R
el at
iv e
in te
ns ity
ra te
d on
r ep
ea te
d sc
al e
(c on
ta in
in g
ve rb
al &
c ar
to on
de sc
ri pt
or s)
D ai
ly to
ta l e
ne rg
y ex
pe nd
itu re
T ot
al e
ne rg
y ex
pe nd
itu re
d ur
in g
sp ec
if ic
ti m
e pe
ri od
s T
ot al
e ne
rg y
ex pe
nd itu
re d
ur in
g sp
ec if
ic ac
tiv iti
es
D es
ig ne
d fo
r ch
ild re
n an
d ad
ol es
ce nt
s C
on te
xt ua
l c ue
s an
d pr
om pt
s in
te nd
ed to
en ha
nc e
m em
or y
of P
A a
nd in
te ns
ity
P A
R [2
, 3 2]
Pa st
w ee
k Sl
ee p
M od
er at
e PA
H ar
d PA
V er
y ha
rd P
A
D ur
at io
n T
ot al
e ne
rg y
ex pe
nd itu
re C
al cu
la tio
ns a
ss um
e th
at th
e un
ac co
un te
d fo
r tim
e w
as s
pe nt
in li
gh t a
ct iv
ity
N ot
e. M
E T
= M
et ab
ol ic
e qu
iv al
en t o
f ta
sk (
1 M
E T
r ep
re se
nt s
3. 5
m l/k
g/ m
in o
xy ge
n co
ns um
pt io
n) [
7] , M
A Q
= M
od if
ia bl
e A
ct iv
ity Q
ue st
io nn
ai re
, P W
M A
Q =
P re
vi ou
s W
ee k
M od
if ia
bl e
A ct
iv ity
Q ue
st io
nn ai
re , P
A R
-Q =
P hy
si ca
l A ct
iv ity
R ea
di ne
ss Q
ue st
io nn
ai re
, R PA
Q =
R ec
en t P
hy si
ca l A
ct iv
ity Q
ue st
io nn
ai re
, I PA
Q -S
= I
nt er
na tio
na l P
hy si
ca l A
ct iv
ity Q
ue st
io nn
ai re
( Sh
or t V
er si
on ),
I PA
Q -L
= In
te rn
at io
na l P
hy si
ca l A
ct iv
ity Q
ue st
io nn
ai re
( L
on g
V er
si on
), P
D PA
R =
P re
vi ou
s D
ay P
hy si
ca l A
ct iv
ity R
ec al
l, PA
R =
7 -d
ay P
hy si
ca l A
ct iv
ity R
ec al
l.
J Acad Nutr Diet. Author manuscript; available in PMC 2015 February 01.
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
Sylvia et al. Page 14
Table 2
Summary of devices to measure physical activity (PA)
Measure Location Data Recorded Output Special Notes
Accelerometers
activPal [119], a Thigh Time spent in sedentary behavior, standing, and walking Count of sit-to-stand transitions Total number of steps for a given period
Energy expenditure per behavior
Distinguishes between standing, sitting, and lying Differentiates various intensities of walking
Tritrac [120], b Hip Composite movement score (vector magnitude)
Energy expenditure per minute of movement Estimate of resting metabolic rate
Questionable validity
TracmorD [121], c Lower back Activity counts perminute Total energy expenditure PA energy expenditure Physical activity level Activity energy expenditure per body mass
Waterproof Comfortable Reduces interference from spontaneous activity
Actigraph [47], d Waist/hip Activity counts (amplitude and frequency of acceleration over each sampling period)
Activity intensity categories Time spent in sedentary, low, moderate, and intense activity
Improves sensitivity to low intensity movement (with the Low- Frequency Extension application) Inaccurate count of steps
Pedometers
Yamax Digi- Walker [122], e
Waist Step counts per minute Distance travelled Total energy expenditure
Underestimates step counts at slow activity speeds Widely used in research studies
StepWatch-3 [123], f Ankle Step counts per minute Distance travelled Total energy expenditure
Degree of accuracy not affected by activity speed or BMI Sensitive to small movements (e.g., fidgeting)
Heart-Rate Monitors
Polar S410 [67], g Wrist and chest (two locations)
Beats per minute Heart-rate per unit time Percentage of the age- based maximum heart-rate estimate Time spent in low, medium, or high intensity activity
No movement measurement
Actiheart [124–125], h Chest (two locations) Beats per minute Activity counts
Physical activity intensity PA energy expenditure
Combines heart-rate and movement sensors Higher noise rates in women
Arm-Band Technology
SenseWear [71], i Upper Arm Beats per minute Temperature
Total energy expenditure Metabolic Equivalent of Task
Objective measure of time worn Algorithms specific to vigorous activity and children
Note.
a PAL Technologies Ltd, Glasgow, UK
J Acad Nutr Diet. Author manuscript; available in PMC 2015 February 01.
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
N IH
-PA Author M anuscript
Sylvia et al. Page 15
b Professional Products, Madison, WI
c Philips New Wellness Solutions, Lifestyle Incubator, the Netherlands
d ActiGraph™, Pensacola, CA
e Yamax Corporation, Tokyo, Japan
f SW-3Ankle; Cymatech Inc., Seattle, WA
g Polar Electro, Inc., Lake Success, NY
h Cambridge Neurotechnology, Cambridge, UK
i BodyMedia, Inc., Pittsburgh, PA, USA. HealthWear (Roche Diagnostics, Indianapolis, IN) and bodybugg (Apex Fitness, San Ramon, CA) are
private label versions of BodyMedia’s SenseWear technology, meaning that they can be used interchangeably with the SenseWear device.
J Acad Nutr Diet. Author manuscript; available in PMC 2015 February 01.
sacn_dietary_reference_values_for_energy(2)(1).pdf
Dietary Reference Values for Energy
2011
Dietary Reference Values for Energy
Scientific Advisory Committee on Nutrition
2011
London: TSO
© Crown copyright 2012
You may re-use this information (excluding logos) free of charge in any format or medium, under the terms of the Open Government Licence. To view this licence, visit http://www.nationalarchives.gov.uk/doc/open-government- licence/ or e-mail: [email protected].
Where we have identified any third party copyright information you will need to obtain permission from the copyright holders concerned.
Any enquiries regarding this publication should be sent to the SACN Secretariat: www.sacn.gov.uk/contact_us/index.html
This document is available from the SACN website at: www.sacn.gov.uk
ISBN: 9780108511370
Printed in the UK by The Stationery Office Limited
ID 2485597 04/12 15707 19585
Printed on paper containing 75% recycled fibre content minimum.
1
Preface
In 1991, the Committee on the Medical Aspects of Food Policy (COMA) provided estimates of energy requirements for the UK population in their report Dietary Reference Values for Food Energy and Nutrients for the United Kingdom. The dietary reference values (DRVs) for energy were based on estimating the total energy expenditure (TEE) for groups of people. TEE provides a measure of the energy requirement at energy balance i.e. when energy intake matches energy expenditure. In this way, an energy requirement can be predicted as the rate of TEE plus any additional needs for growth, pregnancy and lactation.
Since the publication of the report in 1991, the methodology to measure TEE – the doubly labelled water (DLW) method – has advanced and as a result, the evidence base on TEE in a wide variety of population groups has expanded considerably. In addition, the Food and Agriculture Organization of the United Nations, World Health Organization, and United Nations University (FAO/WHO/UNU) and Institute of Medicine (IoM) have updated their recommendations on energy requirements. With the high levels of overweight and obesity currently seen in the UK and the wealth of new data now available, it was considered timely for the Scientific Advisory Committee on Nutrition (SACN) to review recommendations for the UK population.
The present report details the evidence and approaches SACN have considered in order to update the DRVs for energy. The DRVs for energy are based on the estimated average requirements (EARs) of infants, children, adolescents and adults. After much deliberation, SACN agreed that the factorial approach was the most appropriate way to derive energy requirements, whereby TEE is expressed as a multiple of the basal metabolic rate (BMR) and the physical activity level (PAL). Hence, TEE or EAR is equal to BMR x PAL. The PAL is best estimated from measures of DLW, which give higher values than previously estimated by COMA.
SACN noted that in populations like the UK, with a high and increasing proportion of overweight and obese individuals, if energy requirements are estimated at current levels of energy expenditure and body weights, many groups in the population would continue to carry excess weight. This is not desirable since excess body weight is associated with long-term poor health and increased mortality. To address this issue SACN chose a prescriptive approach to estimating energy reference values. That is, suitable reference body weight ranges consistent with long-term good health were used to calculate energy reference values. Thus, BMR values were predicted using healthy reference body weights. For the purposes of calculation, this equates to the 50th centile of UK-WHO growth standards for infants and pre-school children, the 50th centile of UK 1990 reference for school-aged children and for adults at a Body Mass Index (BMI) of 22.5 kg/m2 at the current height of the UK adult population. Using this approach, if overweight groups consume the amount of energy recommended for healthy weight groups, they are likely to lose weight, whereas underweight sections of the population should gain weight
2
towards the healthy body weight range. This approach represents a significant departure from the method used by COMA.
SACN has derived new energy reference values. For most population groups, except for infants and young children, the values have increased. This change reflects the more accurate methods used to assess energy expenditure; the evidence base available to COMA was more limited and as a result energy requirements were underestimated for some age groups. It is important to note that DRVs should be used to assess the energy requirements for large groups of people and populations, but should not be applied to individuals due to the large variation in physical activity and energy expenditure observed between people. Despite SACN’s best efforts to base their recommendations on the most up to date evidence, it should be noted that there is less DLW data available for infants, younger adults (18-30 years) and those aged 80 years. The Committee hopes that this will be addressed in the future.
I would like to thank those who provided comments on the draft version of this report during the public consultation. The process assisted the Committee in refining its approach to setting energy requirements for the UK. This has been a challenging and large undertaking for SACN and I would like to thank the Energy Requirements Working Group and the Secretariat for their great commitment in producing this report. Particular thanks to Professor Joe Millward for his substantial contribution to this report. I would also like to extend my gratitude to the principal investigators of the Beltsville and OPEN studies for allowing the Committee to use their data.
SACN has derived new energy reference values. For most population groups, except for infants and young children, the values have slightly increased. This change reflects the more accurate methods used to assess energy expenditure; the evidence base available to COMA was more limited and as a result energy requirements were underestimated for some age groups. It is important to note that DRVs should be used to assess the energy requirements for large groups of people and populations, but should not be applied to individuals due to the large variation in physical activity and energy expenditure observed between people. Despite SACN’s best efforts to base their recommendations on the most up to date evidence, it should be noted that there is less DLW data available for infants, younger adults (18-30 years) and those aged 80 years. The Committee hopes that this will be addressed in the future. I would like to thank those who provided comments on the draft version of this report during the public consultation. The process assisted the Committee in refining its approach to setting energy requirements for the UK. This has been a challenging and large undertaking for SACN and I would like to thank the Energy Requirements Working Group and the Secretariat for their great commitment in producing this report. Particular thanks to Professor Joe Millward for his substantial contribution to this report. I would also like to extend my gratitude to the principle investigators of the Beltsville and OPEN studies for allowing the committee to use their data.
Professor Alan Jackson Chair of the Energy Requirements Working Group
Professor Alan Jackson Chair of the Energy Requirements Working Group
3
Table of contents
1 Summary 11
Introduction 11
Terms of reference 11
Background 12
Components of energy expenditure 12
Basal metabolic rate 12
Physical activity 12
The physical activity level (PAL) 12
Definition of energy requirements 13
Variability 14
Approaches used to set energy reference values 15
Measurement of total energy expenditure 15
Utilising total energy expenditure measurements to derive energy requirements 15
Predicting TEE with regression equations 16
Predicting TEE from BMR and measured PAL values – the factorial approach 16
Summary of approach used by SaCN to determine revised energy reference values 17
Prescriptive energy reference values for healthy body weights 17
Factorial approach to setting EARs 18
Energy reference values for infants, children and adolescents 19
Energy cost of growth 19
Infants aged 0 – 12 months 19
Children and adolescents aged 1 – 18 years 19
Energy reference values for adults 19
Identifying PAL values 20
Energy reference values for older adults 21
Effects of additional physical activity on energy requirements 22
reference energy values during pregnancy and lactation 22
Energy reference values for pregnancy 22
Energy reference values for lactation 22
4
Comparisons with reference values for energy in the COma 1991 report 23
2 INtrOduCtION 24
Background 24
Terms of reference 25
Report content 25
Methodology 26
Units of energy 26
Energy available from food and drink 26
Components of energy expenditure 28
Basal and resting metabolism 28
Physical activity 28
The physical activity level (PAL) 28
Thermic effect of food (TEF) 30
Other components of energy expenditure 30
Growth 30
Pregnancy and lactation 31
Factors affecting energy expenditure 31
Body size and composition 31
Genetic variation 32
Hormones 32
Illness 32
Ambient temperature 33
Energy balance and storage 33
Obesity 33
the influence of physical activity and diet on the regulation of body weight 34
definition of energy requirement 36
Variability 37
approaches used to estimate energy reference values 39
Measurement of total energy expenditure (TEE) 39
Utilising TEE measurements to derive energy requirements 40
Predicting TEE with regression equations 40
Predicting TEE from BMR and measured PAL values – the factorial approach 41
5
Estimating BMR 42
Calculation of energy requirements 43
Approaches employed to calculate energy requirements in other reports 44
3 EStImatEd avEragE rEquIrEmENtS – apprOaChES uSEd aNd valuES dErIvEd 45
Summary of approach used to determine Ears 45
Prescriptive energy reference values for healthy body weights 46
Energy reference values for infants, children and adolescents 47
Energy cost of growth 47
Infants aged 1 – 12 months of age 47
Energy expenditure 47
Energy deposition 48
Energy reference values for children and adolescents aged 1-18 years 52
Estimating the BMR 52
Identifying PAL values 52
Adjusting for growth 55
Calculating energy reference values for children and adolescents 55
Energy reference values for adults 57
Estimating BMR 57
Identifying PAL values 57
Calculating energy reference values 60
Energy reference values for older adults 62
Energy reference values for children, adolescents and adults outside the expected range of habitual activities 62
Effects of additional physical activity on energy requirements 63
reference energy values during pregnancy and lactation 64
Energy costs of pregnancy 65
Gestational weight gain 65
Basal metabolism in pregnancy 65
Total energy expenditure in pregnancy 66
Calculation of energy requirements for pregnancy 67
Approaches taken in other reports 67
Energy reference values for pregnancy 68
6
Energy costs of lactation 69
Calculation of energy requirements for lactation 69
Energy reference values for lactation 71
Comparisons with reference values for energy in the 1991 COma report 71
4 CONCluSIONS aNd rECOmmENdatIONS 74
recommendations 81
Rationale for recommendations 81
research recommendations 86
Measurements of TEE 86
Specific population groups 86
Prediction of the BMR 87
acknowledgements 87
mEmBErShIp OF ENErgy rEquIrEmENtS WOrkINg grOup 88
mEmBErShIp OF SCIENtIFIC advISOry COmmIttEE ON NutrItION 90
appENdIx 1. SaCN WOrkINg prOCEdurES 92
appENdIx 2. ENErgy yIEldS FrOm SuBStratES 93
Fat 93
Carbohydrate 93
Protein 94
Alcohol 94
appENdIx 3. COmpONENtS OF ENErgy ExpENdIturE aNd FaCtOrS aFFECtINg It 95
Components of energy expenditure 95
Basal and resting metabolism 95
Physical activity 95
Spontaneous physical activity and non-exercise activity thermogenesis 96
Other components of energy expenditure 97
Thermic effect of food 97
Factors affecting energy expenditure 97
Body size and composition 97
Gender 99
Age 99
7
Genetics 100
genetics of obesity 100
Ethnicity 101
Endocrine state 101
Pharmacological agents 102
Environment 103
appENdIx 4. Bmr prEdICtION EquatIONS 104
appENdIx 5. thE phySICal aCtIvIty lEvEl (pal) aNd ItS uSE IN thE prEdICtION OF ENErgy rEquIrEmENtS 108
theoretical aspects of calculation of pal and its factorial prediction 108
PAL as an index of TEE adjusted for BMR 108
Factorial prediction of PAL 112
Observed variation in PAL within specific population groups 113
magnitude and variation in pal within the general population 115
Observed lower and upper limits of PAL 115
Observed magnitude and variation in PAL 116
Observed effect of additional physical activity on PAL 117
Predicting the effect of additional physical activity on PAL 117
pal values in relation to health outcomes 120
utilising pal values to determine reference energy intakes 120
appENdIx 6. dOuBly laBEllEd WatEr (dlW) mEthOd 123
methodology critique 124
Summary 126
appENdIx 7. phySICal aCtIvIty aNd ENErgy BalaNCE 128
Background 128
measuring physical activity 128
Subjective measures of physical activity 128
Objective measures of physical activity 129
assessment of physical activity levels in the uk population 129
UK population PAL values 130
physical activity and body fatness 130
Prospective studies of self-reported physical activity and weight gain 131
Adults 131
8
Children and adolescents 131
Prospective studies of objectively measured physical activity and weight gain 133
Adults 133
Children and adolescents 133
Trials using physical activity as an intervention to prevent weight gain 134
Sedentary behaviour and weight gain 135
Summary 137
appENdIx 8. CharaCtErIStICS OF thE data SEt OF dlW mEaSurES OF ENErgy ExpENdIturE FOr adultS: COmBINEd OpEN/ BEltSvIllE dlW data SEt. 138
appENdIx 9. ENErgy rEquIrEmENtS FOr dISEaSE 144
Introduction 144
Energy intake 144
Energy expenditure 145
diseases in adults 146
diseases in children 147
Summary 147
appENdIx 10. OBESIty prEvalENCE IN thE uk 148
appENdIx 11. a CONSIdEratION OF ENErgy INtakE aNd phySICal aCtIvIty IN rElatION tO WEIght gaIN 153
Background 153
Energy balance and energy flux 153
Exercise and appetite control 155
Summary 155
appENdIx 12. CharaCtErIStICS OF thE data SEt OF dOuBly laBEllEd WatEr (dlW) mEaSurES OF ENErgy ExpENdIturE IN ChIldrEN aNd adOlESCENtS 157
Number of studies and individuals within the studies and general characteristics 158
Grouping by age 159
Influence of gender 160
Variation of PAL with age within the BMR age groups 160
adjustment of pal to account for growth costs 161
9
appENdIx 13. COmparISONS WIth thE ExIStINg drvS FOr ENErgy FrOm thE 1991 COma rEpOrt ‘dIEtary rEFErENCE valuES FOr ENErgy aNd NutrIENtS FOr thE uNItEd kINgdOm’. 168
Overarching differences 168
Infants aged 1-12 months 168
Children and adolescents aged 1-18 years 169
Adults 170
Comparison with COMA report for pregnancy 173
Comparison with COMA report for lactation 173
glOSSary 175
glOSSary OF StatIStICal tErmS 179
aBBrEvIatIONS 181
rEFErENCE lISt 183
11
1 Summary
Introduction The Dietary Reference Values (DRVs) for food energy provide a best estimate of S1. the food energy needs of the UK population and its subgroups and present criteria against which to judge the adequacy of their food energy intakes (Department of Health [DH], 1991). The DRVs for food energy are defined as the Estimated Average Requirement (EAR). In adults, the EAR for energy has previously been set at the level of energy intake required to maintain weight i.e. an energy intake which matches energy expenditure. During infancy and childhood, the requirement also has to meet the needs for healthy growth and development, while during pregnancy and lactation the requirement must meet the needs for development of a healthy baby and supporting adequate lactation. Any sustained imbalance between energy intake and expenditure will lead to progressive gain or loss in weight.
The National Diet and Nutrition Survey (NDNS) series shows average reported S2. energy intakes to be consistently below the level indicated by the EAR for food energy defined in the 1991 report of the Committee on Medical Aspects of Food and Nutrition Policy (COMA) (DH, 1991). In reality, average habitual energy intakes in the United Kingdom (UK) are more likely to exceed energy needs, as evidenced by the increasing number of people classified as overweight or obese. Under-reporting of food intake may explain this paradox as it is known to largely account for a commonly reported discrepancy between measured total energy expenditure (TEE) and the apparent low energy intakes reported in NDNS and other dietary surveys.
These observations, the expanding evidence base on TEE in a wide variety of S3. population groups, together with the publication of the Food and Agriculture Organization of the United Nations, World Health Organization, and United Nations University (FAO/WHO/UNU) updated recommendations for energy intake from the expert consultation on Human Energy Requirements (FAO, 2004), led the Food Standards Agency and the Department of Health to request the Scientific Advisory Committee on Nutrition (SACN) to re-evaluate of the DRVs for food energy. Subsequent to this, the US Institute of Medicine (IoM) published revised Dietary Reference Intakes (DRIs) for energy (IoM, 2005).
Terms of reference The Terms of Reference for the Energy Requirements Working Group were to:S4.
Review and agree on the interpretation of the methods, definitions and • assumptions used by COMA (DH, 1991) and FAO/WHO/UNU expert consultation on Human Energy Requirements (FAO, 2004) to agree energy requirements.
Agree a framework by which to arrive at energy requirements for the UK • population and its subgroups.
12
Agree population-based Dietary Reference Values for energy, and provide • recommendations taking into account age, body size, levels of activity, gender and physiological state (i.e. growth, pregnancy and lactation).
Consider the implications of these recommendations on the requirements for • other nutrients1.
In addressing these Terms of Reference, the report considers: components of and S5. factors affecting energy expenditure; the measurement and estimation of energy expenditure; energy expenditure in representative populations of children and adults; and energy requirements for the UK population in health.
Background
Components of energy expenditure The TEE of an individual can be divided into a number of discrete components that S6. can be determined separately. These are basal metabolic rate (BMR), the energy expended in physical activity, and other components such as the thermic effect of food (TEF) and growth.
Basal metabolic rate Basal metabolic rate (BMR) is a standardised measure of an individual’s metabolism S7. in a basal state: i.e. while awake and resting after all food has been digested and absorbed, and at thermal neutrality. For most individuals, BMR is the largest component of energy expenditure and requirements, ranging from 40-70% depending on age and lifestyle. BMR can be estimated from body size, age and gender. A variety of BMR prediction equations have been described in the literature, with the Schofield equations being primarily used in the COMA (DH, 1991) and FAO/WHO/UNU (FAO, 2004) reports.
Physical activity Physical activity includes a wide range of behaviour and encompasses sitting, S8. standing, walking, and planned and structured exercise which may have the objective of maintaining or improving physical fitness or health. Physical activity- related energy expenditure (PAEE) is quantitatively the most variable component of TEE usually accounting for 25-50% of energy expenditure, up to a maximum of 75% in some unusual circumstances.
The physical activity level (PAL) TEE can be expressed as a multiple of BMR, the physical activity level (PAL). Hence, S9. TEE or EAR is equal to BMR x PAL. PAL is theoretically independent of those factors influencing BMR (weight, age and gender), at least as a first approximation, and consequently, for any PAL value, TEE can be predicted for any group from estimates of the BMR. During growth, pregnancy and lactation the energy cost of tissue deposition also needs to be taken into account.
1 In most cases, energy intake has increased and therefore it was felt that there was no need to undertake this.
13
PAL values are best estimated from direct measures of 24 hour TEE and BMR. S10. Such measurements in free-living populations have indicated that PAL can range from <1.3 in immobile subjects to between 3-4.7 for limited periods of time in, e.g. soldiers on field exercises or elite endurance athletes. Within the general population, however, the overall range of PAL values for individuals in energy balance, leading sustainable lifestyles, is between 1.38 for the most sedentary to 2.5 for the most active.
In previous reports, including COMA (DH, 1991), the duration and energy cost of S11. individual activities (as physical activity ratio (PAR) values) has been summed to provide factorial estimates of PAL. The factorial estimation of PAL assumes that overall energy expenditure within the general population can be predicted using information derived from activity diaries or lifestyle questionnaires. However, there is little evidence that this can be done with sufficient accuracy for two main reasons:
For most lists of activities used to identify lifestyle categories, individual activities • are only defined in general qualitative terms and consequently there can be large inter-individual variations in the energy expended for each activity.
Factorial estimates take no account of variations in spontaneous physical • activity (SPA), a term used to describe all body movements associated with activities of daily living, change of posture and ‘fidgeting’. SPA accounts for a between-individual variation in energy expenditure of ±15%. The potential for variable SPA throughout the range of defined activities adds to the uncertainty that their energy costs can be predicted.
In this report, factorial predictions of PAL values for specific lifestyle categories S12. have not been attempted due to the likely error in estimating actual TEE from listed PAR values of activities and the phenomenon of behavioural phenotypes which exhibit very different rates of energy expenditure. Instead, values for PAL for children, adolescents and adults have been identified from an analysis of the available TEE literature judged to be appropriate for the UK population.
Definition of energy requirements The energy requirement and associated descriptive terminology must be defined S13. with particular care in the context of a population which includes many individuals not defined as healthy because they are overweight or exhibit habitually low levels of physical activity. In 1991, COMA defined requirements in general terms only i.e. intakes of nutrients which were likely to “meet the needs” of some or all within population groups. COMA did not offer a specific definition of the energy requirement in relation to specific body weights within the healthy range or prescribed levels of energy expenditure. COMA set EAR values for population groups calculated for a wide range of physical activity levels and adult body weights which in practice would maintain the status quo in terms of existing body weights. In populations like the UK, with a high and increasing proportion of overweight and obese individuals, application of such energy reference values would therefore maintain body weights in excess of healthy reference
14
body weights. In contrast, the FAO/WHO/UNU reports of 1985 and 2004 used a ‘normative’ or ‘prescriptive’ approach2 to calculate energy requirements. That is, suitable reference body weight ranges consistent with long-term good health were used to calculate energy reference values. Adoption of these prescriptive values by groups with body weights below or above such ranges would tend to mediate weight change towards the healthier, more desirable body weight range as opposed to maintaining body weights which exceed healthy reference values.
In this report, the requirements for energy for all population groups, with the S14. exception of pregnant women, have been set at the level of energy intake required to maintain a healthy body weight in otherwise healthy people at existing levels of physical activity. Allowances are made for any additional physiological needs (i.e. growth, pregnancy and lactation). Thus, the report recognises the increasing prevalence of overweight and obesity in the current UK population, and has adopted the prescriptive terminology and principle (i.e. relating to a desirable standard as distinct from status quo) for body weight, with physical activity set at best estimates of existing levels. Consequently, energy reference values defined in this report are derived for infants, children and adults in relation to body weights which, on the basis of current evidence, are likely to be consistent with long- term good health. This means that for people who are overweight or obese, energy intakes at the reference levels should enable the transition towards the healthy body weight range (i.e. Body Mass Index (BMI) 18.5-24.9kg/m2 for adults). For pregnant women who are overweight or obese a precautionary approach has been adopted and it is recommended that energy requirements for this group are estimated based on actual preconceptional body weights.
As with previous reports, the importance of adequate physical activity is recognised S15. and this report includes advice on desirable physical activity consistent with long- term health. The likely impact of such advice on energy requirement values is also described.
Variability Measurements of energy needs that are used to predict DRVs for a particular S16. population group will, for a number of reasons, exhibit variability.
For most nutrients the population reference value is identified as the Reference S17. Nutrient Intake (RNI)3. However, for dietary energy the DRV is defined differently i.e. it is equal to the average reference value (EAR). The RNI for dietary energy is not used because it represents an excess energy intake for the majority of the population. Energy intakes that consistently exceed requirements lead to weight gain and obesity in the long term. An intake equal to the average reference value for a population group on the other hand is, in theory, associated with similar probabilities of excessive and insufficient energy intakes for any individual within the population.
2 The terms ‘normative’ and ‘prescriptive’ can be used interchangeably.
3 The reference nutrient intake for a nutrient is the amount of the nutrient that is enough, or more than enough, for about 97% of people in a group.
15
In the UK and similar countries, the population is characterised by sedentary S18. lifestyles at all ages, and thus rates of energy expenditure and intakes that maintain energy balance at these levels of energy expenditure are unlikely to be normally distributed. Measurements of energy expenditure within adult populations indicate that the overall distribution is skewed towards sedentary behaviour. For the high proportion of individuals who are relatively inactive, energy expenditure and intakes that maintain energy balance will cluster at the lower end of the range. A small proportion will be more active, with a few individuals exhibiting the highest rates of expenditure that are sustainable. Statistical considerations, therefore, dictate that the appropriate descriptor of the midpoint of the distribution is the median, which is likely to be somewhat lower than the mean, thus the median is used in this report.
Evidence that energy expenditure varies between individuals with similar lifestyles S19. over a wider range than would be expected (at least in adults) presents a second difficulty in estimating energy reference values. This is particularly important since in the past, attempts have been made to define energy reference values in terms of specific physical activity levels for specific lifestyle population groups. Consequently, with the probable exception of those at the extremes of the activity range, lifestyle predictions of energy expenditure and resultant energy reference values cannot be made with the certainty often assumed by users of previous reports such as COMA (DH, 1991).
Approaches used to set energy reference values Measurement or prediction of TEE provides a physiological measure of the energy S20. requirement at energy balance. This is because in most circumstances energy intake must match energy expenditure to achieve energy balance. Thus, the energy requirement can be predicted specifically as the rate of TEE plus any additional needs for growth, pregnancy and lactation.
Measurement of total energy expenditure The S21. doubly labelled water (DLW) method is the most accurate practical means of measuring TEE in free-living individuals over a period of several days to several weeks. In this method, the TEE of individuals is computed from estimates of CO
2
production, which are calculated from the loss of the isotopes 18O and 2H from the body over time following administration of a known dose.
Utilising total energy expenditure measurements to derive energy requirements In recent years, EAR values for populations have been estimated from DLW-derived S22. measures of TEE in reference populations together with estimates of deposited energy. Two potential analytical approaches can be employed.
Predicting TEE with regression equations TEE can be modelled against different anthropometric characteristics of the S23. reference population (such as age and body weights) using regression equations
16
and the regression equations can then be used to predict TEE and resulting EAR values for any population group based on their anthropometric variables. A major limitation to this approach has been the inability of TEE prediction models to account for variation in Physical Activity Energy Expenditure (PAEE), both between- individual and between-group), an important source of variation in TEE, in a transparent way.
Previous reports which have used regression models of TEE to predict energy S24. requirements, the FAO/WHO/UNU report (FAO, 2004)4 and the US report on Dietary Reference Intake values for energy (IoM, 2005) have made special provision for the likely variation in PAEE within the population group described by the regression.
Predicting TEE from BMR and measured PAL values – the factorial approach The second analytical approach to the derivation of energy requirements employs S25. a factorial approach5 based on the assumption that TEE (or EAR) is equal to BMR x PAL. This involves the following steps:
TEE values measured in a reference population are divided by measured or • estimated BMR values from that population to extract PAL values. This means that the reference populations studied by DLW are described primarily by PAL values.
For the population of interest, BMR values are then estimated from BMR • prediction equations using relevant anthropometric data for the population.
The PAL values derived from the reference population can then be used to • estimate TEE and EAR values for the population of interest based on the latter’s estimated BMR values.
When utilising this factorial approach, suitable PAL values appropriate for specific S26. groups and populations should be identified from measurements of DLW-derived TEE obtained from large population studies which are representative of the current UK population and which can be assumed to exhibit similar activity patterns. Once suitable PAL values have been established, their distribution can be evaluated for the population as a whole in order to identify the medians and centile ranges. This approach allows energy reference values to be framed against such distributions, in effect substituting PAL distributions for PAL values defined in terms of lifestyle. Thus energy reference values can be defined as a population EAR (i.e. from the median PAL value) with additional values appropriate to those who are more or less active than the average (i.e. from the 25th and 75th centiles).
4 Regression equations were used to derive energy requirements for infants, children and adolescents. The energy requirements of adults were calculated from factorial estimates of habitual TEE.
5 This approach has also been referred to as the BMR multiple approach and the PAL model.
17
Summary of approach used by SACN to determine revised energy reference values The SACN Framework for the Evaluation of Evidence has been used as the basis S27. to identify and assess published evidence of TEE from which to guide derivation of energy reference values. Only studies using the DLW method to measure TEE were considered. No large-scale population studies of any age group have been conducted to determine TEE in the UK population, and TEE data from the NDNS series are as yet insufficient for the determination of energy reference values. Other data sources (reference populations) have therefore been used to derive the revised EAR values. Priority was given to studies in well-characterised populations (e.g. by age, gender, weight, height or BMI) and where BMR had been measured directly rather than estimated from prediction equations. The majority of studies identified were cross-sectional studies in healthy human populations of infants, children and adults. Those identified as being the most likely to be representative of the UK population were as follows:
For infants, a longitudinal study of healthy American infants comprising similar • numbers of breast fed and breast milk substitute-fed infants (n=76 individuals).
For children and adolescents, a compilation of all identified mean study values • for specific ages for boys and girls (n=170) representing a total of approximately 3500 individual measurements.
For adults, two large data sets of individual TEE values (n=929) were obtained • from the USA, the OPEN and Beltsville studies.
For older adults, no specific representative data set has been identified. •
The limitations of these data sets (reference populations) for deriving estimated S28. EARs for the UK population include the small numbers of participants in the infant and at some ages in the child cohorts and the lack of younger adults aged 18-30 years and older adults aged >80 years in the larger adult data set.
For adults aged 19-65 years and children aged between 3-18 years, the different S29. approaches to setting EARs, namely regression modelling and the factorial approach, were examined. Regression modelling was explored, but the approach was not pursued because the inability of TEE prediction models to account for variation in PAEE in a transparent way was considered a major limitation. A factorial model was therefore adopted in which TEE is predicted from BMR x PAL, following the steps outlined in paragraph S25.
Prescriptive energy reference values for healthy body weights The increasing prevalence of overweight and obesity at all ages in the population S30. raises a clear difficulty in the definition of energy reference values for population groups. Such values, calculated to match TEE, will, for many of the population, maintain overweight and will therefore not be consistent with long-term good health. Even for subjects exhibiting desirable physical activity, excess body weight
18
may be associated with increased risk of mortality. Given this report’s objective of defining prescriptive reference values, healthy reference body weights have been identified and applied.
Factorial approach to setting EARs As described above (paragraphs S11 and S12), in contrast to the approach taken S31. by COMA in 1991 (DH, 1991), the summation of the duration and energy cost of individual activities to derive estimates of PAL has not be used here. Instead, the distribution of PAL values within the reference population has been used to indicate a population average value (median) for PAL as well as the extent to which it is lower (25th centile) or higher (75th centile) for less or more active population groups. The PAL values for adults, as median, 25th and 75th centiles, are those calculated directly from individual TEE values reported in the OPEN and Beltsville data sets. For children over one year of age, the PAL values are those calculated directly from a data set of published DLW studies which were aggregated on the basis of study mean values. Estimates of the likely increase in PAL associated with varying increases in activity level are also given. Due to a lack of substantial new data, the reference values stated in the FAO/WHO/UNU report (FAO, 2004) were used for infants. Pregnant and lactating women were considered separately.
In this report, BMR for children and adults is estimated using the Henry equations S32. applying healthy body weights as indicated by the 50th centile of the UK-WHO Growth Standards (ages 1–4 years), the 50th centile of the UK 1990 reference for children and adolescents aged >4 years, and at weights equivalent to a BMI of 22.5kg/m2 at the appropriate height of the adult population group (see paragraph S36).
COMA reported EAR values for a range of body weights and PAL values for S33. adolescents and adults (DH, 1991). However, the most widely cited values from that report are aggregated EAR values for these groups which were based on low PAL values thought to be in keeping with the sedentary lifestyle of the UK population: i.e. 1.56 for boys, 1.48 for girls and 1.4 for adults. An analysis of the range and distribution of PAL values shows that COMA is likely to have underestimated the PAL values of general sedentary populations because of an under-appreciation of the influence of routine activities of daily living on energy expenditure. Values used by COMA are lower than those observed for 90% of the subjects in the reference adolescent and adult populations examined for this report. Thus, median PAL values identified for adolescents and adults in this new report are 1.75 and 1.63 respectively, with 1.63 representing the median PAL value of a reference adult population in which, like the UK, approximately 60% are overweight or obese.
19
Energy reference values for infants, children and adolescents
Energy cost of growth TEE measured using the DLW method includes the energy expended in tissue S34. synthesis. Therefore, only the cost of energy deposited in newly synthesised tissues should be added when calculating the energy reference values for infants, children and adolescents.
Infants aged 0 – 12 months Following the approach of the FAO/WHO/UNU Expert Consultation (FAO, 2004), S35. the energy reference values for infants are calculated from the energy deposited in new tissue plus TEE. TEE was predicted from a simple equation expressing TEE as a function of weight which was derived from a longitudinal study in which TEE was measured by the DLW method in healthy, well-nourished, non-stunted infants, born at full term with adequate birth weight, and growing along the trajectory of the UK-WHO Growth Standard (Royal College of Paediatrics and Child Health [RCPCH], 2011). Costs of tissue deposition were calculated from an analysis of the body composition of a population of healthy US infants during normal growth. They were then applied to weight increments observed in the WHO Multicentre Growth Reference Study. See Tables 5 and 14 for revised EAR values of infants 1-12 months (pages 51 and 83).
Children and adolescents aged 1 – 18 years The energy reference values for boys and girls aged 1-18 years were calculated as S36. TEE plus deposited energy costs using a factorial model BMR x PAL. The energy value of tissue deposited was assumed to be equivalent to a 1% increase in PAL. The population EAR values are calculated at median PAL values6 for best estimates of healthy body weights i.e. the 50th centiles of the UK-WHO Growth Standards (ages 1–4 years) and the UK 1990 reference for children and adolescents for children aged over four years of age. These reference weights are about 15% lower than current UK weights and thus for those children who are overweight and for any underweight children, energy intakes at these levels will be associated with weight change.
EAR values have also been calculated for less active or more active children by S37. the 25th and 75th centile PAL values. See Tables 8 and 15 for revised EAR values for children aged 1–18 years old (pages 56 and 84).
Energy reference values for adults Energy reference values for adults were derived by the factorial calculation of TEE S38. from BMR x PAL. BMR values are calculated using the Henry equations at weights equivalent to a BMI of 22.5kg/m2, which can be considered to be a healthy body
6 where median PAL =1.4 for children aged 1-<3 years, 1.58 for children aged 3-<10 years, and 1.75 for children aged 10–18 years.
20
weight7, and at the relevant height of the population. For illustration, current mean heights are used for England and Scotland; representative data for Wales and Northern Ireland were not available. PAL values were identified from an analysis of suitable DLW measures of TEE.
Identifying PAL values The approach taken to determine PAL values was to identify a data set of DLW S39. measures of TEE which could serve as a reference distribution of TEE and PAL values from which energy reference values for the UK adult population could be estimated.
An initial survey of all published and other available DLW measures of TEE in S40. healthy adults identified published studies which could be utilised in terms of study means. Two sets of individual data points were also considered:
a DLW data set of individual values drawn from UK national dietary surveys (the • NDNS (n=156) and Low Income Diet and Nutrition Survey (n=36)).
the DLW data set (n=767) assembled for the US DRI report which includes most • of the UK studies published up to the writing of that report.
With the exception of the NDNS data sets, subjects were not recruited to S41. these studies explicitly as a representative sample of the UK or any other adult population; NDNS randomly selects participants while recognising that there may be some recruitment bias. In contrast, the DLW data assembled for the DRI report “were not obtained in randomly selected individuals ... and do not constitute a representative sample of the United States and Canada.” While the NDNS data sets are most appropriate, these are currently too small to serve as a reference.
Since publication of the DRI report in 2005, two large population-based studies S42. from the United States, the OPEN study (n=451) and the Beltsville study (n=478) have been published, in which TEE was measured using DLW. The combined OPEN and Beltsville cohorts contained similar number of men (48%) and women (52%), with levels of overweight and obesity very similar to current levels reported in the Health Surveys for England and Scotland. Mean BMI values were 27.0 kg/m2 (F), 27.5 kg/m2 (M) compared with current UK values of 27 kg/m2 (F and M); 39% were classified as overweight and 25% obese, compared with 38% overweight and 23% obese in England. Demographic characteristics were also similar to the current UK population, although some ethnic differences may exist. The combined OPEN/Beltsville data set has been used as the reference population in this report from which a median PAL value has been identified. Revised EARs for the UK adult population have then been derived employing the factorial approach of BMR x PAL = TEE, in which PAL is that derived from the reference population and BMR is calculated from the Henry equations at healthy body weights equivalent to a BMI of 22.5kg/m2 at the height of the population group (see paragraph S30 above).
7 A body weight equivalent to a BMI of 22.5kg/m2 is at the lower end of the body weight range (BMI 22.5-25kg/ m2) associated with a minimum risk of mortality (based on a collaborative analysis of the influence of BMI on all- cause mortality in 57 prospective studies, n=900,000) and can be considered to represent a healthy body weight.
21
The distribution of PAL values within the combined data set (n=929 individual S43. measures) was 1.01 to 2.61. In healthy, mobile individuals, the overall range of PAL values is normally between 1.38 and 2.5; PAL values can fall as low as 1.27 and may exceed 2.5, but such low and high values are not thought to be sustainable. The combined data set was therefore trimmed for PAL values <1.27 or >2.5. After trimming, the median PAL value used to predict the population EAR was 1.63. The 25th and 75th centile PAL values used to estimate energy reference values for groups judged to be less or more active than average were 1.49 and 1.78 respectively. The report also provides estimates of the probable additional energy needs, defined as increases in PAL, associated with changes in habitual activity involving specific types of activity such as increased walking, running and participation in sport or high-level training regimes.
The derivation of EAR values using PAL values from a predominately overweight S44. and obese population was carefully considered given the report’s objective of identifying prescriptive EAR values consistent with long-term good health i.e. values suitable for weight maintenance at a healthy body weight at existing levels of physical activity. For those subjects within the normal body weight range, the median PAL (1.61) was not significantly different from that of the overweight and obese men or women. Furthermore, investigation of the influence of BMI on PAL values within the combined data set by regression analysis indicated that PAL did not significantly vary with BMI. This indicates that the level of energy expenditure as indicated by the median PAL for the cohort (i.e. 1.63) is not a specific characteristic of overweight or obesity and its use in deriving prescriptive EAR values consistent with healthy body weights is justifiable.
The revised population EAR values for all adults, calculated using a PAL value of S45. 1.63 and BMR values calculated at weights equivalent to a BMI of 22.5kg/m2 at current mean heights for age, are 10.9 MJ/d (2605kcal/d) for men and 8.7 MJ/d (2079kcal/d) for women. See Tables 11, 12 and 16 for revised EAR values for adults (pages 61, 61 and 85).
Energy reference values for older adults Age-related changes in lifestyle and activity are very variable and many older S46. people exhibit relatively high levels of activity. For healthy, mobile older adults energy requirements are unlikely to differ substantially from younger adults and reference values can be described in the same way. However, for those individuals with much reduced mobility it can be assumed that PAL values are likely to be lower. For these individuals and for older people who are not in good general health, energy requirements can be based on the less active, 25th centile PAL value of 1.49, recognising that for some groups of older people with specific diseases or disabilities or for patient groups who are bed-bound or wheelchair bound, the PAL value may be consistently lower than this.
22
Effects of additional physical activity on energy requirements Although the prediction of PAL values associated with a specific lifestyle cannot S47. be made with any certainty, predictions of the likely additional energy cost of well-defined specific activities can be made with reasonable confidence. Examples are given for increments in PAL associated with increased activities ranging from 0.15 for 30 minutes of moderate intensity activity on five or more days of the week, to 0.6 for an intense aerobic exercise programme associated with training for competitive sport daily. These values have been derived primarily from studies on adults but there is no reason to believe that they will be substantially different for children and adolescents.
Reference energy values during pregnancy and lactation The energy requirements for pregnancy and lactation are calculated as increments S48. to be added to the mother’s EAR. These are based on singleton pregnancies reaching term.
Ideally, women should begin pregnancy at a healthy weight (BMI 18.5-24.9 kg/mS49. 2) and the EARs for non-pregnant women identified in this report are set at amounts consistent with maintaining a BMI of 22.5kg/m2. Women who are underweight or overweight at the beginning of pregnancy are at risk of poor maternal and fetal outcomes. Women who are underweight benefit from greater weight gain during pregnancy. For women who are overweight and obese, the consequences of weight change during pregnancy are not completely understood. Given this uncertainty, a precautionary approach has been adopted and weight loss during pregnancy is not advised. The EARs for pregnancy and lactation defined in this report are therefore estimates of the incremental energy intakes likely to be associated with healthy outcomes for mother and child, for women consuming energy intakes which match energy expenditure at the commencement of pregnancy. That is, for women who are overweight the incremental energy intakes should be added to EAR values calculated at preconceptional body weights, rather than at healthy body weights for non-pregnant women.
Energy reference values for pregnancy Energy reference values for pregnancy estimated by the factorial method in previous S50. reports such as the FAO/WHO/UNU (FAO, 2004) and US DRI (IoM, 2005) reports exceed energy intakes observed in well-nourished populations with average birth weight in the healthy range. Adherence to these energy reference values may lead to inappropriate maternal fat gain, some of which may remain after lactation. Consequently, it was not considered necessary to amend the increment of 0.8 MJ/day (191kcal/day) in the last trimester previously recommended by COMA. Women entering pregnancy who are overweight may not require this increment but current data are insufficient to make a recommendation regarding this group.
Energy reference values for lactation Factorial calculation suggests that women with a healthy pre-pregnancy weight S51. (BMI 18.5-24.9 kg/m2) who exclusively breastfeed their infants throughout the
23
first six months require an increment of 2.1 MJ/day (502kcal/day) above the pre-pregnant EAR. However, the factor of 0.8 applied to adjust for efficiency of conversion of maternal energy intake to milk is insecure and likely to overestimate the true synthetic cost. This could explain why the factorial estimate is appreciably greater than that measured in DLW experiments. It is recommended that the US Energy DRIs (IoM, 2005) (based on DLW measurements) of 1.38 MJ /day (330 kcal/ day) in the first six months of lactation are applied. Thereafter, the energy intake required to support breastfeeding will be modified by maternal body composition and the breast milk intake of the infant.
Comparisons with reference values for energy in the COMA 1991 report The energy reference values detailed in this report derive from a methodology S52. which differs to a greater or lesser extent from COMA, the US DRI report, and the FAO/WHO/UNU. As a result of these methodological differences, the revised EAR values differ from COMA in a non-uniform way. For pre-adolescent children (aged 3 months to 10 years) the revised EAR values are lower by between 2-22%. However, the revised EAR values are generally higher for adolescent boys (by 6-9%) and girls (by 15-18%). For adults, including older adults in whom general health and mobility are maintained, the new population EAR values for all men and women are higher by 3% for men and by 7-9% for women due to the combined use of higher PAL values with lower BMI and BMR values. See Tables 38 and 39 for details.
Although for some population groups the revised population EAR values are higher S53. than previous estimates, this should not be interpreted to mean that these groups have increased their activity and therefore need to eat more, but rather that the new values represent a closer approximation of energy needs at current activity levels, estimated using updated methodology.
The revised EARs for dietary energy can be found in Tables 5 and 14 for infants, S54. Tables 8 and 15 for children aged 1- 18 years old, and Tables 11, 12 and 16 for adults.
Gaps in the evidence base and consequent recommendations for future research S55. are detailed in paragraphs 203-206.
24
2 Introduction
Background Dietary Reference Values (DRVs) for food energy describe the requirements for 1. food energy of population groups, and provide criteria against which to judge the adequacy of the food energy intake for the UK population and its subgroups (Department of Health [DH], 1991). These DRVs apply to groups of healthy people and are not appropriate for the definition of requirements for individuals. The DRVs support the maintenance of health in a population and are derived with the assumption that the requirements for all other nutrients are met.
The DRVs for food energy are defined as the Estimated Average Requirement 2. (EAR)8. They are used for various purposes which include informing the provision of food in clinical and institutional settings. In adults, the EAR for energy has been set at the level of energy intake required to maintain body weight i.e. an energy intake which matches energy expenditure. During infancy and childhood, the requirement also has to meet the needs for healthy growth and development, while during pregnancy and lactation the requirement must meet the needs for carrying a healthy baby and supporting adequate lactation. Any sustained imbalance between energy intake and expenditure will lead to progressive gain or loss in body weight.
The National Diet and Nutrition Survey (NDNS)3. 9 series10 shows average reported energy intakes to be consistently below the level indicated by the EAR for food energy defined in the 1991 report of the Committee on Medical Aspects of Food and Nutrition Policy (COMA) (DH, 1991). However, these, and other surveys of the UK population (The NHS Information Centre [NHS IC], 2010; Reilly et al., 2009), also show that the mean Body Mass Index (BMI) and number of people classified as overweight or obese11 is increasing. The latter data indicate that average habitual energy intake does, in fact, exceed energy needs.
One explanation for this apparent paradox is that reported intakes of food energy 4. assessed in the NDNS may be lower than intakes actually consumed (Stephen et al, 2007; Poppitt et al., 1998). Under-reporting of food intake, which is significant and particularly pronounced in people who are overweight and obese (Rennie et al., 2007; Westerterp & Goris., 2002), largely accounts for the apparent low energy intakes observed in NDNS and other dietary surveys.
8 The Estimated Average Requirement (EAR) is an estimate of the average requirement for energy or a nutrient and assumes normal distribution of variability.
9 The National Diet and Nutrition Survey (NDNS) is a UK survey of the food consumption, nutrient intakes and nutritional status of people aged 1.5 years and older living in private households. The NDNS rolling programme is currently collecting data from 2008-2012. Previously it was a series of discrete cross-sectional studies.
10 See publications by Finch et al., 1998; Gregory et al., 1990; Gregory et al., 1995; Gregory et al., 2000; Henderson et al., 2002; Henderson et al., 2003a; Henderson et al., 2003b; Hoare et al., 2004; Nelson et al., 2007; Ruston et al., 2004; .
11 Body Mass Index (BMI) (kg/m²) is often used as a convenient measure of adiposity. In adults, cut-off points for underweight, overweight and obesity are defined as BMI values of <18.5 kg/m2, >25 kg/m2 and >30 kg/m2 respectively.
25
Another possible explanation for the apparent paradox is that the EAR for food 5. energy which has been used to evaluate measured energy intakes in the UK is, in fact, set at too high a value.
The EARs for food energy published by COMA (DH, 1991) were based on limited 6. available evidence at that time. More recent observations on the energy expenditure in a wide variety of population groups together with the publication of the Food and Agriculture Organization of the United Nations, World Health Organization, and United Nations University (FAO/WHO/UNU) updated recommendations for energy intake from the expert consultation on Human Energy Requirements (FAO, 2004) indicate that the EAR values proposed by COMA are unlikely to be too high, at least for some groups.
Even relatively sedentary populations are now thought to exhibit higher rates 7. of energy expenditure than previously estimated (Black, 1996; Black et al., 1996). Consequently, the revised energy reference values published by FAO/WHO/ UNU in 2004 differ from those set by COMA in 1991, being generally higher. For these reasons, it was timely to review the evidence and the Food Standards Agency (FSA) and the Department of Health therefore requested the Scientific Advisory Committee on Nutrition (SACN) to re-evaluate the DRVs for food energy. Subsequent to this, the US Institute of Medicine (IoM) published revised Dietary Reference Intakes (DRIs) for energy (IoM, 2005).
Terms of reference The Terms of Reference for the Energy Requirements Working Group were to:8.
Review and agree on the interpretation of the methods, definitions and • assumptions used by COMA (DH, 1991) and FAO/WHO/UNU expert consultation on Human Energy Requirements (FAO, 2004) to agree energy requirements.
Agree a framework by which to arrive at energy requirements for the UK • population and its subgroups.
Agree population-based Dietary Reference Values for energy, and provide • recommendations taking into account age, body size, levels of activity, gender and physiological state (i.e. growth, pregnancy and lactation).
Consider the implications of these recommendations on the requirements for • other nutrients12.
Report content The report considers: 9.
energy available from food and drink; •
components of and factors affecting total energy expenditure (TEE); •
the measurement and prediction of TEE; •
12 In most cases, energy intake has increased and therefore it was felt that there was no need to undertake this.
26
TEE in representative populations of children and adults; and •
energy requirements for the UK population in health. •
Physical activity and energy balance, and the nature and extent of obesity in the UK population, are also covered.
Methodology The working procedures for the preparation and finalisation of the report are 10. described in Appendix 1.
Units of energy The unit of energy in the International System of Units (SI) is the joule (J) and is the 11. energy expended when an object is moved one metre by a force of one newton in the direction in which the force is applied. A newton is the SI unit of force and one newton will accelerate a mass of one kilogram at the rate of one metre per second. The units normally used to express energy are the kilojoule (kJ = 103 J) and the megajoule (MJ = 106 J). The thermochemical calorie is equivalent to 4.184 J (1 kcal = 4.184 kJ).
Energy available from food and drink Ingested food and drink contains chemical energy in the form of carbohydrate, fat, 12. protein, and alcohol. The maximum amount of energy potentially available from a food by an organism can be determined by measuring the heat released after its complete combustion to carbon dioxide and water. This is the gross energy (GE), but not all GE from food is available for human metabolism because of losses during food utilisation.
Available energy is defined as metabolisable energy (ME) and the ME value of a 13. food or diet can be measured as the difference between energy intake and all losses (mainly in faeces and urine and a small amount in sweat). Energy losses in faeces largely represent incomplete digestion, while energy losses in urine and sweat are due primarily to the urea content which represents the incomplete catabolism of protein.
The ME content is the value quoted as the energy content of foods on food labels 14. and in the UK food composition tables (FSA, 2002). The ME content of a given food can be calculated from the amounts of protein, fat, carbohydrate, and alcohol in the food (determined by chemical analysis) using energy conversion factors (see Table 1). These conversion factors (called the Atwater factors) are estimates of the energy content of each macronutrient and alcohol and have been rounded for practical purposes. Alternatively, UK food composition tables of the ME content of a wide variety of commonly consumed foods can be used; these are based on analytical data. More detail on the energy yielded from different nutrients can be found in Appendix 2.
27
Table 1 Metabolisable energy (ME) conversion factors for nutrients and alcohola
ME factors kJ/g kcal/g
Fat 37 9.0 Carbohydrate Available carbohydrate expressed as monosaccharide 16 3.8 Fermentable non-starch polysaccharides 8 1.9 Organic acids 13 3.1 Polyols 10 2.4 Protein 17 4.0 Alcohol 29 6.9
a Food Standards Agency, 2002
Some ME utilisation results in energy lost as heat i.e. the thermic effect of food 15. (TEF) (Schutz et al., 1984) (see paragraph 31). The extent of this varies with the type of food ingested, but specific amounts are associated with amino acid metabolism, alcohol metabolism and the microbial fermentation of otherwise unavailable carbohydrate. Energy lost via TEF may be subtracted from ME, resulting in an expression of the capacity of food energy to fuel metabolic work or for conversion to stored energy. This is termed the net metabolisable energy (NME) (FAO, 2003). While the ME and NME values are the same or similar for available carbohydrate and fat, NME values are lower for protein, alcohol and fermentable non-starch polysaccharides.
The human body is able to capture some of the chemical energy from food through 16. cellular metabolism, resulting in the generation of an intermediary chemical form, adenosine triphosphate (ATP). ATP acts as an energy source for cellular processes mainly through phosphorylation of proteins and other intermediates. It is regenerated from adenosine diphosphate (ADP) using the energy in food. Cells require chemical energy for three general types of tasks: 1) to drive metabolic reactions that would not occur automatically; 2) for the transport of substances across cell membranes especially, against a concentration gradient; and 3) for mechanical work, e.g. muscle contraction. Energy is also released as heat both in the formation of ATP and during its use in these metabolic processes, and this maintains body temperature. Food energy can also be directly converted to heat if the oxidation pathway is uncoupled from the ATP-producing process.
FAO/WHO/UNU reviewed the case for the use of NME in place of ME values 17. (FAO, 2003). While it was recognised that NME represents the biological ATP- generating potential of foods, it was recommended that, for the present, the ME system should be retained. The FAO/WHO/UNU recommendation (FAO, 2004) is endorsed by this report.
28
Components of energy expenditure The total energy expenditure (TEE) of an individual can be divided into a number of 18. discrete components, which can be determined separately. These are daily energy expended at rest, the basal metabolic rate (BMR), energy expended in physical activity, and other components such as thermogenesis resulting mainly from food intake, and growth. These components, and their variation as a function of body size and composition, and other factors are described in detail in Appendix 3 and are briefly summarised here.
Basal and resting metabolism The basal metabolic rate (BMR) is a standardised measure of an individual’s 19. metabolism in a basal state: i.e. while awake and resting after all food has been digested and absorbed, in a thermoneutral environment. BMR represents the metabolic activity of cells and tissues and the physiological functions essential for life. For most individuals, BMR is the largest component of energy expenditure and requirements, ranging from 40 to 70% depending on age and lifestyle. BMR can be estimated from body size, age and gender; suitable equations are discussed in Appendix 4.
Physical activity Physical activity includes a wide range of behaviour and encompasses sitting, 20. standing, walking, planned exercise, and so on (Wareham & Rennie, 1998). Thus exercise, as planned and structured physical activity, is a subset of physical activity that may have an objective of maintaining or improving physical fitness and/or health. The degree of physical fitness of an individual can be measured with specific tests (Caspersen et al., 1985). Other subsets of physical activity include spontaneous physical activity (SPA), a term used to describe all body movements associated with activities of daily living, change of posture and ‘fidgeting’ (Ravussin et al., 1986).
Physical activity-related energy expenditure (PAEE) is quantitatively the most 21. variable component of TEE usually accounting for 25-50% of energy expenditure and in some unusual circumstances up to a maximum of 75% (Westerterp, 1998). PAEE can have an impact on the metabolic rate beyond the period of a specific exercise for a few hours or even longer; this is described as excess post-exercise oxygen consumption (EPOC).
The energy costs of different physical activities can be expressed as multiples of 22. BMR, called physical activity ratios (PAR), to account for differences in body size. Thus, PAR is used for discreet time periods of several hours or less. The physical activity level (PAL) over 24 hours is used as a measure of daily TEE adjusted for BMR.
The physical activity level (PAL) A detailed discussion of PAL is provided in Appendix 5: an overview is given here. 23. Twenty four hour TEE is a complex function of weight, age, gender, lifestyle and behavioural phenotype, but can be simplified and expressed as a multiple of BMR;
29
this is defined as the physical activity level, PAL (sometimes called the physical activity index, PAI). Thus PAL = TEE/BMR and is an index of 24h TEE adjusted for BMR, and is theoretically independent of those factors influencing BMR (weight, age and gender), at least as a first approximation. In the present context, PAL is important because it can be used to predict TEE, and hence energy reference values, as PAL x BMR (see paragraph 79).
PAL values are best estimated from direct measures of BMR combined with 24. measures of TEE over periods of 24 hours or longer. Such measurements in free- living populations have indicated that PAL can range from <1.3 in immobile subjects to between 3-4.7 for limited periods of time in, e.g. soldiers on field exercises, elite endurance athletes, or Antarctic explorers (Black, 1996; Hoyt & Friedl, 2006). Within the general population, however, the overall range of PAL values for individuals in energy balance, leading sustainable lifestyles, is between 1.38 for the most sedentary to 2.5 for the most active (Prentice, 1995; Schutz, 2000). A change in lifestyle will change an individual’s PAL value by reasonably predictable amounts. For example, an additional 150 minutes of moderate intensity activity a week, as currently recommended for adults in the UK (CMOs, 2011), will raise PAL by about 0.15 units (see Appendix 5 for details).
In previous reports, the duration and energy cost of individual activities (as PAR 25. values) has been summed to provide factorial estimates of PAL (DH, 1991; FAO, 2004; IoM, 2005; WHO, 1985). A comprehensive summary of all published PAR values for a wide range of different physical activities for adults was compiled by Vaz et al (2005).
The factorial estimation of PAL assumes that overall energy expenditure within 26. the general population can be predicted using information derived from activity diaries or lifestyle questionnaires. However, there is little evidence that this can be done with sufficient accuracy for several reasons.
Firstly, for most lists of activities used to identify lifestyle categories, individual 27. activities are only defined in general qualitative terms. Consequently, there can be large inter-individual variations in the energy expended for each activity. In part this is due to variation in activity intensity resulting from variation in fitness (Martins et al., 2007). Also, such listed energy costs may or may not include any thermic effect of food (TEF) or excess post-exercise oxygen consumption (EPOC).
Secondly, factorial estimates generally assume that time not allocated to activities 28. is occupied by basal energy expenditure and take no account of variation in those body movements associated with activities of daily living, change of posture and ‘fidgeting’ (i.e. SPA (Ravussin et al., 1986): see Appendix 3). SPA accounts for between-individual variation in energy expenditure of +15%. The potential for variable SPA throughout the range of defined activities adds to the uncertainty that their energy costs can be predicted.
30
The extent of error in factorial estimates of PAL does not appear to have been 29. systematically evaluated. However, an examination of the doubly labelled water (DLW)-derived TEE literature in Appendix 5 (paragraphs 283-286) identifies considerable evidence that classification of individual lifestyles is a poor predictor of PAL.
In this SACN report, factorial predictions of PAL values for specific lifestyle 30. categories have not been attempted due to the likely error in estimating actual TEE from listed PAR values of activities and the phenomenon of behavioural phenotypes, which exhibit very different rates of energy expenditure. Instead, PAL values for children, adolescents and adults have been identified from an analysis of the available DLW-derived TEE literature judged to be appropriate for the UK population (see paragraphs 104-107 and 113-118). For detail on the DLW method, see paragraph 70 and Appendix 6.
Thermic effect of food (TEF) The TEF or heat increment of feeding reflects the metabolic costs of eating, 31. digestion, absorption and metabolism of food and nutrients. Although TEF varies with food composition it is usually assumed to equate to energy expenditure equivalent to 10% of energy intake (Kleiber, 1975).
Other components of energy expenditure
Growth Physical growth involves an increase in both the size and complexity of body 32. structure, occurring under genetic and endocrine regulation in the presence of adequate nutrient supply. During growth, organs and tissues do not grow at a uniform rate, e.g. in the full-term newborn the brain represents about 12% of body weight, but in the adult is about 2% (Stratz, 1904). Energy is deposited into new tissues, the major part of growth costs, and some is expended during the synthesis of growing tissues. The energy required for growth is highest in the first three months of life when it accounts for about 35% of energy requirements, by 12 months of age this falls to about 3% (Butte et al., 2000a). The energy cost of growth remains low from one year of age to mid-adolescence, then increases slightly during the adolescent growth spurt; by the late teens the amount becomes very small (Tanner, 1990).
Changes in relative organ size influence both energy and protein metabolism. 33. Protein synthesis per unit of body mass proceeds at a high rate in the neonate and declines throughout infancy. High protein turnover contributes to the relatively high energy requirement of the newborn. Gender differences in body composition become more apparent after puberty. The developmental aspects of body composition and whole body metabolism affect the energy cost of growth (Butte et al., 1989; Butte et al., 2000b).
31
Pregnancy and lactation During pregnancy, energy is needed for placental and fetal growth and for the 34. growth of maternal tissues, e.g. uterus, breast and adipose tissue. In addition to these growth costs, there are increased energy costs associated with maintaining the larger tissue mass, along with an increased energy cost of movement, particularly for weight bearing activities after 25 weeks gestation (Butte et al., 2004). However, some of these costs may be offset by adaptive changes in activity and in maternal metabolism.
During lactation, energy is lost as secreted milk and expended in producing the 35. milk. Fat reserves that accumulate during pregnancy provide a variable proportion of this requirement (Butte & King, 2005).
The energy content of any tissue laid down during growth, pregnancy and 36. lactation or of milk produced is not accounted for in TEE measured at this time; these additional costs are estimated from analysis of tissue deposition and milk secretion. The TEE, however, does include the energy required for the metabolic costs of new tissue synthesis or milk production.
Factors affecting energy expenditure A detailed discussion of factors affecting energy expenditure can be found 37. in Appendix 3, a summary is given here. Physical activity is considered in Appendix 7.
Body size and composition Basal and total energy expenditure are related to body size with larger people 38. having more tissue mass and a higher BMR. Both height and weight are determinants of the BMR although the independent influence of height is much less than that of weight. In infants, children and adolescents, there is an increase in BMR with age, due to growth and increasing tissue mass, although the changing relative organ sizes in early life complicates the relationship between body weight and BMR.
After adjustment for body size, inter-individual variation in BMR is mainly determined 39. by variation in the most metabolically active tissue mass of the body, termed fat- free mass (FFM; the non-fat component of body composition comprising muscle, bone, skin and organs). The fat component of the body is termed fat mass (FM) and varies considerably between individuals in terms of the absolute amount. The differences observed in BMR with gender, age and ethnicity, after adjustment for body size, are mainly accounted for by differences in body composition, i.e. FFM (see Appendix 3 for further discussion).
The increasing body size of overweight and obese individuals increases both FM 40. and FFM and the absolute BMR (Das et al., 2004; Prentice et al., 1996a) although the relationship is not linear (Prentice et al., 1996a). Despite this, prediction equations developed for BMR as a function of weight, age, gender and height, provide a reasonably accurate estimate of BMR over quite a wide range of BMI; this is indicated by the similar relationships between BMI and measured and estimated
32
BMR in the adult DLW cohorts used to derive PAL values in this report (see Appendix 8). However, in this report, because prescriptive EAR values are defined (see paragraph 59), BMR values are only calculated for healthy body weights at the height of the population group, and potential errors in estimating BMR values for body weights markedly outside the healthy body weight range will not influence EAR values.
In individuals with higher percentages of body fat, a decrease in the mechanical 41. efficiency of movement can increase the energy expenditure associated with certain types of activity. On a population basis, for adults up to a moderate level of fatness (i.e. overweight, but not obese), such influences of fatness on activity- specific energy expenditure are generally ignored (Durnin, 1996). With obesity these changes may become important because, as discussed in Appendix 3 (paragraphs 233 and 235), PAEE appears to decline only slightly with increasing BMI, and PAL does not vary with BMI; this suggests that any reduction in activity with increasing obesity is offset by the increasing energy cost of such activities with increasing BMI.
Genetic variation While many studies have investigated the role of genetic variation on energy 42. expenditure there have, as yet, been no clearly established relationships between specific gene variants and energy expenditure. For example, a number of studies have investigated the association between the mitochondrial uncoupling protein gene variants and energy expenditure, but results have been equivocal.
Hormones Several hormones, e.g. sex hormones, thyroid hormone, adrenaline and leptin, 43. may affect energy expenditure and have been implicated in the regulation of energy balance. Pharmacological agents, such as glucocorticoids, amphetamines and some anti-obesity drugs have all been shown to increase energy expenditure, while opiates and barbiturates can decrease it. Smoking acutely increases resting energy expenditure to a small extent (e.g. a 3.3% increase in resting metabolic rate (RMR) over a three hour measurement period (Collins et al., 1994)).
Illness The effect of illness on energy expenditure is discussed in more detail in 44. Appendix 9. In patients with a range of conditions (sepsis, degenerative diseases, malignancy, trauma, congenital conditions and others), TEE is usually normal or reduced, partly because of a reduction in body weight and FFM as a result of disease-related malnutrition or neurological causes of wasting, and partly because of reduced physical activity. The latter compensates for any increase in BMR, which is common in acute and many chronic diseases. There are some exceptions, such as subgroups of patients with cystic fibrosis, anorexia nervosa, and congenital heart disease, where TEE has been reported to be increased (Elia, 2005). Specialist clinical advice is essential when considering energy requirements for people with disease.
33
Ambient temperature Body temperature is tightly regulated in order to maintain cell function. A 45. component of this regulation relies on variation in energy expenditure. Differences in environmental temperature affect energy expenditure and have been shown to account for about 2-5 percent of the variation in TEE. Indoor temperatures, however, are typically controlled to remain relatively constant and individuals adjust their clothing to create a relatively constant thermal microenvironment, so in reality, ambient temperature has a minimal effect on energy expenditure.
Energy balance and storage Energy balance is achieved when ME intake is equal to TEE, plus the energy cost of 46. growth in childhood and pregnancy or the energy cost of milk production during lactation. A positive energy imbalance occurs when energy intakes are in excess of these requirements, while negative energy imbalance occurs when energy requirements are not met by intake.
Energy intake should exceed TEE during growth, pregnancy and lactation when 47. new tissues are being laid down or milk produced. In other circumstances, energy intake which exceeds expenditure is stored. Triglycerides within adipose tissue act as the body’s major energy store. Some energy is also stored in the liver and skeletal muscle as glycogen. The amount of energy stored in the adipose tissue of a healthy adult of normal weight is equivalent to approximately one month’s energy requirements (Schutz & Garrow, 2000).
Short-term, day to day energy imbalances are accommodated by the deposition 48. and mobilisation of glycogen and fat. Positive and negative energy imbalances occur in the short term in free-living individuals, so, in terms of weight regulation, it is important to consider the overall energy balance over a prolonged period of time.
Chronic negative energy imbalance results in the utilisation of stored energy from 49. triglyceride in adipose tissue and protein in muscle and viscera, since glycogen stores are limited and are rapidly exhausted. Chronic positive energy imbalances are mostly accommodated by the deposition of adipose tissue triglycerides, together with a small but fixed ratio of lean tissue (Schutz & Garrow, 2000). Thus, an individual in chronic positive energy imbalance stores excess food energy mainly as triglyceride and to a lesser extent protein. Muscle and liver glycogen stores are modest and have a limited capacity; whereas the capacity of the body to store triglycerides in adipose tissue is substantial.
Obesity Obesity results from a long-term positive energy imbalance. The increasing 50. prevalence of obesity must reflect temporal lifestyle changes, since genetic susceptibility remains stable over many generations, although inter-individual differences in susceptibility to obesity may have genetic determinants (Maes et al., 1997). Definitions of obesity and current prevalence in the UK population are discussed in Appendix 10.
34
Obesity increases the risk of a number of diseases (see Table 2 for an overview). 51. In 2009, a collaborative analysis of the influence of BMI on all-cause mortality in 57 prospective studies (900 000 adults), identified a U shaped relationship with minimum risk associated with a BMI of about 22.5-25kg/m2 in men and women (Prospective Studies Collaboration et al., 2009). This relationship and how it has informed the approach used to derive energy reference values in this report will be discussed in more detail in Section 3.
Table 2 Summary of associations observed in prospective studies between obesity and subsequent ill healtha
Association for increased risk Disease outcome
Relative risk >3 Type 2 diabetes
Insulin resistance Hypertension Dyslipidaemia
Breathlessness
Sleep apnoea
Gall bladder disease
Relative risk about 2-3 Coronary heart disease or heart failure Osteoarthritis (knees and hip)
Hyperuricaemia and gout
Complications of pregnancy, e.g. pre-eclampsia
Relative risk about 1-2 Cancer, e.g. oesophagus (adenocarcinoma), colorectum, breast (postmenopausal), endometrium and kidney
Impaired fertility/polycystic ovary syndrome
Low back pain
Increased risk during anaesthesia
Fetal defects associated with maternal obesity
a based on Correa et al., 2008; Haslam et al., 2006; Key et al., 2004; Renehan et al., 2008
The influence of physical activity and diet on the regulation of body weight The influence of physical activity and energy intake in relation to body weight 52. regulation is discussed in detail in Appendix 7 and 11; a summary of the main considerations is given here.
Body weight is gained when energy intake exceeds TEE over time. PAEE is the most 53. variable component of TEE and is amenable to modification, so changes in PAEE may affect risk of weight gain.
Methodological constraints are a severe limitation in defining the role of physical 54. activity in the regulation of body weight, for example, studies and surveys rely mostly on subjective measures of reported physical activity. Proxy measures of population-level physical activity trends suggest that changes in activity in domestic life, work and travel have coincided with the increase in prevalence of obesity. A review of the available evidence from prospective cohort studies
35
suggested that, on balance, increased physical activity and decreased sedentary behaviour were associated with lower relative weight and fatness gains; however, the results were mixed and the identified associations were generally of a small magnitude (Wareham et al., 2005). Evidence from studies using objective measures of physical activity, especially DLW-derived measures of PAEE, and trials of the primary prevention of weight gain, is inconsistent. Available data are insufficient to accurately define a level of energy expenditure or the required frequency, intensity and duration of physical activities required to reduce the risk of unhealthy weight gain.
Methodological constraints are also a severe limitation in defining the role of diet 55. and dietary composition in the regulation of body weight, for example, limitations in the accurate assessment of dietary exposures and under-reporting of dietary intake (Bandini et al., 1990a; Bingham et al., 2001, 2003; Bingham & Day, 2006; Buhl et al., 1995; Champagne et al., 1998; Goris et al., 2000; Lichtman et al., 1992; Livingstone & Black, 2003; Rennie et al., 2005). A prolonged excess energy intake is, however, fundamental to weight gain and the development of obesity. Household purchase data from the Family Food module of the Living Costs and Food Survey13 suggest that there has been a decrease in average daily energy intakes, which seems to date from the 1960s, and coincides with the time period in which there has been a large increase in the prevalence of overweight and obesity. Since under-reporting of food intake is particularly pronounced in the overweight and obese (Rennie et al., 2007), the proportion of the population likely to under-report increases as the population gains weight, thus exacerbating the problem of under-reporting of energy intakes (Bandini et al., 1990a; Rennie et al., 2005). The use of household purchases as a proxy for consumption in these surveys also limits the accuracy with which energy intakes can be estimated, and in addition, reported energy values do not take into account food waste. The NDNS, in which food consumption by individuals has been measured with seven day weighed diaries, a considerably more robust methodology than that used in Family Food, also shows average energy intakes in men, but not women, have decreased between 1986/7 and 2000/1. Under-reporting is also an issue in NDNS (estimated to be approximately 25% of energy needs in both sexes (Rennie et al., 2005)) and there is some indirect evidence of a secular trend towards increasing under-reporting over time (Rennie et al., 2005).
Energy flux or turnover is the rate at which energy in all its forms flows through the 56. body on a daily basis (chemical, work, thermal). For an individual in energy balance, the energy flux is numerically the same as energy expenditure or energy intake. Provided energy intake matches energy expenditure then energy balance will be maintained whether the energy flux itself is at a higher or lower level. However, as the flux of energy might have important physiological relevance in its own right, the level of energy flux at which balance is achieved should be considered. It has
13 The Family Food module of the UK Living Costs and Food Survey was formerly known as the Expenditure and Food Survey (EFS). The EFS was formed in 2001 from the merger of the National Food Survey (NFS) and the Family Expenditure Survey. Under-reporting in Family Food is thought to be lower than in the NFS as it does not only focus on food purchases, but on household expenditure across the board and is largely based on till receipts. Family Food also provides better estimates of food eaten outside the home than did the NFS. The change in methodology when the EFS replaced the NFS, however, makes the estimate of year on year change unreliable between 2000 and 2001-02.
36
long been hypothesised that the mechanisms controlling energy balance may be more sensitive and effective in individuals with higher levels of physical activity and hence energy flux, while sedentary individuals may be below the threshold of physical activity at which these mechanisms become imprecise, leading to obesity (Mayer & Thomas, 1967). There is some indirect evidence for this hypothesis that the coupling between energy expenditure and energy intake may be less precise at low levels of physical activity (Prentice & Jebb, 2004; Schoeller, 1998).
Definition of energy requirement The energy requirement, and associated descriptive terminology, must be defined 57. with particular care in the context of a population which includes many individuals who are overweight or exhibit habitually low levels of physical activity. COMA (DH, 1991) defined requirements in general terms only i.e. intakes of nutrients which were likely to meet the needs of some or all within population groups. There was no specific definition of the energy requirement in relation to specific healthy body weight ranges or prescribed levels of energy expenditure. COMA reported EAR values calculated for a wide range of physical activity levels and adult body weights, which in practice, would maintain the status quo in terms of existing body weights of such population groups. The COMA report did include a widely quoted summary table14 for sedentary younger and older adult men and women at specific body weights which were within the healthy range, albeit within the upper part of this range15.
The definition offered by FAO/WHO/UNU (WHO, 1985; FAO, 2004), in the context 58. of a diet which contains adequate amounts of all essential nutrients, was ‘the amount of food energy needed to balance energy expenditure in order to maintain body size, body composition and a level of necessary and desirable physical activity consistent with long-term good health. This includes the energy needed for the optimal growth and development of children, for the deposition of tissues during pregnancy, and for the production of milk during lactation consistent with the good health of mother and child’ (WHO, 1985; FAO, 2004). Thus this definition includes both healthy body weights and desirable physical activity levels, with the term “normative” used in the 1985 report and “prescriptive” in the 2004 report. Each report calculated energy requirements based on suitable reference body weight ranges, with the 1985 report noting that application of such requirements to groups with body weights below or above such ranges would tend to mediate weight change towards the median. Neither report, however, calculated values in relation to a “desirable” physical activity level.
In this SACN report, the requirements for energy for all population groups have 59. been set at the level of energy intake required to maintain a healthy body weight16
14 Table 2.8 Estimated Average Requirements (EARs) for energy for groups of men and women with physical activity level of 1.4 (DH, 1991).
15 These were median body weight values from a 1980 survey of adult heights and weights in Great Britain, and were equivalent to BMI values of 24.1 and 24. 9 kg/m2 for men and 23.1 and 24.7 kg/m2 for women.
16 Although the healthy body weight range is usually defined as a BMI between 18.5-24.9kg/m2, for the purposes of calculating the revised EAR values, in this report healthy body weights have been identified as those associated with a body mass index of 22.5 kg/m2 which represents the lower end of the minimum mortality range (Prospective Studies Collaboration et al., 2009) (see paragraph 51).
37
in otherwise healthy people at existing levels of physical activity and to allow for any special additional needs (growth, pregnancy and lactation). Thus, the prescriptive17 terminology and principle (i.e. relating to an desirable standard as distinct from status quo) have been adopted, but only for body weight, with physical activity set at best estimates of existing levels. However, as with previous reports, the importance of adequate physical activity is recognised and this report includes advice on desirable physical activity consistent with long-term health (Chief Medical Officers [CMOs], 2011). The likely impact of such advice on energy requirement values is also described.
Energy requirements described here are identified from sufficient measurements 60. of energy expenditure on healthy, well-nourished individuals to demonstrate how energy expenditure varies according to age, gender, body size and composition, pregnancy, lactation and physical activity. These characteristics are used to describe population groups for whom EARs are defined, with values calculated for best estimates of healthy body weights at the heights of each population group and at respective current levels of physical activity.
Variability Measurements of energy needs that are used to predict DRVs for a particular 61. population group will, for a number of reasons, exhibit variability. Consequently, there will be a distribution of energy reference values for each population group and the level set for the DRV should take this into account.
For most nutrients the DRV is identified as the Reference Nutrient Intake (RNI) 62. from the upper bound of the reference ranges i.e. two notional standard deviations above the average reference value (DH, 1991). However, for dietary energy the DRV is defined differently i.e. it is equal to the average reference value (EAR). The RNI for dietary energy is not used because it represents an excess energy intake for the majority of the population. Energy intakes that consistently exceed requirements lead to weight gain and obesity in the long term. An intake equal to the average reference value for a population group on the other hand is, in theory, associated with similar probabilities of excessive and insufficient energy intakes for any individual within the population.
Appetite regulation allows humans in good health to match broadly their energy 63. intake to their requirement so the chances of energy deficiency are low unless food supply is limited. However, this mechanism is not always sufficiently sensitive to prevent small excesses leading to inappropriate weight gain in the long term.
In countries like the UK, the population is characterised by sedentary lifestyles 64. at all ages. As a result, rates of energy expenditure and intakes that maintain energy balance at these levels of energy expenditure are unlikely to be normally distributed. Measurements of energy expenditure within adult populations18 indicate that the overall distribution is skewed towards sedentary behaviour. For
17 Sometimes used interchangeably with ‘normative’.
18 See Appendix 6 for a description of the methodology for measuring energy expenditure and the uncertainties that exist in its application and interpretation.
38
the high proportion of individuals who are relatively inactive, energy expenditure and intakes that maintain energy balance will cluster at the lower end of the range. A small proportion will be more active, with a few individuals exhibiting the highest rates of expenditure that are sustainable. A schematic diagram of the probable distribution within the adult population is shown in Figure 1. Statistical considerations, therefore, dictate that the appropriate descriptor of the midpoint of the distribution is the median, which is likely to be somewhat lower than the mean, thus the median is used in this report.
Figure 1 – Schematic of the likely distribution of energy expenditure (expressed as a multiple of Basal Metabolic Rate – BMR) and hence reference intakes within the adult population
Evidence that energy expenditure varies between individuals with similar lifestyles 65. over a wider range than would be expected (at least in adults) presents a second difficulty in estimating energy reference values. This is particularly important since in the past, attempts have been made to define energy reference values in terms of specific physical activity levels (as multiples of basal metabolic rate, see paragraphs 25 and 26) for specific lifestyle population groups. Consequently, with the probable exception of those at the extremes of the activity range, lifestyle predictions of energy expenditure and resultant energy reference values cannot be made with the certainty assumed in previous reports such as COMA (DH, 1991) (see Appendix 5 for a detailed consideration).
Finally, given the high prevalence of overweight and obesity at all ages in the 66. current population (see Appendix 10), prescriptive energy reference values which are consistent with long-term good health need to be based upon suitable reference body weights, as distinct from existing body weights. If this principle is not adopted, application of such energy reference values would maintain body
29
Deleted: <sp>
64. In countries like the UK, the population is characterised by sedentary lifestyles at all ages. As a result, rates of energy expenditure and intakes that maintain energy balance at these levels of energy expenditure are unlikely to be normally distributed. Measurements of energy expenditure within adult populations15 indicate that the overall distribution is skewed towards sedentary behaviour. For the high proportion of individuals who are relatively inactive, energy expenditure and intakes that maintain energy balance will cluster at the lower end of the range. A small proportion will be more active, with a few individuals exhibiting the highest rates of expenditure that are sustainable. A schematic diagram of the probable distribution within the adult population is shown in Figure 1. Statistical considerations, therefore, dictate that the appropriate descriptor of the midpoint of the distribution is the median, which is likely to be somewhat lower than the mean, and the median is used in this report.
Figure 1. Schematic of the likely distribution of energy expenditure (expressed as a multiple of Basal Metabolic Rate - BMR) and hence reference intakes within the adult population
65. Evidence that energy expenditure varies between individuals with similar lifestyles over a
wider range than would be expected (at least in adults) presents a second difficulty in estimating energy reference values. This is particularly important since in the past, attempts have been made to define energy reference values in terms of specific physical activity levels (as multiples of basal metabolic rate, see paragraph 25 and 26) for specific lifestyle population groups. Consequently, with the probable exception of those at the extremes of the activity range, lifestyle predictions of energy expenditure and resultant energy reference values cannot be made with the certainty assumed in previous reports such as COMA (DH, 1991) (see Appendix 5 for a detailed consideration).
15 See Appendix 2 for a description of the methodology for measuring energy expenditure and the uncertainties that exist in its application and interpretation.
0
sedentary extremely active
Median Mean
Frequency within
population
39
weights, which are in excess of the healthy reference weights for many of the population. Thus, in contrast to COMA’s report (DH, 1991), energy reference values defined in this report will be derived for infants, children and adults in relation to body weights which, on the basis of current evidence, are likely to be associated with long-term good health. This means that for people who are underweight, overweight or obese, energy intakes at the reference levels should enable the transition towards the healthy body weight range (BMI 18.5-24.9 kg/m2) over time.
Approaches used to estimate energy reference values Energy reference values are defined as the Estimated Average Requirement (EAR) 67. for food energy for specific population groups, not for individuals, and are set using two basic approaches:
1) Measurement of energy intake (EI) of healthy reference populations in energy balance, growing appropriately or achieving successful pregnancy and lactation, provides, in theory, a direct estimate of the energy requirement. However, it has not proved possible to make such measurements with sufficient accuracy for them to be useful, especially for groups where under-reporting or failure to report ‘usual’ food and drink intake may be an issue (e.g. overweight adults) (Black et al., 1993; Livingstone, 1995; Prentice et al., 1986). In addition, EI is not a physiological measure of the energy requirement since it is not adequately regulated through the appetite mechanism to match TEE precisely and allow exact energy balance, and EI can be consciously altered by the subject. Also, there is no independent check on whether the measured EI matches TEE and is therefore appropriate for the subject’s need. Thus, in practice, this approach has been largely abandoned in favour of approaches measuring TEE.
2) Measurement or prediction of TEE provides a physiological measure of the energy requirement at energy balance. This is because for energy balance, EI must equal TEE. Thus, the energy requirement can be predicted specifically as the rate of TEE plus any additional energy needs for growth, pregnancy and lactation.
Measurement of total energy expenditure (TEE) Several approaches are used to measure TEE. Short-term measurements under 68. highly defined conditions can be made by calorimetry. Direct calorimetry which measures the rate of heat loss from the subject to the calorimeter is the most accurate method. Indirect calorimetry, the most commonly used approach, measures oxygen consumption and/or carbon dioxide production from which TEE is calculated using standard formulae such as the Weir equation (Mansell & Macdonald, 1990).
The components of TEE (e.g. BMR, PAEE) can be measured separately using direct 69. and indirect calorimetry. TEE can also be measured with large walk-in calorimeters, although the necessary confinement of subjects limits this to short periods, most often 24 hours. The information provided by these approaches is accurate and has provided the energy costs of different physical activities (see paragraph 22 and Appendix 5) and minimal daily rates of TEE. Free-living activities cannot usually
40
be measured by these techniques. Some non-calorimetric techniques can be used to predict free-living TEE by extrapolation from physiological measures, the best example being heart rate monitoring (Levine, 2005). These methods need first to be calibrated against direct or indirect calorimetry before TEE can be calculated.
The DLW method (International Atomic Energy Agency [IAEA], 2009) is generally 70. recognised as the most accurate measure of free-living TEE currently available and is discussed in detail in Appendix 6. DLW measures the rate of carbon dioxide production, and hence TEE, in free-living subjects over a period of several days to several weeks providing more accurate measures of TEE than other non- calorimetric methods, e.g. heart rate monitoring (Levine, 2005).
Whilst the DLW method is the best available, a series of assumptions are made 71. which can affect the predicted TEE values. There are also concerns regarding recruitment bias, since people who are willing to participate in DLW studies may not be representative of the population as a whole. Nevertheless, the DLW method has proved to be the most useful approach. It remains the method of choice and is more accurate than others available. TEE measurements obtained by this method are judged to be representative of the current TEE of the UK population and form the basis of estimates of energy reference values in this report.
Utilising TEE measurements to derive energy requirements The 1985 FAO/WHO/UNU report (WHO, 1985) was the first to use a factorial 72. approach to estimate energy requirements based on the prediction of TEE as BMR x PAL (see paragraphs 23-30). This approach was adopted by COMA in 1991 (DH, 1991) and by the most recent FAO/WHO/UNU report (FAO, 2004) for calculating energy requirements of adults, although not used in the US DRI report (IoM, 2005) (see paragraph 86). The calculations made in these reports estimated BMR from anthropometric measures (as described in paragraphs 80 and 81) and were based on the assumption that values for PAL can be estimated from time-allocated lists of daily activities expressed as PAR values (see paragraph 22). With information about PAL values for specific lifestyles and types of daily activities, appropriate PAL values can then be assigned to particular population groups.
Reservations have been expressed regarding the prediction of PAL through the 73. summation of PAR values (see paragraphs 26-28 and Appendix 5). Determination of energy reference values using measured values of TEE is therefore preferable. There are two potential analytical approaches that can be employed to derive energy reference values from measured TEE:
predicting TEE using regression equations (see paragraphs 74-78); •
employing a factorial approach based on the assumption that TEE is equal to • BMR x PAL (see paragraph 79).
Predicting TEE with regression equations The first approach is to use measured TEE directly to derive regression equations 74. which describe how TEE varies as a function of anthropometric variables (such as weight and height) for defined population groups. The regression equation can
41
then be used to predict TEE and resulting energy reference values for any group on the basis of their anthropometric variables (see Appendix 5 for details).
The analysis of TEE as a function of its potential predictor variables (weight, age and 75. gender) by multiple regression techniques would seem a logical progression from the accumulation of reliably estimated measures of free-living TEE by the DLW method (Goran, 2005). In practice, however, a major limitation to this approach has been the inability of TEE prediction models to account for variation in PAEE, an important source of variation in TEE, in a transparent way.
Variation in PAEE, at least in terms of its duration and intensity, is not physiologically 76. related to anthropometric variables in the same way as the BMR, and any such variation will be lost in a regression model involving just anthropometric variables. The regression equation will, in fact, predict values for TEE which contain a PAEE component comparable to the mean value observed at any given weight or age within the data set used to generate the regression.
Between-individual variability in PAEE in a regression of TEE on anthropometric 77. variables is to some extent indicated by the residuals of the regression. Such residuals (the difference between observed TEE values and those predicted by the regression) will include all TEE not accounted for by the regression, of which variation in PAEE will be the major part. However, any variation in the basal energy expenditure not accounted for by body weight, age and gender, will also be included. Thus, use of residuals in terms of predictors of the likely range of variation in PAEE within population groups is limited and residuals have not been used as such in any previous report (see Appendix 5 for more detail).
Instead, previous reports which have used regression models of TEE to predict 78. energy requirements (FAO/WHO/UNU (FAO, 2004)19 and the US DRI values for energy (IoM, 2005), made special provision for the likely variation in PAEE within the population group described by the regression. The methodology employed, is examined below (paragraphs 85 and 86).
Predicting TEE from BMR and PAL values extracted from measured TEE – the factorial approach The second analytical approach to the derivation of energy requirements from 79. measured TEE employs the factorial approach20 in the following steps:
TEE values measured in a reference population are divided by measured or • estimated BMR to extract PAL values. This means that the reference populations studied by DLW are described primarily by PAL values.
For the population of interest, BMR values are then estimated from BMR • prediction equations using relevant anthropometric data for the population.
19 Regression equations were used to derive energy requirements for infants, children and adolescents. The energy requirements of adults were calculated from factorial estimates of habitual TEE.
20 This approach has also been referred to as the BMR multiple approach and the PAL model.
42
The PAL values derived from the reference population can then be used to • estimate TEE and EAR values for the population of interest based on the latter’s estimated BMR values.
Whilst some criticism of the PAL x BMR approach has been expressed (Goran, 2005) (see Appendix 5), no satisfactory alternative has yet been identified.
Estimating BMR A variety of BMR prediction equations80. are described in the literature and different authors favour different equations in published work especially in the US and Europe. Within the UK COMA report (DH, 1991) and for FAO/WHO/UNU reports (FAO, 2004; WHO, 1985), the Schofield equations have primarily been used (Schofield et al 1985). The Schofield equations are a series of predictive equations for BMR based on height, body weight, age and gender, although simplified equations based on just body weight, age and gender are used in these reports. The equations are derived from an analysis of a compilation of calorimetric measures of BMR values and anthropometric data. These predictive equations, slightly modified, form the basis for FAO/WHO/UNU energy requirements for adults (FAO, 2004) and modified versions were used by COMA for the previous UK EAR for energy in children aged 3-18 years and adults (DH, 1991).
The data set upon which the Schofield prediction equations for BMR were based 81. was compiled mainly from results in West European and North American subjects, with almost half of the subjects being Italian in whom BMR was measured using a closed circuit method21 in the 1930s and 1940s (Henry, 2005). The use of the closed circuit method has been queried as it may overestimate oxygen consumption and consequently energy expenditure (FAO, 2004). Furthermore, the applicability of the Schofield equations to all population groups has been questioned and an alternative more comprehensive database has been assembled and analysed from which a new set of equations has subsequently been derived, i.e. the Henry equations (Henry, 2005). Although estimates of BMR made by the Schofield or Henry prediction equations differ only slightly in children or adults (i.e. the Henry equations are 3-4% lower), an assessment of the validity of different predictive equations for BMR in adults found the Henry prediction equations to be the more accurate (Weijs, 2008). In this report the BMR prediction equations published by Henry (2005) are therefore used to derive EAR values which, because they are prescriptive values, require the identification of reference values for healthy body weights for which BMR is calculated. A more detailed discussion on the BMR equations is provided in Appendix 4, with suitable values for reference body weights described below in paragraphs 90-93.
21 The methods available to measure BMR may be divided into two types; closed and open circuit methods. In the closed circuit indirect calorimetry method, the subject breathes in and out of a closed system, commonly a spirometer, which is sealed to room air. Oxygen, or a mixture of oxygen and nitrogen, is supplied to the spirometer at the rate at which it is consumed. Thus the rate at which oxygen is delivered is the same as oxygen consumption. Heat production may be estimated from oxygen consumption alone, in which case the carbon dioxide produced by the subject is absorbed, for example by soda lime, within the closed breathing circuit. Alternatively, weight gain of the carbon dioxide absorber may be used to derive carbon dioxide production using this method. (From: Green, 1994).
43
Calculation of energy requirements There are two major considerations in the practical application of the factorial 82. approach to calculate energy requirements using PAL and BMR values from appropriate DLW data sets: 1) identifying suitable PAL values appropriate for specific groups and populations; and 2) utilising this information to derive energy reference values.
PAL values can be calculated from DLW-derived measurements of TEE largely 83. obtained from studies conducted in UK populations and comparable populations in the US and other developed countries over the last 20 years, and from measurements or predictions of the BMR. With few exceptions, published DLW studies involve relatively small numbers of healthy subjects who may or may not have been engaged in activities representative of the general population at that time. Although data from these studies can be combined to derive best estimates of PAL values for the whole population based on age and gender, the degree to which these values are representative of the current UK population is uncertain. For example, the data set assembled for the US DRI report (IoM, 2005) (n=767) included many individual studies with subjects exhibiting high levels of physical activity (see Appendix 5). An alternative is to make use of large population studies of randomly selected subjects which can be assumed to exhibit similar activity patterns to that of the current UK population (see paragraph 116 and Appendix 8).
Suitable PAL values can be used to update and improve tables of PAL value ranges 84. for the various ages and lifestyle groups identified in previous reports, allowing energy reference values to be calculated for such groups (see paragraphs 25-30). However, the marked between-individual variation in PAL (which seems to occur independently of any predictable lifestyle) makes such a selection of an appropriate PAL value unreliable. The alternative approach is to evaluate the distribution of PAL values, with medians and centile ranges identified for the population as a whole. This approach allows energy reference values to be framed against such distributions, in effect substituting PAL distributions for PAL values defined in terms of lifestyle. Thus, energy reference values can be defined as a population EAR22 (i.e. from the median PAL value) with additional values appropriate to those who are less or more active than the average (i.e. from the 25th and 75th centiles). Only three activity groups within the population are identified by PAL values using this method, but it is considered unrealistic to judge PAL values more finely. However, additional information on the change in PAL with specified additional activity will allow calculation of the probable additional energy intakes required to support such activities.
22 The word average in the EAR (Estimated Average Requirement) term is used here as a general term embracing both median and mean.
44
Approaches employed to calculate energy requirements in other reports
FAO/UNU/WHO (FAO 2004) The development of regression equations for TEE from measured values forms 85. the basis of the FAO/WHO/UNU (FAO, 2004) report on energy requirements for infants, children and adolescents (but not for adults). Data sets were developed using TEE values from DLW studies for infants, and DLW and heart rate monitoring studies for children and adolescents. Mean study values weighted by the number of subjects were then used to derive the regression equations. For children and adolescents, the predicted TEE values were not used directly but used to predict PAL values. This prediction was done by dividing TEE values from the regression with BMR values predicted as a function of age and weight. The calculated PAL values were assumed to represent activity levels for populations with “average” or “moderate” physical activity. Values for more (“vigorous”) or less (“light”), active lifestyles were calculated as average ± 15%. This approach allowed calculation of energy requirements for the three lifestyles as BMR x PAL. For adults, FAO/WHO/ UNU (FAO, 2004) adopted a factorial model (BMR x PAL) in place of regression equations with PAL values identified for lifestyle categories.
US DRI Institute of Medicine (2005) The US Dietary Reference Intake (DRI) values for energy (IoM, 2005) were 86. calculated from age-range and gender-specific prediction equations for TEE derived by regression analysis of a large (n=767) DLW data set of individual TEE values. These were obtained, with ancillary data, directly from the investigators of each study. However, unlike the FAO/WHO/UNU report (FAO, 2004), variation in physical activity was accommodated within the regression. This involved an activity constant, which could be one of four values, that the user must choose. Each of these four activity constants was identified as equivalent to a range of PAL values appropriate for sedentary, low active, active, and very active lifestyles. Thus, the user of the regression equations is instructed to estimate which activity/ lifestyle category the subject or population group belongs to, and assign the appropriate activity term within the regression, together with age, weight, and height, to calculate the EAR value. Separate regression equations were defined for each gender and age range. In practice these equations differ little from prediction equations of the form BMR x PAL with four values of PAL identified for the four activity ranges.
45
3 Estimated average requirements – approaches used and values derived
Summary of approach used to determine EARs The SACN Framework for the Evaluation of Evidence (SACN, 2002) was used as the 87. basis to identify and assess published evidence of total energy expenditure (TEE) from which to guide derivation of energy reference values. Only studies using the doubly-labelled water (DLW) method to measure TEE were considered, as this represents the most accurate method for assessing TEE in free-living populations. No large-scale population studies of any age group have been conducted to determine DLW-derived TEE in the UK population and although TEE values for subjects of all ages obtained from the UK NDNS series were examined, this data set alone (n=156 adults) was not large enough to serve as a sole reference for the determination of energy reference values. Consequently, other data sources (reference populations) were sought. Priority was given to studies in well- characterised populations (e.g. by age, gender, weight, height or BMI) and where BMR had been measured directly rather than estimated from prediction equations. Studies of extreme energy expenditure, such as those involving elite athletes, and studies based on populations which are not commonly represented in the UK (such as ethnic groups not widely seen in the UK) were given less emphasis. The majority of studies identified were cross-sectional studies in healthy human populations of infants, children and adults and provided mean TEE values for the study population. Two large data sets of individual TEE values for adults were subsequently obtained from the USA, the OPEN and Beltsville studies (as described in paragraph 116 and Appendix 8). Some prospective cohort studies were also found and any potential link between energy intake and risk of ill health has been mainly drawn from these.
For adults aged 19-65 years and children aged between 3-18 years, the different 88. approaches to setting EARs described above (see paragraphs 72-86) were examined. Regression modelling was explored and mean TEE values from the data set of DLW studies were used to develop regression equations of TEE against age, weight and gender for different age groups. However, as discussed above (see paragraphs 74-78) this approach is limited because of the inability of TEE prediction models to account for PAEE, an important source of variation in TEE, in a transparent way. As a result, the regression approach was not pursued and a factorial model was adopted, in which TEE is predicted from BMR x PAL (see paragraph 79).
Thus, new EAR values have been derived for children and adults. However, for 89. infants the reference values stated in the FAO/WHO/UNU report (FAO, 2004) were used due to a lack of substantial new data. Pregnant and lactating women were considered separately.
46
Prescriptive energy reference values for healthy body weights The high prevalence of overweight and obesity at all ages in the UK population (see 90. Appendix 10) raises a clear difficulty in the definition of energy reference values for population groups. Such values, calculated to match energy expenditure, will, for many of the population, maintain overweight and will therefore not be consistent with long-term good health. Given the objective of defining prescriptive reference values, such values need to be based on healthy body weights.
For infants and children, healthy body weights are difficult to define but the UK-91. WHO Growth Standards (Royal College of Paediatrics and Child Health [RCPCH], 2011) for infants and preschool children are considered to be appropriate as the pattern of growth represented is associated with favourable health outcomes. For school children, the UK 1990 reference values for child growth (Freeman et al., 1995) indicate body weights which are on average about 15% lower than recent (2009) UK values (NHS IC, 2010). These body weights can reasonably be assumed to be a better indication of healthy weights than current values.
For adults, the normal body weight range is generally defined as a BMI between 18.5 92. and 24.9 kg/m2 (WHO, 1998). As already indicated (paragraph 51), a collaborative analysis of the influence of BMI on all-cause mortality in approximately 900,000 adults (Prospective Studies Collaboration et al., 2009) identified a U-shaped relationship with minimum risk associated with a BMI of about 22.5-25kg/m2. Above this range, positive associations were recorded for several specific causes (vascular mortality, diabetic, renal, and hepatic mortality and neoplastic mortality) with each 5 kg/m2 higher BMI associated with about 30% higher overall mortality. Below 22.5-25 kg/m2, the overall inverse association with BMI was predominantly due to strong inverse associations for smoking-related respiratory disease (including cancer).
On this basis, energy reference values were calculated in this report at the 5093. th centile of the UK-WHO Growth Standards for children up to four years (RCPCH, 2011) and using the UK 1990 reference values for child growth for children aged over four years (Freeman et al., 1995). In adults, the healthy body weights that equate to a BMI at the lower end of the minimum mortality range (as discussed in paragraph 92) i.e. BMI 22.5 kg/m2, have been adopted. Thus, EAR values should be calculated for a body weight equivalent to a BMI of 22.5 kg/m2 at the height of the population group. Illustrative EAR values are shown calculated from current estimates of heights of the UK population23. Energy intakes matching the revised EAR values will be less than those which would maintain weight for overweight groups and can therefore form the basis of intakes designed to achieve healthier body weights. In contrast, energy intakes matching the revised EAR values will be greater than intakes that would maintain weight for those who are underweight.
23 It is important to note that a weight equivalent to a BMI of 22.5 kg/m2 does not represent a precise target body weight to which everyone should aspire, rather that a single figure was required for the purpose of calculating prescriptive EARs based on healthy body weights. A body weight equating to a BMI within the range of 18.5 – 24.9 kg/m2 is generally considered ‘normal’ or ‘healthy’.
47
Energy reference values for infants, children and adolescents
Energy cost of growth TEE measured using the DLW method includes the energy expended in tissue 94. synthesis. Thus, only the cost of energy deposited in growing tissues should be added when calculating the energy reference values for infants, children and adolescents. For infants, such costs are relatively high and change during the first year of life (see paragraphs 100 and 101 and Tables 3 and 4). For older children and adolescents, growth costs are much less and the energy deposited can be accounted for by a simple adjustment to the factorial prediction of TEE from BMR x PAL (see paragraph 108 and Table 7).
Infants aged 1 – 12 months of age Following the approach of the FAO/WHO/UNU report (FAO, 2004), the energy 95. reference values for infants (Table 5) were estimated from TEE (measured by the DLW method in healthy, well-nourished infants born at full term within the range of normal birth weight (Butte, 2005)), plus the energy deposited during growth, estimated from measured protein and fat deposits (Table 3). These energy deposition values were then applied to weight increments taken from the UK-WHO Growth Standards (RCPCH, 2011) (Table 5) to derive EARs for breastfed, breast milk substitute-fed and for those infants where feeding is mixed or unknown. These standards describe the growth pattern of healthy infants living in non-deprived circumstances who were exclusively or predominantly breastfed for at least four months and introduced to complementary foods at a mean of 5.4 months (WHO, 2006). Since this pattern of growth is associated with favourable health outcomes, it is considered an indicator of optimal growth applicable to all infants and young children. In 2009, standards for children aged 0-4 years of age were adopted by the UK (Scientific Advisory Committee on Nutrition & the Royal College of Paediatrics and Child Health [SACN/RCPCH], 2007) as the growth standards to be used for clinical monitoring and population surveillance.
Energy expenditure The FAO/WHO/UNU report (FAO, 2004) describes several equations relating TEE 96. to infant body weights. One simple prediction equation for TEE as a function of body weight was derived from a longitudinal study of TEE with DLW measures conducted at three month intervals for the first two years of life on 76 infants (40 breastfed and 36 breast milk substitute-fed) (Butte et al., 2000a).
TEE (MJ/day) = 0.371 kg – 0.416 n = 320, r = 0.85, see = 0.456 MJ/day (109 kcal/day)
This was very similar to the relationship between TEE and weight derived from an 97. analysis of 13 published studies with DLW performed on a total of 417 infants aged 0-12 months using the mean values for TEE and body weight (Butte, 2005).
Exclusive breastfeeding to the age of about six months with continued 98. breastfeeding as part of a progressively varied diet is recommended for all healthy
48
infants (SACN/RCPCH, 2007). The energy expenditure of breast milk substitute- fed infants has been shown to be higher than that of breastfed infants (Butte et al., 1990; Jiang et al., 1998; Butte et al., 2000a;), indicating differences in TEE between feeding groups over the first year of life that diminish thereafter.
When predicting TEE for breastfed and for breast milk substitute-fed infants from 99. body weight, because of the differences described above, separate regression equations for TEE as a function of weight were derived for these two groups (Butte, 2005). The between individual coefficient of variation (CV) of TEE ranged from 15 to 21% (18% average) or from 13 to 17% for TEE/kg (15% average). The equations used to predict TEE from body weight are as follows:
TEE for breastfed infants (FAO, 2004):
TEE (MJ/day) = 0.388 Weight (kg) – 0.635; standard error of estimate (SEE) = 0.453 MJ/day (108kcal/day)
TEE (kcal/day) = 92.8 Weight (kg) – 152; SEE = 0.108
TEE for breast milk substitute-fed infants (FAO, 2004):
TEE (MJ/day) = 0.346 Weight (kg) – 0.122; SEE = 0.463 MJ/day (110kcal/day)
TEE (kcal/day) = 82.6 Weight (kg) – 29.0; SEE = 0.110
Weights substituted in these equations were derived from the WHO Child Growth Standards (WHO, 2006) discussed below (paragraph 101).
Energy deposition The energy stored in new tissue was calculated as the deposited energy accrued 100. during normal growth. These costs were estimated from a multi-component body composition model (total body water, total body potassium and bone mineral content) (Butte et al., 2000b) based on a modified version of Fomon’s term infant reference (Fomom et al., 1982) describing changes in body composition during growth. Estimates of protein and fat gain over three month periods were used to predict energy accrued per gram of weight gain, which was then used to predict growth costs at monthly intervals.
49
Table 3 Energy content of tissue deposition of infantsa
Age interval (months) Protein gain
(g/d) Fat mass gain
(g/d)
Energy deposited in growing tissues
(kJ/g)
Boys
0-3 2.6 19.6 25.1
3-6 2.3 3.9 11.6
6-9 2.3 0.5 6.2
9-12 1.6 1.7 11.4
Girls
0-3 2.2 19.7 26.2
3-6 1.9 5.8 15.6
6-9 2.0 0.8 7.4
9-12 1.8 1.1 9.8 aButte, 2005 Gross energy equivalents: 1g protein = 23.6kJ (5.65 kcal); 1g fat = 38.7kJ (9.25kcal)
Using this model, the estimate of energy deposited in new tissue fell from about 101. 25.6 kJ/g (6.3 kcal/g) at 0-3 months to about 10.6 kJ/g (2.5 kcal/g) at 9-12 months (see Table 3). These values were applied to the weight velocities observed in the UK- WHO Growth Standards (RCPCH, 2011) to estimate the rates of energy deposition at monthly intervals (see Table 4 for weight velocities). These predictions of energy deposited during growth derive from a relatively small study by Butte et al (2000a) which was “validated” against other data sets (Butte, 2005). It is assumed that these values for the energy deposited in new tissue are appropriate for children growing according to the WHO weight velocity values, even though in the original study (Butte et al, 2000a) the pattern of breastfeeding followed was not fully described and the growth of infants did not fully reflect the WHO growth trajectory (WHO, 2006). In the study by Butte et al (2000a) no significant differences in the body composition of breastfed and breast milk substitute-fed infants were noted, but further information on this point is lacking.
50
Table 4 Weights, growth and energy deposition rates for infants 1–12 months of age
Age (months) Weight
(kg)a Weight velocity
(g/day)b
Energy deposition
(kJ/g)
Energy deposition
(kJ/day)
Boys
1 4.47 37.1 25.1 931
2 5.56 35.8 25.1 899
3 6.37 26.6 25.1 668
4 7 20.7 11.6 240
5 7.51 16.8 11.6 195
6 7.93 13.8 11.6 160
7 8.3 12.2 6.2 76
8 8.61 10.2 6.2 63
9 8.9 9.5 6.2 59
10 9.16 8.5 11.4 97
11 9.41 8.2 11.4 93
12 9.65 7.9 11.4 90
Girls
1 4.19 31.6 26.2 828
2 5.13 30.9 26.2 810
3 5.84 23.3 26.2 610
4 6.42 19.1 15.6 298
5 6.9 15.8 15.6 246
6 7.3 13.1 15.6 204
7 7.64 11.2 7.4 83
8 7.95 10.2 7.4 75
9 8.22 8.9 7.4 66
10 8.48 8.5 9.8 83
11 8.72 7.9 9.8 77
12 8.95 7.6 9.8 74 a 50th percentile weight for age of the UK-WHO Growth Standards (RCPCH, 2011) b 50th percentile weight increment of the UK-WHO Growth Standards (RCPCH, 2011)
51
Table 5 Estimated Average Requirement (EAR) values for infants 0–12 months of age
Age (months)
Total energy expenditure (kJ/day) EARe
Breast- feda
Breast milk sub- stitute-
fedb
Feeding mixed or unknown
c,d
Breast-fed Breast milk
substitute-fed Feeding mixed or
unknown
kJ/day kJ/kg
per day kJ/day kJ/kg
per day kJ/day kJ/kg
per day
Boys
1 1099 1425 1242 2030 454 2356 527 2173 486
2 1522 1802 1647 2421 435 2701 486 2546 458
3 1837 2082 1947 2505 393 2750 432 2615 411
4 2081 2300 2181 2321 332 2540 363 2421 346
5 2279 2476 2370 2474 329 2671 356 2565 342
6 2442 2622 2526 2602 328 2782 351 2686 339
7 2585 2750 2663 2661 321 2826 340 2739 330
8 2706 2857 2778 2769 322 2920 339 2841 330
9 2818 2957 2886 2877 323 3016 339 2945 331
10 2919 3047 2982 3016 329 3144 343 3079 336
11 3016 3134 3075 3109 330 3227 343 3168 337
12 3109 3217 3164 3199 332 3307 343 3254 337
Girls
1 991 1328 1138 1819 434 2156 514 1966 469
2 1355 1653 1487 2165 422 2463 480 2297 448
3 1631 1899 1751 2241 384 2509 430 2361 404
4 1856 2099 1966 2154 336 2397 373 2264 353
5 2042 2265 2144 2288 332 2511 364 2390 346
6 2197 2404 2292 2401 329 2608 357 2496 342
7 2329 2521 2418 2412 316 2604 341 2501 327
8 2450 2629 2533 2525 318 2704 340 2608 328
9 2554 2722 2634 2620 319 2788 339 2700 328
10 2655 2812 2730 2738 323 2895 341 2813 332
11 2748 2895 2819 2825 324 2972 341 2896 332
12 2838 2975 2904 2912 325 3049 341 2978 333 a Total Energy Expenditure (TEE) (MJ/day) = 0.388 Weight (kg) – 0.635 b TEE (MJ/day) = 0.346 Weight (kg) – 0.122 c These figures should be applied for infants when the mode and proportions of feeding are uncertain d TEE (MJ/day) = 0.371 Weight (kg) – 0.416 e Calculated as TEE + energy deposition (kJ/day) as in Table 4
52
Energy reference values for children and adolescents aged 1-18 years The energy reference values for boys and girls aged 1-18 years were calculated 102. as TEE plus deposited energy costs using a factorial model BMR x PAL, with PAL adjusted for growth in terms of a 1% increase (see paragraph 108).
Estimating the BMR The BMR values for children and adolescents were estimated from the Henry 103. prediction equations (Henry, 2005) (see Appendix 4 for details), using weights and heights indicated by the 50th centiles of the UK-WHO Growth Standards (RCPCH, 2011) for ages 1 to 4 years and the UK 1990 reference for children and adolescents (Freeman et al., 1995) for ages 5 to 18 years.
Identifying PAL values PAL values were identified from an analysis of DLW measures of TEE. The objective 104. of the analysis was to identify specific age ranges of children within which variation with age was less than variation between individuals. This approach would allow PAL to be defined for these age-range groups in terms of its distribution: i.e. as the median, 25th and 75th centile range values. The analysis also examined the need to identify gender-specific PAL values.
A data set was compiled of all published DLW studies of children aged over one 105. year (see Appendix 12 for details). All studies were tabulated according to mean values for boys and girls for specific age groups. This resulted in 170 data points as study means representing a total of 3502 individual measurements (2082 females, 1420 males).
For all studies which did not report BMR, BMR values were calculated from the 106. Henry equations for weight and height or weight alone if no height was reported (see Appendix 4). PAL values were calculated from TEE and either the reported or calculated BMR values.
The analysis revealed no influence of gender but an increase in PAL values with age 107. as shown in Figure 2. From an early age, however, there was a wide range of study mean PAL values so that variation in PAL at any age was much greater than variation with age itself. Indeed, on the basis of the association of PAL values with activity levels (as shown in Table 22 Appendix 5), many of the study mean PAL values for younger school children (PAL<1.5) would imply very low activity levels compared with other children of the same age. Although one option was to exclude some of these studies, there was insufficient evidence to do this and all studies shown in Figure 2 were therefore included in the analysis. Thus, three age groups were identified within which the distribution of PAL values could be identified: 1-3 years, >3-<10 years and 10-18 years. To some extent, these age ranges reflect important periods of growth and development that could influence behaviour and energy requirements. These age ranges also correspond to the age ranges within which BMR prediction equations have been generated for both the Schofield and Henry
53
prediction equations. Median, 25th centile and 75th centile PAL values were identified for each of these groups and are shown in Table 6.
Figure 2 – Physical Activity Level (PAL) values for children and adolescents (aged 1-18 years) as a function of age.
44
Deleted: <sp>
Figure 2. Physical Activity Level (PAL) values for children and adolescents (aged 1-18 years) as a function of age.
group1 group 2 group 3 1.39 1.57 1.73
0 2 4 6 8 10 12 14 16 18 20
Age
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
PAL
Female Male
Median values for the indicated age ranges are shown. Each point represents a single age group reported in a single publication or in publications which list mean values for each of several age groups. Boys and girls are shown separately. The publications from which these data points and the mean values in Table 6 were derived are detailed in Table 38, Appendix 12.
Table 6. Physical Activity Level (PAL) values for the age groups 1-3, >3-<10 and 10-18 years derived from the DLW data from studies of children aged >1 year*
Age group (years)
Age (years) PAL
N mean sd Min Max mean sd Min 10th
centile Q25 Median Q75 90th
centile Max
1-3 14 2.3 0.54 1.5 3.0 1.39 0.06 1.26 1.30 1.35 1.39a 1.43 1.45 1.46 >3-<10 85 7.0 1.87 4.0 9.3 1.56 0.14 1.21 1.35 1.42 1.57b 1.69 1.77 1.98 10-18 71 13.0 2.38 10.0 18.0 1.75 0.13 1.42 1.58 1.66 1.73c 1.85 1.91 2.19
* PAL values calculated from DLW-derived TEE and measured or estimated BMR (using Henry equations).
Adjusting for growth 108. The gross energy deposited in new tissue should be added to the energy expended as a cost
of tissue deposition. Only the latter is part of the observed TEE. Deposited energy accounts for a relatively small overall proportion of the total energy needs of children at all ages after the first year of life (see Appendix 12). In the context of a factorial model of requirements the simplest approach is an adjustment of PAL, as suggested by FAO/WHO/UNU (FAO, 2004), which in effect represents growth as a fixed proportion of
Median values for the indicated age ranges are shown. Each point represents a single age group reported in a single publication or in publications which list mean values for each of several age groups. Boys and girls are shown separately. The publications from which these data points and the mean values in Table 6 were derived are detailed in Table 37, Appendix 12.
54
Table 6 Physical Activity Level (PAL) values for the age groups 1-3, >3-<10 and 10-18 years derived from the DLW data from studies of children aged >1 yeara
Age group (years)
Age (years) PAL
N mean sd Min Max mean sd Min 10th
centile Q25 Median Q75 90th
centile Max
1-3 14 2.3 0.54 1.5 3.0 1.39 0.06 1.26 1.30 1.35 1.39 1.43 1.45 1.46
>3-<10 85 7.0 1.87 4.0 9.3 1.56 0.14 1.21 1.35 1.42 1.57 1.69 1.77 1.98
10-18 71 13.0 2.38 10.0 18.0 1.75 0.13 1.42 1.58 1.66 1.73 1.85 1.91 2.19 a PAL values calculated from DLW-derived TEE and measured or estimated BMR (using Henry equations).
55
Adjusting for growth The gross energy deposited in new tissue should be added to the energy expended 108. as a cost of tissue deposition. Only the latter is part of the observed TEE. Deposited energy accounts for a relatively small overall proportion of the total energy needs of children at all ages after the first year of life (see Appendix 12). In the context of a factorial model of requirements, the simplest approach is an adjustment of PAL, as suggested by FAO/WHO/UNU (FAO, 2004), which in effect represents growth as a fixed proportion of the overall energy requirement. Growth costs calculated as a percentage of energy requirement are on average (1-16 years of age) 0.98% (min=0.4%, boys, 0.05%, girls; max= 1.37%, boys, 1.59%, girls). This indicates that a single average growth value calculated as 1% of energy requirements could be used throughout the age range. Use of this 1% value will result in underestimation of overall energy requirements during the peak growth phase for older children of up to 0.6% for girls or 0.4% for boys. There will be an overestimate by similar amounts as growth slows in late adolescence. Such errors were not viewed to be significant given the overall variation in PAL values between children. In this report growth costs were therefore accounted for by a 1% adjustment of PAL values for each age group. These adjusted values are shown in Table 7.
Table 7 Physical Activity Level (PAL) values for use in calculation of energy requirements of children and adolescents, adjusted for growth
Age group (years) PALa
Q25c Medianb Q75c
1-<3 1.36 1.40 1.45
3-<10 1.43 1.58 1.70
10-18 1.68 1.75 1.86 a PAL adjusted for growth (=PALx1.01) b Population Estimated Average Requirement (EAR) estimates for children and adolescents with average levels of physical activity c 25th and 75th centile PAL values used to calculate energy requirements for children with lesser or greater activity levels
Calculating energy reference values for children and adolescents Energy reference values can be calculated for boys and girls aged 1-18 years as 109. adjusted PAL x BMR. BMR values are estimated from the Henry equations (see Appendix 4), using weights and heights indicated by the 50th centiles of the UK- WHO Growth Standards (RCPCH, 2011) (ages 1 to 4 years) and the UK 1990 reference for children and adolescents (Freeman et al., 1995). Growth-adjusted PAL values representing the 25th (less active), median and 75th (more active) centiles of the PAL distributions for each age group (1-<3, 3-<10 and 10-18 years), are then used to calculate energy reference values for boys and girls, for each year (see Table 8). The reference weights used represent best estimates of healthy body weights and are about 15% lower than current UK weights (see Appendix 10). Thus, for those children who are overweight and for any underweight children, energy intakes at these levels will be associated with weight change.
56
Although it is difficult to predict and make judgements about the different activity 110. patterns likely to be associated with the three PAL bands shown in Table 8, it can be assumed that the 75th centile “more active” PAL value will represent a more desirable level of physical activity for the maintenance of health and that the 25th centile “less active” PAL value will represent a less desirable level.
Table 8 Estimated Average Requirement (EAR) values for energy based on median weights and heights from the WHO growth standardsa (ages 1 to 4) and the UK 1990 reference for children and adolescentsb for children aged >4 years
Boys Girls Energy requirements MJ/d
Age (years)
Boys Girls
Weight kg
height cm
BMRc MJ/d
Weight kg
height cm
BMRc MJ/d
less actived
Popula- tione
more activef
less actived
Popula- tione
more activef
1 9.6 76 2.29 9.0 74 2.12 3.1 3.2 3.3 2.9 3.0 3.1
2 12.2 87 3.02 11.5 86 2.78 4.1 4.2 4.4 3.8 3.9 4.0
3 14.4 97 3.46 13.9 96 3.23 4.7 4.9 5.0 4.4 4.5 4.7
4 16.3 104 3.67 16.0 103 3.43 5.3 5.8 6.3 4.9 5.4 5.8
5 18.6 110 3.89 18.2 109 3.63 5.6 6.2 6.6 5.2 5.7 6.2
6 21.0 117 4.14 21.0 117 3.88 5.9 6.6 7.1 5.6 6.2 6.6
7 23.0 123 4.34 23.0 123 4.07 6.2 6.9 7.4 5.8 6.4 6.9
8 26.0 129 4.61 26.0 129 4.32 6.6 7.3 7.9 6.2 6.8 7.4
9 29.0 134 4.87 29.0 134 4.57 7.0 7.7 8.3 6.5 7.2 7.8
10 31.5 138 4.83 32.0 139 4.63 8.1 8.5 9.0 7.8 8.1 8.6
11 34.5 143 5.09 35.9 144 4.84 8.5 8.9 9.5 8.1 8.5 9.0
12 38.0 149 5.38 40.0 149 5.05 9.0 9.4 10.0 8.5 8.8 9.4
13 43.0 155 5.77 46.0 154 5.34 9.7 10.1 10.8 9.0 9.3 10.0
14 49.0 164 6.26 51.0 160 5.60 10.5 11.0 11.7 9.4 9.8 10.4
15 55.5 170 6.75 53.0 162 5.70 11.3 11.8 12.6 9.6 10.0 10.6
16 60.2 173 7.09 55.3 163 5.80 11.9 12.4 13.2 9.7 10.1 10.8
17 64.0 175 7.36 57.0 163 5.87 12.3 12.9 13.7 9.8 10.3 10.9
18 66.2 176 7.52 57.2 163 5.88 12.6 13.2 14.0 9.9 10.3 11.0 a RCPCH, 2011 b Freeman et al., 1995 c Calculated from the Henry equations based on weight and height (see Appendix 4). Because the Henry equations have overlapping age bands (0-3, 3-10, 10-18 years), choice of equation is ambiguous at the age boundaries. Here the BMR equation for 3-10 year olds is used for the 3 year olds and the equation for 10-18 year olds is used for those aged 10 years on the basis of a smoother transition in the plot of BMR/kg against age. d Calculated with the 25th centile PAL value adjusted for growth shown in Table 7. e Calculated with the median PAL value adjusted for growth shown in Table 7. f Calculated with the 75th centile PAL value adjusted for growth shown in Table 7.
57
Energy reference values for adults The approach adopted for the determination of energy reference values for adults 111. was to utilise a factorial model based on BMR x PAL. BMR is estimated for healthy body weights, i.e. weights equivalent to a BMI of 22.5 kg/m2 at the height of the population group (as discussed in paragraph 92). Values listed are the current mean heights (for England and Scotland24) at various ages. The PAL values are identified from an analysis of suitable DLW measures of TEE.
Estimating BMR As already indicated there are several different equations currently used to 112. estimate BMR (see paragraphs 80 and 81, and Appendix 4 for details). In this report, the prediction equations of Henry (2005) have been used.
Identifying PAL values The approach taken to determine PAL values was to identify a data set of DLW 113. measures of TEE which could serve as a reference distribution of TEE and PAL values from which energy reference values for the UK adult population could be predicted. An initial survey of all published and other available DLW measures of TEE in healthy adults identified published studies which could be utilised in terms of study means. Two sets of individual data points were also considered:
a DLW data set of individual values drawn from the NDNS (n=66) (Henderson • et al., 2003a) and Low Income Diet and Nutrition Survey (n=36) (Nelson et al., 2007), and values from the unpublished NDNS comparison study (n= 90 adult)
the DLW data set (n=767) assembled for the US DRI report (IoM, 2005) which • includes most of the UK studies published up to the writing of that report.
With the exception of the NDNS data sets, subjects were not recruited to these 114. studies explicitly as a representative sample of the UK or any other adult population; NDNS randomly selects participants to be representative of the UK population while recognising that there may be recruitment bias into the DLW subsample (as discussed in paragraph 71). While the NDNS data sets are most appropriate, these are currently too small to serve as a reference. All of the published DLW studies included a wide range of BMI and a reasonable age distribution, but several involved investigations of physical activity measurement devices (e.g. accelerometers) and specifically recruited subjects following relatively high activity lifestyles. Although these studies illustrated the overall extent of the variation in PAL values, especially its upper and lower limits and its likely response to changes associated with specific activity programmes, the suitability of a combined data set was a cause for concern.
Subsequent to publication of the DRI report in 2005, two large population-115. based studies of energy expenditure measured using DLW have been published, the OPEN study (Subar et al., 2003; Tooze et al., 2007) and the Beltsville study
24 Representative data for Wales and Northern Ireland were not available.
58
(Moshfegh et al., 2008) (see Appendix 8 for details). The OPEN study involved healthy volunteers (n=451; 245 men and 206 women) aged 40–69 years; about 85% of volunteers were white with the remainder mainly black or Asian. The Beltsville study involved healthy volunteers (n= 478) aged 30–69 years. The subjects were predominately non-Hispanic white and were distributed evenly by sex and approximately by age. Both studies comprised an urban population with subjects recruited from the Washington DC metropolitan area. The combined OPEN and Beltsville cohorts contained similar numbers of men (48%) and women (52%), with levels of overweight and obesity very similar to current UK levels reported in the UK Health Surveys (see Appendix 10). Thus, mean BMI values were 27.0 kg/m2 (F), 27.5 kg/m2 (M) compared with current UK values of 27kg/m2 (M and F); 39% were classified as overweight and 25% obese, compared with 38% overweight and 23% obese in England in 2009 (NHS IC, 2010). From the perspective of ethnic mix and body weights, this population is therefore similar to the current UK population. However, no objective measures of physical activity were made in either study so it is not known how representative the distribution of TEE and PAL values is of the UK population. Individual PAL values from both studies were made available for this report25. BMR was measured in the Beltsville study but not in the OPEN study, so in the latter case BMR has been calculated using the Henry BMR prediction equations (Henry, 2005) based on weight and height. As shown in Figure 9 (Appendix 8), predicting BMR for an overweight/obese population did not appear to introduce bias since the regressions of PAL on BMI for the two cohorts were very similar.
Table 9 Distribution of Physical Activity Level (PAL) values in the Beltsville and OPEN studiesa,b
Distribution boundaries OPEN (n=451; age 40-69 years)
Beltsville (n=478; age 30-69 years)
Minimum 1.01 1.01
10th centile 1.40 1.32
lower quartile 1.49 1.46
median 1.61 1.62
upper quartile 1.77 1.78
90th centile 1.92 1.96
Maximum 2.61 2.34 a Beltsville data set (Moshfegh et al, 2008) b OPEN data set (Subar et al., 2003; Tooze et al., 2007)
The distribution of PAL values within the combined data set (n=929 individual 116. measures), was 1.01 to 2.61, as shown in Table 9. Investigation of the range of PAL values observed in healthy, mobile individuals with overall levels of physical activity which are sustainable, indicated a range of 1.38-2.5 (see Appendix 5) with a value of 1.27 representing a minimal survival requirement: i.e. minimal movement in waking hours as suggested by FAO/WHO/UNU in 1985 (WHO, 1985). Thus, values below 1.27 can be assumed to reflect either methodological error or to be non ambulatory
25 TEE data from the OPEN study were obtained from Amy Subar. TEE data from the Beltsville study was provided by Alanna Moshfegh.
59
individuals, highly dependent on others, or those with an unsustainable lifestyle. Similarly, subjects with PAL values >2.5 can be assumed to be participating in very high activity levels during the measurement period which are unrepresentative of usual activity. The combined data set was therefore trimmed for PAL values ≤1.27 or >2.5. This removed one subject with a PAL value >2.5 and 38 subjects with PAL values <1.27. However, the trimming had a minimal influence on the distribution characteristics increasing the median PAL value from 1.62 to 1.63 and the mean PAL value from 1.64 to 1.66 (see Table 10). The skewed distribution with subjects clustered at the lower end of the range is clearly apparent in Figure 3. The PAL values selected to predict energy reference values for the adult population are highlighted. The use of the 25th and 75th centile values is discussed below (see paragraph 119-122).
The derivation of EAR values using PAL values from a predominantly overweight 117. and obese population was carefully considered given the objective of identifying prescriptive EAR values consistent with long term health: i.e. values suitable for weight maintenance at a healthy body weight and at existing physical activity levels. For those subjects within the normal body weight range i.e. BMI 18.5 to 24.9kg/m2 (n=322, 36% total), the median PAL (1.61) was not significantly different from that of the overweight or obese subjects for men or women (see Table 30, Appendix 8). Furthermore, investigation of the influence of BMI on PAL values within the combined data set by regression analysis indicated that PAL did not significantly vary with BMI (p=0.64 for slope: R2 less than 0.1%: see Appendix 8 for detail). This implies that the rate of energy expenditure, as indicated by the median PAL for the cohort (i.e. 1.63), is not a specific characteristic of overweight or obesity and its use in deriving prescriptive EAR values consistent with healthy body weights is justifiable.
Regression analysis did show that PAL values decrease slightly with age. However, 118. age explained <1% of the variance (R2 =0.004), and since the slope is shallow, age has a minor influence on PAL, i.e. PAL = 1.69 at 30 years and 1.63 at 70 years. This indicates that energy reference values can be defined independently of age at least to the age of 70 years.
Table 10 Physical Activity Level (PAL) value statistics for the combined Beltsville and OPEN data setsa,b
N Mean SD Min 10th
centile 25th
centile Median 75th
centile 90th
centile Max
All 929 1.64 0.23 1.01 1.36 1.48 1.62 1.78 1.95 2.61
Trimmedc 890 1.66 0.21 1.27 1.40 1.49 1.63 1.78 1.96 2.50 a Beltsville data set (Moshfegh et al, 2008) b OPEN data set (Subar et al., 2003; Tooze et al., 2007) c Subjects with PAL values < 1.27(n=38) and >2.5 (n=1) excluded
60
Figure 3 – Distribution of Physical Activity Level (PAL) values for trimmed combined OPEN and Beltsville data setsa,b
50
Deleted: <sp>
Figure 3. Distribution of Physical Activity Level (PAL) values for trimmed combined OPEN and Beltsville data setsa,b
a Beltsville data set (Moshfegh et al, 2008) bOPEN data set (Subar et al., 2003; Tooze et al., 2007)
Calculating energy reference values 119. Tables 11 and 12 show energy reference values for men and women respectively. Specific
values of TEE and the consequent EAR are calculated from BMR x PAL as a function of age, gender, height and BMI. BMR is calculated with the Henry equations for healthy body weights equivalent to a BMI of 22.5kg/m2. For illustrative purposes, current mean heights of the population are shown for the population at the various age ranges as indicated by the Health Surveys for England (data for 2009) (NHS IC, 2010) and Scotland (Reilly et al., 2009). PAL is independent of gender and the change with age is too small to have any significant influence. Thus, a single set of PAL values calculated as the median (1.63), 25th (1.49) and 75th (1.78) centile boundaries of the reference population is used (see Table 31 and Appendix 8).
120. In the absence of other information on activity, the population EAR values are those derived using the median PAL (1.63) as the assumed population activity level. Reference values for population groups of men and women thought to be less or more active than average are those calculated from the 25th (1.49) and 75th (1.78) centile boundary PAL values and the BMR values shown in Tables 11 and 12. For population groups of different heights, weights can be calculated at a BMI of 22.5 kg/m2, as weight = 22.5 x height2. BMR can be calculated from the Henry (2005) equations (see Appendix 4), and population EAR values can be calculated as BMR x PAL (see Tables 11 and 12).
PAL (Physical Activity level) 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1 2.2 2.3 2.4 2.5
0
20
40
60
80
100
120
140
160
180
200 N
o. o
f o bs
er va
tio ns
.
median 1.63
25th centile 1.49
75th centile 1.78
a Beltsville data set (Moshfegh et al, 2008) b OPEN data set (Subar et al., 2003; Tooze et al., 2007)
Calculating energy reference values Tables 11 and 12 show energy reference values for men and women, respectively. 119. Specific values of TEE and the consequent EAR are calculated from BMR x PAL as a function of age, gender, height and BMI. BMR is calculated with the Henry equations for healthy body weights equivalent to a BMI of 22.5kg/m2. For illustrative purposes, current mean heights of the population are shown for the population at the various age ranges as indicated by the Health Surveys for England (data for 2009) (NHS IC, 2010). PAL is independent of gender and the change with age is too small to have any significant influence. Thus, a single set of PAL values calculated as the median (1.63), 25th (1.49) and 75th (1.78) centile boundaries of the reference population is used (see Table 31 and Appendix 8).
In the absence of other information on activity, the population EAR values are 120. those derived using the median PAL (1.63) as the assumed population activity level. Reference values for population groups of men and women, thought to be less or more active than average, are those calculated from the 25th (1.49) and 75th (1.78) centile boundary PAL values and the BMR values shown in Tables 11 and 12. For population groups of different heights, weights can be calculated at a BMI of 22.5 kg/m2, as weight = 22.5 x height2. BMR can be calculated from the Henry (2005) equations (see Appendix 4), and population EAR values can be calculated as BMR x PAL (see Tables 11 and 12).
61
Table 11 Estimated Average Requirement (EAR) values for energy for groups of men at various ages, weights and physical activity levels1, at current mean height for age and a Body Mass Index (BMI) of 22.5 kg/m2
EAR MJ/d
Age range (years) Heighta cm Weightb kg
BMRc MJ/day less actived populatione more activef
19-24 178 71.5 7.1 10.6 11.6 12.6
25-34 178 71.0 7.1 10.5 11.5 12.6
35-44 176 69.7 6.7 10.0 11.0 12.0
45-54 175 68.8 6.7 9.9 10.8 11.8
55-64 174 68.3 6.6 9.9 10.8 11.8
65-74 173 67.0 6.0 9.0 9.8 10.7
75+ 170 65.1 5.9 8.8 9.6 10.5
all adults 175 69.2 6.7 10 10.9 11.9 a Values for illustration derived from mean heights reported in the Health Survey for England 2009 (NHS IC, 2010) b At BMI= 22.5 kg/m2: i.e. weight = 22.5 x height2 c Calculated from the Henry prediction equations based on weight and height, where BMR = coefficient x weight (kg) + coefficient x height (m) + constant (see Appendix 4) d 25th centile Physical Activity Level (PAL) =1.49 e Median PAL= 1.63, f 75th centile PAL=1.78
Table 12 Estimated Average Requirement (EAR) values for energy for groups of women at various ages and physical activity levels, at current mean height for age and a Body Mass Index (BMI) of 22.5 kg/m2
EAR MJ/d
Age range (years) Heighta cm Weightb kg
BMRc MJ/day less actived populatione more activef
19-24 163 59.9 5.6 8.4 9.1 10.0
25-34 163 59.7 5.6 8.3 9.1 10.0
35-44 163 59.9 5.4 8.1 8.8 9.7
45-54 162 59.0 5.4 8.0 8.8 9.6
55-64 161 58.0 5.3 7.9 8.7 9.5
65-74 159 57.2 4.9 7.3 8.0 8.7
75+ 155 54.3 4.7 7.0 7.7 8.4
all adults 162 58.7 5.4 8 8.7 9.5 a Values for illustration derived from mean heights reported in the Health Survey for England 2009 (NHS IC, 2010) b At BMI= 22.5 kg/m2: i.e. weight = 22.5 x height2 c Calculated from the Henry prediction equations based on weight and height, where BMR = coefficient x weight (kg) + coefficient x height (m) + constant (see Appendix 4) d 25th centile Physical Activity Level (PAL) =1.49 e Median PAL= 1.63, f 75th centile PAL=1.78
62
The EAR values shown in Tables 11 and 12 have been calculated for subjects with 121. healthy body weights equivalent to a BMI of 22.5kg/m2, the lower end of the body weight range associated with minimum mortality (see paragraph 92). For subjects with BMI values greater than 22.5kg/m2, energy intakes at these EAR levels would be associated with weight change towards a healthier body weight.
As with the EAR reference values for children, although it is difficult to predict 122. and make judgements about the different activity patterns likely to be associated with the three PAL bands shown in Tables 11 and 12, it can be assumed that the 75th centile “more active” PAL value will represent a more desirable level of physical activity for the maintenance of health and that the 25th centile “less active” PAL value will represent a less desirable level.
Energy reference values for older adults Age-related changes in lifestyle and activity are very variable and there is little 123. evidence that older people have decreased physical activity while they remain mobile and in good health. In a cohort (n= 302) of community-dwelling US older adults (aged 70-82 years) (Manini et al., 2006) who are described as high- functioning, able to independently perform activities of daily living, and with no evidence of life-threatening illnesses, a wide variation of PAL values was observed. The mean PAL values for tertiles of PAEE were 1.48, 1.68 and 1.94. The overall mean PAL value for this group, 1.70, was slightly higher than that of the combined OPEN and Beltsville cohort of younger adults (mean PAL =1.64).
In advanced age, PAL values can be very low. In a group of 21 Swedish men and 124. women aged 91-96 years of age, all free- and independently-living, and termed healthy, all living a quiet life (some not having been out of doors for years), PAL values were on average 1.38 (1.13-1.65 after trimming for PAL values <1.1 on the basis of comments that BMR used as the divisor of TEE to calculate PAL may have been overestimated in some cases (Rothenberg et al., 2000)). Another study in free- living British men (n=23; all over 75 years of age) observed a mean PAL value of 1.5 (Fuller et al., 1996).
Taken together, these data indicate that it is difficult to generalise about the energy 125. requirements of older adults other than the requirements are unlikely to differ from younger adults whilst general health and mobility are maintained. However, in advanced age for those individuals with much reduced mobility it can be assumed that PAL values are likely to be lower. For these individuals and for older people who are not in good general health, energy requirements can be based on the less active, 25th centile PAL value of 1.49, recognising that for some groups of older people with specific diseases or disabilities or for patient groups who are bed- bound or wheelchair bound, the PAL value may be consistently lower than this.
Energy reference values for children, adolescents and adults outside the expected range of habitual activities The PAL values identified in the preceding sections have not been derived in 126. relation to any specific level of physical activity or lifestyle because, as discussed
63
in Appendix 5, it has proved very difficult to predict the PAL values of individuals as a function of lifestyle or even measured physical activity. The median PAL values for children (see Table 7) and adults (see Table 10) represent the midpoint of the distribution observed in the reference population and, as such, provide the best estimate of the average activity level for the population. In the absence of any other information on activity this value is the assumed population level and is used to calculate the population EAR. Based on the literature (as summarised in Table 22, Appendix 5), the median PAL values can probably be judged to represent a light activity lifestyle for school children and adults. The 25th and 75th centile PAL values have been identified for subject groups who are less or more active than the average and probably represent a sedentary, less desirable or moderate activity, more desirable lifestyle. The difference between these centile values and the median is 9% for adults and for children aged >3-<10 years, 5% for adolescents and only 3% for the youngest children. These values can be compared with the slightly greater range of ± 10% identified as the basis for the higher or lower than average activity levels by FAO/WHO/UNU in their analysis of energy requirements of children and adolescents (FAO, 2004). Judging when to assign subjects with the 25th less active and 75th more active centile values as opposed to the median population PAL values poses a difficult problem. As shown in Table 13, the difference of ±0.15 PAL units from the median for adults corresponds to ±30 minutes of moderate intensity activity five times per week, all other activities being the same. However, to date, it has not proved possible to identify from questionnaires or activity logs the actual behavioural differences which determine such variation in TEE in terms of identifiable exercise activities. Hopefully this will improve with the increasing use of objective methods of physical activity assessment (e.g. accelerometers) which can measure and record a much wider range of human movement including ‘incidental’ movement not captured by questionnaires.
Effects of additional physical activity on energy requirements Although the prediction of PAL values associated with a specific lifestyle cannot 127. be made with any certainty, predictions of the likely additional energy cost of well defined specific activities can be made with reasonable confidence (see Appendix 5). Estimates of the likely changes in PAL for a range of activities derived from theoretical calculations and from observed effects to the change in PAL for ten minutes and one hour of activity, or following the adoption of a high level physical activity training programme, are summarised in Appendix 5. Some examples are shown in Table 13. These examples are the likely increases in TEE (and hence requirements) expressed as the change in PAL for the listed activities for individuals leading lives with little physical activity, but maintaining body weight, and provided that no compensatory reduction in other activities or increase in energy intake occurs. These values have been derived primarily from studies on adults, but there is no reason to believe that they will be substantially different for children and adolescents.
64
Table 13 Examples of changes in Physical Activity Level (PAL) associated with increased activity (IoM, 2005)a,b
Change in PAL Activity
0.15 30 minutes of moderate intensity activity on 5 or more days of the weekc
0.2 60 minutes brisk walking (brisk =>6<7.5kmph, (=>4<5mph)) daily
0.3 60 minutes of active sport, (i.e. jogging at 9km/hr (6mph)), 5 times per week
0.4 60 minutes jogging at 9km/hr (6 mph) daily
0.6 An intense aerobic exercise programme associated with training for competitive sport daily
a More examples are shown in Appendix 5 b IoM, 2005 c Equivalent to previous recommendations from the Chief Medical Officer for England regarding physical activity in adults (DH, 2004)
Reference energy values during pregnancy and lactation The energy requirements for pregnancy and lactation are calculated as increments 128. to be added to the mother’s EAR. These are based on singleton pregnancies reaching term.
Ideally, women should begin pregnancy at a healthy body weight (BMI 18.5-24.9 kg/129. m2); the EARs for non-pregnant women identified in this report are set at amounts consistent with maintaining a BMI of 22.5kg/m2 (see paragraphs 90-93). Women who are overweight or underweight at the beginning of pregnancy are at risk of poor maternal and fetal outcomes (Han et al., 2011; McDonald et al., 2010; March of Dimes, 2002; Stothard et al., 2009). For such women, the relationships between weight change during pregnancy and fetal outcome are not completely understood. Given this uncertainty, a precautionary approach has been adopted and weight loss during pregnancy is not advised (NICE, 2010). The EARs for pregnancy and lactation defined in this report are therefore estimates of the incremental energy intakes likely to be associated with healthy outcomes for mother and child. These incremental energy intakes should be added to EAR values calculated at actual preconceptional body weights, rather than at healthy body weights for non- pregnant women.
The energy requirements for pregnancy also need to take into account the 130. protection of vulnerable groups. Adolescents who become pregnant must meet the dietary requirements imposed by growth, in addition to the demands of pregnancy and lactation. This is a complex issue. For example, consuming the extra energy needed to cover the costs of pregnancy does not in itself guarantee a better outcome (Kramer & Kakuma, 2003). Also, those under 18 years of age are at greater risk than older women of giving birth to infants who are of low birth weight by virtue of pre-term delivery or small size for gestational age (FAO, 2004). Thus, pregnancies up to 18 years are qualitatively different and must be considered differently.
65
Energy costs of pregnancy The energy costs associated with the maintenance of a normal pregnancy arise 131. from increases in maternal and feto-placental tissue mass, the rise in energy expenditure attributable to increased basal metabolism and changes in the energy cost of physical activity. Additional energy is also required to ensure energy stores are sufficient to support adequate lactation following delivery (Butte & King, 2005).
Gestational weight gain Gestational weight gain (GWG) is the major determinant of the incremental energy 132. needs during pregnancy. GWG determines not only energy deposition, but also the increase in BMR and the increase in TEE resulting from the energy cost of moving a larger body mass. The WHO Collaborative Study on Maternal Anthropometry and Pregnancy Outcomes (WHO, 1995) reported birth weights and maternal weight gains associated with lower risk of fetal and maternal complications. Birth weights between 3.1 and 3.6 kg (mean, 3.3 kg) were associated with the optimal ratio of maternal and fetal health outcomes. The range of GWG associated with birth weights greater than 3 kg was 10–14 kg (mean, 12 kg).
The tissue deposited during pregnancy includes the products of conception 133. (fetus, placenta, and amniotic fluid), maternal tissues (uterus, breasts, blood, and extracellular fluid) and maternal energy reserves as fat. A theoretical model has previously been used to estimate the energy requirements during pregnancy (Hytten & Chamberlain, 1991). This assumed an average GWG of 12.5 kg (0.9 kg protein, 3.8 kg fat, and 7.8 kg water), an efficiency of energy utilization of 90% and a mean birth weight of 3.4 kg. The total energy cost of pregnancy was estimated to be about 330 MJ (80,000 kcal) (WHO, 1985). Since publication of this model in 1991, a number of longitudinal studies in developed and developing countries have facilitated the revision of these theoretical estimates.
Longitudinal studies of body composition during pregnancy in well-nourished134. 26 women from the UK, USA, Netherlands and Sweden have observed a mean GWG of 11.9 kg at 36 weeks gestation. Extrapolating the calculations to 40 weeks of gestation suggests a total mean weight gain of 13.8 kg (Butte & King, 2005).
Basal metabolism in pregnancy In studies of healthy, well-nourished women with adequate weight gain during 135. pregnancy who gave birth to infants with adequate weights, average increases in BMR over pre-pregnancy values have been observed to be around 5, 10 and 25% in the first, second and third trimesters, respectively (Butte & King, 2005). There is, however, considerable variation in the cumulative increase in BMR (Prentice et al., 1996b).
In under-nourished populations such as The Gambia, adaptive changes in BMR, 136. and reduction in the amount of additional maternal fat stored during gestation,
26 The terms ‘well-nourished’ and ‘under-nourished’ are generally not defined in the original studies which inform reference energy values during pregnancy and lactation.
66
can make a profound difference to the overall energy needs of pregnancy (Prentice & Goldberg, 2000). The extent of adaptive changes in BMR that occur in well-nourished populations is unclear, but the increase in BMR during pregnancy has been observed to vary in response to pre-pregnancy body fat content. Larger increases in BMR have been observed in those having a higher percent fat mass, while increased energetic efficiency in the basal state (i.e. a depressed BMR) has been observed in thinner women (Butte et al., 2004; Prentice et al., 1996b).
Total energy expenditure in pregnancy The TEE of pregnancy has been measured longitudinally using DLW techniques 137. in well-nourished, free-living women in Sweden, the UK and the USA (Butte et al., 2004; Forsum et al., 1992; Goldberg et al., 1991a, 1993; Kopp-Hoolihan et al., 1999). TEE increased throughout pregnancy in proportion to the increase in body weight. TEE increased by about 1, 6 and 19%, and weight increased by 2, 8, and 18% over baseline in the first, second and third trimesters, respectively. The estimated increments in TEE (0.1 (25), 0.4 (95) and 1.5 (360) MJ/day (kcal/day) in the first, second and third trimesters, respectively) are similar to the increments observed by 24-hour calorimetry (Butte & King, 2005). The average GWG was 13.8 kg, but for a mean GWG of 12 kg (observed in the WHO collaborative study (WHO, 1995)) the corresponding values would be 0.08 (20), 0.35 (85) and 1.30 (310) MJ/day (kcal/ day).
In the latter half of pregnancy, increases in body weight result in increased energy 138. costs for activities; however, women may compensate for this by reducing the pace or intensity with which the activity is performed (Butte & King, 2005; Prentice et al., 1996b). The extent to which women are able to modify habitual physical activity patterns during pregnancy will be determined by socioeconomic and cultural factors specific to the population; women who are sedentary prior to pregnancy will have little flexibility to reduce their level of physical activity further.
Changes in PAEE (TEE–BMR) during pregnancy are highly variable, but when 139. measured longitudinally by DLW in well-nourished women averaged -2, +3 and +6% in the first, second and third trimesters, respectively, relative to pre-pregnancy values (Butte & King, 2005). Because of the larger increment in BMR, PAL declined from 1.73 prior to pregnancy to 1.60 in late gestation in well-nourished women (Butte & King, 2005).
In another study of women of varying pre-pregnancy BMI140. 27, significant reductions in PAEE and PAL were also observed in all BMI groups as pregnancy progressed, while BMR increased gradually throughout pregnancy (Butte et al., 2004). The same study found that excessive GWG was mainly due to fat mass gain and not protein accretion (Butte et al., 2003a): GWG within the IoM recommendations (Institute of Medicine Food and Nutrition Board, 1990) was associated with appropriate birth weights and moderate postpartum fat retention; whereas women who gained weight above the IoM recommendations had significantly higher excessive
27 Participants were grouped into healthy underweight (BMI<19.8), normal weight (BMI 19.8-26.0) and overweight (BMI >26.0) according to 1990 Institute of Medicine recommended GWG ranges (Institute of Medicine Food and Nutrition Board, 1990).
67
fat retention at 27 weeks postpartum. Calculation of total energy costs indicate that due to the higher GWGs, maternal fat deposition, and increments in BMR, increases in dietary energy intakes are required as pregnancy progresses; increases in BMR and energy deposited in maternal and fetal tissues are not fully offset by reductions in physical activity. Studies such as this highlight the problems which can arise when energy costs of pregnancy are based on the sum of TEE and energy deposition, since the latter may reflect excessive fat deposition which is retained postpartum.
Calculation of energy requirements for pregnancy As discussed above, the energy cost of pregnancy is not evenly distributed over 141. the gestational period and this must be considered when calculating the energy requirements for pregnancy (Butte & King, 2005). Most weight is gained in the second and third trimesters, at rates of 0.45 kg per week and 0.40 kg per week respectively, compared with a GWG of 1.6 kg in the whole first trimester. The increases in BMR and TEE are most pronounced in the second half of pregnancy.
The total energy cost of pregnancy can be estimated both from the increment in 142. BMR and energy deposition, and from the increment in TEE and energy deposition. These two factorial approaches provide slightly different distributions, but when applied to the WHO Collaborative Study on Maternal Anthropometry and Pregnancy Outcomes mean GWG of 12 kg (WHO, 1995), the estimated extra energy cost of pregnancy is 321 MJ (77,000 kcal). This is divided into approximately 0.35 (85), 1.2 (285) and 2.0 (475) MJ/day (kcal/day) for the first, second and third trimesters, respectively (FAO, 2004). This is based on the assumption that increments in BMR and TEE are proportional to gestational weight gain.
Most dietary studies in well-nourished women, however, have revealed no or 143. only minor increases in energy intake that only partially covered the estimated energy cost of pregnancy. An analysis of available data from longitudinal studies in populations with average birth weights greater than 3 kg revealed a cumulative reported intake of only 85 MJ (around 20,000 kcal; or 0.3 MJ/day, 72kcal/day) over the whole of pregnancy (Prentice et al., 1996b). This is only 25% of the estimated needs or about 40% if allowances are made for under-reporting of energy intake85.
Approaches taken in other reports The FAO/WHO/UNU report (FAO, 2004) recommends an increase in food intake 144. of 1.5 MJ/day (360 kcal/day) in the second trimester and 2.0 MJ/day (475 kcal/day) in the third, based on a GWG of 12 kg and specific for women in societies with a high proportion of non-obese women who do not seek prenatal advice before the second and third month of pregnancy. Energy deposited as fat represents 45% of these energy costs.
The 1.5 MJ/day (360kcal/day) increment for the second trimester represents a 145. summing and rounding for ease of calculation of 1.2 MJ/day (285 kcal/day) and 0.35MJ/day (85 kcal/day) (for first trimester) on the grounds that women may not
68
know they are pregnant until the second trimester. This is pragmatic, but there is no evidence that retrospective energy supplementation of this type alters outcome and in circumstances of unconstrained food intake, increased requirements may have already been achieved through an increase in appetite.
The US energy DRI for pregnancy (IoM, 2005) was established using a DLW 146. database of individual energy expenditure measures of pregnant women with pre- pregnancy BMI values of 18.5 up to 25 kg/m2; the measures were obtained from studies in British (Goldberg et al., 1991b, 1993) Swedish (Forsum et al., 1992) and North American (Kopp-Hoolihan et al., 1999) women. GWG was between 11.6 and 13.5 kg. The median change in TEE was 33.5 kJ (8 kcal) per week of gestation with a large range of -238 kJ (-57 kcal) to +448 kJ (+107 kcal) per week. The value of 33.5 kJ (8 kcal) per week was supported by a subsequent study (Butte et al., 2004) which reported that TEE, as measured by DLW, increased linearly at a similar mean rate of 31 ± 42.7kJ (7.4 ± 10.2 kcal) per week of gestation in women with pre-pregnancy BMI values of 19.8 to 26.0 kg/m2. The total energy deposition during pregnancy was estimated to be 753 kJ/day (180 kcal/day). Thus, the estimated requirement for energy during pregnancy was derived from the sum of the TEE of the woman in the non-pregnant state plus a median change in TEE of 33.5 kJ/week (8 kcal/week) plus the energy deposited during pregnancy. For the first trimester, no increase in energy intake was recommended, on the basis of only small changes in TEE and weight gain. For the second and third trimesters, the energy deposition rate of 753 kJ/day (180 kcal/day) was added to the increase in TEE expected at 20 weeks (672 kJ; 160 kcal) and 34 weeks (1142 kJ; 272 kcal) to give the incremental values of 1.43 MJ/day (340 kcal/day) and 1.9 MJ/day (452 kcal/day), respectively. As with the FAO/WHO/UNU recommendations (FAO, 2004), maternal fat deposition accounts for a considerable part of these recommendations.
The COMA DRV report (DH, 1991) set an increment in EAR of 0.8 MJ/day (191 kcal/147. day) above the pre-pregnant EAR only during the last trimester, which is equivalent to an extra total intake of 75 MJ (around 18,000 kcal). It was noted that women who were underweight at the beginning of pregnancy, and women who did not reduce activity, may have a higher EAR.
An increasing proportion of women in the UK enter pregnancy at a weight exceeding 148. the healthy range and, especially for those who are obese, this may place them and their babies at increased risk (McDonald et al., 2010; March of Dimes, 2002; Stothard et al., 2009). In 2009, between 40% (16-24 years) and 53% (35-44 years) of all women of child-bearing age were either overweight or obese in the UK (NHS IC, 2010) and corresponding mean BMI values (kg/m2) were 24.8 and 27.2 (see Appendix 10). A study of UK pregnant women (Heslehurst et al., 2007) found that the incidence of maternal obesity at the start of pregnancy had increased from 9.9% to 16.0% between 1990 and 2004, and predicted this would rise to 22% by 2010 if the trend continued.
Energy reference values for pregnancy Current evidence suggests that, in general, it is unlikely that women in the UK require 149. extra energy in the first trimester of pregnancy and compensatory reductions in
69
PAEE during the second and third trimesters are likely to reduce the demand for extra energy intake at this time. Furthermore, although careful studies of energy expenditure and deposition indicate an apparent need for additional energy (Butte et al., 2003a), such calculations include fat deposition. This is not a determinant of birth weight (Butte et al., 2003a) and some fat deposition remains six months postpartum. Such fat deposition may contribute to weight gain in women through successive pregnancies and could be undesirable, though there is uncertainty about this. On the other hand, there is strong evidence that inadequate GWG is associated with decreased birth weight and fetal growth (small for gestational age) (Siega-Riz et al., 2009). Recommendations for constraining weight gain may also be inappropriate for pregnancies in vulnerable groups such as teenagers. Taken together and in the absence of sufficient evidence to revise the recommendation made by COMA (DH, 1991), in this SACN report it was considered prudent to retain the EAR for pregnancy set by COMA i.e. an additional intake of 0.8MJ/day (191 kcal/day) during the last trimester.
Energy costs of lactation The amount of milk produced and secreted, the energy content of the milk and 150. the energetic efficiency of milk synthesis theoretically determine the energy cost of lactation. There is little evidence of energy conservation, i.e. changes in BMR, thermogenesis or the energy cost of certain physical activities, compensating for these energy costs during lactation in well-nourished women (Butte & King, 2005). Fat stores that accumulate during pregnancy may cover part of the additional energy needs in the first few months of lactation.
Variation in the energy content of human milk is principally attributable to 151. fluctuation in milk fat concentration which shows complex diurnal, within-feed and between-breast changes. Twenty-four hour milk sampling schemes have been developed which minimally interfere with the secretion of milk flow and capture the diurnal and within-feed variation (Garza & Butte, 1986). The mean gross energy content of representative 24-hour milk samples analysed in a number of studies of well-nourished women was 2.80 kJ/g or 0.67 kcal/g from 1 to 24 months (Butte & King, 2005).
The biological efficiency of converting dietary energy into human milk has been 152. conservatively estimated to be about 80 to 85% (Butte & King, 2005). Based on a WHO-sponsored review (Butte & King, 2002), mean milk intakes through six months postpartum measured by the test-weighing technique were 769g/day for women exclusively breastfeeding. This value is an indicative average for infants of both sexes over the whole six months of exclusive breastfeeding. Correction of the mean milk intakes for the infant’s insensible water loss during a feed (assumed to be equal to 5%) gives a mean milk intake over the first six months postpartum of 807g/day (FAO, 2004) for exclusive breastfeeding.
Calculation of energy requirements for lactation The average total energy requirements associated with lactation can be estimated 153. by the factorial approach, whereby the cost of milk production (estimated from
70
the amount of milk produced, energy density of milk and the energetic efficiency of milk synthesis is added to the energy requirements of non-pregnant women, with an allowance made for energy mobilisation from tissue stores, if replete. In well-nourished women it has been estimated that on average 0.72 MJ/day of tissue stores may be utilised to support lactation during the first six months postpartum (Butte & King, 2002), based on a rate of weight loss of -0.8kg per month (Butte & Hopkinson, 1998). This will vary depending on the amount of fat deposited during pregnancy, lactation pattern and duration.
For exclusive breastfeeding during the first six months of life, the mean energy 154. cost of lactation over the six month period is 2.8 MJ/day (675 kcal/day) based on a mean milk production of 807g/day, energy density of milk of 2.8kJ/g (0.67kcal/g), and energetic efficiency of 0.80. This may be subsidised by energy mobilisation from tissues in the order of 0.72 MJ/day (172 kcal/day) (Butte & King, 2002), resulting in a net increment of 2.1 MJ/day (505 kcal/day) over pre-pregnancy energy requirements.
The estimated average energy requirement of breastfed infants in the first six 155. months of life is 2.28 MJ/day (545kcal/day) (see Table 5). If this is met by maternal supply and the energetic efficiency of milk synthesis is assumed to be 0.8, this also equates to a maternal energy cost of about 2.8 MJ/day (669kcal/day).
The total energy requirements may also be estimated from the sum of TEE plus 156. milk energy output, minus the energy mobilised from tissues. The measurement of TEE by DLW techniques circumvents any assumptions regarding the energetic efficiency of milk synthesis or activity energy expenditure, since they are included in TEE. This approach was taken in four studies between one and six months postpartum of well-nourished women who exclusively breastfed their infants (Butte et al., 2001; Forsum et al., 1992; Goldberg et al., 1991b; Lovelady et al., 1993). Milk energy output averaged 2.15 MJ/day (514kcal/day).
The US energy DRI for lactation (IoM, 2005) is based on this approach and uses a 157. DLW database of individual energy expenditure measures of lactating women with a pre-pregnancy BMI value of 18.5 up to 25 kg/m2. The measured TEE, milk energy output and estimated energy mobilisation from tissue stores are used to estimate energy requirements. Based on a milk energy output rounded to 500kcal/d and an average weight loss of 0.8kg/month, which is equivalent to 170kcal/day, the recommendation is for an increment of 1.38MJ/day (330kcal/day; 2.1MJ/day - 0.72 MJ/day (500 - 170kcal/day)) for the first six months. After the first six months, the energy cost of lactation will depend on the amount of breast milk, which is likely to be diminishing.
The FAO/WHO/UNU report (FAO, 2004) employs the factorial approach and 158. recommends an increase in food intake by 2.1MJ/day (505kcal/day) for the first six months of lactation in well-nourished women. It notes that energy requirements for milk production in the second six months are dependent on rates of milk production that are highly variable among women and populations.
71
Energy reference values for lactation Having reviewed the available evidence, it was159. agreed that the approach outlined above (paragraph 157) adopted for the US Energy DRI report (IoM, 2005), should be adopted for this report, i.e. an increment of 1380kJ/day (330kcal/day) for the first six months during which time exclusive breastfeeding is recommended. Thereafter, energy intake required to support breastfeeding will be modified by maternal body composition and the breast milk intake of the infant.
Comparisons with reference values for energy in the 1991 COMA report The energy reference values detailed in this SACN report derive from a methodology 160. which differs from the COMA report (DH, 1991). In contrast to the COMA report:
Energy reference values have been calculated from rates of TEE assessed by the • DLW method. This has been used either directly as a measure of TEE for infants or, in all other cases, to identify suitable PAL values which have been employed within a factorial calculation as BMR x PAL.
BMR has been estimated using the Henry prediction equations, resulting in • slightly lower BMR values than would have been estimated from the modified Schofield equations.
The revised population PAL values are higher. In this report, a value of 1.63 has • been used as the population PAL value for adults compared to a PAL value of 1.4 used by COMA in their commonly quoted summary table (Table 2.8; DH, 1991).
A prescriptive approach has been adopted, calculating energy reference values • on the basis of healthy body weights which are slightly lower than those used by COMA (DH, 1991).
The details of these differences compared with previous reports are outlined below. The use of the Henry BMR prediction equations is discussed in Appendix 4; the use of PAL values to determine energy reference values is discussed further in Appendix 5; and a comparison of actual EAR values in this and the 1991 COMA report (DH, 1991) is made in Appendix 13.
As a result of these methodological differences, the revised EAR values differ 161. from COMA in a non-uniform way for the various age groups. Compared to mean summary values reported by COMA (see Tables 2.6 and 2.8 (DH, 1991)), the revised EAR values are higher by 6-9% and 15-18% for adolescent boys and girls respectively, and for adults are 3% higher for men and 7-9% higher for women. Table 39 (Appendix 13) provides a comparison of the SACN EAR values against the COMA EAR values presented at the same body weight and PAL values as those used by SACN.
In the COMA report (DH, 1991) insufficient DLW data were available to enable 162. calculation of energy reference values for any age group. Such data that were available for infants up to 30 months were used together with energy intake data
72
to derive EAR values for infants up to 36 months. Energy intake data alone were used for all pre-adolescent children. A factorial approach was adopted for children aged 10-18 years, predicting TEE from BMR x PAL, and assigning PAL values of 1.56 for boys and 1.48 for girls with small additions made for growth costs. In this SACN report a median PAL value of 1.75 has been used for both boys and girls aged 10- 18 years, which includes the growth costs. This is reflected in the higher energy reference values for adolescents, especially girls, in this report.
For adults, COMA (DH, 1991) discussed a matrix of PAL values varying by occupational 163. and leisure activities from 1.4-1.9 for men and 1.4-1.7 for women, on the basis that judgements could be made about lifestyles of population groups. EAR values were also defined for various weights at nine PAL values between 1.4 and 2.2 (in Table 2.7, DH, 1991). A better understanding of the extent and nature of the variation in TEE within populations and lifestyle groups has clearly indicated the impracticality of predicting lifestyle-dependent PAL values with any precision. For adults in this SACN report, PAL values of 1.49, 1.63 and 1.78 equate to the 25th, median and 75th centile. These values represent the less active, typically active, and the more active, and can only be equated to lifestyles in very general categories: i.e. sedentary, low and moderate activity. However, simple recommendations for the likely influence of changes in activity on PAL values can be made. This approach, adopted here for children and adults, is similar in principle to that recommended for children and adolescents by FAO/WHO/UNU (FAO, 2004): i.e. identifying average PAL values from the DLW data with a ±15% variation for more or less active children. However, here, PAL values for broad age ranges (1-3, >3-<10, 10-18 years) have been aggregated, whereas FAO/WHO/UNU (FAO, 2004) predicted PAL values for each year of age by regression analysis of the DLW data.
Although not discussed as such, COMA164. (DH, 1991) identified weight-maintaining energy reference values. The present SACN report has taken into account the high proportion of the population who are overweight and obese and the generally low levels of physical activity in the UK, and has adopted a prescriptive approach (see paragraphs 90-93). This approach is similar in principle to that taken in the FAO/WHO/UNU report (FAO, 2004) which was also prescriptive (or “normative”), recommending that reference EAR values should be identified in relation to height and listing values which would maintain BMI between 18.5 and 24.9 kg/m2.
COMA recommended that the general adult population could be assumed to 165. have an inactive lifestyle. Accordingly, the population EAR was based on a PAL value of 1.4, which is lower than the 25th centile PAL identified here (i.e. 1.49) but at the time was thought to be an appropriate value. The updated EAR values, estimated from the published DLW-derived TEE data which have become available since 1991 and calculated using a population PAL value of 1.63, better reflect the energy requirements of the current UK population. Despite being calculated using a PAL value about 16% higher than that used by COMA, the revised all-adult energy reference values reported here are 2.8% higher for men and 7-9% higher for women. This is explained by the use of 5-7% lower average body weight for the all-age category of adult men and women and the slightly lower BMR values (see
73
Table 40 in Appendix 13 for details). It should be noted that these revised values are within the range of measurement error/uncertainty.
The energy reference value identified in this report for pregnancy, a single 166. daily increment of 0.8 MJ/day (191kcal/day) in the last trimester, is the same as previously recommended by COMA (DH, 1991), but considerably lower than that recommended by either FAO/WHO/UNU (FAO, 2004) (1.5MJ/day (359kcal/ day) during the second trimester and 2.0MJ/day (478kcal/day) during the third trimester), or in the US DRI report (IoM, 2005) (1.43MJ/day (340kcal/day) during the second trimester and 1.9MJ/day (454kcal/day) during the third trimester) (see paragraphs 144-148). As stated above (paragraphs 148-149) there are cogent arguments for avoiding excessive weight gain, especially fat gain, in pregnancy and evidence for adaptive changes in energy expenditure (paragraph 138) which reduce the need for additional food energy intake. Although such arguments were not taken into account in the previous reports, they are considered to be valid and sufficiently important because of the increasing prevalence of overweight and obesity to guide the recommendation in this SACN report.
The energy reference value identified here for lactation, an increment of 1.38MJ/167. day (330kcal/day) for the first six months, is the same as previously recommended in the US DRI report (IoM, 2005), but considerably lower than that recommended for the first six months by either COMA (DH, 1991), an increment of 1.9-2.4MJ/day (454-574kcal/day) or by FAO/WHO/UNU (FAO, 2004) (2.1MJ/day (502kcal/day). In all cases the values derive from estimated milk energy contents less the energy mobilised from maternal stores, with different calculations for the magnitude of these two components. The higher value recommended by FAO/WHO/UNU (FAO, 2004) is mainly due to the inclusion of a scaling factor for milk energy content of +25% to take into account an assumed inefficiency of dietary energy utilisation to provide for milk energy. Since such energy costs would appear as increased TEE in lactating women and because such increases are not observed in practice, the inclusion of such a scaling factor is unwarranted.
74
4 Conclusions and recommendations
The requirement for energy for all population groups was previously set at the level 168. of energy intake required to maintain body weight in otherwise healthy people at their existing levels of physical activity and to allow for any special additional needs (i.e. growth, pregnancy and lactation). Measurements of total energy expenditure (TEE) together with estimates of deposited energy provide the basis for estimating energy reference values. In the absence of growth, an individual’s TEE is the sum of daily energy used to maintain basal metabolic rate (BMR; metabolism at rest), energy expended in physical activity, and thermogenesis deriving mainly from food intake. TEE can be expressed as a multiple of BMR, the physical activity level (PAL); hence TEE or EAR is equal to BMR x PAL. BMR is predictable as a function of age, weight and gender, while PAL, which is a descriptor of lifestyle and/or behaviour as determinants of energy expenditure, is independent of these factors, at least as a first approximation. Thus for any PAL value, TEE can be predicted for any group from estimates of the BMR. The BMR is calculated for each population group from their heights and reference body weights at these heights, which are specifically chosen as those likely to be associated with long term good health. During growth, pregnancy and lactation the energy cost of tissue deposition also needs to be taken into account.
The most accurate practical means of measuring TEE in free-living individuals 169. integrated over days and weeks is the doubly labelled water (DLW) method. In this method the TEE of individuals is computed from estimates of CO
2 production,
which are derived from the rate of loss of the isotopes 18O and 2H from the body over a period of days following the administration of a known dose. The computation relies on a series of assumptions (e.g. about water losses and metabolic fuel use) the validity of which varies between and within individuals, but the limits of such variability are known and considered generally acceptable for the purpose of estimating EARs.
Energy requirements for populations have been estimated from DLW-derived 170. measures of TEE in reference populations together with estimates of deposited energy. There are two basic approaches: TEE can be modelled against different anthropometric characteristics of the reference population (e.g. age, healthy body weights) using regression equations which can then be used to predict TEE and EAR values for population groups with defined anthropometric values. Alternatively, measured TEE values and measured or predicted BMR values from the reference populations can be used to derive PAL values. These PAL values can then be used to estimate TEE and EAR values for population groups on the basis of their predicted BMR values at healthy body weights. This is a factorial approach to setting EARs. Both approaches are limited by the validity of the DLW method and, to a larger extent, by the representativeness of the study participants compared with those
75
groups to whom the EAR will be applied. Nevertheless, the DLW method remains the method of choice and is more accurate than others available.
As no large randomised population-based UK studies of any age group have 171. been conducted using DLW, and with measurements from the National Diet and Nutrition Survey as yet insufficient to serve as a reference population, a variety of other data sources (reference populations) have been used to derive the revised EAR values. Those identified as being the most likely to be representative of the UK population were as follows:
For infants, a longitudinal study of healthy American infants comprising similar • numbers of breast fed and breast milk substitute-fed infants (n=76 individuals) (Butte et al., 2000a).
For children and adolescents, a compilation of all identified mean study values • for specific ages for boys and girls (n=170 mean study values and about 3500 individual measurements) (Appendix 12).
For adults, two large population-based studies (Moshfegh • et al., 2008; Tooze et al., 2007) of adult urban populations in the US (n = 890 individuals; Appendix 8). These have similar demographic and anthropometric characteristics as the current UK population although some cultural differences may exist.
For older adults, no specific representative data set has been identified (see • paragraphs 123-125 for discussion).
The limitations of these data sets (reference populations) for deriving estimated 172. EARs for the UK population include the small numbers of participants in the infant and at some ages in the child cohorts and the lack of younger adults aged 18-30 years and older adults aged >80 years in the larger adult data set.
The application of these data sets to the determination of EAR values is as 173. follows:
For infants from birth to 12 months • , FAO/WHO/UNU energy requirement recommendations (Butte, 2005; FAO, 2004) were recalculated on the basis of more recent infant growth data from the WHO Multicentre Growth Reference Study (WHO Multicentre Growth Reference Study Group, 2006).
For pregnancy and lactation • , the COMA DRV (DH, 1991) and US DRIs (IoM, 2005) have been used respectively, and reference values given as incremental energy requirements for pregnancy and lactation.
For children, adolescents and adults • , revised reference values have been determined using a factorial method of estimating TEE from BMR x PAL, with BMR predicted for reference body weights which can be assumed to be associated with long term health.
For older adults • , evidence indicates EAR values estimated for adults should also be applied whilst general health and mobility are maintained (see paragraphs 124-126).
76
The distribution of PAL values within the reference population has been used to 174. indicate a population average value for PAL, as well as the extent to which it is lower or higher for less or more active population groups. Thus, three PAL values are identified for age bands of children, adolescents, and for adults. These are the median which represents the best estimate of the average level of activity for the population, and the 25th and 75th centiles which represent those who are less active and more active than average. In addition, estimates of the likely increase in PAL associated with varying increases in activity levels (for example, increased walking, running and participation in sport or high-level training regimes) are given. The median PAL values identified for adolescents and adults in this report are 1.75 and 1.63 respectively, with 1.63 representing the median PAL value of a reference population in which, like the UK, approximately 60% are overweight or obese. When energy expenditure is impaired due to immobility or chronic illness (e.g. in extreme old age), energy requirements can be based on the less active, 25th centile PAL value of 1.49. For some groups of older people with specific diseases or for patient groups who are bed-bound or wheelchair bound, the PAL value may be consistently lower than this (Tables 11 and 12).
BMR values have been calculated on the basis of healthy body weights as indicated 175. by the 50th centile of the WHO Child Growth Standards (RCPCH, 2011) (ages 1 to 4 years), the 50th centile of the UK 1990 reference for children and adolescents (Freeman et al., 1995) aged over four years, and equivalent to a BMI of 22.5kg/m2 at current mean heights for adults, which are consistent with long term health28. This approach was adopted in the context of the high prevalence of overweight and obesity in the UK and consequently, in contrast to the COMA EARs, the revised EAR values for both children and adults are prescriptive. The exception to this is pregnant women, where EARs are based on actual pre-conceptual body weights rather than prescriptive body weights.
Due to changes in the methodology for determining both PAL values and BMR, 176. the revised EAR values in this report differ from those of COMA in a non-uniform way. The median PAL values identified in this report are higher than those used by COMA in their summary tables i.e. 1.56 for boys, 1.48 for girls and 1.4 for adults. An analysis of the range and distribution of PAL values shows that COMA was likely to have underestimated the PAL values of generally sedentary populations because of an under-appreciation of the influence of routine activities of daily living on energy expenditure. The use of prescriptive body weights which are lower than those used by COMA in their summary tables means that the BMR values used in the SACN calculation of the revised EARs are correspondingly lower. This is in addition to the slightly lower BMR values predicted by the Henry prediction equations used here compared with the Schofield equations used by COMA and by FAO/WHO/UNU (FAO, 2004). A detailed comparison of the revised SACN values with those from COMA can be found in Appendix 13.
28 In adults, the healthy body weight range is defined as a body mass index (BMI) between 18.5-24.9kg/m2. For the purposes of calculating the revised EAR values, healthy body weights have been identified as those associated with a BMI of 22.5kg/m2.
77
The reference population identified for estimation of the adult energy reference 177. values comprises predominantly overweight and obese individuals (mean BMI=28 kg/m2: range 18-43 kg/m2), and is similar to the current UK population in this respect. It is important to recognise that this reference population was used to identify activity levels (as PAL values), rather than TEE values, and equivalent energy intakes. In fact, the median PAL value, 1.63, is at the upper end of the sedentary-light activity range identified by WHO/FAO/UNU (FAO, 2004) (1.4-1.69). The similarity in distribution of BMI to that in the current UK population enables it to serve as a reference in terms of energy expenditure. However, the derived population PAL value should not be assumed to specifically represent that of overweight and obese subjects; this is because no relationship was observed between the distribution of BMI and physical activity levels within the reference population group. Also, those subjects with healthy body weights exhibited the same average PAL value as those overweight and obese subjects. This observation indicates that the risk of overweight and obesity i.e. a positive energy imbalance, is independent of measured PAL values (see Appendix 8, Table 30). Nevertheless, the importance of proposals for increased physical activity for the general health of the population, referred to in paragraph 181 and 187 below, should not be underestimated.
These new population EAR values are prescriptive and calculated for body weights 178. thought to be associated with health benefits i.e. weights calculated at current average heights equivalent to a BMI of 22.5kg/m2 which represents the lower end of the minimum mortality range. Consequently, EAR values will be less than energy intakes which would maintain weight for those who are overweight. For such individuals, reference energy intakes will be associated with weight loss. For adults with a BMI of 30 kg/m2 with average levels of physical activity, intakes at the new population EAR values for energy would be less than their energy intakes for weight maintenance by on average 1.8MJ/day (426Kcal/day) for men and 1.1MJ/ day (260Kcal/day) for women.
On the basis of current evidence, it is not possible to define the extent or nature 179. of the impact of diet and physical activity on the risk of weight gain in any detail (see Appendix 11) and as indicated above (see paragraph 177), observed variation in physical activity levels in populations are not directly related to variation in levels of overweight or obesity. Nevertheless, it is clear that both physical activity (CMOs, 2011) and nutrient composition of the diet have key roles in maintaining good health (SACN, 2008). Weight gain is only possible when energy is consumed in excess of requirements, so it is important that individuals are aware of their energy intakes relative to their energy expenditure. Higher physical activity energy expenditures, balanced by a higher energy intake in the form of a nutritionally balanced diet could be expected to yield health benefits, by both increasing fitness and overall nutrient intakes. The EAR values calculated for a healthy body weight and with desirable physical activity consistent with long term health can be assumed to be those based on the “more active” PAL value (i.e. 75th centile). For overweight individuals who undertake this increased activity, energy intakes at this level should mediate weight change towards a more desirable weight.
78
The following general statement describes the energy requirements of adults, 180. with the principle illustrated by Figure 4. The statement should, however, be viewed with caution given the limitations of the data (outlined in paragraph 182). The population estimated average energy requirement is an energy intake from a nutritionally-balanced diet which, in subjects with healthy body weights, balances a TEE of about 1.63 (PAL) x BMR, with BMR varying with body weight, age and gender; this value can be assumed to be maintained in old age as long as subjects remain mobile and in good health. Rates of energy expenditure and consequent requirements vary markedly between individuals, mainly as a result of lifestyle, but also through individual behavioural characteristics. Within the population the overall range is normally between about 1.38-2.5 (PAL) x BMR; PAL values can fall as low as 1.27 (PAL) x BMR and may exceed 2.5, but such low and high values are not thought to be sustainable. Although it is inherently difficult to relate rates of energy expenditure of individuals or small population groups to identifiable activity levels, the population (median) PAL value of 1.63 can be used. For those judged to be either less or more active than average, representative PAL values of 1.49 and 1.78 are appropriate.
For maintenance of healthy body weights, individuals may require less or more than 181. these reference values, but individual appetite generally helps match intakes with expenditure. For those leading lives with little physical activity, but maintaining body weight, health benefits would follow from increasing energy expenditure and hence requirements by about 0.15 (PAL) x BMR. This can be achieved by 150 minutes of moderate intensity exercise per week (CMOs, 2011); a brisk walk of one hour/day would result in similar benefit. For very old people in whom activities become limited, energy expenditure and requirement are likely to be at the low end of the physiological range. For groups with average activity levels and body weights greater than values considered to be healthy, intakes at the population EAR level should mediate weight change towards a more healthy weight.
79
Figure 4 – Schematic representation of energy expenditure and consequent energy requirements of men and women as a function of the physical activity (PAL) valuea
Lifestyle/phenotype/activity: overall range sedentary extremely active
18 sedentary extremely active
Frequency within
population
150mins moderate intensity activity per week
Sport/strenuous leisure activitySport/strenuous leisure activity 1hr/d 5 times/week
0 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 2 0 2 1 2 2 2 3 2 4 2 5
Competitive training 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1 2.2 2.3 2.4 2.5
Physical Activity Values
1.49 less
ti
1.78 more
ti
1.63 population
active (25th
centile)
active (75th
centile) (median)
a The distribution of PAL values is that of the reference population discussed in Appendix 8. The likely increases in PAL (and consequent energy needs) shown for various activities are those identified in Table 13 and include the current UK recommendations that adults participate in at least 150 minutes of moderate intensity activity per week (CMOs, 2011). This would result in an increase in PAL of 0.15. These increases in activity are shown as they would affect initial PAL values at the 25th centile and median values, with the increases indicated by the length of the arrows. Although for those with PAL values at the 25th centile the recommendations for increased activity will only increase PAL to the median level, this can still be expected to have health benefits. As explained in Table 25, Appendix 5, the actual increase in PAL associated with a specific activity will vary to a small extent with the initial PAL, i.e. with the intensity of the activity replaced. This means that the actual increases will be slightly less than the values shown, especially for subjects with high initial PAL values.
Although these revised EAR values are based on a much larger body of evidence of 182. measured TEE than was available to COMA in 1991 (DH, 1991), it is important to note the insecurities which remain. These include the applicability of the data sets and reference populations from which they are derived to the current UK population; the paucity of data for younger adults aged 18-30 years and for older adults aged >80 years; the methods used to extract PAL values from measured TEE values when BMR has not been measured; the accuracy of the BMR prediction equations; and the measurement of TEE itself (see Appendix 5), especially the possibility of methodological bias influencing the values for the 3-10 year old children (see Appendix 12). Individual activity patterns and consequent rates of TEE can vary considerably over time, so that measures collected over 1-2 week periods provide no more than a snapshot of habitual activity.
The approaches used in this report, and previously by COMA (DH, 1991), have 183. produced reference values that ‘band’ population subgroups according to the mean EAR value for that group. This approach results in the reference value being too low for some people and too high for others. Caution is therefore required when applying these EAR values to groups since a mismatch between energy intake and energy expenditure (unlike most nutrients) has major public health implications; a
80
proportion of any group would gain weight inappropriately, while others would become under-nourished if everyone received the same energy intake.
The COMA EAR values (DH, 1991) of 10.6MJ/day (2550kcal/day) and 8.1MJ/184. day (1940kcal/day) as average values for all men and women respectively,29 are commonly quoted, e.g. in healthy eating messages to consumers. These figures were rounded to obtain the commonly cited values of 2500kcal/day for men and 2000kcal/day for women. The continuation of use of these values will require thorough deliberation. The COMA values were derived using average weights and with a PAL value of 1.4. The revised EAR values would indicate higher average values for all men and women respectively: 10.9MJ/day (2605kcal/day) and 8.7MJ/ day (2079kcal/day) (Table 16) i.e. 2.8% and 7-9% higher respectively. Given this relatively small difference, especially in the context of the uncertainty arising from the assumptions made in estimating these values, and given the concerns about increasing obesity, careful consideration should be given to the risks and benefits of updating the values particularly as the revised figures fall within the bounds of measurement error/uncertainty.
A question arises as to whether a reduction in population energy intake, as 185. recommended in this report, has implications for intakes of other nutrients. The revised EAR values for energy are based on the assumption of both a healthy body weight and consumption of a balanced diet which will supply adequate intakes of all nutrients. For population groups with healthy body weights, these new EAR values should not pose any risk of nutrient inadequacy. For those population groups showing a high prevalence of overweight or obesity, the new EAR values indicate a need to reduce food energy intake if physical activity is unaltered. Whenever food energy intakes are reduced with the intention of weight change towards a healthier, lower body weight, particular care should be taken to maintain a balanced diet to minimise risk of nutrient deficiency; micronutrient deficiencies have been observed for some UK population groups with unbalanced diets (SACN, 2008). In the case of obese individuals the reductions in energy intakes implied by the new prescriptive EAR values will be considerable and recommendations for such changes should be accompanied by expert dietetic advice on dietary nutrient balance.
For adults, the revised EAR values are broadly similar to the COMA EAR values 186. (DH, 1991) while for infants aged 0-3 months and adolescents they are higher. Consequently, the impact on the achievement of adequate nutrient intakes should be negligible. The revised EAR values for young children are lower than those previously set, but requirements for micronutrients in this group are generally low in comparison to energy requirements because the rate of growth slows considerably after the first year of life. Thus, the new recommendations should have minimal implications for this group. To meet micronutrient requirements, individuals should be encouraged to consume a healthy balanced diet containing
29 Table 2.8 Estimated Average Requirements (EARs) for energy for groups of men and women with physical activity level of 1.4 (DH, 1991).
81
a variety of foods including plenty of fruit, vegetables, starchy foods and some protein-rich and dairy foods.
Finally, it is important to recognise that where this new report indicates higher 187. than previous population EAR estimates, i.e. for adolescent boys and especially girls and for adult women, this does not mean that these groups are considered to have increased their activity. Instead, the new values represent a closer estimation of energy needs at current activity levels for the UK population. In fact, for a considerable proportion of all population groups who are overweight, the new reference intakes calculated for desirable body weights will be less than their current intakes and should help in achieving a healthier body weight. There is consistent evidence to support current public health recommendations from the UK health departments regarding physical activity; adults are recommended to have at least 150 minutes of moderate intensity physical activity a week (which can be achieved by 30 minutes of moderate activity on five or more days of the week) (CMOs, 2011). For children and young people, a total of at least 60 minutes each day of at least moderate to vigorous intensity physical activity is recommended. If the UK population responds to such recommendations, PAL values would increase: i.e. closer to the current 75th centile identified for the “more active” population for whom an increase in EAR to maintain energy balance may be appropriate.
Recommendations Rationale for recommendations The Dietary Reference Values (DRVs) for food energy describe the requirements 188. for food energy of the UK population and its subgroups in health. The Estimated Average Requirements (EAR) for energy for the UK population and its subgroups, have been calculated from measurements of total energy expenditure (TEE) in free-living people. In the absence of growth, TEE is primarily the sum of daily energy used to maintain basal metabolic rate (BMR; metabolism at rest) and the energy expended in physical activity. TEE can be expressed as a multiple of BMR, the physical activity level (PAL). Therefore,
TEE (or EAR) = BMR x PAL
During growth, pregnancy and lactation, the energy cost associated with tissue deposition or milk secretion must also be taken into account.
Body weight will be maintained with energy balance i.e. when energy intake 189. matches energy balance. Any sustained imbalance between energy intake and expenditure will lead to progressive weight gain or loss. In the UK, a substantial proportion of the population are overweight and obese. Obesity increases the risk for a number of diseases (such as type 2 diabetes, some cancers, hypertension and coronary heart disease) and adverse pregnancy outcomes, and represents a major public health problem for the UK.
In recognition of the high prevalence of overweight and obesity, the revised EAR 190. values for all population groups (with the exception of pregnant women), are prescriptive in relation to body weight. This means that they are set at the level
82
of energy intake needed to maintain a healthy body weight. In adults, the healthy body weight range is defined as equivalent to a body mass index (BMI30) between 18.5 - 24.9 kg/m2. For the purposes of calculating the revised EAR values, healthy body weights have been identified as those associated with a body mass index of 22.5 kg/m2. The prescriptive nature of the revised EAR values means that for population groups who are overweight, the EAR values are lower than the energy intakes required to maintain weight. Consequently, matching energy intake to the EAR values will facilitate weight reduction towards the healthy body weight range. However, it should be noted that the revised EAR values are set at existing levels of physical activity, i.e. a prescriptive PAL has not been used. EAR values set in line with current public health recommendations regarding increased physical activity (see paragraph 201), would therefore be higher than the revised values.
The 1991 COMA EAR values for energy (DH, 1991) were based on limited available 191. evidence. Since then, better measurements of TEE in a wide variety of population groups have substantially expanded and improved the evidence base.
Following a careful consideration of this evidence, SACN recommend a revision 192. to the EAR values for food energy for infants, children, adolescents and adults. They are intended only for use in healthy populations and are not intended for individuals or groups that require clinical management.
BMR should be calculated using the Henry prediction equations (see Appendix 4).193.
Exclusive breastfeeding is recommended for about the first six months of 194. life, however, recognising that infants are fed in a variety of ways, separate recommendations are therefore made for exclusively breast-fed, breast milk substitute-fed infants and where the mode of feeding is mixed or not known. Reference values for infants have been calculated from measurements of TEE plus allowances for energy deposited in new tissue. The latter were calculated by combining weight incremental data from the WHO Multicentre Growth Reference Study (2006) with estimates of energy deposited in new tissue.
30 Body Mass Index (BMI). An index of weight adequacy and obesity of older children and adults, calculated as weight in kilograms divided by the square of height in metres.
83
Table 14 Revised Estimated Average Requirements (EAR) for infants 1-12 months
Age (months) EAR a
Breast-fed Breast milk substitute-fed
Mixed feeding or unknown b
MJ/kg per day (kcal/ kg per day)
MJ/day (kcal/day)
MJ/kg per day (kcal/ kg per day)
MJ/day (kcal/day)
MJ/kg per day (kcal/ kg per day)
MJ/day (kcal/day)
Boys
1-2 0.4 (96) 2.2 (526) 0.5 (120) 2.5 (598) 0.5 (120) 2.4 (574)
3-4 0.4 (96) 2.4 (574) 0.4 (96) 2.6 (622) 0.4 (96) 2.5 (598)
5-6 0.3 (72) 2.5 (598) 0.4 (96) 2.7 (646) 0.3 (72) 2.6 (622)
7-12 0.3 (72) 2.9 (694) 0.3 (72) 3.1 (742) 0.3 (72) 3.0 (718)
Girls
1-2 0.4 (96) 2.0 (478) 0.5 (120) 2.3 (550) 0.5 (120) 2.1 (502)
3-4 0.4 (96) 2.2 (526) 0.4 (96) 2.5 (598) 0.4 (96) 2.3 (550)
5-6 0.3 (72) 2.3 (550) 0.4 (96) 2.6 (622) 0.3 (72) 2.4 (574)
7-12 0.3 (72) 2.7 (646) 0.3 (72) 2.8 (670) 0.3 (72) 2.7 (646) a Calculated as TEE + energy deposition (kJ/day) as in Table 5. b These figures should be applied for infants when there is mixed feeding and the proportions of breast milk and breast milk substitute are not known.
The population PAL values estimated by COMA in 1991 (i.e. 1.56 for boys, 1.48 for 195. girls, and 1.4 for adults) were underestimated. Even relatively sedentary populations exhibit higher rates of PAEE than previously thought. Median PAL values of 1.75 for adolescents and 1.63 for adults are thought to reflect current UK average activity levels better.
The revised population EARs for children aged 1-18 years old, based on the 50196. th centile of weights from the UK-WHO Growth Standards (ages 1-4 years) (RCPCH, 2011) and the UK 1990 reference for children and adolescents (ages 5-18 years) (Freeman et al., 1990) are:
84
Table 15 Revised population Estimated Average Requirements (EAR) for children aged 1-18 years olda
EAR MJ/d (kcal/d)
Age (years) PALb Boys Girls
1 1.40 3.2 (765) 3.0 (717)
2 1.40 4.2 (1004) 3.9 (932)
3 1.40 4.9 (1171) 4.5 (1076)
4 1.58 5.8 (1386) 5.4 (1291)
5 1.58 6.2 (1482) 5.7 (1362)
6 1.58 6.6 (1577) 6.2 (1482)
7 1.58 6.9 (1649) 6.4 (1530)
8 1.58 7.3 (1745) 6.8 (1625)
9 1.58 7.7 (1840) 7.2 (1721)
10 1.75 8.5 (2032) 8.1 (1936)
11 1.75 8.9 (2127) 8.5 (2032)
12 1.75 9.4 (2247) 8.8 (2103)
13 1.75 10.1(2414) 9.3 (2223)
14 1.75 11.0 (2629) 9.8 (2342)
15 1.75 11.8 (2820) 10.0 (2390)
16 1.75 12.4 (2964) 10.1 (2414)
17 1.75 12.9 (3083) 10.3 (2462)
18 1.75 13.2 (3155) 10.3 (2462) a Calculated from BMR x PAL. BMR values are calculated from the Henry equations, using weights and heights indicated by the 50th centiles of the UK-WHO Growth Standards (ages 1-4 years) and the UK 1990 reference for children and adolescents. b Physical Activity Level
The revised population EAR values for adults, calculated using a PAL value of 1.63 197. and BMR values calculated at weights equivalent to a BMI of 22.5kg/m2 at current mean heights for age are:
85
Table 16 Revised population Estimated Average Requirement (EAR) values for adults
Age range (years) Men Women
Height cma
EAR MJ/d (kcal/d)b
Height cma
EAR MJ/d (kcal/d)b
19-24 178 11.6 (2772) 163 9.1 (2175)
25-34 178 11.5 (2749) 163 9.1 (2175)
35-44 176 11.0 (2629) 163 8.8 (2103)
45-54 175 10.8 (2581) 162 8.8 (2103)
55-64 174 10.8 (2581) 161 8.7 (2079)
65-74 173 9.8 (2342) 159 8.0 (1912)
75+ 170 9.6 (2294) 155 7.7 (1840)
All adults 175 10.9 (2605) 162 8.7 (2079) a Values for illustration derived from mean heights in 2009 for England (Health Survey for England 2009) (NHS IC, 2010) b Median PAL= 1.63
These revised values apply to all adults, unless energy expenditure is impaired due 198. to immobility or chronic illness, e.g. in extreme old age. For this group a lower PAL value of 1.49 should be used.
For some population groups, the revised population EAR values are higher than 199. previous estimates. This should not be interpreted to mean that these groups have increased their activity and thus need to eat more, but rather that the new values represent a closer approximation of energy needs at current activity levels, estimated using updated methodology.
For pregnancy and lactation, the COMA DRV (DH, 1991) and US DRIs (IoM, 2005)200. are advised, respectively. For pregnancy an increment of 0.8MJ/d (191kcal/d) above the pre-pregnancy EAR31, during the last trimester only, is recommended. This assumes that the majority of women in the UK enter pregnancy adequately nourished and that there are decreases in physical activity during pregnancy that may partly compensate for the increased energy costs. It also assumes that the woman has completed her growth; this may not be the case for adolescents entering pregnancy. For lactation, an increment of 1.4MJ/d (335kcal/d) in the first six months is recommended.
There is consistent evidence that increasing physical activity is associated with 201. reductions in the risk of chronic diseases such as coronary heart disease, stroke and type 2 diabetes, and the risk of preventable death (DH, 2004). It may also reduce the risk of becoming obese provided energy expenditure is greater than energy intake. If the UK population respond to recommendations to increase physical activity, PAL values would increase and would be closer to the current 75th centile identified for the ‘more active’ population. EAR values would also increase as a result.
31 It is important to note that energy requirements for pregnant women are based on actual pre-conceptional body weights rather than healthy body weights for non-pregnant women (see paragraph 129).
86
These recommendations offer the best estimates of EAR values for population 202. subgroups. Values for individuals will vary considerably. Those population groups with BMI values greater than 25kg/m2 are likely to benefit from reduced food energy intakes. Increased physical activity is also likely to benefit health, and may help reduce body weight particularly if combined with a reduction in energy intake.
Research Recommendations Scientific evidence has expanded considerably since the previous COMA report 203. (DH, 1991). However, gaps in the data to enable estimation of energy requirements remain and in particular, data are lacking for younger adults aged 18-30 years and for older adults aged >80 years, and are limited for infants and children.
The PAL values identified in this report derive from data sets in which TEE, and 204. in most cases BMR, have been measured. This can be considered to be the best and most appropriate available evidence. Nevertheless, uncertainties exist about some of the data which has been used and specific UK population groups are also unrepresented within these data sets. Notably, the adult PAL values used in this report are derived from USA study populations and further investigation is needed both to establish PAL values for the UK population and the health outcomes associated with them.
In order to improve future recommendations for energy requirements research in 205. the following areas is required.
Measurements of TEE There is insufficient understanding of the variation in energy expenditure, and 206. hence energy requirements, within individuals over time and between individuals because of behavioural differences (i.e. spontaneous physical activity) are not currently described in lifestyle assessments. It is also not clear why the average value and distribution of PAL appears to be the same in subjects who are obese compared with those with normal weight. Therefore:
Intra-individual variation in DLW measures of TEE needs to be characterised. •
Improved understanding in all age groups of the relationship between patterns • of physical activity, in terms of its intensity and amount, and TEE, body weight maintenance and long term health is needed. This will enable the identification of reference values for physical activity which in turn will allow EAR values to be defined in terms of prescriptive values in relation to physical activity.
Identifying new and improving existing objective measures of energy expenditure, • especially those which allow integrated measurements of TEE comparable to the DLW method, is required.
Specific population groups The database of DLW-derived TEE values for children, adolescents, adults aged • 18-30 years and those aged >80 years should be expanded.
87
Further investigation is required to improve understanding of the marked • variation in physiological changes in energy requirements, expenditure and body composition in relation to pregnancy outcomes for mother and child especially in relation to maternal fat gain and its post-pregnancy retention.
There is a lack of evidence to establish energy requirements for overweight • and obese women during pregnancy. The effectiveness and safety of maternal weight management during pregnancy needs to be clarified to inform advice.
Further data on the relationship between energy intake and the quality and • quantity of growth in infants, regardless of the mode of feeding, should be collected.
Understanding of the potential interaction of diet composition and physical • activity in body weight regulation and the development and maintenance of obesity should be improved.
Activity patterns and values for TEE in older age groups likely to be associated • with the maintenance of mobility and reduced risk of morbidity and mortality need to be identified.
Prediction of the BMR The predictive accuracy of the Henry equations for BMR within the current UK • population needs to be established.
Acknowledgements The Committee would like to thank the principal investigators from the OPEN
and Beltsville studies, Amy Subar (National Cancer Institute, USA) and Alanna Moshfegh (United States Department of Agriculture), for providing data for this report. Thanks to Tim Cole (University College London Institute of Child Health, London) and Kirsten Rennie (University of Ulster) for their contributions to Working Group meetings.
88
Membership of Energy Requirements Working Group
Chair Professor Alan Jackson Professor of Human Nutrition, University of Southampton
Members Professor Marinos Elia Professor of Clinical Nutrition and Metabolism at (External Expert) the University of Southampton and Honorary Consultant Physician at Southampton General Hospital
Professor Ian Macdonald Professor of Metabolic Physiology at the University of Nottingham and Director of Research in the Faculty of Medicine and Health Sciences
Professor Joe Millward Emeritus Professor Human Nutrition, Faculty of (External Expert and Consultant) Health and Medical Sciences, University of Surrey
Professor Andrew Prentice Head of MRC International Nutrition Group and (External Expert) Professor of International Nutrition at the London School of Hygiene and Tropical Medicine
Professor Chris Riddoch Professor of Sport and Exercise Science, University (External Expert) of Bath
Dr Anita Thomas Consultant Physician in Acute Medicine, Plymouth Hospitals NHS Trust
Dr Anthony Williams Reader in Child Nutrition and Consultant in Neonatal Paediatrics, St George’s, University of London
Dr Stella Walsh Postgraduate Programme Leader, Leeds (Consumer representative) Metropolitan University
Dr Robert Fraser Reader, Reproductive and Developmental Medicine, (Co-opted from SACN’s Sub- University of Sheffield group on Maternal and Child Nutrition to advise on energy requirements during pregnancy and lactation)
89
External Consultant Dr Peter Sanderson
Secretariat Dr Alison Tedstone (Scientific)
Dr Sheela Reddy (Scientific)
Ms Rachel Elsom (Scientific)
Ms Emma Peacock (Scientific)
Mr Andrew James (Statistics)
Ms Verity Kirkpatrick (Scientific)
90
Membership of Scientific Advisory Committee on Nutrition
Chair Dr Ann Prentice Director, MRC Human Nutrition Research, (From June 2010) Cambridge
Professor Alan Jackson Professor of Human Nutrition, University of (Until June 2010) Southampton
Members Professor Peter Aggett Honorary Professor, School of Medicine and Health, Lancaster University, and Emeritus Professor and Past Head of School of Postgraduate Medicine and Health
Professor Annie Anderson Professor of Food Choice, Centre for Public Health (Until February 2011) Nutrition Research, University of Dundee
Late Professor Sheila Bingham Formerly Director, Medical Research Council’s Dunn (Until March 2009) Human Nutrition Unit, Cambridge.
Mrs Christine Gratus Retired advertising and marketing research director. (Lay representative)
Dr Paul Haggarty Head of Lifelong Health, Rowett Institute of Nutrition and Health, University of Aberdeen
Professor Timothy Key Professor in Epidemiology and Deputy Director of Cancer Epidemiology Unit, University of Oxford
Professor Peter Kopelman Principal, St George’s, University of London (Until November 2010)
Professor Susan Lanham-New Head, Nutritional Sciences Division and Reader in (From November 2009) Nutrition, Faculty of Health and Medical Sciences at the University of Surrey
Professor Julie Lovegrove Reader in Nutritional Metabolism and Deputy (From November 2009) Director of the Institute of Cardiovascular & Metabolic Research at the University of Reading
Professor Ian Macdonald Professor of Metabolic Physiology at the University of Nottingham and Director of Research in the Faculty of Medicine and Health Sciences
91
Professor Harry McArdle Deputy Director of Science and the Director of (From November 2009) Academic Affairs at the Rowett Institute of Nutrition and Health, University of Aberdeen
Dr David Mela Science Leader, Unilever R&D Vlaardingen, The (Industry representative) Netherlands
Professor Hilary Powers Professor of Nutritional Biochemistry and Head of (From November 2009) Human Nutrition Unit, University of Sheffield
Dr Anita Thomas Consultant Physician in Acute Medicine, Plymouth (Until February 2011) Hospitals NHS Trust
Professor Angus Walls Professor of Restorative Dentistry and Director of (From November 2009) The Centre for Oral Health Research, Newcastle Biomedicine, Newcastle University
Dr Stella Walsh Postgraduate Programme Leader, Leeds (Consumer representative) Metropolitan University
Dr Anthony Williams Reader in Child Nutrition and Consultant in Neonatal Paediatrics, St George’s, University of London
Professor Ian Young Professor of Medicine and Director of the Centre (From November 2009) for Public Health at Queen’s University Belfast
Observers Dr Alison Tedstone Department of Health
Dr Sheela Reddy Department of Health
Dr Fergus Millan Scottish Government, Health Department
Mrs Maureen Howell The Welsh Assembly, Health Promotion Division
Dr Naresh Chada Department of Health, Social Services and Public Safety, Northern Ireland
Secretariat Dr Elaine Stone (Scientific Secretary)
Mrs Vicki Pyne (Scientific)
Ms Rachel Elsom (Scientific)
Ms Rachel White (Scientific)
Mr Michael Griffin (Administrative)
92
Appendix 1 – SACN working procedures
Meetings
Energy Requirements Working Group meetings The SACN Energy Requirements Working Group was established in 2005. The 207. Working Group met three times in 2005, twice in 2006 and 2007, three times again in 2008, then twice in 2009 and one final time in 2010. At the final meeting, the Working Group considered the comments received in response to the consultation on the draft report (see below). The minutes of all the meetings are available on the SACN website (www.sacn.gov.uk).
Main SACN meetings The draft report was considered by the full Committee in its February 2009 and 208. June 2010 meetings and by correspondence.
Consultation The draft report was published on the SACN website on 5 November 2009. 209. Interested parties were invited to submit comments relating to the science of the report by 11 February 2010.
Submissions were received from the following organisations and individuals. 210.
1. British Dietetics Association (BDA)
2. British Nutrition Foundation (BNF)
3. Professor Elisabet Forsum, Linköping University, Sweden
4. Professor Jeya Henry, Oxford Brookes University
5. Ki Performance (Consultants) Limited
6. Medical Research Council Human Nutrition Research (MRC-HNR), Cambridge
7. Public Health Nutrition Research Group, University of Aberdeen
8. Safefood
9. School Food Trust
10. Sugar Bureau
11. Professor JT Winkler, London Metropolitan University
The Working Group’s response to all the submissions is available on the SACN 211. website (www.sacn.gov.uk).
93
Appendix 2 – Energy yields from substrates
This Appendix provides details on the energy provided by different macronutrients 212. and alcohol
Fat The amount of energy yielded from fat in food varies slightly with the type of 213. food, mainly according to the chain length and degree of saturation of constituent fatty acids in dietary triglycerides. It is generally assumed that the digestibility of all dietary fat is the same (≈95%). In practice, because dietary fat incorporates a mixture of different fatty acids these differences are ignored and the metabolisable energy (ME) content of dietary fat (the general Atwater factor) is assumed to be 37kJ (9.0 kcal/g), i.e. equal to the digestible energy content at 95% of gross energy (GE).
Carbohydrate The amount of energy yielded from different carbohydrates in food varies according 214. to the molecular form i.e. glucose, disaccharides and starch, the actual available energy content per unit weight is 15kJ (3.6 kcal/g), 16kJ (3.8 kcal/g) and 17kJ (4.0 kcal/g) respectively. FAO/WHO/UNU (FAO, 2004), however, has recommended that when carbohydrate is expressed as monosaccharide equivalents, a conversion factor of 16 kJ/g (3.8 kcal/g) should be used, and when determined by direct analysis, this should be expressed as the weight of the carbohydrate with a conversion factor of 17 kJ/g (4.0 kcal/g); the latter value being an estimated average of the different forms of carbohydrate in food. It should be noted that this method is used in the UK but not in all other countries. In the USA, for example, carbohydrate is calculated by difference, i.e. from total dry weight of food minus protein, fat and ash. This method includes unavailable carbohydrate (i.e. dietary fibre) and other non-carbohydrate components (e.g. lignin, organic acids, tannins, waxes, Maillard products) in the calculated value. This means that for an average diet, total energy intakes can be 12% higher and carbohydrate intake 14% higher when carbohydrate by difference is used in dietary analysis compared with available carbohydrate analysed directly. As a result, estimates of the extent of under-reporting in dietary surveys in the USA (Subar et al, 2003) are often lower than those in the UK (Dr Alison Lennox32, personal communication).
Other carbohydrates may also provide energy. Fermentation of non-starch 215. polysaccharides (NSPs) in the colon results in the formation of short-chain fatty acids, some of which are absorbed into the blood stream and are used as energy. A conversion factor of 8 kJ/g (1.9 kcal/g) has been suggested (FAO, 2003). The
32 Dr Alison Lennox is Head of Population Nutrition Research at the Medical Research Council Human Nutrition Research, Cambridge.
94
UK food composition table energy values only include carbohydrate expressed as monosaccharide (FSA, 2002).
Protein The amount of energy yielded from protein in food can vary according to both 216. the quality and the digestibility of the protein. These differences can mean that the energy yield from some highly digestible animal proteins like egg may be 40% greater than for some less digestible plant proteins. Also, in relation to the net metabolisable energy (NME), the immediate metabolic fate of amino acids absorbed into the body includes deamination and other reactions which contribute to the TEF. However, this energy loss is usually ignored. The ME content of dietary protein (the general Atwater factor) is assumed to be 17kJ (4.0 kcal/g), a value which is slightly lower than the actual value for most animal proteins and higher than that for most plant proteins. The average energy yield will usually be close to the assumed ME content because dietary protein usually represents a mixture of several animal and plant protein sources.
Alcohol Alcohol yields 29 kJ (6.9 kcal) per gram consumed. Alcohol oxidation starts rapidly 217. after absorption, and alcohol is eventually completely eliminated by oxidation (Prentice, 1995; Schutz, 2000) so that it does not directly add to body energy stores. Consumption of alcohol can modestly activate the hepatic de novo lipogenesis pathway, but some of the acetate produced in the liver by alcohol dehydrogenase (the major quantitative fate of ingested ethanol (Siler et al., 1999)) is released into plasma and inhibits adipose tissue lipolysis. This influence on tissue fuel selection can reduce fat oxidation and contribute to an increase in adiposity. The results from studies on the magnitude of the TEF after alcohol consumption vary, with reported values ranging between 9% and 28% (Prentice, 1995; Raben et al., 2003).
95
Appendix 3 – Components of energy expenditure and factors affecting it
Components of energy expenditure
Basal and resting metabolism The basal metabolic rate (BMR) is a standardised measure of an individual’s 218. metabolism in a basal, postabsorptive state: i.e. while awake and resting. This is a standardised metabolic state corresponding to the situation at thermoneutrality when food and physical activity have minimal influence on metabolism. BMR and resting metabolic rate (RMR) are often used interchangeably, but if RMR is not measured at the standardised metabolic state, it can include additional metabolic activity (e.g. thermic effect of food and excess post-exercise oxygen consumption, see paragraph 228), and would therefore be greater than the BMR. Because the BMR includes a small component associated with arousal, it is slightly higher than the sleeping metabolic rate.
Physical activity Physical activity is a complex and multi-dimensional behaviour taking place in a 219. variety of domains: in transportation, domestic life, occupation and recreation (Wareham & Rennie, 1998). The dimensions of a specific physical activity are defined as its volume, frequency, intensity, time and type.
‘Physical activity’, ‘exercise’, and ‘physical fitness’ are terms that describe different 220. concepts. Physical activity is defined as any bodily movement produced by the contraction of skeletal muscles resulting in energy expenditure (Caspersen et al., 1985). Exercise is a subset of physical activity that is planned, structured, and repetitive; its final or intermediate objective is the improvement or maintenance of physical fitness and/or health. Physical fitness is a set of attributes that are health and/or performance-related. The degree to which people have these attributes can be measured with specific tests (Caspersen et al., 1985).
Physical activity-related energy expenditure (PAEE) is quantitatively the most 221. variable component of total energy expenditure (TEE), usually accounting for 25- 50% of energy requirements and up to 75% of energy requirements (Westerterp, 1998) in extreme situations which are highly unusual and not sustainable. Due to differences in body size, skill and training there is a large inter-individual variation in the energy expended for any given activity.
The energy costs of different physical activities can be expressed as multiples of 222. BMR to account for differences in body size, i.e. the physical activity ratio (PAR). Such energy costs can also be expressed as multiples of the metabolic energy equivalent (MET): a fixed rate of oxygen consumption assumed to represent that of an adult measured at supine rest (defined as 3.5 ml oxygen consumption/kg body
96
weight/minute). 1 MET is similar to the BMR, but its precise relationship will vary as a function of body weight, age and gender, which each influence BMR per kg body weight (IoM, 2005). The physical activity level (PAL), the ratio of daily energy expenditure to BMR, used in TEE calculations is discussed in Appendix 5.
The role of physical activity in raising TEE depends on the intensity and duration 223. of the activity undertaken and whether this affects the degree to which other physical activities are performed, i.e. an increase in one component in PAEE may be balanced by a compensatory decrease in another. In addition, many studies of human subjects indicate a short-term elevation in RMR following single exercise events (generally termed the excess post-exercise oxygen consumption; EPOC). This EPOC appears to have two phases, one lasting less than two hours and a smaller, more prolonged effect lasting up to 48 hours (Speakman & Selman, 2003). The EPOC varies with exercise intensity and duration. The effects of long term exercise training on BMR, however, are less clear (Speakman & Selman, 2003).
Spontaneous physical activity and non-exercise activity thermogenesis Spontaneous physical activity (SPA) is a term used to describe all body movements 224. associated with activities of daily living, change of posture and ‘fidgeting’ (Ravussin et al., 1986). SPA accounts for a between-individual variation in energy expenditure of ± 15% and has been shown to be highly reproducible within individuals, a familial trait, and to significantly correlate with free-living TEE measured by doubly labelled water (DLW) (Snitker et al., 2001; Zurlo et al., 1992). SPA displays an inverse relationship with future weight gain (Zurlo et al., 1992) and it has therefore been described as a putative obesity subphenotype (Snitker et al., 2001). It has been argued that because SPA has only been quantified within a calorimeter it cannot be regarded as a component of free-living TEE in its own right in comparison to BMR and the thermic effect of food (TEF, see paragraph 228). It is described as “a useful paradigm to quantify an individual’s propensity to locomotion under standardized conditions” (Snitker et al., 2001).
Non-exercise activity thermogenesis (NEAT) is another term used to describe the 225. additional energy expenditure attributable to spontaneous physical activities other than volitional exercise, during everyday activities (Levine et al., 1999). Like SPA, NEAT has been implicated in energy balance regulation following observations in overfeeding studies that subjects exhibiting high levels of NEAT (Levine et al., 1999) displayed a resistance to fat gain. More recently, however, the term NEAT has been applied by the same author to all ambulatory activity, apart from specific sports, including relatively high level exertion activities such as dancing (Levine, 2004), and with walking identified as the predominant component. In this context, NEAT ceases to be an appropriate term since easily identifiable exercises have been included.
Although both SPA and NEAT may be used to describe “fidgeting” in popular parlance, 226. they are each likely to involve a much wider range of behaviours influencing energy expenditure. Thus, SPA as an “individual’s propensity to locomotion” (Snitker et
97
al., 2001) is likely to exhibit itself in terms of the non volitional choices between low and high expenditure activities which are a continuous feature of daily living (e.g. standing or walking up escalators, using lifts or stairs). In this report the term SPA, rather than NEAT, will be used in the wider sense to represent spontaneous physical activity that can occur within any domain. As such, SPA is potentially quantifiable as the technology for measuring movement and activity improves.
SPA, with its potential magnitude, and large between-individual but small within-227. individual variation, represents a behavioural phenotype which is likely to significantly influence individual energy expenditure independently of lifestyle categories.
Other components of energy expenditure
Thermic effect of food The thermic effect of food (TEF) or heat increment of feeding can be attributed 228. to the metabolic processes associated with ingestion, digestion and absorption of food, and the intermediary metabolism and deposition of nutrients. TEF varies mainly according to the amount and composition of dietary macronutrients and is usually assumed to account for 10% of energy intake (Kleiber, 1975). This means that for any individual the energy used in TEF varies in absolute terms according to overall rates of energy intake, and will usually be greater than 10% of BMR over a 24 hour period in individuals eating a mixed diet.
Factors affecting energy expenditure
Body size and composition Energy expenditure is related to body size, mainly in terms of weight and to a 229. lesser extent height. In general, larger people have more tissue mass than smaller people and therefore have a higher BMR.
The metabolically active tissue mass of the body is termed fat-free mass (FFM) 230. and comprises muscle, bone, skin and organs. FFM is the principal determinant of inter-individual variation in BMR and RMR, after adjustment for body size (Byrne et al., 2003; Heymsfield et al., 2002; Illner et al., 2000; Johnstone et al., 2005; Muller et al., 2004; Nelson et al., 1992; Wang et al., 2000; Weinsier et al., 1992). Fat mass (FM) has been observed to account for a small amount of the inter-individual variation in BMR and RMR in most studies (Cunningham, 1991; Ferraro & Ravussin, 1992; Fukagawa et al., 1990; Johnstone et al., 2005; Karhunen et al., 1997; Muller et al., 2004; Nelson et al., 1992; Svendsen et al., 1993; Weinsier et al., 1992), but not all (Bogardus et al., 1986; Segal et al., 1987).
In adults, variation in the composition of FFM, after adjustment for body size, may 231. also account for a small amount of the inter-individual variation in BMR and RMR (Gallagher et al., 1998; Garby & Lammert, 1994; Heymsfield et al., 2002; Illner et al., 2000; Sparti et al., 1997).
98
In infants, children and adolescents, there is an increase in BMR with age, due to 232. growth and increasing body size. Body composition changes during growth. At birth, the newborn is about 11% FM. Progressive fat deposition in the early months results in a peak in the percentage of FM (about 31%) at three to six months, which declines to about 27% by 12 months of age (Butte et al., 2000b). During infancy and childhood, girls grow more slowly than boys and girls have slightly more body fat. During adolescence, the gender differences in body composition are accentuated (Tanner, 1990). In boys, there is a rapid increase in FFM, coinciding with the rapid growth spurt in height, and a modest increase in FM during early puberty, followed by a decline; in girls, adolescence is characterised by a modest increase in FFM and a continual accumulation of FM. The pubertal increase in FFM ceases after about 18 years of age (Tanner, 1990).
Body mass and body composition have been shown to impact on PAEE and TEE in 233. children (Ekelund et al., 2004a; Johnson et al., 1998) and adults (Black et al., 1996; Butte et al., 2003b; Carpenter et al., 1995; Goran et al., 1993a; Mâsse et al., 2004; Plasqui et al., 2005; Roberts & Dallal, 1998; Rush et al., 1999; Schulz & Schoeller, 1994).
In adults, however, on a population basis and up to a moderate level of fatness 234. (BMI <30 kg/m2), the relative proportions of FFM and FM are probably unlikely to influence TEE in ways other than through their impact on body weight (Durnin, 1996). In adults with higher percentages of body fat composition, an effect on the mechanical efficiency of movement can increase the energy expenditure associated with weight-bearing activities (Prentice et al., 1996c).
Overweight and obese individuals have been shown to have higher absolute TEE 235. than normal weight individuals, because of the effect of a higher BMR associated with increased body size (FM and FFM (Das et al., 2004; Prentice et al., 1996a)). Because weight gain in men contains a higher proportion of FFM than in women, the increase in BMR with BMI is greater in men than women. For example, within the Beltsville DLW cohort examined in this report (Moshfegh et al., 2008) (Appendix 8), the slope of the regression of measured BMR on BMI in men was twice that in women. The observed increase in TEE is not in direct proportion to body weight since, when expressed per kg, both TEE and PAEE decline with increasing BMI (Prentice et al., 1996a). Nevertheless, the decline in PAEE/kg with increasing BMI is relatively shallow, decreasing in the combined OPEN (Subar et al., 2003; Tooze et al., 2007) and Beltsville (Moshfegh et al., 2008) adult DLW data sets used in this report by only 1% for each one unit increase in BMI. Because of this, even though this cohort included a wide range of BMI values (14.4 to 51 kg/m2: mean=27 kg/ m2), BMI explained less than 5% of the variance in PAEE and was not significantly correlated with either PAL values or the residuals from a multiple regression of age, weight, height and gender on TEE. This lack of any obvious difference in behaviour in terms of PAEE with increasing obesity is probably explained by any reduction in weight-bearing activity with increasing obesity cancelling out the increasing energy cost of such activities.
99
Gender Gender differences in TEE largely reflect differences in body size and composition, 236. with less FFM and more FM/kg in women than men. Thus, absolute and per kg TEE is lower in women. In most studies, BMR and TEE are observed to be lower in girls (Goran et al., 1994a; Goran et al., 1995; Kirkby et al., 2004) and women (Arciero et al., 1993; Carpenter et al., 1998; Dionne et al., 1999; Ferraro et al., 1992; Poehlman et al., 1997) even after adjustment for body size and composition; however, some studies have observed no differences in TEE between sexes after adjustment for FFM in children (Ekelund et al., 2004a) and adults (Blanc et al., 2004; Klausen et al., 1997).
In pre-menopausal women, a small increase in BMR, RMR, sleeping metabolic rate 237. and TEE during the luteal phase of the menstrual cycle, has been observed in several studies (Bisdee et al., 1989; Day et al., 2005; Ferraro et al., 1992; Hessemer & Bruck, 1985; Howe et al., 1993; Lariviere et al., 1994; Meijer et al., 1992; Melanson et al., 1996; Pelkman et al., 2001; Reimer et al., 2005; Solomon et al., 1982; Webb, 1986), although not all (Diffey et al., 1997; Kimm et al., 2001; Li et al., 1999; Piers et al., 1995). This suggests an effect on energy expenditure by sex hormones; in premenopausal women, pharmacological suppression of oestrogen and progesterone release has been observed to reduce RMR (Day et al., 2005). There have been no longitudinal studies of energy expenditure in women across the menopausal transition to determine whether the natural withdrawal of sex hormones influences energy expenditure. It has been speculated that suppression of ovulation with contraceptives could prevent the increase in energy expenditure observed in the luteal phase (Bisdee et al., 1989), but results from studies investigating the effect of contraceptive drug use on energy expenditure have been equivocal (Bisdee et al., 1989; Day et al., 2005; Diffey et al., 1997; Eck et al., 1997; Ferraro et al., 1992; Hessemer & Bruck, 1985; Howe et al., 1993; Kimm et al., 2001; Lariviere et al., 1994; Li et al., 1999; Meijer et al., 1992; Melanson et al., 1996; Pelkman et al., 2001; Piers et al., 1995; Reimer et al., 2005; Solomon et al., 1982; Webb, 1986).
Age There is a decline in BMR with older age and this is mainly attributable to the 238. progressive loss of FFM observed with aging (Kyle et al., 2001); however, a small decline in BMR with age, independent of any age-related changes in FFM, has been observed in most cross-sectional studies (Bosy-Westphal et al., 2003; Johnstone et al., 2005; Kim et al., 2002; Klausen et al., 1997; Pannemans & Westerterp, 1995; Piers et al., 1998; Poehlman et al., 1991; Roberts et al., 1995b; Vaughan et al., 1991; Visser et al., 1995), but not all (Cunningham, 1980; Das et al., 2001). The age-related decline in BMR was fully accounted for by a reduction in FFM and proportional changes in its metabolically active components in one study in healthy subjects (Bosy-Westphal et al., 2003).
It is unclear whether changes in the BMR with age are entirely a result of changes 239. in body composition or whether this is related to other factors, e.g. a decline in sodium-potassium ATPase activity (see paragraph 16), decreased muscle protein turnover, and changes in mitochondrial membrane protein permeability (Wilson
100
& Morley, 2003). A difficulty encountered in studies of the effects of aging on the decline in BMR is the differentiation of the aging process itself from common age- associated diseases and the subsequent effects on organ metabolic rates e.g. left ventricular hypertrophy (Bosy-Westphal et al., 2003).
PAEE has been observed to decline with aging (Roberts & Rosenberg, 2006; Wilson 240. & Morley, 2003), but the results from studies investigating an effect of aging on diet-induced thermogenesis have been inconsistent (Kunz et al., 2000; Melanson et al., 1998).
Genetics Genetic inheritance potentially influences all factors affecting inter-individual 241. variation in energy expenditure, e.g. body size and composition, and a familial influence on RMR, independent of FFM, age, and sex, has been reported (Bogardus et al., 1986; Bouchard et al., 1989). Genotype association studies have largely been restricted to candidate genes whose dysfunction might reasonably be expected to have an effect on energy expenditure.
Overall, the results do not show a consistent effect on energy expenditure of any 242. of the different genotypes studied thus far: the adrenoceptors (ADRB1 (Dionne et al., 2002; Nagai et al., 2003); ADRB2 (Oomen et al., 2005); ADRB3 (Dionne et al., 2001; Gagnon et al., 1996; Hojlund et al., 2006; Rawson et al., 2002; Rissanen et al., 1997; Shiwaku et al., 2003; Tchernof et al., 1999) the leptin receptor gene (Loos et al., 2006; Stefan et al., 2002; Wauters et al., 2002), the mitochondrial uncoupling protein genes (UCP1 (Oppert et al., 1994; Ukkola et al., 2001; Valve et al., 1998); UCP2 (Astrup et al., 1999; Kimm et al., 2001; Klannemark et al., 1998; Maestrini et al., 2003; Ukkola et al., 2001; Walder et al., 1998; Yanovski et al., 2000); UCP3 (Kimm et al., 2001; Ukkola et al., 2001) sodium-potassium ATPase genes (ATP1A1, ATP1BL1) (Deriaz et al., 1994) or the intestinal fatty acid binding protein 2 gene (Kim et al., 2001; Sipilainen et al., 1997).
Several other individual studies have found associations between energy 243. expenditure and genotypes for the glucocorticoid receptor gene (Di Blasio et al., 2003), interleukin-6 gene (Kubaszek et al., 2003), melanocortin-4 receptor gene (Rutanen et al., 2004; Krakoff et al., 2008) and the dopamine D2 receptor gene (Tataranni et al., 2001). Genome wide association studies, in several populations, have detected significant linkage on several chromosomes with measures of energy expenditure (Cai et al., 2008; Wu et al., 2004; Norman et al., 1998).
There have, as yet, been no clearly established relationships between specific 244. genotypes and energy expenditure.
Genetics of obesity Many studies have examined possible hereditary factors predisposing to human 245. obesity that influence energy expenditure. It is notable however that, thus far, all monogenic defects identified as causing human obesity disrupt hypothalamic pathways and have an effect on satiety and food intake (O’Rahilly & Farooqi, 2006). Genome-wide association studies have shown genetic variation in the FTO
101
gene (a 2-oxoglutarate-dependent nucleic acid demethylase gene (Gerken et al., 2007) to be associated with FM and obesity across multiple populations (Dina et al., 2007; Frayling et al., 2007; Hinney et al., 2007; Hunt et al., 2008; Kring et al., 2008; Scuteri et al., 2007), although only accounting for a small amount of the variation. In 13 cohorts, with a total of 38,759 participants from the UK and Finland, the 16% of adults who had the AA genotype for the FTO gene weighed about 3 kg more and had a 1.67 fold increased odds of obesity compared with the non-carriers (TT genotype) (Frayling et al., 2007). The genetic variation in the FTO gene has also been implicated in appetite control (Cecil et al., 2008; Timpson et al., 2008; Wardle et al., 2008; Wardle et al., 2009), but not energy expenditure, after adjustment for body size (Berentzen et al., 2008; Cecil et al., 2008; Do et al., 2008). The association between FTO variants and obesity has been observed to be reduced in those who are more physically active (Rampersaud et al., 2008) and accentuated in those who are less physically active (Andreasen et al., 2008). Inter- individual differences in susceptibility to obesity, therefore, may be determined, in part, by genetic variants impacting on appetite control in the presence of environmental exposures.
Ethnicity Differences in body composition and FFM composition exist between different 246. ethnic groups, e.g. between white and black (Jones, Jr. et al., 2004) and white and Asian populations (Soares et al., 1998; Wouters-Adriaens & Westerterp, 2008).
Most studies, although not all (Blanc247. et al., 2004; Kushner et al., 1995; Lawrence et al., 1988; Nicklas et al., 1997; Sun et al., 1998), suggest that BMR, adjusted for differences in FFM and FM, is lower in black subjects than white subjects (Albu et al., 1997; Blanc et al., 2004; Carpenter et al., 1998; Chitwood et al., 1996; Forman et al., 1998; Foster et al., 1997; Foster et al., 1999; Gannon et al., 2000; Jakicic & Wing, 1998; Kaplan et al., 1996; Kimm et al., 2001; Lovejoy et al., 2001; Morrison et al., 1996; Sharp et al., 2002; Sun et al., 2001; Treuth et al., 2000a; Weinsier et al., 2000; Weyer et al., 1999; Wong et al., 1999; Yanovski et al., 1997). This difference, however, may be due to racial differences in the composition of FFM i.e. metabolically active organ mass, rather than ethnic differences in metabolism (Byrne, 2003; Gallagher, 2006; Hunter, 2000; Jones et al., 2004; Tershakovec, 2002).
Racial differences in BMR, adjusted for body weight, between Asian and white 248. subjects/populations appear to be accounted for by differences in FFM and FM (Soares et al., 1998; Wouters-Adriaens & Westerterp, 2008). Differences in body composition, therefore, appear to be mainly responsible for the reported differences in energy expenditure between ethnic groups.
Endocrine state As discussed above (paragraph 237), sex hormones may affect energy expenditure. 249. Other hormones have also been implicated in the regulation of energy expenditure.
102
Thyroid status is a major determinant of metabolic rate. Hyperthyroidism increases 250. while hypothyroidism decreases RMR (Danforth & Burger, 1984). It is unclear, however, whether variation within the normal physiological range of plasma thyroid hormone, tri-iodothyronine (T
3 ), concentration is associated with variation
in BMR, independently of FFM (Al Adsani et al., 1997; Astrup et al., 1992; Bernstein et al., 1983; Johnstone et al., 2005; Muller et al., 1989; Obarzanek et al., 1994; Onur et al., 2005; Rosenbaum et al., 2000; Svendsen et al., 1993; Toubro et al., 1996; Van Wymelbeke et al., 2004; Welle et al., 1990).
Plasma noradrenaline concentration has been observed to be associated with 251. RMR, adjusted for FFM (Rosenbaum et al., 2000; Toubro et al., 1996), but not all studies have found this association (Obarzanek et al., 1994).
The hormone leptin is involved in energy balance and is produced primarily 252. in white adipose tissue; it is subject to acute regulation, particularly by the sympathetic nervous system (Trayhurn, 2001). It is unclear whether variation in plasma leptin concentration is associated with variation in RMR, adjusted for FFM and FM (Bobbioni-Harsch et al., 1999; Filozof et al., 2000; Haas et al., 2005; Johnstone et al., 2005; Jorgensen et al., 1998; Kennedy et al., 1997; Nagy et al., 1997; Neuhauser-Berthold et al., 2000; Nicklas et al., 1997; Pauly et al., 2000; Roberts et al., 1997; Salbe et al., 1997; Satoh et al., 2003; Toth et al., 1997; Wauters et al., 2002). Some of the discrepancies observed may reflect problems in accounting for the confounding effects of FM on plasma leptin concentration (Neuhauser- Berthold et al., 2000). Administration of leptin has not been shown to affect RMR (Hukshorn et al., 2000; Hukshorn et al., 2003a; Hukshorn et al., 2003b; Mackintosh & Hirsch, 2001; Rosenbaum et al., 2002), but during weight loss (negative energy imbalance) leptin may reduce the increased work efficiency of skeletal muscle, seen in response to energy restriction, and thereby reduce the observed decline in PAEE (Rosenbaum et al., 2003; Rosenbaum et al., 2005). The influence of leptin on energy expenditure is, therefore, unclear.
Metabolic stress and fever have also been observed to increase BMR; this is 253. discussed in Appendix 9.
Pharmacological agents Smoking has been shown to increase energy expenditure to a small extent 254. probably through the sympathoadrenal activation by nicotine (Collins et al., 1994; Kimm et al., 2001; Perkins, 1992). Caffeine also increases energy expenditure to a small extent (Arciero et al., 1995; Astrup et al., 1990) with an additive thermogenic effect to nicotine (Arciero et al., 1995; Collins et al., 1994; Jessen et al., 2003; Perkins et al., 1994). For alcohol, while no acute effect has been observed (Perkins et al., 1996), alcoholics have a higher RMR, adjusted for FFM, than healthy, social drinking controls (Addolorato et al., 1998) which falls with abstinence from alcohol (Addolorato et al., 1998; Levine et al., 2000).
Administration of glucocorticoids (Chong255. et al., 1994; Tataranni et al., 1996), adrenaline (Diepvens et al., 2007; Fellows et al., 1985), amphetamines and some anti-obesity drugs (Heal et al., 1998) have all been shown to increase energy
103
expenditure; whereas, the administration of opiates (Swinamer et al., 1988) and barbiturates (Dempsey et al., 1985) reduce energy expenditure. Growth hormone administration may increase energy expenditure (Wallace et al., 2002), but this may be partly explained by increased FFM (Hansen et al., 2005). β-blockers (β-adrenergic antagonists) have also been shown to reduce RMR (Jung et al., 1980).
Environment With cold exposure, increasing energy expenditure can occur by shivering 256. thermogenesis and increasing muscular activity (Haman, 2006), although this is unlikely to be a significant contributor to energy expenditure for those living at relatively constant ambient temperatures.
Studies of adult subjects executing a standard daily activity protocol at different 257. ambient temperatures (16 to 28°C) varied over the short-term, identified an inverse association of ambient temperatures with sedentary TEE (Blaza & Garrow, 1983; Buemann et al., 1992; Dauncey, 1981; Valencia et al., 1992; van Marken Lichtenbelt et al., 2001; van Marken Lichtenbelt et al., 2002; Warwick & Busby, 1990; Westerterp- Plantenga et al., 2002). Whether ambient temperature influences RMR, however, is less clear (Blaza & Garrow, 1983; Buemann et al., 1992; Dauncey, 1981; Lean et al., 1988; Valencia et al., 1992; van Marken Lichtenbelt et al., 2001; van Marken Lichtenbelt et al., 2002; Warwick & Busby, 1990; Westerterp-Plantenga et al., 2002).
Seasonal variation in measures of energy expenditure, adjusted for FFM have also 258. been reported, with increased energy expenditure during colder months (Bitar et al., 1999; Goran et al., 1998a; Plasqui et al., 2003; Plasqui & Westerterp, 2004). Seasonal differences in physical activity levels may also occur (Haggarty et al., 1994).
Overall, these studies suggest that the variation in energy expenditure due to 259. differences in environmental temperature can account for about 2-5% of TEE. The maintenance of indoor temperatures to within 20-25°C and the use of clothes to control body heat loss, however, mean that changes in ambient temperature are unlikely to have much impact on energy requirements in the UK.
104
Appendix 4 – Basal metabolic rate prediction equations
The factorial calculation of energy reference values as a basal metabolic rate 260. (BMR) multiple for any population group identified in relation to age, gender and size, requires appropriate BMR prediction equations. The FAO/WHO/UNU report (FAO, 2004) calculated BMR with Schofield equations (Schofield et al., 1985) (see Table 17). It has been shown that Schofield prediction equations may overestimate BMR in many communities; therefore alternative prediction equations, the Henry equations (see Tables 18 and 19), based on either weight or both weight and height have been developed (Henry, 2005).
The validity of all published prediction equations for resting energy expenditure 261. was tested in US and Dutch cohorts (separately) of overweight and obese adults (body mass index (BMI) range 25-40) aged 18–65 years (Weijs, 2008). The US cohort was a subset of those in the Dietary Reference Intakes (DRI) doubly labelled water (DLW) database (IoM, 2005). Against the US cohort, the accuracy of the Henry equations (% of subjects predicted within ±10% of the measured resting metabolic rate, (RMR)) was 79%, which was as good as, or better than, all other equations tested, including the Schofield equations (69% accurate). Against the Dutch cohort however, all prediction equations performed less well.
The two sets of prediction equations (Schofield 262. et al., 1985 and Henry 2005) were tested against a larger US cohort DRI DLW database (IoM, 2005) in which BMR values were measured for most subjects (n=767, age 20-96, BMI 18.5-62 kg/m2, men n=334; women n=433). The calculated statistics included accuracy, the root mean squared error (RMSE), and the mean, minimum and maximum % difference (bias) between estimated and measured RMR, as well as mean values for BMR and physical activity level (PAL).
The comparison is shown in Table 20.263. The differences between these prediction equations were small, but both Henry prediction equations, especially those based on weight and height, performed slightly better than the Schofield equations. The inclusion of height, as well as weight, is particularly appropriate as the range of healthy adult body weights used in this report was generated from BMI and height categories. Height was included in the original Schofield equations (Schofield et al., 1985), however, height is not generally included when these equations are used. Although there does not appear to have been any recent validation within the current UK population, the Henry equations based on weight and height were chosen to estimate BMR in this report.
105
Table 17 Prediction equations for BMRa: Schofieldb
Gender Age (years) BMR (MJ/day) BMR (kcal/day)
Weight coefficient constant
Weight coefficient constant
Males <3 0.249 -0.127 59.51 -30.4
3-10 0.095 2.110 22.706 504.3
10-18 0.074 2.754 17.686 658.2
18-30 0.063 2.896 15.057 692.2
30-60 0.048 3.653 11.472 873.1
>60 0.049 2.459 11.711 587.7
Females <3 0.244 -0.130 58.317 -31.1
3-10 0.085 2.033 20.317 485.9
10-18 0.056 2.898 13.38 692.6
18-30 0.062 2.036 14.818 486.6
30-60 0.034 3.538 8.126 845.6
>60 0.038 2.755 9.082 658.5 a Coefficients and constants shown for equations of the form BMR = weight coefficient x weight (kg) + constant b Equations as quoted by FAO/WHO/UNU (FAO, 2004) which derive from Schofield et al., 1985. These differ slightly from those quoted by Henry (2005) which derive from the modified Schofield equations (DH, 1991).
Table 18 Prediction equations for BMRa: Henryb weight and height
Gender Age (years) BMR (MJ/day) BMR (kcal/day)
Weight coefficient
Height coefficient constant
Weight coefficient
Height coefficient constant
Males <3 0.118 3.59 -1.55 28.2 859 -371
3-10 0.0632 1.31 1.28 15.1 313 306
10-18 0.0651 1.11 1.25 15.6 266 299
18-30 0.0600 1.31 0.473 14.4 313 113
30-60 0.0476 2.26 -0.574 11.4 541 -137
>60 0.0478 2.26 -1.070 11.4 541 -256
Females <3 0.127 2.94 -1.2 30.4 703 -287
3-10 0.0666 0.878 1.46 15.9 210 349
10-18 0.0393 1.04 1.93 9.40 249 462
18-30 0.0433 2.57 -1.180 10.4 615 -282
30-60 0.0342 2.1 -0.0486 8.18 502 -11.6
>60 0.0356 1.76 0.0448 8.52 421 10.7 a Coefficients and constants shown for equations of the form BMR = weight coefficient x weight (kg) + height coefficient x height (m) + constant b Henry, 2005
106
Table 19 Prediction equations for BMRa: Henry weightb
Gender Age (years) BMR (MJ/day) BMR (kcal/day)
Weight coefficient constant
Weight coefficient constant
Males <3 0.255 -0.141 61.0 -337
3-10 0.0937 2.15 23.3 514
10-18 0.0769 2.43 18.4 581
18-30 0.0669 2.28 16.0 545
30-60 0.0592 2.48 14.2 593
>60 0.0563 2.15 13.5 514
60-70 0.0543 2.37 13.0 567
>70 0.0573 2.01 13.7 481
Females <3 0.246 -0.0965 58.9 -23.1
3-10 0.0842 2.12 20.1 507
10-18 0.0465 3.18 11.1 761
18-30 0.0546 2.33 13.1 558
30-60 0.0407 2.90 9.7 694
>60 0.0424 2.38 10.1 569
60-70 0.0429 2.39 10.2 572
>70 0.0417 2.41 10 577 a Coefficients and constants shown for equations of the form BMR = weight coefficient x weight (kg) + constant b Henry, 2005
Table 20 Comparison of BMR prediction equations against the US DRI data seta
BMR PAL Accuracyb RMSE Bias%
Mean sd Mean sd % Mean Mean Min Max
kcal/d kcal/d
Reported 1524 300 1.71 0.29
Predicted values
Henry weight and heightc 1514 273 1.71 0.29 73 114 -0.32 -33 35
Henry weightc 1509 292 1.72 0.30 70 121 -0.32 -33 35
Schofield/FAOd 1531 278 1.69 0.28 69 123 1.33 -36 43 a IoM, 2005 b percentage of subjects predicted within ±10% of the RMR measured c Henry, 2005 d FAO, 2004
107
The Henry BMR prediction equations are based on the age bands 18-30 years, 264. 30-60 years and >60 years. The age bands used in this report do not exactly match those presented for the Henry BMR prediction equations and there is no Henry prediction equation for BMR for all ages. The following Henry BMR prediction equation age bands were therefore used to calculate predicted BMR for the age bands used in this report.
Table 21 Clarification of the Henry BMR prediction equation age bands applied in this report
SACN EAR age bands (years) Henrya BMR prediction equation age bands
19-24 18-30
25-34 18-30
35-44 30-60
45-54 30-60
55-64 30-60
65-74 >60
75+ >60
All Adults 30-60 aHenry, 2005
When creating BMR prediction equations, there is insufficient data to allow 265. equations to be generated for narrow age bands. Consequently, age bands tend to be wide and it is difficult to avoid the overlap between the age bands used for the Henry equations and those used in this SACN report, as seen in Table 21. The approach outlined in this table was considered reasonable because the BMR values estimated using the Henry age bands generally describe average BMR for each age band in this report.
In this report, one of the two large data sets of DLW measures of total energy 266. expenditure (TEE) (Subar, 2003) did not measure BMR values. As a result, it was necessary to estimate BMR values using the Henry equations and these BMR values were then used to extract PAL values as measured TEE/estimated BMR = PAL.
108
Appendix 5 – The physical activity level (PAL) and its use in the prediction of energy requirements
Theoretical aspects of calculation of PAL and its factorial prediction
PAL as an index of TEE adjusted for BMR The daily rate of total energy expenditure (TEE) for individuals or groups can be 267. expressed as a multiple of basal metabolic rate (BMR), which has been defined as the physical activity level (PAL). This allows prediction of TEE and consequent energy reference values as PAL x BMR, each of which represents a physiologically generalisable and predictable term. The introduction of the concept by FAO/ WHO/UNU in 1985 (WHO, 1985) was considered a simplifying approach to the determination of energy reference values. Thus BMR, a relatively fixed function of body composition, which is predictable as a function of weight, age and gender, is separated from all other components of TEE which are assumed to be variable. These other components reflect dietary intake, through the heat increment of feeding, miscellaneous thermogenic influences and lifestyle in terms of physical activity. Thus, in principle PAL is an index of TEE adjusted for BMR, which should mean it is independent of weight, age and gender. While most have embraced the concept, important reservations have been expressed (Carpenter et al., 1995; Goran, 2005) with preference given to an alternative approach based on multiple regression techniques to develop prediction models of TEE as a function of measured predictor variables such as body weight or age. These reservations need to be addressed.
Firstly, it has been implied (Goran, 2005) that this factorial approach is not 268. sufficiently evidence based. This presumably refers to the approach as introduced by FAO/WHO/UNU (WHO, 1985) when in the absence of an extensive database of measures of TEE, not only BMR but also PAL was predicted from time allocated calculations of physical activity ratio (PAR) values (activity cost/BMR). As discussed below (paragraphs 278-282), there are limitations with PAL values determined in this way but direct assessment of PAL from measured TEE and BMR is evidence based.
Secondly, it is argued that body weight predicts more of the variance in TEE 269. than BMR and consequently “the regression based approach provides a more physiologically appropriate model for TEE, as compared to the PAL approach which assumes TEE is composed of multiples of BMR” (Goran, 2005). Body weight could explain more of the variance in TEE than BMR when a) variability of physical activity is relatively small, and b) any influence of weight on physical activity is very marked. In fact, within the data sets examined here, there is little evidence
109
that these conditions apply. For the adult data set (Appendix 8), the regression of TEE on weight, height, age and gender has the same R2 value (0.62) as that for TEE on BMR (0.63), with BMR capturing all the variance in TEE associated with weight, height, age and gender in a factorial regression model. For adults in the DLW data sets published in the US DRI report (IoM, 2005), in a multiple regression with weight and BMR, the BMR coefficient (regression coefficient 0.71, p value = <0.0001) was more influential than weight (regression coefficient: 0.05, p = 0.18), which may indicate that BMR explained a larger proportion of the variance of TEE than weight. For children (Carpenter et al., 1995), although variance in physical activity energy expenditure (PAEE) is less marked than in adults, nevertheless within the US DRI data set for children, BMR explained slightly more of the variance in TEE than weight (regression coefficient = 0.59 for BMR and 0.37 for weight). Thus, it is not the case that body weight predicts more variance in TEE than BMR.
Since the difference between TEE and BMR is mainly PAEE, which will account for 270. the variance not explained by the BMR, TEE can be assumed to be composed of multiples of BMR, especially in adults but also in children. On this basis, regression models of TEE with body weight are arguably physiologically limited since they fail to account for an important between-individual source of variation in energy expenditure, which is physical activity energy expenditure (PAEE). Regression residuals indicate the range of between-individual variation in TEE and are highly correlated with PAEE and PAL but also include any variation in the BMR component of TEE as a function of the anthropometric variables. Within the adult DLW- determined TEE data set (see Appendix 8), the residuals of a regression of TEE on gender, weight, height and age account for much less of the variance in PAEE (TEE – BMR) in women (68%) compared with men (91%), when examined by linear regression. Given that there are no gender differences in PAL values, the most likely explanation of this relates to the gender differences in BMR as a function of weight, and the adequacy of partitioning differences in the BMR component of the TEE between men and women with a single term in the regression equation. This shows that whereas PAEE is a physiologically transparent quantity, the regression residuals are not. Furthermore, the absolute magnitude of the residual range is less than that of PAEE because residuals comprise only the difference between average PAEE predicted in the regression and individual values. Their maximum range as a fraction of maximum PAEE is 77%, which is a similar value to the slope of their regression on PAEE. This means that the residuals cannot strictly be used as an unambiguous measure of variance in PAEE in terms of their magnitude and distribution. As a result, the use of residuals in terms of predictors of the likely range of variation in PAEE within population groups is limited and they have not been used as such in any previous report.
Thirdly, a theoretical argument against the BMR multiple approach has been 271. presented (Goran, 2005) that: “the PAL model assumes a linear relationship between TEE and BMR that has a slope equivalent to PAL and a zero intercept” and that “the presence of significant and variable intercepts in the regression equations relating total energy expenditure to either BMR or weight invalidates the use of the traditionally used ratios (i.e. TEE/BMR or TEE/body mass) for expressing total energy expenditure data”.
110
The BMR multiple approach as used in this report makes no assumptions about 272. the regression relationships between TEE and BMR and such relationships are not strictly relevant to the discussion of the use of PAL within the factorial model of describing TEE. This is because it has never been suggested that PAL should be calculated as the slope of a linear regression of TEE on BMR for a population group. With information on TEE and BMR, individual subject PAL values can be calculated. From this, the distribution of PAL values for the population group can be calculated and the way in which PAL varies with age, body weight and any other demographic variable can be usefully examined. Indeed in the case of gender, which could influence TEE as some have suggested through gender differences in PAEE as well as through its known influence on BMR, a multiple regression approach to the role of gender on TEE would be unable to distinguish between these possibilities. Only the factorial approach based on measured BMR, measured TEE and calculated PAL allows any gender influences on PAL to be examined in a straightforward way. This has been the approach adopted for this report.
Much of the criticisms and confusion over the BMR multiple approach arise from 273. attempting to fit linear regressions to DLW data derived from studies of children. In this case, there does appear to be somewhat less variance in physical activity at any particular age or weight, and in addition, there is on average an increase in physical activity and consequent PAL values with age and weight. This means that the relationship between TEE and BMR is not linear. With small data sets analysed by linear regression the between-individual variation in PAEE will give varying slopes and intercepts which will be inversely correlated. It is therefore not surprising that this has been observed by Carpenter et al (1995). In the FAO/ WHO/UNU report (FAO, 2004), while a simple linear regression of TEE on weight was satisfactorily applied to infants in the first year of life, for children aged 1-18 years a quadratic polynomial regression equation of TEE on weight was derived from the DLW TEE data. PAL values were then extracted from this analysis by comparing TEE with BMR for each age group. These PAL values increased with age and the variance in PAEE at any age was expressed by calculating PAL values ±15% for children after the age of five years. This change with age in PAL is apparent in the data set of mean TEE, BMR and PAL values assembled in this report for children (see Appendix 12). There is an obvious and marked increase in PAL after the age of three years and then to a lesser extent in older children (Figure 12 Appendix 12) i.e. the plot of TEE on BMR is clearly curvilinear as PAL increases with age. Only when weight becomes a minor predictor of PAEE but explains most of the variance in BMR will the slope of TEE on BMR tend towards the mean PAL and the intercept tend towards zero. This is observed within the combined OPEN-Beltsville adult data set (Moshfegh et al., 2008; Tooze et al., 2007) assembled for this report. As shown in Figure 8 (Appendix 8), the intercept of the line of best fit of TEE on BMR is not significantly different from zero and the slope is similar to the median PAL.
The question of variation of PAL with body weight i.e. the influence of body weight 274. on the energy cost of specific activities is nevertheless an important issue that needs to be examined separately. When the use of PAL was introduced by FAO/ WHO/UNU (WHO, 1985) PAL values were estimated with factorial calculations
111
of time allocated energy costs of individual activities, expressed as PAR values. These were used to identify categories of activities based on lifestyles, so that those engaged in identifying energy requirements could make such calculations for specific population groups. To this end, lists of PAR values for activities are reproduced in the recent FAO/WHO/UNU report (FAO, 2004) (on the basis of a comprehensive review (Vaz et al., 2005). MET values for various activities are listed in the US DRI report (IoM, 2005). Thus, each of these reports assumes that having derived a PAL value for a particular lifestyle, it can apply equally to individuals regardless of body weight. Careful calorimetric studies have shown, however, that this is not strictly the case: i.e. PAR values increased with weight (Haggarty et al., 1997). The directly measured energy costs of a fixed programme of work expressed as a BMR multiple increased with body weight between 48 and 80kg by about 13% of the mean cost of the activity. This was consistent with theoretical calculations reported by the authors. The implications of this are that the energy requirements of adults engaged in similar tasks will be higher in larger compared with smaller adults. The authors showed that for a 40kg adult, the energy requirement calculated as PAL x BMR with PAL derived from measurements of PAR values for the specific activities in a 70kg adult will be an overestimation of about 10%.
The effect of body weight on PAR values for specific activities is only of practical 275. importance, however, when it is assumed that PAL can be calculated with some confidence from a factorial, time-allocated list of the PAR values. As discussed below (paragraphs 278-282), such predictions of PAL values are unlikely to be accurate and are not recommended in this report.
The final issue of general importance for this report is whether there is an 276. influence of gender on PAL. Gender differences in PAL and hence the predicted energy requirement was a feature of the 1985 FAO/WHO/UNU energy report (WHO, 1985) and the 1991 COMA DRV report (DH, 1991). Gender differences in behaviour are said to influence overall TEE, and therefore PAL values, for similar lifestyles or activities (Erlichman et al., 2002). However, most studies have failed to identify differences between men and women in PAEE or PAL, and both recent reports argue that average energy costs of activities expressed as a multiple of BMR, or PAR, should be similar for men and women (FAO, 2004; IoM, 2005). Within one meta-analysis of DLW-derived TEE values, PAL values were observed to be 11% lower in women than men, but it was not possible to identify whether this reflected an absence of subjects recruited from more active groups or a general tendency of women to be less involved in strenuous activity (Black et al., 1996). In the DLW data sets examined in this report, no differences in PAL with gender were identified for children or adults.
In summary, therefore, none of the concerns expressed above regarding the use of 277. the BMR-multiple approach are likely to be of importance within the framework of this SACN report. As developed below, estimates of PAL values for population groups are best derived from individual measurements of TEE and BMR and not by regression approaches. Within the adult data set assembled here, there is no evidence of any significant variation of PAL with either weight or gender (Appendix 8). For children, individual PAL values do show some increase with age and weight but
112
these are almost certainly behavioural changes with development that in no way detract from the validity of the use of the BMR-multiple approach.
Factorial prediction of PAL When the concept of PAL was introduced by FAO/WHO/UNU in 1985 (WHO, 278. 1985) it was implicit that PAL values could and would be estimated from the time allocated summation of individual activity energy costs expressed as PAR values. This formed the basis of the 1991 COMA report on DRVs (DH, 1991), and the principle is embodied within the FAO/WHO/UNU report (FAO, 2004) and the US DRI report (IoM, 2005). This approach is not adopted in this report for the following reasons.
Firstly, there is disagreement about how to use published PAR values for the 279. calculation of PAL. The FAO/WHO/UNU (FAO, 2004) used PAR values from a list compiled specifically for the report (Vaz et al., 2005). The authors of the US DRI report (IoM, 2005) express concern that factorial calculations from published PAR values may underestimate daily TEE because of a lack of specific inclusion of energy expenditure associated with feeding, the themic effect of feeding (TEF) and/or excessive post exercise oxygen consumption (EPOC). In the US DRI report (IoM, 2005), PAR values were derived from a list of MET values for specific activities (1MET = a fixed rate of oxygen consumption/kg). The MET values were converted to the BMR multiple PAR values (a ≈7% reduction) and then increased by 26.5% to account for TEF (10%) and EPOC (15%: i.e. increase by 1.15*1.10, see legend to Table 24). Thus, the factorial calculations in the US DRI report were made from values that were about 26.5% greater than those in the FAO/WHO/UNU report (FAO, 2004). Since there is no evidence that this is justifiable in all cases, some PAR values may be overestimated by up to 26.5%.
Secondly, even if PAR or 280. ΔPAL values are known with certainty, variation in spontaneous physical activity (SPA) (Snitker et al., 2001) (see paragraphs 224-227 Appendix 3), both within and between specific designated activities can introduce considerable error in factorial predictions of PAL. The concept of SPA or non exercise activity thermogenesis (NEAT) (Levine, 2007) is that energy expenditure is variable between individuals because of the variable expression of a behavioural phenotype associated with high levels of physical activity: i.e. SPA is an “inherent propensity to locomotion” (Snitker et al., 2001). This concept derived from observations of greater than expected variation in TEE during calorimeter studies where “workout” type activity is restricted. In these circumstances, activity is limited to daily living and postural changes yet PAL values ranged from 1.2-1.724, (Zurlo et al., 1992; Snitker et al., 2001) (see paragraph 283). In free-living circumstances the potential for such behaviour is more marked (Levine, 2007) and calorimeter- measured PAL values predict the range of higher free-living PAL values (Snitker et al., 2001). It is highly unlikely that subjects with a high SPA phenotype would ever exhibit the range of PAL values associated with the sedentary category of 1.40-1.69 identified by FAO/ WHO/UNU (FAO, 2004) even if they had the seated occupations assumed for this category.
113
Thirdly, some individuals may have what is to some extent the opposite of the 281. SPA phenotype, exhibiting marked compensatory reduced activity after periods of intense activity. Such a phenotype has not been widely investigated but has been reported in at least one study in which quite variable PAL values were observed in subjects engaged in highly controlled work activities (Haggarty et al., 1997). PAL varied markedly between 1.53 and 2.08 mainly because of marked differences in discretionary activities. Thus, subjects with the lowest DLW-determined PAL values (1.58, n=5) reverted to the basal state between-work periods exhibiting mean PAR values for discretionary energy expenditure of 1.02. The rest of the group exhibited discretionary PAR values averaging 1.89 (n=8) with higher average PAL values (1.95).
The existence of such behavioural phenotypes is increasingly being recognised 282. within the context of energy balance regulation (Levine et al., 1999; Zurlo et al., 1992), although their frequency within populations is currently unknown. Nevertheless, for such phenotypes, TEE is unpredictable within factorial models that have been applied to date. In other words, there may be a continuous spectrum of behavioural activities ranging from, on the one hand fidgeting while being otherwise stationary and choosing to walk up escalators, to on the other hand, adopting a resting posture whenever the possibility presents. The extent of this is indicated in studies reviewed below showing the degree of variation in individual PAL values within specific activity categories.
Observed variation in PAL within specific population groups Many published studies of PAL values of population subgroups only include mean 283. values, but where individual values are reported a much greater range of PAL values is often observed than would be anticipated. As already indicated (paragraph 280), 24 hour restricted activity studies within a calorimeter have repeatedly shown wide variation in PAL values: from 1. 15-1.7 in one report of 177 subjects (Ravussin et al., 1986) and from 1.20-1.65 in a second report (Snitker et al., 2001), in the latter case with individual values correlating with free-living DLW-measured PAL (1.35-2.15). PAL values of urban Chinese adults with manual or sedentary occupations classified into activity categories varied within these categories by up to 0.7 (Yao et al., 2002). Within a group of inactive men and women studied prior to a training programme the range of PAL values was 1.5-2.2 (Westerterp et al., 1992). In a study of older people investigating energy expenditure, self reported physical activity, and mortality, there was no relationship between self reported activity and PAL even though PAL predicted mortality (Manini et al., 2006).
The difficulty of predicting PAL values in terms of reported or measured physical 284. activity is exemplified by a study of men with sedentary occupations but varying levels of physical activity (Haggarty et al., 1994). Figure 5 shows a comparison of PAL with either measured activity time or activity categories predicted by the authors from the activity diaries. The relationship between measured PAL and recorded physical activity is only moderate. Thus, total active leisure time and activity level category explained only 41 and 48% of the variation in PAL. Subjects with a PAL ≥2, recorded leisure activities ranging from 35 to 229 minutes per day and were categorised into categories two to five. Within three of the five
114
categories individual PAL values varied by 0.6 and the difference in PAL between the lowest subjects in category five and the highest in category one was only 0.17 PAL units. The authors of this study emphasised the large differences that can occur between highly active individuals with a large exercise capacity compared with untrained individuals in the potential energy expended in tasks that are only loosely defined within activity diaries. This explains some of the poor correlation shown in Figure 5. Whatever the explanation for the discrepancy, it is clear that if the energy requirements of these subjects were individually predicted from their activity categories, there would be substantial errors.
Figure 5 – Relationship between PAL and measured leisure time activity and categorised activity levela
103
Deleted: <sp>
284. The difficulty of predicting PAL values in terms of reported or measured physical activity is exemplified by a study of men with sedentary occupations but varying levels of physical activity (Haggarty et al., 1994). Figure 5 shows a comparison of PAL with either measured activity time or activity categories predicted by the authors from the activity diaries. The relationship between measured PAL and recorded physical activity is only moderate. Thus, total active leisure time and activity level category explained only 41 and 48% of the variation in PAL. Subjects with a PAL ≥2, recorded leisure activities ranging from 35 to 229 minutes per day and were categorised into categories two to five. Within three of the five categories individual PAL values varied by 0.6 and the difference in PAL between the lowest subjects in category five and the highest in category one was only 0.17 PAL units. The authors of this study emphasised the large differences that can occur between highly active individuals with a large exercise capacity compared with untrained individuals in the potential energy expended in tasks that are only loosely defined within activity diaries. This explains some of the poor correlation shown in Figure 5. Whatever the explanation for the discrepancy, it is clear that if the energy requirements of these subjects were individually predicted from their activity categories, there would be substantial errors.
Figure 5. Relationship between PAL and measured leisure time activity and categorised activity levela
aHaggarty et al., 1994
285. Within a data set of DLW studies in healthy adults assembled for this report, all of the studies which reported PAL values (n=63) were examined for descriptions of the activities/lifestyles of the subjects. Just over half the studies (n=33) provided no information, but for those that did (n=30), the mean PAL values were assigned to categories of light, moderate or heavy activity, with those studies with a mixture of activities assigned to moderate. The values in Table 22 show that, notwithstanding the somewhat arbitrary assignment to activity categories, the range of mean study PAL values is considerable, especially in the “light” category (1.23-1.98).
Measured and predicted energy expenditure in men
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
0 1 2 3 4 5 Activity level category
P A
L (f
ro m
D LW
T E
E )
Measured energy expenditure and total active leisure time in men
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
0 50 100 150 200 250 Total active leisure (min/d)
P A
L (f
ro m
D LW
T E
E )
aHaggarty et al., 1994
Within a data set of DLW studies in healthy adults assembled for this report, all of 285. the studies which reported PAL values (n=63) were examined for descriptions of the activities/lifestyles of the subjects. Just over half the studies (n=33) provided no information, but for those that did (n=30), the mean PAL values were assigned to categories of light, moderate or heavy activity, with those studies with a mixture of activities assigned to moderate. The values in Table 22 show that, notwithstanding the somewhat arbitrary assignment to activity categories, the range of mean study PAL values is considerable, especially in the “light” category (1.23-1.98).
Table 22 Mean PAL values from DLW studies assigned (where possible) to activity groups
Number of studies Activity categorya PAL value
mean range
33 none 1.73 1.42-1.97
15 light 1.67 1.23-1.98
12 moderate 1.84 1.63-2.01
3 heavy 2.09 1.91-2.48
63 all 1.75 1.23-2.48
a Activity categories assigned based on written comments by study authors
115
Although an exhaustive review has not been conducted for this report, it is clear 286. that considerable variation in overall rates of TEE and PAL occur within subjects who would be classified as exhibiting similar lifestyles. As the values shown in Table 22 are mean study values, the range of individual values will, in fact, be much wider. The extent of inter-individual variability in PAL is hard to judge, however, because so few studies report individual PAL values within occupational or lifestyle categories. Thus, with the exception of groups at the extreme of the range (either very restricted or with high levels of physical activity), the relationship between PAL and lifestyle, or even self-reported physical activity, appears to be weak. Consequently, predictions of PAL values are unlikely to be accurate and are not recommended in this report.
Magnitude and variation in PAL within the general population
Observed lower and upper limits of PAL The FAO/WHO/UNU 1985 report (WHO, 1985) identified the lower limit of PAL, a 287. ‘survival’ value, to be 1.27, which is consistent with studies assembled elsewhere in non-ambulatory chair-bound and non-exercising subjects confined to a calorimeter (where PAL values of 1.17-1.27 were observed) (Black, 1996).
The lower limit of energy expenditure in subjects who are ambulatory but 288. only exhibiting the minimal activities associated with daily living (e.g. grooming, showering and dressing), equates to a PAL value between 1.35 and 1.4 (Alfonzo- Gonzalez et al., 2004; Goran et al., 1994b; Warwick, 2006). In grossly obese subjects confined to a whole-body calorimeter following a standardised sedentary protocol apart from two 30 minute exercise periods (30 minutes of cycling at 25W and 30 minutes of stepping on and off a 20-cm block at a rate 40/minute), mean PAL values were 1.35 (1.27-1.42) (Gibney et al., 2003). In the healthy, most elderly, the mean PAL value (after trimming for PAL values <1.1) was 1.38 (Rothenberg et al., 2000).
The US DRI report compiled a list of activities of daily living, which accounts 289. for about four hours in total and amounts to a ΔPAL of 0.29 (IoM, 2005). This is reported to equate to a sedentary PAL of 1.39, i.e. 1+0.1(TEF) +0.29. Such calculations, however, can only be relevant for subjects in the basal state for 20 hours per day. Further, the listed activities may overestimate actual costs since they include additions of energy expenditure (≈28%) to allow for corrections in TEF and EPOC.
Overall, a PAL value of 1.38 seems the likely lower limit for free-living individuals. 290. This indicates that the upper limit of the sedentary PAL range (1.4) suggested in the US DRI report is too low.
The upper limit to human physical activity is that exhibited for limited periods 291. of time by elite endurance athletes and soldiers on field exercises, for whom PAL values between 3 and 4.7 have been reported (Black et al., 1996; Hoyt & Friedl, 2006). The maximum PAL value associated with a sustainable lifestyle within the general population, however, appears to be about 2.5 (Black et al., 1996; Westerterp
116
& Plasqui, 2004). This value has been supported by subsequent studies in long- term exercising women (Withers et al., 1998) and physically active men (Black et al., 1996; Davidson et al., 1997; Haggarty et al., 1994; Westerterp & Plasqui, 2004).
Observed magnitude and variation in PAL Several large data sets of TEE measures containing individual PAL values have been 292. made available for this SACN report. These include the US DRI data set (IoM, 2005), the Beltsville study (Moshfegh et al., 2008) and the OPEN study (Subar et al., 2003; Tooze et al., 2007). Although the OPEN study did not measure BMR, this has been estimated using the Henry BMR prediction equations based on weight and height (Henry 2005). The distribution statistics for these three large data sets are shown below in Table 23. Details of the OPEN and Beltsville cohorts are given in Appendix 8.
The US DRI data set (IoM, 2005) of adults aged 18 or more years (n=767, 360 293. overweight or obese and 407 normal weight) includes the data in the Black et al. (1996) analysis. There is a fall in PAL with age, most notably in terms of lower values in the small sample at ages greater than 80 years for both the normal and overweight/obese populations. Gender differences in mean PAL values are only apparent within the overweight/obese group, but the difference is small. The characteristics of this data set shown in Table 23 are those after trimming to exclude the very old (aged >80 years). The median PAL (1.72) is higher than the median PAL values for the OPEN and Beltsville studies. The US DRI data set is, however, a collection of many individual studies with much smaller sample sizes, and cannot be considered to be representative of a normal adult population. There may be an over representation of subjects with PAL values >2, i.e. very physically active subjects, as indicated by the 90th centile value of 2.10 compared with 1.96 and 1.92 for the Beltsville and OPEN studies, respectively.
Table 23 Distribution of PAL values of large data setsa,b,c
Distribution boundaries US DRIa
(n=724; age 18-80y) OPENb
(n=451; age 40-69y) Beltsvillec
(n=478; age 30-69y)
10th centile 1.38 1.40 1.32
lower quartile 1.55 1.49 1.46
median 1.72 1.61 1.62
upper quartile 1.92 1.77 1.78
90th centile 2.10 1.92 1.96 a IoM, 2005 b OPEN data set (Subar et al., 2003; Tooze et al., 2007) c Beltsville data set (Moshfegh et al, 2008)
In the OPEN study (Subar 294. et al., 2003), mean PAL values fall slightly with age (from 1.70 at 40 years to 1.57 at 70 years), are slightly higher for women than men (by 0.04 units), but are independent of weight or BMI. The distribution pattern of PAL values is clearly shifted downwards compared with the US DRI data set values.
117
In the Beltsville study (Moshfegh 295. et al., 2008), there is little change in PAL with age or with BMI (by regression or with category). Thus, the possibility that the energy cost of physical activity may be affected by adiposity (Byrne et al., 2005; Forsum et al., 2006; Leenders et al., 2001) does not seem to influence the range of PAL values within normal and obese subjects. The distribution is similar to the OPEN study for the lower and upper quartiles and median, which, to some extent, lends confidence to the calculation of BMR in the OPEN data. For the Beltsville study, however, the 10th centile PAL value is slightly lower (1.32) and the 90th centile PAL value is slightly higher (1.96), than in the OPEN study (1.40 and 1.92).
The characteristics of PAL values within the combined OPEN-Beltsville cohort are 296. shown in Appendix 8.
Observed effect of additional physical activity on PAL The energy cost of additional amounts of sport or strenuous leisure activity has 297. been considered in terms of ΔPAL values. One widely quoted meta-analysis of DLW studies (Black et al., 1996) reported that mean PAL values increase from 1.63 to 1.99 with imposed activity (physical training) on a low activity background.
Adults with normally sedentary occupations who did not exercise or play sport on 298. a regular basis increased their PAL value by 0.41 (from 1.59 with a range 1.43-1.68, to 1.99 with a range 1.66-2.42) with a nine week incremental programme of jogging up to one hour/day, five days/week (Bingham et al., 1989). Inactive men and women following a 40 week incremental running programme sufficient to enable running a half marathon, increased their mean PAL values by 0.44 PAL units above their initial range of 1.5-2.2 (Arthur et al., 1987): obese boys achieved a mean increase of 0.27 with a four week training program of cycling for 45 minutes five times/week at 50-60% of VO
2 max (i.e. from 1.77 (±0.15) to 2.04 (±0.15) (Blaak et al., 1992). The
implications of these studies in the light of theoretical calculations of the likely response of PAL to exercise are considered at the end of this section.
Individuals involved in competitive sport or who have habitual high levels of 299. physical activity exhibit PAL values up to 0.6 units higher than those who do not take regular exercise. In a small number of healthy men, there was a mean difference in PAL of 0.64 between competitive runners (PAL = 2.26) and those reporting no leisure activity (Davidson et al., 1997). In women aged 50-70 years there was a mean difference in PAL of 0.61 between long-term exercisers (PAL= 2.48: range 1.60–3.43) or long-term non-exercisers, (PAL= 1.87: range 1.63–2.25) (Withers et al., 1998).
Predicting the effect of additional physical activity on PAL The US DRI report (IoM, 2005) lists the energy cost of various activities in terms of 300. METs and ΔPAL values. Selected examples, including those for walking, are given in Table 24, together with actual additional rates of energy expenditure associated with the activities calculated for a standard woman (57kg age 30-60 years) and man (70kg age 30-60 years) (Henry, 2005). As indicated above, such ΔPAL values are calculated from METs after adjustment of the equivalent PAR value to include an additional 26.5% to account for TEF and EPOC. The calculations are similar but
118
not exactly the same (see footnote b of table 24) and could be overestimates if the MET values were not obtained in subjects in the basal state. They will be higher than equivalent values quoted by FAO/WHO/UNU (FAO, 2004) and others (Vaz et al., 2005). The ΔPAL values are the increases in the daily PAL expected when the one hour of the activity (mean PAR-1/24) replaces the BMR. The ΔPAL values shown in Table 24 are slightly lower than those in the US DRI report (IoM, 2005) because the reference BMR values are slightly different and the additions for EPOC and TEF are made a little differently (x 1.15 x 1.10 =+26.5% in this report and x 1.15 x 1/0.9=+27.8% in the US DRI report). The PAR values can be compared with directly measured values in men with light occupations and variable leisure activities while walking at moderate pace (2.54), walking briskly or carrying a load (4.09) and jogging or running (13.1) (Haggarty et al., 1994). By and large the values compare well.
Table 24 Influence of activities on PAL
Activity
METsa
PARb
ΔPAL/10
min kJ(kcal)/10 min ΔPAL /h kJ(kcal)/h
woman man woman man
Walking (2mph) 2.5 2.9 0.013 71(17) 91(22) 0.08 436(104) 559(134)
Walking (3 mph) 3.3 3.9 0.02 109(26) 140(33) 0.12 654(157) 839(201)
Walking (4mph) 4.5 5.3 0.03 164(39) 210(50) 0.18 981(235) 1258(301)
Tennis (doubles) 5 5.9 0.034 185(44) 238(57) 0.2 1091(261) 1398(335)
Dancing 6 7.1 0.042 229(55) 294(70) 0.25 1363(326) 1748(418)
Roller Skating 6.5 7.7 0.046 251(60) 322(77) 0.28 1527(365) 1957(469)
Swimming 7 8.2 0.05 273(65) 350(84) 0.3 1636(392) 2097(502)
Walking (5 mph) 8 9.4 0.058 316(76) 405(97) 0.35 1908(457) 2447(586)
Jogging (6 mph) 10 12 0.08 436(104) 559(134) 0.46 2508(600) 3216(770)
Rope skipping 12 14.1 0.091 496(119) 636(152) 0.55 2999(718) 3845(920)
Squash 12 14.1 0.091 496(119) 636(152) 0.55 2999(708) 3845(920) a 1 MET =0.0175 kcal/minute/kg b Energy expenditure as a multiple of the BMR, calculated from METs as follows.
1. Conversion to PAR BMR/multiple values: BMR is calculated as mean BMR value for reference women (57kg age 30-60 years)=0.0159kcal/min/kg), and reference man (70kg age 30-60 years) =0.0166kcal/min/kg: Henry BMR prediction equations (Henry, 2005). The resulting PAR value is 7% lower than the MET value.
2. Addition of 26.5% for EPOC and the thermic effect of feeding. Note that ΔPAL values are slightly lower than those in the DRI report because the reference BMR values are slightly different and because the additions for EPOC and TEF are made slightly differently (x 1.15 x 1.10 =+26.5% here and x 1.15 x 1/0.9=+27.8% DRI report (IoM, 2005)).
Overall these changes mean that the PAR value is 17.7% greater than the MET value.
119
It is important to recognise that the overall effect of an additional activity at some 301. fixed rate will have a variable effect on the PAL value according to the magnitude of energy expenditure which it replaces, even if no other changes in energy expenditure occur.
In the US DRI report (IoM, 2005) the factorial calculations of PAL values from MET 302. and ΔPAL values for various activities, such as those shown in Table 24 (e.g. walking (4mph) ≡ 0.18 ΔPAL) involve an assumption that each activity replaces the BMR. This implies in turn that its effect on PAL will be the same for all individuals. In practice, much of the non-sleeping time energy expenditure is greater than BMR and may even be much higher for those exhibiting a high SPA phenotype. The effect of an additional activity, such as one hour/day walking at 4mph, will be to replace activity likely to be at a higher rate than the BMR and will, therefore, result in a lower increase in PAL. Thus, individuals with a low rate of background discretionary activity will experience a higher overall increase in PAL for the extra activity, than those with a higher background rate.
This is shown in303. Table 25, where one hours walking at 4 and 5 mph or jogging 10 minute miles (ΔPAL values of 0.18, 0.35 and 0.46, PAR values of 5.3, 9.4 or 12.0), replace one hours activity in subjects exhibiting PAL values of 1.5, 1.6 or 1.7. Such PAL values in subjects sleeping for eight hours a day imply average discretionary PAR values of 1.75, 1.9 and 2.05. Substitution of an average hour of discretionary activity with one hour of the increased physical activity will result in actual new PAL and ΔPAL values which are 10-25% less than the listed ΔPAL value for the activity. For high SPA phenotypes additional planned activity may have little effect on overall energy expenditure and PAL. Only if the new activity replaced a period of complete inactivity (PAR=1) would the expected increase be observed. If the additional activity resulted in a period of compensatory reduced activity as discussed above, the overall effect would be even less than indicated. These calculations show the difficulty of calculating a planned change in energy expenditure. In practice, however, because the variation in the magnitude of ΔPAL with mean discretionary PAR is small, and given the insecurities involved in such calculations, this uncertainty is not considered in this report.
Table 25 Influence of specific additional activities on PAL in subjects with varying initial PAL values
1 hours additional planned activity
4mph walking 5mph walking 10min mile jogging
PARa 5.3 9.4 12.0
ΔPALb 0.18 0.35 0.46
Initial PAL Mean discretionary PARc new PALd (actual ΔPALe)
1.5 1.75 1.65 (0.148) 1.82 (0.320) 1.93 (0.427)
1.6 1.9 1.74 (0.142) 1.91 (0.313) 2.02 (0.421)
1.7 2.05 1.84 (0.135) 2.01 (0.307) 2.11 (0.415) a Values from Table 24 b = (PAR-1)/24: assuming the activity replaces a period at the BMR c Assuming 8 hrs at PAR =1(sleeping) d Calculated as the1hr activity replacing 1 hr at the mean discretionary rate e New PAL-initial PAL
120
Overall, these theoretical calculations indicate that PAL can be increased by about 304. 0.2 for one hours brisk walking (with brisk defined as slightly faster than 4mph), about 0.4 by one hours jogging at 6 mph, and up to about 0.6 by an intense aerobic exercise programme associated with training for competitive sport. This allows a re-examination of a widely quoted statement derived from a DLW meta-analysis (Black, 1996; Black et al., 1996) that 30-60 minutes of active sport 4-5 times per week can raise PAL by about 0.3 units. On the basis of the ΔPAL values in Table 25 and if ‘active sport’ is equivalent to jogging at ΔPAL per hour of about 0.43, then 30 minutes a day for four days a week and 60 minutes a day for five days a week would raise PAL on average by about 0.12 and 0.31, respectively. Thus, a more accurate statement would be that 60 minutes of active sport five times per week can raise PAL by about 0.3 units and this is generally consistent with the observed effects discussed in paragraphs 297-299 above.
PAL values in relation to health outcomes The US DRI report discusses a ‘Physical Activity Level consistent with a normal Body 305. Mass Index’ in terms of one hour of moderately intensive physical activity (walking at 4 miles per hour) resulting in an increase of 0.2 PAL units (IoM, 2005). The FAO/ WHO/UNU report identified a desirable PAL value of 1.75 or more (FAO, 2004).
Information on the relationship between PAL and mortality has been published in 306. a prospective study of healthy older adults (n=302; aged 70 to 79 years) (Manini et al., 2006). TEE was determined using the DLW technique and BMR by indirect calorimetry. Over an average of 6.15 years of follow-up, participants in the upper tertile of PAEE (PAL greater than 1.78) had a reduced risk of all-cause mortality (HR 0.43, 95% CI 0.21-0.88; P
trend = 0.02) compared to those in the lowest tertile
(PAL less than 1.57). Thus, this objectively measured free-living PAEE was strongly associated with lower risk of all-cause mortality in these healthy older adults. The published Kaplan-Meier Survival Plots indicate that the separation of survival statistics between the tertiles did not occur until after two years following the initial measures. This suggests that the increased mortality in the lowest PAL tertile was not due to reverse causality (i.e. subjects exhibiting low PAL values because they were ill and at increased risk of mortality), but this cannot be ruled out. Although the intensity and type of physical activity was not objectively measured, physical activity questionnaires suggested that the proportion of individuals who reported high-intensity exercise and walking for exercise, in terms of both duration and intensity, was similar across tertiles of free-living activity energy expenditure. This implies that simply expending energy through any activity may influence survival in older adults and that specific, high intensity exercise per se may not be required to produce health benefits.
Utilising PAL values to determine reference energy intakes Notwithstanding the uncertainties of predicting PAL for population groups, 307. it remains the case that all previous dietary energy recommendations have recognised the need to include a variable physical activity factor in the derivation
121
of energy requirements for school-aged children and adults. Although criticisms have been levelled at the BMR x PAL approach, no satisfactory alternative has been identified. There are two major considerations in the practical application of this approach: identifying suitable PAL values appropriate for groups and populations; and utilising this information to derive energy reference values.
In the past, starting with the FAO/WHO/UNU report (WHO, 1985), a range of PAL 308. values has been determined by factorial calculations which equate to occupation and lifestyle. This has been with the intention of providing guidance to the health- care professional using the reports in making a judgement about the appropriate PAL value for individuals and population groups. The 1991 COMA DRV report (DH, 1991) required a choice from a 3 x 3 matrix of PAL values (three occupations, three leisure activities), for each gender over a range of 1.4-1.9 for men and 1.4-1.7 for women. In addition, a table of EAR values calculated for nine PAL values from 1.4 to 2.2 is was given. The FAO/WHO/UNU report (FAO, 2004) classified the intensity of a population’s habitual physical activity into three categories identified by a range of PAL values for each category: sedentary or light activity 1.40-1.69, active or moderately active 1.70-1.99, vigorous 2.00-2.40, with worked examples using the midpoint of these three ranges. The report then listed tables of daily average energy requirements calculated for six PAL values (1.45, 1.60, 1.75, 1.90, 2.05 and 2.20) with further worked examples for individuals or groups with PAL values not included in the list (e.g. PAL =1.8). The implication of both the COMA (DH, 1991) and FAO/WHO/UNU (FAO, 2004) reports is that it is possible to utilise the information on PAR values for activities to predict PAL values for specific population groups to within 0.05-0.1 of a PAL value.
The US DRI report (IoM, 2005) included physical activity factors in the prediction 309. equations for TEE for men and women based on weight, height, age and the physical activity coefficient (PA)33.
The physical activity (PA) coefficient, which scales the weight and height factors, 310. results in the equation containing a form of PAL x BMR, since the height and weight factors should capture the BMR data (in fact, calculations with the published DLW data set within the report, show that the prediction equations calculate the same values for TEE as PAL x estimated BMR). The PA coefficient is not PAL per se, but constants assigned to each of the four PA categories (sedentary, low active, active, or very active). For men these were: 1.0, 1.11, 1.25 or 1.48; and for women: 1.0, 1.12, 1.27, or 1.45. In effect, this results in four parallel prediction equations for each gender, one for each PAL category. The user of the report must choose the appropriate PA coefficient by locating the individual or population group under consideration within one of four PAL ranges. It is not clear how these PAL ranges have been derived, but factorial calculations of PAL are derived in the report from lists of ΔPAL values for various activities. The examples shown are for PAL values of 1.39 (sedentary), 1.49 (low active), 1.75 and 1.77 (active) and 2.06 (very active). Assuming
33 PA is the physical activity coefficient, which depends on whether the individual is estimated to be in the sedentary (PAL ≥ 1.0-1.4), low-active (PAL ≥ 1.4<1.6), active (PAL ≥ 1.6<1.9) or very active (PAL ≥ 1.9 < 2.5) PAL categories.
122
that the sedentary PA factor of 1.0 is equivalent to a PAL of 1.39, then the other PA factors in the equation are equivalent to PAL values of 1.54, 1.74 and 2.06 (men) and 1.56, 1.77 and 2.02 (women), which generally correspond to the worked examples.
The user of the US DRI report (IoM, 2005) is required only to locate an individual 311. within a PAL range rather than a precise PAL value, and consequently this makes it, to some extent, simpler than FAO/WHO/UNU (FAO, 2004) procedure. Indeed, with a lower limit of PAL for free-living individuals of about 1.38 (see paragraph 290), few subjects in any population would be assigned to the sedentary category leaving the choice to be made between one of the other three categories. In this SACN report, the difficulty in choosing between adjacent PAL categories has been highlighted. Miscategorisation between adjacent categories would result in predicted TEE, and hence energy reference values, being 13-18% too high or low for subjects in the middle of each category, or up to 40% for individuals at the top or bottom of ranges.
It is clear from the above discussion that the inter-individual variability of PAL 312. is such that it is an unrealistic expectation that PAL values can be predicted for specific lifestyle-dependent population groups with the precision implicit in any of these previous reports (FAO, 2004; IoM, 2005). Furthermore, little can be said about the health implications of specific PAL values apart from the general desirability for values to be within the higher rather than lower range of values. Nevertheless, the definition of TEE and the energy requirement for a population group on the basis of PAL x BMR remains a sound principle and an alternative approach to its use is therefore required.
The simplest approach is to decide that TEE for a population group can only be 313. predicted on the basis of direct measurements of TEE for reference populations from which average PAL values and their distribution can be identified. Thus, medians and ranges of PAL values observed in the reference populations can be applied to similar population groups in terms of age, BMI and gender, assuming only that reference population groups are appropriate. TEE is then predicted as a function of the BMR. Dietary reference values can then be framed against such distributions. Such framing could involve an average (median) reference intake for the population together with lower (e.g. 25th centile) and higher (e.g. 75th centile) values identified for those representing the less active or more active sections of the population and with additional amounts of energy likely to be needed for lifestyle changes associated with additional activities. Identifying objective measures of activity allowing assignment of populations to the less or more active sections is an important research task.
Such advice is clearly a major departure from previous approaches for adults 314. although it is in principle similar to the method adopted by FAO/WHO/UNU (FAO, 2004) in their recommendations for children. However, it is an approach which recognises the reality that prediction of rates of TEE for individuals or population groups is inherently uncertain.
123
Appendix 6 – Doubly labelled water (DLW) method
In 2009, the International Atomic Energy Agency (IAEA) published information on 315. the theoretical background, as well as the practical application of state of the art methodologies, to monitor total energy expenditure (TEE) using stable isotopes (and changes in body composition). IAEA reviewed recent advances in analytical techniques developed by an international group of experts which readers are referred for more detailed information (IAEA, 2009). The doubly labelled water (DLW) method is a minimally invasive stable isotopic technique of measuring carbon dioxide (CO
2 ) production in free-living subjects over a period of several
weeks. The subject drinks a weighed amount of the DLW containing known amounts of the stable isotopes of hydrogen (2H) and oxygen (18O
2 ), based on their
body weight. The isotopically labelled water equilibrates with normal body water and a sample is taken typically after about five hours to measure the initial isotope enrichment, which also indicates the size of the total body water pool from isotope dilution. As water is lost from the body in urine, sweat and evaporation from the lungs during normal water turnover, the labelled water containing 2H and 18O
2 is lost. However, 18O
2 in water exchanges with the oxygen in CO
2 because of
the carbonic anhydrase reaction. This means that CO 2 excretion will also result
in an additional loss of 18O 2 as C18O
2 . This means that 18O
2 leaves the body faster
than 2H, the difference being proportional to CO 2 production. Loss of the two
isotopes from body water is assessed by measurement of the rate of decline in concentration of the isotope in a sample of the subject’s urine or saliva, collected during the study period, and measured by isotope ratio mass spectrometry. The difference between the elimination rates of the two isotopes reflects the rate at which CO
2 is produced from metabolism. TEE can then be estimated from the
CO 2 production rate after assigning an energy value to CO
2 calculated from the
assumed average respiratory quotient (RQ) value (ratio of CO 2 produced to the
O 2 consumed), which is determined by the balance of macronutrients oxidised
during the period. This in turn is assumed to reflect the composition of the dietary intake.
The accuracy and precision of the DLW method for measuring TEE is influenced by 316. isotope fractionation during evaporative water loss and CO
2 excretion. Corrections
for isotopic fractionation of water lost in breath and (non-sweat) transcutaneous loss need to be made when using labelled water to measure water turnover or CO
2 production (Schoeller et al., 1986). The technique, therefore, is based on
assumptions about the amount of water lost from the body by evaporation and the extent of incorporation of 2H and 18O
2 into body tissues, especially during
growth. This technique, however, provides an indirect measure of TEE and is the most accurate available measure in free-living subjects. The estimated TEE is the energy expended during a time period, including the energy required for tissue synthesis, but does not include the energy content of tissue laid down (growth,
124
pregnancy, weight gain) or milk produced during lactation; these are estimated from analysis of tissue deposition and milk secretion.
Methodology critique Although the DLW method has become established as the method of choice for 317. the estimation of free-living TEE, uncertainties exist about its application and, consequently, in the interpretation of published DLW studies. As this SACN report relies almost entirely on information on TEE determined by the DLW method, some of the important issues in relation to the reliability of DLW studies are briefly discussed.
The application of the DLW technique has not been standardised, with different 318. approaches taken in both laboratory isotopic analysis and in the experimental design of the measurement of 18O
2 and 2H
2 turnover in body water. The three most
widely used approaches are the ‘slope-intercept’, ‘2-point’, and ‘modified’ methods of calculation. This lack of standardisation means its potential high precision (the within-individual coefficient of variation (CV) for the validation of TEE from DLW against respiratory gas exchange) is not always realised. A double-blind between-laboratory variability study identified substantial between-laboratory variability in results, in some cases with precision as low as 35% and with some reporting physiologically impossible results (Roberts et al., 1995a). This seems to reflect analytical error rather than methods of calculation or the approach used to assess isotope decay rates. The different approaches are in principle equally valid although some reviewers suggest that collecting samples repeatedly over the measurement period rather than by collecting them only before and after the measurement period may decrease the error to a small degree. Of the data sets assembled for this report, the adult values from the Beltsville study (Moshfegh et al., 2008) involved the multipoint slope-intercept approach while the OPEN study (Trabulsi et al., 2003) involved the 2-point approach. However, the data set assembled for children and adolescents involves reports from many different laboratories so that between-investigator errors could represent cause for concern in our identification of energy reference values for these population groups.
Estimates of measurement precision vary between investigators.319. Not all investigators document error terms for the dose, background determinations, or uncertainties in fractionated evaporative water loss and RQ. A change in background 18O
2 enrichment in water during the study resulting from travel or a
change in the dietary water source could affect the final CO 2 production rate,
especially in subjects where very low isotopic dosing is used. Observed variation in repeated measurement studies will reflect both methodological error as well as actual variation in TEE due to a change in behaviour, making interpretation difficult. In one study, careful duplicate measurement of TEE in six adult women at a six month interval indicated that the within-subject CV for TEE was 7.8%, most of which was estimated by the investigators to be physiological variation (6.4%) due to variation in activity (Schoeller & Hnilicka, 1996). They also reviewed 16 studies with at least two DLW measurements, which indicated the reliability of the method to be 7.8%, except under conditions of high water flux.
125
Within the OPEN study, repeated isotopic analysis and repeated DLW measurement 320. in a subset of 25 subjects identified an overall CV of the TEE measurement of 5.1%, of which 2.9% was due to analytical variation and 4.2% was due to within- individual physiological variation (Trabulsi et al., 2003). Others have identified a value for analytical variation of 4% (Elia et al., 2000). Within the Beltsville study, repeat DLW measurements were conducted in a subset of 32 subjects and an overall CV of 12.6% was identified. As the repeat measures were more than one year after the initial measurements (unlike the OPEN subset study which compared repeat measures after two weeks) it might be expected that variation would be larger. Some of the individuals in the Beltsville subset study (Moshfegh et al., 2008) exhibited TEE in the second study of only 50% of the first. This highlights the difficulty of identifying rates of TEE within free-living populations of relatively small sample sizes.
The application of the DLW approach usually involves 7-14 day measurements; 321. however it is unknown if this is representative of TEE in the longer term. The Beltsville subset study points to potential errors in this assumption e.g. the R2 of the second measure compared with the first was only 0.35. Individuals are likely to change their physical activity with time, due to differences in season for example, resulting in potentially considerable within-person variability. Concerns have been expressed about this issue (Willett, 2003) and about the design of the OPEN Study due to the within-person CV of TEE (5.1%). This value appears too low when compared with values obtained from a quantitative review of the reproducibility of TEE measured by DLW in 25 studies with repeated measurements (Black & Cole, 2000). In this latter review, estimates of 8% for within-subject variation in DLW measurements were reported. This estimate included analytical errors plus inherent within-subject biological variation in TEE due to changes in weight, season and physical activity. This biological variation increased as might be expected with increased time between measurements to about 15% at a time span of 12 months. The authors of the OPEN study (Kipnis et al., 2003) subsequently reanalysed the data from studies examined in the review and found that for studies of only free- living subjects (as in the OPEN study) within-person variation did not increase with time. The Beltsville study, however, shows a large increase in the CV when replicate studies are conducted over a longer time period.
Another potential concern is the influence of growth on measurements due to 322. sequestration of isotope within body tissues during the study. Deuterium can be incorporated into tissues during reductive biosynthesis especially of fat and cholesterol. Such sequestration would decrease the difference in 2H
2 and 18O
2 decay
rates and lead to an under-estimation of CO 2 production and TEE. The extent to
which this is a problem is a difficult question to resolve. Technical problems in assessing the relative influences of isotope sequestration and isotope fractionated water loss, which have opposite influences on the estimation of TEE, make the effect on TEE measures in rapidly growing infants difficult to assess. A consensus review (NAHRES-4, 1990) concluded that under extreme anabolic conditions and using pessimistic assumptions regarding de novo fat synthesis, the maximum error in estimation of TEE due to 2H sequestration could be 5% but that “it seems unlikely that the error would be as high as this under many circumstances”.
126
The extent to which increasing body fat can influence the method is also an issue. 323. One study in obese and lean subjects identified an underestimation of TEE by DLW of 0.285 MJ/day for each additional 10 kg of fat (Ravussin et al., 1991). Others, however, have failed to observe such an effect (Gibney et al., 2003).
Another potential problem is recruitment bias. Subject selection in any study 324. requires subjects to either volunteer or to agree to participate when randomly approached and this raises the possibility of the healthy volunteer effect with study subjects atypical of the general population, in this case more physically active with higher than average rates of TEE. In a validation study for NDNS (Ruston et al., 2004), DLW studies were conducted in a small adult population (n=66) indicating mean PAL values of 1.74 (range 1.36-2.2) for women and 1.88 (range 1.37-2.50) for men. These values are on average higher than those of the two large population studies (i.e. OPEN and Beltsville) used in this report to identify adult population PAL values. This may be because the NDNS DLW subsample exhibited the healthy volunteer effect, or that the men in the NDNS DLW sub-sample were heavier than those in the OPEN-Beltsville data set and in the NDNS main study, or it may simply reflect the smaller cohort size. For the OPEN study, volunteers were derived from a random sample of 5000 households so the healthy volunteer effect is likely to be less than in the Beltsville study in which subjects were recruited through advertisements and letters of invitation. The fact that the distribution of TEE was so similar in these two studies would tend to suggest that recruitment bias was not a significant factor.
Summary It is likely that recent DLW studies have benefited from the experience gained 325. with the method over 25 years of its use. The data set of DLW studies used to derive the EAR for children and adolescents in this SACN report, however, includes a wide range of studies assembled over many years and by many investigators. Although technical problems can be minimised with careful investigators, some caution must be used in examining this data set.
More important, however, is the difficulty posed by true, within-individual 326. variation in physical activity and consequent TEE. To illustrate this, the repeat DLW assessments of TEE conducted more than one year after the first measurement in the Beltsville study are given below (see Table 26) (Moshfegh et al., 2008). The within-subject average CV for TEE of 12.6% was just over half the between- subject average value of 22.8% for the larger whole cohort. Some of this latter figure reflects between-subject variation in TEE with size due to variation in basal metabolic rate (BMR). After adjusting for BMR and calculating PAL there is a lower CV of 15.4%. During the within-subject repeat measurements of TEE, the subjects were generally weight stable and the BMR would not be expected to have changed. This means that most of the variance must reflect change in physical activity. The within-subjects CV and the CV of the between-subject PAL values are similar: 12.6 % compared with 15.4%. This indicates the difficulty of identifying energy requirements with any certainty for individuals and small groups of subjects, which are likely to reflect average values for extended periods of time.
127
Table 26 Beltsvillea Doubly Labelled Water (DLW) data set: within and between subject variability
within subject CVa % between subject CVb %
TEE n TEE n PAL n
Normal 12.1 19 20.2 210 16.2 203
Overweight 12.4 14 22.4 186 14.4 179
Obese 15.5 9 20.8 101 15.4 96
All subjects 12.6 42 22.8 497 15.4 478 a As reported by Moshfegh et al (2008) b Calculated directly from the data set
128
Appendix 7 – Physical activity and energy balance
Background Physical activity energy expenditure (PAEE) is the most variable component of total 327. energy expenditure (TEE) and is amenable to modification. Decreases in PAEE may affect an individual’s ability to maintain energy balance and consequently maintain a stable body weight. The rising prevalence of obesity has been attributed in part to population-level changes in PAEE (WHO, 1998; Royal College of Physicians, 2004; Foresight, 2007).
The aim of this Appendix is to consider the energy expended during physical 328. activity in relation to risk of weight gain and obesity, and to determine whether a specific level of physical activity can be defined that is protective against a positive energy imbalance, and hence weight gain. Data on the current physical activity levels of the UK population are also considered.
Measuring physical activity Studies investigating physical activity have employed either subjective measures 329. based on self-report and questionnaires, objective measures based on measures of physiological response (e.g. heart rate) or bodily movement (e.g. accelerometers), or TEE measured by the doubly labelled water (DLW) method over longer periods.
Subjective measures of physical activity Many types of physical activity questionnaires have been used in surveys and 330. epidemiological studies. Information obtained is often converted into a summary measure that is then used to categorise or rank the physical activity level of subjects. Questionnaires can detail physical activities performed during a specified period, but their accuracy is limited, especially in assessing non-exercise physical activity (Sesso, 2007).
Children are less likely than adults to make an accurate self-reported physical 331. activity assessment (Welk et al., 2000) and in children of younger age groups it is virtually impossible to obtain valid self-reported physical activity data (Rennie et al., 2006).
Even when physical activity questionnaires are logically constructed with 332. attention to the different domains of activity, they are still relatively imprecise as a measure of PAEE (Wareham et al., 2002). In adults, subjective measures of physical activity have proved sufficient to demonstrate associations with many disease outcomes, to monitor compliance with physical activity guidelines, and have allowed the estimation of dose-response effects. Objective methods are more accurate, however, and allow more precise estimations of dose-response
129
effects. It has been suggested that many physical activity questionnaires have an arbitrary grading in their classification of relative activities, which can result in an overestimation (Erlichman et al., 2002).
Objective measures of physical activity Techniques such as heart rate monitoring (HRM) and accelerometry provide 333. minute-by-minute data and give information on the total levels of physical activity, as well as its intensity, duration and frequency. Accelerometry measures body movement, usually in one (vertical) or three (vertical, lateral and anterior- posterior) planes, but is limited in its ability to measure activities such as swimming and cycling. By applying movement count cut-off points, minute-by-minute data from accelerometers can be summated into time spent in low- moderate- and vigorous-intensity activity. Accelerometers overcome some of the problems of measuring children’s physical activity through subjective measures. There is uncertainty, however, in defining cut-offs for different intensity levels, which causes problems when comparing studies (Freedson et al., 2005; Guinhouya et al., 2006). The HRM method is limited in its ability to differentiate between modest increases in heart rate (HR) above resting levels and increases in HR associated with stress or other causes. The combining of HRM with movement sensors addresses these issues and improves accuracy (Rennie et al., 2006). PAEE can be estimated in groups using HRM and accelerometry (Corder et al., 2007), but the DLW method provides a more accurate assessment.
The DLW method measures TEE over several days and in conjunction with 334. measures, or estimates, of basal metabolic rate (BMR) or resting metabolic rate (RMR) can be used to measure PAEE indirectly. The DLW method is considered the most suitable method for measuring TEE under free-living conditions. It does not, however, give day-to-day information nor does it give information on the forms, frequency and intensity of physical activity undertaken (Rennie et al., 2006). The DLW method is discussed further in Appendix 6.
Assessment of physical activity levels in the UK population Data on physical activity levels of the UK population are available from the 335. national Health Surveys for England, Scotland, Wales and Northern Ireland (Aresu et al., 2009; Corbett et al., 2010; Statistics for Wales/Ystadegau ar gyfer Cymru, 2010; Northern Ireland Statistics and Research Agency, 2006; NHS IC, 2010), the National Diet and Nutrition Survey (NDNS) of adults aged 19-64 years (Ruston et al., 2004), the Low Income Diet and Nutrition Survey (LIDNS) (Nelson et al., 2007) which included adults and children, and the unpublished NDNS comparison study. The Health Surveys use a seven-day recall method to assess physical activity. The adults NDNS used a seven-day diary method and DLW data was also collected in a small subset of survey participants (see paragraph 113), while LIDNS used a four day recall method.
130
Both the NDNS and the Health Surveys provided estimates of physical activity 336. as the proportion of the survey population who reported achieving the physical activity recommendations for England published in 2004 (DH, 2004)34. The surveys consistently show that the majority of people do not meet these physical activity recommendations. These estimates are based on self-reported physical activity questionnaires which are likely to over-report physical activity in the population and have limited accuracy in representing habitual levels.
In children, additional studies assessing physical activity levels via accelerometry 337. and the proportions meeting physical activity recommendations have been published (Basterfield et al., 2008; Mattocks et al., 2007; Riddoch et al., 2007; van Sluijs et al., 2008; Trayers et al., 2006). However, the results are mixed as they are dependent on the threshold used to define moderate-to-vigorous activity.
UK population PAL values It is not possible to derive PAL values from the Health Surveys of England, Scotland, 338. Wales and Northern Ireland because they do not record the amount of reported time spent on all types of activities. In NDNS, PAL values for adults were derived in a validation study conducted prior to the main adult survey (Ruston et al., 2004). Physical activity was assessed in a sample of 66 adults using both seven day physical activity questionnaires and DLW assessment of TEE. The mean PAL values from the questionnaires were 1.76 and 1.66 for men and women respectively, while the PAL values derived from the DLW data were 1.88 and 1.74. However, correlation between the two sets of values was weak, with the activity values explaining only 14% of the variation in the DLW data, indicating that estimates of PAL derived from self-reported physical activity questionnaires resulted in considerable misclassification of activity levels (see paragraph 323). Also, the PAL values are higher than you would expect for the general population, suggesting that this sub-sample is not representative.
Physical activity and body fatness Physical activity has long been considered an integral component in the treatment 339. of those who are obese and in the prevention of weight regain in those who have lost weight (Astrup, 2006; Miller et al., 1997; Shaw et al., 2006). Physical activity alone appears a relatively inefficient means for losing weight, but appears to be an important factor in the successful maintenance of weight loss and in improving insulin sensitivity and cardiovascular health (Astrup, 2006; Atlantis et al., 2006; Bensimhon et al., 2006; CMOs, 2011; Wing & Hill, 2001).
This section focuses on the role of physical activity in the primary prevention of 340. weight gain and obesity.
34 In the UK, adults are recommended to have at least 150 minutes of moderate intensity physical activity a week. Previous recommendations were based on a pattern of activity encompassing 30 minutes of moderate activity on five or more days of the week (DH, 2004); however, the new recommendations (Chief Medical Officers of England, Scotland, Wales and Northern Ireland, 2011) recognise that the 150 minutes can be achieved in a variety of ways. For children and young people, a total of at least 60 minutes each day of at least moderate to vigorous intensity physical activity is recommended.
131
Prospective studies of self-reported physical activity and weight gain Prospective studies relating to physical activity and weight change in both adults 341. and children have been systematically reviewed (Fogelholm & Kukkonen-Harjula 2000; Molnar & Livingstone 2000 and Wareham et al., 2005).
Adults The Fogelholm and Kukkonen-Harjula systematic review (2000) included 16 342. prospective studies investigating the relationship between self-reported physical activity and weight change in adults. It concluded that there was inconsistent evidence of a predictive effect of higher levels of physical activity at baseline being associated with less weight gain over time. The association between weight gain and change in activity was observed to be stronger, although still modest.
A follow-up systematic review by Wareham343. et al (2005) included 12 prospective studies investigating the relationship between self-reported physical activity and weight change. Nine studies reported a negative association between baseline physical activity and subsequent weight gain (Bell et al., 2001; Drøyvold et al., 2004; Hu et al., 2003; Koh-Banerjee et al., 2003; Macdonald et al., 2003; Schmitz et al., 2000; Sherwood et al., 2000; Wagner et al., 2001; Wenche et al., 2004) and two found no association (Ball et al., 2002; Rainwater et al., 2000). One study reported an inverse association suggesting higher baseline levels of BMI predicted physical inactivity (Petersen et al., 2004). The majority of studies suggested that low levels of physical activity were associated with future weight gain, but the effect size was small. The more recent studies included in this review (Wareham et al., 2005) had at least 500 participants, whereas the previous review (Fogelholm & Kukkonen- Harjula, 2000) included five studies with less than 500 participants and, therefore, less power to detect small differences. Improvements in study design could be a factor, as could publication bias, in determining why the follow-up systematic review reported more consistent results.
Wareham344. et al (2005) concluded that in longitudinal cohort studies, individuals who reported higher levels of leisure-time physical activity tended to be less likely to gain weight, but studies varied in their conclusions due to issues of confounding, measurement error and reverse causality, i.e. obesity may lead to physical inactivity (Petersen et al., 2004).
Studies published after these systematic reviews have observed leisure-time 345. physical activity to be inversely associated with BMI (Wilsgaard et al., 2005), waist circumference (Waller et al., 2008) and weight gain, particularly in those with a larger baseline weight (Gordon-Larsen et al., 2009), suggesting a favourable effect of physical activity on weight maintenance.
Children and adolescents The Molnar and Livingstone systematic review (2000) identified two prospective 346. studies that investigated the influence of self-reported physical activity on the change in relative BMI. One study found increases in children’s leisure activity at
132
follow-up to be associated with decreases in subsequent weight gain (Klesges et al., 1995), while the other found no association (Maffeis et al., 1998).
The Wareham347. et al systematic review (2005) identified a further 11 studies. All studies except one (Tammelin et al., 2004), used reported change in BMI or sum of skinfolds as the outcome. Five of the studies did not observe an association between physical activity or sedentary behaviour and weight gain (Bogaert et al., 2003; Davison & Birch, 2001; Francis et al., 2003; Kimm et al., 2001; Mamalakis et al., 2000). The other six studies found an inverse association between higher levels of physical activity and weight gain or a positive association between weight gain and sedentary activities (Berkey et al., 2000; Berkey et al., 2003; Hancox et al., 2004; Horn et al., 2001; O’Loughlin et al., 2000; Proctor et al., 2003; Tammelin et al., 2004).
Overall, the results were mixed and it was concluded that, as in the adult studies, 348. the measures of association tended to be small (Wareham et al., 2005). Another review of prospective studies (Must and Tybor, 2005) also concluded that the results were mixed and that the associations identified were generally of a small magnitude.
Of the studies that have reported subsequent to the Wareham349. et al review (2005), two studies (Kimm et al., 2005; Mundt et al., 2006) have observed physical activity to attenuate increases in fat mass development in boys, but not in girls. One observed no difference in BMI changes or the percentage of students classified as obese between schools with higher and lower frequency of physical education (Wardle et al., 2007). Another study observed reported physical activity and inactivity to be related to accrual of body fat, particularly among children with at least one overweight parent (Must et al., 2007).
Most obese children remain obese as adults (Magarey350. et al., 2003), a progression that is referred to as ‘tracking’ of overweight. Several studies have examined whether adolescent physical activity affects subsequent weight gain through to adulthood. Some prospective studies do provide evidence that a decline in reported physical activity between adolescence and adulthood may increase risk of weight gain and obesity, but these associations are generally weak and inconsistent (Boreham et al., 2004; Kvaavik et al., 2003; Parsons et al., 2006; Pietilainen et al., 2008; Tammelin et al., 2004; Twisk et al., 2000; Yang et al., 2006; Yang et al., 2007).
On balance, the available evidence from prospective cohort studies suggests that 351. increased physical activity and decreased sedentary behaviour may be protective against relative weight and fatness gains; however, the results are mixed and the associations that are identified are generally of a small magnitude. It is likely that imprecise measurement of activity exposures weakens the observed relationships (Must & Tybor, 2005). Measurement error is probably an important factor as most studies rely on subjective measures of reported physical activity and assess fatness using BMI, which is limited in its ability to determine fat and lean tissue mass across the normal range in adults (Wells et al., 2007a) and in children (Wells et al., 2002).
133
Prospective studies of objectively measured physical activity and weight gain
Adults The Wareham352. et al systematic review (2005) identified two studies investigating objectively measured physical activity in relation to weight gain. This has been updated by the Wilks et al systematic review (2011) of studies including objectively measured physical activity, which identified 16 prospective studies, six in adults and ten in children.
In adults, the length of follow up varied between 1.5 and 5.6 years and three of the 353. six studies were carried out in the USA. Three adult studies reported no association between baseline PAEE and subsequent change in body weight. The other three observed favourable associations with increased PAEE inversely correlated with weight outcomes; however, due to methodological issues two of the latter studies were not truly prospective (Bailey et al., 2007; Weinsier et al., 2002).
Children and adolescents The Molnar and Livingstone systematic review (2000) identified five prospective 354. studies that investigated the association between objectively measured physical activity or PAEE and change in indices of body fatness in children and adolescents.
One study observed children with low levels of physical activity (assessed using 355. accelerometry) to gain substantially more subcutaneous fat than more active children (Moore et al., 1995). One small study (n=18) observed reduced TEE, and particularly PAEE, to be associated with weight gain (Roberts et al., 1988), while other, larger studies have found no association (Davies et al., 1991a; Goran et al., 1998b).
The Wareham356. et al (2005) systematic review identified five subsequent studies that had investigated the relationship between physical activity and body weight. The children included in these studies were mostly younger than 10 years and the duration of follow-up ranged from 2 to 8 years.
One study found increased physical activity, assessed using accelerometry, to be 357. associated with smaller gains in BMI and subcutaneous fat (Moore et al., 2003). Overall, however, the results from the other studies using DLW methods to assess energy expenditure were inconsistent (Figueroa-Colon et al., 2000; Johnson et al., 2000; Treuth et al., 2003; Wells & Ritz, 2001).
The Wilks358. et al., review (2011) identified 10 studies in children. All studies were carried out in the USA and the majority of children were aged between 2-12 years at baseline. The duration of follow-up ranged between 1 and 8 years. Six studies found no association between physical activity and adiposity, three found a negative association, and one found a positive association.
134
Overall, the results from prospective studies using objective measures of physical 359. activity in children, adolescents and adults were inconsistent and the associations identified were generally of small magnitude (Must & Tybor, 2005; Wareham et al., 2005; Wilks et al., 2011). For example, the degree of variance in BMI attributed to physical activity in several studies was less than one percent (Ekelund et al., 2004b; Styne, 2005).
The lack of consistent associations between DLW-derived measures of PAEE and 360. measures of body fatness could be interpreted as evidence that energy intake is a more important determinant of excess fat mass gain. There are difficulties in the interpretation of these data, however, because of the controversy regarding the means of comparing TEE and PAEE among individuals of different sizes (Dietz, 1998). It has been suggested that when studies evaluate associations between PAEE or PAL and percentage body fat, the differences between energy expended in physical activity are likely to be overestimated between leaner and fatter children and the differences in body fatness to be underestimated, resulting in associations being biased towards null (Rennie et al., 2006). The use of DLW measures to identify how much PAEE is necessary to prevent obesity is complex; even if appropriate adjustment for body composition is made, comparisons between populations are difficult. It is also important to note that the energy expended in activity may not be the same as the amount of physical activity required to prevent excess FM gain; thus, assessment of physical activity by methods such as HRM and accelerometry is also required (Rennie et al., 2006).
The potential impact of exercise intensity on change in body weight and FM 361. remains unclear (Grediagin et al., 1995; Lemura & Maziekas, 2002; Tremblay et al., 1994; Yoshioka et al., 2001) and it is not known which, if any, of the subcomponents of free-living physical activity contributes more to change in body weight and FM.
Trials using physical activity as an intervention to prevent weight gain Interventions aimed at weight reduction or at preventing weight regain are not 362. included in this consideration. A systematic review by Hardeman et al (2000) identified nine interventions (eleven publications) using physical activity as the primary prevention against weight gain. Interventions lasted from 6 weeks to 36 months. Four interventions took place in the community (Fitzgibbon et al., 1995; Forster et al., 1988; Jeffery & French, 1997; Sherwood et al., 1998; Stolley & Fitzgibbon, 1997) and five were school based (Caballero et al., 1998; Cairella et al., 1998; Davis et al., 1993; Donnelly et al., 1996; Gittelsohn et al., 1998; Simonetti D’Arca et al., 1986).
It was concluded that overall the results suggested mixed effects and, for various 363. methodological reasons, they were uncertain in their conclusions about whether increasing physical activity was effective in preventing weight gain. Effectiveness appeared to be greater among older, male and high-income participants, and lower among low-income participants, school students and smokers. Where diet
135
and physical activity were described, positive effects were usually obtained, but the validity of this was limited as they were measured by self-report.
This systematic review (Hardeman364. et al., 2000) was subsequently updated with a further seventeen trials (Wareham et al., 2005). A total of six trials aimed at increasing physical activity and preventing weight gain in adults were identified. The interventions took place in populations at risk of weight gain or in whom a public health intervention might be targeted. Interventions lasted from 12 weeks to 5 years. In the four trials where differences in body composition between intervention and control group were observed, two found an increase in body weight in the control group and weight stability in the intervention group (Littrell et al., 2003; Simkin-Silverman et al., 2003), one found a weight reduction in the intervention group (Muto et al., 2001) and the other observed decreases in both groups (Proper et al., 2003). Two trials observed no effect on weight gain (Burke et al., 2003; Polley et al., 2002).
A total of eleven trials were identified in children aimed at preventing unhealthy 365. weight gain by increasing physical activity or reducing sedentary behaviour (Wareham et al., 2005). Nine trials were school-based and the others home or family-based. Interventions lasted from 12 weeks to 3 years. Three of the trials reported a small intervention effect at follow-up (Kain et al., 2004; McMurray et al., 2002; Sallis et al., 2003), with two of them reporting effects in boys only (Kain et al., 2004; Sallis et al., 2003). The other eight trials reported no significant effects on body weight or composition at follow-up (Baranowski et al., 2003; Caballero et al., 2003; Dennison et al., 2004; Neumark-Sztainer et al., 2003; Pangrazi et al., 2003; Robinson et al., 2003; Sahota et al., 2001; Warren et al., 2003).
Wareham366. et al (2005) concluded that there were still relatively few trials aimed at the primary prevention of weight gain using physical activity as an intervention and that there remained insufficient evidence on which to base conclusions regarding which approaches were effective.
A subsequent review by Wilks367. et al (2011) identified five intervention trials using objective measures of physical activity, one in adults and four in children. The adult study investigated the effect of a home based exercise programme on adoposity, with compliance measured by accelerometry. However, no change in body weight was observed in either the control or intervention group (Cooper et al., 2000). Of the studies in children (aged between 4-12 years), two were community based, one was nursery based and one was school based. Only one physical activity intervention found a siginificant effect on adiposity (Verstraete et al., 2007).
Additionally, another school-based intervention (where physical activity was 368. assessed by questionnaires) observed a slower gain in BMI, especially in non- overweight adolescents, with increasing physical activity (Simon et al., 2008).
Sedentary behaviour and weight gain Sedentary behaviour is a different concept to physical activity with a different 369. physiology and different determinants. Many behaviours are largely sedentary,
136
including those where sitting or lying are the predominant activity (e.g. listening to the radio or music, watching television, using computers and reading).
It has been suggested that the increased use of information and communication 370. technology is a sedentary factor affecting obesity prevalence (Fox & Hillsdon, 2007; Kautiainen et al., 2005). The use of computers, both at home and at work, has been one of the most rapidly expanding activities in the past 20 years and could potentially impact on overall physical activity levels.
A meta-analysis has been conducted of prospective studies and trials investigating 371. the relation between television viewing and video/computer game use and body fatness and physical activity in children and adolescents (Marshall et al., 2004). The only significant relationship observed was between television viewing and body fatness, but it was concluded that this was likely to be too small to be of substantial clinical relevance and that media-based inactivity may be unfairly implicated in recent epidemiologic trends of overweight and obesity among children and adolescents. It was also noted that relationships between sedentary behaviour and health were unlikely to be explained using single markers of inactivity, such as television viewing or video/computer game use. Physical activity and sedentary behaviours are regulated through a complex series of decision-making mechanisms and restricting television viewing alone may not be effective in increasing physical activity (Nelson et al., 2005).
Several prospective studies conducted since this meta-analysis have also observed 372. positive associations between television viewing in children and subsequent weight gain (Davison et al., 2006; Hancox et al., 2004; Hancox & Poulton, 2006; Jago et al., 2005; Parsons et al., 2008; Reilly et al., 2005; Viner & Cole, 2005). It has been suggested that although the effect size appears small for time spent watching television as a predictor of weight gain in childhood, it is larger than the effect sizes commonly reported for dietary intake and physical activity; thus, television viewing could be an important contributing factor to childhood obesity (Hancox & Poulton, 2006).
The issue of measurement error in these studies and the need to select measures 373. of television viewing that are valid and reliable to examine with greater accuracy the influence of television viewing on childhood overweight, has been highlighted (Bryant et al., 2007).
Most studies examining the prospective and longitudinal associations between 374. sedentary behaviour and BMI have relied on self-reported data. One prospective population-based cohort study measured sedentary behaviour by individually calibrated HRM in 393 healthy adults (Ekelund et al., 2008). At 5.6 years follow- up, sedentary time did not predict any of the obesity indicators (body weight, BMI, fat mass and waist circumference); however, the obesity indicators predicted sedentary time at follow-up after adjustment.
A systematic review has been conducted of trials to reduce sedentary behaviour 375. among children, either alone or in combination with other health messages (Demattia et al., 2007). The interventions ranged from four weeks to four years,
137
with six of the studies targeting clinic-based populations that were overweight or at risk of overweight. A further six were population-based prevention studies. The magnitude of change in weight parameters was modest and was difficult to interpret, as normal BMI ranges vary with age and development in children. The z-BMI score (BMI normalised to age and sex) was only reported in a few of the studies. Virtually all of the interventions, however, consistently resulted in slowing of the increase in the subjects’ BMI relative to similar aged controls. As the sedentary behaviour messages in these interventions are often combined with other health information (e.g. healthy eating and exercise), it was not possible to estimate the magnitude of the weight influences due to sedentary behaviour messages alone.
Summary The assessment of UK population physical activity levels, i.e. the Health Surveys 376. and the NDNS and LIDNS, is currently dependent upon subjective measures of physical activity. It is likely that these give an overestimation of physical activity in the population. To enable the accurate determination of habitual physical activity and PAEE in the UK population, surveys need to employ objective measures.
Methodological constraints are a severe limitation in defining the role of physical 377. activity in the regulation of body weight. Most studies rely on subjective measures of reported physical activity and assess fatness using BMI, which is of limited value in determining fat and lean tissue mass across the normal range in adults (Wells et al., 2007a) and children (Wells et al., 2002). Error is introduced on both sides of the relationship thereby reducing the ability to detect any change. The application of more precise methods for the measurement of physical activity and body fatness is required to define their interrelationship.
On balance, the available evidence from prospective cohort studies suggests 378. that increased leisure time physical activity may be protective against relative weight and fatness gains; however, the findings are mixed and the associations that are identified are generally of a small magnitude. Prospective studies also suggest lengthy television watching may be a predictor for weight gain, but again the associations are weak and inconsistent. Evidence from studies employing objective measures of PAEE and trials of the primary prevention of weight gain through increased physical activity is also inconsistent.
The issue of whether there is a specific level of physical activity required to 379. prevent unhealthy weight gain is complex, and available data is insufficient to reach a definitive conclusion (Blair et al., 2004; Wareham et al., 2005).
138
Appendix 8 – Characteristics of the data set of DLW measures of energy expenditure for adults: combined OPEN/Beltsville DLW data set.
The two data sets of total energy expenditure (TEE) measures used were the OPEN 380. study (Subar et al., 2003; Tooze et al., 2007) (individual data obtained from Amy Subar, National Cancer Institute USA) and the Beltsville underreporting study (Moshfegh et al., 2008) (individual data obtained from Alanna Moshfegh, US Department of Agriculture (USDA), Agricultural Research Service).
The two USA populations studied were similar in terms of anthropometry, ethnicity 381. and social class. However, there is no indication in either study of the range of lifestyles within the cohorts in terms of physical activity levels. Because of this and in the absence of any other cross-sectional population study of TEE, it is not possible to determine how representative the distribution of TEE and physical activity level (PAL) values is of the US or, most importantly, the UK population. However, on the basis of the doubly labelled water (DLW) data examined in the production of this report, it is likely that the median PAL value identified here, 1.63, represents a light activity population. As identified in Appendix 5, this PAL value is lower than either 1.72, the median value of the data base assembled for the US DRI report on energy requirements (IoM, 2005), or 1.75 the mean value of published studies tabulated in Table 22 (Appendix 5). It is also lower than the median value for the NDNS data set i.e. 1.75 for 156 subjects aged 19-75years, with a mean body mass index (BMI) of 27.2kg/m2. In Table 22, study populations categorised as exhibiting light activity exhibited a mean PAL of 1.67; FAO/WHO/UNU (FAO, 2004) identified a PAL range associated with sedentary or light activity lifestyle to be 1.40-1.69.
The OPEN study (n=451) involved healthy volunteers aged 40–69 years recruited 382. from a random sample of 5000 households in the metropolitan area of Washington, DC (Montgomery County, MD). The cohort comprised 245 men and 206 women, of which 85% were white with the rest mainly black or Asian. Most (87%) had some college schooling with 63% college graduates or post graduates. The distribution by BMI groups was 31% normal (18.5 to less than 25 kg/m2), 41% overweight (25 to less than 30 kg/m2), and 29% obese (30 kg/m2 or more).
BMR was not measured in the OPEN study, but BMR was estimated using the 383. Mifflin predictions based on weight, height and age (Tooze et al., 2007). For reasons discussed in Appendix 4, the data presented in this SACN report are calculated using the Henry prediction equations based on weight and height (Henry, 2005). The validity of these Mifflin BMR prediction equations is indicated by the fact that the regression of BMR on BMI within the OPEN cohort is almost identical to that for the Beltsville study (in which BMR was measured) for men and quite similar for
139
women. Consequently, BMR values estimated by BMI within these two cohorts differed by less than 0.5% in men and by less than 2% in women. Also, as shown in Table 27 below, the distribution of PAL values within the OPEN cohort is the same.
The Beltsville study (n=525) involved a study cohort of volunteers aged 30–69 384. years residing in the greater Washington, DC metropolitan area. These volunteers were recruited through advertisements in local newspapers and on websites; announcements sent to employees of USDA (Beltsville, MD), local industries, and offices; and the use of a Beltsville Human Nutrition Research Center database of persons known to be interested in participating in human studies.
The subjects were predominately non-Hispanic white and were distributed evenly 385. by sex and approximately by age. Only 8% of subjects had not attended college. Approximately 21% of the subjects (both sexes) were obese (BMI greater than 30 kg/m2). More females (48%) than males (36%) were considered normal weight. Only 5% of the men and 6% of the women were current smokers. BMR was measured.
The characteristics of the distribution of the PAL values for the two data sets are 386. shown in Table 27. They were very similar. The combined OPEN and Beltsville data set was trimmed for PAL values less than 1.27 or greater than 2.5, on the grounds that these are the limits of sustainable PAL values within a healthy population and that values outside this range are unphysiological. This removed one subject with a PAL value greater than 2.5 and 38 subjects with PAL values less than 1.27 (mean 1.17, range 1.01- 1.269). The effect of trimming was to increase the median PAL value from 1.62 to 1.63.
Table 27 Physical activity level (PAL) value statistics for individual and combined Beltsvillea and OPENb data sets
PAL n Mean SD Min
10th centile
25th centile
Median centile
75th centile
90th centile Max
Beltsvillea 478 1.63 0.25 1.01 1.32 1.46 1.62 1.78 1.96 2.34
OPENb 451 1.64 0.21 1.01 1.40 1.49 1.61 1.77 1.92 2.61
All 929 1.64 0.23 1.01 1.36 1.48 1.62 1.78 1.95 2.61
Trimmed 890 1.66 0.21 1.27 1.40 1.49 1.63 1.78 1.96 2.50 a Beltsville data set (Moshfegh et al, 2008) b OPEN data set (Subar et al., 2003; Tooze et al., 2007)
The distribution of PAL by gender and age, and in terms of frequency, is shown 387. in Figure 6. The distribution of subjects by BMI and age is shown in Figure 7. The regression of TEE on BMR is shown in Figure 8 indicating the very small intercept (33.5 kcal/kg (95% CI:-104 & 171; p=0.6), which is discussed further in Appendix 5.
140
Figure 6 – Distribution of physical activity level (PAL) by gender and age and frequency in the OPEN and Beltsville data sets
Figure 7 – Distribution of subjects by body mass index (BMI) ( kg/m2) and age of OPEN and Beltsville participants.
Figure 8 – Regression of total energy expenditure TEE on basal energy expenditure (BEE): combined OPEN and Beltsville data seta a PAL values from 1.27 to 2.5. TEE (kcal/d) = 33.5206 + 1.6341*X
5000 BEE (kcal/d):TEE (kcal/d): r 2 = 0.6112; r = 0.7818, p = 00.0000
4000
5000
3000
kc al
/d )
2000T EE
(k
1000
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 2600
BEE (kcal/d)
0
Combined OPEN Beltsville data set PAL values from 1.27 to 2.5 TEE (kcal/d) = 33.5206+1.6341 *x
Scatterplot: physical activity level and agep p y y g
2.4
2.6 Male Female
2.2
1.8
2.0
L
1 4
1.6
PA L
1.2
1.4
25 30 35 40 45 50 55 60 65 70 75 1.0
PAL
1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1 2.2 2.3 2.4 2.5 0
20
40
60
80
100
120
140
160
180
200 Histogram: physical activity level
N o.
o f o
bs er
va tio
ns
Age (years)
200
160
180
120
140
160
at io
ns
80
100
120
er o
f o bs
er va
60
80
N um
be
20
40
14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52
BMI (kg/m
0
2)
70
80
60
io ns
40
50
of o
bs er
va ti
20
30
N um
be r
10
20
24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70
Age (years)
0
141
The population demographics for the combined data sets are shown in Table 28, 388. and the distribution of BMI and its relationship to PAL is shown in Table 29 and Table 30.
Table 28 Population demographics for combined Beltsvillea and OPENb data sets
n Mean SD Min Max
Age F 424 51 10 30 70
years M 466 52 10 30 69
All Groups 890 51 10 30 70
Heightc F 206 1.63 0.06 1.47 1.82
meters M 239 1.77 0.07 1.58 1.94
All Groups 445 1.70 0.10 1.47 1.94
Weightc F 206 73 17 45 126
kg M 239 88 16 54 138
All Groups 445 81 18 45 138
BMI F 424 27 6 17 51
kg/m2 M 466 28 4 18 43
All Groups 890 27 5 17 51
BMR F 424 1393 166 982 1960
kcal/d M 466 1757 217 1279 2522
All Groups 890 1584 266 982 2522
TEE F 424 2290 386 1442 3731
kcal/d M 466 2923 516 1889 5061
All Groups 890 2621 557 1442 5061 a Beltsville data set (Moshfegh et al, 2008) b OPEN data set (Subar et al., 2003; Tooze et al., 2007) c Beltsville data set contains only BMI, not height or weight F=female, M=male
Table 29 Distribution of body mass index (BMI) and its relationship to physcial activity level (PAL) for the combined Beltsvillea and OPENb data sets
BMI kg/m2 PAL
N mean sd mean median sd Min Max
Normal All 322 22.7 1.69 1.65 0.21 1.61 1.27 2.37
Female 181 22.4 1.76 1.64 0.22 1.60 1.27 2.37
Male 141 23.1 1.51 1.67 0.21 1.64 1.29 2.34
Overweight All 348 27.2 1.45 1.66 0.20 1.65 1.29 2.50
Female 136 27.1 1.41 1.66 0.21 1.63 1.29 2.50
Male 212 27.3 1.47 1.66 0.20 1.66 1.29 2.31
Obese All 220 34.1 3.94 1.66 0.23 1.63 1.28 2.37
Female 107 34.7 4.37 1.65 0.21 1.63 1.28 2.15
Male 113 33.5 3.41 1.67 0.24 1.63 1.29 2.37
All 890 27.3 4.97 1.66 0.21 1.63 1.27 2.50 a Beltsville data set (Moshfegh et al, 2008) b OPEN data set (Subar et al., 2003; Tooze et al., 2007)
142
Table 30 Distribution of physcial activity level (PAL) according to body mass index (BMI) categories for the combined Beltsvillea and OPENb data sets
Status N Mean Std.Dev. Min 10th Q25 Median Q75 90th Max
Normalc 322 1.65 0.21 1.27 1.40 1.49 1.61 1.78 1.95 2.37
Overweightc 348 1.66 0.20 1.29 1.41 1.50 1.65 1.79 1.92 2.50
Obesec 220 1.66 0.23 1.28 1.40 1.47 1.63 1.80 1.99 2.37
All 890 1.66 0.21 1.27 1.40 1.49 1.63 1.78 1.96 2.50 a Beltsville data set (Moshfegh et al, 2008) b OPEN data set (Subar et al., 2003; Tooze et al., 2007) c Using the following BMI definitions: Underweight: less than 18.5 kg/m2; normal: 18.5 to less than 25kg/m2; overweight: 25 to less than 30 kg/m2; obese 30 kg/m2 or more; overweight including obese 25 kg/m2 or more; morbidly obese: 40 kg/m2 or more
Within the cohort, there were similar numbers of overweight and obese subjects 389. (36% normal, 39% overweight and 25% obese). Mean PAL values for BMI categories did not differ significantly (p=0.91) and neither did the distribution in terms of quartile boundaries and 10th and 90th centile values. The regression of PAL on BMI was also non-significant (p=0.64 for slope: R2 less than 0.1%). Finally, as shown below in Figure 9, the regressions of PAL on BMI for the two cohorts were very similar even though PAL was calculated from a measured BMR for the Beltsville data set compared with an estimated BMR for the OPEN data. Thus, estimating BMR for an overweight/obese population did not appear to introduce bias. This lack of influence of BMI on PAL was also observed in the NDNS sample.
Figure 9 – Relationship of physical activity level (PAL) values with body mass index (BMI) for the combined Beltsvillea and OPENb data sets
133
Deleted: <sp>
Figure 9. Relationship of physical activity level (PAL) values with body mass index (BMI) for the combined Beltsvillea and OPENb data sets
a Beltsville data set (Moshfegh et al, 2008) bOPEN data set (Subar et al., 2003; Tooze et al., 2007)
390. PAL fell with age (PAL=1.74-0.0016 x age, R 2=0.004, p=<0.03); however, the shallow slope means that the fall with age is small and age explains only 0.4% of the variance, i.e. PAL = 1.69 at 30 and 1.63 at 70. Also, as shown in Table 31, a comparison of PAL values between the younger subjects (aged ≤35 years) with those aged >35 years shows the absolute values of mean and median PAL values to be slightly lower in the younger subjects, although not significantly different (p=0.51 for mean values). This suggests that the lack of adult subjects aged <30 years is unlikely to represent a significant source of bias when the data set is used to represent PAL values for all adults. This is supported by the NDNS DLW data set (see paragraph 113) for which between the ages of 19 and 75 years there was no significant change in PAL with age (R2 = 0.0153; p = >0.1; PAL = 1.85 - 0.0018*age: a fall of <0.1PAL units between 20 and 75 years).
Table 31. Comparison of physical activity level (PAL) values between the youngest group (aged≤35 years) with those >35 years for the combined Beltsvillea and OPENb data sets
Age N Mean Std.Dev. Min 10th
centile Q25 Median Q75 90th
centile Max
≤35 56 1.64 0.21 1.30 1.41 1.46 1.61 1.78 1.97 2.12 >35 834 1.66 0.21 1.27 1.40 1.49 1.63 1.78 1.96 2.50
a Beltsville data set (Moshfegh et al, 2008) bOPEN data set (Subar et al., 2003; Tooze et al., 2007)
Scatterplot: BMI vs. PAL (OPEN data) PAL = 1.6218 + .88E-3 * BMI
Correlation: r = .02267
15 20 25 30 35 40 45 50 55
BMI
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
P A
L
95% confidence
Scatterplot: BMI vs. PAL (Beltsville data) PAL = 1.7240 - .0022 * BMI
Correlation: r = -.0451
15 20 25 30 35 40 45 50 55
BMI
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
P A
L
95% confidence
a Beltsville data set (Moshfegh et al, 2008) b OPEN data set (Subar et al., 2003; Tooze et al., 2007)
PAL fell with age (PAL=1.74-0.0016 x age, R390. 2=0.004, p=<0.03); however, the shallow slope means that the fall with age is small and age explains only 0.4% of the variance, i.e. PAL = 1.69 at 30 and 1.63 at 70. Also, as shown in Table 31, a comparison of PAL values between the younger subjects (aged ≤35 years) with those aged >35 years shows the absolute values of mean and median PAL values to be slightly lower in the younger subjects, although not significantly different (p=0.51 for mean values). This suggests that the lack of adult subjects aged <30 years is unlikely to represent a significant source of bias when the data set is used to represent PAL
143
values for all adults. This is supported by the NDNS DLW data set (see paragraph 113) for which there was no significant change in PAL with age, between the ages of 19 and 75 years (R2 = 0.015, Regression Coefficient for age on PAL: — 0.0018). This would be a fall in PAL of around 0.1 units.
Table 31 Comparison of physical activity level (PAL) values between the youngest group (aged≤35 years) with those >35 years for the combined Beltsvillea and OPENb data sets
Age N Mean Std.
Dev. Min 10th
centile Q25 Median Q75 90th
centile Max
<35 56 1.64 0.21 1.30 1.41 1.46 1.61 1.78 1.97 2.12
>35 834 1.66 0.21 1.27 1.40 1.49 1.63 1.78 1.96 2.50 a Beltsville data set (Moshfegh et al, 2008) b OPEN data set (Subar et al., 2003; Tooze et al., 2007)
144
Appendix 9 – Energy requirements for disease
Introduction There is a continuum in physical function between health and disease, and the 391. dividing line between them can be difficult to define. The proportion of older individuals in the UK population is increasing (Walker et al., 2001) and the incidence of many diseases and disabilities also increases with age (Elia et al., 2000). In the 2001 UK General Household Survey, one in ten respondents under the age of 45 years reported a longstanding illness that limited physical activity, compared with a third of older respondents (Walker et al., 2001). Less severe disabilities may produce effects on physical activity that are difficult to distinguish from normal activity. The high prevalence of disease in older people makes it difficult to define ‘normal’, especially in extreme old age. A consideration of physical activity level (PAL) values in older adults is given in Appendix 5.
Determining the energy requirements of patients with acute and chronic diseases 392. is more complex than for those in good health. Energy requirements in illness are greatly influenced by the type (acute or chronic), severity, and phase of the disease (acute or recovery phase). They are also affected by the presence of other physical and psychological disabilities, which may vary over time, the treatment of the disease, and the nutritional state of the patient (e.g. the presence of prior malnutrition). In some circumstances, it may not be appropriate to treat the disease or the associated malnutrition, as when a patient is approaching death and feeding is a burden.
The energy requirements of a number of severe acute diseases were thought 393. to be increased (Elia, 1995), as were the requirements of individuals with spastic disorders, such as spastic cerebral palsy (Eddy et al., 1965). It is now understood that this is not usually the case.
Energy intake Individuals with disease may be in substantial energy imbalance for days, weeks, 394. and sometimes months. This imbalance is often caused by anorexia which is a common consequence of traumatic, infective, malignant, or inflammatory diseases. Clearly, accurate measurements of energy intake in disease do not necessarily reflect the energy required to maintain energy balance or the energy required to achieve optimal health. Furthermore, although in under-nourished subjects, additional energy is required for repletion and improvement in tissue function, health and well-being, the reverse applies to obese individuals. If absorption of nutrients is reduced due to illness or if urinary losses occur (e.g. in diabetes), then the nutritional value of ingested foods will be less than in an unaffected individual.
145
The timing of nutritional support is also important. Aggressive overfeeding in the 395. acute phase of injury, for example, can cause metabolic disturbances, such as hyperglycaemia (diabetes of injury) and increased CO2 production, which can be detrimental to patients with respiratory failure. In contrast, very slow repletion during recovery can prolong the period of ill health, and will also have a negative impact on work performance and quality of life. Finally, nutritional intake in disease may differ from that in health because it can be provided artificially using an enteral tube or venous catheter (sometimes for life), independent of appetite.
Energy expenditure Basal metabolic rate (BMR) is more variable in disease than in health. Apart 396. from being influenced by the type, severity and phase of the illness, BMR is also influenced by the nutritional status of the patient and a wide range of treatments, which may vary from surgical interventions, immobilisation and artificial ventilation, to blood transfusions and drug therapy. Standard reference tables or equations for estimating BMR (from weight, height and age) were established for use in healthy subjects without malnutrition and disease, and without dehydration and oedema, all of which can have substantial effects on body weight. The estimation of BMR, therefore, is more likely to be in error in disease than in health.
Whether measurements of total energy expenditure (TEE) (or measurements of 397. BMR) are regarded as normal or abnormal relative to a healthy group may depend on the way in which the results are expressed. This can be problematic for patients with diseases associated with abnormal body composition and body proportions who preferentially lose or preserve particular tissues or organs (e.g. muscle wasting in certain neurological conditions). Different investigators have used over 20 indices to express energy expenditure, including absolute energy expenditure, energy expenditure per kilogram body weight, energy expenditure per kilogram fat free mass (FFM), and energy expenditure per m2 (Elia, 1997). In children, energy expenditure has also been expressed as a percentage of the BMR values obtained in healthy children of the same age, the same surface area, or the same height. In infants with growth failure, there may be a decrease in both the whole body BMR and the tissue specific BMR, while BMR per kilogram body weight may be increased due to preferential preservation of the brain, which has a high metabolic rate (Elia, 1997). Similarly, TEE may be low when expressed in relation to values obtained in children of the same age, and normal when it is expressed in relation to healthy children of the same weight. The hypermetabolic effect of disease and the hypometabolic effect of weight loss alter BMR to a greater extent than in health, even when age, weight and height are taken into account. This affects the accuracy of expressing TEE as a ratio to BMR (i.e. PAL) (Elia, 1992). Consequently, PAL values based on measured BMR and estimated BMR (based on values established for healthy subjects) can vary widely, especially in acute disease which typically increases BMR.
As with energy intake, TEE does not necessarily indicate the requirements of the 398. under-nourished patient who is in need of repletion, or the requirements of the over-nourished patient, who is in need of depletion of excess fat.
146
Measurements of TEE have helped establish important concepts about energy 399. requirements in disease. Acute and chronic diseases are associated with simultaneous changes in BMR, physical activity and TEE. These factors, and the methodological limitations associated with their estimation, represent important practical and theoretical issues that should be considered when determining energy requirements in pathological states.
Diseases in adults For many diseases, both TEE and energy requirements are decreased mainly because 400. of a reduction in physical activity energy expenditure (PAEE). This reduction in PAEE can occur because disease produces lethargy and restricts physical activity (e.g. pain due to claudication). It also occurs because treatment, especially in hospital, usually requires restricted physical activity. Loss of weight will contribute to the reduction in TEE, partly because it reduces the energy expended in physical activity and partly because it reduces BMR.
TEE is normal or decreased in chronic diseases (Elia, 2005). In those with advanced 401. disease or with disease-related weight loss, both PAEE and TEE are usually decreased, despite a possible increase in BMR. The weight loss that may occur in such diseases is more likely to be due to a reduction in energy intake than an increase in energy expenditure. A possible exception to this general rule concerns subgroups of patients with anorexia nervosa, who have been reported to have increased PAEE and TEE, when adjusted for weight (Casper et al., 1991). However, some studies of women with anorexia nervosa show no increase in TEE, after adjustment for weight (Marken Lichtenbelt et al., 1997). When no adjustment for weight is made, TEE is likely to be lower than in healthy women of the same age.
Acute diseases increase resting energy expenditure (REE) above that predicted for 402. healthy individuals of the same age, weight and height by up to 100% (Bessey & Wilmore, 1988), although usually by 0-40%. Both the magnitude and duration of the increase in energy expenditure are dependent on the severity of disease. The effect of ‘injury’ on BMR is also influenced by age. For example, the increase in BMR may last a few days following elective surgery in adults and only a few hours in infants (Elia, 2000). Acute or sub-acute diseases usually cause a decrease in lean body mass. After the early phase of an illness, therefore, BMR may decline below pre-illness BMR before beginning to return to normal in the recovery phase.
In studies where the TEE of bed-bound artificially ventilated patients has been 403. measured (e.g. those with head injuries or other critical illness), BMR is frequently increased. TEE is usually not elevated however, primarily because of the concomitant reduction in physical activity (Elia, 2005). The overall result is that TEE is normal or even decreased compared to values obtained in healthy subjects in free-living circumstances, with the exception of the most severe acute diseases, such as burns, when TEE may be transiently elevated above normal (Bessey & Wilmore, 1988).
147
In summary, in most of the chronic conditions examined BMR adjusted for weight 404. is usually normal or slightly increased (about 10% in HIV/AIDS for example), while in most acute conditions it is usually increased (0-40% and occasionally more). This increase in BMR is counteracted by the decrease in PAEE, with the overall result that TEE is usually normal or decreased. More severe acute disease produces greater increments in BMR and greater reductions in PAEE.
Diseases in children Assessment of the energy requirements of children is more difficult than in adults 405. because of the need to consider growth and the most appropriate way to express energy expenditure.
In children, most chronic disease conditions do not produce an increase in TEE 406. adjusted for body weight, despite a possible increase in BMR (e.g. cystic fibrosis (Magoffin et al., 2008). There are some exceptions, such as subgroups of patients with cystic fibrosis (cystic fibrosis genotype) and some children with congenital heart disease (Kuip et al., 2003). The general situation is analogous to that observed in adults, as is that for acute disease conditions, where TEE is not increased (actually decreased) because of the reduction in PAEE.
Summary The lack of accurate information on TEE in a large number of diseases, does not 407. allow a comprehensive assessment of the field. In those conditions which have been investigated, TEE is usually normal or reduced, partly because of a reduction in body weight and FFM (due to disease-related malnutrition or neurological causes of wasting), and partly because of reduced physical activity. The reduced physical activity compensates for any increase in BMR, which is common in acute diseases. There are some exceptions, such as subgroups of patients with cystic fibrosis, anorexia nervosa, and congenital heart disease, where TEE has been reported to be increased. Such patients are often underweight, and therefore weight adjustments are necessary to demonstrate differences compared to control groups. Without such adjustments, TEE is again typically not increased.
In determining energy requirements in disease there is a need to consider not only 408. the energy required to maintain energy balance, but also the energy required to change body composition at different rates in both under-nourished and over- nourished patients.
148
Appendix 10 – Obesity prevalence in the UK
Body Mass Index (BMI) (kg/m409. 2) is often used as a convenient measure of adiposity, with recognised limitations. It has been used since the 1960s to assess obesity in adults (Keys et al., 1972) and more recently in children (Cole et al., 2005; Dietz & Robinson, 1998). While BMI is a good measure of weight independent of height, it fails to distinguish between adipose and non-adipose body mass. When used as a proxy for fat mass (FM) or adiposity, assumptions are made about absolute and relative body composition, which may not be valid. BMI fails to reflect body shape, and hence fat distribution, and is not strongly related to central adiposity, which is considered to be most harmful to health (Wells et al., 2007a).
In adults, cut-off points for underweight, overweight and obesity are defined as 410. BMI values of ≤18.5 kg/m2, ≥25 kg/m2 and ≥30 kg/m2, respectively (NICE, 2006). Insufficient energy intakes are uncommon in healthy free-living individuals in the UK and do not generally arise from insufficient food supplies, but from accompanying physical or psychological conditions. In the 2009 Health Survey for England (HSE) (NHS IC, 2010), 2.3% of adults were classed as underweight. The Low Income Diet and Nutrition Survey (Nelson et al., 2007) also found the prevalence of underweight to be low in adults living in low income households (2% of both men and women). In contrast, there is a high prevalence of overweight and obesity in the UK population resulting from a chronic excess of dietary energy intake over energy expenditure.
In children, BMI measures require cautious interpretation when comparing across 411. groups that differ in age or when predicting a specific individual’s total or percent body fat (Pietrobelli et al., 1998). Children of the same age and gender have been shown to have a two-fold range of FM for a given BMI value, which is also observed in those who are obese (Wells et al., 2007b). BMI normally changes during growth and this requires careful interpretation through the use of appropriate growth references or standards for age and sex. In the UK, the UK 1990 growth references (Cole, 1995) were used until 2009 for all children up to 18 years of age. From May 2009, for children aged between 0-4 years, revised standards have been adopted based on the WHO Child Growth Standards (SACN/RCPCH, 2007; RCPCH, 2011). For the purpose of population monitoring, children with a BMI over the 85th centile for age of the reference population are categorised as overweight and those with a BMI over the 95th centile for age as obese (NHS IC, 2010). In UK clinical practice, however, children over the 91st centile for the reference population are categorised as overweight and those over the 98th as obese (Hall & Elliman, 2003). Further thresholds have been proposed for the purpose of international comparison (Cole et al., 2000); these allow estimation of the proportion of children at each age expected to exceed a BMI of 25 or 30 at the age of 18 years. The use of different
149
definitions that are not directly comparable can lead to difficulty in interpreting different surveys or studies of childhood obesity.
Temporal changes in the estimated prevalence of overweight and obesity in England 412. are tabulated below for children aged 2-15 years based on the UK 1990 BMI growth reference (Cole, 1995) (Table 32), and for adults (NHS IC, 2010) (Table 33).
In 2009, 31% boys and 28% girls in England were overweight or obese (NHS IC, 413. 2010). Obesity (as defined by a BMI over the 95th centile for age) increased from 11% and 12% in 1995 to 16% and 15% in 2009, in boy and girls respectively. Despite the overall increase since 1995, the proportion of girls aged 2 to 15 years who were obese decreased from 18% to 15% between 2005 and 2006 and remained 15% in 2009. There was no significant decrease among boys over the same period, with 17% classed as obese in 2006 and 2008.
For adults in England, the prevalence of obesity has increased from 15% in 1993 to 414. 23% in 2009 (NHS IC, 2010), with 66% men and 57% women overweight and obese. Fewer people are in the normal weight range and the proportion of morbidly obese individuals has more than doubled. Importantly for women of childbearing age, for whom overweight and obesity results in increased clinical risk (McDonald et al., 2010; Stothard et al., 2009), the prevalence of overweight and obesity is substantial and increases with age. Thus, for the age groups 16 – 24 years, 25 – 34 years and 35 – 44 years, mean BMI values were 24.3, 25.8 and 27.2 and the prevalence of overweight and obesity was 37%, 48% and 62%, respectively. In a comparison of the HSE between 1993 and 2003, both BMI and central adiposity (waist circumference) increased more in the upper part of the distribution, with intermediate increases in the middle and little change at the lower end of the distribution (Wardle & Boniface, 2008). The observed temporal gains in central adiposity were not equivalent across the BMI distribution. Thinner people were almost as thin as they were 10 years earlier, but fatter people were considerably fatter.
The 2009 Scottish Health Survey (Corbett et al., 2010) shows there has also been 415. a steady upward trend in the prevalence of overweight and obesity among both sexes since 1995 for adults and children. Most adults aged 16 years or over are either overweight or obese i.e. 66% of men and 58% of women. Overall obesity prevalence was 27% for men and 28% for women, and of these 1.0% and 3.5% were morbidly obese respectively, a prevalence which has stabilised since 2003. For children aged 2 to 15 years, boys were more likely than girls to be overweight or obese (29% versus 27%).
The Health Surveys for Scotland and England involve objective measures of weight 416. and height, but the respective Health Surveys for Wales and Northern Ireland use self-reported measures and are therefore less accurate.
In the 2009 Welsh Health Survey (Statistics for Wales/Ystadegau ar gyfer Cymru, 417. 2010), 62% of men were classified as overweight or obese compared with 52% of women. In men, 41% were overweight and 21% obese, while in women 31% were
150
overweight and 21% obese. In children, 34% were estimated to be overweight or obese, including 19% obese.
The 2005/6 Health and Social Wellbeing Survey in Northern Ireland (Northern 418. Ireland Statistics and Research Agency, 2006) observed 64% men and 59% women to be either overweight or obese. In men, 39% were overweight and 25% obese, while in women 35% were overweight and 24% obese. In boys, 18% were overweight and 20% obese and in girls, 16% were overweight and 15% obese.
151
Table 32 Overweight and obesity prevalence among children, by year, 1995 to 2009, in the Health Survey for Englanda
All Children (aged 2-15 years)
Percentages
Unweightedb Weightedb
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
Overweightc 13.3 13.4 13.2 14.3 14.1 12.7 15.5 14 15 15.4 14.8 14 14.3 14.3 14.2
Obesec 11.7 12.2 12.8 13.7 15.5 14.5 15.2 17.4 16.9 18.9 18.6 16.3 16.8 16 15.7
Overweight including obesec 25 25.6 25.9 28 29.7 27.2 30.7 31.4 31.9 34.3 33.4 30.3 31.1 30.3 29.8 a NHS IC, 2010 b All years were weighted to adjust for the probability of selection, and from 2003 non-response weighting was also applied. c Categories are independent, i.e. overweight does not include those who are obese. Overweight was defined as ≥ 85th < 95th UK BMI percentile; obese was defined as ≥ 95th UK BMI percentile
152
Table 33 Body Mass Index (BMI)a among adultsb, 1993 to 2009 in the Health Survey for Englanda
All adultsb Percentages
Unweightedc Weightedc
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
Underweightd 1.6 1.7 1.8 1.7 1.5 1.7 1.7 1.5 1.4 1.7 1.8 1.6 1.6 1.6 1.6 1.8 2.3
Normald 45.5 45.2 43.7 42.2 41.6 40.6 40.4 38.6 37.0 37.7 37.8 36.7 37.9 36.8 37.7 36.8 36.4
Overweightd 38.0 37.4 38.1 38.7 38.5 38.3 38.0 38.8 39.2 38.1 37.9 38.8 37.3 37.6 36.7 36.9 38.3
Obesed 14.9 15.7 16.4 17.5 18.4 19.4 20.0 21.2 22.4 22.5 22.6 22.9 23.2 23.9 24.0 24.5 23.0
Overweight including obesed
52.9 53.1 54.5 56.2 56.9 57.7 57.9 60.0 61.6 60.6 60.5 61.8 60.5 61.6 60.8 61.4 61.3
Morbidly obesed 0.8 1.0 0.9 0.9 1.6 1.3 1.4 1.5 1.7 1.8 1.9 1.7 1.8 2.1 1.8 2.0 2.4
a NHS IC, 2010 b Adults aged 16 and over with a valid height and weight measurement c Data from 2003 onwards have been weighted for non-response. d Using the following BMI definitions: Underweight: less than 18.5 kg/m2; normal: 18.5 to less than 25kg/m2; overweight: 25 to less than 30 kg/m2; obese 30 kg/m2 or more; overweight including obese 25 kg/m2 or more; morbidly obese: 40 kg/m2 or more
153
Appendix 11 – A consideration of energy intake and physical activity in relation to weight gain
Background Weight gain is dependent on the relationship between energy intake and energy 419. expenditure, and it is therefore necessary to consider energy balance and energy flux, rather than energy intake and energy expenditure in isolation. This is the focus of this appendix.
It is not currently possible to accurately define the frequency, intensity and 420. duration of physical activities that reduce the risk of weight gain. Available evidence suggests the influence of physical activity on body weight is weak, although methodological constraints are clearly an issue in interpreting the data (see Appendix 7). Furthermore, in experimental studies it is difficult to control for changes in energy intake which may occur during periods of altered physical activity.
The evidence investigating whether diet composition affects risk of weight gain is 421. also weak (Jebb, 2007), but clearly a mismatch between energy intakes and energy expenditure is fundamental to weight gain. Methodological constraints have hampered the elucidation of the role of physical activity and diet composition in the development of overweight and obesity. Weight gain and obesity must result from a chronic positive energy imbalance, which implies a failure of auto-regulatory homeostatic responses to maintain energy balance. The asymmetry between the hunger and satiety arms of human appetite control has been implicated in this process (Prentice & Jebb, 2004).
The average daily weight gain in modern populations is relatively small and even 422. in morbidly obese people the lifetime error in daily energy balance regulation is surprisingly small (Prentice & Jebb, 2004). Adult weight gain indicated in national surveys in the US is up to 8g/d (90th centile) (Hill et al., 2003; Hill, 2009) and up to twice this rate for excess weight gain in children (Butte & Ellis, 2003). Estimates of the energy cost of excess weight gain (mean growth costs in overweight less mean growth costs in normal weight children) are about 550kJ/d (130kcal/d) but this is equivalent to only 3-4% of estimated energy intakes (Butte & Ellis, 2003).
Energy balance and energy flux It is a widely discussed hypothesis that the mechanisms controlling energy balance 423. may be more accurate at higher levels of physical activity, with some suggesting a threshold of physical activity or energy flux below which these mechanisms become imprecise and dysregulated leading to a positive energy imbalance (Blair
154
et al., 2004) and obesity (Mayer & Thomas, 1967). Some evidence that the coupling between energy expenditure and energy intake may be less efficient at low levels of physical activity (Prentice & Jebb, 2004; Schoeller, 1998) is reviewed here.
Metabolic studies lasting between one and two weeks have covertly manipulated 424. the energy density of foods supplied to volunteers fed ad libitum (Prentice, 1998; Prentice & Jebb, 2004; Stubbs et al., 1995a; Stubbs et al., 1995b) to investigate the impact on energy balance in subjects with different activity levels. For those fed a 40% fat energy diet, less physically active lean subjects (confined to whole-body calorimeters) had a positive energy imbalance of +850 kJ per day compared with a negative energy imbalance of -1,800kJ per day observed in those under free-living conditions.
Another two day whole-body calorimeter study showed that for lean subjects 425. fed ad libitum the imposition of sedentary behaviour resulted in a positive energy imbalance especially on a high fat (covertly manipulated) diet whereas planned exercise enabled energy balance stability (Murgatroyd et al., 1999).
A further seven day whole-body calorimeter study showed that lean subjects 426. fed ad libitum consumed the same energy intake when sedentary (1.4 x resting metabolic rate – RMR) compared with moderately active (1.8 x RMR) exhibiting a positive energy imbalance of 15.2 MJ over the seven days of the sedentary regimen (Stubbs et al., 2004).
Some investigators have proposed a threshold of physical activity below which 427. appetite control is ineffective (Mayer, 1966). Based on a doubly labelled water (DLW) study of previously obese women (Schoeller, 1998), a physical activity level (PAL) value of 1.75 has been suggested as the threshold, although the evidence supporting this is not strong (Prentice & Jebb, 2004).
While these studies do provide some support for increased physical activity 428. enabling better energy balance, the evidence is only short term (one to two weeks), and the concept of a threshold of energy flux above which body weight regulation is more sensitive has not been demonstrated convincingly.
Energy flux can be increased through weight gain, which increases basal metabolic 429. rate (BMR) independently from any change in physical activity energy expenditure (PAEE). Indeed, the increasing cost of exercise with weight gain means that energy flux could increase in the obese even when the range and extent of actual activities falls. This is suggested by the lack of any obvious relationship between PAL and BMI (see Appendix 8). It is not clear whether the apparent improved energy balance regulation with increased physical activity, described above for lean adults, occurs in the obese (Hill, 2006). It may be, however, that the increasing energy costs of PAEE with weight gain represent a barrier to achieving sufficient increases in PAEE to improve energy balance regulation.
While physical activity alone seems to be a relatively inefficient means for losing 430. weight in overweight individuals, it appears it can be a factor in the successful maintenance of weight loss (Astrup, 2006) (see Appendix 7). The efficacy of
155
increased physical activity in maintaining weight loss could be partly due to the increased energy flux; the psychological effects of exercise may also be important by enhancing well-being and status of control, and hence compliance with a restrictive dietary regimen (Prentice & Jebb, 2004).
Even when increasing PAEE and energy flux does not lead to a loss of body weight 431. but a compensatory increase in energy intake, if energy balance is maintained this may also result in beneficial nutritional effects through the increased intake of other food constituents, such as micronutrients, which reduce the risk of nutritional deficiencies and poor nutritional status (Melzer et al., 2005).
Exercise and appetite control There appears to be a spontaneous reduction in hunger associated with 432. participation in exercise programmes (Elder & Roberts, 2007). Short-term (1-2 day) and medium-term (7-16 day) physical activity intervention studies show substantial initial negative energy balances but energy intake subsequently increases to provide partial compensation for about 30% of the energy expended in activity (Blundell et al., 2003), although the extent of compensation varies between individuals (King et al., 2007). However, a study investigating the effect of an imposed sedentary routine on appetite, energy intake, energy balance and nutrient balance in lean men over seven days found no equivalent compensatory reduction in energy intake, leading to a significantly positive energy imbalance (Stubbs et al., 2004).
It seems, therefore, that it might take considerable time for energy intake to fully 433. adjust to changes in PAEE. However, there is also evidence that eventual increases in food intake do not follow the same pattern in obese as in lean individuals (Melzer et al., 2005).
There is some evidence to suggest that exercise may modulate appetite control 434. by improving the sensitivity of the physiological satiety signalling system (Blundell et al., 2003; Martins et al., 2008a; Martins et al., 2008b). Compensation for a high- carbohydrate preload was observed to be more accurate in habitual exercisers than non-exercisers (Long et al., 2002), as well as in those who had completed a six week moderate-intensity exercise intervention (Martins et al., 2007). Following a bout of exercise, subjects were observed to discriminate more accurately between the energy content of different beverages (King et al., 1999).
Summary While it may not be possible to accurately define the nature of the relationship 435. between physical activity, diet and weight gain, evidence reviewed here suggests that the matching of energy intake and expenditure may be improved at higher expenditure levels.
Most previous research has focused exclusively on the effects of one type of 436. diet or physical activity, or has examined diet composition and physical inactivity independently, but few studies have compared the effects of a combination of diet composition and physical activity. Any effect of diet composition may be
156
dependent on the pattern of physical activity. Conversely, any effect of physical activity may depend, in part, on diet composition. A better understanding of these two factors and how they interact may help explain why obesity is so prevalent in the UK, while allowing for the fact that not everyone within that environment is obese.
157
Appendix 12 – Characteristics of the data set of doubly labelled water (DLW) measures of energy expenditure in children and adolescents
The available doubly labelled water (DLW) data from studies of children aged more 437. than one year were examined. A data set of 845 individual values was assembled for the US DRI report (IoM, 2005), probably limited to studies published before 2002. The distribution by age was uneven and included large numbers of infants aged under two years, 4-5 and 8-10 year olds, with fewer 2-3 and 10-18 year olds.
The FAO/WHO/UNU report (FAO, 2004) was based on studies assembled by 438. Torun (2005) which were listed in terms of mean study values for children aged over one year. This involved many more individual children with a more even age spread than that examined in the US DRI report (IoM, 2005).
For this SACN report, a data set of all published DLW studies of children aged over 439. one year was compiled (the SACN children’s dataset). This included all those studies assembled by Torun (2005) and other studies published up to 2006 (Anderson et al., 2004; Arvidsson et al., 2005; Bandini et al., 2002; Bratteby et al., 2005; Craig et al., 1996; DeLany et al., 2006; Goran et al., 1993b; Hernandez-Triana et al., 2002; Hoos et al., 2003; Lindquist et al., 2000; Lopez-Alarcon et al., 2004; McGloin et al., 2002; Montgomery et al., 2005; Perks et al., 2000; Roemmich et al., 2000; Rush et al., 2003; Sjoberg et al., 2003; Sun et al., 1999; Wong et al., 1999). All studies were tabulated according to study mean values for boys and girls for specific age groups (see below Table 37). This resulted in 170 data points (study means) representing a total of 3502 individual measurements (females=2082, males=1420). They included four studies from Central and South America (Brazil, Chile, Guatemala and Mexico) but all children were considered well-nourished. The remaining studies were mainly from the UK or the USA with single studies from Canada, Denmark, the Netherlands and Sweden. Those from the USA involved Caucasian-American, African-American and native-American children. UK studies (500 individual measurements) included only 15% of the total children studied and these were mainly infants and younger children. Only 6% of the total adolescents studied were from the UK.
Physical activity level (PAL) values and basal metabolic rate (BMR) values were not 440. included in many of the studies. For all studies which did not report BMR, BMR values were estimated using the Henry equations (Henry, 2005) for weight and height or just weight if no height data were reported. PAL values were then derived from total energy expenditure (TEE) and BMR (see paragraph 79).
In addition, a data set of individual values obtained from the UK NDNS unpublished 441. comparison study was examined. This included 65 subjects between the ages of
158
4 – 18 years. Whilst this sample size was too small to represent a reference data set, because it was a uniform sample in terms of the measurement methodology and is exclusively UK data it was able to provide useful comparative information (see Figure 12 and paragraph 444).
Number of studies and individuals within the studies and general characteristics
442. Mean PAL study values for the SACN children’s data sat are shown in Figure 10. The tabulated data can be found in Table 37 at the end of this Appendix. The data set includes a reasonable overall age spread in terms of number of studies, although there are five or less studies for the age groups of 3, 11, 13, 15 and 16 years. Thus, there were relatively few subjects studied at the ages of 3, 4, 12, 14, 16-18 years and a relative excess of subjects (>600) at 10 years of age. Older adolescents (>15 years of age) are under-represented. Although most data points do relate to discreet age groups, in some cases data points are the mean of an age range. For example, three data points for eight year olds include PAL values from a total of 149 children with ages ranging from 5-10.5 years.
Figure 10 – Mean physical activity level (PAL) study values in the SACN children’s data set as a function of age and gender
Physical Activity Level values with age
0 2 4 6 8 10 12 14 16 18 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4
P hy
si ca
l A ct
iv ity
L ev
el c
al cu
la tio
n
Female Male
Age (years)
Physical activity level as a function of agePhysical activity level as a function of age
2.2
2.4 Median 25%-75% Min-Max
2.0
1.8
ac tiv
ity le
ve l
1 4
1.6
P hy
si ca
l
1.2
1.4
1.0 e 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18g
Subjects 16 187 39 84 256 112 222 469 300 656 331 58 274 65 286 71 20 56 Studies 2 8 4 7 18 7 12 22 19 20 8 5 11 3 11 5 2 6
Age (years)
159
The scatterplot of TEE on BMR is shown in Figure 12. The increase in PAL after the 443. age of four years means that the slope of the relationship of TEE with BMR increases after this time. This relationship is discussed in more detail in Appendix 5.
Figure 11 – Scatterplot of total energy expenditure (TEE) on basal metabolic rate (BMR) (SACN children’s data set)
Total Energy Expenditure versus Basal Metabolic Rate Children and adolescents (study mean values)
16
18
12
14
J/ d
8
10
12
xp en
di tu
re M
J
6
8
al E
ne rg
y E
x
2
4To t
0 1 2 3 4 5 6 7 8 9
Basal Metabolic Rate MJ/d
0
Figure 12 – Physical activity level (PAL) values in the NDNS children's data set as a fucntion of age and gender
1
1
1.2
1.4
1.6
1.8
2
2.2
2.4
0 5 10 15 20
Ph ys
ic al
a ct
iv it
y le
ve l (
PA L)
Age (years)
Boys
Girls
Grouping by age Overall, PAL appears to increase with age (see Figure 12) from 1.33 to 1.81 between 444. the ages of 1 and 18 years. However, there was a clustering of the youngest children (aged 1-3 years) at the lower range (PAL ≈ 1.4), with an increase in the overall range starting at school age and with fewer studies with mean PAL values below 1.6 (n = 2) in adolescents. From an early age (≈5 years) there is a wide range of mean study PAL values for boys and girls around the regression and as discussed below (see paragraph 448), it is not entirely clear whether the apparent change with age for school-age children is real. One possibility was to exclude those studies with particularly low mean study values but it was decided that there was insufficient
160
evidence to do this. Therefore, all studies shown in Figure 10 were included in the analysis. The NDNS data set, of which individual values are shown in figure 12 and the median values for 4-9 and 10-18 year olds are shown in Figure 13, also displays an increase in PAL between these two age groups (although as shown in both Figures, the increase is slightly less than in the main data set). Grouping of all study means into three age groups was identified as a valid way of expressing the increase in PAL with age, thereby simplifying the calculation of energy reference values for children and adolescents.
Grouping of PAL values within the ages for which BMR is estimated (i.e. 1-3, 3-445. <10 and 10-18 years) is shown in Figure 13 and Table 34. Median PAL values for the three age groups are 1.39 (≤3 years), 1.57 (>3-<10 years) and 1.73 (>10-18 years). For the age groups ≤3 years and >10-18 years and for the ages of 5.5-8 years, the median values fall within or close to the 95th CI values of the regression. Clearly, the step changes in PAL and consequent energy reference values at the age group boundaries are a disadvantage to this approach which needs to be balanced against the simplifying effect of not using age specific PAL values predicted from the regression.
Figure 13 – Grouping of physical activity level (PAL) values by ages 0-3, >3-<10 and >10-18 years (SACN children’s data set)
148
Figure 12. Grouping of physical activity level (PAL) values by ages 0–3, >3–<10 and >10–18 years (SACN children’s data set)
Table 34. Grouping of physical activity level (PAL) values within the ages ≤3, >3-<10 and 10- 18 years (SACN children’s data set)
Age PAL N mean sd Min Max mean sd Min Max Q25 Median Q75 1 14 2.3 0.54 1.5 3.0 1.39 0.06 1.26 1.46 1.35 1.39 1.43 2 85 7.0 1.87 4.0 9.3 1.56 0.14 1.21 1.98 1.42 1.57 1.69 3 71 13.0 2.38 10.0 18.0 1.75 0.13 1.42 2.19 1.66 1.73 1.85
Influence of gender 446. Factorial ANOVA of PAL by age group and gender shows that gender is not a significant
influence on PAL but there are significant differences between each of the three age groups. No gender differences were observed within the NDNS data set.
Table 35. Influences of gender on physical activity level (PAL) values within the ages 1-3,>3- <10 and 10-18 years (SACN children’s data set)
Gender Age PAL
Group N Means Std.Dev. Means Std.Dev. P values Girls 1 5 2.0 0.34 1.41 0.06 Gender 0.28
2 23 5.4 1.23 1.50 0.18 Group <0.0001 3 63 11.2 2.94 1.67 0.15 Gender x group 0.55
Boys 1 5 2.0 0.34 1.40 0.04 2 25 5.4 1.24 1.54 0.18 3 49 11.5 3.19 1.71 0.20
Box & Whisker plot: PAL
Median 25%-75% Min - Max
1 2 3
Age group
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
P A
L
Children and adolescents age 1-18 PAL values: study means for boys and girls separately NDNS data: median values of individual subjects for age ranges shown.
0 2 4 6 8 10 12 14 16 18 20
Age (years)
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
P A
L
group1 group 2 group 3 1.39 1.57 1.73
NDNS 1.64 1.72
Table 34 Grouping of physical activity level (PAL) values within the ages ≤3, >3-<10 and 10-18 years (SACN children’s data set)
Age PAL
N Mean sd Min Max Mean sd Min Max Q25 Median Q75
1 14 2.3 0.54 1.5 3.0 1.39 0.06 1.26 1.46 1.35 1.39 1.43
2 85 7.0 1.87 4.0 9.3 1.56 0.14 1.21 1.98 1.42 1.57 1.69
3 71 13.0 2.38 10.0 18.0 1.75 0.13 1.42 2.19 1.66 1.73 1.85
Influence of gender Factorial ANOVA of PAL by age group and gender shows that gender is not a 446. significant influence on PAL but there are significant differences between each of the three age groups. No gender differences were observed within the NDNS data set, shown in Figure 12.
161
Table 35 Influences of gender on physical activity level (PAL) values within the ages 1-3,>3-<10 and 10-18 years (SACN children’s data set)
Gender Age PAL
Group N Means Std.Dev. Means Std.Dev. P values
Girls 1 5 2.0 0.34 1.41 0.06 Gender 0.28
2 23 5.4 1.23 1.50 0.18 Group <0.0001
3 63 11.2 2.94 1.67 0.15 Gender x group 0.55
Boys 1 5 2.0 0.34 1.40 0.04
2 25 5.4 1.24 1.54 0.18
3 49 11.5 3.19 1.71 0.20
Variation of PAL with age within the BMR age groups PAL varies with age within both the 3-447. <10 and 10-18 year age groups but the influence of age is minor compared with the overall variability (R2=0.098 and 0.057) especially in the 10-18 year group (p=0.04). The slopes imply increases in PAL of 0.14 and 0.1 PAL units over the age ranges (3-<10 and 10-18 years respectively), a very small change compared with the wide within-group overall range as shown above (Table 35).
It is by no means certain whether the changes with age for school-age children are 448. real or reflect selection bias or other between-study factors. Indeed, on the basis of the association of PAL values with activity levels as shown in Table 22 (Appendix 5), many of the mean PAL values for younger school children (PAL<1.5) would imply very low activity levels compared with other children at the same age. An analysis of the data set by DLW methodology (multipoint, two point or other) indicates that the PAL values for those involving the two point method, are almost invariably lower (≈ 0.2 PAL units) than the other methodologies and include all the very low activity groups. Because these studies represent a considerable fraction of studies on children aged from 5-9 years but relatively few involve adolescents, their inclusion in the data set does tend to increase the age-related changes in PAL. However, as discussed in Appendix 6 (paragraph 318), although some reviewers suggest that the multipoint approach may decrease the error to a small degree, overall the different approaches are in principle equally valid.
Examination of the few studies which report the range of PAL values in randomly 449. selected pre-adolescent school children shows that such children do exhibit the same wide range of physical activity as the adult population. This is observed in the NDNS data set from the age of 7 (see Figure 12). In a study of 47 randomly selected Australian school children aged 5-10.5 years, PAL varied from 1.32-2.18 (mean =1.71: tertiles of 1.32-1.63, 1.64-1.80, 1.81-2.18) (Abbott and Davies, 2004). The same group reported mean PAL values of a group of 106 school children aged 6.0-9.6 years (Ball et al., 2001) which varied over a very wide range of 1.2-2.32, mean = 1.70. The combined data from these two studies (n=149) indicated a median PAL of 1.69 (range 1.19-2.34, 25th and 75th centiles of 1.56 and 1.83) with ages from 5-10.5 years. Although individual PAL values are not shown as a function of age, the authors
162
Table 36 PAL values for use in calculation of energy requirements of children and adolescents, adjusted for growth
Age group PALa
mean Q25 Median Q75
1-3 1.36 1.40 1.45
>3-<10 1.43 1.58 1.70
10-18 1.68 1.75 1.86 a PAL as indicated in Table 34 adjusted for growth (=PALx1.01)
analysis of variation in PAL values (in relation to adiposity), implied that age was not an influence.
Adjustment of PAL to account for growth costs 450. Actual growth costs calculated by SACN differ only slightly from those reported
by FAO/WHO/UNU (FAO, 2004). In the US DRI report (IoM, 2005) a simplified estimate is reported with single values for the 3-7 and 8-18 year age groups. Within the FAO/WHO/UNU report (FAO, 2004) growth costs are also reported as an adjustment to average PAL values involving an increase of 1% i.e. an assumption that growth costs (deposition) are equivalent to 1% of the energy requirement.
451. It is the case that if growth costs are calculated as a percentage of the energy requirement, the values indicated are an overall average (1-16 years) of 0.98% of the energy requirement (range, boys: 0.4% – 1.37%; girls: 0.05% – 1.59% girls). If a single average growth value as a percentage of the energy requirement was used throughout the age range, the maximum errors would be during the peak growth phase for older children where growth would be underestimated by up to 0.6% of the energy requirement for girls or 0.4% for boys and then overestimated by similar amounts as growth slows at the end of adolescence. Relatively, these are very small amounts (i.e. 0.01PAL units) given the variation in PAL and overall TEE which occurs at all ages.
452. Thus, growth costs can be included within a factorial model as an adjustment to PAL values of an increase of 1% (i.e. =1.01 x PAL).
Table 37 SACN data set of all published doubly-labelled water (DLW) studies of children aged over one year
Study n Sex Age
years Weight
kg Height
cm BMI TEE
MJ/d BMR MJ/d PAL
Abbott et al., 2004 24 F 8 30.1 131.0 17.4 7.90 4.61 1.71
Abbott et al., 2004 23 M 9 30.0 133.0 16.7 8.30 4.92 1.69
Anderson et al., 2004 172 F 10 33.1 141.0 16.6 8.10 5.14 1.58
Arvidsson et al., 2005 16 F 16 56.4 164.0 21.0 9.10 5.70 1.60
Arvidsson et al., 2005 17 M 16 64.1 175.0 21.0 11.30 6.90 1.64
Ball et al., 2001 54 F 8 28.1 127.0 17.2 7.50 4.45 1.69
Ball et al., 2001 52 M 8 27.6 128.0 16.7 7.90 4.70 1.68
163
Study n Sex Age
years Weight
kg Height
cm BMI TEE
MJ/d BMR MJ/d PAL
Bandini et al., 1990b 14 F 14 55.7 162.0 21.2 10.00 6.02 1.66
Bandini et al., 1990b 14 M 15 56.4 167.0 20.2 13.00 7.28 1.79
Bandini et al., 2002 123 F 10 30.1 137.0 16.0 7.70 4.90 1.57
Bandini et al., 2002 73 F 11 37.9 146.0 17.8 8.80 5.50 1.60
Bratteby et al., 1997 25 F 15 58.4 167.0 20.9 10.70 5.97 1.79
Bratteby et al., 1997 25 M 15 61.3 174.0 20.2 14.10 7.31 1.93
Bratteby et al., 2005 89 F 15 56.0 165.0 20.4 11.00 5.85 1.88
Bratteby et al., 2005 71 M 15 60.7 175.0 19.7 14.10 7.14 1.97
Butte et al., 2000a 43 F 2 10.9 82.0 16.3 3.50 1.88 1.86
Butte et al., 2000a 33 M 2 11.4 83.0 16.8 3.90 2.01 1.94
Butte et al., 2000a 43 F 2 12.0 88.0 16.6 4.10 2.85 1.44
Butte et al., 2000a 33 M 2 12.5 88.0 16.2 4.10 2.95 1.39
Champagne et al., 1998 31 F 10 39.3 144.0 19.0 9.70 5.34 1.82
Champagne et al., 1998 27 F 10 39.2 146.0 18.0 9.00 4.99 1.80
Champagne et al., 1998 29 M 10 46.4 147.0 21.0 10.60 6.14 1.73
Champagne et al., 1998 31 M 10 42.9 148.0 19.8 10.80 5.93 1.82
Champagne et al., 1998 15 F 10 28.9 8.28 4.52 1.83
Champagne et al., 1998 15 M 10 36.8 10.66 5.26 2.03
Craig et al., 1996 49 F 10 32.6 140.0 16.4 8.40 4.86 1.73
Davies et al., 1991b 16 F 5 18.5 6.18 3.68 1.68
Davies et al., 1991b 12 M 5 18.9 6.88 3.92 1.75
Davies et al., 1991b 15 F 7 26.0 8.17 4.31 1.90
Davies et al., 1991b 10 M 7 24.6 8.15 4.46 1.83
Davies et al., 1991b 15 F 9 29.1 7.57 4.57 1.66
Davies et al., 1991b 14 M 9 29.5 8.95 4.91 1.82
Davies et al., 1991b 10 F 12 49.3 10.53 5.47 1.92
Davies et al., 1991b 8 M 12 39.7 10.48 5.48 1.91
Davies et al., 1991b 11 F 15 58.0 10.12 5.88 1.72
Davies et al., 1991b 12 M 15 60.1 13.47 7.05 1.91
Davies et al., 1991b 11 F 18 62.4 11.09 6.08 1.82
Davies et al., 1991b 12 M 18 71.6 15.05 7.94 1.90
Davies & White 1995 11 M 2 12.7 88.0 16.3 4.50 3.11 1.45
Davies & White, 1995 12 F 2 13.0 89.0 16.4 4.40 3.07 1.43
Davies & White, 1995 16 F 3 14.9 96.0 16.2 4.70 3.51 1.34
Davies & White, 1995 15 M 3 15.0 96.0 16.3 5.10 3.67 1.39
Davies & White, 1995 16 M 4 16.9 104.0 15.6 5.40 3.71 1.46
Davies & White, 1995 11 F 4 17.1 5.30 3.56 1.49
Table 37 (continued) SACN data set of all published doubly-labelled water (DLW) studies of children aged over one year
164
Study n Sex Age
years Weight
kg Height
cm BMI TEE
MJ/d BMR MJ/d PAL
DeLany et al., 2002 65 F 11 37.7 144.0 18.4 8.80 5.63 1.56
DeLany et al., 2002 66 M 11 43.9 146.0 20.1 10.30 6.38 1.61
DeLany et al., 2004 53 F 13 49.1 156.0 23.7 9.20 6.24 1.47
DeLany et al., 2004 61 M 13 55.3 159.0 25.6 11.10 7.17 1.55
DeLany et al., 2006 28 F 11 37.9 145.0 18.0 9.10 5.38 1.69
DeLany et al., 2006 25 F 11 39.4 147.0 18.9 9.70 5.55 1.75
DeLany et al., 2006 31 M 11 45.0 146.0 21.0 10.80 5.83 1.85
DeLany et al., 2006 29 M 11 42.6 143.0 19.5 10.80 6.42 1.68
Ekelund et al., 2001 11 F 9 37.0 139.0 19.2 8.30 5.14 1.61
Ekelund et al., 2001 15 M 9 33.0 140.0 16.8 8.90 5.20 1.71
Ekelund et al., 2002 10 F 17 61.4 166.0 22.3 10.40 6.00 1.73
Ekelund et al., 2002 8 M 18 73.7 180.0 22.7 13.20 7.20 1.83
Ekelund et al., 2004 11 F 10 37.0 139.0 19.2 8.20 5.10 1.61
Ekelund et al., 2004 15 M 10 33.0 140.0 16.8 8.90 5.20 1.71
Ekelund et al., 2004 10 F 17 62.3 166.0 22.6 10.40 6.20 1.68
Ekelund et al., 2004 15 M 18 71.4 179.0 22.3 14.40 8.00 1.80
Fontvieille et al., 1993 15 M 5 21.1 114.0 16.2 5.90 4.34 1.36
Fontvieille et al., 1993 13 F 6 18.9 112.0 15.1 5.60 4.00 1.40
Goran et al., 1993b 14 F 5 21.0 113.0 16.4 5.50 4.54 1.21
Goran et al., 1993b 16 M 5 20.3 112.0 16.2 6.00 4.74 1.27
Goran et al., 1995 34 F 5 20.1 110.0 16.4 5.20 4.20 1.24
Goran et al., 1995 36 M 5 20.1 111.0 16.1 5.80 4.60 1.26
Goran et al., 1998c 11 M 5 21.3 113.0 16.7 6.60 4.90 1.35
Goran et al., 1998c 11 F 6 21.5 114.0 16.5 5.70 4.57 1.25
Goran et al., 1998c 11 M 6 25.2 120.0 17.5 7.50 5.30 1.42
Goran et al., 1998c 11 F 7 24.8 121.0 16.9 7.60 4.80 1.58
Goran et al., 1998c 11 M 9 37.8 138.0 19.8 8.70 5.90 1.47
Hernandez-Triana et al., 2002
5 M 5 18.0 110.0 15.0 6.30 3.72 1.69
Hernandez-Triana et al., 2002
6 F 6 19.5 6.10 3.76 1.62
Hoffman et al., 2000 14 F 10 30.9 8.08 4.72 1.71
Hoffman et al., 2000 14 M 10 32.0 9.03 4.89 1.85
Hoos et al., 2003 8 F 6 18.6 109.0 15.7 5.70 3.98 1.43
Hoos et al., 2003 3 M 9 21.8 121.0 14.9 7.30 4.34 1.68
Kaskoun et al., 1994 23 F 6 20.7 5.63 3.86 1.46
Kaskoun et al., 1994 22 M 6 19.5 5.84 3.98 1.47
Lindquist et al., 2000 17 M 9 41.2 139.0 20.9 7.90 5.70 1.38
Table 37 (continued) SACN data set of all published doubly-labelled water (DLW) studies of children aged over one year
165
Study n Sex Age
years Weight
kg Height
cm BMI TEE
MJ/d BMR MJ/d PAL
Lindquist et al., 2000 13 F 10 42.9 141.0 20.8 7.90 5.56 1.42
Livingstone et al., 1992 6 M 3 16.4 100.1 5.4 3.98 1.36
Livingstone et al., 1992 2 F 3 15.9 103.0 4.83 3.85 1.26
Livingstone et al., 1992 6 F 5 18.1 110.2 5.38 3.63 1.48
Livingstone et al., 1992 6 M 5 17.9 111.6 6.8 3.87 1.76
Livingstone et al., 1992 6 F 7 23.5 120.4 7.27 4.08 1.78
Livingstone et al., 1992 6 M 7 25.4 125.8 7.98 4.53 1.76
Livingstone et al., 1992 6 F 9 32.2 134.0 7.89 4.78 1.65
Livingstone et al., 1992 6 M 9 30.2 135.2 9.68 4.96 1.95
Livingstone et al., 1992 6 M 12 44.5 152.1 10.98 5.84 1.88
Livingstone et al., 1992 6 F 12 44.8 157.3 10.1 5.33 1.90
Livingstone et al., 1992 6 F 15 57.2 162.2 10.39 5.86 1.77
Livingstone et al., 1992 6 M 15 56.4 173.1 13.03 6.84 1.90
Livingstone et al., 1992 5 F 18 63.9 159.9 10.26 6.10 1.68
Livingstone et al., 1992 5 M 18 78.5 181.3 16.73 8.37 2.00
Livingstone et al., 1992 5 F 8 23.5 120.0 16.3 7.10 4.08 1.74
Livingstone et al., 1992 6 M 8 25.4 126.0 16.0 8.00 4.54 1.76
Livingstone et al., 1992 4 F 9 33.4 134.0 18.6 8.10 4.86 1.67
Livingstone et al., 1992 5 M 9 30.2 135.0 16.6 9.80 4.96 1.98
Livingstone et al., 1992 5 M 13 43.8 152.0 19.0 10.70 5.79 1.85
Livingstone et al., 1992 5 F 13 45.1 158.0 18.1 11.70 5.35 2.19
Livingstone et al., 1992 3 M 15 50.7 11.00 6.33 1.74
Livingstone et al., 1992 3 F 16 55.4 9.60 5.76 1.67
Lopez-Alarcon et al., 2004
5 F 4 21.3 111.0 17.3 6.00 3.85 1.56
Lopez-Alarcon et al., 2004
7 M 5 19.7 109.0 16.6 6.20 3.95 1.57
Lopez-Alarcon et al., 2004
10 M 5 20.0 112.0 15.9 6.10 4.01 1.52
Lopez-Alarcon et al., 2004
7 F 5 20.7 113.0 16.2 5.70 3.83 1.49
Montgomery et al., 2005
31 F 6 19.6 112.0 15.4 6.00 3.75 1.60
Montgomery et al., 2005
32 M 6 20.6 115.0 16.3 6.90 4.09 1.69
Nagy et al., 1997 22 M 7 31.6 129.0 18.4 7.10 5.20 1.37
Nagy et al., 1997 9 F 8 28.8 125.0 18.2 6.70 4.80 1.40
Nagy et al., 1997 25 F 8 32.8 129.0 19.1 6.90 4.90 1.41
Nagy et al., 1997 20 M 8 29.7 130.0 17.4 6.90 5.10 1.35
Table 37 (continued) SACN data set of all published doubly-labelled water (DLW) studies of children aged over one year
166
Study n Sex Age
years Weight
kg Height
cm BMI TEE
MJ/d BMR MJ/d PAL
O’Connor et al., 2001 22 M 7 25.0 123.0 16.4 7.60 4.47 1.70
O’Connor et al., 2001 25 F 8 26.5 124.0 17.1 7.20 4.31 1.67
Perks et al., 2000 27 F 13 49.4 154.0 20.5 9.60 5.70 1.68
Perks et al., 2000 23 M 13 45.1 140.0 18.3 10.10 5.86 1.72
Prentice et al., 1988 8 M 2 11.2 3.89 2.72 1.43
Prentice et al., 1988 8 F 2 11.2 3.89 2.66 1.46
Prentice et al., 1988 6 F 2 13.3 4.51 3.18 1.42
Prentice et al., 1988 6 M 2 13.3 4.51 3.25 1.39
Rennie et al., 2005 21 F 7 22.4 122.0 15.1 6.40 4.02 1.59
Rennie et al., 2005 19 F 7 25.0 123.0 16.7 7.10 4.20 1.69
Rennie et al., 2005 29 M 7 24.5 126.0 15.5 7.80 4.48 1.74
Rennie et al., 2005 31 M 7 26.1 126.0 16.3 8.10 4.58 1.77
Roemmich et al., 1998 12 F 11 36.6 8.07 4.88 1.65
Roemmich et al., 1998 18 M 11 35.3 8.91 5.14 1.73
Roemmich et al., 1998 18 F 14 51.8 9.67 5.59 1.73
Roemmich et al., 1998 11 M 15 54.5 11.28 6.62 1.70
Roemmich et al., 2000 13 F 10 34.7 137.0 18.5 8.90 5.09 1.75
Roemmich et al., 2000 14 M 11 34.8 143.0 17.0 9.10 5.20 1.75
Roemmich et al., 2000 18 F 13 51.2 158.0 20.5 9.40 5.70 1.65
Roemmich et al., 2000 14 M 13 52.0 162.0 19.8 10.70 6.80 1.57
Rush et al., 2003 13 M 10 35.4 140.0 17.5 9.80 5.90 1.66
Rush et al., 2003 13 F 10 35.5 140.0 17.4 8.10 5.05 1.60
Salazar et al., 2000 14 M 4 17.6 6.19 3.80 1.63
Salazar et al., 2000 14 F 5 17.6 5.73 3.60 1.59
Saris et al., 1989 10 F 8 28.2 8.06 4.49 1.79
Saris et al., 1989 9 M 9 30.9 9.00 5.05 1.78
Sjoberg et al., l 2003 17 F 16 56.4 164.0 20.8 10.10 5.85 1.73
Sjoberg et al., 2003 18 M 16 64.1 176.0 20.8 12.70 7.38 1.72
Spadano et al., l 2005 28 F 10 33.7 141.0 16.9 8.20 5.20 1.58
Spadano et al., 2005 28 F 12 45.3 154.0 19.0 9.40 5.92 1.59
Spadano et al., 2005 24 F 15 58.4 165.0 21.6 10.40 5.85 1.78
Sun et al., 1998 30 M 7 30.5 128.0 18.6 7.20 5.29 1.36
Sun et al., 1998 29 F 8 34.5 129.0 20.7 7.20 5.06 1.42
Sun et al., 1998 21 M 8 29.3 130.0 17.3 7.10 5.20 1.37
Sun et al., 1998 18 F 8 41.2 132.0 23.6 8.10 5.35 1.51
Sun et al., 1999 30 F 8 31.6 128.0 18.7 7.00 4.80 1.46
Sun et al., 1999 34 M 8 34.9 133.0 21.8 7.30 5.60 1.30
Table 37 (continued) SACN data set of all published doubly-labelled water (DLW) studies of children aged over one year
167
Study n Sex Age
years Weight
kg Height
cm BMI TEE
MJ/d BMR MJ/d PAL
Sun et al., 1999 30 M 9 34.4 132.0 19.3 7.10 5.60 1.27
Sun et al., 1999 18 F 9 40.5 134.0 19.2 7.70 5.60 1.38
Treuth et al., 1998 12 F 8 28.5 129.0 17.1 6.59 4.50 1.46
Treuth et al., 1998 12 F 9 46.5 134.0 25.9 8.40 5.40 1.56
Treuth et al., 2000b 30 F 9 27.2 130.0 15.9 7.10 4.41 1.61
Treuth et al., 2000b 44 F 9 28.0 130.0 16.5 7.40 4.47 1.66
Treuth et al., 2000b 27 F 9 29.6 131.0 17.2 7.50 4.58 1.64
Trowbridge et al., 1997 18 F 8 40.3 7.40 5.00 1.48
Trowbridge et al., 1997 27 F 8 33.8 7.40 5.10 1.45
Trowbridge et al., 1997 13 M 8 28.5 7.80 5.40 1.44
Trowbridge et al., 1997 17 M 8 30.8 7.70 5.50 1.40
Valencia et al., 1995 10 M 8 26.9 6.60 4.67 1.41
Valencia et al., 1995 10 M 8 27.2 7.49 4.70 1.59
Vasquez et al., 2006 12 F 4 23.1 108.0 20.6 6.20 4.00 1.55
Vasquez et al., 2006 12 M 4 22.3 108.0 19.1 6.80 4.23 1.61
Wong et al., 1994 9 F 13 43.3 155.0 17.8 9.70 5.24 1.85
Wong et al., 1999 41 F 13 57.5 160.0 22.5 10.10 5.57 1.81
Wong et al., 1999 40 F 14 53.2 159.0 20.9 11.80 5.90 2.00
Wren 1997 8 M 5 19.1 5.80 3.94 1.47
Wren 1997 8 F 5 18.5 5.22 3.68 1.42
Table 37 (continued) SACN data set of all published doubly-labelled water (DLW) studies of children aged over one year
168
Appendix 13 – Comparisons with the existing DRVs for energy from the 1991 COMA report ‘Dietary Reference Values for Energy and Nutrients for the United Kingdom’.
Overarching differences The energy reference values detailed in this SACN report derive from a methodology 453. which differs from that employed in 1991 by the Committee on Medical Aspects of Food Policy (COMA) (Department of Health (DH) 1991) in several ways. In contrast to the COMA report (DH, 1991):
Energy reference values have been calculated from rates of total energy • expenditure (TEE) assessed by the doubly labelled water (DLW) method. This has been used either directly as a measure of TEE for infants or, in all other cases, to identify suitable physical activity level (PAL) values which have been employed within a factorial calculation as basal metabolic rate (BMR) x PAL.
BMR has been estimated using the Henry prediction equations (Henry 2005), • resulting in slightly lower BMR values than would have been estimated from the Schofield equations (Schofield et al., 1985) used by COMA in 1991.
The population PAL values identified are higher. • In this report, a value of 1.63 has been used as the population PAL value for adults compared to a PAL value of 1.4 commonly cited from the COMA report (Table 2.8 DH, 1991).
A prescriptive approach has been adopted, calculating energy reference values • on the basis of healthy body weights (i.e. using a desirable body mass index – BMI) which gives slightly lower values than those used by COMA.
Infants aged 1-12 months COMA defined energy reference values for infants aged 0-12 months on the basis 454. of available evidence on TEE and energy intakes. Energy reference values for infants were only defined for breast milk substitute-fed infants as it was felt that a reference value for breast-fed infants was meaningless in practice.
SACN followed the approach of the FAO/WHO/UNU (Butte, 2005; FAO, 2004) 455. using more recent infant growth data from the UK-WHO Growth Standards (RCPCH, 2011). Separate values are provided for breast-fed and breast milk substitute-fed infants, and values are also given for when the method of feeding is mixed or not known. The new energy reference values described in this SACN report are
169
10-14% higher at 0-3 months but are lower by between 7-18% for infants after three months of age compared to the COMA values (see Table 38).
Children and adolescents aged 1-18 years For children aged 1-10 years, in the absence of sufficient TEE data, COMA based 456. its reference values on energy intake data. For older children and adolescents (aged 10-18 years), COMA defined energy reference values with a factorial model assigning PAL values of 1.56 for boys and 1.48 for girls with small additions made for growth costs. The PAL values used were lower for girls compared to boys, with the gender differences based on the view that girls exhibited lower levels of physical activity than boys. Average weights and heights from studies available at the time were used by COMA to calculate BMR using the Schofield equations.
SACN calculated energy reference values for boys and girls aged 1-18 years using a 457. factorial model. PAL values were identified for three age groups (1-<3, 3-<10 10-18 years) from an analysis of DLW measures of TEE and adjusted for growth in terms of a 1% increase. As discussed in Appendix 5 (paragraph 276), in the DLW data sets examined in this SACN report no differences in PAL with gender were identified for children (or adults) and the same PAL values are therefore used for boys and girls. Three PAL values are presented in the current report for the three age groups of children representing “less active”, “typically active” and “more active” (see paragraph 461 for further detail under the adults section). The new energy reference values were calculated from BMR values estimated using Henry equations based on what can be considered best estimates of healthy body weights i.e. mean weights from the UK-WHO Growth Standards (RCPCH, 2011) (ages 1 to 4 years) and from the 50th centile of the UK 1990 reference for children and adolescents (Freeman et al., 1995) for children aged more than four years.
Table 38 shows energy reference values for infants, children and adolescents in 458. the current SACN report based on median PAL values for the various age groups compared with summary values reported by COMA (in Table 2.6 DH, 1991), with the change in this report shown as a percentage. Some of the body weights at the various ages used to calculate values in the two reports vary slightly and this explains some of the differences although most are due to the different methods of calculation. The overall pattern of differences is for, lower values from 3 months to 10 years, with higher values for adolescents, especially girls (up to 16% higher) due to use of a higher PAL value.
170
Table 38 Energy reference values for infants, children and adolescents in the current report compared with values reported by COMAa
Age Energy reference values (MJ/d)
COMA (1991)a SACN 2011 Change (±%)
Boys Girls Boys Girls Boys Girls
0-3 months 2.3 2.2 2.6b 2.4b 14 10
4-6 months 2.9 2.7 2.7b 2.5b -8 -7
7-9 months 3.4 3.2 2.9b 2.7b -15 -16
10-12 months 3.9 3.6 3.2b 3.0b -16 -18
1-3 years 5.2 4.9 4.1 3.8 -20 -22
4-6 years 7.2 6.5 6.2 5.8 -14 -11
7-10 years 8.2 7.3 7.6 7.2 -8 -2
11-14 years 9.3 7.7 9.8 9.1 6 18
15-18 years 11.5 8.8 12.6 10.2 9 15 a DH, 1991 b Using the comparable values for breast milk substitute fed infants (see Table 5).
Adults The COMA report listed EAR values for energy for men and women aged 19-29 459. years and 30-59 years calculated as BMR x PAL. BMR was calculated from modified Schofield equations (Schofield et al., 1985) for a wide range of body weights, and EAR values were listed for nine PAL values ranging from 1.4 to 2.2. Summary values were then presented for men and women aged 19-49 years and 50-59 years using a the PAL value of 1.4 (Table 2.8 DH, 1991).
COMA derived its PAL values from an analysis of PAR values for specific activities 460. because at that time more reliable information on PAL values was unavailable. COMA assumed that much of the population had an inactive lifestyle and recommended that when activity patterns were unknown, energy reference values for the adult population should be based on a PAL value of 1.4.
This SACN report employs a different approach to identifying PAL values and 461. to formulating specific recommendations for population subgroups. Thus, the magnitude and distribution of PAL values reflect measured values of TEE and BMR in reference populations. For adults in this report, PAL values of 1.49, 1.63 and 1.78 equate to the 25th, median and 75th centile. These values represent the less active, typically active, and the more active, and can only be equated to lifestyles in very general categories: i.e. sedentary, low and moderate activity. In practice, some sedentary individuals may have energy requirements slightly lower than implied by the less active PAL value and some active individuals may have energy requirements higher than implied by the more active PAL value. Judgements must be made where it is deemed necessary to identify energy requirements more precisely. The likely extra energy required for specific activities can aid such judgements (see Table 13).
171
The EAR values for adults reported by COMA were calculated from median 462. body weights from a 1980 survey of adult heights and weights in Great Britain equivalent to BMI values of 24.1 and 24. 9 kg/m2 for men and 23.1 and 24.7 kg/m2 for women, for ages 19-49 and 50-59 years respectively. Although not discussed as such, COMA identified weight-maintaining energy reference values and gave only brief consideration to a healthy PAL value (i.e. an additional 0.1 PAL was suggested for exercise associated with maintenance of cardiovascular health, increasing PAL from 1.4 in a previous non-active group to 1.5).
In contrast, in this SACN report the increasing prevalence of overweight and 463. obesity has been recognised and the new energy reference values are prescriptive i.e. they are calculated for body weights which are consistent with long-term good health. In adults, a healthy body weight range is generally defined as a BMI between 18.5-24.9kg/m2, however, for the purposes of calculating the revised EAR values, healthy body weights are identified as body weights equivalent to a BMI of 22.5kg/m2 from current mean heights for men and women (see paragraph 92). This approach is similar in principle to that taken in the FAO/WHO/UNU report (FAO, 2004) which was also prescriptive (or “normative”), recommending that reference EAR values should be identified in relation to height and listing values which would maintain BMI between 18.5 and 24.9 kg/m2.
In this report, the influence of age on energy requirements is limited to that 464. associated with the fall with age in BMR per kg body weight. The recommendations in terms of PAL values, therefore, are the same for healthy mobile free-living older adults as for younger adults. This is in recognition that an increasing fraction of older adults can and do remain physically active. Evidence suggests that individuals who are able to maintain higher levels of physical activity of any sort will gain benefit in terms of lower mortality (Manini et al., 2006). With an increasing lack of mobility, energy expenditure and requirements will fall so that PAL values at or below the lower quartile (i.e. PAL=1.49) will become more appropriate. For those who are immobile, falling food intakes associated with reduced energy requirements increase the potential for nutrient deficiencies. As a result nutrient dense food becomes particularly important for this population group.
Despite being calculated using a PAL value about 16% higher than that used by 465. COMA (DH, 1991), the revised all-adult energy reference values reported here are 3% higher for men and 7-9% higher for women (see Table 39). This is explained by the use of a 5-7% lower average body weight for the all-age category of adult men and women. It should be noted that these revised values are within the range of measurement error.
172
Table 39 Comparison of COMAa and new SACN mean adult population EAR values
Age group (years)
Weight (kg)
COMA EAR (MJ/d)c
SACN EAR (MJ/d)d
% change for SACN values
All men 69.2b – 10.9 –
19-49 74.0 10.6 – 2.8
50-59 74.0 10.6 – 2.8
All women 58.7b – 8.7 –
19-49 60.0 8.1 – 7.4
50-59 63.0 8.0 – 8.7 a DH, 1991 b Weights calculated for a Body Mass Index of 22.5kg.m2 at mean heights reported in the 2009 Health Survey for England (NHS IC, 2010). c From Table 2.8 COMA (DH, 1991) based on PAL=1.4 and BMR calculated from mean bodyweights from National Diet and Nutrition Survey 1986/1987 d New SACN summary values for adults at PAL = 1.63
Table 40 compares the new SACN EAR values with values calculated using the 466. approach adopted by COMA at the same healthy weights and PAL of 1.63 used by SACN to allow a more like for like comparison of the different methodologies used. The differences are the result of the Henry BMR prediction equations (Henry, 2005) used in the present SACN report (based on height and weight) providing slightly lower estimates of BMR than the weight-based Schofield equations (about 4% for men under 65 years and 3% for women).
173
Table 40 Comparison of the new SACN EAR values with values calculated using the approach adopted by COMAa at the same weights and PAL value used by SACN
Age group (years)
Height (cm) Weight (kg) BMIb =
22.5 kg/m2
COMA EAR (MJ/d)c
SACN EAR (MJ/d)d
% change for SACN values
Men
19-24 178 71.5 12.1 11.6 -4.1
25-34 178 71.0 12.0 11.5 -4.2
35-44 176 69.7 11.4 11.0 -3.5
45-54 175 68.8 11.3 10.8 -4.4
55-64 174 68.3 11.3 10.8 -4.4
65-74 173 67.0 9.4 9.8 4.3
75+ 170 65.1 9.2 9.6 4.3
Women
19-24 163 59.9 9.4 9.1 -3.2
25-34 163 59.7 9.4 9.1 -3.2
35-44 163 59.9 9.1 8.8 -3.3
45-54 162 59.0 9.0 8.8 -2.2
55-64 161 58.0 9.0 8.7 -3.3
65-74 159 57.2 8.0 8.0 0
75+ 155 54.3 7.9 7.7 -2.5 a DH, 1991 b Body Mass index (BMI) c EAR= PAL (1.63) x Basal Metabolic Rate (BMR), with BMR calculated using the modified Schofield equations (DH, 1991) d EAR= PAL (1.63) x BMR, with BMR calculated using the Henry equation (see Appendix 4)
Comparison with COMA report for pregnancy The COMA report recommended an increment in EAR of 0.8MJ/d above the pre-467. pregnant EAR only during the last trimester. In the absence of sufficient evidence to revise this recommendation, SACN considered it prudent to retain the EAR for pregnancy set by COMA. For pregnant women, unlike other adults, SACN has set energy requirements (EAR’s) at actual pre-conceptional body weights, rather than at healthy body weights (see paragraph 129).
Comparison with COMA report for lactation The COMA report recommended an increment of 1.9-2.4MJ/day (454-574kcal/day 468. additional energy for the first six months of lactation and recognised two distinctive groups of breastfeeding mothers firstly, women who practised exclusive or almost exclusive breastfeeding until the baby was 3-4 months old and then progressively introduced complementary foods as part of an active complementary feeding process which often lasted only a few months (Group 1). Secondly, women who introduced only limited complementary feeds after 3-4 months and whose intention was that breast milk should provide the primary source of nourishment for 6 months or more (Group 2).
174
The energy reference value identified in this report for lactation, an increment 469. of 1.38MJ/day (330kcal/day) for the first six months, is the same as previously recommended in the US DRI report (IoM, 2005), but considerably lower than that recommended for the first six months by COMA. The COMA values are higher due to the application of an efficiency factor of 0.8, to adjust for the efficiency of conversion of maternal energy intake to milk, which is insecure and likely to overestimate the true synthetic cost.
175
Glossary
Accelerometry A non-calorimetric method of assessing free living energy expenditure by monitoring activity and movement.
Adenosine triphosphate (ATP)
The cofactor that acts as an intermediate between catabolic, anabolic and energy expenditure reactions
Adipose tissue Body fat storage tissue. Distributed under the skin and around body organs. Composed of cells that synthesise and store fat, releasing it for metabolism in fasting.
Anthropometry Body measurements made non-invasively to assess body composition, physiological development and nutritional status.
Average Used here as a general term embracing both median and mean
Basal Metabolic Rate (BMR)
Rate at which the body uses energy when it is at complete rest, when food and physical activity have minimal influence on metabolism.
Body Mass Index (BMI)
An index of fatness and obesity of older children and adults. Weight in kilograms divided by the square of height in meters.
Dietary Reference Value (DRV)
A term used to define the various expressions of estimated dietary requirements in individuals and population groups. DRVs comprise 3 levels of intake Lower Reference Nutrient Intake, Reference Nutrient Intake and Estimated Average Requirement (see glossary entry).
Direct calorimetry Calorimetry is a method of energy expenditure measurement. Direct calorimetry is a measure of heat output from the body, as an index of energy expenditure.
Doubly labelled water (DLW)
The stable (non radioactive) isotope method for estimating energy expenditure in free living individuals over extended periods up to several weeks. Subjects consume water containing isotopes hydrogen (2H2) and oxygen (18O).
Energy balance The difference between metabolisable energy intake and total energy expenditure. A neutral energy balance occurs when energy intake is equal to energy expenditure.
176
Energy flux Energy flux or turnover is the rate at which energy in all its forms flows through the body on a daily basis (chemical, work, thermal). For an individual in energy balance the energy flux is numerically the same as energy expenditure or energy intake. Provided energy intake matches energy expenditure then energy balance will be maintained whether the energy flux itself is at a higher or lower level.
Estimated Average Requirement (EAR)
Estimated Average Requirement of a group of people for energy or protein or a vitamin or mineral. About half of a defined population will usually need more than the EAR, and half less.
Excess Post exercise oxygen consumption (EPOC)
Excess oxygen consumption following exercise. Small increase in energy expenditure following exercise which persists for some time after the exertion itself has been completed.
Fat Mass The component of body composition made up of fat.
Fat Free Mass (FFM) The non fat component of body composition comprising muscle, bone, skin and organs.
Food Frequency Questionnaire
A method for assessing past dietary intake. A questionnaire asking the frequency of consumption of foods over a day/ week/month etc.
Gross Energy (GE) The total maximum amount of energy contained within food, determined by measuring the heat released after complete combustion to carbon dioxide and water.
Glycogen The storage carbohydrate in the liver and muscles. A branched polymer of glucose units.
Heart rate monitoring (HRM)
A method for measuring free living energy expenditure. Heart rate is monitored minute by minute throughout the day using portable meters. The energy expenditure at a given heart rate can be estimated using an individual linear regression line of the relationship between oxygen consumption and heart rate.
Heat Increment of Feeding
See Themic Effect of Food.
Homeostasis The control of key components, (such as temperature and blood constituent concentrations) to ensure consistency and physiological normalisation.
Hyperglycaemia Elevated plasma concentration of glucose, caused by failure of the normal hormonal mechanisms of blood glucose control.
Hypoglycaemia Abnormally low concentration of plasma glucose.
177
Indirect calorimetry Calorimetry is the measurement of energy expenditure. Indirect calorimetry is the most commonly used approach. It is the calculation of energy expenditure by the measurement of oxygen consumption and carbon dioxide production.
Insulin resistance Reduction in the biological activity of insulin sensitive peripheral tissues, which results in reduced disposal of glucose from plasma for any given concentration of insulin.
Kilocalorie (kcal) Units used to measure the energy value of food, 1kcal = 4.18kJ
kilojoule (kJ)/ megajoule (MJ)
Units used to measure the energy value of food, 1kJ=1000 joules, 1MJ = 1 million joules
Lipolysis The breakdown of fat molecules e.g. triglycerides.
Metabolisable energy (ME)
The energy contained within food that is available to human metabolism as ATP after digestion and absorption.
Metabolic equivalent (MET)
Specific unit of measurement of energy expenditure by the body: equivalent to the BMR of an average adult; 1 MET = 3.5 ml O2. kg-1. min-1
Net Metabolisable energy (NME)
The ATP producing capacity of foods which is available to the body excluding unavoidable energy use in nutrient absorption and excretion of waste products.
Non exercise activity thermogenesis (NEAT)
Increase in energy expenditure due to non volitional activity/ behaviours such as fidgeting, muscle tone, posture maintenance.
Normative See prescriptive.
Physical Activity Level (PAL)
Daily total energy expenditure (TEE) expressed as multiple of Basal metabolic rate (BMR). It is calculated as TEE divided by BMR.
Physical activity- related energy expenditure (PAEE)
The component of energy expenditure related to physical activity.
Physical Activity Ratio (PAR)
Energy cost of different physical activities per unit of time expressed as a multiple of BMR.
Prescriptive Relating to an ideal standard: e.g. of body weight or physical activity. Sometimes used interchangeably with normative. As distinct from status quo.
178
Reference Nutrient Intake (RNI)
The Reference Nutrient Intake for a nutient is the amount of the nutrient that is enough, or more than enough, for about 97% of people in a group. If the average intake of a group is at the RNI, then the risk of deficiency in the group is very small.
Resting metabolic rate (RMR)
Rate at which the body uses energy when it is at rest. Sometimes used interchangeably with BMR but RMR is not measured at the standardised metabolic state and can include TEF and EPOC.
Thermic Effect of Food (TEF)
The increase in heat production by the body after eating, due to both the metabolic energy cost of ingestion, digestion and the energy cost of forming tissue reserves of fat, glycogen and protein.
Total Energy Expenditure (TEE)
The sum of all the energy expended by an individual over the course of one day. It includes BMR, PAEE and TEF and represents the average amount of energy spent in a typical day.
US DRI (dietary reference intakes)
US term for dietary reference values (includes the terms average requirement, Recommended Daily Amount and tolerable upper levels for supplements)
179
Glossary of statistical terms
Confidence intervals Gives a range around the estimate of a mean from a sample such that if samples of the same size are taken repeatedly from the same population and a confidence interval is calculated for each sample then 95% of these intervals should contain the true population mean.
Determination, Coefficient of
Also referred to as R-squared value. The square of the product moment correlation between two variables, so called because it expresses the proportion of the variance of one variable, Y, given by the other, X, when Y is expressed as a linear regression on X. More generally, if a dependent variable has multiple correlation R with a set of independent variables, R-squared is known as the coefficient of determination.
Linear Regression Analysis
Simple linear regression – A statistical method that attempts • to explain the relationship between a dependent variable and a single independent variable by fitting a linear equation to the observed data Multiple linear regression – The regression of a dependent • variable on more than one independent variable.
Mean The sum of the observations divided by the number of observations: similar to the median only if the data set is normally distributed
Unweighted mean The mean of a set of observations in which no weights are attached to them, except in the trivial sense that each is weighted equally
Median The midpoint or 50th centile of a distribution
Multiple regression techniques
Techniques by which multiple linear regression models are assessed to determine the independent variables that have the greatest influence on the dependent variable
Post-hoc Analysis Refers to statistical analysis after the data has been collected and investigating trends that were not specified before the study was conducted.
Residuals A term denoting a quantity remaining after some other quantity has been subtracted. In terms of regression modelling the values by which the observations differ from the model values are called residuals.
180
Standard Error of Estimates (SEE)
A measure (estimate) of the accuracy of predictions (the estimated standard deviation of the error in the model). Note that the true value is unknown, by definition, so the standard error of an estimate is itself an estimate.
181
Abbreviations
ATP Adenosine Triphosphate BMI Body Mass Index BMR Basal Metabolic Rate COMA Committee on Medical Aspects of Food and Nutrition Policy CV Coefficient of Variation DLW Doubly Labelled Water DRI Dietary Reference Intake DRV Dietary Reference Value EAR Estimated Average Requirement EE Energy Expenditure EFS Expenditure Food Survey EI Energy Intake EPOC Excess Post Exercise Oxygen Consumption FAO/WHO/UNU Food and Agriculture Organization/World Health
Organization/United Nations University FFM Fat Free Mass FFQ Food Frequency Questionnaire FM Fat Mass GDA Guideline Daily Amount GE Gross Energy GWG Gestational Weight Gain HR Heart Rate HRM Heart Rate Monitor HSE Health Survey for England J Joule kJ Kilojoule ME Metabolisable Energy MET Metabolic Energy Equivalent MJ Megajoule NDNS National Diet and Nutrition Survey NFS National Food Survey NME Net Metabolisable Energy NEAT Non-Exercise Activity Thermogenesis PA Physical Activity PAEE Physical Activity Energy Expenditure PAL Physical Activity Level PAR Physical Activity Ratio REE Resting Energy Expenditure RMR Resting Metabolic Rate RQ Respiration Quotient SACN Scientific Advisory Committee on Nutrition SEE Standard Error of Estimate
182
SI System of Units SPA Spontaneous Physical Activity TEE Total Energy Expenditure
TEF Thermic Effect of Food
183
Reference List
Abbott RA, Davies PS (2004) Habitual physical activity and physical activity intensity: their relation to body composition in 5.0-10.5-y-old children. Eur J Clin Nutr 58(2): 285-291.
Addolorato G, Capristo E, Greco AV, Stefanini GF & Gasbarrini G (1998) Influence of chronic alcohol abuse on body weight and energy metabolism: is excess ethanol consumption a risk factor for obesity or malnutrition? Journal of Internal Medicine 244, 387-395.
Al Adsani H, Hoffer LJ & Silva JE (1997) Resting Energy Expenditure is Sensitive to Small Dose Changes in Patients on Chronic Thyroid Hormone Replacement. J Clin Endocrinol Metab 82, 1118-1125.
Albu J, Shur M, Curi M, Murphy L, Heymsfield SB & Pi-Sunyer FX (1997) Resting metabolic rate in obese, premenopausal black women. Am J Clin Nutr 66, 531-538.
Alfonzo-Gonzalez G, Doucet E, Almeras N, Bouchard C & Tremblay A (2004) Estimation of daily energy needs with the FAO/WHO/UNU 1985 procedures in adults: comparison to whole-body indirect calorimetry measurements. Eur J Clin Nutr 58, 1125-1131.
Anderson SE, Bandini LG, Dietz WH & Must A (2004) Relationship between temperament, nonresting energy expenditure, body composition, and physical activity in girls. Int J Obes Relat Metab Disord 28, 300-306.
Andreasen CH, Stender-Petersen KL, Mogensen MS et al (2008) Low Physical Activity Accentuates the Effect of the FTO rs9939609 Polymorphism on Body Fat Accumulation. Diabetes 57(1), 95-101.
Arciero PJ, Gardner AW, Calles-Escandon J, Benowitz NL & Poehlman ET (1995) Effects of caffeine ingestion on NE kinetics, fat oxidation, and energy expenditure in younger and older men. Am J Physiol 268, E1192-E1198.
Arciero PJ, Goran MI & Poehlman ET (1993) Resting metabolic rate is lower in women than in men. J Appl Physiol 75, 2514-2520.
Aresu M, Bécares L, Brage S, et al. (2009) Health Survey for England - 2008: Physical activity and fitness: The NHS Information Centre for Health and Social Care.
Arthur PG, Hartmann PE & Smith M (1987) Measurement of the milk intake of breast-fed infants. J Pediatr Gastroenterol Nutr 6, 758-763.
Arvidsson D, Slinde F & Hulthen L (2005) Physical activity questionnaire for adolescents validated against doubly labelled water. Eur J Clin Nutr 59, 376-383.
Astrup A (2006) How to maintain a healthy body weight. Int J Vitam Nutr Res 76, 208-215.
184
Astrup A, Buemann B, Christensen NJ, Madsen J, Gluud C, Bennett P & Svenstrup B (1992) The contribution of body composition, substrates, and hormones to the variability in energy expenditure and substrate utilization in premenopausal women. J Clin Endocrinol Metab 74, 279-286.
Astrup A, Toubro S, Cannon S, Hein P, Breum L & Madsen J (1990) Caffeine: a double- blind, placebo-controlled study of its thermogenic, metabolic, and cardiovascular effects in healthy volunteers. Am J Clin Nutr 51, 759-767.
Astrup A, Toubro S, Dalgaard LT, Urhammer SA, Sorensen TI & Pedersen O (1999) Impact of the v/v 55 polymorphism of the uncoupling protein 2 gene on 24-h energy expenditure and substrate oxidation. Int J Obes Relat Metab Disord 23, 1030-1034.
Atlantis E, Barnes EH & Singh MAF (2006) Efficacy of exercise for treating overweight in children and adolescents: a systematic review. Int J Obes 30, 1027-1040.
Bailey BW, Tucker LA, Peterson TR & LeCheminant JD (2007) A prospective study of physical activity intensity and change in adiposity in middle-aged women. Am J Health Promot 21, 492-497.
Ball EJ, O’Connor J, Abbott R, Steinbeck KS, Davies PS, Wishart C, Gaskin KJ & Baur LA (2001) Total energy expenditure, body fatness, and physical activity in children aged 6-9 y. Am J Clin Nutr 74, 524-528.
Ball K, Mishra G & Crawford D (2002) Which aspects of socioeconomic status are related to obesity among men and women? Int J Obes Relat Metab Disord 26, 559-565.
Bandini LG, Must A, Spadano JL & Dietz WH (2002) Relation of body composition, parental overweight, pubertal stage, and race-ethnicity to energy expenditure among premenarcheal girls. Am J Clin Nutr 76, 1040-1047.
Bandini LG, Schoeller DA, Cyr HN & Dietz WH (1990a) Validity of reported energy intake in obese and nonobese adolescents. Am J Clin Nutr 52, 421-425.
Bandini LG, Schoeller DA & Dietz WH (1990b) Energy expenditure in obese and nonobese adolescents. Pediatr Res 27, 198-203.
Baranowski T, Baranowski JC, Cullen KW, Thompson DI, Nicklas T, Zakeri IE & Rochon J (2003) The Fun, Food, and Fitness Project (FFFP): the Baylor GEMS pilot study. Ethn Dis 13, S30-S39.
Basterfield L, Adamson AJ, Parkinson KN, Maute U, Li PX, Reilly JJ & the Gateshead Millennium Study Core Team (2008) Surveillance of physical activity in the UK is flawed: validation of the Health Survey for England Physical Activity Questionnaire. Arch Dis Child 93, 1054-1058.
Bell AC, Ge K & Popkin BM (2001) Weight gain and its predictors in Chinese adults. Int J Obes Relat Metab Disord 25, 1079-1086.
Bensimhon DR, Kraus WE & Donahue MP (2006) Obesity and physical activity: a review. Am Heart J 151, 598-603.
185
Berentzen T, Kring SII, Holst C, Zimmermann E, Jess T, Hansen T, Pedersen O, Toubro S, Astrup A & Sorensen TIA (2008) Lack of Association of Fatness-Related FTO Gene Variants with Energy Expenditure or Physical Activity. J Clin Endocrinol Metab 93, 2904-2908.
Berkey CS, Rockett HR, Field AE, Gillman MW, Frazier AL, Camargo CA, Jr. & Colditz GA (2000) Activity, dietary intake, and weight changes in a longitudinal study of preadolescent and adolescent boys and girls. Pediatrics 105, E56.
Berkey CS, Rockett HRH, Gillman MW & Colditz GA (2003) One-Year Changes in Activity and in Inactivity Among 10- to 15-Year-Old Boys and Girls: Relationship to Change in Body Mass Index. Pediatrics 111, 836-843.
Bernstein RS, Thornton JC, Yang MU, Wang J, Redmond AM, Pierson RN, Jr., Sunyer FX & Van Itallie TB (1983) Prediction of the resting metabolic rate in obese patients. Am J Clin Nutr 37, 595-602.
Bessey PQ & Wilmore DW (1988) The burned patient. In Nutrition and Metabolism in Patient Care, pp. 672-700. London: W.B. Saunders.
Bingham SA & Day N (2006) Commentary: Fat and breast cancer: time to re-evaluate both methods and results? Int J Epidemiol 35, 1022-1024
Bingham SA, Goldberg GR, Coward WA, Prentice AM & Cummings JH (1989) The effect of exercise and improved physical fitness on basal metabolic rate. Br J Nutr 61, 155-173.
Bingham SA, Luben R, Welch A, Wareham N, Khaw KT & Day N (2003) Are imprecise methods obscuring a relation between fat and breast cancer? Lancet 362, 212-214.
Bingham SA, Welch AA, McTaggart A, Mulligan AA, Runswick SA, Luben R, Oakes S, Khaw KT, Wareham N & Day NE (2001) Nutritional methods in the European Prospective Investigation of Cancer in Norfolk. Public Health Nutrition 4, 847-858.
Bisdee JT, James WP & Shaw MA (1989) Changes in energy expenditure during the menstrual cycle. Br J Nutr 61, 187-199.
Bitar A, Fellmann N, Vernet J, Coudert J & Vermorel M (1999) Variations and determinants of energy expenditure as measured by whole-body indirect calorimetry during puberty and adolescence. Am J Clin Nutr 69, 1209-1216.
Blaak EE, Westerterp KR, Bar-Or O, Wouters LJ & Saris WH (1992) Total energy expenditure and spontaneous activity in relation to training in obese boys. Am J Clin Nutr 55, 777-782.
Black AE, Coward WA, Cole TJ & Prentice AM (1996) Human energy expenditure in affluent societies: an analysis of 574 doubly-labelled water measurements. Eur J Clin Nutr 50, 72-92.
Black AE (1996) Physical activity levels from a meta-analysis of doubly labeled water studies for validating energy intake as measured by dietary assessment. Nutr Rev 54, 170-174.
186
Black AE & Cole TJ (2000) Within- and between-subject variation in energy expenditure measured by the doubly-labelled water technique: implications for validating reported dietary energy intake. Eur J Clin Nutr 54, 386-394.
Black AE, Prentice AM, Goldberg GR, Jebb SA, Bingham SA, Livingstone MB & Coward WA (1993) Measurements of total energy expenditure provide insights into the validity of dietary measurements of energy intake. J Am Diet Assoc 93, 572-579.
Blair SN, LaMonte MJ & Nichaman MZ (2004) The evolution of physical activity recommendations: how much is enough? Am J Clin Nutr 79, 913S-9920.
Blanc S, Schoeller DA, Bauer D, Danielson ME, Tylavsky F, Simonsick EM, Harris TB, Kritchevsky SB & Everhart JE (2004) Energy requirements in the eighth decade of life. Am J Clin Nutr 79, 303-310.
Blaza S & Garrow JS (1983) Thermogenic response to temperature, exercise and food stimuli in lean and obese women, studied by 24 h direct calorimetry. Br J Nutr 49, 171-180.
Blundell JE, Stubbs RJ, Hughes DA, Whybrow S & King NA (2003) Cross talk between physical activity and appetite control: does physical activity stimulate appetite? Proc Nutr Soc 62, 651-661.
Bobbioni-Harsch E, Assimacopoulos-Jeannet F, Lehmann T, Munger R, Allaz AF & Golay A (1999) Leptin plasma levels as a marker of sparing-energy mechanisms in obese women. Int J Obes Relat Metab Disord 23, 470-475.
Bogaert N, Steinbeck KS, Baur LA, Brock K & Bermingham MA (2003) Food, activity and family – environmental vs biochemical predictors of weight gain in children. Eur J Clin Nutr 57, 1242-1249.
Bogardus C, Lillioja S, Ravussin E, Abbott W, Zawadzki JK, Young A, Knowler WC, Jacobowitz R & Moll PP (1986) Familial dependence of the resting metabolic rate. N Engl J Med 315, 96-100.
Boreham C, Robson PJ, Gallagher AM, Cran GW, Savage JM & Murray LJ (2004) Tracking of physical activity, fitness, body composition and diet from adolescence to young adulthood: The Young Hearts Project, Northern Ireland. Int J Behav Nutr Phys Act 1, 14.
Bosy-Westphal A, Eichhorn C, Kutzner D, Illner K, Heller M & Muller MJ (2003) The Age- Related Decline in Resting Energy Expenditure in Humans Is Due to the Loss of Fat-Free Mass and to Alterations in Its Metabolically Active Components. J Nutr 133, 2356-2362.
Bouchard C, Tremblay A, Nadeau A, Despres JP, Theriault G, Boulay MR, Lortie G, Leblanc C & Fournier G (1989) Genetic effect in resting and exercise metabolic rates. Metabolism 38, 364-370.
Bratteby LE, Sandhagen B, Fan H & Samuelson G (1997) A 7-day activity diary for assessment of daily energy expenditure validated by the doubly labelled water method in adolescents. Eur J Clin Nutr 51, 585-591.
187
Bratteby LE, Sandhagen B & Samuelson G (2005) Physical activity, energy expenditure and their correlates in two cohorts of Swedish subjects between adolescence and early adulthood. Eur J Clin Nutr 59, 1324-1334.
Bryant MJ, Lucove JC, Evenson KR & Marshall S (2007) Measurement of television viewing in children and adolescents: a systematic review. Obes Rev 8, 197-209.
Buemann B, Astrup A, Christensen NJ & Madsen J (1992) Effect of moderate cold exposure on 24-h energy expenditure: similar response in postobese and nonobese women. Am J Physiol 263, E1040-E1045.
Buhl KM, Gallagher D, Hoy K, Matthews DE & Heymsfield SB (1995) Unexplained disturbance in body weight regulation: diagnostic outcome assessed by doubly labeled water and body composition analyses in obese patients reporting low energy intakes. J Am Diet Assoc 95, 1393-1400.
Burke V, Giangiulio N, Gillam HF, Beilin LJ & Houghton S (2003) Physical activity and nutrition programs for couples: a randomized controlled trial. J Clin Epidemiol 56, 421-432.
Butte NF (2005) Energy requirements of infants. Public Health Nutr 8, 953-967.
Butte NF & Ellis KJ (2003) Comment on “Obesity and the Environment: Where Do We Go from Here?”. Science 301, 598b.
Butte NF, Ellis KJ, Wong WW, Hopkinson JM, Smith EO (2003a) Composition of gestational weight gain impacts maternal fat retention and infant birth weight. Am J Obstet Gynecol 189(5):1423-1432.
Butte NF & Hopkinson JM (1998) Body composition changes during lactation are highly variable among women. J Nutr 128, 381S-385S.
Butte NF, Hopkinson JM, Wong WW, Smith EO & Ellis KJ (2000b) Body composition during the first 2 years of life: an updated reference. Pediatr Res 47, 578-585.
Butte NF & King JC (2002) Energy requirements during pregnancy and lactation. Energy background paper prepared for the joint FAO/WHO/UNU Consultation on Energy in Human Nutrition.
Butte NF & King JC (2005) Energy requirements during pregnancy and lactation. Public Health Nutr 8, 1010-1027.
Butte NF, Treuth MS, Mehta NR, Wong WW, Hopkinson JM & Smith EO (2003b) Energy requirements of women of reproductive age. Am J Clin Nutr 77, 630-638.
Butte NF, Wong WW, Ferlic L, Smith EO, Klein PD & Garza C (1990) Energy expenditure and deposition of breast-fed and formula-fed infants during early infancy. Pediatr Res 28, 631-640.
Butte NF, Wong WW & Garza C (1989) Energy cost of growth during infancy. Proceedings of the Nutrition Society 48, 303-312.
188
Butte NF, Wong WW & Hopkinson JM (2001) Energy requirements of lactating women derived from doubly labeled water and milk energy output. J Nutr 131, 53-58.
Butte NF, Wong WW, Hopkinson JM, Heinz CJ, Mehta NR & Smith EO (2000a) Energy requirements derived from total energy expenditure and energy deposition during the first 2 y of life. Am J Clin Nutr 72, 1558-1569.
Butte NF, Wong WW, Treuth MS, Ellis KJ & O’Brian Smith E (2004) Energy requirements during pregnancy based on total energy expenditure and energy deposition. Am J Clin Nutr 79, 1078-1087.
Byrne NM, Hills AP, Hunter GR, Weinsier RL & Schutz Y (2005) Metabolic equivalent: one size does not fit all. J Appl Physiol 99, 1112-1119.
Byrne NM, Weinsier RL, Hunter GR, Desmond R, Patterson MA, Darnell BE & Zuckerman PA (2003) Influence of distribution of lean body mass on resting metabolic rate after weight loss and weight regain: comparison of responses in white and black women. Am J Clin Nutr 77, 1368-1373.
Caballero B, Davis S, Davis CE, Ethelbah B, Evans M, Lohman T, Stephenson L, Story M, White J (1998) Pathways: A school-based program for the primary prevention of obesity in American Indian chidlren J Nutr Biochem 9, 535-543.
Caballero B, Clay T, Davis SM, et al. (2003) Pathways: a school-based, randomized controlled trial for the prevention of obesity in American Indian schoolchildren. Am J Clin Nutr 78, 1030-1038.
Cai G, Cole SA, Butte NF, Voruganti VS, Comuzzie AG (2008) Genome-wide scan revealed genetic loci for energy metabolism in Hispanic children and adolescents. Int J Obes 32(4):579-585.
Cairella G, Romagnoli F, Cantarelli P, Valentini P & Tarsitani G (1998) School oriented intervention on dietary education: Results of phase 1. Int J Obes Relat Metab Disord 22, S254 (Abstract).
Carpenter WH, Fonong T, Toth MJ, Ades PA, Calles-Escandon J, Walston JD & Poehlman ET (1998) Total daily energy expenditure in free-living older African-Americans and Caucasians. Am J Physiol Endocrinol Metab 274, 96-101.
Carpenter WH, Poehlman ET, O’Connell M & Goran MI (1995) Influence of body composition and resting metabolic rate on variation in total energy expenditure: a meta-analysis. Am J Clin Nutr 61, 4-10.
Casper RC, Schoeller DA, Kushner R, Hnilicka J, Gold ST (1991) Total daily energy expenditure and activity level in anorexia nervosa. Am J Clin Nutr 53(5), 1143-1150.
Caspersen CJ, Powell KE & Christenson GM (1985) Physical activity, exercise, and physical fitness: definitions and distinctions for health-related research. Public Health Rep 100, 126-131.
189
Cecil JE, Tavendale R, Watt P, Hetherington MM & Palmer CNA (2008) An Obesity- Associated FTO Gene Variant and Increased Energy Intake in Children. N Engl J Med 359, 2558-2566.
Champagne CM, Baker NB, DeLany JP, Harsha DW & Bray GA (1998) Assessment of energy intake underreporting by doubly labeled water and observations on reported nutrient intakes in children. J Am Diet Assoc 98, 426-433.
Chief Medical Officers of England, Scotland, Wales and Northern Ireland (2011). Start Active, Stay Active. A report on physical activity for health from the four home countries’ Chief Medical Officers. Department of Health, Welsh Government, The Scottish Government, Department of Health SSaPS.
Chitwood LF, Brown SP, Lundy MJ & Dupper MA (1996) Metabolic propensity toward obesity in black vs white females: responses during rest, exercise and recovery. Int J Obes Relat Metab Disord 20, 455-462.
Chong PK, Jung RT, Scrimgeour CM & Rennie MJ (1994) The effect of pharmacological dosages of glucocorticoids on free living total energy expenditure in man. Clin Endocrinol (Oxf) 40, 577-581.
Cole TJ (1995) Conditional reference charts to assess weight gain in British infants. Arch Dis Child 73, 8-16.
Cole TJ, Bellizzi MC, Flegal KM, Dietz WH (2000) Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ 320 (7244), 1240-1243.
Cole TJ, Faith MS, Pietrobelli A & Heo M (2005) What is the best measure of adiposity change in growing children: BMI, BMI %, BMI z-score or BMI centile? Eur J Clin Nutr 59, 419-425.
Collins LC, Cornelius MF, Vogel RL, Walker JF & Stamford BA (1994) Effect of caffeine and/or cigarette smoking on resting energy expenditure. Int J Obes Relat Metab Disord 18, 551-556.
Cooper AR, Moore LA, McKenna J & Riddoch C (2000) What is the magnitude of blood pressure response to a programme of moderate intensity exercise? Randomised controlled trial among sedentary adults with unmedicated hypertension. Br J Gen Pract 50, 958-962.
Corbett J, Dobbie F, Doig M, et al. (2010) Scottish Health Survey 2009 - Volume 1: Main report: The Scottish Government.
Corder K, Brage S, Mattocks C, Ness A, Riddoch C, Wareham NJ & Ekelund U (2007) Comparison of two methods to assess PAEE during six activities in children. Med Sci Sports Exerc 39, 2180-2188.
Correa A, Gilboa SM, Besser LM et al. Diabetes mellitus and birth defects. (2008) Am J Obstet Gynecol 199(3):237-239.
190
Craig SB, Bandini LG, Lichtenstein AH, Schaefer EJ & Dietz WH (1996) The impact of physical activity on lipids, lipoproteins, and blood pressure in preadolescent girls. Pediatrics 98, 389-395.
Cunningham JJ (1980) A reanalysis of the factors influencing basal metabolic rate in normal adults. Am J Clin Nutr 33, 2372-2374.
Cunningham JJ (1991) Body composition as a determinant of energy expenditure: a synthetic review and a proposed general prediction equation. Am J Clin Nutr 54, 963-969.
Danforth E Jr & Burger A (1984) The role of thyroid hormones in the control of energy expenditure. Clin Endocrinol Metab 13, 581-595.
Das SK, Moriguti JC, McCrory MA, Saltzman E, Mosunic C, Greenberg AS & Roberts SB (2001) An Underfeeding Study in Healthy Men and Women Provides Further Evidence of Impaired Regulation of Energy Expenditure in Old Age. J Nutr 131, 1833-1838.
Das SK, Saltzman E, McCrory MA, Hsu LK, Shikora SA, Dolnikowski G, Kehayias JJ & Roberts SB (2004) Energy expenditure is very high in extremely obese women. J Nutr 134, 1412-1416.
Dauncey MJ (1981) Influence of mild cold on 24 h energy expenditure, resting metabolism and diet-induced thermogenesis. Br J Nutr 45, 257-267.
Davidson L, McNeill G, Haggarty P, Smith JS & Franklin MF (1997) Free-living energy expenditure of adult men assessed by continuous heart-rate monitoring and doubly- labelled water. Br J Nutr 78, 695-708.
Davies PS, Day JM & Lucas A (1991a) Energy expenditure in early infancy and later body fatness. Int J Obes 15, 727-731.
Davies PS, Livingstone MB, Prentice AM, Coward WA, Jagger SE, Stewart CM & Strain JJ (1991b) Total energy expenditure during childhood and adolescence. Proc Nutr Soc 50, 1-14A.
Davies PS & White A (1995) Energy expenditure in children aged 1.5 to 4.5 years: a comparison with current recommendations for energy intake. Eur J Clin Nutr, 360-364.
Davis S, Gomez Y, Lambert L & Skipper B (1993) Primary prevention of obesity in American Indian children. In Pevention and treatment of childhood obesity, pp. 167-180 [CL Williams and SYS Kimm, editors]. New York: New York Academy of Sciences.
Davison KK & Birch LL (2001) Child and parent characteristics as predictors of change in girls’ body mass index. Int J Obes Relat Metab Disord 25, 1834-1842.
Davison KK, Marshall SJ & Birch LL (2006) Cross-sectional and longitudinal associations between TV viewing and girls’ body mass index, overweight status, and percentage of body fat. J Pediatr 149, 32-37.
191
Day DS, Gozansky WS, Van Pelt RE, Schwartz RS & Kohrt WM (2005) Sex hormone suppression reduces resting energy expenditure and {beta}-adrenergic support of resting energy expenditure. J Clin Endocrinol Metab 90, 3312-3317.
DeLany JP, Bray GA, Harsha DW & Volaufova J (2002) Energy expenditure in preadolescent African American and white boys and girls: the Baton Rouge Children’s Study. Am J Clin Nutr 75, 705-713.
DeLany JP, Bray GA, Harsha DW & Volaufova J (2004) Energy expenditure in African American and white boys and girls in a 2-y follow-up of the Baton Rouge Children’s Study. Am J Clin Nutr 79, 268-273.
DeLany JP, Bray GA, Harsha DW & Volaufova J (2006) Energy expenditure and substrate oxidation predict changes in body fat in children. Am J Clin Nutr 84, 862-870.
Demattia L, Lemont L & Meurer L (2007) Do interventions to limit sedentary behaviours change behaviour and reduce childhood obesity? A critical review of the literature. Obes Rev 8, 69-81.
Dempsey DT, Guenter P, Mullen JL, Fairman R, Crosby LO, Spielman G & Gennarelli T (1985) Energy expenditure in acute trauma to the head with and without barbiturate therapy. Surg Gynecol Obstet 160, 128-134.
Dennison BA, Russo TJ, Burdick PA & Jenkins PL (2004) An intervention to reduce television viewing by preschool children. Arch Pediatr Adolesc Med 158, 170-176.
Department of Health (1991) Dietary Reference Values for Food Energy and Nutrients for the United Kingdom no. 41. London: HMSO.
Department of Health (2004) At least five a week. Evidence on the impact of physical activity and its relationship to health. A report from the Chief Medical Officer. London: Department of Health.
Deriaz O, Dionne F, Perusse L et al. (1994) DNA variation in the genes of the Na,K- adenosine triphosphatase and its relation with resting metabolic rate, respiratory quotient, and body fat. J Clin Invest 93(2):838-843.
Di Blasio AM, van Rossum EF, Maestrini S et al (2003) The relation between two polymorphisms in the glucocorticoid receptor gene and body mass index, blood pressure and cholesterol in obese patients. Clin Endocrinol (Oxf) 59(1), 68-74.
Diepvens K, Westerterp KR & Westerterp-Plantenga MS (2007) Obesity and thermogenesis related to the consumption of caffeine, ephedrine, capsaicin, and green tea. Am J Physiol Regul Integr Comp Physiol 292, R77-R85.
Dietz WH (1998) Does energy expenditure affect changes in body fat in children? Am J Clin Nutr 67, 190-191.
Dietz WH & Robinson TN (1998) Use of the body mass index (BMI) as a measure of overweight in children and adolescents. J Pediatr 132, 191-193.
192
Diffey B, Piers LS, Soares MJ & O’Dea K (1997) The effect of oral contraceptive agents on the basal metabolic rate of young women. Br J Nutr 77, 853-862.
Dina C, Meyre D, Gallina S, et al. (2007) Variation in FTO contributes to childhood obesity and severe adult obesity. Nat Genet 39, 724-726.
Dionne I, Despres JP, Bouchard C & Tremblay A (1999) Gender difference in the effect of body composition on energy metabolism. Int J Obes Relat Metab Disord 23, 312-319.
Dionne IJ, Garant MJ, Nolan AA, Pollin TI, Lewis DG, Shuldiner AR & Poehlman ET (2002) Association between obesity and a polymorphism in the beta(1)-adrenoceptor gene (Gly389Arg ADRB1) in Caucasian women. Int J Obes Relat Metab Disord 26, 633-639.
Dionne IJ, Turner AN, Tchernof A, Pollin TI, Avrithi D, Gray D, Shuldiner AR & Poehlman ET (2001) Identification of an interactive effect of beta3- and alpha2b-adrenoceptor gene polymorphisms on fat mass in Caucasian women. Diabetes 50, 91-95.
Do R, Bailey SD, Desbiens K, Belisle A, Montpetit A, Bouchard C, Perusse L, Vohl MC & Engert JC (2008) Genetic Variants of FTO Influence Adiposity, Insulin Sensitivity, Leptin Levels, and Resting Metabolic Rate in the Quebec Family Study. Diabetes 57, 1147-1150.
Donnelly JE, Jacobsen DJ, Whatley JE, Hill JO, Swift LL, Cherrington A, Polk B, Tran ZV & Reed G (1996) Nutrition and physical activity program to attenuate obesity and promote physical and metabolic fitness in elementary school children. Obes Res 4, 229-243.
Drøyvold WB, Holmen J, Midthjell K & Lydersen S (2004) BMI change and leisure time physical activity (LTPA): an 11-y follow-up study in apparently healthy men aged 20-69 y with normal weight at baseline. Int J Obes Relat Metab Disord 28, 410-417.
Durnin JV (1996) Energy requirements: general principles. Eur J Clin Nutr 50 Suppl 1, S2-S9.
Eck LH, Bennett AG, Egan BM, Ray JW, Mitchell CO, Smith MA & Klesges RC (1997) Differences in macronutrient selections in users and nonusers of an oral contraceptive. Am J Clin Nutr 65, 419-424.
Eddy TP, Nicholson AL & Wheeler EF (1965) Energy expenditures and dietary intakes in cerebral palsy. Dev Med Child Neurol 7, 377-386.
Ekelund U, Aman J, Yngve A, Renman C, Westerterp K & Sjostrom M (2002) Physical activity but not energy expenditure is reduced in obese adolescents: a case-control study. Am J Clin Nutr 76, 935-941.
Ekelund U, Brage S, Besson H, Sharp S, Wareham NJ (2008) Time spent being sedentary and weight gain in healthy adults: reverse or bidirectional causality? Am J Clin Nutr 88(3), 612-617.
Ekelund U, Sardinha LB, Anderssen SA, Harro M, Franks PW, Brage S, Cooper AR, Andersen LB, Riddoch C & Froberg K (2004b) Associations between objectively assessed physical activity and indicators of body fatness in 9- to 10-y-old European children: a
193
population-based study from 4 distinct regions in Europe (the European Youth Heart Study). Am J Clin Nutr 80, 584-590.
Ekelund U, Yngve A, Brage S, Westerterp K & Sjostrom M (2004a) Body movement and physical activity energy expenditure in children and adolescents: how to adjust for differences in body size and age. Am J Clin Nutr 79, 851-856.
Ekelund U, Sjostrom M, Yngve A et al. Physical activity assessed by activity monitor and doubly labeled water in children. Med Sci Sports Exerc 2001;33(2):275-281.
Elder SJ & Roberts SB (2007) The effects of exercise on food intake and body fatness: a summary of published studies. Nutr Rev 65, 1-19.
Elia M (1992) Organ and tissue contribution to metabolic rate. In Energy Metabolism: Tissue Determinants and Cellular Corollaries, pp. 61-79. New York: Raven Press.
Elia M (1995) Changing concepts of nutrient requirements in disease: implications for artificial nutritional support. Lancet 345, 1279-1284.
Elia M (1997) Tissue distribution and energetics in weigh loss and undernutrition. In Physiology,Stress and Malnutrition: Functional Correlates and Nutritional Intervention [HN Tucker and JM Kinney, editors]. New York: Lipcott-Raven.
Elia M (2000) Hunger disease: metabolic response to starvation, injury and sepsis. Clin Nutr 19, 379-386.
Elia M (2005) Insights into energy requirements in disease. Public Health Nutr 8, 1037-1052.
Elia M, Ritz P & Stubbs RJ (2000) Total energy expenditure in the elderly. Eur J Clin Nutr 54 Suppl 3, S92-103.
Erlichman J, Kerbey AL & James WP (2002) Physical activity and its impact on health outcomes. Paper 2: Prevention of unhealthy weight gain and obesity by physical activity: an analysis of the evidence. Obes Rev 3, 273-287.
FAO (2003) Food energy - methods of analysis and conversion factors. Report of a Technical Workshop no. 77. Rome: FAO.
FAO (2004) Human energy requirements. Report of a Joint FAO/WHO/UNU Expert Consultation no. 1. Rome: FAO.
Fellows IW, Bennett T & Macdonald IA (1985) The effect of adrenaline upon cardiovascular and metabolic functions in man. Clin Sci (Lond) 69, 215-222.
Ferraro R, Lillioja S, Fontvieille AM, Rising R, Bogardus C & Ravussin E (1992) Lower sedentary metabolic rate in women compared with men. J Clin Invest 90, 780-784.
Ferraro R & Ravussin E (1992) Fat mass in predicting resting metabolic rate. Am J Clin Nutr 56, 460-461.
194
Figueroa-Colon R, Arani RB, Goran MI & Weinsier RL (2000) Paternal body fat is a longitudinal predictor of changes in body fat in premenarcheal girls. Am J Clin Nutr 71, 829-834.
Filozof CM, Murua C, Sanchez MP, Brailovsky C, Perman M, Gonzalez CD & Ravussin E (2000) Low plasma leptin concentration and low rates of fat oxidation in weight-stable post-obese subjects. Obes Res 8, 205-210.
Finch S, Doyle W, Lowe C, Gregory J, Bates CJ, Prentice A, Smithers G & Clarke PC (1998) National Diet and Nutrition Survey: people aged 65 years and over. Volume 1: report of the diet and nutrition survey. London: The Stationary Office.
Fitzgibbon ML, Stolley MR & Kirschenbaum DS (1995) An obesity prevention pilot program in African-American mothers and daughters. J Nutr Educ 27, 93-97.
Fogelholm M & Kukkonen-Harjula K (2000) Does physical activity prevent weight gain – a systematic review. Obes Rev 1, 95-111.
Fomom SJ, Haschke F, Ziegler EE & Nelson SE (1982) Body composition of reference children from birth to age 10 years. Am J Clin Nutr 35, 1169-1175.
Fontvieille AM, Harper IT, Ferraro RT, Spraul M & Ravussin E (1993) Daily energy expenditure by five-year-old children, measured by doubly labeled water. J Pediatr 123, 200-207.
Food Standards Agency (2002) McCance and Widdowson’s the composition of foods, 6 ed. Cambridge: Royal Society of Chemistry.
Foresight (2007) Tackling obesities: future choices. Government Office of Science. www. foresight.gov.uk/Obesity/14.pdf.
Forman JN, Miller WC, Szymanski LM & Fernhall B (1998) Differences in resting metabolic rates of inactive obese African-American and Caucasian women. Int J Obes Relat Metab Disord 22, 215-221.
Forster JL, Jeffery RW, Schmid TL & Kramer FM (1988) Preventing weight gain in adults: a pound of prevention. Health Psychol 7, 515-525.
Forsum E, Kabir N, Sadurskis A & Westerterp K (1992) Total energy expenditure of healthy Swedish women during pregnancy and lactation. Am J Clin Nutr 56, 334-342.
Forsum E, Lof M & Schoeller DA (2006) Calculation of energy expenditure in women using the MET system. Med Sci Sports Exerc 38, 1520-1525.
Foster GD, Wadden TA, Swain RM, Anderson DA & Vogt RA (1999) Changes in resting energy expenditure after weight loss in obese African American and white women. Am J Clin Nutr 69, 13-17.
Foster GD, Wadden TA & Vogt RA (1997) Resting energy expenditure in obese African American and Caucasian women. Obes Res 5, 1-8.
195
Fox KR & Hillsdon M (2007) Physical activity and obesity. Obes Rev 8 Suppl 1, 115-121.
Francis LA, Lee Y & Birch LL (2003) Parental weight status and girls’ television viewing, snacking, and body mass indexes. Obes Res 11, 143-151.
Frayling TM, Timpson NJ, Weedon MN, et al. (2007) A common variant in the FTO gene is associated with body mass index and predisposes to childhood and adult obesity. Science 316, 889-894.
Freedson P, Pober D & Janz KF (2005) Calibration of accelerometer output for children. Med Sci Sports Exerc 37, S523-S530.
Freeman JV, Cole TJ, Chinn S, Jones PR, White EM & Preece MA (1995) Cross sectional stature and weight reference curves for the UK, 1990. Arch Dis Child 73, 17-24.
Fukagawa NK, Bandini LG & Young JB (1990) Effect of age on body composition and resting metabolic rate. Am J Physiol 259, E233-E238.
Fuller NJ, Sawyer MB, Coward WA, Paxton P & Elia M (1996) Components of total energy expenditure in free-living elderly men (over 75 years of age): measurement, predictability and relationship to quality-of-life indices. Br J Nutr 75, 161-173.
Gagnon J, Mauriege P, Roy S, Sjostrom D, Chagnon YC, Dionne FT, Oppert JM, Perusse L, Sjostrom L & Bouchard C (1996) The Trp64Arg mutation of the beta3 adrenergic receptor gene has no effect on obesity phenotypes in the Quebec Family Study and Swedish Obese Subjects cohorts. J Clin Invest 98, 2086-2093.
Gallagher D, Belmonte D, Deurenberg P, Wang Z, Krasnow N, Pi-Sunyer FX & Heymsfield SB (1998) Organ-tissue mass measurement allows modeling of REE and metabolically active tissue mass. Am J Physiol 275, E249-E258.
Gallagher D, Albu J, He Q et al. Small organs with a high metabolic rate explain lower resting energy expenditure in African American than in white adults. Am J Clin Nutr 2006;83(5):1062-1067.
Gannon B, DiPietro L & Poehlman ET (2000) Do African Americans have lower energy expenditure than Caucasians? Int J Obes Relat Metab Disord 24, 4-13.
Garby L & Lammert O (1994) Between-subjects variation in energy expenditure: estimation of the effect of variation in organ size. Eur J Clin Nutr 48, 376-378.
Garza C & Butte NF (1986) Energy concentration of human milk estimated from 24-h pools and various abbreviated sampling schemes. J Pediatr Gastroenterol Nutr 5, 943-948.
Gerken T, Girard CA, Tung YC et al (2007) The obesity-associated FTO gene encodes a 2-oxoglutarate-dependent nucleic acid demethylase. Science 318(5855), 1469-1472.
Gibney ER, Murgatroyd P, Wright A, Jebb S & Elia M (2003) Measurement of total energy expenditure in grossly obese women: comparison of the bicarbonate-urea method with
196
whole-body calorimetry and free-living doubly labelled water. Int J Obes Relat Metab Disord 27, 641-647.
Gittelsohn J, Evans M, Helitzer D, Anliker J, Story M, Metcalfe L, Davis S & Iron CP (1998) Formative research in a school-based obesity prevention program for Native American school children (Pathways). Health Educ Res 13, 251-265.
Goldberg GR, Black AE, Jebb SA, Cole TJ, Murgatroyd PR, Coward WA & Prentice AM (1991a) Critical evaluation of energy intake data using fundamental principles of energy physiology: 1. Derivation of cut-off limits to identify under-recording. Eur J Clin Nutr 45, 569-581.
Goldberg GR, Prentice AM, Coward WA, Davies HL, Murgatroyd PR, Sawyer MB, Ashford J & Black AE (1991b) Longitudinal assessment of the components of energy balance in well-nourished lactating women. Am J Clin Nutr 54, 788-798.
Goldberg GR, Prentice AM, Coward WA, Davies HL, Murgatroyd PR, Wensing C, Black AE, Harding M & Sawyer M (1993) Longitudinal assessment of energy expenditure in pregnancy by the doubly labeled water method. Am J Clin Nutr 57, 494-505.
Goran MI (2005) Estimating energy requirements: regression based prediction equations or multiples of resting metabolic rate. Public Health Nutr 8, 1184-1186.
Goran MI, Beer WH, Wolfe RR, Poehlman ET & Young VR (1993a) Variation in total energy expenditure in young healthy free-living men. Metabolism 42, 487-496.
Goran MI, Calles-Escandon J, Poehlman ET, O’Connell M & Danforth E Jr (1994b) Effects of increased energy intake and/or physical activity on energy expenditure in young healthy men. J Appl Physiol 77, 366-372.
Goran MI, Carpenter WH & Poehlman ET (1993b) Total energy expenditure in 4- to 6-yr- old children. Am J Physiol 264, E706-E711.
Goran MI, Gower BA, Nagy TR & Johnson RK (1998c) Developmental changes in energy expenditure and physical activity in children: evidence for a decline in physical activity in girls before puberty. Pediatrics 101, 887-891.
Goran MI, Kaskoun M & Johnson R (1994a) Determinants of resting energy expenditure in young children. J Pediatr 125, 362-367.
Goran MI, Kaskoun M, Johnson R, Martinez C, Kelly B & Hood V (1995) Energy expenditure and body fat distribution in Mohawk children. Pediatrics 95, 89-95.
Goran MI, Nagy TR, Gower BA, Mazariegos M, Solomons N, Hood V & Johnson R (1998a) Influence of sex, seasonality, ethnicity, and geographic location on the components of total energy expenditure in young children: implications for energy requirements. Am J Clin Nutr 68, 675-682.
Goran MI, Shewchuk R, Gower BA, Nagy TR, Carpenter WH & Johnson RK (1998b) Longitudinal changes in fatness in white children: no effect of childhood energy expenditure. Am J Clin Nutr 67, 309-316.
197
Gordon-Larsen P, Hou N, Sidney S, Sternfeld B, Lewis CE, Jacobs DR, Jr. & Popkin BM (2009) Fifteen-year longitudinal trends in walking patterns and their impact on weight change. Am J Clin Nutr 89, 19-26.
Goris AH, Westerterp-Plantenga MS & Westerterp KR (2000) Undereating and underrecording of habitual food intake in obese men: selective underreporting of fat intake1. Am J Clin Nutr 71, 130-134.
Grediagin A, Cody M, Rupp J, Benardot D & Shern R (1995) Exercise intensity does not effect body composition change in untrained, moderately overfat women. J Am Diet Assoc 95, 661-665.
Green JH. Asessment of energy requirements (1994). In: Consensus in Clinical Nutrition. Ed. Heatley RV, Green JH, Losowsky MS. Cambridge University Press.
Gregory J, Bates CJ, Lowe S, Prentice A, Jackson LV, Smithers G, Wenlock R & Farron M (2000) National diet and nutrition survey: Young people aged 4 to 18 years: Volume 1: Report of the diet and nutrition survey, Volume 1 ed. London: The Stationery Office.
Gregory JR, Collins DL, Davies PSW, Hughes JM & Clarke PC (1995) National Diet and Nutrition Survey: children aged 1½-4½ years. Volume 1: report of the diet and nutrition survey. London: HMSO.
Gregory J, Foster K, Tyler H & Wiseman M (1990) The Dietary and Nutritional Survey of British Adults. London: HMSO.
Guinhouya CB, Hubert H, Soubrier S, Vilhelm C, Lemdani M & Durocher A (2006) Moderate-to-vigorous physical activity among children: discrepancies in accelerometry- based cut-off points. Obesity (Silver Spring) 14, 774-777.
Haas V, Onur S, Paul T, Nutzinger DO, Bosy-Westphal A, Hauer M, Brabant G, Klein H & Muller MJ (2005) Leptin and body weight regulation in patients with anorexia nervosa before and during weight recovery. Am J Clin Nutr 81, 889-896.
Haggarty P, McNeill G, Manneh MK, Davidson L, Milne E, Duncan G & Ashton J (1994) The influence of exercise on the energy requirements of adult males in the UK. Br J Nutr 72, 799-813.
Haggarty P, Valencia ME, McNeill G, Gonzales NL, Moya SY, Pinelli A, Quihui L, Saucedo MS, Esparxa J, Ashton J, Milne E, James WPT (1997) Energy expenditure during heavy work and its interaction with body weight. Br J Nutr 77, 359-373.
Hall MB & Elliman D (2003) Health for all children. 4th Edition. Oxford University Press.
Haman F (2006) Shivering in the cold: from mechanisms of fuel selection to survival. J Appl Physiol 100, 1702-1708.
Han Z, Mulla S, Beyene J, Liao G & McDonald SD (2011) Maternal underweight and the risk of preterm birth and low birth weight: a systematic review and meta-analyses. Int J Epidemiol 40, 65-101.
198
Hancox RJ, Milne BJ & Poulton R (2004) Association between child and adolescent television viewing and adult health: a longitudinal birth cohort study. Lancet 364, 257-262.
Hancox RJ & Poulton R (2006) Watching television is associated with childhood obesity: but is it clinically important? Int J Obes (Lond) 30, 171-175.
Hansen M, Morthorst R, Larsson B, Flyvbjerg A, Rasmussen MH, Orskov H, Astrup A, Kjaer M & Lange KHW (2005) Effects of 2 wk of GH administration on 24-h indirect calorimetry in young, healthy, lean men. Am J Physiol Endocrinol Metab 289, E1030-E1038.
Hardeman W, Griffin S, Johnston M, Kinmonth AL & Wareham NJ (2000) Interventions to prevent weight gain: a systematic review of psychological models and behaviour change methods. Int J Obes Relat Metab Disord 24, 131-143.
Haslam D, Sattar N & Lean M (2006) Obesity--time to wake up. BMJ 333, 640-642.
Heal DJ, Aspley S, Prow MR, Jackson HC, Martin KF & Cheetham SC (1998) Sibutramine: a novel anti-obesity drug. A review of the pharmacological evidence to differentiate it from d-amphetamine and d-fenfluramine. Int J Obes Relat Metab Disord 22 Suppl 1, S18-S28.
Henderson L, Gregory J, Irving K & Swan G (2003a) The National Diet and Nutrition Survey: adults aged 19 to 64 years. Volume 2: Energy, protein, carbohydrate, fat and alcohol intake. London: TSO.
Henderson L, Gregory J & Swan G (2002) The National Diet and Nutrition Survey: adults aged 19 to 64 years. Volume 1: Types and quantities of foods consumed. London: TSO.
Henderson L, Irving K, Gregory J, Bates CJ, Prentice A, Perks J, Swan G & Farron M (2003b) The National Diet and Nutrition Survey: adults aged 19 to 64 years. Volume 3: Vitamin and mineral intake and urinary analytes. London: TSO.
Henry CJ (2005) Basal metabolic rate studies in humans: measurement and development of new equations. Public Health Nutr 8, 1133-1152.
Hernandez-Triana M, Salazar G, Diaz E, Sanchez V, Basabe B, Gonzalez S & Diaz ME (2002) Total energy expenditure by the doubly-labeled water method in rural preschool children in Cuba. Food Nutr Bull 23, 76-81.
Heslehurst N, Ells LJ, Simpson H, Batterham A, Wilkinson J, Summerbell CD (2007) Trends in maternal obesity incidence rates, demographic predictors, and health inequalities in 36,821 women over a 15-year period. BJOG 114(2):187-194.
Hessemer V & Bruck K (1985) Influence of menstrual cycle on thermoregulatory, metabolic, and heart rate responses to exercise at night. J Appl Physiol 59, 1911-1917.
Heymsfield SB, Gallagher D, Kotler DP, Wang Z, Allison DB & Heshka S (2002) Body- size dependence of resting energy expenditure can be attributed to nonenergetic homogeneity of fat-free mass. Am J Physiol Endocrinol Metab 282, E132-E138.
199
Hill JO (2006) Understanding and Addressing the Epidemic of Obesity: An Energy Balance Perspective. Endocr Rev 27, 750-761.
Hill JO (2009) Can a small-changes approach help address the obesity epidemic? A report of the Joint Task Force of the American Society Nutrition, Institute of Food Technologists, and International Food Information Council. Am J Clin Nutr 89, 477-484.
Hill JO, Wyatt HR, Reed GW & Peters JC (2003) Obesity and the Environment: Where Do We Go from Here? Science 299, 853-855.
Hinney A, Nguyen TT, Scherag A, et al. (2007) Genome wide association (GWA) study for early onset extreme obesity supports the role of fat mass and obesity associated gene (FTO) variants. PLoS ONE 2, e1361.
Hoare J, Henderson L, Bates CJ, Prentice A, Birch M, Swan G & Farron M (2004) The National Diet and Nutrition Survey: adults aged 19 to 64 years. Volume 5: Summary report. London: TSO.
Hoffman DJ, Sawaya AL, Coward WA, Wright A, Martins PA, de Nascimento C, Tucker KL & Roberts SB (2000) Energy expenditure of stunted and nonstunted boys and girls living in the shantytowns of Sao Paulo, Brazil. Am J Clin Nutr 72, 1031.
Hojlund K, Christiansen C, Bjornsbo KS, Poulsen P, Bathum L, Henriksen JE, Lammert O & Beck-Nielsen H (2006) Energy expenditure, body composition and insulin response to glucose in male twins discordant for the Trp64Arg polymorphism of the beta3- adrenergic receptor gene. Diabetes Obes Metab 8, 322-330.
Hoos MB, Plasqui G, Gerver WJ & Westerterp KR (2003) Physical activity level measured by doubly labeled water and accelerometry in children. Eur J Appl Physiol 89, 624-626.
Horn OK, Paradis G, Potvin L, Macaulay AC & Desrosiers S (2001) Correlates and predictors of adiposity among Mohawk children. Prev Med 33, 274-281.
Howe JC, Rumpler WV & Seale JL (1993) Energy expenditure by indirect calorimetry in premenopausal women: variation within one menstrual cycle. Journal of Nutritional Biochemistry 4, 268-273.
Hoyt RW & Friedl KE (2006) Field studies of exercise and food deprivation. Curr Opin Clin Nutr Metab Care 9, 685-690.
Hu FB, Li TY, Colditz GA, Willett WC & Manson JE (2003) Television watching and other sedentary behaviors in relation to risk of obesity and type 2 diabetes mellitus in women. JAMA 289, 1785-1791.
Hukshorn CJ, Menheere PP, Westerterp-Plantenga MS & Saris WH (2003a) The effect of pegylated human recombinant leptin (PEG-OB) on neuroendocrine adaptations to semi- starvation in overweight men. Eur J Endocrinol 148, 649-655.
Hukshorn CJ, Saris WHM, Westerterp-Plantenga MS, Farid AR, Smith FJ & Campfield LA (2000) Weekly Subcutaneous Pegylated Recombinant Native Human Leptin (PEG-OB) Administration in Obese Men. J Clin Endocrinol Metab 85, 4003-4009.
200
Hukshorn CJ, Westerterp-Plantenga MS & Saris WH (2003b) Pegylated human recombinant leptin (PEG-OB) causes additional weight loss in severely energy- restricted, overweight men. Am J Clin Nutr 77, 771-776.
Hunt SC, Stone S, Xin Y, Scherer CA, Magness CL, Iadonato SP, Hopkins PN & Adams TD (2008) Association of the FTO gene with BMI. Obesity (Silver Spring) 16, 902-904.
Hunter GR, Weinsier RL, Darnell BE, Zuckerman PA, Goran MI. Racial differences in energy expenditure and aerobic fitness in premenopausal women1. Am J Clin Nutr 2000;71(2):500-506.
Hytten FE & Chamberlain G (1991) Clinical physiology in obstetrics. Oxford, United Kingdom: Blackwell Scientific Publications.
IAEA (2009) Assessment of body composition and total energy expenditure in humans using stable isotope techniques. Vienna: International Atomic Energy Agency.
Illner K, Brinkmann G, Heller M, Bosy-Westphal A & Muller MJ (2000) Metabolically active components of fat free mass and resting energy expenditure in nonobese adults. Am J Physiol Endocrinol Metab 278, E308-E315.
Institute of Medicine Food and Nutrition Board (1990) Nutrition During Pregnancy: Part I: Weight Gain, Part II: Nutrient Supplements. Washington DC: National Academy Press.
Institute of Medicine (2005) Dietary Reference Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids. Washington, D.C.: National Academy Press.
Institute of Medicine Food and Nutrition Board (1990) Nutrition During Pregnancy: Part I: Weight Gain, Part II: Nutrient Supplements. Washington DC: National Academy Press.
Jago R, Baranowski T, Baranowski JC, Thompson D & Greaves KA (2005) BMI from 3-6 y of age is predicted by TV viewing and physical activity, not diet. Int J Obes (Lond) 29, 557-564.
Jakicic JM & Wing RR (1998) Differences in resting energy expenditure in African- American vs Caucasian overweight females. Int J Obes Relat Metab Disord 22, 236-242.
Jebb SA (2007) Dietary determinants of obesity. Obes Rev 8 Suppl 1, 93-97.
Jeffery RW & French SA (1997) Preventing weight gain in adults: design, methods and one year results from the Pound of Prevention study. Int J Obes Relat Metab Disord 21, 457-464.
Jessen AB, Toubro S & Astrup A (2003) Effect of chewing gum containing nicotine and caffeine on energy expenditure and substrate utilization in men. Am J Clin Nutr 77, 1442-1447.
Jiang Z, Yan Q, Su Y, Acheson KJ, Thelin A, Piguet-Welsch C, Ritz P & Ho ZC (1998) Energy expenditure of Chinese infants in Guangdong Province, south China, determined with use of the doubly labeled water method. Am J Clin Nutr 67, 1256-1264.
201
Johnson MS, Figueroa-Colon R, Herd SL, Fields DA, Sun M, Hunter GR & Goran MI (2000) Aerobic Fitness, Not Energy Expenditure, Influences Subsequent Increase in Adiposity in Black and White Children. Pediatrics 106, e50.
Johnson RK, Russ J & Goran MI (1998) Physical activity related energy expenditure in children by doubly labeled water as compared with the Caltrac accelerometer. Int J Obes Relat Metab Disord 22, 1046-1052.
Johnstone AM, Murison SD, Duncan JS, Rance KA & Speakman JR (2005) Factors influencing variation in basal metabolic rate include fat-free mass, fat mass, age, and circulating thyroxine but not sex, circulating leptin, or triiodothyronine. Am J Clin Nutr 82, 941-948.
Jones A, Jr., Shen W, St Onge MP, Gallagher D, Heshka S, Wang Z & Heymsfield SB (2004) Body-composition differences between African American and white women: relation to resting energy requirements. Am J Clin Nutr 79, 780-786.
Jorgensen JO, Vahl N, Dall R & Christiansen JS (1998) Resting metabolic rate in healthy adults: relation to growth hormone status and leptin levels. Metabolism 47, 1134-1139.
Jung RT, Shetty PS & James WP (1980) The effect of beta-adrenergic blockade on metabolic rate and peripheral thyroid metabolism in obesity. Eur J Clin Invest 10, 179-182.
Kain J, Uauy R, Albala, Vio F, Cerda R & Leyton B (2004) School-based obesity prevention in Chilean primary school children: methodology and evaluation of a controlled study. Int J Obes Relat Metab Disord 28, 483-493.
Kaplan AS, Zemel BS & Stallings VA (1996) Differences in resting energy expenditure in prepubertal black children and white children. J Pediatr 129, 643-647.
Karhunen L, Franssila-Kallunki A, Rissanen A, Kervinen K, Kesaniemi YA & Uusitupa M (1997) Determinants of resting energy expenditure in obese non-diabetic caucasian women. Int J Obes Relat Metab Disord 21, 197-202.
Kaskoun MC, Johnson RK & Goran MI (1994) Comparison of energy intake by semiquantitative food-frequency questionnaire with total energy expenditure by the doubly labeled water method in young children. Am J Clin Nutr 60, 43-47.
Kautiainen S, Koivusilta L, Lintonen T, Virtanen SM & Rimpela A (2005) Use of information and communication technology and prevalence of overweight and obesity among adolescents. Int J Obes (Lond) 29, 925-933.
Kennedy A, Gettys TW, Watson P, Wallace P, Ganaway E, Pan Q & Garvey WT (1997) The metabolic significance of leptin in humans: gender-based differences in relationship to adiposity, insulin sensitivity, and energy expenditure. J Clin Endocrinol Metab 82, 1293-1300.
Key TJ, Schatzkin A, Willett WC, Allen NE, Spencer EA & Travis RC (2004) Diet, nutrition and the prevention of cancer. Public Health Nutr 7, 187-200.
202
Keys A, Fidanza F, Karvonen MJ, Kimura N & Taylor HL (1972) Indices of relative weight and obesity. J Chronic Dis 25, 329-343.
Kim CH, Yun SK, Byun DW, Yoo MH, Lee KU, Suh KI (2001) Codon 54 polymorphism of the fatty acid binding protein 2 gene is associated with increased fat oxidation and hyperinsulinemia, but not with intestinal fatty acid absorption in Korean men. Metabolism 50(4), 473-476.
Kim J, Wang Z, Heymsfield SB, Baumgartner RN & Gallagher D (2002) Total-body skeletal muscle mass: estimation by a new dual-energy X-ray absorptiometry method. Am J Clin Nutr 76, 378-383.
Kimm SY, Barton BA, Obarzanek E, McMahon RP, Sabry ZI, Waclawiw MA, Schreiber GB, Morrison JA, Similo S & Daniels SR (2001) Racial divergence in adiposity during adolescence: The NHLBI Growth and Health Study. Pediatrics 107, E34.
Kimm SY, Glynn NW, Obarzanek E, Kriska AM, Daniels SR, Barton BA & Liu K (2005) Relation between the changes in physical activity and body-mass index during adolescence: a multicentre longitudinal study. Lancet 366, 301-307.
King NA, Appleton K, Rogers PJ & Blundell JE (1999) Effects of sweetness and energy in drinks on food intake following exercise. Physiol Behav 66, 375-379.
King NA, Caudwell P, Hopkins M, Byrne NM, Colley R, Hills AP, Stubbs JR & Blundell JE (2007) Metabolic and Behavioral Compensatory Responses to Exercise Interventions: Barriers to Weight Loss. Obesity Res 15, 1373-1383.
Kipnis V, Subar AF & Midthune D (2003) Structure of dietary measurement error: results of the OPEN biomarker study. Am J Epidemiol 158, 14-21.
Kirkby J, Metcalf BS, Jeffery AN, O’Riordan CF, Perkins J, Voss LD & Wilkin TJ (2004) Sex differences in resting energy expenditure and their relation to insulin resistance in children (EarlyBird 13). Am J Clin Nutr 80, 430-435.
Klannemark M, Orho M & Groop L (1998) No relationship between identified variants in the uncoupling protein 2 gene and energy expenditure. Eur J Endocrinol 139, 217-223.
Klausen B, Toubro S & Astrup A (1997) Age and sex effects on energy expenditure. Am J Clin Nutr 65, 895-907.
Kleiber M (1975) The Fire of Life. An Introduction to Animal Energetics. New York: Robert E. Krieger Publishing.
Klesges RC, Klesges LM, Eck LH & Shelton ML (1995) A longitudinal analysis of accelerated weight gain in preschool children. Pediatrics 95, 126-130.
Koh-Banerjee P, Chu NF, Spiegelman D, Rosner B, Colditz G, Willett W & Rimm E (2003) Prospective study of the association of changes in dietary intake, physical activity, alcohol consumption, and smoking with 9-y gain in waist circumference among 16 587 US men. Am J Clin Nutr 78, 719-727.
203
Kopp-Hoolihan LE, van Loan MD, Wong WW & King JC (1999) Longitudinal assessment of energy balance in well-nourished, pregnant women. Am J Clin Nutr 69, 697-704.
Krakoff J, Ma L, Kobes S et al (2008) Lower Metabolic Rate in Individuals Heterozygous for Either a Frameshift or a Functional Missense MC4R Variant. Diabetes 57(12), 3267-3272.
Kramer MS & Kakuma R (2003) Energy and protein intake in pregnancy (Review). Cochrane Database of Systematic Reviews, Art. No.:CD000032. DOI: 10.1002/14651858/ CD000032.
Kring SII, Holst C, Zimmermann E, Jess T, Berentzen T, Toubro S, ansen T, strup A, edersen O & orensen T (2008) FTO Gene Associated Fatness in Relation to Body Fat Distribution and Metabolic Traits throughout a Broad Range of Fatness. PLoS ONE 3, e2958.
Kubaszek A, Pihlajamaki J, Punnonen K, Karhapaa P, Vauhkonen I, Laakso M (2003) The C-174G promoter polymorphism of the IL-6 gene affects energy expenditure and insulin sensitivity. Diabetes 52(2), 558-561.
Kuip M, Hoos MB, Forget PP, Westerterp KR, Gemke RJ & Meer KD (2003) Energy expenditure in infants with congenital heart disease, including a meta-analysis. Acta Paediatrica 92, 921-927.
Kunz I, Klaus S, Kallies B, Schorr U & Sharma AM (2000) Kinetic analysis of the thermic effect of food and its relationship to body composition in humans. Metabolism 49, 1340-1345.
Kushner RF, Racette SB, Neil K & Schoeller DA (1995) Measurement of physical activity among black and white obese women. Obes Res 3 Suppl 2, 261s-265s.
Kvaavik E, Tell GS & Klepp KI (2003) Predictors and tracking of body mass index from adolescence into adulthood: follow-up of 18 to 20 years in the Oslo Youth Study. Arch Pediatr Adolesc Med 157, 1212-1218.
Kyle UG, Genton L, Hans D, Karsegard L, Slosman DO & Pichard C (2001) Age-related differences in fat-free mass, skeletal muscle, body cell mass and fat mass between 18 and 94 years. Eur J Clin Nutr 55, 663-672.
Lariviere F, Moussalli R & Garrel DR (1994) Increased leucine flux and leucine oxidation during the luteal phase of the menstrual cycle in women. Am J Physiol Endocrinol Metab 267, E422-E428.
Lawrence M, Thongprasert K & Durnin JV (1988) Between-group differences in basal metabolic rates: an analysis of data collected in Scotland, the Gambia and Thailand. Eur J Clin Nutr 42, 877-891.
Lean ME, Murgatroyd PR, Rothnie I, Reid IW & Harvey R (1988) Metabolic and thyroidal responses to mild cold are abnormal in obese diabetic women. Clin Endocrinol (Oxf) 28, 665-673.
204
Leenders NY, Sherman WM, Nagaraja HN & Kien CL (2001) Evaluation of methods to assess physical activity in free-living conditions. Med Sci Sports Exerc 33, 1233-1240.
Lemura LM & Maziekas MT (2002) Factors that alter body fat, body mass, and fat-free mass in pediatric obesity. Med Sci Sports Exerc 34, 487-496.
Levine JA (2005) Measurement of energy expenditure. Public Health Nutr 8, 1123-1132.
Levine JA (2007) Nonexercise activity thermogenesis-liberating the life-force. J Intern Med 262, 273-287.
Levine JA, Eberhardt NL & Jensen MD (1999) Role of nonexercise activity thermogenesis in resistance to fat gain in humans. Science 283, 212-214.
Levine JA, Harris MM & Morgan MY (2000) Energy expenditure in chronic alcohol abuse. Eur J Clin Invest 30, 779-786.
Levine JA (2004) Nonexercise activity thermogenesis (NEAT): environment and biology. Am J Physiol Endocrinol Metab 286, E675-E685.
Li ET, Tsang LB & Lui SS (1999) Resting metabolic rate and thermic effects of a sucrose- sweetened soft drink during the menstrual cycle in young Chinese women. Can J Physiol Pharmacol 77, 544-550.
Lichtman SW, Pisarska K, Berman ER, Pestone M, Dowling H, Offenbacher E, Weisel H, Heshka S, Matthews DE & Heymsfield SB (1992) Discrepancy between self-reported and actual caloric intake and exercise in obese subjects. N Engl J Med 327, 1893-1898.
Lindquist CH, Cummings T & Goran MI (2000) Use of tape-recorded food records in assessing children’s dietary intake. Obes Res 8, 2-11.
Littrell KH, Hilligoss NM, Kirshner CD, Petty RG & Johnson CG (2003) The effects of an educational intervention on antipsychotic-induced weight gain. J Nurs Scholarsh 35, 237-241.
Livingstone MB (1995) Assessment of food intakes: are we measuring what people eat? Br J Biomed Sci 52, 58-67.
Livingstone MB & Black AE (2003) Markers of the validity of reported energy intake. J Nutr 133 Suppl 3, 895S-920S.
Livingstone MB, Prentice AM, Coward WA, Strain JJ, Black AE, Davies PS, Stewart CM, McKenna PG & Whitehead RG (1992) Validation of estimates of energy intake by weighed dietary record and diet history in children and adolescents. Am J Clin Nutr 56, 29-35.
Long SJ, Hart K & Morgan LM (2002) The ability of habitual exercise to influence appetite and food intake in response to high- and low-energy preloads in man. Br J Nutr 87, 517-523.
205
Loos RJF, Rankinen T, Chagnon Y, Tremblay A, Perusse L & Bouchard C (2006) Polymorphisms in the leptin and leptin receptor genes in relation to resting metabolic rate and respiratory quotient in the Quebec Family Study. Int J Obes Relat Metab Disord 30, 183-190.
Lopez-Alarcon M, Merrifield J, Fields DA, Hilario-Hailey T, Franklin FA, Shewchuk RM, Oster RA & Gower BA (2004) Ability of the actiwatch accelerometer to predict free- living energy expenditure in young children. Obes Res 12, 1859-1865.
Lovejoy JC, Champagne CM, Smith SR, de Jonge L & Xie H (2001) Ethnic differences in dietary intakes, physical activity, and energy expenditure in middle-aged, premenopausal women: the Healthy Transitions Study. Am J Clin Nutr 74, 90-95.
Lovelady CA, Meredith CN, McCrory MA, Nommsen LA, Joseph LJ & Dewey KG (1993) Energy expenditure in lactating women: a comparison of doubly labeled water and heart-rate-monitoring methods. Am J Clin Nutr 57, 512-518.
Macdonald HM, New SA, Campbell MK & Reid DM (2003) Longitudinal changes in weight in perimenopausal and early postmenopausal women: effects of dietary energy intake, energy expenditure, dietary calcium intake and hormone replacement therapy. Int J Obes Relat Metab Disord 27, 669-676.
Mackintosh RM & Hirsch J (2001) The Effects of Leptin Administration in Non-obese Human Subjects. Obesity Res 9, 462-469.
Maes HH, Neale MC & Eaves LJ (1997) Genetic and environmental factors in relative body weight and human adiposity. Behav Genet 27, 325-351.
Maestrini S, Podesta F, Di Blasio AM, Savia G, Brunani A, Tagliaferri A, Mencarelli M, Chiodini I & Liuzzi A (2003) Lack of association between UCP2 gene polymorphisms and obesity phenotype in Italian Caucasians. J Endocrinol Invest 26, 985-990.
Maffeis C, Talamini G, Tato L (1998) Influence of diet, physical activity and parents’ obesity on children’s adiposity: a four-year longitudinal study. Int J Obes Relat Metab Disord 22(8), 758-764.
Magarey AM, Daniels LA, Boulton TJ & Cockington RA (2003) Predicting obesity in early adulthood from childhood and parental obesity. Int J Obes Relat Metab Disord 27, 505-513.
Magoffin A, Allen JR, McCauley J, Gruca MA, Peat J, Van Asperen P & Gaskin K (2008) Longitudinal analysis of resting energy expenditure in patients with cystic fibrosis. J Pediatr 152, 703-708.
Mamalakis G, Kafatos A, Manios Y, Anagnostopoulou T & Apostolaki I (2000) Obesity indicies in a cohort of primary school children in Crete: a six year prospective study. International Journal of Obesity and Related Metabolic Disorders 24, 765-771.
206
Manini TM, Everhart JE, Patel KV, Schoeller DA, Colbert LH, Visser M, Tylavsky F, Bauer DC, Goodpaster BH & Harris TB (2006) Daily Activity Energy Expenditure and Mortality Among Older Adults. JAMA 296, 171-179.
Mansell PI & Macdonald IA (1990) Reappraisal of the Weir equation for calculation of metabolic rate. Am J Physiol 258(6 Pt 2), R1347-54.
March of Dimes (2002) Nutrition Today Matters Tomorrow. A Report from The March of Dimes Task Force on Nutrition and Optimal Human Development.: March of Dimes Fulfillment Center, Pennsylvania.
Marken Lichtenbelt WD, Heidendal GA & Westerterp KR (1997) Energy expenditure and physical activity in relation to bone mineral density in women with anorexia nervosa. Eur J Clin Nutr 51, 826-830.
Marshall SJ, Biddle SJ, Gorely T, Cameron N & Murdey I (2004) Relationships between media use, body fatness and physical activity in children and youth: a meta-analysis. Int J Obes Relat Metab Disord 28, 1238-1246.
Martins C, Morgan L & Truby H (2008a) A review of the effects of exercise on appetite regulation: an obesity perspective. Int J Obes (Lond) 32, 1337-1347.
Martins C, Robertson MD & Morgan LM (2008b) Effects of exercise and restrained eating behaviour on appetite control. Proceedings of the Nutrition Society 67, 28-41.
Martins C, Truby H & Morgan LM (2007) Short-term appetite control in response to a 6-week exercise programme in sedentary volunteers. Br J Nutr 98, 834-842.
Mâsse LC, Fulton JE, Watson KL, Mahar MT, Meyers MC & Wong WW (2004) Influence of body composition on physical activity validation studies using doubly labeled water. J Appl Physiol 96, 1357-1364.
Mattocks C, Leary S, Ness A, Deere K, Saunders J, Kirkby J, Blair SN, Tilling K & Riddoch C (2007) Intraindividual variation of objectively measured physical activity in children. Med Sci Sports Exerc 39, 622-629.
Mayer J (1966) Some aspects of the problem of regulation of food intake and obesity. N Engl J Med 274, 722-731.
Mayer J & Thomas DW (1967) Regulation of food intake and obesity. Science 156, 328-337.
McDonald SD, Han Z, Mulla S & Beyene J (2010) Overweight and obesity in mothers and risk of preterm birth and low birth weight infants: systematic review and meta-analyses. BMJ 341, c3428.
McGloin AF, Livingstone MB, Greene LC, Webb SE, Gibson JM, Jebb SA, Cole TJ, Coward WA, Wright A & Prentice AM (2002) Energy and fat intake in obese and lean children at varying risk of obesity. Int J Obes Relat Metab Disord 26, 200-207.
207
McMurray RG, Harrell JS, Bangdiwala SI, Bradley CB, Deng S, Levine A (2002) A school- based intervention can reduce body fat and blood pressure in young adolescents. J Adolesc Health 31(2), 125-132.
Meijer GA, Westerterp KR, Saris WH & Ten Hoor F (1992) Sleeping metabolic rate in relation to body composition and the menstrual cycle. Am J Clin Nutr 55, 637-640.
Melanson KJ, Saltzman E, Russell R & Roberts SB (1996) Postabsorptive and postprandial energy expenditure and substrate oxidation do not change during the menstrual cycle in young women. J Nutr 126, 2531-2538.
Melanson KJ, Saltzman E, Vinken AG, Russell R & Roberts SB (1998) The effects of age on postprandial thermogenesis at four graded energetic challenges: findings in young and older women. J Gerontol A Biol Sci Med Sci 53, B409-B414.
Melzer K, Kayser B, Saris WH & Pichard C (2005) Effects of physical activity on food intake. pp. 885-895.
Miller WC, Koceja DM & Hamilton EJ (1997) A meta-analysis of the past 25 years of weight loss research using diet, exercise or diet plus exercise intervention. Int J Obes Relat Metab Disord 21, 941-947.
Molnar D & Livingstone B (2000) Physical activity in relation to overweight and obesity in children and adolescents. Eur J Pediatr 159 Suppl 1, S45-S55.
Montgomery C, Reilly JJ, Jackson DM, Kelly LA, Slater C, Paton JY & Grant S (2005) Validation of energy intake by 24-hour multiple pass recall: comparison with total energy expenditure in children aged 5-7 years. Br J Nutr 93, 671-676.
Moore LL, Gao D, Bradlee ML, Cupples LA, Sundarajan-Ramamurti A, Proctor MH, Hood MY, Singer MR & Ellison RC (2003) Does early physical activity predict body fat change throughout childhood? Prev Med 37, 10-17.
Moore LL, Nguyen US, Rothman KJ, Cupples LA & Ellison RC (1995) Preschool Physical Activity Level and Change in Body Fatness in Young Children: The Framingham Children’s Study. Am J Epidemiol 142, 982-988.
Morrison JA, Alfaro MP, Khoury P, Thornton BB & Daniels SR (1996) Determinants of resting energy expenditure in young black girls and young white girls. J Pediatr 129, 637-642.
Moshfegh AJ, Rhodes DG, Baer DJ, et al. (2008) The US Department of Agriculture Automated Multiple-Pass Method reduces bias in the collection of energy intakes. Am J Clin Nutr 88, 324-332.
Muller MJ, Bosy-Westphal A, Klaus S, et al. (2004) World Health Organization equations have shortcomings for predicting resting energy expenditure in persons from a modern, affluent population: generation of a new reference standard from a retrospective analysis of a German database of resting energy expenditure. Am J Clin Nutr 80, 1379-1390.
208
Muller MJ, Burger AG, Ferrannini E, Jequier E & Acheson KJ (1989) Glucoregulatory function of thyroid hormones: role of pancreatic hormones. Am J Physiol 256, E101-E110.
Mundt CA, Baxter-Jones AD, Whiting SJ, Bailey DA, Faulkner RA & Mirwald RL (2006) Relationships of activity and sugar drink intake on fat mass development in youths. Med Sci Sports Exerc 38, 1245-1254.
Murgatroyd PR, Goldberg GR, Leahy FE, Gilsenan MB & Prentice AM (1999) Effects of inactivity and diet composition on human energy balance. Int J Obes Relat Metab Disord 23, 1269-1275.
Must A, Bandini LG, Tybor DJ, Phillips SM, Naumova EN, Dietz WH (2007) Activity, inactivity, and screen time in relation to weight and fatness over adolescence in girls. Obesity (Silver Spring) 15(7), 1774-1781.
Must A & Tybor DJ (2005) Physical activity and sedentary behavior: a review of longitudinal studies of weight and adiposity in youth. Int J Obes (Lond) 29 Suppl 2, S84-S96.
Muto T, Yamauchi K (2001) Evaluation of a multicomponent workplace health promotion program conducted in Japan for improving employees’ cardiovascular disease risk factors. Prev Med 33(6), 571-577.
Nagai N, Sakane N, Ueno LM, Hamada T & Moritani T (2003) The -3826 A-->G variant of the uncoupling protein-1 gene diminishes postprandial thermogenesis after a high fat meal in healthy boys. J Clin Endocrinol Metab 88, 5661-5667.
Nagy TR, Gower BA, Shewchuk RM & Goran MI (1997) Serum Leptin and Energy Expenditure in Children. J Clin Endocrinol Metab 82, 4149-4153.
NAHRES-4 (1990) The Doubly-labelled Water Method for Measuring Energy Expenditure: A consensus Report by the IDECG working group. Vienna: IAEA.
National Health Service (NHS) Information Centre (2010) Health Survey for England - 2009 trend tables. The NHS Information Centre for health and social care.
Nelson KM, Weinsier RL, Long CL & Schutz Y (1992) Prediction of resting energy expenditure from fat-free mass and fat mass. Am J Clin Nutr 56, 848-856.
Nelson MC, Gordon-Larsen P, Adair LS & Popkin BM (2005) Adolescent physical activity and sedentary behavior: patterning and long-term maintenance. Am J Prev Med 28, 259-266.
Nelson MJ, Erens B, Bates B, Church S & Boshier T (2007) Low income diet and nutrition survey. London: TSO.
Neuhauser-Berthold M, Herbert BM, Luhrmann PM, Sultemeier AA, Blum WF, Frey J & Hebebrand J (2000) Resting metabolic rate, body composition, and serum leptin concentrations in a free-living elderly population. Eur J Endocrinol 142, 486-492.
209
Neumark-Sztainer D, Story M, Hannan PJ & Rex J (2003) New Moves: a school-based obesity prevention program for adolescent girls. Prev Med 37, 41-51.
NICE (2006) Obesity: the prevention, identification, assessment and management of overweight and obesity in adults and children.
NICE (2010) Weight management before, during and after pregnancy.
Nicklas BJ, Toth MJ, Goldberg AP & Poehlman ET (1997) Racial Differences in Plasma Leptin Concentrations in Obese Postmenopausal Women. J Clin Endocrinol Metab 82, 315-317.
Norman RA, Tataranni PA, Pratley R et al (1998) Autosomal genomic scan for loci linked to obesity and energy metabolism in Pima Indians. Am J Hum Genet 62(3):659-668.
Northern Ireland Statistics and Research Agency (2006) Health and Social Wellbeing Survey. Topline results 2005/06.
O’Connor J, Ball EJ, Steinbeck KS, Davies PS, Wishart C, Gaskin KJ & Baur LA (2001) Comparison of total energy expenditure and energy intake in children aged 6-9 y. Am J Clin Nutr 74, 643-649.
O’Loughlin J, Gray-Donald K, Paradis G & Meshefedjian G (2000) One- and two-year predictors of excess weight gain among elementary schoolchildren in multiethnic, low- income, inner-city neighborhoods. Am J Epidemiol 152, 739-746.
O’Rahilly S & Farooqi IS (2006) Genetics of obesity. Philos Trans R Soc Lond B Biol Sci 361, 1095-1105.
Obarzanek E, Lesem MD & Jimerson DC (1994) Resting metabolic rate of anorexia nervosa patients during weight gain. Am J Clin Nutr 60, 666-675.
Onur S, Haas V, Bosy-Westphal A, Hauer M, Paul T, Nutzinger D, Klein H & Muller MJ (2005) L-Tri-iodothyronine is a major determinant of resting energy expenditure in underweight patients with anorexia nervosa and during weight gain. Eur J Endocrinol 152, 179-184.
Oomen JM, Waijers PM, van Rossum C, Hoebee B, Saris WH & van Baak MA (2005) Influence of beta(2)-adrenoceptor gene polymorphisms on diet-induced thermogenesis. Br J Nutr 94, 647-654.
Oppert JM, Vohl MC, Chagnon M, Dionne FT, Cassard-Doulcier AM, Ricquier D, Perusse L & Bouchard C (1994) DNA polymorphism in the uncoupling protein (UCP) gene and human body fat. Int J Obes Relat Metab Disord 18, 526-531.
Pangrazi RP, Beighle A, Vehige T & Vack C (2003) Impact of Promoting Lifestyle Activity for Youth (PLAY) on children’s physical activity. J Sch Health 73, 317-321.
Pannemans DL & Westerterp KR (1995) Energy expenditure, physical activity and basal metabolic rate of elderly subjects. Br J Nutr 73, 571-581.
210
Parsons TJ, Manor O & Power C (2006) Physical activity and change in body mass index from adolescence to mid-adulthood in the 1958 British cohort. Int J Epidemiol 35, 197-204.
Parsons TJ, Manor O & Power C (2008) Television viewing and obesity: a prospective study in the 1958 British birth cohort. Eur J Clin Nutr 62, 1355-1363.
Pauly RP, Lear SA, Hastings FC & Birmingham CL (2000) Resting energy expenditure and plasma leptin levels in anorexia nervosa during acute refeeding. Int J Eat Disord 28, 231-234.
Pelkman CL, Chow M, Heinbach RA & Rolls BJ (2001) Short-term effects of a progestational contraceptive drug on food intake, resting energy expenditure, and body weight in young women. Am J Clin Nutr 73, 19-26.
Perkins KA (1992) Metabolic effects of cigarette smoking. J Appl Physiol 72, 401-409.
Perkins KA, Sexton JE & DiMarco A (1996) Acute thermogenic effects of nicotine and alcohol in healthy male and female smokers. Physiol Behav 60, 305-309.
Perkins KA, Sexton JE, Epstein LH, DiMarco A, Fonte C, Stiller RL, Scierka A & Jacob RG (1994) Acute thermogenic effects of nicotine combined with caffeine during light physical activity in male and female smokers. Am J Clin Nutr 60, 312-319.
Perks SM, Roemmich JN, Sandow-Pajewski M, Clark PA, Thomas E, Weltman A, Patrie J & Rogol AD (2000) Alterations in growth and body composition during puberty. IV. Energy intake estimated by the youth-adolescent food-frequency questionnaire: validation by the doubly labeled water method. Am J Clin Nutr 72, 1455-1460.
Petersen L, Schnohr P & Sorensen TI (2004) Longitudinal study of the long-term relation between physical activity and obesity in adults. Int J Obes Relat Metab Disord 28, 105-112.
Piers LS, Diggavi SN, Rijskamp J, van Raaij JM, Shetty PS & Hautvast JG (1995) Resting metabolic rate and thermic effect of a meal in the follicular and luteal phases of the menstrual cycle in well-nourished Indian women. Am J Clin Nutr 61, 296-302.
Piers LS, Soares MJ, McCormack LM & O’Dea K (1998) Is there evidence for an age- related reduction in metabolic rate? J Appl Physiol 85, 2196-2204.
Pietilainen KH, Kaprio J, Borg P, Plasqui G, Yki-Jarvinen H, Kujala UM, Rose RJ, Westerterp KR & Rissanen A (2008) Physical Inactivity and Obesity: A Vicious Circle. Obesity (Silver Spring) 16, 409-414.
Pietrobelli A, Faith MS, Allison DB, Gallagher D, Chiumello G, Heymsfield SB (1998) Body mass index as a measure of adiposity among children and adolescents: a validation study. J Pediatr 132(2), 204-210.
Plasqui G, Joosen AM, Kester AD, Goris AH & Westerterp KR (2005) Measuring free- living energy expenditure and physical activity with triaxial accelerometry. Obes Res 13, 1363-1369.
211
Plasqui G, Kester ADM & Westerterp KR (2003) Seasonal variation in sleeping metabolic rate, thyroid activity, and leptin. Am J Physiol Endocrinol Metab 285, E338-E343.
Plasqui G & Westerterp KR (2004) Seasonal Variation in Total Energy Expenditure and Physical Activity in Dutch Young Adults. Obesity Res 12, 688-694.
Poehlman ET, Melby CL & Badylak SF (1991) Relation of age and physical exercise status on metabolic rate in younger and older healthy men. J Gerontol 46, B54-B58.
Poehlman ET, Toth MJ, Ades PA & Calles-Escandon J (1997) Gender differences in resting metabolic rate and noradrenaline kinetics in older individuals. Eur J Clin Invest 27, 23-28.
Polley BA, Wing RR & Sims CJ (2002) Randomized controlled trial to prevent excessive weight gain in pregnant women. Int J Obes Relat Metab Disord 26, 1494-1502.
Poppitt SD, Swann D, Black AE, Prentice AM (1998) Assessment of selective under- reporting of food intake by both obese and non-obese women in a metabolic facility. Int J Obesity 22, 303-31.
Prentice AM (1995) Alcohol and obesity. Int J Obes Relat Metab Disord 19 Suppl 5, S44-S50.
Prentice AM (1998) Manipulation of dietary fat and energy density and subsequent effects on substrate flux and food intake. Am J Clin Nutr 67, 535S-5541.
Prentice AM, Black AE, Coward WA & Cole TJ (1996a) Energy expenditure in overweight and obese adults in affluent societies: an analysis of 319 doubly-labelled water measurements. Eur J Clin Nutr 50, 93-97.
Prentice AM, Black AE, Coward WA, Davies HL, Goldberg GR, Murgatroyd PR, Ashford J, Sawyer M & Whitehead RG (1986) High levels of energy expenditure in obese women. Br Med J (Clin Res Ed) 292, 983-987.
Prentice AM & Goldberg GR (2000) Energy adaptations in human pregnancy: limits and long-term consequences. Am J Clin Nutr 71, 1226S-1232
Prentice AM, Goldberg GR, Murgatroyd PR & Cole TJ (1996c) Physical activity and obesity: problems in correcting expenditure for body size. Int J Obes Relat Metab Disord 20, 688-691.
Prentice AM & Jebb SA (2004) Energy intake/physical activity interactions in the homeostasis of body weight regulation. Nutr Rev 62, S98-104.
Prentice AM, Lucas A, Vasquez-Velasquez L, Davies PS & Whitehead RG (1988) Are current dietary guidelines for young children a prescription for overfeeding? Lancet 2, 1066-1069.
Prentice AM, Spaaij CJ, Goldberg GR, Poppitt SD, van Raaij JM, Totton M, Swann D & Black AE (1996b) Energy requirements of pregnant and lactating women. Eur J Clin Nutr 50 Suppl 1, S82-110.
212
Proctor MH, Moore LL, Gao D, Cupples LA, Bradlee ML, Hood MY & Ellison RC (2003) Television viewing and change in body fat from preschool to early adolescence: The Framingham Children’s Study. Int J Obes Relat Metab Disord 27, 827-833.
Proper KI, Hildebrandt VH, van der Beek AJ, Twisk JW, van Mechelen W (2003) Effect of individual counseling on physical activity fitness and health: a randomized controlled trial in a workplace setting. Am J Prev Med 24(3), 218-226.
Prospective Studies Collaboration, Whitlock G, Lewington S, Sherliker P, Clarke R, Emberson J, Halsey J, Qizilbash N, Collins R & Peto R (2009) Body-mass index and cause- specific mortality in 900 000 adults: collaborative analyses of 57 prospective studies Lancet 373, 1083-1096.
Raben A, Agerholm-Larsen L, Flint A, Holst JJ & Astrup A (2003) Meals with similar energy densities but rich in protein, fat, carbohydrate, or alcohol have different effects on energy expenditure and substrate metabolism but not on appetite and energy intake. Am J Clin Nutr 77, 91-100.
Rainwater DL, Mitchell BD, Comuzzie AG, VandeBerg JL, Stern MP & Maccluer JW (2000) Association among 5-year changes in weight, physical activity, and cardiovascular disease risk factors in Mexican Americans. Am J Epidemiol 152, 974-982.
Rampersaud E, Mitchell BD, Pollin TI et al (2008) Physical Activity and the Association of Common FTO Gene Variants With Body Mass Index and Obesity. Arch Intern Med 168(16), 1791-1797.
Ravussin E, Harper IT, Rising R & Bogardus C (1991) Energy expenditure by doubly labeled water: validation in lean and obese subjects. Am J Physiol 261, E402-E409.
Ravussin E, Lillioja S, Anderson T, Christin L & Bogardus C (1986) Determinants of 24- hour energy expenditure in man. Methods and results using a respiratory chamber. pp. 1562-1578.
Rawson ES, Nolan A, Silver K, Shuldiner AR & Poehlman ET (2002) No effect of the Trp64Arg beta(3)-adrenoceptor gene variant on weight loss, body composition, or energy expenditure in obese, caucasian postmenopausal women. Metabolism 51, 801- 805.
Reilly JJ, Armstrong J, Dorosty AR, Emmett PM, Ness A, Rogers I, Steer C & Sherriff A (2005) Early life risk factors for obesity in childhood: cohort study. BMJ 330, 1357.
Reilly JJ, Corbett J, Given L, Gray L, Leyland A, MacGregor A, Marryat L, Miller M & Reid S (2009) The Scottish Health Survey 2008. Edinburgh: The Scottish Government.
Reimer RA, Debert CT, House JL & Poulin MJ (2005) Dietary and metabolic differences in pre- versus postmenopausal women taking or not taking hormone replacement therapy. Physiol Behav 84, 303-312.
213
Renehan AG, Tyson M, Egger M, Heller RF & Zwahlen M (2008) Body-mass index and incidence of cancer: a systematic review and meta-analysis of prospective observational studies. Lancet 371, 569-578.
Rennie KL, Coward A & Jebb SA (2007) Estimating under-reporting of energy intake in dietary surveys using an individualised method. Br J Nutr 97, 1169-1176.
Rennie KL, Jebb SA, Wright A & Coward WA (2005) Secular trends in under-reporting in young people. Br J Nutr 93, 241-247.
Rennie KL, Wells JC, McCaffrey TA & Livingstone MB (2006) The effect of physical activity on body fatness in children and adolescents. Proc Nutr Soc 65, 393-402.
Riddoch CJ, Mattocks C, Deere K, Saunders J, Kirkby J, Tilling K, Leary SD, Blair SN & Ness AR (2007) Objective measurement of levels and patterns of physical activity. Arch Dis Child 92, 963-969.
Rissanen J, Kuopusjarvi J, Pihlajamaki J, Sipilainen R, Heikkinen S, Vanhala M, Kekalainen P, Kuusisto J & Laakso M (1997) The Trp64Arg polymorphism of the beta 3-Adrenergic receptor gene. Lack of association with NIDDM and features of insulin resistance syndrome. Diabetes Care 20, 1319-1323.
Roberts SB & Dallal GE (1998) Effects of age on energy balance. Am J Clin Nutr 68, 975S-979S.
Roberts SB, Dietz W, Sharp T, Dallal GE & Hill JO (1995a) Multiple laboratory comparison of the doubly labeled water technique. Obes Res 3 Suppl 1, 3-13.
Roberts SB, Fuss P, Heyman MB & Young VR (1995b) Influence of age on energy requirements. Am J Clin Nutr 62, 1053S-1058S.
Roberts SB, Nicholson M, Staten M, Dallal GE, Sawaya AL, Heyman MB, Fuss P & Greenberg AS (1997) Relationship between circulating leptin and energy expenditure in adult men and women aged 18 years to 81 years. Obesity Res 5, 459-463.
Roberts SB & Rosenberg I (2006) Nutrition and Aging: Changes in the Regulation of Energy Metabolism With Aging. Physiol Rev 86, 651-667.
Roberts SB, Savage J, Coward WA, Chew B & Lucas A (1988) Energy expenditure and intake in infants born to lean and overweight mothers. N Engl J Med 318, 461-466.
Robinson TN, Killen JD, Kraemer HC, et al. (2003) Dance and reducing television viewing to prevent weight gain in African-American girls: the Stanford GEMS pilot study. Ethn Dis 13, S65-S77.
Roemmich JN, Clark PA, Mai V, Berr SS, Weltman A, Veldhuis JD & Rogol AD (1998) Alterations in growth and body composition during puberty: III. Influence of maturation, gender, body composition, fat distribution, aerobic fitness, and energy expenditure on nocturnal growth hormone release. J Clin Endocrinol Metab 83, 1440-1447.
214
Roemmich JN, Clark PA, Walter K, Patrie J, Weltman A & Rogol AD (2000) Pubertal alterations in growth and body composition. V. Energy expenditure, adiposity, and fat distribution. Am J Physiol Endocrinol Metab 279, E1426-E1436.
Rosenbaum M, Goldsmith R, Bloomfield D, Magnano A, Weimer L, Heymsfield S, Gallagher D, Mayer L, Murphy E & Leibel RL (2005) Low-dose leptin reverses skeletal muscle, autonomic, and neuroendocrine adaptations to maintenance of reduced weight. J Clin Invest 115, 3579-3586.
Rosenbaum M, Hirsch J, Murphy E & Leibel RL (2000) Effects of changes in body weight on carbohydrate metabolism, catecholamine excretion, and thyroid function. Am J Clin Nutr 71, 1421-1432.
Rosenbaum M, Murphy EM, Heymsfield SB, Matthews DE & Leibel RL (2002) Low Dose Leptin Administration Reverses Effects of Sustained Weight-Reduction on Energy Expenditure and Circulating Concentrations of Thyroid Hormones. J Clin Endocrinol Metab 87, 2391.
Rothenberg EM, Bosaeus IG, Westerterp KR & Steen BC (2000) Resting energy expenditure, activity energy expenditure and total energy expenditure at age 91-96 years. Br J Nutr 84, 319-324.
Rosenbaum M, Vandenborne K, Goldsmith R, et al. (2003) Effects of experimental weight perturbation on skeletal muscle work efficiency in human subjects. Am J Physiol Regul Integr Comp Physiol 285, R183-R192.
Royal College of Paediatrics and Child Health (2011) UK-WHO Growth Charts: early years. Available at: http://www.rcpch.ac.uk/growthcharts
Royal College of Physicians (2004) Storing up problems: The medical case for a slimmer nation. London: RCP.
Rush EC, Plank LD & Coward WA (1999) Energy expenditure of young Polynesian and European women in New Zealand and relations to body composition. Am J Clin Nutr 69, 43-48.
Rush EC, Plank LD, Davies PS, Watson P & Wall CR (2003) Body composition and physical activity in New Zealand Maori, Pacific and European children aged 5-14 years. Br J Nutr 90, 1133-1139.
Ruston D, Hoare J, Henderson L, Gregory J, Bates CJ, Prentice A, Birch M, Swan G & Farron M (2004) The National Diet and Nutrition Survey: adults aged 19 to 64 years. Volume 4: Nutritional Status (anthropometry and blood analytes), blood pressure and physical activity. London: TSO.
Rutanen J, Pihlajamaki J, Karhapaa P et al (2004) The Val103Ile polymorphism of melanocortin-4 receptor regulates energy expenditure and weight gain. Obes Res 12(7), 1060-1066.
215
Sahota P, Rudolf MC, Dixey R, Hill AJ, Barth JH & Cade J (2001) Randomised controlled trial of primary school based intervention to reduce risk factors for obesity. BMJ 323, 1029-1032.
Salazar G, Vio F, Aguirre E & Coward WA (2000) Energy requirements in Chilean infants. Arch Dis Child Fetal Neonatal Ed 83, F123.
Salbe AD, Nicolson M & Ravussin E (1997) Total energy expenditure and the level of physical activity correlate with plasma leptin concentrations in five-year-old children. J Clin Invest 99, 592-595.
Sallis JF, McKenzie TL, Conway TL, Elder JP, Prochaska JJ, Brown M, Zive MM, Marshall SJ & Alcaraz JE (2003) Environmental interventions for eating and physical activity: a randomized controlled trial in middle schools. Am J Prev Med 24, 209-217.
Saris WHM, Emons HJG, Groenenboom DC & Westerterp KR (1989) Discrepancy between FAO/WHO energy requirements and actual energy expenditure in healthy 7-11 year old children. Lueven, Belgium: 14th International Seminar on Pediatric Work Physiology.
Satoh Y, Shimizu T, Lee T, Nishizawa K, Iijima M & Yamashiro Y (2003) Resting energy expenditure and plasma leptin levels in adolescent girls with anorexia nervosa. Int J Eat Disord 34, 156-161.
Schmitz KH, Jacobs DR, Jr., Leon AS, Schreiner PJ & Sternfeld B (2000) Physical activity and body weight: associations over ten years in the CARDIA study. Coronary Artery Risk Development in Young Adults. Int J Obes Relat Metab Disord 24, 1475-1487.
Schoeller DA (1998) Balancing energy expenditure and body weight. Am J Clin Nutr 68, 956S-961S.
Schoeller DA & Hnilicka JM (1996) Reliability of the doubly labeled water method for the measurement of total daily energy expenditure in free-living subjects. J Nutr 126, 348S-354S.
Schoeller DA, Leitch CA & Brown C (1986) Doubly labeled water method: in vivo oxygen and hydrogen isotope fractionation. Am J Physiol Regul Integr Comp Physiol 251, R1137-R1143.
Schofield WN, Schofield C & James WPT (1985) Basal metabolic rate - review and prediction, together with an annotated bibliography of source material. Hum Nutr Clin Nutr 39C Suppl 1, 1-96.
Schulz LO & Schoeller DA (1994) A compilation of total daily energy expenditures and body weights in healthy adults. Am J Clin Nutr 60, 676-681.
Schutz Y (2000) Role of substrate utilization and thermogenesis on body-weight control with particular reference to alcohol. Proceedings of the Nutrition Society 59, 511-517.
216
Schutz Y, Bessard T & Jequier E (1984) Diet-induced thermogenesis measured over a whole day in obese and nonobese women. Am J Clin Nutr 40, 542-552.
Schutz Y & Garrow JS (2000) Energy and substrate balance and weight regulation. In Human nutrition and dietetics, pp. 137-148 [JS Garrow, WPT James, and A Ralph, editors]. Edinburgh: Churchill Livingstone.
Scientific Advisory Committee on Nutrition (2002) A framework for the Evaluation of Evidence no. http://www.sacn.gov.uk/pdfs/sacn_02_02a.pdf.
Scientific Advisory Committee on Nutrition (2008) The Nutritional Wellbeing of the British Population. London: TSO.
Scientific Advisory Committee on Nutrition & The Royal College of Paediatrics and Child Health (2007) Application of WHO Growth Standards in the UK. Report prepared by the Joint SACN/RCPCH Expert Group on Growth Standards. London: TSO.
Scuteri A, Sanna S, Chen WM, et al. (2007) Genome-wide association scan shows genetic variants in the FTO gene are associated with obesity-related traits. PLoS Genet 3, e115.
Segal KR, Gutin B, Albu J & Pi-Sunyer FX (1987) Thermic effects of food and exercise in lean and obese men of similar lean body mass. Am J Physiol 252, E110-E117.
Sesso HD (2007) Invited Commentary: A Challenge for Physical Activity Epidemiology. Am J Epidemiol 165, 1351-1353.
Sharp TA, Bell ML, Grunwald GK, Schmitz KH, Sidney S, Lewis CE, Tolan K & Hill JO (2002) Differences in Resting Metabolic Rate between White and African-American Young Adults. Obesity Res 10, 726-732.
Shaw K, Gennat H, O’Rourke P & Del Mar C (2006) Exercise for overweight or obesity. Cochrane Database Syst Rev, CD003817.
Sherwood NE, Jeffery RW, French SA, Hannan PJ & Murray DM (2000) Predictors of weight gain in the Pound of Prevention study. Int J Obes Relat Metab Disord 24, 395-403.
Sherwood NE, Morton N, Jeffery RW, French SA, Neumark-Sztainer D & Falkner NH (1998) Consumer preferences in format and type of community-based weight control programs. Am J Health Promot 13, 12-18.
Shiwaku K, Nogi A, Anuurad E, Kitajima K, Enkhmaa B, Shimono K & Yamane Y (2003) Difficulty in losing weight by behavioral intervention for women with Trp64Arg polymorphism of the beta3-adrenergic receptor gene. Int J Obes Relat Metab Disord 27, 1028-1036.
Siega-Riz AM, Viswanathan M, Moos MK, Deierlein A, Mumford S, Knaack J, Thieda P, Lux LJ & Lohr KN (2009) A systematic review of outcomes of maternal weight gain according to the Institute of Medicine recommendations: birthweight, fetal growth, and postpartum weight retention. Am J Obstet Gynecol 201, 339-343.
217
Siler SQ, Neese RA & Hellerstein MK (1999) De novo lipogenesis, lipid kinetics, and whole-body lipid balances in humans after acute alcohol consumption. Am J Clin Nutr 70, 928-936.
Simkin-Silverman LR, Wing RR, Boraz MA & Kuller LH (2003) Lifestyle intervention can prevent weight gain during menopause: results from a 5-year randomized clinical trial. Ann Behav Med 26, 212-220.
Simon C, Schweitzer B, Oujaa M, Wagner A, Arveiler D, Triby E, Copin N, Blanc S & Platat C (2008) Successful overweight prevention in adolescents by increasing physical activity: a 4-year randomized controlled intervention. Int J Obes (Lond) 32, 1489-1498.
Simonetti D’Arca A, Tarsitani G, Cairella M, Siani V, De Filippis S, Mancinelli S, Marazzi MC & Palombi L (1986) Prevention of obesity in elementary and nursery school children. Public Health 100, 166-173.
Sipilainen R, Uusitupa M, Heikkinen S, Rissanen A, Laakso M (1997) Variants in the human intestinal fatty acid binding protein 2 gene in obese subjects. J Clin Endocrinol Metab;82(8), 2629-2632.
Sjoberg A, Slinde F, Arvidsson D, Ellegard L, Gramatkovski E, Hallberg L & Hulthen L (2003) Energy intake in Swedish adolescents: validation of diet history with doubly labelled water. Eur J Clin Nutr 57, 1643-1652.
Snitker S, Tataranni PA & Ravussin E (2001) Spontaneous physical activity in a respiratory chamber is correlated to habitual physical activity. Int J Obes Relat Metab Disord 25, 1481-1486.
Soares MJ, Piers LS, O’Dea K & Shetty PS (1998) No evidence for an ethnic influence on basal metabolism: an examination of data from India and Australia. Br J Nutr 79, 333-341.
Solomon SJ, Kurzer MS & Calloway DH (1982) Menstrual cycle and basal metabolic rate in women. Am J Clin Nutr 36, 611-616.
Spadano JL, Bandini LG, Must A, Dallal GE & Dietz WH (2005) Longitudinal changes in energy expenditure in girls from late childhood through midadolescence. Am J Clin Nutr 81, 1102-1109.
Sparti A, DeLany JP, de la Bretonne JA, Sander GE & Bray GA (1997) Relationship between resting metabolic rate and the composition of the fat-free mass. Metabolism 46, 1225-1230.
Speakman JR & Selman C (2003) Physical activity and resting metabolic rate. Proc Nutr Soc 62, 621-634.
Statistics for Wales/Ystadegau ar gyfer Cymru (2010) Welsh Health Survey 2009: Welsh Assembly Government.
Stefan N, Vozarova B, Del Parigi A, Ossowski V, Thompson DB, Hanson RL, Ravussin E & Tataranni PA (2002) The Gln223Arg polymorphism of the leptin receptor in Pima Indians:
218
influence on energy expenditure, physical activity and lipid metabolism. Int J Obes Relat Metab Disord 26, 1629-1632.
Stephen AM, Teucher B, Bates BJ, Bluck LJ, Mander AP (2007) UK National Diet and Nutrition Survey: a study to decide the dietary assessment method for the new rolling programme (unpublished).
Stolley MR & Fitzgibbon ML (1997) Effects of an obesity prevention program on the eating behavior of African American mothers and daughters. Health Educ Behav 24, 152-164.
Stothard KJ, Tennant PW, Bell R & Rankin J (2009) Maternal overweight and obesity and the risk of congenital anomalies: a systematic review and meta-analysis. JAMA 301, 636-650.
Stratz CH (1904) Der Korper des Kindes: Fur Eltern, Arzta und Kunstler. Ferdinland Enke. Stuttgart: Germany.
Stubbs RJ, Harbron CG, Murgatroyd PR & Prentice AM (1995b) Covert manipulation of dietary fat and energy density: effect on substrate flux and food intake in men eating ad libitum. Am J Clin Nutr 62, 316-329.
Stubbs RJ, Hughes DA, Johnstone AM, Horgan GW, King N & Blundell JE (2004) A decrease in physical activity affects appetite, energy, and nutrient balance in lean men feeding ad libitum. Am J Clin Nutr 79, 62-69.
Stubbs RJ, Ritz P, Coward WA & Prentice AM (1995a) Covert manipulation of the ratio of dietary fat to carbohydrate and energy density: effect on food intake and energy balance in free-living men eating ad libitum. Am J Clin Nutr 62, 330-337.
Styne DM (2005) Obesity in childhood: what’s activity got to do with it? Am J Clin Nutr 81, 337-338.
Subar AF, Kipnis V, Troiano RP, et al. (2003) Using intake biomarkers to evaluate the extent of dietary misreporting in a large sample of adults: the OPEN study. Am J Epidemiol 158, 1-13.
Sun M, Gower BA, Bartolucci AA, Hunter GR, Figueroa-Colon R & Goran MI (2001) A longitudinal study of resting energy expenditure relative to body composition during puberty in African American and white children. Am J Clin Nutr 73, 308-315.
Sun M, Gower BA, Nagy TR, Bartolucci AA & Goran MI (1999) Do hormonal indices of maturation explain energy expenditure differences in African American and Caucasian prepubertal children? Int J Obes Relat Metab Disord 23, 1320-1326.
Sun M, Gower BA, Nagy TR, Trowbridge CA, Dezenberg C & Goran MI (1998) Total, resting, and activity-related energy expenditures are similar in Caucasian and African- American children. Am J Physiol 274, E232-E237.
219
Svendsen OL, Hassager C & Christiansen C (1993) Impact of regional and total body composition and hormones on resting energy expenditure in overweight postmenopausal women. Metabolism 42, 1588-1591.
Swinamer DL, Phang PT, Jones RL, Grace M & King EG (1988) Effect of routine administration of analgesia on energy expenditure in critically ill patients. Chest 93, 4-10.
Tammelin T, Laitinen J & Nayha S (2004) Change in the level of physical activity from adolescence into adulthood and obesity at the age of 31 years. Int J Obes Relat Metab Disord 28, 775-782.
Tanner JM (1990) Foetus into man: physical growth from conception to maturity, Second ed.: Castlemead Publications, Ware.
Tataranni PA, Baier L, Jenkinson C, Harper I, Del Parigi A, Bogardus C (2001) A Ser311Cys mutation in the human dopamine receptor D2 gene is associated with reduced energy expenditure. Diabetes 50(4), 901-904.
Tataranni PA, Larson DE, Snitker S, Young JB, Flatt JP & Ravussin E (1996) Effects of glucocorticoids on energy metabolism and food intake in humans. Am J Physiol Endocrinol Metab 271, E317-E325.
Tchernof A, Starling RD, Walston JD, Shuldiner AR, Dvorak RV, Silver K, Matthews DE & Poehlman ET (1999) Obesity-related phenotypes and the beta3-adrenoceptor gene variant in postmenopausal women. Diabetes 48, 1425-1428.
Tershakovec AM, Kuppler KM, Zemel B, Stallings VA. Age, sex, ethnicity, body composition, and resting energy expenditure of obese African American and white children and adolescents. Am J Clin Nutr 2002;75(5):867-871.
Timpson NJ, Emmett PM, Frayling TM, Rogers I, Hattersley AT, McCarthy MI & Davey Smith G (2008) The fat mass-and obesity-associated locus and dietary intake in children. Am J Clin Nutr 88, 971-978.
Tooze JA, Schoeller DA, Subar AF, Kipnis V, Schatzkin A & Troiano RP (2007) Total daily energy expenditure among middle-aged men and women: the OPEN Study. Am J Clin Nutr 86, 382-387.
Torun B (2005) Energy requirements of children and adolescents. Public Health Nutr 8, 968-993.
Toth MJ, Gottlieb SS, Fisher ML, Ryan AS, Nicklas BJ & Poehlman ET (1997) Plasma leptin concentrations and energy expenditure in heart failure patients. Metabolism 46, 450-453.
Toubro S, Sorensen TI, Ronn B, Christensen NJ & Astrup A (1996) Twenty-four-hour energy expenditure: the role of body composition, thyroid status, sympathetic activity, and family membership. J Clin Endocrinol Metab 81, 2670-2674.
Trabulsi J, Troiano RP, Subar AF, Sharbaugh C, Kipnis V, Schatzkin A & Schoeller DA (2003) Precision of the doubly labeled water method in a large-scale application: evaluation
220
of a streamlined-dosing protocol in the Observing Protein and Energy Nutrition (OPEN) study. Eur J Clin Nutr 57, 1370-1377.
Trayers T, Cooper AR, Riddoch CJ, Ness AR, Fox KR, Deem R & Lawlor DA (2006) Do children from an inner city British school meet the recommended levels of physical activity? Results from a cross sectional survey using objective measurements of physical activity. Arch Dis Child 91, 175-176.
Trayhurn P (2001) Biology of leptin--its implications and consequences for the treatment of obesity. Int J Obes Relat Metab Disord 25 Suppl 1, S26-S28.
Tremblay A, Simoneau JA & Bouchard C (1994) Impact of exercise intensity on body fatness and skeletal muscle metabolism. Metabolism 43, 814-818.
Treuth MS, Butte NF, Puyau M & Adolph A (2000b) Relations of parental obesity status to physical fitness of prepubertal girls. Pediatrics 106.
Treuth MS, Butte NF & Sorkin JD (2003) Predictors of body fat gain in nonobese girls with a familial predisposition to obesity. Am J Clin Nutr 78, 1212-1218.
Treuth MS, Butte NF & Wong WW (2000a) Effects of familial predisposition to obesity on energy expenditure in multiethnic prepubertal girls. Am J Clin Nutr 71, 893-900.
Treuth MS, Figueroa-Colon R, Hunter GR, Weinsier RL, Butte NF & Goran MI (1998) Energy expenditure and physical fitness in overweight vs non-overweight prepubertal girls. Int J Obes Relat Metab Disord 22, 440-447.
Treuth MS, Butte NF & Sorkin JD (2003) Predictors of body fat gain in nonobese girls with a familial predisposition to obesity. Am J Clin Nutr 78, 1212-1218.
Trowbridge CA, Gower BA, Nagy TR, Hunter GR, Treuth MS & Goran MI (1997) Maximal aerobic capacity in African-American and Caucasian prepubertal children. Am J Physiol 273, E809-E814.
Twisk JW, Kemper HC & van Mechelen W (2000) Tracking of activity and fitness and the relationship with cardiovascular disease risk factors. Med Sci Sports Exerc 32, 1455-1461.
Ukkola O, Tremblay A, Sun G, Chagnon YC & Bouchard C (2001) Genetic variation at the uncoupling protein 1, 2 and 3 loci and the response to long-term overfeeding. Eur J Clin Nutr 55, 1008-1015.
Valencia ME, McNeill G, Brockway JM & Smith JS (1992) The effect of environmental temperature and humidity on 24 h energy expenditure in men. Br J Nutr 68, 319-327.
Valencia ME, McNeill G, Haggarty P, Moya SY, Pinelli A, Quihui L & Davalos R (1995) Energetic consequences of mild Giardia intestinalis infestation in Mexican children. Am J Clin Nutr 61, 860-865.
Valve R, Heikkinen S, Rissanen A, Laakso M & Uusitupa M (1998) Synergistic effect of polymorphisms in uncoupling protein 1 and beta3-adrenergic receptor genes on basal metabolic rate in obese Finns. Diabetologia 41, 357-361.
221
van Marken Lichtenbelt WD, Schrauwen P, van de Kerckhove S & Westerterp-Plantenga MS (2002) Individual variation in body temperature and energy expenditure in response to mild cold. Am J Physiol Endocrinol Metab 282, E1077-E1083.
van Marken Lichtenbelt W, Westerterp-Plantenga MS & van Hoydonck P (2001) Individual variation in the relation between body temperature and energy expenditure in response to elevated ambient temperature. Physiol Behav 73, 235-242.
van Sluijs EM, Skidmore PM, Mwanza K, et al. (2008) Physical activity and dietary behaviour in a population-based sample of British 10-year old children: the SPEEDY study (Sport, Physical activity and Eating behaviour: Environmental Determinants in Young people). BMC Public Health 8, 388.
Van Wymelbeke V, Brondel L, Marcel Brun J & Rigaud D (2004) Factors associated with the increase in resting energy expenditure during refeeding in malnourished anorexia nervosa patients. Am J Clin Nutr 80, 1469-1477.
Vasquez F, Salazar G, Andrade M, Vasquez L & Diaz E (2006) Energy balance and physical activity in obese children attending day-care centres. Eur J Clin Nutr 60, 1115-1121.
Vaughan L, Zurlo F & Ravussin E (1991) Aging and energy expenditure. Am J Clin Nutr 53, 821-825.
Vaz M, Karaolis N, Draper A & Shetty P (2005) A compilation of energy costs of physical activities. Public Health Nutr 8, 1153-1183.
Verstraete SJ, Cardon GM, de Clercq DL & De Bourdeaudhuij IM (2007) A comprehensive physical activity promotion programme at elementary school: the effects on physical activity, physical fitness and psychosocial correlates of physical activity. Public Health Nutr 10, 477-484.
Viner RM & Cole TJ (2005) Television viewing in early childhood predicts adult body mass index. J Pediatr 147, 429-435.
Visser M, Deurenberg P, Van Staveren WA & Hautvast JG (1995) Resting metabolic rate and diet-induced thermogenesis in young and elderly subjects: relationship with body composition, fat distribution, and physical activity level. Am J Clin Nutr 61, 772-778.
Wagner A, Simon C, Ducimetiere P, et al. (2001) Leisure-time physical activity and regular walking or cycling to work are associated with adiposity and 5 y weight gain in middle-aged men: the PRIME Study. Int J Obes Relat Metab Disord 25, 940-948.
Walder K, Norman RA, Hanson RL, et al. (1998) Association between uncoupling protein polymorphisms (UCP2-UCP3) and energy metabolism/obesity in Pima indians. Hum Mol Genet 7, 1431-1435.
Walker A, Maher J, Coulthard M, Goddard E & Thomas M (2001) Living in Britain: Results from the 2000/01 General Household Survey. London: The Stationery Office.
222
Wallace JD, Abbott-Johnson WJ, Crawford DHG, Barnard R, Potter JM & Cuneo RC (2002) GH Treatment in Adults with Chronic Liver Disease: A Randomized, Double- Blind, Placebo-Controlled, Cross-Over Study. J Clin Endocrinol Metab 87, 2751-2759.
Waller K, Kaprio J, Kujala UM (2008) Associations between long-term physical activity, waist circumference and weight gain: a 30-year longitudinal twin study. Int J Obes (Lond) 32(2), 353-361.
Wang Z, Heshka S, Gallagher D, Boozer CN, Kotler DP & Heymsfield SB (2000) Resting energy expenditure-fat-free mass relationship: new insights provided by body composition modeling. Am J Physiol Endocrinol Metab 279, E539-E545.
Wardle J & Boniface D (2008) Changes in the distributions of body mass index and waist circumference in English adults, 1993//1994 to 2002//2003. Int J Obes 32, 527-532.
Wardle J, Brodersen NH & Boniface D (2007) School-based physical activity and changes in adiposity. Int J Obes.
Wardle J, Carnell S, Haworth CM, Farooqi IS, O’Rahilly S & Plomin R (2008) Obesity- associated genetic variation in FTO is associated with diminished satiety. J Clin Endocrinol Metab 93, 3640-3643.
Wardle J, Llewellyn C, Sanderson S & Plomin R (2009) The FTO gene and measured food intake in children. Int J Obes 33, 42-45.
Wareham NJ, Jakes RW, Rennie KL, Mitchell J, Hennings S & Day NE (2002) Validity and repeatability of the EPIC-Norfolk Physical Activity Questionnaire. Int J Epidemiol 31, 168-174.
Wareham NJ & Rennie KL (1998) The assessment of physical activity in individuals and populations: why try to be more precise about how physical activity is assessed? Int J Obes Relat Metab Disord 22 Suppl 2, S30-S38.
Wareham NJ, van Sluijs EM & Ekelund U (2005) Physical activity and obesity prevention: a review of the current evidence. Proc Nutr Soc 64, 229-247.
Warren JM, Henry CJ, Lightowler HJ, Bradshaw SM & Perwaiz S (2003) Evaluation of a pilot school programme aimed at the prevention of obesity in children. Health Promot Int 18, 287-296.
Warwick PM (2006) Factorial estimation of daily energy expenditure using a simplified method was improved by adjustment for excess post-exercise oxygen consumption and thermic effect of food. Eur J Clin Nutr 60, 1337-1340.
Warwick PM & Busby R (1990) Influence of mild cold on 24 h energy expenditure in ‘normally’ clothed adults. Br J Nutr 63, 481-488.
Wauters M, Considine RV, Chagnon M, Mertens I, Rankinen T, Bouchard C & Van Gaal LF (2002) Leptin levels, leptin receptor gene polymorphisms, and energy metabolism in women. Obes Res 10, 394-400.
223
Webb P (1986) 24-hour energy expenditure and the menstrual cycle. Am J Clin Nutr 44, 614-619.
Weijs PJ (2008) Validity of predictive equations for resting energy expenditure in US and Dutch overweight and obese class I and II adults aged 18-65 y. Am J Clin Nutr 88, 959-970.
Weinsier RL, Hunter GR, Desmond RA, Byrne NM, Zuckerman PA & Darnell BE (2002) Free-living activity energy expenditure in women successful and unsuccessful at maintaining a normal body weight. Am J Clin Nutr 75, 499-504.
Weinsier RL, Hunter GR, Zuckerman PA, Redden DT, Darnell BE, Larson DE, Newcomer BR & Goran MI (2000) Energy expenditure and free-living physical activity in black and white women: comparison before and after weight loss. Am J Clin Nutr 71, 1138-1146.
Weinsier RL, Schutz Y & Bracco D (1992) Reexamination of the relationship of resting metabolic rate to fat-free mass and to the metabolically active components of fat-free mass in humans. Am J Clin Nutr 55, 790-794.
Welk GJ, Corbin CB & Dale D (2000) Measurement issues in the assessment of physical activity in children. Res Q Exerc Sport 71, S59-S73.
Welle S, Jozefowicz R & Statt M (1990) Failure of dehydroepiandrosterone to influence energy and protein metabolism in humans. J Clin Endocrinol Metab 71, 1259-1264.
Wells JC, Chomtho S, Fewtrell MS (2007b) Programming of body composition by early growth and nutrition. Proc Nutr Soc 66(3), 423-434.
Wells JC, Coward WA, Cole TJ, Davies PS (2002) The contribution of fat and fat-free tissue to body mass index in contemporary children and the reference child. Int J Obes Relat Metab Disord 26(10), 1323-1328.
Wells JC & Ritz P (2001) Physical activity at 9-12 months and fatness at 2 years of age. Am J Hum Biol 13, 384-389.
Wells JC, Treleaven P & Cole TJ (2007a) BMI compared with 3-dimensional body shape: the UK National Sizing Survey. Am J Clin Nutr 85, 419-425.
Wenche DB, Holmen J, Kruger O & Midthjell K (2004) Leisure time physical activity and change in body mass index: an 11-year follow-up study of 9357 normal weight health women 20-49 years old. J Womens Health (Larchmt ) 13, 55-62.
Westerterp KR & Goris AHC (2002) Validity of the assessment of dietary intake: problems of mis-reporting. Current Opinion in Clinical Nutrition Metabolic Care. 5, 489-493.
Westerterp KR (1998) Alterations in energy balance with exercise. Am J Clin Nutr 68, 970S-974S.
Westerterp KR, Meijer GA, Janssen EM, Saris WH & Ten Hoor F (1992) Long-term effect of physical activity on energy balance and body composition. Br J Nutr 68, 21-30.
224
Westerterp KR & Plasqui G (2004) Physical activity and human energy expenditure. Curr Opin Clin Nutr Metab Care 7, 607-613.
Westerterp-Plantenga MS, van Marken Lichtenbelt WD, Strobbe H & Schrauwen P (2002) Energy metabolism in humans at a lowered ambient temperature. Eur J Clin Nutr 56, 288-296.
Weyer C, Snitker S, Bogardus C & Ravussin E (1999) Energy metabolism in African Americans: potential risk factors for obesity. Am J Clin Nutr 70, 13-20.
WHO Multicentre Growth Reference Study Group (2006) Complementary feeding in the WHO Multicentre Growth Reference Study. Acta Paediatrica Suppl 450, 27-37.
Wilks DC, Besson H, Lindroos AK & Ekelund U (2011) Objectively measured physical activity and obesity prevention in children, adolescents and adults: a systematic review of prospective studies. Obes Res 12, e119-e129.
Willett WC (2003) Invited Commentary: OPEN Questions. Am J Epidemiol 158, 22-24.
Wilsgaard T, Jacobsen BK & Arnesen E (2005) Determining Lifestyle Correlates of Body Mass Index using Multilevel Analyses: The Tromso Study, 1979-2001. Am J Epidemiol 162, 1179-1188.
Wilson MM & Morley JE (2003) Invited Review: Aging and energy balance. J Appl Physiol 95, 1728-1736.
Wing RR & Hill JO (2001) Successful weight loss maintenance. Annu Rev Nutr 21, 323-341.
Withers RT, Smith DA, Tucker RC, Brinkman M & Clark DG (1998) Energy metabolism in sedentary and active 49- to 70-yr-old women. J Appl Physiol 84, 1333-1340.
Wong WW (1994) Energy expenditure of female adolescents. J Am Coll Nutr 13, 332-337.
Wong WW, Butte NF, Ellis KJ, Hergenroeder AC, Hill RB, Stuff JE & Smith EO (1999) Pubertal African-American girls expend less energy at rest and during physical activity than Caucasian girls. J Clin Endocrinol Metab 84, 906-911.
World Health Organization (1985) Energy and protein requirements. Report of a joint FAO/WHO/UNU expert consultation no. 724. Geneva: WHO.
World Health Organization (1995) Maternal Anthropometry and Pregnancy Outcomes – A WHO Collaborative Study. Bulletin of the World HealthOrganization 73, S1-S69.
World Health Organization (1998) Obesity – Preventing and Managing the Global Epidemic no. 894. Geneva: World Health Organization.
World Health Organization (2006) WHO Child Growth Standards: Methods and development: Length/height-for-age, weight-for-age, weight-for-length, weight-for- height and body mass index-for-age. Geneva, Switzerland: World Health Organization.
Wouters-Adriaens MP & Westerterp KR (2008) Low resting energy expenditure in Asians can be attributed to body composition. Obesity (Silver Spring) 16, 2212-2216.
225
Wren RE, Blume H, Mazariegos M, Solomons N, Alvarez JO & Goran MI (1997) Body composition, resting metabolic rate, and energy requirements of short- and normal- stature, low-income Guatemalan children. Am J Clin Nutr 66, 406-412.
Wu X, Luke A, Cooper RS et al (2004) A genome scan among Nigerians linking resting energy expenditure to chromosome 16. Obes Res 12(4):577-581.
Yang X, Telama R, Leskinen E, Mansikkaniemi K, Viikari J & Raitakari OT (2007) Testing a model of physical activity and obesity tracking from youth to adulthood: the cardiovascular risk in young Finns study. Int J Obes (Lond) 31, 521-527.
Yang X, Telama R, Viikari J & Raitakari OT (2006) Risk of obesity in relation to physical activity tracking from youth to adulthood. Med Sci Sports Exerc 38, 919-925.
Yanovski JA, Diament AL, Sovik KN, Nguyen TT, Li H, Sebring NG & Warden CH (2000) Associations between uncoupling protein 2, body composition, and resting energy expenditure in lean and obese African American, white, and Asian children. Am J Clin Nutr 71, 1405-1420.
Yanovski SZ, Reynolds JC, Boyle AJ & Yanovski JA (1997) Resting metabolic rate in African-American and Caucasian girls. Obes Res 5, 321-325.
Yao M, McCrory MA, Ma G, Li Y, Dolnikowski GG, Roberts SB (2002) Energy requirements of urban Chinese adults with manual or sedentary occupations, determined using the doubly labelled water method. Eur J Clin Nutr 56, 575-584.
Yoshioka M, Doucet E, St Pierre S, Almeras N, Richard D, Labrie A, Despres JP, Bouchard C & Tremblay A (2001) Impact of high-intensity exercise on energy expenditure, lipid oxidation and body fatness. Int J Obes Relat Metab Disord 25, 332-339.
Zurlo F, Ferraro RT, Fontvielle AM, Rising R, Bogardus C & Ravussin E (1992) Spontaneous physical activity and obesity: cross-sectional and longitudinal studies in Pima Indians. Am J Physiol 263, E296-E300.
- Preface
- Table of contents
- 1 Summary
- Introduction
- Background
- Summary of approach used by SACN to determine revised energy reference values
- Energy reference values for infants, children and adolescents
- Energy reference values for adults
- Reference energy values during pregnancy and lactation
- Comparisons with reference values for energy in the COMA 1991 report
- 2 Introduction
- Background
- Energy available from food and drink
- Components of energy expenditure
- Factors affecting energy expenditure
- Energy balance and storage
- Obesity
- The influence of physical activity and diet on the regulation of body weight
- Definition of energy requirement
- Approaches used to estimate energy reference values
- 3 Estimated average requirements – approaches used and values derived
- Summary of approach used to determine EARs
- Energy reference values for infants, children and adolescents
- Energy reference values for adults
- Reference energy values during pregnancy and lactation
- Comparisons with reference values for energy in the 1991 COMA report
- 4 Conclusions and recommendations
- Recommendations
- Research Recommendations
- Acknowledgements
- Membership of Energy Requirements Working Group
- Membership of Scientific Advisory Committee on Nutrition
- Appendix 1 – SACN working procedures
- Appendix 2 – Energy yields from substrates
- Appendix 3 – Components of energy expenditure and factors affecting it
- Components of energy expenditure
- Factors affecting energy expenditure
- Genetics of obesity
- Appendix 4 – Basal metabolic rate prediction equations
- Appendix 5 – The physical activity level (PAL) and its use in the prediction of energy requirements
- Theoretical aspects of calculation of PAL and its factorial prediction
- Magnitude and variation in PAL within the general population
- PAL values in relation to health outcomes
- Utilising PAL values to determine reference energy intakes
- Appendix 6 – Doubly labelled water (DLW) method
- Methodology critique
- Summary
- Appendix 7 – Physical activity and energy balance
- Background
- Measuring physical activity
- Assessment of physical activity levels in the UK population
- Physical activity and body fatness
- Sedentary behaviour and weight gain
- Summary
- Appendix 8 – Characteristics of the data set of DLW measures of energy expenditure for adults: combined OPEN/Beltsville DLW data set.
- Appendix 9 – Energy requirements for disease
- Introduction
- Energy intake
- Energy expenditure
- Diseases in adults
- Diseases in children
- Summary
- Appendix 10 – Obesity prevalence in the UK
- Appendix 11 – A consideration of energy intake and physical activity in relation to weight gain
- Background
- Energy balance and energy flux
- Exercise and appetite control
- Summary
- Appendix 12 – Characteristics of the data set of doubly labelled water (DLW) measures of energy expenditure in children and adolescents
- Number of studies and individuals within the studies and general characteristics
- Adjustment of PAL to account for growth costs
- Appendix 13 – Comparisons with the existing DRVs for energy from the 1991 COMA report ‘Dietary Reference Values for Energy and Nutrients for the United Kingdom’.
- Overarching differences
- Glossary
- Glossary of statistical terms
- Abbreviations
- Reference List
henry 2005(1)(1).pdf
Basal metabolic rate studies in humans: measurement and development of new equations
CJK Henry* School of Biological and Molecular Sciences, Oxford Brookes University, Gipsy Lane Campus, Oxford OX3 0BP, UK
Abstract
Objective: To facilitate the Food and Agriculture Organization/World Health Organization/United Nations University Joint (FAO/WHO/UNU) Expert Consultation on Energy and Protein Requirements which met in Rome in 1981, Schofield et al. reviewed the literature and produced predictive equations for both sexes for the following ages: 0–3, 3–10, 10–18, 18–30, 30–60 and .60 years. These formed the basis for the equations used in 1985 FAO/WHO/UNU document, Energy and Protein Requirements.
While Schofield’s analysis has served a significant role in re-establishing the importance of using basal metabolic rate (BMR) to predict human energy requirements, recent workers have subsequently queried the universal validity and application of these equations. A survey of the most recent studies (1980–2000) in BMR suggests that in most cases the current FAO/WHO/UNU predictive equations overestimate BMR in many communities. The FAO/WHO/UNU equations to predict BMR were developed using a database that contained a disproportionate number – 3388 out of 7173 (47%) – of Italian subjects. The Schofield database contained relatively few subjects from the tropical region.
The objective here is to review the historical development in the measurement and application of BMR and to critically review the Schofield et al. BMR database presenting a series of new equations to predict BMR. Design: This division, while arbitrary, will enable readers who wish to omit the historical review of BMR to concentrate on the evolution of the new BMR equations. Setting: BMR data collected from published and measured values. Subjects: A series of new equations (Oxford equations) have been developed using a data set of 10 552 BMR values that (1) excluded all the Italian subjects and (2) included a much larger number (4018) of people from the tropics. Results: In general, the Oxford equations tend to produce lower BMR values than the current FAO/WHO/UNU equations in 18–30 and 30–60 year old males and in all females over 18 years of age. Conclusions: This is an opportune moment to re-examine the role and place of BMR measurements in estimating total energy requirements today. The Oxford equations’ future use and application will surely depend on their ability to predict more accurately the BMR in contemporary populations.
Keywords Universal validity
Basal metabolic rate Energy metabolism
Energy requirements Body mass index
Introduction
Since the last Food and Agriculture Organization/World
Health Organization/United Nations University
(FAO/WHO/UNU) Expert Committee on Energy and
Protein Requirements met in 1981, a considerable amount
of work has been reported on the use and validity of the
FAO/WHO/UNU1 equations to predict basal metabolic rate
(BMR). This paper is divided into two parts – one will
review the historical development in the measurement and
applicationofBMR; and the secondwill critically review the
Schofield BMR database and then present a series of
new equations (Oxford equations) to predict BMR.
This division, while arbitrary, will enable readers who
wish to omit the historical review of BMR to concentrate on
the evolution of the new BMR equations.
Work concerning energy metabolism may be traced
back to 1783 and the classical experiments of Lavoisier and
Laplace. The principles of calorimetry laid down by these
founding fathers over 200 years ago are still valid today.
The development and subsequent apparatus used to
measure respiratory exchange were based on the
principles of calorimetry. The term ‘basal’ was used to
distinguish between the energy expended while perform-
ing physical activity and being at rest. BMR represents the
integration of minimal activity of all the tissues in the body
q The Author 2005*Corresponding author: Email [email protected]
Public Health Nutrition: 8(7A), 1133–1152 DOI: 10.1079/PHN2005801
under steady state conditions. It is usually expressed as
heat production or oxygen consumption per unit body
size. A more succinct definition of BMR was presented by
Mitchell2 who said, ‘Basal metabolism of an animal is the
minimal rate of energy expenditure compatible with life’.
In order to begin our analysis, it is appropriate to briefly
review the historical developments in the study of BMR.
This approach will enable us to appreciate the primary
purpose of the early measures of BMR and how its
application has evolved with time.
BMR is the daily rate of energy metabolism an individual
needs to sustain in order to preserve the integrity of vital
functions. It must be measured under conditions, which,
as far as possible, avoid the influence of the external
environment, for example heat, or cold, physical move-
ment and the effects of food or drugs. If these conditions
are observed, the result of the measurement is considered
to represent the physiological and biochemical integrity of
the individual concerned. In normal free-living persons,
the amount of energy provided by their diet must cover the
demands of basal metabolism plus additional amounts
needed for the physical activity associated with essential
bodily needs, and also those imposed by the physical
work involved in occupation, in engaging in social
exchanges and in leisure activities.
Historical background: origins of the term ‘basal
metabolism’
Magnus-Levy coined the term Grundumsatz or ‘basal
metabolism’ in 1899. This term was of great value to the
early investigators, as it emphasised the need to conduct
the experiments under strictly standardised conditions.
These included the following: (1) absence of gross
muscular activity; (2) post-absorptive state; (3) minimal
emotional disturbance; (4) wakefulness; (5) normal
nutritive condition; (6) absence of disease or infection;
and (7) thermo-neutral environment. In practice, however,
it was impossible to impose all of the above conditions.
For example, many of the early studies in humans
reported by DuBois, Lusk and Rubner during the years
1900–1920 did not strictly meet the requirement of a
thermoneutral environment, leading to a slightly elevated
BMR. Moreover, many of the values reported by Aub and
DuBois3,4 were obtained in anxious, untrained subjects.
For this reason, the Aub–DuBois standards tended to be
higher than other BMR standards.
The term ‘basal metabolism’ is often misunderstood to
imply the lowest level of energy expenditure, which it
clearly is not. During sleep and in conditions of under-
nutrition, metabolism may be lower than that observed
under basal conditions. To avoid this confusion, Krogh5
coined the term ‘standard metabolism’. In order to secure
comparable results, the imposition of strict conditions for
the measurement of BMR is essential.
Conditions to be met while measuring BMR
The concept of basal metabolism arose from the need to
standardise measurements so that accurate comparisons
could be made between individuals. This is achieved by
measuring a minimum rate of heat production free of the
effects of any consumption of food and ‘extreme’ physical
environments6.
All BMR measurements must therefore meet the
following conditions:
1. The subject should be completely rested, both before
and during the measurements. They should be lying
down and fully awake.
2. The subjects should be fasted for at least 10–12 hours
before the measurements are taken.
3. The environment in which the measurements are taken
should be thermo-neutral (22–268C) so that there is no
thermoregulatory effect on heat production.
4. The subject should be free from emotional stress and
familiar with the apparatus used.
Ambient temperature during BMR measurements
Theambient temperature atwhich energyexpenditure is at a
minimumwas termed the ‘critical temperature’ by Rubner at
the turn of the 20th century. Themore commonly used term
was ‘zone of thermal neutrality’. This was defined as the
ambient temperature above or below which resting
metabolism of subjects begins to rise. The lowest ambient
temperature at which an organism can maintain ‘resting’ or
basal metabolic rate (without an increase in energy
expenditure) is called the lower critical temperature. Work
carried out on humans suggests the lower critical
temperature to be between 22 and 278C7,8. Numerous
publishedworksonBMRwereconductedat temperatures as
low as 9–158C9. Indeed many of the early studies paid little
attention to maintaining the subjects at thermoneutrality.
Clinical and physiological standards
During the early studies on BMR, there were two schools
of thought on how BMR values should be represented.
One group, called the ‘clinical standard’, assembled data
on first tests on supposedly ‘normal’ subjects. As is now
well known, first tests are usually higher in untrained
subjects. Therefore, these values and standards led to
values that were usually higher than those of the
‘physiological standards’. On the basis of extensive review
and observations of BMR values at that time, Roth and
Buckingham10 made the point succinctly as follows, ‘more
than one authority has stated on the basis of extensive
travel and observation, that as many as 70% of the basal
metabolism reports made today by the average operator
may not be worth the paper on which they are written’.
CJK Henry1134
Early use of BMR measurements
BMR measurement in humans attracted considerable
interest during the early part of the 20th century. They
were primarily used for the diagnosis of hypo- and
hyperthyroidism. BMR tests marked a new era in clinical
medicine. It is interesting to note that until the mid-1950s
the only reliable measure to diagnose thyroid dysfunction
was to determine a subject’s BMR. The study of basal
metabolism represents an example of the early collabor-
ation between physiologists and clinicians. BMR measure-
ments were routinely made during clinical examinations
and were believed to be instrumental in the diagnosis of
thyroid disease, diabetes and leukaemia11.
Use of BMR to predict energy requirements
While BMR measurements were used in clinical diagnosis
during the early part of the 20th century, the first
comprehensive study to use BMR as the basis to estimate
human energy requirements and hence food requirements
was described by Bedale12. She studied a group of 45 boys
and 55 girls aged 7–18 years. This was a significant
departure from previous work as the early work on BMR
was primarily intended to serve as metabolic reference
values in clinical nutrition, notably in the diagnosis of
hypo- and hyperthyroidism. It is, therefore, of interest to
note that the 1985 FAO/WHO/UNU1 approach to the
estimation of energy requirements is a refinement of the
method first described by Bedale12.
FAO studies on calorie requirements
The FAO nutrition studies No. 1513 published in 1957,
entitled Calorie Requirements, represented a landmark
both in approach and analysis. Many of the previous
reports on energy requirements proposed by Voit in 1890,
Atwater in 1895, Lusk in 1918 and the NRC (USA)
published in 1943 and revised in 195313, were based on
food intake. The FAO13 publication proposed for the first
time the use of energy expenditure to calculate energy
requirements.
Two simplified empirical equations to predict energy
requirements were presented:
Males: E ¼ 152W0.73 Females: E ¼ 123.47W0.75
This was further simplified to:
Males: E ¼ 815 þ 36.6W Females: E ¼ 580 þ 31.1W
where, E represents total energy requirements
(kcal day21) and W represents weight.
These simple linear equations to predict total energy
requirements bear close resemblance to the linear
equations used to predict BMR today. It is instructive to
record that the concept of using energy expenditure to
estimate energy requirements historically originated from
the 1957 report of the FAO13.
Current views on the use and application of BMR
The resurgence of interest in BMR can be directly ascribed to
a confluence of factors, notably two events. Firstly, a desire
to understand the biology and aetiology of obesity and,
secondly, the publication of the FAO/WHO/UNUdocument
Energy and Protein Requirements in 1985 that proposed, for
the first time, the use of energy expenditure (hence BMR)
rather than food intake to calculate energy requirements.
This new approach to estimate energy requirements
emphasised the need to estimate accurately BMR in
populations livingunder various climatic and environmental
conditions.Under-oroverestimationofBMRwould severely
affect the overall estimation of energy requirements.
If BMR measurements are to be used in estimating
energy requirements, it is important to have some details
on the apparatus used and the techniques adopted by
various investigators during the past 80–90 years. The
following section will, therefore, review the methods and
apparatus used to measure BMR between 1900 and 2000.
Description of methodology: development of
apparatus to measure BMR
With the growing importance of determining BMR in the
diagnosis and treatment of endocrine disorders (notably
thyroid disorders), the demand and use of calorimetry
rapidly expanded between 1910 and 1950. It was
customary to use indirect calorimetry to measure BMR.
The methods available to measure BMR may be divided
into two types: closed and open circuit methods. In the
closed circuit methods, the CO2 produced is absorbed
within the system. Oxygen is added to maintain the
volume of the gas constant. Benedict in 1918 initially
devised a method where the amount of CO2 absorbed by
soda-lime was carefully replaced by O2 which could be
measured. Later, Krogh5 and Roth14 developed an
instrument that measured O2 consumption from the
reduction in the volume of the gas by using a spirometer.
An interesting feature related to the pioneering studies
on BMR was that, until 1919, calorimetry was confined to
experimental laboratories under the control of highly
trained scientists and technicians. With the advent of the
Benedict-Roth spirometer (in the 1920s), which was a
simple, portable calorimeter, the use and application
spread widely. While this portable calorimeter was
encouraged by some15, others were more critical. Roth
and Buckingham10 commented ‘many of these technicians
educated overnight merely to man the machine, lacked
training and experience necessary to face the multiplicity
of problems to be otherwise encountered. . .Their work
Basal metabolic rate studies in humans 1135
and reports were generally unquestioned because there
were few capable of checking and passing judgement’.
While it is impossible to gauge the impact such practices
had on BMR measurements at that time, it nevertheless
reinforces the need to examine more carefully and
critically the methods used to estimate BMR during the
early days of BMR collection. It is important to recognise
that closed circuit was the more widely used method to
measure BMR in the early days of its study.
The most commonly used closed circuit apparatus to
measure BMR during the period 1910–1950 were the
following:
1. Krogh spirometer
2. Benedict spirometer (universal apparatus)
3. Benedict-Roth spirometer
4. Knipping apparatus
5. Fleish metabometer-metabograph
The major drawback of the closed circuit method was that
the absorption of CO2 did not allow the value of RQ to be
calculated. As a rule, a value of 0.82–0.85 was assumed,
leading to an error of up to ^6% since the food
undergoing oxidation had not been defined16.
Problems related to the use of closed circuit
apparatus
The accurate determination of BMR requires firstly, that the
subject be in the basal resting state (for either open or
closed circuit calorimetry) and secondly for the rate of O2 consumption to be measured accurately. In closed circuit
tests, the subject rebreathes from a spirometer that contains
a CO2 absorber. It also contains O2 at a partial pressure
much greater than the atmosphere. The period of
rebreathing is usually 5–10minutes as subjects become
uneasy if prolonged. The rate of O2 consumption in the
subject is calculated from the average rate of decrease in
volume from the spirometer. Numerous difficulties in
obtaining accurate values using the spirometer have been
reported. These include (1) completeness of CO2 absorbed,
(2) volumetric calibration, and (3) Kymographic accuracy.
Both Krogh5,17 and Benedict18 commented on the
importance of the lung volume remaining effectively
a constant at the beginning and end of the experimental
period. If not, the changes in spirometer volume will
represent not only the changes in O2 consumption but also
changes in lung volume.
In a series of studies comparing the closed to the open
circuit method to estimate BMR, Lewis et al.19–21 reported
that the closed circuit overestimated BMR in 12 out of 25
adults and showed no difference in BMR measurements in
the other subjects. Willard and Wolf22 reported another
source of error with closed circuit calorimetry. This
involved changes in chest volume during respiration.
Changes in the expiratory position of the chest occurring
during the experimentation had a marked effect on the
slope of the tracing. Thus, a change in chest volume by its
impact on the slope of the spirogram may lead to a falsely
high apparent metabolic rate. A further reason for the
elevated BMR values reported in the early literature was
due to the usual practice at that time to record values from
the first test. An additional source of error relates to the
sample collection of O2 for a very short period of time
(5–10minutes). If hyperventilation had occurred (which is
quite common in untrained subjects) an overestimation of
energy expenditure would occur16. Moreover, these early
studies did not maintain standard BMR conditions prior to
testing. It must be remembered that the early BMR studies
were used for clinical diagnostic purposes and not for any
other metabolic usage.
BMRmeasurements using open circuit calorimetry show
little difference due to type of equipment used23–26 using
a range of methods, notably Douglas Bag, Oxylog, HB
metabolator, ventilated hood, canopy and whole body
calorimeter, showed marginal difference in BMR between
methods used (see Table 1).
In contrast to the numerous studies comparing different
methods of measuring BMR using open circuit calorimetry,
few studies have compared closed circuit with open circuit
calorimetry. Such comparisons as there are suggest that
closed circuit calorimetry usually overestimates BMR (see
Table 2).
More recently, Clark and Hoffer27 who measured
BMR in 18–30 year old men either using a Deltatrac
(open circuit) or a 9 litre respirometer (similar to a
Benedict-Roth closed circuit apparatus), found that the
ventilated hood measurements produced a BMR of
6.87 ^ 0.619MJ/24 h (1643 ^ 148 kcal/24 h) compared
to 7.19 ^ 0.606MJ/24 h (1721 ^ 145 kcal/24 h) in the
Table 1 Measurement reliability and reproducibility using indirect calorimetry
Oxylog (kcal day21)
HB metabolator (kcal day21)
Ventilated hood (kcal day21)
Ventilated tent (kcal day21)
Whole-body calorimeter (kcal day21) Difference (%)
Power of t test
Protocol 1 (n 6) 1386.23 ^ 83.65 – – 1367.11 ^ 81.26 – þ1.6 ^ 2.5 0.07 Protocol 2 (n 6) – 1460.33 ^ 64.53 1515.30 ^ 93.21 – – 23.1 ^ 2.5 0.15 Protocol 3 (n 6) 1364.72 ^ 59.75 1379.10 ^ 62.14 – – – 20.9 ^ 1.9 0.05 Protocol 4 (n 6) – 1412.52 ^ 59.75 1410.13 ^ 74.09 – – þ0.4 ^ 2.5 0.03 Protocol 5 (n 10) – – 1321.70 ^ 45.41 – 1367.11 ^ 57.36 23.1 ^ 2.8 0.15
CJK Henry1136
closed circuit respirometer. These values indicate that
the closed circuit apparatus produced a BMR 5.6%
higher than the open circuit.
Summary
1. Closed circuit calorimetry was widely used in the
measurement of BMR during the first half of the 20th
century.
2. Closed circuit calorimetry tended to produce higher
BMR values.
3. The elevated values were ascribed to:
(a) Small leaks causing a larger error in the closed
circuit calorimetry28.
(b) The respiration of pure oxygen tended to elevate
BMR16.
(c) Changes in chest volume during respiration
tended to alter the slope of the spirogram reading,
leading to an apparently higher BMR value22.
Brief review of BMR standards and predictive
equations
This section briefly reviews the predictive equations for
BMR in man beginning with the ‘surface area law’ at the
turn of the 20th century to the more recent analysis by
Schofield et al.29
DuBois height–weight formula chart
While surface area may be calculated using various
anthropometric parameters, DuBois and DuBois30 pro-
duced an equation relating weight and height to surface
area as follows:
A ¼ W0:425 £H0:715 £ 71:84 where, A ¼ surface area in cm2; W ¼ weight in kilograms and H ¼ height in centimeters.
Later, Aub and DuBois3,4, applying the surface law
principle to man, published a table of BMRm22 per hour
from 14 to 80 years of age (Table 3). These formulae are
still widely used despite being based on a group with only
nine subjects and one cadaver!
Harris–Benedict standards
While the ‘surface law’ remained a dominant concept since
its introduction early in the 20th century, it was
nevertheless strongly challenged by Harris and Benedict11
who embarked on a detailed biometric analysis of BMR
which culminated in the publication of their monumental
work entitled A Biometric Study of Basal Metabolism in
Man. BMR measurements were made on 136 males and
103 females at the Carnegie Nutrition Laboratory in
Boston. Using rigorous statistical concepts, they devel-
oped the following equations to predict BMR:
males h ¼ 66:4730þ 13:7516W þ 5:0033S2 6:7750A
females h ¼ 665:0955þ 9:5634W þ 1:8496S2 4:6756A
where, h ¼ kcal day21; W ¼ weight in kilograms; S ¼ stature in centimeters; A ¼ age in years.
Harris and Benedict’s analysis marked a significant
departure from previous work. Firstly, it introduced for the
first time biometric principles in its analysis. Secondly,
they used subjects that were maintained under strict
experimental conditions prior to the measurements. Using
partial correlation coefficients, they also showed that both
stature and weight have an independent effect on BMR.
While these equations were useful and valuable aids to
predicting BMR, they were not above criticism. For
example, the constant in the equation showed a ten-fold
difference between males and females (66 versus 665).
Benedict himself later recognised and expressed concern
that the equations overestimated BMR, ‘particularly in
those young women’. Daly et al.31 confirmed that the
Harris–Benedict equations overestimated BMR by about
10–15%. Despite this, the simplicity of the Harris–
Benedict equation made it a popular equation in wide
use. Even today, many clinicians in North America use it
routinely32.
Table 2 Comparison between closed and open circuit BMR
Authors No. of subjects Difference
Krogh and Rasmussen (1922)
5 of 19 6–11% of open circuit values
Hunt (1926) 20 of 25 12% of open circuit values
Lewis et al. (1943) 13 of 25* 5% of open circuit values
Willard and Wolf (1951) 8 of 18* 10% of open circuit values
Fowler et al. (1957) 52* SD 7% of open circuit values
Harmin (1953) – SD 7% of open circuit values
Abbreviation: BMR – basal metabolic rate. Source: Consolazio et al.26
*Adults.
Table 3 DuBois normal standards for BMR (Cal m22 per hour)
Age (y) Males Females
14–15 46.0 43.0 16–17 43.0 40.0 18–19 41.0 38.0 20–29 39.5 37.0 30–39 39.5 36.5 40–49 38.5 36.0 50–59 37.5 35.0 60–69 36.5 34.0 70–79 35.5 33.0
Abbreviation: BMR – basal metabolic rate.
Basal metabolic rate studies in humans 1137
Boothby and Sandiford or ‘Mayo standards’
Scientists at the Mayo clinic commenced collecting BMR
data systematically in a variety of subjects from 1917. The
investigators used a combination of normal, free-living
subjects and ‘hospital normal’ subjects. While the subjects
admitted to hospital were not seriously ill, they never-
theless highlight the point that the subjects were not all
‘normal’ free-living subjects33,34. Boothby et al.15 made a
careful study of BMR in 639 males and 828 females.
Quenouille standards
Quenouille et al.’s35 analysis in 1951 was the first
comprehensive survey of all the available BMR studies
conducted and represented over 8600 subjects (4300 aged
between 17 and 39 years, 800 over 40 years and 3520 less
than 1 year of age). We need to pay tribute to these
investigators who were ‘pioneers’ in the systematic
collection of BMR and they statistically analysed the data
prior to the advent of computers. Quenouille et al.’s35
extensive review of the early literature on BMR has also
been a major source of valuable information for both the
Schofield and Oxford databases. For the first time,
Quenouille et al.35 also included BMR measurements
from people living in the tropics. Their analysis attempted
to examine the role of ethnicity and climate on BMR. This
made it the first large-scale study of the world literature on
BMR. While they considered temperature and humidity as
important factors in predicting BMR, sadly their equations
were not used extensively. Given below is an example of
their equation for men in Northern Europe.
M ¼ 2:975Hþ 8:90W þ 11:7Sþ 3:0h2 4:0tþ 293:8 where, M ¼ kcal day21; W ¼ weight in kilograms; H ¼ height in centimeters; S ¼ surface area from DuBois; t ¼ temperature and h ¼ humidity.
Schofield equations (FAO/WHO/UNU equations):
issues and analysis
Note here that the term FAO/WHO/UNU equations and
Schofield equations will be used interchangeably. To
facilitate the 1981 FAO/WHO/UNU expert consultation on
Energy and Protein Requirements, Durnin16 surveyed the
literature on BMR and assembled BMR values and
anthropometric data on 2238 subjects. Durnin16 presented
tables to predict BMR based on body weight, age and
gender. Subsequently, the FAO/WHO/UNU requested
Schofield et al.29 to extend this analysis and produce a
series of predictive equations. Schofield et al.29 reviewed
the literature and produced predictive equations for both
sexes for the following ages: 0–3, 3–10, 10–18, 18–30,
30–60 and .60 years. These formed the basis for the
equations used in the FAO/WHO/UNU document Energy
and Protein Requirements1. The Schofield database
comprised 114 published studies of BMR, totalling 7173
data points. Although their database comprised almost
11 000 BMR values (including group mean values), most of
the results were obtained from European and North
American subjects. An interesting feature that emerged
from their analysis was that the BMR of Asiatic Indians was
overestimated by 10–11% by their equations. This issue
was further highlighted by the FAO/WHO/UNU report. At
the time of their analysis, there was insufficient data to
ascertainwhether the effect noted in Indianswas unique or
whether it reflected a general pattern of metabolism in
tropical peoples. Indeed, the observation that BMRmay be
different in peoples living in the tropics was first reported
by de Almeida36. He showed that BMR in Brazilians was
approximately 24% lower than the Aub–DuBois standards.
Subsequently, Henry and Rees37 showed that the FAO/
Table 4 The percentages by which the FAO/WHO/UNU equations overestimate (þ) or underestimate (2) the actual BMR in different ethnic groups
Age group (y) Mean % No. of subjects
Males (all ethnicities) 3–10 þ1.9 196
10–18 þ7.1 409 18–30 þ10.3 1174 30–60 þ11.2 274 3–60 þ9.0 2053
Females (all ethnicities) 3–10 þ1.5 88
10–18 þ7.6 233 18–30 þ3.8 350 30–60 þ9.7 98 3–60 þ5.4 769
All ethnicities, all ages, both sexes þ8.0 2822
Abbreviations: FAO/WHO/UNU – Food and Agriculture Organization/World Health Organization/United Nations University; BMR – basal metabolic rate. Source: Henry and Rees37.
Table 5 The percentage by which the FAO/WHO/UNU equations overestimate (þ ) or underestimate (2) BMR in different ethnic groups by sex, all ages 3–60 years
Male Female
Ethnicity Mean % Sample size Mean % Sample size
Philippino þ9.5 172 þ1.1 31 Indian þ12.8 50 þ12.9 7 Japanese þ5.8 202 þ4.6 152 South American þ9.4 941 þ4.8 227 Chinese þ7.6 274 þ3.8 190 Malayan þ9.3 62 No data Javanese þ5.0 86 No data Mayan þ1.5 76 No data Ceylonese þ22.4 125 þ12.5 100 African þ6.5 20 No data Hawaiian þ7.2 19 þ4.5 62 Samoan þ3.3 21 No data All þ9.0 2053 þ5.4 769
Abbreviations: FAO/WHO/UNU – Food and Agriculture Organization/World Health Organization/United Nations University; BMR – basal metabolic rate. Source: Henry and Rees37.
CJK Henry1138
WHO/UNU equations overestimated BMR in a range of
tropical populations (see Tables 4 and 5).
While Schofield’s analysis has served a significant role in
re-establishing the importance of using BMR to predict
human energy requirements, recent workers have sub-
sequently queried the universal validity and application of
these equations38–43. Other authors44,45 have questioned
the equations’ continued use in present day populations
with their secular changes in body weight and body
composition46–48. In contrast to the previous observation
that the FAO/WHO/UNU equations overestimated BMR in
tropical people49, further analysis shows that the FAO/
WHO/UNU equations tend to overestimate BMR in most
populations and these overestimations are not small or
insignificant (see Tables 6 and 7) . A detailed analysis of the
over- and underestimation in BMR using the FAO/WHO/
UNUequationsobserved in children aged 2.5–18 yearswas
reported by several investigators50–53. These childrenwere
studied in a range of countries, including Columbia,
Guatemala and China. While all of the above studies
reported an overestimation of BMR in these children, the
studies by Livingstone et al.54 reported an underestimation
andBandini et al.55 no significant differencewhen the FAO/
WHO/UNU equations were used. BMR studies conducted
more recently have also shown that the predicted values
using the FAO/WHO/UNU equations overestimate BMR in
Asian and Chinese subjects56,57. For example, in the study
by Leung et al.57 the FAO/WHO/UNU equation over-
estimated BMR from the measured values by up to
456 ^ 67 kJ (measured BMR 5.025 ^ 0.967MJ; predicted
Table 6 Comparison of BMR values from the literature: observed values vs. predicted values using the Schofield equations
Investigator Year Subjects Age range (y) Male Female Total % difference
Owen et al. 1986 Mixed Race 18–65 – 34 34 þ11 Americans
Owen et al. 1987 Mixed Race 18–82 48 – 48 þ5 Caucasians African-American Oriental
Soares and Shetty 1988 Indian 18–30 123 123 þ9.3 Urban upper socio-econ 47 þ6.6 Urban lower socio-econ 36 þ12.9 Rural 40 Range þ5.5–12.6
Mifflin et al. 1990 Mixed Race 19–78 251 247 498 þ6.2 (males) Americans þ2.2 (females)
Spurr et al. 1992 Colombian 2–16 153 93 246 Range Males 3–10 þ2.7–9.4
10–18 þ2.5 Females 3–10 þ9.4
10–18 þ1.4 Maffeis et al. 1993 Italian 6–10 62 68 130 No
Males Non-obese 48 – Difference Obese 14 – þ15.3 Females Non-obese – 49 þ8.0 Obese – 19 þ20.0
Rieper et al. 1993 German 14–15 – 11 11 No difference Valencia et al. 1993 Mexican 18–40 32 – 32 þ8.2 Fontville and Ravussin 1993 Mixed Race 7–12
Caucasian 21 21 42 No difference Pima Indian 22 21 43 þ6.7
Piers and Shetty 1993 Indian 18–30 – 60 60 þ9.2 Past Indian Study – 52 þ10.5 European/American – 52 þ4.1
Molnar et al. 1995 Hungarian 10–16 193 178 371 Males Non-obese 116 – þ7.8 Obese 77 – þ13.4 Females Non-obese 119 – þ8.0 Obese 59 – þ5.8
Liu et al. 1995 Chinese 20–78 102 121 223 þ15.1 (males) þ17.9 (females)
Wong et al. 1996 Mixed Race 8–17 – 118 118 Caucasian – 76 No difference African-American – 42 þ8
Piers et al. 1997 Australian 18–30 39 89 128 þ5.3 (males) þ2.2 (females)
Abbreviation: BMR – basal metabolic rate.
Basal metabolic rate studies in humans 1139
5.481 ^ 0.845MJ). Recently, Piers et al.58 reported that the
FAO/WHO/UNU equations overestimated BMR in male
and female Australians.
A survey of the most recent studies (1980–2000) in BMR
suggests that in most cases the current FAO/WHO/UNU
predictive equations overestimate BMR in many commu-
nities. The few exceptions to this general trend are the
reports by Bandini et al.,55 Livingstone et al.,54 Ferro-Luzzi
et al.59 and Yamauchi et al.60. These authors either showed
an underestimation or good agreement with the FAO/
WHO/UNU equations. Contrary to the popular view that
only people in the tropics may have lower BMR, the recent
study by Wong et al.43 showed that BMR in African-
American children aged 8–17 years was 7% lower than
that predicted by the FAO/WHO/UNU equations. What-
ever the reasons for these varied observations, it is clear
that the present FAO/WHO/UNU equation tends to
overpredict BMR in many communities.
Another significant feature of the Schofield database
was that for males aged between 10 and 60 years, over
3000 (50%) data points come from Italian subjects. The
Italian group appear to have a higher BMR per kilogram
than any other Caucasian group49,41. More importantly,
the inclusion of this disproportionately large Italian group
with a higher BMR per kilogram may have artificially
elevated the Schofield predictive equations. Indeed this
view was first expressed by Schofield29. He wrote:
‘The equation for adult males and females (18–30) were
recalculated excluding Italian subjects. The new equations
were:
Males: BMR(MJ/24 h) ¼ 0.0582W þ 3.2399 SEE ¼ 0.6148 Females: BMR(MJ/24 h) ¼ 0.0545W þ 2.5135 SEE ¼ 0.4813 When these equations were used to predict BMR for
Italian subjects there was a highly significant lack of fit for
both males and females’.
The applicability and use of body weight to predict BMR
in various populations depends on the assumption that a
similarity in body composition exists between the surveyed
database and the test population applied to. It is increas-
ingly clear that the subjects from whom the BMR database
(originally from data gathered over 80–90 years ago) was
assembled had a different body composition to that seen
today (Norgan)61. The discussion and debate surrounding
the use and application of the BMR equations were
summarised by Soares et al.44 as follows:
‘Whatever the reason it is becoming increasingly evident
that the equations of Schofield derived from measure-
ments made over 60 years ago are not at present valid for
the precise prediction of BMR of population groups
worldwide’.
Contribution of Italian subjects to the Schofield
database
A closer examination of the Italian data points in the
Schofield database reveals some issues of major concern.
The Italian group represented 3388 BMR data points from
a total of 7173 values in the Schofield database. These 3388
BMR values for Italians came from just nine papers that
were published between 1936 and 1942 and contributed
60–64% of the Schofield database, depending on the age
group under consideration (see Table 7).
A significant feature of the Schofield database was that it
shared a large proportion of the original database initially
identified by Quenouille et al.35 Table 8 illustrates this
point.
The Schofield database contained the same nine Italian
papers first identified by Quenouille et al.35 Quenouille
analysis35 delineated four populations: North European
and Americans; Italians; Asians; and a ‘residual mixed
group’. Within these groups, Italians had the highest BMR
per kilogram. Schofield et al.29 also noted a similar higher
BMR per kg in the Italians. Since the 18–30 year old male
group in the Schofield database contained the largest
number of Italian subjects (1740 out of a total of 2879), this
group may be further analysed when matched for body
size and body mass index (BMI).
Table 9 shows weight, height, BMR and BMR per
kilogram in BMI-matched Italian and North European and
American subjects in the Schofield database. What is clear
is that the Italian subjects show an elevated BMR
(MJ day21 or kJ kg21 per day) when also matched for
BMI62.
Table 7 Contribution of Italian subjects to Schofield database
Gender Age (y) Italians (n) Schofield (n) Percentage
of total
Males 0–3 0 162 0.0 3–10 158 338 46.7
10–18 472 734 64.3 18–30 1740 2879 60.4 30–60 392 646 60.7 60þ 0 50 0.0 Total 2762 4809 57.4
Females 0–3 0 137 0.0 3–10 220 413 53.3
10–18 167 575 29.0 18–30 135 829 16.3 30–60 106 372 28.5 60þ 6 38 15.8 Total 634 2364 26.8
Table 8 Papers and data points shared by Schofield29 and Quenouille et al.35
Investigator Number
of papers Number of data points
Quenouille 89 7434 Schofield 114 7173 Papers common to Quenouille and Schofield
50 6124
CJK Henry1140
The results of this analysis suggest that previous studies
appear to have either overlooked or ignored such large
population differences in BMR per kilogram observed in
the Italians. The nine papers from the Italian investigators
used in the Schofield analysis are shown in Table 10.
All of the Italian studies used the Benedict–Roth
spirometer, an indirect closed circuit method. Our earlier
review has suggested that closed circuit methods are more
likely to lead to higher BMR values compared to open
circuit. The subjects used in the Italian studies were
primarily young, themales tending to lead physically active
lives. This is especially so in the case of labourers and
miners; these occupations are known to have very high
rates of daily energy expenditure. The work of Poehlman
et al.64 indicates that physically active individuals have
higher BMRs compared to sedentary lifestyles. Whatever
the reasons for the elevated BMR noted in the Italians, their
numerical dominance in the Schofield databasemay largely
explain why the Schofield equations (FAO/WHO/UNU
equations) overestimate BMR in present day populations.
Summary
A need to re-assess the FAO/WHO/UNU equations to
predict BMR is prompted by the following:
1. The current FAO/WHO/UNU equations appear to
overestimate BMR in many populations (both in
tropical and temperate regions). Some studies,
however, show good agreement with the FAO/
WHO/UNU equations.
2. If BMR values are to be used to estimate total energy
requirements for prescriptive and diagnostic pur-
poses (in both individuals and populations), any
overestimation in BMR is likely to produce mislead-
ing estimates of energy requirements.
3. The FAO/WHO/UNU equations to predict BMR were
developed using a database that contained a
disproportionate number – 3388 out of 7173 (47%)
– of Italian subjects.
4. These Italian subjects had a higher BMR per kilogram
than any other group in the Schofield database.
5. The numerical dominance of Italians in the Schofield
database and their apparent elevated BMR recorded
may largely explain why the Schofield equations
overestimate BMR in most populations today.
6. It is recommended that the Italian database be
removed from any future analysis as (1) they have
unusually high BMR values (even when normalised
for body weight and BMI); (2) they are over-
represented in the Schofield database; and (3) their
over-representation has meant by definition an
under-representation of other world populations.
7. The Schofield database contained relatively few
subjects from the tropical region (322 Indian and 615
tropical residents) making it a poor representation of
the global population.
8. The primary purpose of collecting BMR measures
during the early part of the 20th century was to
diagnose hypo- and hyperthyroidism.
9. The primary purpose of collecting BMR measures in
recent years has been to estimate total energy
requirements or to better understand energy
regulation.
Table 9 Wilcoxon U-tests of weight, height, BMR and BMR kg21
in BMI matched Italian and North European and American (NE&A) subjects in the Schofield database
BMI range Group n Weight
(kg) Height
(m) BMR
(MJ day21)
BMR (kJ kg21
per day)
18.0–18.9 NE&A 36 54.8 1.72 6279 115 Italian 29 54.0 1.70 6476 120
19.0–19.9 NE&A 63 58.4 1.73 6467 111 Italian 78 56.5* 1.70* 6890*** 122***
20.0–20.9 NE&A 78 61.8 1.74 6635 107 Italian 218 59.6*** 1.70*** 6966** 117***
21.0–21.9 NE&A 85 66.2 1.75 6932 105 Italian 313 62.1*** 1.70*** 7007 113***
22.0–22.9 NE&A 59 68.1 1.74 6932 102 Italian 314 65.3*** 1.71*** 7204* 110***
23.0–23.9 NE&A 38 71.8 1.75 7259 101 Italian 298 67.9*** 1.70*** 7259 107
24.0–24.9 NE&A 18 74.7 1.75 7217 97 Italian 155 70.6* 1.70 7480 106**
25.0– 25.9 NE&A 8 73.9 1.70 7208 98 Italian 53 72.3 1.69 7468 103
Abbreviations: BMR – basal metabolic rate; BMI – body mass index. Significant difference: *P , 0.05; **P , 0.01; ***P , 0.001. Source: Hayter and Henry63.
Table 10 Italian data used in the Schofield database29
Study n Gender Age (y) Subject details
Felloni (1936) 532 Male 19–25 Students of the Royal Fascist Academy Granti and Busca (1941–1942) 186 Male 16–55 Labourers and miners on shift work Lafratta (1937) 213 Male 14–20 Students of Naples Royal Military College Lenti (1937) 525 Male 20–25 Military servicemen Occhiuto and Pepe (1939) 247 Female 20–67 Different social groups Occhiuto and Pepe (1940) 571 Male 22–54 Police officers Pepe (1938) 252 Male 18–24 Students of Royal Naval Academy Pepe and Perrelli (1937) 267 Male 5–16 No details
235 Female 5–12 No details Pepe and Rinaldi (1936) 217 Male 6–16 No details
143 Female 5–12 No details Total 3388
Basal metabolic rate studies in humans 1141
10. This change of emphasis and role of BMR has placed
it within a different nutritional paradigm.
11. If BMR equations are to be used and applied
worldwide, the database must contain a more
representative sample of the world population.
The objective of all consultations is to scientifically progress
and identify fresh ideas and issues. Many of the
technological advances that have emerged during the past
few years have enabled the measurement of BMR to be
conducted with ease and reproducibility. This is an
opportune moment to re-examine the role and place of
BMRmeasurements in estimating total energy requirements
today, using a more representative world population base.
Development of the Oxford database
Initial selection criteria for BMR data in the Oxford
Database: preliminary screening
There has been considerable disagreement in the literature
as to the ‘best’ way in which BMR data should be selected
and collected.
The four methods used to ‘accept’ values for BMR
reported in the literature include:
1. Mean of all determinations (the BMR of the subject was
taken as the mean of all determinations taken on the
subject).
2. First determination (only the first observation was
considered).
3. Lowest of all determinations (only the lowest value
was chosen).
4. Mean of the lower of three duplicates (the lower values
in each of the 3 days were averaged and taken as the
BMR of the subject).
While an element of training has been considered by
several investigators as an important factor in BMR
determinations, the BMR report by Robertson and Reid65
from a large study in Britain has been excluded from wide
use as the investigation used the lowest values recorded
after several bouts of collection. Durnin16 suggested that
no significant effect was produced by the method of
selection of BMR data with the exception of including
lowest BMR results. Apart from excluding BMR values that
were reported as lowest values, all other BMR values were
initially considered for inclusion in the Oxford survey,
prior to further screening.
Computation of new BMR equations (Oxford
equations)
Oxford data compendium
Once such papers were identified (as outlined above), the
literature search produced numerous papers on BMR.
Much of the identification of the early studies on BMR was
done by hand search. This was supplemented by using
MEDLINE. As papers began to accumulate, it rapidly
became clear that the quantity of data gathered was
uneven. The research methods presented the description
of subjects and conditions varied from complete details to
no information. To preserve uniformity and to meet the
criteria defining basal metabolism, it was decided to
include data for further analysis only if the following
information was provided in the papers:
1. Age, weight, gender of subjects
2. Description of experimental conditions and equipment
used to measure BMR.
3. Post-absorptive, rested subjects.
4. Subjects that were described as ‘healthy’ (i.e. not
suffering from any illness).
5. Location/ethnicity of subjects.
Reasons for rejecting data for further analysis included the
following:
1. BMR presented only in terms of surface area (no height
or weight provided), therefore BMR/24 hours could
not be calculated.
2. BMR presented as a percentage deviation of other
standards (usually Harris–Benedict and DuBois).
3. BMR measured on malnourished or sick subjects.
4. BMR measured below 188C.
5. BMR measured at high altitudes.
6. BMR measured in subjects who had eaten or drunk
coffee.
It was not possible to obtain information on the ambient
temperature at the timeof BMRmeasurements in all papers.
However, in papers where temperature was described, if
the ambient temperature was below 188C, such papers
were rejected. In most papers, BMR was expressed as
kcal day21, kJ day21, kcal kg21 per day or kcalm22. When
BMRwas expressed as O2 consumption, with no RQ values
reported, an energy equivalence of 4.9 was used. By the
time the compilation was complete, data were available for
10 552 subjects (5794 males and 4702 females). The data
came from166 separate investigations. In this analysis, only
individual data points were used. Several studies that
presented data as group means were excluded.
In the cases of researchers such as Benedict18,66–75,
Lewis20,76,77, Mason78–81, Nakagawa82–85, Wardlaw86–88
and Wang90–93, even if descriptive details were not
provided in all their papers, they were included for further
analysis as their protocol was detailed (and acceptable) in
the first of their papers.
The Oxford database also excluded all the Italian
subjects due to their unusually high BMR values. To ensure
quality data for the equations to estimate BMR, further
screening took place. All individual data was screened to
identify errors of data input and transcription. Screening
also allowed outlying or extreme cases to be identified and
removed, if appropriate, from the database. As well as
CJK Henry1142
screening data on an individual basis, screening also took
place at study level.
The value of a large database that draws on information
collected by a wide range of investigators rests on the
assumption that all investigators adopted a ‘standard’
practice to measure BMR – which clearly they did not.
Although strict inclusion criteria had been used to develop
the Oxford database, very similar to those adopted by
Schofield et al.29, the fact that such criteria must, of
necessity, rely on published reports of methods and
protocol needs to be recognised.
Computation of equations from Oxford database
A series of plots of BMR against body weight were
performed at six different age groups (0–3 years, 3–10
years, 10–18 years, 18–30 years, 30–60 years and .60
years) for males and females. Representative examples are
shown in Figs 1–4 . These represent BMR vs. body weight
in the Oxford database and compares themwith the Italian
subjects for illustrative purposes. It is evident that the
Italians once again show considerable difference with the
Oxford trendline (Italian trendline – top line on figures;
Oxford regression – top left-hand corner on figures).
To further substantiate why the Italian subjects have
been excluded from the Oxford database, Table 11 shows
descriptive statistics between the Italian subjects and the
rest of the Oxford database. The age bands 10–18 years,
18–30 years and 30–60 years only were chosen for
analysis as they contain the largest number of Italian
subjects. It is evident that the Italian subjects show
significant differences in BMR, even when expressed as
BMRday21 or BMR/kg/body weight.
Table 12 contains the equations for predicting BMR from
weight alone and descriptive statistics for the Oxford
equations.
Equations to predict BMR from weight for six separate
age groups and gender are presented in Table 13, along
with the FAO/WHO/UNU equations for comparison.
Given that a reasonably large number of BMR values
from elderly subjects were available, it was decided to
Fig. 1 Basal metabolic rate (BMR) vs. body weight – males 18–30 years
Fig. 2 Basal metabolic rate (BMR) vs. body weight – males 30–60 years
Fig. 3 Basal metabolic rate (BMR) vs. body weight – females 18–30 years
Fig. 4 Basal metabolic rate (BMR) vs. body weight – females 30–60 years
Basal metabolic rate studies in humans 1143
break down the elderly group into ‘young elderly’ and
‘older elderly’ (see Table 14).
Figures 5–8 illustrate the relationship between body
weight and BMR in the elderly groups.
While Table 13 contains equations for predicting
BMR from weight alone, Table 15 contains equations
using height and weight. To enable the comparison of the
effect of including height as a second variable for males
and females, the equations were re-calculated using
weight/height as independent variables across the entire
age range. Table 16 gives mean values and standard
deviations for the various age groups in the Oxford
analysis.
Any improvement in using height and weight to predict
BMR was tested. Table 17 shows that no significant
advantage was afforded in predicting BMR with the
inclusion of height.
Applications of new BMR equations
Using the Oxford equations, BMR was calculated for a
range of body weights and ages (0–3, 3–10, 10–18,
18–30, 30–60, .60 years) for both males and females
(Tables 18 and 19). It is interesting to note that in
adults the greatest differences in BMR were found in
18–30 and 30–60 year age groups within the lower
body weight ranges (,60 kg) in males. There was a
steadily increasing difference in BMR with decreasing
body weight in these two age groups (18–30 and
30–60 years). For females, differences in BMR were
seen in most of the body weight range notably in the
age groups 18–30, 30–60 and .60 years. In the lower
age groups (0–3 years), the FAO/WHO/UNU equations
appear to underestimate BMR, both in males and
females. At 3–10 years of age, the FAO/WHO/UNU
equations underestimate BMR in males and over-
estimate BMR in females. The differences in BMR
assume significance when one considers that the lower
body weights (50–60 kg) are those that are commonly
observed in many developing countries. Moreover, the
ages (18–60 years) that show the greatest differences in
BMR are ages at which most populations are in a
productive stage of occupational activity. Over- and
under-prediction of BMR in these groups may have
significant influences on estimating their energy
requirements and hence food needs. Table 20
summarises these differences in BMR at various ages.
It is significant to note that the ages at which differences
in BMR were recorded between the new Oxford equations
Table 11 Comparison of descriptive statistics (mean ^ SD) for Italian data vs. all data in Oxford database
Gender Age (y) Source Weight (kg) Height (m) BMI BMR (kJ kg21 per day) BMR (MJ day21)
Males 10–18 All1 40.0 ^ 12.5 1.49 ^ 0.146** 17.7 ^ 2.65*** 143 ^ 22.6*** 5.51 ^ 1.11*** Italian 41.5 ^ 15.4 1.47 ^ 0.171 18.4 ^ 3.10 151 ^ 26.8 5.89 ^ 1.27
18–30 All1 61.0 ^ 11.4*** 1.70 ^ 0.0872 20.9 ^ 2.84*** 106 ^ 12.8*** 6.36 ^ 1.00*** Italian 64.8 ^ 6.83 1.70 ^ 0.0658 22.4 ^ 1.79 111 ^ 11.9 7.15 ^ 0.767
30–60 All1 65.3 ^ 13.0 1.69 ^ 0.0942* 22.8 ^ 3.2* 98.7 ^ 13.6*** 6.35 ^ 1.03*** Italian 65.6 ^ 10.3 1.68 ^ 0.0538 23.3 ^ 3.35 109 ^ 14.5 7.04 ^ 0.772
Females 10–18 All1 43.4 ^ 12.9*** 1.50 ^ 0.113*** 18.8 ^ 3.64*** 126 ^ 24.1*** 5.20 ^ 0.797*** Italian2 28.4 ^ 4.42 1.33 ^ 0.0690 16.0 ^ 1.55 155 ^ 17.5 4.36 ^ 0.475
18–30 All1 53.2 ^ 10.0*** 1.60 ^ 0.0755*** 20.7 ^ 3.18*** 99.8 ^ 12.7*** 5.24 ^ 0.786*** Italian 58.0 ^ 10.3 1.57 ^ 0.0612 23.4 ^ 3.97 106 ^ 14.3 6.04 ^ 0.688
30–60 All1 59.1 ^ 13.7 1.59 ^ 0.0792*** 23.3 ^ 4.48*** 92.0 ^ 14.3*** 5.30 ^ 0.804*** Italian 60.7 ^ 13.0 1.56 ^ 0.605 25.0 ^ 5.00 98.4 ^ 17.8 5.72 ^ 0.605
Significant difference: *P , 0.05; **P , 0.01; ***P , 0.001. 1Excludes Italians 2Italian girls all just over 10 years.
Table 12 Descriptive equations and statistics (mean ^ SD) of Oxford predictive equations for BMR
Gender Age (y) MJ day21 kcal day21 SE n r
Males 0–3 0.255W 2 0.141 61.0W 2 33.7 0.255 277 0.954 3–10 0.0937W þ 2.15 23.3W þ 514 0.328 289 0.827
10–18 0.0769W þ 2.43 18.4W þ 581 0.566 863 0.861 18–30 0.0669W þ 2.28 16.0W þ 545 0.652 2821 0.760 30–60 0.0592W þ 2.48 14.2W þ 593 0.693 1010 0.742 60 þ 0.0563W þ 2.15 13.5W þ 514 0.685 534 0.776
Females 0–3 0.246W 2 0.0965 58.9W 2 23.1 0.242 215 0.960 3–10 0.0842W þ 2.12 20.1W þ 507 0.360 403 0.820
10–18 0.0465W þ 3.18 11.1W þ 761 0.525 1063 0.752 18–30 0.0546W þ 2.33 13.1W þ 558 0.564 1664 0.700 30–60 0.0407W þ 2.90 9.74W þ 694 0.581 1023 0.690 60 þ 0.0424W þ 2.38 10.1W þ 569 0.485 334 0.786
Abbreviation: BMR – basal metabolic rate.
CJK Henry1144
and FAO/WHO/UNU equations coincide with the ages at
which a disproportionate number of Italian subjects were
included in the Schofield database.
Practical examples of how the use of Oxford BMR
equations influence energy requirements
On the basis of the new BMR equations, it is now possible
to predict energy requirements in subjects performing
different tasks.
Tables 21 and 22 predict total energy requirements at
light, moderate and heavy physical activity levels in
18–30 year old males and females. It is apparent, when
applying the Oxford equation for BMR, that a reduction
in total energy requirements ranges from 396 kJ (95 kcal)
to 841 kJ (201 kcal) per day for males, and 202 kJ
(48 kcal) to 863 kJ (206 kcal) per day for females. As
illustrative examples, Tables 23–28 show the effect an
alteration in BMR will have on total energy requirements
and cereal needs in various occupations. In the case of a
subsistence farmer, moderately active, weighing 58 kg,
the reduction in energy needs per day turns out to be
676 kJ (162 kcal) and a reduction in cereal requirement
of 41 g day21 or 15 kg per year. In the case of a 35-year
old male, engaged in heavy physical activity, weighing
65 kg, the reduction in energy needs become 1064 kJ
Fig. 5 Basal metabolic rate (BMR) vs. body weight – males 60–70 years
Fig. 6 Basal metabolic rate (BMR) vs. body weight – males 70þ years
Fig. 7 Basal metabolic rate (BMR) vs. body weight – females 60–70 years
Fig. 8 Basal metabolic rate (BMR) vs. body weight – females 70þ years
Table 13 New Oxford equations with FAO/WHO/UNU equations for comparison
Gender Age (y) BMR Oxford (MJ day21) BMR FAO (MJ day21)
Males 0–3 0.255W 2 0.141 0.255W 2 0.226 3–10 0.0937W þ 2.15 0.0949W þ 2.07
10–18 0.0769W þ 2.43 0.0732W þ 2.72 18–30 0.0669W þ 2.28 0.0640W þ 2.84 30–60 0.0592W þ 2.48 0.0485W þ 3.67 60þ 0.0563W þ 2.15 0.0565W þ 2.04
Females 0–3 0.246W 2 0.0965 0.255W 2 0.214 3–10 0.0842W þ 2.12 0.0941W þ 2.09
10–18 0.0465W þ 3.18 0.0510W þ 3.12 18–30 0.0546W þ 2.33 0.0615W þ 2.08 30–60 0.0407W þ 2.90 0.0364W þ 3.47 60þ 0.0424W þ 2.38 0.0439W þ 2.49
Abbreviations: FAO/WHO/UNU – Food and Agriculture Organization/World Health Organization/United Nations University; BMR – basal metabolic rate.
Table 14 Descriptive equations and statistics (mean ^ SD) of Oxford predictive equations for BMR in the elderly
Gender Age (y) MJ day21 kcal day21 SE n r
Males 60–70 0.0543W þ 2.37 13.0W þ 567 0.697 270 0.766 70 þ 0.0573W þ 2.01 13.7W þ 481 0.667 264 0.779
Females 60–70 0.0429W þ 2.39 10.2W þ 572 0.476 185 0.798 70 þ 0.0417W þ 2.41 10.0W þ 577 0.518 155 0.746
Abbreviation: BMR – basal metabolic rate.
Basal metabolic rate studies in humans 1145
(254 kcal) per day with a reduction in cereal needs of
approximately 65 g. As a final comparison, the energy
requirements of a rural woman in a developing country,
weighing 50 kg are presented.
In summary, the Oxford equations produced lower
BMR values than the FAO/WHO/UNU equations in the
18–30 and 30–60 year old males and in all females over 18
years of age. In the examples cited above (where the
newly calculated BMR was used to estimated energy
requirements), the Oxford BMR equation produced a
significant reduction in total energy and cereal require-
ments per day. Possible explanations for these differences
in BMR when applying the Oxford equations may be
because the Oxford database (1) did not include any of the
elevated BMR values of the Italian subjects and (2)
included a much larger number of people from the
tropical region (see Table 29).
Discussion and areas for future research
One practical use of BMR is in the estimation of energy
requirements for population groups and subsequently
their food needs. The FAO/WHO/UNU report Energy
and Protein Requirements1 made clear for the first time
two main purposes of determining energy requirements.
The first was for prescriptive purposes, i.e. for making
recommendations about the level of consumption that
ought to be maintained in a population; the second, for
diagnostic purposes, i.e. the assessment of the
Table 15 Descriptive equations and statistics (mean ^ SD) of Oxford prediction equations for BMR using height and weight
Gender Age (y) MJ day21 kcal day21 SE n r
Males 0–3 0.118W þ 3.59H 2 1.55 28.2W þ 859H 2 371 0.246 246 0.959 3–10 0.0632W þ 1.31H þ 1.28 15.1W þ 74.2H þ 306 0.322 289 0.835
10–18 0.0651W þ 1.11H þ 1.25 15.6W þ 266H þ 299 0.562 863 0.864 18–30 0.0600W þ 1.31H þ 0.473 14.4W þ 313H þ 113 0.648 2816 0.764 30–60 0.0476W þ 2.26H 2 0.574 11.4W þ 541H 2 137 0.678 1006 0.756 60 þ 0.0478W þ 2.26H 2 1.07 11.4W þ 541H 2 256 0.668 533 0.789
Females 0–3 0.127W þ 2.94H 2 1.20 30.4W þ 703H 2 287 0.232 201 0.964 3–10 0.0666W þ 0.878H þ 1.46 15.9W þ 210H þ 349 0.357 403 0.825
10–18 0.0393W þ 1.04H þ 1.93 9.40W þ 249H þ 462 0.521 1063 0.758 18–30 0.0433W þ 2.57H 2 1.18 10.4W þ 615H 2 282 0.542 1655 0.724 30–60 0.0342W þ 2.10H 2 0.0486 8.18W þ 502H 2 11.6 0.564 1023 0.713 60 þ 0.0356W þ 1.76H þ 0.0448 8.52W þ 421H þ 10.7 0.472 324 0.805
Abbreviation: BMR – basal metabolic rate.
Table 16 Descriptive statistics (mean ^ SD) of data in Oxford database
Gender Age (y) Height (m) Weight (kg) BMI BMR (MJ day21)
Males 0–3 0.4 ^ 0.62 0.65 ^ 0.13 6.3 ^ 3.20 15.1 ^ 2.01 1.47 ^ 0.86 3–10 6.6 ^ 2.04 1.17 ^ 0.13 21.4 ^ 5.14 15.5 ^ 1.26 4.17 ^ 0.58
10–18 12.7 ^ 2.07 1.49 ^ 0.15 40.0 ^ 12.48 17.7 ^ 2.65 5.51 ^ 1.11 18–30 22.7 ^ 2.87 1.70 ^ 0.09 61.0 ^ 11.40 20.9 ^ 2.84 6.36 ^ 1.00 30–60 40.8 ^ 8.72 1.69 ^ 0.09 65.3 ^ 12.98 22.8 ^ 3.24 6.35 ^ 1.03 60 þ 70.9 ^ 7.60 1.70 ^ 0.09 71.3 ^ 14.94 24.6 ^ 4.13 6.17 ^ 1.09 Females
0–3 0.5 ^ 0.71 0.65 ^ 0.14 6.7 ^ 3.40 15.0 ^ 2.31 1.54 ^ 0.87 3–10 7.1 ^ 1.77 1.22 ^ 0.14 23.6 ^ 6.14 15.7 ^ 1.59 4.10 ^ 0.63
10–18 13.0 ^ 2.35 1.50 ^ 0.11 43.4 ^ 12.91 18.8 ^ 3.64 5.20 ^ 0.80 18–30 22.4 ^ 3.01 1.60 ^ 0.08 53.2 ^ 10.04 20.7 ^ 3.18 5.24 ^ 0.79 30–60 41.6 ^ 8.18 1.59 ^ 0.08 59.1 ^ 13.65 23.3 ^ 4.48 5.31 ^ 0.80 60 þ 69.8 ^ 6.88 1.56 ^ 0.09 60.0 ^ 14.52 24.3 ^ 4.78 4.93 ^ 0.78
Abbreviations: BMI – body mass index; BMR – basal metabolic rate.
Table 17 Prediction of BMR from weight and height or weight alone (mean ^ SD)
BMR (MJ day21)
Gender Age (y) Weight alone Weight þ height %
difference P
Males 0–3 1.474 ^ 0.86 1.564 ^ 0.82 26.12 0.169 3–10 4.168 ^ 0.58 4.168 ^ 0.49 20.02 0.465
10–18 5.506 ^ 1.11 5.505 ^ 0.96 þ0.01 0.767 18–30 6.364 ^ 1.00 6.366 ^ 0.77 20.02 0.364 30–60 6.347 ^ 1.03 6.349 ^ 0.78 20.01 0.808 60þ 6.173 ^ 1.09 6.178 ^ 0.86 20.08 0.615
Females 0–3 1.544 ^ 0.87 1.598 ^ 0.84 23.56 0.245 3–10 4.100 ^ 0.63 4.096 ^ 0.52 0.12 0.617
10–18 5.202 ^ 0.80 5.199 ^ 0.60 0.05 0.995 18–30 5.239 ^ 0.79 5.232 ^ 0.57 þ0.14 0.800 30–60 5.306 ^ 0.80 5.307 ^ 0.57 20.02 0.658 60þ 4.931 ^ 0.78 4.934 ^ 0.64 20.05 0.640
Abbreviation: BMR – basal metabolic rate.
CJK Henry1146
adequacy or otherwise of the food needs in a
population. In the factorial estimation of total energy
expenditure (FAO/WHO/UNU)1, a major feature and
component was the estimation of BMR. The measure-
ment and prediction of BMR thus took on a greater
significance. It is important to recognise that the
primary purpose of the early measures of BMR
(1900–1950) was to diagnose hypo- or hyperthyroidism,
not to estimate energy requirements.
The Schofield database comprised 114 published
studies of BMR, totalling 7173 data points. These formed
the basis for the equations used in the FAO/WHO/UNU
document Energy and Protein Requirements1. While
Schofield’s analysis has served a significant role in
re-establishing the importance of using BMR to predict
human energy requirements, recent workers have
subsequently queried the universal validity and appli-
cation of these equations. A survey of the most recent
Table 18 Comparison of Oxford and FAO/WHO/UNU BMR equations in males at various ages (MJ day21)
Age Weight (kg) BMR (Ox) BMR (FAO) Difference
from FAO (%)a
0–3 3 0.624 0.5390 215.77 5 1.134 1.0490 28.10
10 2.409 2.3240 23.66 12 2.919 2.8340 23.00 15 3.684 3.5990 22.36 18 4.449 4.3640 21.95 20 4.959 4.8740 21.74
3–10 12 3.2744 3.2088 22.04 15 3.5555 3.4935 21.77 18 3.8366 3.7782 21.55 22 4.2114 4.1578 21.29 25 4.4925 4.4425 21.13 30 4.961 4.9170 20.89 35 5.4295 5.3915 20.70 45 6.3665 6.3405 20.41 50 6.835 6.8150 20.29
10–18 25 4.3525 4.5500 þ4.34 30 4.737 4.9160 þ3.64 35 5.1215 5.2820 þ3.04 45 5.8905 6.0140 þ2.05 55 6.6595 6.7460 þ1.28 65 7.4285 7.4780 þ0.66 75 8.1975 8.2100 þ0.15 95 9.7355 9.6740 20.64
105 10.5045 10.4060 20.95 18–30 50 5.6250 6.0400 þ6.87
55 5.9595 6.3600 þ6.30 60 6.294 6.6800 þ5.78 65 6.6285 7.0000 þ5.31 75 7.2975 7.6400 þ4.48 85 7.9665 8.2800 þ3.79 95 8.6355 8.9200 þ3.19
105 9.3045 9.5600 þ2.67 30–60 50 5.440 6.0950 þ10.75
55 5.736 6.3375 þ9.49 60 6.032 6.5800 þ8.33 65 6.328 6.8225 þ7.25 75 6.92 7.3075 þ5.30 85 7.512 7.7925 þ3.60 95 8.104 8.2775 þ2.10
105 8.696 8.7625 þ0.76 60 þ 55 5.2465 5.1475 21.92
60 5.528 5.4300 21.80 65 5.8095 5.7125 21.70 75 6.3725 6.2775 21.51 85 6.9355 6.8425 21.36 95 7.4985 7.4075 21.23
100 7.78 7.6900 21.17
Abbreviations: FAO/WHO/UNU – Food and Agriculture Organization/World Health Organization/United Nations University; BMR – basal metabolic rate. a þ indicates that FAOs formulae give higher values, and 2 indicates lower values.
Table 19 Comparison of Oxford and FAO/WHO/UNU BMR equations in females at various ages (MJ day21)
Age Weight (kg) BMR (Ox) BMR (FAO) Difference from
FAO (%)a
0–3 3 0.6415 0.5510 216.42 5 1.1335 1.0610 26.83
10 2.3635 2.3360 21.18 12 2.8555 2.8460 20.33 15 3.5935 3.6110 þ0.48 18 4.3315 4.3760 þ1.02 20 4.8235 4.8860 þ1.28
3–10 12 3.1304 3.2192 þ2.76 15 3.383 3.5015 þ3.38 18 3.6356 3.7838 þ3.92 22 3.9724 4.1602 þ4.51 25 4.225 4.4425 þ4.90 30 4.646 4.9130 þ5.43 35 5.067 5.3835 þ5.88 45 5.909 6.3245 þ6.57 50 6.33 6.7950 þ6.84
10–18 25 4.3425 4.3950 þ1.19 30 4.575 4.6500 þ1.61 35 4.8075 4.9050 þ1.99 45 5.2725 5.4150 þ2.63 55 5.7375 5.9250 þ3.16 65 6.2025 6.4350 þ3.61 75 6.6675 6.9450 þ4.00 95 7.5975 7.9650 þ4.61
105 8.0625 8.4750 þ4.87 18–30 50 5.060 5.1550 þ1.84
55 5.333 5.4625 þ2.37 60 5.606 5.7700 þ2.84 65 5.879 6.0775 þ3.27 75 6.425 6.6925 þ4.00 85 6.971 7.3075 þ4.60 95 7.517 7.9225 þ5.12
105 8.063 8.5375 þ5.56 30–60 50 4.935 5.2900 þ6.71
55 5.1385 5.4720 þ6.09 60 5.342 5.6540 þ5.52 65 5.5455 5.8360 þ4.98 75 5.9525 6.2000 þ3.99 85 6.3595 6.5640 þ3.12 95 6.7665 6.9280 þ2.33
105 7.1735 7.2920 þ1.63 60 þ 55 4.712 4.9045 þ3.92
60 4.924 5.1240 þ3.90 65 5.136 5.3435 þ3.88 75 5.56 5.7825 þ3.85 85 5.984 6.2215 þ3.82 95 6.408 6.6605 þ3.79
100 6.62 6.8800 þ3.78
Abbreviations: FAO/WHO/UNU – Food and Agriculture Organization/World Health Organization/United Nations University; BMR – basal metabolic rate. a þ indicates that FAOs formulae give higher values, and 2 indicates lower values.
Basal metabolic rate studies in humans 1147
studies (1980–2000) in BMR suggests that in most cases
the current FAO/WHO/UNU predictive equations over-
estimate BMR in many communities.
It is concluded that the over-representation of BMR
values obtained from Italian subjects – 3388 out of 7173 –
(who had a higher BMRkg21) in the Schofield1 database
may have resulted in the FAO/WHO/UNU predictive
equations to overestimate BMR in contemporary popu-
lations.
A series of new equations (Oxford equations) have been
developed using a data set of 10 552 BMR values that (1)
excluded all the Italian subjects and (2) included a much
larger number (4018) of people from the tropics. In general,
the Oxford equations tend to produce lower BMR values
than the current FAO/WHO/UNU equations in 18–30 and
30–60 year oldmales and in all females over 18 years of age
(see pages 33–42 for detailed discussion). The objective of
all consultations is to scientifically progress and identify
fresh ideas and issues. This is an opportune moment to re-
examine the role and place of BMR measurements in
estimating total energy requirements today. The Oxford
equations use and future application will surely depend on
their ability to predict more accurately the BMR in
contemporary populations.
Table 20 Summary of difference between the Oxford and FAO equations
Sex and age band (male/female, years)
Oxford equation (MJ day21)
Schofield (1985) equation (MJ day21) Difference
Males, 0–3 0.225W 2 0.141 0.255W 2 0.226 Large difference at low body weights Females, 0–3 0.246W 2 0.0965 0.255W 2 0.214 Large difference at low body weights Males, 3–10 0.0937W þ 2.15 0.0949W þ 2.07 No significant difference Females, 3–10 0.0842W þ 2.12 0.0941W þ 2.09 Difference at high body weights Males, 10–18 0.0769W þ 2.43 0.0732W þ 2.72 Difference at low body weights Females, 10–18 0.0465W þ 3.18 0.0510W þ 3.12 Difference at high body weights Males, 18–30 0.0669W þ 2.28 0.0640W þ 284 Large difference at body weight , 75 kg Females, 18–30 0.0546W þ 2.33 0.0615W þ 2.08 Large difference at body weight , 75 kg Males, 30–60 0.0592W þ 2.48 0.0485W þ 3.67 Very large difference at body weight , 75 kg Females, 30–60 0.0407W þ 2.90 0.0364W þ 3.47 Large difference at body weight , 65 kg Males, 60 þ 0.0563W þ 2.15 0.0565W þ 2.04 No significant difference Females, 60 þ 0.0424W þ 2.38 0.0439W þ 2.49 Difference across body weights
Abbreviation: FAO – Food and Agriculture Organization.
Table 21 Differences in total EE for light, moderate and high activity levels in males (18–30 years) using Oxford and FAO/WHO/UNU equations (MJ day21)
Weight BMR (Ox)
BMR (FAO)
EE (Ox)
EE (FAO)
Difference from FAO (kJ)
Light Activity (1.55) 55 5.9595 6.3600 9.237225 9.858 2620.77 60 6.2940 6.6800 9.7557 10.354 2598.30 65 6.6285 7.0000 10.27418 10.85 2575.82 75 7.2975 7.6400 11.31113 11.842 2530.88 85 7.9665 8.2800 12.34808 12.834 2485.93 95 8.6355 8.9200 13.38503 13.826 2440.98 105 9.3045 9.5600 14.42198 14.818 2396.02 Moderate Activity (1.76) 55 5.9595 6.3600 10.48872 11.1936 2704.88 60 6.2940 6.6800 11.07744 11.7568 2679.36 65 6.6285 7.0000 11.66616 12.32 2653.84 75 7.2975 7.6400 12.8436 13.4464 2602.80 85 7.9665 8.2800 14.02104 14.5728 2551.76 95 8.6355 8.9200 15.19848 15.6992 2500.72 105 9.3045 9.5600 16.37592 16.8256 2449.68 High Activity (2.10) 55 5.9595 6.3600 12.51495 13.356 2841.05 60 6.2940 6.6800 13.2174 14.028 2810.60 65 6.6285 7.0000 13.91985 14.7 2780.15 75 7.2975 7.6400 15.32475 16.044 2719.25 85 7.9665 8.2800 16.72965 17.388 2658.35 95 8.6355 8.9200 18.13455 18.732 2597.45 105 9.3045 9.5600 19.53945 20.076 2536.55
Abbreviations: EE – energy expenditure; FAO/WHO/UNU – Food and Agriculture Organization/World Health Organization/United Nations Univer- sity; BMR – basal metabolic rate.
Table 22 Differences in total EE for light, moderate and high activity levels in females (18–30 years) using Oxford and FAO/WHO/UNU equations (MJ day21)
Weight BMR (Ox)
BMR (FAO)
EE (Ox)
EE (FAO)
Difference from FAO (kJ)
Light activity (1.56) 55 5.3330 5.4625 8.32 8.5215 2202.02 60 5.6060 5.7700 8.75 9.0012 2255.84 65 5.8790 6.0775 9.17 9.4809 2309.66 75 6.4250 6.6925 10.02 10.4403 2417.3 85 6.9710 7.3075 10.87 11.3997 2524.94 95 7.5170 7.9225 11.73 12.3591 2632.58 105 8.0630 8.5375 12.58 13.3185 2740.22 Moderate activity (1.64) 55 5.3330 5.4625 8.75 8.9585 2212.38 60 5.6060 5.7700 9.19 9.4628 2268.96 65 5.8790 6.0775 9.64 9.9671 2325.54 75 6.4250 6.6925 10.54 10.9757 2438.7 85 6.9710 7.3075 11.43 11.9843 2551.86 95 7.5170 7.9225 12.33 12.9929 2665.02 105 8.0630 8.5375 13.22 14.0015 2778.18 High activity (1.82) 55 5.3330 5.4625 9.71 9.94175 2235.69 60 5.6060 5.7700 10.20 10.5014 2298.48 65 5.8790 6.0775 10.70 11.06105 2361.27 75 6.4250 6.6925 11.69 12.18035 2486.85 85 6.9710 7.3075 12.69 13.29965 2612.43 95 7.5170 7.9225 13.68 14.41895 2738.01 105 8.0630 8.5375 14.67 15.53825 2863.59
Abbreviations: EE – energy expenditure; FAO/WHO/UNU – Food and Agriculture Organization/World Health Organization/United Nations Univer- sity; BMR – basal metabolic rate.
CJK Henry1148
Areas for further research
1. It is recommended that a more detailed analysis of BMR
in children aged between 10 and 18 years and from
different communities be undertaken. A break down of
the age band into more physiologically acceptable
ranges, e.g. 10–12, 12–15 and 15–18 years, is
recommended.
2. There is an urgent need to develop age and gender
specific BMR equations taking into account the
stages in pubertal development (Tanner rating).
Because of the rapid changes during puberty
(changes in body composition, hormone levels,
growth), even small age differences may cause large
changes in metabolic rate.
3. There is a glaring absence of BMR data from mainland
China and Africa. It is recommended that BMR values
are collected from China and other developing
countries, especially from young children and the
elderly.
4. While BMR data collection in the elderly living in
developing countries should be encouraged, the
present age band for the elderly should be further
refined to the following groups: 60–75, 76–85 and
.85 years.
Table 23 Energy requirement of a subsistence farmer (moderate activity work) using FAO/WHO/UNU equations (age: 25 years, weight: 58 kg, height: 1.61 m, BMI: 22.4)
Hours kcalth kJ
In bed at 1.0 £ BMR 8 520 2170 Occupational activities at 2.7 £ BMR 7 1230 5150 Discretionary activities:
–Socially desirable and household tasks at 3.0 £ BMR 2 390 1630 –Cardiovascular and muscular maintenance–not needed if moderately active –
For residual time, energy needs at 1.4 £ BMR 7 640 2680 Total ¼ 1.78 £ BMR 2780 11630
Abbreviations: FAO/WHO/UNU – Food and Agriculture Organization/World Health Organization/United Nations Univer- sity; BMI – body mass index; BMR – basal metabolic rate. Source: FAO/WHO/UNU1. Estimated BMR: 65 kcalth (273 kJ)/h.
Table 24 Energy requirement of a subsistence farmer (moderate activity work) using Oxford equations (age: 25 years, weight: 58 kg, height: 1.61 m, BMI: 22.4)
Hours kcalth kJ
In bed at 1.0 £ BMR 8 491 2052 Occupational activities at 2.7 £ BMR 7 1160 4849 Discretionary activities:
–Socially desirable and household tasks at 3.0 £ BMR 2 368 1539 –Cardiovascular and muscular maintenance–not needed if moderately active –
For residual time, energy needs at 1.4 £ BMR 7 601 2514 Total ¼ 1.78 £ BMR 2621 10 954
Abbreviations: BMI – body mass index; BMR – basal metabolic rate. Reduction in energy requirements per day ¼ 676 kJ (162 kcal). Reduction in cereal requirements per day ¼ 41 g (assuming energy value of 16.3 kJ g21 – raw rice (McCance and Wid- dowson94)) Reduction in cereal requirements per year ¼ 15.0 kg. Estimated BMR: 61.5 kcalth (257 kJ) per hour.
Table 25 Energy requirement for a male engaged in heavy work using the Oxford equations (age: 35 years, weight: 65 kg, height: 1.72 m, BMI: 22)
Hours kcalth kJ
In bed at 1.0 £ BMR 8 545 2280 Occupational activities at 3.8 £ BMR 8 2070 8660 Discretionary activities at 3.0 £ BMR 1 205 860 For residual time, maintenance energy needs
at 1.4 £ BMR 7 670 2800
Total ¼ 2.14 £ BMR 3490 14 580
Abbreviations: BMI – body mass index; BMR – basal metabolic rate. Source: FAO/WHO/UNU1. Estimated BMI: 68 kcalth (284 kJ) per hour.
Table 26 Energy requirement for a male engaged in heavy work using the Oxford equations (age 35 years, weight 65 kg, height 1.72 m, BMI 22)
Hours kcalth kJ
In bed at 1.0 £ BMR 8 505 2112 Occupational activities at 3.8 £ BMR 8 1920 8025 Discretionary activities at 3.0 £ BMR 1 189 792 For residual time, maintenance energy needs
at 1.4 £ BMR 7 619 2587
Total ¼ 2.14 £ BMR 3233 13 516
Abbreviations: BMI – body mass index; BMR – basal metabolic rate. Reduction in energy requirements per day ¼ 1064 kJ (254 kcal). Reduction in cereal requirements per day ¼ 65 g (assuming energy value of 16.3 kJ g21 – raw rice (McCance and Widdowson94). Reduction in cereal requirements per year ¼ 23.7 kg. Estimated BMR: 63 kcalth (264 kJ) per hour.
Basal metabolic rate studies in humans 1149
Acknowledgements
The author wishes to thank IDECG Switzerland, Nestle
Foundation and FAO Rome, for supporting this analysis.
The efforts of David Martineau, Darren Massey and
Helen Lightowler are also gratefully acknowledged.
Special thanks to Drs. Butte, Chen, Goldberg, Jones,
Muhilal, Poehlman, Pullicino, Reilly, Soares, Shetty,
Westerterp and Wong for sending their raw data on BMR
so readily.
References
1 FAO/WHO/UNU. Energy and Protein Requirements. Report of a Joint FAO/WHO/UNU Expert Consultation. Technical Report Series No. 724. Geneva: World Health Organization, 1985.
2 Mitchell HH. Comparative Nutrition of Man and Domestic Animals. Vol. I. New York: Academic Press, 1964.
3 Aub JC, DuBois EF. Clinical calorimetry: the basal metabolism of dwarfs and legless men with observations on the specific dynamic action of protein. Archives of Internal Medicine 1917; 19: 840–64.
4 Aub JC, DuBois EF. The basal metabolism of old men. Archives of Internal Medicine 1917; 19: 823–31.
5 Krogh A. Determination of standard (basal) metabolism of patients by recording apparatus. Boston Medical Surgery Journal 1923; 189: 313–7.
6 Blaxter K. Energy Metabolism in Animals and Man. Cambridge: Cambridge University Press, 1989.
7 DuBois EF, Ebaugh FG, Hardy JD. Basal heat production and elimination of thirteen normal women at temperatures from 22–358C. Journal of Nutrition 1952; 48: 257–93.
8 Wilkerson JE, Raven PB, Horvath SM. Critical temperature of unacclimatized male Caucasians. Journal of Applied Physi- ology 1972; 33: 451–5.
9 Pi-Suner J. Studies in racial metabolism: basal metabolism of the Araucanian Mapuches. American Journal of Physiology 1933; 105: 383–8.
10 Roth P, Buckingham PE. The criteria of a dependable basal metabolism report. American Journal of Clinical Pathology 1939; 9: 79–92.
11 Harris JA, Benedict FG. A Biometric Study of Basal Metabolism in Man. Publication No. 279. Washington: Carnegie Institute of Washington, 1919.
12 Bedale EM. Energy expenditure and food requirements of children at school. Proceedings of the Royal Society of London 1923; 94: 368–404.
13 FAO. Calorie Requirements. Report of the Second Commit- tee on Calorie Requirements. Nutritional Studies No. 15. Rome: Food and Agriculture Organization, 1957.
14 Roth PR. Portable respiration apparatus for clinical use. Boston Medical Surgery Journal 1922; 186: 457.
15 Boothby WM, Berkson J, Dunn HL. Studies of the energy metabolism of normal individuals: a standard for basal metabolism, with a nomogram for clinical application. American Journal of Physiology 1936; 116: 468–84.
16 Durnin JVGA. Basal Metabolic Rate in Man. Working paper submitted to the Joint FAO/WHO/UNU Expert Consultation on Energy and Protein Requirements. Rome: Food and Agriculture Organization, 1981
17 Krogh A. Respiratory Exchange of Animals and Man. New York: Longmans, 1916.
18 Benedict FG. Portable respiration apparatus for clinical use. Boston Medical Surgery Journal 1918; 178: 667.
19 Lewis RC, Iliff A, Duval AM. Further consideration of the effect of altitude on basal metabolism. The Journal of Nutrition 1943; 26: 175–85.
20 Lewis RC, Iliff A, Duval AM. The comparative accuracy of the closed circuit bedside method and the open circuit chamber procedure for the determination of basal metab- olism. Journal of Laboratory Clinical Medicine 1943; 28: 1238–45.
Table 27 Energy requirement of a rural woman in a developing country using FAO/WHO/UNU equations (age: 35 years, weight: 50 kg, height: 1.6 m, BMI: 19.5)
Hours kcalth KJ
In bed at 1.0 £ BMR 8 425 1780 Occupational activities:
–Housework, preparing food, etc, at 2.7 £ BMR
3 430 1800
–Working in fields, at 2.8 £ BMR 4 595 2490 Discretionary activities at 2.5 £ BMR 2 265 1110 For residual time, energy needs
at 1.4 £ BMR 7 520 2180
Total ¼ 1.76 £ BMR 2235 9360
Abbreviations: FAO/WHO/UNU – Food and Agriculture Organization/World Health Organization/United Nations University; BMI – body mass index; BMR – basal metabolic rate. Source: FAO/WHO/UNU1. Estimated BMR: 53 kcalth (220 kJ) per hour.
Table 28 Energy requirement of a rural woman in a developing country using Oxford equations (age: 35 years, weight: 50 kg, height: 1.6 m, BMI: 19.5)
Hours kcalth KJ
In bed at 1.0 £ BMR 8 392 1640 Occupational activities:
–Housework, preparing food, etc, at 2.7 £ BMR
3 397 1660
–Working in fields, at 2.8 £ BMR 4 549 2296 Discretionary activities at 2.5 £ BMR 2 245 1025 For residual time, energy needs
at 1.4 £ BMR 7 480 2009
Total ¼ 1.76 £ BMR 2063 8630
Abbreviations: BMI – body mass index; BMR – basal metabolic rate. Reduction in energy requirements per day ¼ 730 kJ (175 kcal). Reduction in cereal requirements per day ¼ 45 g (assuming energy value of 16.3 kJ g21 – raw rice (McCance and Widdowson94) Reduction in cereal requirements per year ¼ 16.4 kg Estimated BMR: 49 kcalth (206 kJ) per hour.
Table 29 Comparison between Oxford database and Schofield database
Number of papers
Number of data points %
Oxford database 166 10 552 Common to Schofield and
Oxford database 77 4039
New in Oxford database 89 6513 Tropical subjects in
Schofield database 937
Percentage of tropical subjects in Schofield database
13
Tropical subjects in Oxford database
4018
Percentage of tropical subjects in Oxford database
38
CJK Henry1150
21 Lewis RC, Duval AM, Iliff A. Standards for the basal metabolism of children from 2 to 15 years of age, inclusive. Journal of Pediatrics 1943; 23: 1–18.
22 Willard HN, Wolf GA Jr. A source of error in the determination of basal metabolism by the closed-circuit technique. Annual International Medicine 1951; 34: 148–62.
23 Segal KR. Comparison of indirect calorimetric measurement of resting energy expenditure with a ventilated hood, face mask, and mouthpiece. American Journal of Clinical Nutrition 1987; 45: 1420–3.
24 Shetty PS, Soares MJ, Sheela ML. Basal Metabolic Rates of South Indian Males. Bangalore: Food and Agriculture Organization, 1986.
25 Soares MJ, Sheela ML, Kurpad AV, Kulkarni RN, Shetty PS. The influence of different methods on basal metabolic rate measurements in human subjects. American Journal of Clinical Nutrition 1989; 50: 731–6.
26 Consolazio CF, Johnson RE, Pecora LJ. Physiological Measurements of Metabolic Function in Man. New York: McGraw-Hill, 1963.
27 Clark HD, Hoffer LF. Reappraisal of the resting metabolic rate of normal young men. American Journal of Clinical Nutrition 1991; 53: 21–6.
28 Fowler WS, Blackburn CM, Helmholz HF Jr. Determination of basal rate of oxygen consumption by open and closed- circuit methods. Journal of Clinical Endocrinological Metabolism 1957; 17: 786–96.
29 Schofield WN, Schofield C, James WPT. Basal metabolic rate – review and prediction, together with an annotated bibliography of source material. Human Nutrition Clinical Nutrition 1985; 39C: 5–96.
30 DuBois D, DuBois EF. The measurements of the surface area of man. Archives of Internal Medicine 1915; 15: 868–75.
31 Daly JM, Heymsfield SB, Head CA, Harvey LP, Nixon DW, Katzeff H, Grossman GD. Human energy requirements: over estimation by widely used prediction equation. The American Journal of Clinical Nutrition 1985; 42: 1170–4.
32 Alpers PH, Stevenson WF, Bier DM. Manual of Nutritional Therapeutics. Philadelphia: Lippincott Williams and Wilkins, 1995.
33 Boothby WM, Sandiford I. A comparison of the DuBois and the Harris and Benedict normal standards for the estimation of basal metabolic rate. Journal of Biological Chemistry 1922; 54: 767–81.
34 Boothby WM, Sandiford I. Summary of the basal metabolic data on 8614 subjects with special reference to the normal standards for the estimation of the basal metabolic rate. Journal of Biological Chemistry 1922; 54: 783–803.
35 Quenouille MH, Boyne AW, Fisher WB, Leitch I. Statistical Studies of Recorded Energy Expenditure of Man. Basal Metabolism Related to Sex, Stature, Age, Climate, and Race. Commonwealth Bureau of Animal Nutrition. Technical Communication No. 17. Aberdeen: Commonwealth Agri- cultural Bureau, 1951.
36 de Almeida AO. L’émission de chaleur. Le métabolisme basal et le métabolisme mimimum de l‘homme noir tropical. Journal de Physiologie et Pathologie Géneral 1921; 18: 958–64.
37 Henry CJK, Rees DG. New predictive equations for the estimation of basal metabolic rate in tropical peoples. European Journal of Clinical Nutrition 1991; 45: 177–85.
38 de Boer JO, van Es AJH, Voorrips LE, Blockstra F, Vogt JE. Energy metabolism and requirements in different ethnic groups. European Journal Clinical Nutrition 1988; 42: 983–97.
39 Arciero PJ, Goran MI, Gardner AM, Ades PA, Tyzbir RS, Poehlman ET. A practical equation to predict resting metabolic rate in older females. Journal of American Geriatric Society 1993; 41: 389–95.
40 Piers LS, Shetty PS. Basal metabolic rates of Indian women. European Journal of Clinical Nutrition 1993; 47: 586–91.
41 Hayter JE, Henry CJK. A re-examination of basal metabolic rate predictive equations: the importance of geographic origin of subjects in sample selection. European Journal of Clinical Nutrition 1994; 48: 702–7.
42 Valencia ME, Moya SY, McNeill G, Haggarty P. Basal metabolic rate and body fatness of adult men in northern Mexico. European Journal Clinical Nutrition 1994; 48: 205–11.
43 Wong WW, Butte NF, Hergenroeder AC, Hill RB, Stuff JE, Smith EO. Are basal metabolic rate prediction equations appropriate for female children and adolescents? Journal of Applied Physiology 1996; 81: 2407–14.
44 Soares MJ, Francis DG, Shetty PS. Predictive equations for basal metabolic rates of Indian males. European Journal of Clinical Nutrition 1993; 47: 389–94.
45 Cruz CM, daSilva AF, dos Anjos LA. A taxa metabica basal e superestimade pelas equa co es preditivas em universitaras do Rio de Janeiro, Brasil. Archivio Latino americano Nutricion 1999; 49: 232–7.
46 Dietz WH, Bandini LG, Schoeller DA. Estimates of metabolic rate in obese and non-obese adolescents. Journal of Pediatrics 1991; 118: 140–9.
47 Cunningham JJ. A reanalysis of the factors influencing basal metabolic rate in normal adults. American Journal of Clinical Nutrition 1980; 33: 2372–3374.
48 Cunningham JJ. Body composition as a determinant of energy expenditure: a synthetic review and a proposed general prediction equation. American Journal of Clinical Nutrition 1991; 54: 963–9.
49 Hayter J, Henry CJK. Basal metabolic rate in human subjects migrating between tropical and temperate regions – a longitudinal study and review of previous work. European Journal of Clinical Nutrition 1993; 47: 724–34.
50 Ho Z-C, Zi HM, Bo L, Ping H. Energy expenditure of pre school children in a subtropical area. World Reviews in Nutrition Dietetics 1988; 57: 75–94.
51 Min Q, Ho Z-C. The basal metabolism rate of adolescent girls in the sub-tropical areas of China. Acta Nutrimenta Sinica 1991; 13: 252–8.
52 Spurr GB, Reina JC, Hoffmann RG. Basal metabolic rate of Colombian children 2–16 years of age: ethnicity and nutritional status. American Journal of Clinical Nutrition 1992; 56: 623–9.
53 Torun B, Davies PSW, Livingstone MBE, Paolisso M, Sackett R, Spurr GB. Energy requirements and dietary energy recommendations for children and adolescents 1 to 18 years old. European Journal of Clinical Nutrition 1996; 50(Suppl. 1): S37–S81.
54 Livingstone MBE, Coward WA, Prentice AM, Davies PSW, Strain JJ, McKenna PG, Mahoney CA, White JA, Stewart CM, Kerr MJJ. Daily energy expenditure in free living children: comparison of heart-rate monitoring with the doubly labelled water (2H2
18O) method. American Journal of Clinical Nutrition 1992; 56: 343–52.
55 Bandini LG, Schoeller DA, Dietz WH Jr. Energy expenditure in obese and non-obese adolescents. Pediatric Research 1990; 27: 198–203.
56 Case KO, Brahler CJ, Heiss C. Resting energy expenditures in Asian women measured by indirect calorimetry are lower than expenditures calculated from prediction equations. Journal of American Dietetic Association 1997; 97: 1288–92.
57 Leung R, Woo J, Chan D, Tang N. Validation of prediction equations for basal metabolic rate in Chinese subjects. European Journal of Clinical Nutrition 2000; 54: 551–4.
58 Piers LS, Diffey B, Soares MJ, Frandsen SL, McCormack LM, Lutschini MJ, O’Dea K. The validity of predicting the basal metabolic rate of young Australian men and women. European Journal of Clinical Nutrition 1997; 51: 333–7.
Basal metabolic rate studies in humans 1151
59 Ferro-Luzzi A, Petracchi C, Kuriyan R, Kurpad AV. Basal metabolism of weight-stable chronically undernourished men and women: lack of metabolic adaptation and ethnic differences. American Journal of Clinical Nutrition 1997; 66: 1086–93.
60 Yamauchi T, Ohtsuka R. Basal metabolic rate and energy costs at rest and during exercise in rural- and urban-dwelling Papua New Guinea Highlanders. European Journal of Clinical Nutrition 2000; 54: 494–9.
61 Norgan NG. Population differences in body composition in relation to the Body Mass Index. European Journal of Clinical Nutrition 1994; 48(Suppl. 3): S10–S27.
62 Shetty PS, Henry CJK, Black AE, Prentice AM. Energy requirements of adults: an update on basal metabolic rates (BMRs) and physical activity levels (PALs). European Journal of Clinical Nutrition 1996; 50(Suppl. 1): S11–S23.
63 Hayter J, Henry CJK. Basal metabolic rate in human subjects migrating between tropical and temperate regions – a longitudinal study and review of previous work. European Journal of Clinical Nutrition 1993; 47: 724–34.
64 Poehlman ET, Melby CL, Badylak SF. Resting metabolic rate and post-prandial thermogenesis in highly trained and untrained males. American Journal of Clinical Nutrition 1988; 47: 793–8.
65 Robertson JD, Reid DD. Standards for the basal metabolism of normal people in Britain. The Lancet 1952; 1: 940–9.
66 Benedict FG, Garven HSD. The basal metabolism of male Chinese in Manchuria. Chinese Journal of Physiology 1936; 10: 141–8.
67 Benedict FG, Meyer MH. The basal metabolism of American- born Chinese girls. Chinese Journal of Physiology 1933; 7: 45–60.
68 Benedict FG, Miles WR, Roth P, Monmouth Smith H. Human Vitality and Efficiency Under Prolonged Restricted Diet. Washington: Carnegie Institution of Washington, 1919, 280.
69 Benedict FG, Talbot FB. The Gaseous Metabolism of Infants. Washington: Carnegie Institution of Washington, 1914, 201.
70 Benedict FG. Age and basal metabolism of adults. Journal of Biological Chemistry 1928; 80: 650–64.
71 Benedict FG. Basal metabolism data on normal men and women (series II) with some considerations on the use of prediction standards. Journal of Biological Chemistry 1928; 8: 607–20.
72 Benedict FG. Factors affecting basal metabolism. Journal of Biological Chemistry 1915; 20: 263–99.
73 Benedict FG. Metabolism and Growth from Birth to Puberty. Washington: Carnegie Institution of Washington, 1921, 302.
74 Benedict FG. Old age and basal metabolism. New England Journal of Medicine 1935; 212: 1111–23.
75 Benedict FG. The basal metabolism of girls 12 to 17 years of age. Proceedings of the National Academy of Science 1921; VII: 10–13.
76 Lewis RC, Kinsman GM, Iliff A. The basal metabolism of normal boys and girls from two to twelve years old, inclusive. American Journal of Diseases in Childhood 1937; 53: 348–428.
77 Lewis WH. Changes with age in the basal metabolic rate in adult men. American Journal of Physiology 1934; 12: 502–17.
78 Mason ED, Benedict FG. The basal metabolism of South Indian Women. Independent Journal of Medical Research 1931; 19: 75–98.
79 Mason ED, Benedict FG. The effect of sleep on human basal metabolism, with particular reference to South Indian women. American Journal of Physiology 1934; 108: 377–83.
80 Mason ED, Jacob M. Variations in basal metabolic rate: responses to changes between tropical and temperate climates. Human Biology 1972; 44: 141–72.
81 Mason ED. The basal metabolism of European women in South India and the effect of change of climate on European and South Indian Women. Journal of Nutrition 1934; 8: 695–713.
82 Nakagawa I. Growth and basal metabolism. Basal metab- olism of preschool children. American Journal of Diseases in Childhood 1934; 47: 963–9.
83 Nakagawa I. Growth and basal metabolism. Basal metab- olism of elementary school children. American Journal of Diseases in Childhood 1934; 48: 34–8.
84 Nakagawa I. Growth and basal metabolism. Changes in the basal metabolism of children during a year. American Journal of Diseases in Childhood 1935; 49: 1232–9.
85 Nakagawa I. Growth and basal metabolism. Basal metab- olism of high school children. American Journal of Diseases in Childhood 1937; 51: 985–9.
86 Wardlaw HSH, Davies HW, Joseph MR. Energy metabolism and insensible perspiration of Australian Aborigines. Australian Journal of Experimental Biology 1934; 12: 63–73.
87 Wardlaw HSH, Horsley CH. The basal metabolism of some Australian Aborigines. Australian Journal of Experimental Biology 1928; 5: 263–72.
88 Wardlaw HSH, Lawrence WJ. Further observations on the basal metabolism of Australian Aborigines. Australian Journal of Experimental Biology 1932; 10: 157–65.
89 Wang CC, Hawks JE. Basal metabolism of twenty-one Chinese children reared or born and reared in the United States. American Journal of Diseases in Childhood 1932; 44: 69–80.
90 Wang CC, Kaucher M, Wing M. Metabolism of adolescent girls. Basal metabolism and energy exchange. American Journal of Diseases in Childhood 1936; 51: 801–15.
91 Wang CC, Kern R, Frank M, Hays BB. Metabolism of undernourished children. Basal metabolism. American Journal of Diseases in Childhood 1926; 32: 350–9.
92 Wang CC. Basal metabolism and preformed total creatinine in urine of seventy children. American Journal of Diseases in Childhood 1939; 57: 838–50.
93 Wang CC. Basal metabolism of American-born Chinese girls and of American girls of the same age. American Journal of Diseases in Childhood 1934; 48: 1041.
94 Holland B, Welch AA, Unwin ID, Buss DH, Paul AA, Southgate DAT. McCance and Widdowson’s the Composition of Foods, 5th edn. Cambridge: Royal Society of Chemistry, 1991.
CJK Henry1152
Elia techniques for the study of energy balance(1).pdf
��������� ���� ����� �� ��������� ���������������� ��� � �������� �����������
������������������
������� ��� �������� !��"��#$��%&�"'(����� ��������������� � ��� ���)������* �("����!( ��) %�+,,����-�� �,������. (!�*���!( /����"� 0��1�
2�3��"���" �(�" !������0��'("#���)�����(�(�"���0(��$���������� 455��6�����(�(�"���0(��$������5����������"��#$�7 ! "0��("��� !��� "'�'(�� ��*���!( �� ��� ��� ��� 6�����(�(�"���0(��$�����
��������������� ���� ��������������� � �������� �
* �("����!( �8��9�7�00 ���� ���"�� "'�: .������77�� ��� � � ������������� �� ����������� � �������� ���� ������� ���� ������������� �� ���
�������� �������� ���� ����!�"�"#$���% &'() !�� �*� � � �������� � � ���+��, -�����'-�������'+&��.�+���%
����)������* �("����!( ��) %�+,,����-�� �,������. (!�*���!( /����"� 0��1
�"��#$�7 ! "0��0 "�7�����(. ��'�("��(�������7�'$���#.�"����("'(;('� !���7<�0��������)�0����)����
&����"�� ��(0!����#���&���)���7<�0��� "'��;�"���0(��(����2� "#���("�7�'$�0�.&��(�(�"�("�("'(;('� !
��7<�0���0 "�7���� "�! ��'�("���0� "#���("������"��#$�0�"��"���)�����7�'$��7�����(��.����'�(��!(.(��'
7$�����&��0(�(�"��)�������0�"(=���������&��0(�(�"�)���.� ���("#�) �� "'�) �4)����. ���0 "�7�� ��!�>
�� �?� 1#� >��"� 0��� ("� ��)���"0�� ��0�"(=���� ��� ���'� ��$'��'�"�(��.���$�� (�4'(�&! 0�.�"�
&!���$�.�#� &�$�� '� !4�"��#$� @4� $� 7���&�(�.���$��� "'� &&��%(. ��!$� �?� 1#� )��� 0� "#��
7��>��"��>���(.��&�("������0�"(=���� ���0( ��'�>(��� �.� ����.�"���������)��?� 1#�)���0� "#��
("� ) �� "'� ) �4)����. ��� � &&��%(. ��!$��-*:�� ��� �)� !(��!�����"��; !��� )��� 0 !0�! �("#� �"��#$
7 ! "0���;���������&��(�'���)��(.���7������$�. $�7���)���.��; !����;���!�"#�&��(�'���)��(.����-*:
�;�����$� ��0�����&�"'����� "� ;�� #��' (!$��"��#$�7 ! "0���)��� 1:��>�(0��(��A��B��)�����"��. !
'(�� �$��"��#$�("� 1����3�'$�0�.&��(�(�"�.� ����.�"���0 "� !���7�����)�!�("�0 !0�! �("#�0� "#��
("��"��#$�7 ! "0��>��"�����0� "#��� ("�7�'$�>�(#��� "'�0�.&��(�(�"� ���! �#�����#��C� �� 1#�
��� � .��&�("0(&!���0 "�7�� &&!(�'�������� �����.�"���)��"��#$�7 ! "0��("�7�'$���#.�"�����("#
'� !4�"��#$�@4� $� 7���&�(�.���$���"��#$�7 ! "0��0 "�7���7� ("�'��;���&��(�'�� �������� �� �)�>
.("������ ��#�� '��("#�.� ����.�"��� �)� 3*9�� ���� ; �( 7(!(�$� ("� 3*9� 7��>��"� ("'(;('� !�� �)
�(.(! �� #���>�(#��� "'���(#��� "'�#�"'���(�� 7����� �B��.�����)�>�(0��(���)�7(�!�#(0 !���(#("
� ����� �� "�.� ����.�"���������>�(0�� (�� 7���� ��B��*� ����.�"���)� ��� !� �"��#$��%&�"'(����
'��("#��� �; �(�"��"���"��#$�("� 1���0 "� !���7�����'�������(. ����"��#$�7 ! "0��("� �>��!�47�'$
0 !��(.������("�& �(�"���("�("��"�(;��0 ����"(���7�("#� ��()(0( !!$�;�"�(! ��'� "'�7$��� 0�����0�"(=����
���� &��0(�(�"� �)� ������ ��0�"(=���� ; �(��� )��.� �� ��� ��� B�� ��� 7!(��("#� �"��#$� 7 ! "0�� 7$
.� ���("#�����'(�0��& "0$�7��>��"��"��#$�("� 1�� "'��%&�"'(������ ������ 1��("���0�"�('�� �(�"
���� 0�.7("�'� ; !('(�$� "'� ��!( 7(!(�$� �)� 7���� 0�.&�"�"���� ����.� ����.�"�� ������ )��� '(�� �$
("� 1�� . $� 7�� �� !�>� �� � ��B� ("� 0 ��)�!!$� 0�"���!!�'� �";(��".�"���� ("� >�(0�� ��7<�0��� ��
&��;('�'� �"!$�>(��� 0��� ("� )��'� (��.�� "'� 7�.7� 0 !��(.���$� 0 "� 7�� �"'��� 1�"� �"� '�&!(0 ��
� .&!����)�����'(����9�!( 7!������!���0 "� !���7���7� ("�'�("����&(� !(��'�& �(�"�����0�(;("#��"��� !
��7��)��'("#����& ��"��� !�"���(�(�"� �������"!$�����0���)�"���(�(�"��D"��!( 7(!(�$�("0�� ������� "
�"1"�>"� �%��"�� ("� )���4!(;("#� ��7<�0��� � �("#� � .(%�'� "'� ; �(�'� '(��E� ������ (.&��;�'
.����'�!�#$�(��"��'�'�)����������'$��)��"��#$�7 ! "0��
��������� �� ��������� ����� ��� �!� ������� � ���������
�������� !��"��#$��%&�"'(����
���� 0�"0�&�� �)� �"��#$� 7 ! "0�� (�� �(.&!�� ��"��#$ 7 ! "0� F �"��#$�("� 1��−��"��#$��%&�"'(�������7��� 00�� �� "'���!( 7!��.� ����.�"���("�)���4!(;("#�0�"'(�(�"��. $�7� ) �� )��.�� �$� "'� ��7<�0�� ��� &��7!�.��>(��� ("���&��� �(�"� ������ ��.�� <���()(0 �(�"� )��� �"'��� 1("#� �"��#$� 7 ! "0� ���'(���(����=�(��'����&�0( !!$�()�����.� ����.�"��� ���!(1�!$ ��� 7�� �(.�40�"��.("#� "'� ! 7���� ("��"�(;��� "�� �)� ��� &��&����� �)� �"'��� 1("#� �"��#$� 7 ! "0�� ���'(��� (�� ��� ��!& ��� 7!(���0�"0�&���0�"0��"("#��"��#$���=�(��.�"���("�7���
�� !��� "'�'(�� �����"�0!("(0 !�&� 0�(0��& �(�"���. $�� ��;��$ !(��!�� �� � ����!�� �)� &���� &&��(��� ��� '())(0�!�(��� ("� � �("#� �> !!�>("#����'(#���("#�)��'����0(�(�"�� 7�������� '�=� 0$ �)�'(�� �$�("� 1��. $�'�&�"'��"�(��!$��"������� !���&��)��4 �(�" !�� ������� "�����& �(�"��� "'������)������.��1"�>!�'#� (�� ��=�(��'� 7���� �"��#$� ��=�(��.�"��� "'� ��>� ������ �� ��� 7!(���'��G�>�;����(������!'�7���������'��� �� �����"��#$ ("� 1�� "�0��� �$� ��� . ("� ("� �"��#$� 7 ! "0�� (�� "�� "�0��� �(!$�����.���� &&��&�( ���("� 1����"�'�&!���'�& �(�"��
��� *���!( �� ���/
� &��(�(;�� �"��#$� 7 ! "0�� (�� �)��"� ��=�(��'�� "��� �"!$� �� ��&!���� ���� ��'�0�'� �"��#$� �������� 7��� !��� ��� ��&!���� !� " �(������ � &��(�(;�� �"��#$� 7 ! "0�� ) ;����� � &��(�(;�� � 7 ! "0�����"�0�"�� ����("��7����("'(;('� !�� �"�# �(;���"��#$ 7 ! "0��(��'��(� 7!�������'�0�������(�1��)�'�;�!�&("#�0 �'(�4 ; �0�! �� '(�� ���� '( 7����� "'� �$&����"�(�"�� �;�"� (" ("'(;('� !�� >(��� � '��(� 7!�� 3*�� "'� #����� 7�'$� 0�.&�4 �(�(�"�� >��� ��� ("� �"��#$� 7 ! "0��� ���� ("� 1�� ��=�(��'� �� . ("� ("� 7 ! "0�� '���� "��� "�0��� �(!$� (.&!$� �� �� (�� (� �&�(. !�� �)� ���� ("'(;('� !�� ��� ;��$� ��'�"� �$� ���$� >��!' 7�"�)(��)��.�("0�� ��'�&�$�(0 !� 0�(;(�$��>�(0��0 "�(.&��;� 0 �'(�4���&(� ���$� )�"0�(�"��>�!!47�("#� "'� ��'�0�� ���� �(�1 �)� '�;�!�&("#� �G�� �H(�� � � ���� ������� ��(�� ("0�� ��� (" &�$�(0 !� 0�(;(�$� >(!!� �)� 0������ !��� ("0�� ��� �"��#$ ��=�(��.�"���
�"������ �� ��"� )��� �"'��� 1("#� �"��#$� 7 ! "0�� (�� �� �7� ("� (")��. �(�"� 7���� .�0� "(�.�� �)� '�&!��(�"� "' ��&!��(�"��(")��. �(�"��"������$&���)��(�������� ��. $�7��!��� ���# ("�'���!( ��� �-��� ��� "'�("�(#����("���&�$�(�!�#(0 ! ' &� �(�"�����0� "#���("��"��#$�("� 1�� "'�����%&�"'(����� I����% .&!��� �'��(� 7!����#(.�"�)���&��'�0("#�>�(#���!��� ("������7�����"�# �(;���"��#$�7 ! "0���(���"���� ��(.&��;�� 7�'$�0�.&��(�(�"� "'�)�"0�(�"���%0���(;��!�����)�!� "�7�'$ . ���(��"���"��. !!$�'��(� 7!�����&�0( !!$�()�(��(�� ���0( ��' >(���'����(�� �(�"��)��(�����)�"0�(�"�������$�J�: .����� ,E �!( ����������"�0�"�� ����("��"'��"���(���'�& �(�"�������"���(4 �(�" !� ��� �.�"�� (.�� ��� 00����� 7���� !� "� "'� ) �� �(������ ��&�0( !!$� ()� ���$� !� '� ��� (.&��;�.�"��� ("� 7�'$� )�"0�(�"� 9�#(.�"���� ������!���"!$�("�# ("��)�) ��. $�&��'�0��!(��!���� "��)�"0�(�" !�7�"�)(����("0���"��#$�7 ! "0��0 "�7�����(. ��' 7$�.� ���("#�0� "#���("�7�'$�0�.&��(�(�"������(")��. �(�" 0 "� 7�� ���'� ��� ������ >������� ���� 0�.&��(�(�"� �)� >�(#�� 0� "#��'�&�"'���"������$&���)���� �.�"�� "'�����. #"(��'� �)������"��#$�(.7 ! "0��� �����7 ! "0�����'(���0 "�&��;('� (")��. �(�"� 7���� ���� �%��"�� ��� >�(0�� �"��#$� ("� 1� 0�.&�"� ����)���0� "#���("�&�$�(0 !� 0�(;(�$� "'�;(0��;��� � >�(0�� (�� (.&��� "�� 7���� )��.� � &�$�(�!�#(0 !� "'� 0!("(0 ! &���&�0�(;��� ��.�� �% .&!��� �)� ��>� 0�"0�&��� �)� �"��#$ ��.���� �(��0 "�7���7� ("�'�)��.�'())���"���$&����)��"��#$ 7 ! "0�����'(��� ���(!!���� ��'�7�!�>���������"��#$�7 ! "0� ���'(���0 "�7��0! ��()(�'� 00��'("#���������$&���)�.� ����4 .�"���"��'�'����"���"��'�'�
.� ����.�"��� �)� �"��#$� �%&�"'(����� �"!$�� �� �; �(�" ���'(����"���"��#$�("� 1��E .� ����.�"����)�0� "#���("�7�'$�0�.&��(�(�"��"!$��"� .� ����.�"����)��"��#$�("� 1������%&�"'(����E .� ����.�"����)��"��#$�("� 1�� "'��"��#$��%&�"'(����E .� ����.�"��� �)� �"��#$� ("� 1����� �"��#$� �%&�"'(���� ����7����� "'�0� "#���("�7�'$�0�.&��(�(�"�
� 0���$&���)����'$�� �� '; "� #��� "'�'(� '; "� #����>�(0� ��! ����������=� !(�$� "'�1("'��)�(")��. �(�"�(��&��;('���
�������� � ����������� ������� ������������ � ����� ��������
���� #�"�� !� �"��#$� 7 ! "0�� �=� �(�"� ��"��#$ 7 ! "0� F �"��#$� ("� 1�− �"��#$� �%&�"'(������ 0 "� 7� �(.&!()(�'�>��"� ������ (�� "�� '(�� �$� �"��#$� ("� 1�� ��"��#$ 7 ! "0� F − �"��#$� �%&�"'(������� �("0�� .� ����.�"��� �) 3*9� ��� �"'��� 1�"� )���� "� �;��"(#��� ) ���� ���$� ��
����"�( !!$� .� ����.�"��� �)� "�# �(;�� �"��#$� 7 ! "0�� ����&� "'( !� .� ����.�"��� �)� �"��#$� �%&�"'(����� !�� ��&����"�� .� ����.�"��� �)� "�# �(;�� �"��#$� 7 ! "0�� G�>�;���� !����#���������.�K�"��#$�7 ! "0�L�0 "�7�� &&!(�' ���&��(�'��� "#("#�)��.�A�.("���� �!()��(.���(��(��"������ !!$ &&!(�'�����������%���.��� "#����)��(.����"��"��#$�7 ! "0� ���'$��)�("���.�'( ���'�� �(�"�(������0! ��(0��� '��� �; �(�" ��( !��) 3�"�'(0���� �����("�>�(0�� ���7<�0��> �����'(�'�("� >��!�47�'$� 0 !��(.������ D��� �)� ���� ��(#(" !� ' � � !!�>� 0 !0�! �(�"��)�����0�.�! �(;��"�# �(;���"��#$�7 ! "0������ & ����"� �)� 0�.�! �(;�� �"��#$� !���� )��.� ���� 7�'$� �;�� ��� �� '�(�����>"�("�I(#���� � "'�7���>�(0�� !������>����� 0�"��(7��(�"� �)� 3*9� �I(#�� �� ��� "'� �)� &����("� �%(' �(�" �I(#�� �� �7��� ��� ���� 0�.�! �(;�� �"��#$� !����� ��("#� ��&����' ' (!$� ; !���� �)� 3*9� "'� ��(" �$��� ; !����� ���.("#� � # ��(" �$� �� ��&����"��� ��5 1:� '��(;�'� )��.� �%(' �(�"� �) &����("� ��!( � J� H(;���$�� � ���� ���� 0�.�! �(;�� �"��#$ '��(;�'� )��.� &����("� �%(' �(�"� ���?�*:�� �;��� ���� �"�(�� �� '� 00��"��'�)�����? �B��)�������� !��"��#$�!����)��.���� 7�'$���-�*:E�� ?-�B�>��"��%&�����'���! �(;���������3*9� >�(0�� > ��5�?�*:��� ��� !� �"��#$� �%&�"'(����� ������
0 20
40
60 80
100
120
140
160 180
200
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31
E n
er g
y ex
p en
d it
u e
(M J)
(a)
0 20
40
60 80
100
120
140
160 180
200
E n
er g
y ex
p en
d it
u e
(M J)
(b)
Period of starvation (d)
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31
Period of starvation (d)
���� �� ������������ � �� ��� ��� �� ������� ��������� �� ���
���������� �������� ������ ������� ����������� ������ �����������
������ �� � � ����������� �u4u !���� ������ �����������" �w4w ������������� ��� ��# ��� ��� ������� ���������� �������������
����� �� �������� �$�%�
�"��#$�7 ! "0��("��� !��� "'�'(�� �� ���
3*9� "'���(" �$����%0���(�"���"'�'����'�0�� ���>(����(.�� 7���>��"�����������!���>�����%&�����'���! �(;������ 0������� ������ > �� !(��!�� 0� "#��� � 7!�� �� (!!���� ���� ������ ' � � 7$ 0�.& �("#� ���� ����!����7� ("�'�'��("#� ���� )(���� "'���0�"' � !;����)�������.(4�� �; �(�"�&��(�'������&��&���(�"��)���� �"��#$��%&�"'(�����'��(;�'�)��.�&����("��%(' �(�"�(���(.(! � ��� �� �� �7���;�'� ("� � "�.7��� �)� ������ ���'(��� �)� !� " ��7<�0���� 7��� 0�"�('�� 7!$� �(#���� �� "� �� �� �7� ("�'� )��. �7������7<�0����"'��#�("#��� �; �(�"��"���"��#$�("� 1���)�� C��>��1�����(����'�0�'�!�;�!�("������7����0 "�7����# �'�' � "� ' &� �(�"� �� �� !!�>�� &��!�"#�'� ���;(; !�� �)� '��("# �� �; �(�"��7������7<�0���>��������7� ("������ .��&��&���(�" �)��"��#$�)��.�&����("� ��!� "���7<�0�������$�>��!'�� &('!$ '�&!�������(��!� "��(������ "'�'(��>(�����7�� "�( !� .��"����) ) ����. ("("#���� � ��
D"'��� 1("#� �"��#$� 7 ! "0�� ���'(��� ("� >��!�47�'$ 0 !��(.������ !!�>�� #��'� 00�� 0$� "'� &��0(�(�"� ��� 7� ��� 7!(���'������&&����, ������7������$�'��"�����)!�0��)���4 !(;("#� 0�"'(�(�"��� ���� ;��$� !�>� &�$�(0 !� 0�(;(�$� !�;�! �����3*9���>�(0��> ���"!$��?���("������� �; �(�"����'$��) 3�"�'(0�� �� 7!�� ���� &��7 7!$� ��)!�0��� ���� '(��0�� �))�0��� �) �� �; �(�"� ��>�!!� ������ ��()(0( !��%&��(.�"� !�0�"'(�(�"���) ���� ���'$�� �"� �"��#$� 7 ! "0�� ��0�"(=��� �� �� �;��0�.�� ��(� &��7!�.� (";�!;��� .� ����.�"�� �)� 0� "#��� ("� 7�'$ 0�.&��(�(�"� 7��>��"� �>�� �(.�� &�("���� '��("#� >�(0�� ��� ��7<�0����"'��� 1�����(��"��. !�' (!$� 0�(;(�(���
�������� � ����������� ������� ������� ������������� ������� ���
2� "#��� ("� ) �� "'� ) �4)���� . ���� .� ����'� ��("#� 7�'$ 0�.&��(�(�"� ��0�"(=����� 0 "� 7�� ���'� ��� ��� 7!(��� �"��#$ 7 ! "0�� 7$� ��(#"("#� �"��#$� ; !���� ��� ���� ) �� "'� ) �4)��� 7�'$� �� �� (�� !���� ��� # ("�'�� ���� �"��#$� ; !��� �)� 7�'$� ) � (� >�!!���� 7!(���'�� &&��%(. ��!$�� ?,*:�1#�� ���.("#� .�� 7�!(� 7(!(�$��)��?�E��!( �J�H(;���$��� ����7���������(� .���� �"0��� ("�$� 7���� ���� �"��#$� '�"�(�$� �)� ���� ) �4)��� �(�������� �� ���!�������# ("�'��G�>�;����()�(��(�� ���.�'��� � &����("���-?�*:�1#�� !���)��� �.�� 7�!(� 7(!(�$��)��?�E��!( J�H(;���$��� ��� 00��"���)��� 7�������B��)�����0� "#��(" ���� ) �4)���� 7�'$�� ���"� �����"��#$�'�"�(�$� �)� ) �4)����. �� �� ��(��# ("�'����!������?�*:�1#��(���"!$� 7�������B��)��� � �)� ���� ) ��. ���� ������ ("� #�"�� !��.� ����.�"�� �)� �"��#$ 7 ! "0��7$�7�'$�0�.&��(�(�"���0�"(=����'�&�"'��&��'�.(4 " "�!$��"����� 00�� 0$� "'�&��0(�(�"��)�.� ����.�"���)�) � . ��� ����� & ��(�(�"� �)� 7�'$� >�(#��� ("��� ) �� "'� ) �4)���
. ������ ������� "��"������"��#$�'�"�(�$��)�����) �4)����. �� �)���.����'�� (!�'�'(�0���(�"������&&����, ������2� "#���(" #!$0�#�"� ������� !��� . 1�� � 0�"��(7��(�"� ��� ���� �"��#$ 7 ! "0���7������(���))�0��� ����)��"��. !!����&�0( !!$�("�!�"#4 ���.� ���'(��� (";�!;("#� . <��� 0� "#��� ("� 7�'$� >�(#��� M!$0�#�"��>�(0�� (��& ����)� ���� ) �4)����. ����� �� ��(.(! � �"��#$� '�"�(�$� ��� &����("� "'� (�� !���� )��.� ���� 7�'$ (" &&��%(. ��!$� �=�(; !�"�� .��"��� �)� > ���� ��� �� � ���0( ��'�>(���!�����)�&����("�
����0! ��(0���.(4�� �; �(�"����'$�� ����*(""���� ����'$� �)�N�$��� ������� ����(�����'����(!!���� �������; !����)�7�'$ 0�.&��(�(�"� ��0�"(=���� ("� ��� 7!(��("#� �"��#$� 7 ! "0��� �� > �� 0����"� & ��!$� 7�0 ���� ���� 0� "#��� ("� ) �� "'� ) �4)��� . ���>����! �#��� "'�& ��!$�7�0 ����; �(����������.� ����4 .�"���>�����7� ("�'�'��("#�����0�������)��������'$��>�(0� 0 "� 7�� 0�.7("�'� ��� &��;('�� ("�(#���� ("��� 0�"0�&��� 7��� �"��#$� ��.���� �(�� ������� ������ .� ����.�"��� >(!!� 7� &����"��'� ("� ���� ��7��=��"�� ��0�(�"�� ���� &&�� ��� ��,��� �" ��(�����'$����(��$4�>����7<�0���>������.(4�� �;�'�)��� �&��(�' �)� �,�>��1�� �"'��� 0 ��)�!!$� 0�"���!!�'� 0�"'(�(�"��� '��("# >�(0�����$�!���� �.� "��)��,�B��)����(��("(�( !�7�'$�>�(#��� ����0� "#���("�7�'$�>�(#��� "'�0�.&��(�(�"��� ���00����' '��("#� ���� )(���� "'� ��0�"'�� !;����)� ���� �� �; �(�"�&��(�' ������>"� ("�� 7!�����*�����)� ����>�(#��� !����> ��'��� �� !�����)�) �4)����. ����5��B��;�� !!���7���.�����)�����"�# �(;� �"��#$�7 ! "0��> ��'������!�����)�) ��. ����--�B��)�������� !� ���.("#��� �������"��#$��=�(; !�"����)�) �� "'�) �4)����. �� !���� ��� � ?,� "'� �?�*:�1#� ���&�0�(;�!$�� �� ��##����' � �!(����� ������ �$&��� �)� 0 !0�! �(�"�� � ;�� 7��"� &&!(�'� �� '())���"��#���&���)�!� "� "'��7����("'(;('� !��!��("#�>�(#�� �� �����!���)���'�0�'�)��'�("� 1���� 7!�������>����������!�� �)� ����*(""���� ����'$�� ��#������>(��� �������)� ����������� ���'(����I�!!���� ������ �E�; "�O(<1��� E�M(7"�$�������� ������' � � ���("�#�"�� !� #���.�"��>(�������!����7� ("�'�(" � � "#�� �)� ������ ���'(��� "'� '�.�"��� ��� �� �� ���� 0�"��(4 7��(�"��)�!� "��(������������ !�>�(#���!����(��#�� ����("�!� " ��7<�0����� "�("��7������7<�0������������!��� !���(.&!$��� � ���� 0�"��(7��(�"� �)� ) �4)���� . ��� �. ("!$� &����("�� ��� ��� ! �"��#$� !���� )��.� ���� 7�'$� (�� !��� #�� ���� ("� !� "� ��7<�0�� �� "�("��7������7<�0���
������ �����;�� !�!(.(� �(�"��("� �����("#��"��#$�7 ! "0� )��.�0� "#���("�7�'$�0�.&��(�(�"�� "���)�������!(.(� �(�"� 0�"0��"�� ����&��0(�(�"��)� ����.����'��>�(0�� (�� 0�"�('���' ! ���� �&&�� ��, ������ �"������ !(.(� �(�"� 0�"0��"�� ��� 00�� 0$� �)� 7�'$� 0�.&��(�(�"� ��0�"(=���E� )��� �% .&!�� > ���4'(!��(�"���0�"(=���� ���.���� �������$'� �(�"�)� 0�(�" �)� ���� ) �4)���� . ��� (�� 0�"�� "�� "'� �$'��'�"�(��.���$
��� �� &������� ������ ������� ���������� ��� ������� �� ����
����������� ������ ��� ���� ��� ������ ������ � � ������ �
'( )��*� � ����+���������� ��� ��������� ������� ��� �� ������
���� ��� ��� ����,
-����� � �����
�)��*�
.��
��/
-������
��/
.�� 0 �������
��/
12�'
�'2'(
12'( ������
'1345
514�
'534$
'%4'
��46
�341
'�'46
3'4�
�1(4$
,.�� ������� ��� ����������� ��� �� %��" ���� ���������� ��� 7��� ������ ��$%1 �
��� �� ���������� � ���������� �������� ������ ������� ������
������ �����������" !88 ��� ������� ������� ������ ��� ���� ���
������ ������ � � ������ � �� � � ����� ���������� ��� ������
������ �� ���� ��� �������� �$�%
�2�% � �52�� � �2�� �
!88 ��/
��� ��/
������������ � ��� �� !88 �9
-������� �������� ����� �!88:���
-������ ���������; �/
9 !88
9 ���
$%4%
6%4%
6$4%
�4�'
�34(
�64'
'14(
6(4%
3341
$�4�
�4�1
�(46
�34%
�$4'
�6141
�5'4%
$14�
�4��
�'4'
�34$
�$46
��� ����� �������
��� ��� �� �� ���� ���� ��� ��� ���� � ������� ��� �� ����� ��� �� ���������� ��� ������������������ ���� ����� ����� ����� � � � ����� ����� �� ���� � ��� ������ � ���� ���� ����� �� ������� ��������������� ��� � ��������� ��� ����� ������������ ���������� ���� � ���� ����������������� ��� ��� �������� ���� �������� ��� �� �� ��!""��#�$������������ !""�%��&��� �������� � � ��� ���� ��������� ��� � �������� � ���������������� ��� ���� � ��� ��������� ��� � � ���� �� �������'�� � � ������������$����( �������������� ��� ����� ���� �� � ��� �������� � ��' � ��� ����'�� � � ���� � �� ����������'���� �'��� ������'�� �� )������(���� ����� ��������� �� �� � ��������(���������� ������������������ ��� �������� �'�������� ���� ������������������� �� ��������*� ������ ��� �� �� ��� ������ � ������ �� �� ��� ���� ������ ��� �� )�� � ����� � ���� ����� ��� ����'�� �� )�� ��� �(����� ������ +�������� �� �� ���� ���� ��� � ����� ���� ��� ����� �� �������� ������ ������������������ ������ ���� �'��� � ������� ��� � ��� ���� ����� ��� ������ � ������ ������� ���,��-%�
�������� ������ ���� ���������� ������������ ��� ��� ������������������
������ �� �� ��� � ��� ������������ ��������'������ ����� �' ����'�� �� )�� ����(���� ������*�� ����� ���� ������'��� � ��� ���� ���� ������� � ���� ������������'�� �� )�� ����(����� ����������������� ����������� �� � �� ���������� � �� �������*� ������������� ���������� ������� ����� ������ ������ ��.
!� ����'�� �(���� ����� � �� ��� � ������ ���� ���� ���� ��� �������� ���� �� ������������ � ��� ����� �� �� � ������ � �'� ���� � ���� �� �'� �� ���� ������� ��� ����� ���� �� ����� �������� ����� �'� �� ���� ��� �� ����� ������� ������ ��� �� � ��� � � ����/ � ����� ������ � �������� ������ ������'�� �� )�� � ���� ������ �� �� ���� �� � �� ��� �� � � � ���� ������ ����� �� )�� � �� �� �� ������ ��������������� �� �������� � �'����� � ���� �������� � ��� ��� ���� �������� ���������������� ���� ���� �������� �� ������� �� �����-,���%#
�� ���� �������� ������ � �� ��� �� )�� ��������'�� � �� � ������ ������ �� ���� � ���� ���� ��� ������ �� �� � �� �� 0 �'� �� !""!#� /� ������ ��� ���� !""1%�� ������� � ���� �� ��� �� � ��� �� ���� �������� � ���� ������������'� �(���� ������ �����'�� ������ � �� ����� � ����� ��� � ����� ' � �������� ������ � � ��#
�� � ���� ����� ��� ����'�� � � ���� �� ���� ���������� ���� � ����'���� ����������� �������� ������ �� �������� � ������ � �'� ���� � ����� ������ �� ���� � '���� �� � �) �� �� � �� � �� ����� �'����������� ������'��� � ���� �
������ � �'����� � ������������������� ���� �������� � ��� '������� ���� �� ������'� ������ �����������2������������ ������������ ���� � ������ �����'� � ��������� ������� ���� �������� ���������������� �� ��� �� )�� ��� �� � � � ��� ��� � ��� � � �'�� ������ ������ �� � ����� ��� ��� �� �� ������� ��������'�� �������� � �'����� � ���#
-� ������ ��� ���� � �������� �� ���� ����� ��� ���������� � �� ��� ��� �� ��� �� )�� ����(���� ����� ��������$�� �( ����������� �������� ���� �������� �� ���� ����� ���������� ����������� �� )��������� '��������� �� )� ������ �������� ��� ���� �� �� ���� ��� ���� �� � �� � �� �������������� �'��� �� ��� ������� ���������� �' �' ��� � ��� � �'�� ��������+�������������� ����������� � � �������������� ��� �� ����������� �� ���������� �' ���� ���� � ���� �� ���� ��� ���� ����� ���� � ����� ��� ��� �� )�� ������� 0 �'� �� !"31%�� / � ���� ������� ���� � ��� �'� ����'�� �(���� ����� � �� ���� ���� � �������� ���$����( ����� ������� ��������� �� �� � ��� � ����� ��� ���� � ����� �� �� ����'� �(���� ������ ����� � ���� ��� ���� ���� ����� ���� �� ��� ����'���(�������������2����������� ) �'�� ��� ��� � � ���� ���� �� �'�������� ��������� '����� � �'� �� �� � ��� ��'�� � �) �'%� ��� ������ '����� � �'� �� � � �� ��'��� �� �'���2�������� �'��� ������ �����)%��������� ����� ���� ��� �� ��� �'���� � ���������'���(���� ���� �� ���� ������ � �� ��� ���� ��� ����� ��� � � � � ��� �� �� ���� ��������������2������� ���������� ) �'� ��� ��� �� ���� �� � ����������� � �� � ������ �� � �� �����2���� ����� ����� �� ��������� �'���� � ����(��� ������ ��� � ��� '����� � �'� �� � ��� �� �� �������� ���� � ������ ��� ����'�������������� � �� �� � ��� ������ '���� ������� �� '����� � �'� �� � � ��� ������� ��� ���� ����� ��� ���) ����� ����� �� � � ��������� )���������2����%��4 � �� � � ������ �'��� ����(��� ��� �� ���������� � ������������ 2����� � ��� � �� �� � �� ���� ����� ���� ������ � �� ������ ���$������ ������� ��������� ��������� � ���� �'� ����'�� �(���� ����� � ��� ���� ����� ���� ���� � � � ������� �'�� � ���� '�����������' ���� ���� ��� ����� ����� �� �� � ����� � �� ��� �� )�� ��� ����'� �(���� ���������������� �� ���� ��� ������� � �������� � ���� � ���� �������� ������� � ��������� �� � ����� � � �'�� �� ����� ��� ���2����� �� ������ � �'� ���� � ���� +�������������� ����� ��� ��� �� ����� ���� �������� ������ ��� �������� ������ ����������������� ��������� ��� ����� ��� ��� ��� ��� �� ������$����( ������- � �� ����� � ��� ����� � �� ��� � �� ��� ���� �� �� � � ���� ���� � ����� � ������� �'���� ������������ '��������� �� ��� �������� � ���� ��� 5��� �� ���� ���� ��� ���� ��� �� ������� ���)������� ��� � ��� ��� � � �������
����� �� ������������ � ��� � ���� ����� �� � ���� ���� ��� � ��� ���� �� � �� ��� �� � ���� ��� ������ � � ���� � � ��������������
������� ��
�!�"�#�
$����� � ��
���� �!��
% �� ���� �� ����
��� ���
% �� ���� � ��� ����
��� ��� ����� ��� & � ��
' ��
(� �
#)*+
#)*,
--*.
/ ,.
)0*1
,*0
).*)
,)*2
0.
31
)2
#.
)#
))
#
#
4 �� �����5 �)23.�
6��� � �#..#�
��� 7��! �)222�
���� � �����5 �)223�
�$�� � ����� � � ������ � ��� 8� ��� ���� ����� � � � �������������9 8� �� �� ����� � ���� � �����5 �)223� �� �� � ���� ���9 ����� ��� ��� � �� ����� � �� ���� ����������� �� ����������� � ����������� ��� ������ ������������5
Energy balance in health and disease 533
It is often not practical to obtain long-term measurements of both energy intake and expenditure in free-living or even in laboratory conditions. For example, no formal direct measurements of TEE were made before, during and after 6 months of semi-starvation in the Minnesota Study (Keys et al. 1950). However, it is possible to calculate energy expenditure and physical activity by combining energy balance measurements with those of energy intake. Conversely, energy intake can be calculated from measure- ments of body composition and energy expenditure. It is also possible to subdivide energy expenditure into its components to obtain insights into the factors responsible for energy imbalance and energy homeostasis.
Combining measurements of body composition with energy intake or expenditure
The total negative energy balance during 24 weeks of semi- starvation in the Minnesota Study of Keys et al. (1950), which was estimated through changes in body composition (Table 3), can be converted to the mean daily balance by dividing the total negative energy balance by 168 (24 × 7 d). Similarly, the values for the two 12-week periods can be obtained by dividing the changes in balance obtained during this period by 84 (12 × 7 d). These values are shown in Table 4. The energy intake of the individuals was carefully recorded during the entire 6 months of semi-starvation (also shown in Table 4). TEE during these periods, calculated as the sum of the negative energy balance and the energy intake values, averaged 8·38 MJ/d, and was greater during the first 12 weeks (9·55 MJ/d) than during the second 12 weeks (7·21MJ/d).
Table 4 also shows the changes in physical activity energy expenditure, BMR and dietary-induced thermo- genesis. The BMR measurements were made on three occasions, once before starvation and at 12 and 24 weeks of semi-starvation. The BMR value assigned to the first half of the study was the average of the first two measurements, the value for the second half of the study was the average of the last two measurements, and the overall value for the whole study was the average of all three measurements. The calcu- lated BMR was based on measurements of O2 consumption, which was converted to energy expenditure using an energy equivalent of 4·8 kcal/l O2 (19·94 kJ/l O2). Dietary-induced thermogenesis can be assumed to be about 10 % of energy intake, and physical activity energy expenditure can then be calculated as the difference. If no assumptions are made
about dietary-induced thermogenesis, physical activity energy expenditure plus dietary-induced thermogenesis can be estimated as a single entity. Table 4 shows that the value during the first 12 weeks (3·26 MJ/d) was greater than that during the second 12 weeks (2·25 MJ/d; overall value for the 24-week period 2·61 MJ/d). Whilst this information is inter- esting, it does not indicate the pattern of changes that occur during the first or second halves of the study. Furthermore, they do not allow comparison with baseline (pre-starvation) measurements of TEE and physical activity, which makes it difficult to assess the extent of adaptation changes that may have been induced by dietary restriction. Since only one body composition measurement was made before starvation, it is obviously not possible to calculate changes in energy stores using body composition techniques. However, it is possible to estimate energy expenditure by assuming stability in body composition. The group was studied for several weeks immediately before starvation was started, and their weight was found to be stable whilst ingesting a diet of known energy content. For example, during the 3 weeks before semi-starvation the subjects lost only about 0·10 kg/week, and so it could be assumed that the group was close to energy balance. Similarly, during the last 3 weeks of semi-starvation (weeks 22–24) it could be assumed that the group of subjects was close to energy balance because mean body weight decreased by only 0·11 kg/week. If allowances are made for this weight loss, using the energy densities of fat and fat-free tissues and the proportions in which they were lost in the study as a whole, the effect on TEE and energy balance would be <0·30 MJ/d, both before and during the last 3 weeks of semi-starvation. Since the dietary intake and BMR were measured before starvation was started and at 24 weeks, it is possible to calculate TEE (assuming energy balance for simplicity) and subdivide it into its components, as before. The results (Table 5 and 6) show that the reduction in TEE between the beginning and end of the weight-loss period was predominantly due to a reduction in physical activity, from 6·46 kJ/d to 1·91 kJ/d, rather than BMR, which decreased from 6·69 MJ/d to 4·09 MJ/d. Thus, most of the energy conservation was achieved by behavioural change rather than a change in basal metabolism. Since the subjects followed a set schedule of activities, which included housekeeping and other labo- ratory activities, and walking 20 miles each week, it is reasonable to conclude that there was a major reduction in
Table 4. Components of negative energy balance (MJ/d) during the first and second halves of a period of 24 weeks of semi-starvation*
Weeks 0–12
Weeks 12–24
Weeks 0–24
Dietary intake Negative energy balance Total energy expenditure
Dietary-induced thermogenesis BMR Physical activity energy expenditure
6·78 2·77 9·55 0·68 5·61 3·26
6·35 0·86 7·21 0·64 4·32 2·25
6·57 1·81 8·38 0·66 5·11 2·61
*For details and assumptions, see p. 533; calculated from Keys et al. (1950).
Table 5. Components of energy balance and energy expendi- ture (MJ/d) during the 3 weeks immediately before semi- starvation (−3–0 weeks) and during the last 3 weeks of semi-
starvation (weeks 22–24)*
Weeks −3–0
Weeks 22–24 Change
Dietary intake Total energy expenditure
Dietary-induced thermogenesis BMR Physical activity energy expenditure
14·61 14·61 1·46 6·69 6·46
6·66 6·66 0·66 4·09 1·91
7·95 7·95 0·80 2·06 4·55
*Energy balance assumed; for details and other assumptions, see p. 533; calculated from keys et al. (1950).
534 M. Elia et al.
the energy expended in discretionary physical activities. Indeed, the subjects indicated that they felt weak and tired easily, as if they were ‘old men’. The energy cost of walking on a treadmill at 3·5 miles/h at a 10 % gradient was measured before, in the middle and at the end of the semi- starvation period (n 10), and was found to be unchanged at 0·464, 0·460 and 0·464 kJ/kg per min respectively. This finding indicates that the energy cost of this weight-bearing activity was reduced in proportion to body weight, which progressively decreased during the starvation period. If it can be assumed that the energy cost of other activities was also reduced in proportion to body weight, which is a reasonable first approximation for the type of activities undertaken by the subjects (this approximation may not apply when heavy prolonged physical work involves lifting objects), then it can be calculated that 64 % of the reduction in physical activity energy expenditure was the result of a reduction in tasks, and 34 % to a reduced cost of under- taking these tasks (i.e. a direct consequence of reduced body weight).
The decrease in BMR can also be partitioned into that due to the reduction in body cell mass (65 %) and that due to a reduction in BMR on a per kg body cell mass basis (35 %). Overall, 57 % of the reduction in TEE between the basal pre-starvation period and the end of the starvation period was due to a reduction in the energy cost of physical activity (about two-thirds of which was due to a reduction in physical activities undertaken) and 33 % to a reduction in BMR (about two-thirds of which was due to a reduction in body cell mass). The remaining 10 % is attributed to a reduction in dietary-induced thermogenesis.
In these examples, TEE was used to estimate energy intake during a period of weight stability. More recent studies have estimated energy intake from measurements of TEE obtained by tracer techniques during periods of weight stability. Two tracer techniques have been used to estimate TEE in free-living conditions, the doubly-labelled water technique, which typically provides single estimates over about 10–20 d in adults, and the bicarbonate–urea method, which provides estimates of energy expenditure on a daily basis. Both techniques measure net CO2 production, which has to be converted to energy expenditure using an appro- priate energy equivalent for CO2. It has been argued that tracer techniques are more accurate and/or reliable than many dietary intake methods, when the latter are used in isolation. Major discrepancies between dietary intake and expenditure during periods of weight stability have been attributed to underreporting of dietary intake, especially in
certain groups of subjects, such as the obese. TEE:energy intake during periods of weight stability has been used to establish cut-off points for underreporting (Goldberg et al. 1991). However, in order to establish the cut-off points it is necessary to define weight stability, the period of time over which weight stability is measured and the precision and accuracy of the methods used. The issues on precision and accuracy of energy intake and expenditure methods are considered next, because they determine the accuracy and precision of energy balance. The precision of energy balance, assessed by changes in body composition, is also considered.
Precision and accuracy of energy balance techniques
Changes in body composition
The precision of various body composition techniques can vary considerably depending on the type of technique used, the compliance of the subject and the experience of the operator. With some reference body composition tech- niques, such as dual-energy X-ray absorptiometry, air- displacement plethysmography and hydrodensitometry, and possibly water dilution, fat and fat-free mass may be estimated with a precision (1 SD) of about 0·5 kg. Even a four-component model, which is based on a combination of air-displacement plethysmography (density), dual-energy X-ray absorptiometry (bone mineral) and water dilution (water space), can give a precision close to 0·5 kg, which translates to 19·7 MJ. However, when fat mass is over- estimated by 0·5 kg, fat-free mass is underestimated by 0·5 kg, and vice versa. If it is assumed that 0·5 kg fat-free mass has an energy content of 1·9 MJ (see earlier), the precision for the energy content of the body is 17·9 MJ (19·7–1·9 MJ). For assessment of changes in energy stores (energy balance) by two independent body composition measurements the precision is 25·4 MJ (√(17·9)2 + (17·9)2). This precision is equivalent to two to three times the energy intake when the changes in body composition are measured over 1 d, but <1 % of the energy intake over 1 year. Fig. 2 is used to illustrate: (a) how this precision decreases when it is divided by the number of days between the two measure- ments; (b) how the precision can be improved by making two and four measurements per time point. For example, for a 2 d interval the precision improves from 12·7 MJ/d (one measurement per time point) to 8·98 MJ/d (two measure- ments per time point) and 6·35 MJ/d (four measurements per time point). Over a 20 d period the precision remains the same (8·98 MJ/d), but mean daily values are ten times lower
Table 6. Changes in total energy expenditure and its components between the baseline 3-week period before starvation and during the last 3 weeks of semi-starvation*
Change between baseline and end of semi-starvation period
Baseline (MJ/d) MJ/d % baseline % total reduction
Dietary-induced thermogenesis BMR Physical activity energy expenditure Total
1·46 6·69 6·46
14·61
0·80 2·60 4·55 7·95
55 39 70 54
10 33 57
100
*Energy balance during each 3-week period; for details and other assumptions, see p. 533; calculated from keys et al. (1950).
Energy balance in health and disease 535
(1·27, 0·90 and 0·64 MJ/d for one, two and four measure- ments per time point respectively).
The CV for indirect calorimetry is about 2 % (same calorimeter). The analytical precision for measuring CO2 production by the doubly-labelled water technique has been found to range from 3 % to 6 % (Schoeller & Hnilica, 1986; Elia et al. 2000), and for the bicarbonate–urea method from 2 % to 3 %. Other methods for measuring energy expend- iture, such as activity diaries, have a greater imprecision. For energy intake the CV may be as low as 2–3 % when duplicate food samples are collected under controlled conditions (e.g. in a metabolic facility or in studies of whole-body calor- imetry) and analysed by bomb calorimetry. Reliable results can also be obtained in hospitalised patients receiving enteral tube feeding and parenteral nutrition. However, unreliability increases to an unknown extent in free-living subjects eating a mixed and varied diet. The extent is likely to vary depending on the technique used (e.g. dietary recall, weighed-food records), transcription errors and the type of subject. In recent years there has been increasing scepticism about the accuracy of dietary assessment, particularly in free-living subjects. Our recent studies (O’Reilley, 2002; Stubbs et al. 2003) have shown that when subjects are asked to record their intake in a metabolic suite, where independent and accurate methods for assessing dietary intake can be undertaken, they change their feeding behaviour (observation effect). They also misreport their actual intake (reporting effect). Thus, ‘misreporting’ comprises two separate but synchronous processes. There is also evidence that the problems can be considerably greater in free-living conditions, making assessment of energy balance more difficult than in the laboratory.
In view of this uncertainty it is difficult to establish the overall reliability of energy balance from measurements of both energy intake and energy expenditure. Thus, Fig. 3 shows the results obtained at three levels of precision of energy expenditure (CV of 3, 5 and 10 %, corresponding to 0·36, 0·6 and 1·2 MJ/d respectively), over a range of CV of
energy intake of 3–15 %. The precision varies from about 0·5 MJ/d to >2·0 MJ/d.
Estimates of precision do not of course reflect the accuracy of a technique, since there may be a systematic bias between the measurement and the true value. For example, Fig. 4 shows the results of twelve studies (seven in adults (n 50) and five in children (n 34)) when compared with indirect calorimetry, which is regarded as a reference ‘gold’ standard. The values shown represent the bias (systematic difference from indirect calorimetry), and the standard deviation of this difference. For the group as a whole the results obtained by the doubly-labelled water method differed from indirect calorimetry by −1·3 (SD 10·9) % (n 84), with similar results in adults (−0·6 (SD 11·0) %; n 50) and children (−2·2 (SD 10·9) %; n 34). The extent of disagreement between the two techniques was not related to gender or the duration of study. One study found that as the amount of fat increased, the doubly-labelled water method progressively underestimated energy expenditure obtained by indirect calorimetry (r − 0·81; P < 0·002; Ravussin et al. 1991). It should also be noted that the doubly-labelled-water
2 4 6 8 10 12 14 16 18 20 0
2
4
6
8
10
12
14
P re
ci si
o n
in e
n er
g y
b al
an ce
( M
J/ d
)
Interval between two measurements (d)
Fig. 2. Precision in energy balance (MJ/d), estimated from changes in body composition (25·4 MJ), in relation to the duration of the study and number of measurements at each time point: (u4u), one measurement per time point; (n4n), two measurements per time point; (l4l), four measurements per time point.
10
5
3
1514131211109876543 0
0.5
1.0
1.5
2.0
2.5
P re
ci si
o n
in e
n er
g y
b al
an ce
( M
J/ d
)
CV in energy intake (%)
Fig. 3. Precision in estimated energy balance based on independ- ent measurements of energy expenditure (at three levels of precision of 0·36 (uRRRu), 0·60 (n--n) and 1·2 (s4s) MJ) and energy intake (precision ranging from 0·4 to 1·8 MJ/d).
0−10−20−30−40−50−60 10 20 TOTAL
Schoeller et al. (1986) Schoeller & Webb (1984)
*Roberts et al. (1986) *Jensen et al. (1992)
Westerterp et al. (1988) Westerterp et al. (1984)
*Jones et al. (1987) Coward et al. (1984)
Ravussin et al. (1991) *Jones et al. (1988)
Parkinson (1990)
Klein et al. (1984)
Difference from indirect calorimetry
Fig. 4. Percentage difference in energy expenditure estimated by doubly-labelled water and indirect calorimetry (100 × doubly-labelled water/indirect calorimetry). Values are means and standard devia- tions represented by vertical bars. * Studies in infants; other studies involved adults.
��5 *���!( �� ���/
.����'� ���.����� ������� �����)�2 �����"�;���������(R���) ���� ��� !� > ���� &��!� "'� ���� 7 01#���"'� �"�(0�.�"�� �) (����&��� ���0�"�� "���;����(.���"�"���)�>�(0��(��!(1�!$����7� ������ �("0�� ������ ) 0����� ��� !(1�!$� ��� ; �$� .���� ("� )���4 !(;("#� 0�"'(�(�"�� �� "� ("� 0 ��)�!!$40�"���!!�'� .�� 7�!(0 ���'(����� ��� ;���������������� 7!(���0�"�� "0$������ 00�� 0$ "'� &��0(�(�"� �)� ����!��� �7� ("�'� ("� )���4!(;("#� 0�"'(�(�"� ���!(1�!$����7��>����������'(�0��& "0$�7��>��"�>��!�47�'$ ("'(��0��0 !��(.���$� "'�����7(0 �7�" �� ��� �.����'��−�?� �����?���B����−�?��� ���� �?���*:�'�� � �� 7��"� )��"'� ��� 7� �. !!�� 7��� ������ ; !���� ��&����"�� ����!��� �7� ("�'� �"!$� 7$ ����#���&�
����������
���� 0�"0�&�� �)� �"��#$� 7 ! "0�� 0 "� 7�� &&!(�'� ��� ��. " �(������������������("'(;('� !���7<�0�������#���&���)���7<�0�� "'��;�"������0(��(������#���"��#$�7 ! "0��'��("#�'�;�!�&.�"� "'� ��0�;��$� )��.� ) .("��� 7$�.� ���("#� ���� '(�0��& "0$ 7��>��"����(. ��'��"��#$���=�(��.�"��� "'�'(� &&� � "0���) )��'���&&!$���������'())���"��0 ��#��(����)��"��#$�7 ! "0� .� ����.�"�����=�(���'())���"��������)��&�0()(0� ���.&�(�"� "'�0 !0�! �(�"��)��.�������("'(0 ��'�("�����&����"��& &���)�� ("'(;('� !���. "���7<�0���
����.���� 00�� ��� "'�&��0(��������4���.�.� ����.�"�� �)��"��#$�7 ! "0�� ������ !!$��7� ("�'�("� ��()(0( !��";(��"4 .�"����>�(0��0 ""���7��#�"�� !(��'� ��� �� !4!()��0�"'(�(�"�� I����% .&!���'(��0��.� ����.�"����)�# �������%0� "#�� �� "��� "��. !!$� �"'��� 1�"� ("� �� !4!()�� 0�"'(�(�"��� !����#� ���$� 0 "� 7�� .� ����'� �;��� �, �� �"'��� "��. !� 0(�0�.4 �� "0��� ("� ��()(0( !!$4;�"�(! ��'� & �(�"��� >��� ��� (")���' >(��� � 1"�>"� .��"�� �)� �"��#$� �(����� �"��� !!$� �� & ��"��� !!$�� �"� ��0�"�� $� ��� '; "0��� � ;�� 7��"�. '�� (" .� ���("#��"��#$��%&�"'(�����("�)���4!(;("#���7<�0�����("# �� 0��� ��0�"(=����� 7��� (�� (�� "�0��� �$� ��� � 1�� ("��� 00��"� ���(�� 00�� 0$� "'� &��0(�(�"�� �"� ''(�(�"�� ���(. �(�"� �) �"��#$� 7 ! "0�� ��=�(���� .� ����.�"��� �)� �"��#$� ("� 1�� >�(0��� �� "��"0��� ("� 00�� 0$����&��0(�(�"�("�)���4!(;("# 0�"'(�(�"��� & ��(0�! �!$� ("� '())���"�� #���&�� �)� ("'(;('� !� >���. $�7�� ��))��("#� )��.�; �(����'(�� ������'; "0��� (" 7�'$� 0�.&��(�(�"� ��0�"(=���� 0 "� !��� &��;('�� 7����� (")��. �(�"� 7���� 0� "#��� ("� �"��#$� ������� 7��>��"� �>� �(.�� &�("���� 7��� ���(�� &��0(�(�"� (�� (" '�=� ��� )��� ��� 74 !(��("#������4���.�0� "#���("��"��#$�7 ! "0���� "#("#�)��. � '���� �)�>�' $���("�("'(;('� !���7<�0�������$� ����)�.��� ; !��� )��� #���&�� �)� ��7<�0���� )��� ! �#�� 0� "#��� ("� 7�'$ 0�.&��(�(�"� "'� )��� ���'(��� �)� !�"#��� '�� �(�"�� G�>�;��� ���� �����.�"���)��"��#$�7 ! "0���7$�.� ���("#�0� "#���(" 7�'$� 0�.&��(�(�"�� '���� "��� &��;('�� ("�(#���� ("��� ��� '$" .(0�� "'� ("�����! �(�"��(&�� �)� �"��#$� ("� 1��� �"��#$ �%&�"'(����� "'� ���(�� 0�.&�"�"���� ������ (�� � "��'� �� '�;�!�&�7������.����'�!�#$�)����������'$��)��"��#$�7 ! "0� ("�)���4!(;("#���7<�0����>(��� "'�>(������'(�� ���
$���������
3�"�'(0�� IM� �� ���� '� � ���� ��� ������ ��� 0� �� /� (���� �� �� � � �����1� ��� �����-���� ����"��������O ��("#��"���2� 2 �"�#(���"��(������)�O ��("#��"�
3("#� .����� -�������'(�� �$� �����.�"���)�("'(;('� !�E�.����'�� 00�� 0$�� "�>� ��0�"(=���� "'� ��0�..�"' �(�"��� �� �� ��� '- ��� ����� ����* ��%&����� �,��
3("#� .� ��� �� ��� H(.(� �(�"�� �)� ���� ; �(���� .����'�� )�� 0�!!�0�("#� '(�� �$� ("� 1�� ' � ��'���� � ����� �� ����2� �-��� � '%����� ����
2�> �'� O��� ���"�(0�� �*�� *��# ���$'� �9�� M�!'7��#� M9� G !!(' $����* 0� . � � :���� ;(���GH��2�!���:�J���>���* �� -���*� ����.�"���)�2 �� "'�> ����&��'�0�(�"�� ��� ("�. " ��("#� �G�� �- 4! 7�!!�'�G� E� 0�.& �(��"�� 7��>��"� 0 !��(.���� "'� (����&�� ; !����� �"� ������ )��� �� 2� �-��� ��� ��� ���� '� ��� �� ���� )��� �� )3����� ���� 2�� ������ � ��� )�������4 �� ����� � ������+� ��������$���� ������������ �(������� ��/ ���� � ��� ���)(�1��, ���� � �1� ���� ���� ������ ������� 5 )���4�� �� ���� �"������&&����5 ��-�S�:GQ���1���'(���T����" G #������������! "'���2��4M�#�;�"��N�"("1!(<1��
�!( �*��� �����"��; !� �(�"��)��>�� "'�.������ "��>��0�.&�"�"� .�'�!���)�7�'$�0�.&��(�(�"��(���������� �� �����((����, ����
�!( �*� �� �-�� �# "� "'� �(����� 0�"��(7��(�"� ���.�� 7�!(0� � ��� �" )��� �� 2� �-��� �5� �� ��� $� ������� � ���� (������� (��������� �� &&�� 5� � � S:*�N(""�$� "'�G����01���� �'(����T� ��>�P��1��9 ;�"�������
�!( �*� �� ��� �(����� '(���(7��(�"� "'� �"��#��(0�� ("� >�(#��� !��� "' �"'��"���(�(�"�� �"� ��� ���� ��� � �� � ���� 2���� �� ���/ 0��� ������(������ � ������� �� �������� ����� �����&&��-� ,�� S:*�N(""�$� "'�G����01���� �'(����T�� ��>�P��1�� H(&&("0���4 9 ;�"���7!(������
�!( �*� ��'(����� �������$� �� ���� ����2��� ���� � ��� �������4 �� ������� ���(������ �/�'� ���� �-������2���� �� ����'��� ��� ������ 6'� 1��,�� � ������ ��� ���� +�� � �� ' ���� ���� ��� ����� ���������)� ������� �� ���7��* ('�"�� '��3��1����3�(�(�� ����0( �(�"�)���� ��"��� !� "'��"��� !�����(�(�"�
�!( �*�J�H(;���$�M��� ����"��#$��%&�"'(����� "'�)��!���!�0�(�" ("� 7(�!�#(0 !� �$���.��� ���� �����$� "'� &� 0�(0�� �)� 0 !0�! �(�"� 7 ��'��"�("'(��0��0 !��(.���$� "'��� 0���.����'����� ���� ����� 8������������ �� ��������$�� � �� �&)��5- ����
�!( � *�� 9(�R� �� J� ���77�� 9:� ������� �"��#$� �%&�"'(����� ("� ��� �!'��!$��)�������� 8������� ��� (�������� �� �� ���� %*�� ��&&!�� �� � � �����
I�!!����:�� :�77�����H �1�$�*���2�> �'�O��J��!( �*� �� �� I���40�.&�"�"��.�'�!�)������� �����.�"���)�7�'$�0�.&��(�(�" ("���. "���0�.& �(��"�>(��� !���" �(;��.����'�� "'��; !� �(�" �)� ����'�"�(�$� "'��$'� �(�"��)� ) �4)����. ����(��������������� +���5-� 5 ��
I�!!����:��� >$���*3�J��!( �*��� ���3�'$�0�.&��(�(�"�0� "#�� )����;���(0 !�7 "'�# ����&! ��$�("��7����)�. !������������� ��� ����� �� ��������� ��%*�� ���
M(7"�$� �9� ������� ���� &�$�(0 !�� &�$0��!�#(0 !� "'� .�� 7�!(0 �))�0��� �)� "���(�(�" !� '�&!��(�"� "'� ��7��=��"�� ��&!��(�"�� ��� ����(���D"(;���(�$��)�2 .7�('#��
M�!'7��#� M9�� 3! 1� ���� :�77� ���� 2�!�� �:�� *��# ���$'� �9� 2�> �'�O��J����"�(0���*��� ���2�(�(0 !��; !� �(�"��)��"��#$ ("� 1�� ' � � ��("#� )�"' .�"� !� &�$�(�!�#(0 !� &�("0(&!��� �) �"��#$ &�$�(�!�#$�� ��� ���(; �(�"� �)� 0���))� !(.(��� ��� ('�"�()$ �"'��4��0��'("#�� )�������� 8������� ��� (�������� �� �� ���� � *%� �5 �-��
:�"��"�2H��3������I��O�"#�OO�J�*��"�:N��� ��������.("("# �"��#$��%&�"'(�����("�&�����.�(") "����0�.& �(��"��)��G����- .����'� "'� ("'(��0�� 0 !��(.���$�� '�������� 8������� �� ��� ���� ����'��95-� 95 ��
:�"��� �:�� O("����&� �H�� �0���!!��� ���� I(!!��� 9*�� �>$��� �9� �.(�� :�J�G�(.����� --���; !� �(�"��)�'��7!$� ! 7�!�'�> ��� )�� .� ���("#� �"��#$� �%&�"'(����� '��("#� 0� "#("#� "���(�(�"� '��������8����������(���������� �� ����*&��� -�,�
:�"����:��O("����&��H���0���!!��������>$����9���.(���:��I(!!�� 9*�J�G�(.��� �� -���Q !(' �(�"� �)� '��7!$� ! 7�!�'�> ���� )��
�������� ������� � � � ������� �� ���
���������������������������������� �������������� ������� �� ��������
�������� �!"�#�$��%���� � ����&��#� ����'�(�) � !��%*�+,-�./ ��� ������� �� ����� ������������ ���� 0,������ &���� �! ��� &12�3��4�������!��&�����!� �5�����
� ����56�� $ 7���85)��8!���88�� 9�4����:;��&��� ��!���5<� : ���� � &�� 6 !��!� %&�� � ���� �� (� 1�� ! �� *� +,-0�/ : !��7������4 �� ��!��!��� ���!�� �= �� ���> ����7�� !���!� �����7�� ��!��!��������������������� ���7 �������� ��������� �������� ������������������,.��
*���5�����! ��&�� � ��#���9��%!> ����$��%��#� �?��9�� "�>�#��� �� ���� +�..,/� ) �� �..,� : � �� �� : ���!4 ��� �� ;!����� �!������������� ������� ����!��� ��7 � ��7���� ������4����!� !�� � ���� � ������������� ������� �� ���������� ��� �������
'@<�� ���*�+�..�/��&��=���!������!���!!����� #�����3�� �� ����5 6 ) ������3��4�������!��:! �� ����
5 �#���!�� ;�� +,--./� � ����� 7� ����7���� !�� � ����� ��� �!�� �!7�!����!���5 6�) ������3��4�������!��: 7�������
< 4������ ��� % ����� 9�� <������ <� (� !� ����� :� +,--,/� ������ ����������������!�� �� �� ���> ���2�4 �� ��!����� � �� ���!���� ���A������!������� ������� �� ������������������.����.-�
<!������; ��:!> ���8���;� �������������=%��1! �� �?�(�*�� � �� +,-0�/� :!7� ���!�� !�� � �� �!�� �� �� ��� > ���� +�%�,0'/ 7�� !��>�� ����������� !��7����� ��� ���������=� �����������!� ��7� � ��!��������7�� ��!��!����������������������> ������� #� ���7�� �! �" � �� ������� ��� #�� ��� ������7� ��� �����!������� ������� �� �������� ���������� ���,������
;� !� ��� 6�� (� %�� �� � $&� +,-0�/� <� � �� ���� !�� � �� �!�� � �� ��� > ���� 7�� !�� �!�� � �� 7� ����7���� !�� �!� � ������ ������������ ��� ����= �4���� ���A������ ������� �� ���������� ���� ��0;����;�
;� !� ���6���< 4���������;� ��"�B���� ��!���$�� ����� ��5�( $�C�������+,-0�/������������������������!�� �� �� ���> ���2 4 �� ��!�� ��� �7 ���� !������� ������� �� ����������� � �� <0���<0�.�
;� !� ���6��(�8����5�+,-0�/�D�4�=� ���!7� ���!��!��� ���!�� � �� ���> ����7�� !��>�� ������� �!���� ����� �����!������� ������� �� �������� ���������� ���,���,�0�
; �����5;�(�$ 7���85)�+,--�/����� "��� ���#$ ! "������ �� ������� %����� &��������� �� !�����' (!) (��� ��� ��������� ��*����!����������,�����<!7�2�D�'�
;� !��� 66�� ������� �$�� ��� 7� ;�� (� & �7!�� &E� +,--�/ 6��� ��� �����7���� ���8 ��� � 992� � �� ��� ������ !�� ���!����� �� �� !�� �� ����� �!��!��!�!7��� 4 �� ��!�� ��� ��������� ��� #��� %���*��� ������� �� �������� ����������� ���0,��0���
;������<$��'@<�� ���*��D� ���F��%!�� ��E��&�����:��6� ��� 9� ����������<��"�5��&� �����(�$ 7���85)�+�..�/���&�������� ��� "�������� "��+��*������ �� (��� ���,� �� !�������*!��!�2�D!!� ;� �� �����������
4 ��8�A#�&:8�+,---/� �) �� �����7����!���!����!7�!����!�� �� !��������A�������������!� ������ ��7��� ����������!��!����5 6 ) ������3��4�������!�����������
8���������� �<�� �!���� D�� ; ���� 8%&� (� %!!�� D5� +,-00/ :!7� ���!�� !�� �!�� �� �� ��� > ���� >�� � ������!7����� � !> ��� �� � ���4���� �4� ���������� �� !**���� ������������� � ������
8�����������<��;� !��� ���5D&��; ����8%&�(�����%!!��D�+,-0�/ &� ����7���� !�� ������� ������������ ������ �!�� �� �� �� > ����� �4 �� ��!�� ������� �*��� �� �������� ���� �� -������� ���� ,�-�,�,�� & ����� ��� ) �� 1�� �� ���2� 9��������� 4!!�� �� ?!������
par epic(1)(1).pdf
International Journal of Epidemiology 2002;31:168–174
Physical activity is a complex and difficult exposure to assess in epidemiological studies.1 Improving the accuracy of its assess- ment is important so that the sub-dimension of this multi- dimensional exposure that is most closely associated with a particular health outcome can be identified.2 This precision is necessary to develop appropriate interventions that target the
specific sub-dimension of interest. Obtaining better information about the strength of association will lead to better estimates of the benefit that may result from successful behaviour change.
For reasons of practicality, most epidemiological studies have relied on questionnaires to assess activity.3–5 Many have focused on recreational rather than total activity, probably because it is easier to recall repeated discrete activities that are undertaken for a limited period of time and for which a conscious choice is made prior to engagement. The focus on recreational activity may be appropriate in studies which are occupationally defined or which have samples drawn from restricted socioeconomic groups, because in such studies there will be little variability in
© International Epidemiological Association 2002 Printed in Great Britain
Validity and repeatability of the EPIC-Norfolk Physical Activity Questionnaire Nicholas J Wareham, Rupert W Jakes, Kirsten L Rennie, Jo Mitchell, Susie Hennings and Nicholas E Day
Background Physical activity is an important lifestyle which is often poorly assessed in epidemiological studies. The European Prospective Investigation into Cancer Study-Norfolk cohort (EPIC-Norfolk), a large population-based cohort study, has developed a comprehensive questionnaire to assess activity in different domains of life aimed at assessing total energy expenditure. We report the repeatability of this instrument and its validity against repeated objective measures of fitness and energy expenditure undertaken throughout the time frame of reference of the questionnaire.
Methods The validity of the instrument was measured in 173 individuals randomly selected from a continuing population-based cohort study. Energy expenditure was assessed by four separate episodes of 4-day heart-rate monitoring, a method previously validated against whole body calorimetry and doubly-labelled water. Cardio-respiratory fitness was assessed by four repeated measures of sub-maximum oxygen uptake. At the end of the 12-month period, participants completed the physical activity questionnaire that assesses past-year activity at home, work and in recreation. Repeatability was assessed in a separate group of 399 randomly selected participants in EPIC who completed the physical activity questionnaire twice with a 3-month interval.
Results The age- and sex-adjusted correlation between the objective measure of daytime energy expenditure and the sum of recreational and occupational reported physical activity (in MET h per week) was 0.28 (P , 0.001). The reported time spent in vigorous activity was correlated with cardio-respiratory fitness (0.16, P , 0.05) and with the proportion of time when energy expenditure was more than five times basal (0.17, P , 0.05). The repeatability of the sum of recreational and occupational reported activity was high, r = 0.73.
Conclusions The indices of physical activity derived from this questionnaire have levels of validity and repeatability comparable to other physical activity instruments that are used in large epidemiological studies and which have undergone such intense development and testing.
Keywords Questionnaire, validity, reliability, exercise, epidemiologic measurements
Accepted 1 August 2001
Department of Public Health and Primary Care, University of Cambridge Institute of Public Health, Robinson Way, Cambridge CB2 2SR, UK.
Correspondence: Dr Nicholas J Wareham. E-mail: [email protected]. ac.uk
168
occupational activity and most of the heterogeneity between individuals will be in recreational activity. Such an argument would not hold, however, in populations with greater variation in work-related physical activity. The focus on recreational activity may, however, represent that element of activity that is most easily modifiable.
Although the totality of physical activity may be the aetiologic- ally important exposure for some endpoints,6,7 this has rarely been assessed by questionnaires. In their review of physical activity instruments, Montoye et al. noted that few question- naires had been designed to assess physical activity in work, recreation and domestic life.4
This paper describes the development, validation and repeatability testing of a comprehensive instrument designed to measure the different sub-dimensions of physical activity in the Norfolk cohort of the European Prospective Investigation into Cancer (EPIC-Norfolk).8 As this study is designed to measure associations of lifestyle factors with major chronic diseases in mid-to-late life, including cancer, cardiovascular disease, type 2 diabetes and osteoporosis, the goal was to develop a question- naire capable of assessing the key sub-dimensions of activity likely to be related to these endpoints; aerobic intensity, overall energy expenditure and load bearing. The resulting EPIC Physical Activity Questionnaire (EPAQ) goes under the name of EPAQ2 to distinguish it from the questionnaire used in the main EPIC study throughout Europe which was a short global assessment of activity derived from a Dutch questionnaire.9
Few previous physical activity validation studies have used objective methods as comparisons. The validation instruments selected for this study include a measure of cardio-respiratory fitness and of total energy expenditure as assessed by heart rate monitoring with individual calibration. This method has pre- viously been compared to the gold standard methods of doubly- labelled water and whole body calorimetry, with which it has a high correlation (0.93).10 Both the assessment of cardio- respiratory fitness and the level and pattern of energy expend- iture were repeated three times at 4-monthly intervals throughout the 12-month frame of reference of the questionnaire. The validation study was undertaken in a sub- group of participants in a population-based cohort study in Ely, Cambridgeshire. The participants in this study were of the same age, social class and ethnicity as the Norfolk population for whom the questionnaire was designed. The assessment of reliability took place within the EPIC-Norfolk cohort itself.
Methods Questionnaire development
The EPAQ2 is a self-completed questionnaire that collects past-year self-reported physical activity behaviours in a dis- aggregated way such that the information may be re-aggregated according to the dimension of physical activity that is of interest. The questionnaire consists of three sections: activity at home, work and recreation. In each case questions are closed rather than open-ended, to make them easy to complete and to facil- itate large-scale data entry. A copy of the questionnaire may be obtained from the EPIC-Norfolk website at http://www.srl.cam. ac.uk/epic/questionnaires/epaq2/. The EPAQ2 asks about activities in the context in which they are undertaken. As an example, other physical activity questionnaires ask people to
report how far they walk. This may not be a construct to which most people can easily relate. Instead EPAQ2 asks people about walking in the different sections of their life: in and around home, to work, how much at work, and walking for leisure. The questionnaire went through nine iterations with a series of groups of unselected volunteers in a continuing population- based study. Although the final version of questionnaire has closed questions with ordered categories of continuous variables that the volunteer has to select from, the selection of these categories was determined by the range of responses in early versions where the question was open framed. In the recreation section, which is derived from the Minnesota Leisure Time Activity questionnaire,11,12 activities are ordered according to the previ- ously reported frequency in a UK population.13 Where different intensities of activity are possible within a specific category e.g. running, then the more intense (e.g. competitive running) is asked before the less intense (jogging). The occupational activity is derived from the Modified Tecumseh Occupational Activity questionnaire that has been validated elsewhere.14
The validity of the questionnaire was assessed in a sub-group of 173 participants in the existing population-based Ely study and its reliability was measured in a sub-group of 399 partici- pants within the EPIC-Norfolk study itself.
Validity study
Selection of study participants The volunteers in the validity study were all participants in the Ely cohort, a continuing population-based study in Ely, Cambridgeshire, the detailed design of which has been described previously.6,15 In the second phase of the Ely Study, 790 individuals completed measurements of anthropometry, cardio- respiratory fitness and 4-day energy expenditure by heart rate monitoring.6 A random sub-set of 200 individuals from this cohort were asked to re-attend a further three times over the following year when all tests were repeated. Overall, 173 of the participants completed all measurements and therefore had four measures of cardio-respiratory fitness and four measures of 4-day energy expenditure by heart rate monitoring (Figure 1) completed across one year. At the final visit, participants com- pleted the EPAQ2 questionnaire that asks about physical activity in the past year. Ethical permission for the study was granted by the Cambridge Local Research Ethics Committee.
Measurement protocol On each of the four visits, height and weight were measured in light clothing and body fat percentage was obtained using a
EPIC-NORFOLK PHYSICAL ACTIVITY QUESTIONNAIRE 169
Figure 1 Time intervals for the assessment of objective measures of physical activity and assessment of past year physical activity by the questionnaire (EPAQ2)
standard impedance technique (Bodystat, Isle of Man). Body circumferences were measured using a metal tape. The waist circumference was measured at the mid-point between the lower costal margin and the level of the anterior superior iliac crest. Hip circumference was measured at the level of the greater trochanter. The protocol for undertaking the individual calibration between heart rate and energy expenditure has been reported previously.6,16 The oxygen consumption-heart rate relationship was assessed at rest with the subject lying prone and then seated, using an oxygen analyser calibrated daily using 100% nitrogen and fresh air as standard gases. To provide the slope and the intercept of the line relating energy expenditure to heart rate, each participant cycled at 50 revolutions per minute and the workload was progressively increased from 0 W, through 37.5 W, 75 W and 125 W in stages each lasting 5 minutes. At each workload three separate readings were made of heart rate, minute volume and expired air oxygen concentra- tion. The 125 W level was only undertaken if the heart rate had not reached 120 beats per minute by the end of the 5 minutes at 75 W. The oxygen concentration in the expired air and minute volume data were used to calculate oxygen consumption after correction for standard temperature and pressure. Energy ex- penditure (kJ/min) was calculated at each time point as oxygen consumption (ml/min) × 20.35.17 Mean resting energy expend- iture was taken as the average of the lying and sitting values. Flex heart rate, the empirical point at which the distinction between rest and exercise is made, was calculated as the mean of the highest resting pulse rate and the lowest on exercise. Finally the slope and intercept of the least squares regression line of the exercise points were calculated. Maximum oxygen uptake (VO2max) was measured from the linear regression as predicted oxygen consumption at maximal heart rate (220-age) and is expressed in the results per unit body weight. The volunteers wore the heart rate monitor (Polar Electro, Finland) continuously during the waking hours over the following 4 days. Heart rate readings were directly downloaded into a computer via a serial interface and the individual calibration data were used to predict minute energy expenditure for each person. The energy expenditure data were summed over the day to create an estimate of daytime energy expenditure (Daytime EE). Finally the minute-by-minute energy expenditure data was analysed to calculate the proportion of time when the physical activity ratio (PAR), the ratio of minute energy expenditure to minute basal metabolic rate (BMR), was >5.18 This was used in the analysis as an indicator of the pattern of physical activity. Basal metabolic rate was calculated from prediction equations.18
A cut-off of five was used because this energy intensity ratio discriminates moderate and highly intense activities from those that are light.19 For each individual the means of body mass index (BMI), percentage body fat, waist-hip ratio (WHR), day- time energy expenditure and VO2max on the four occasions were calculated and are used in the analysis as the measures of the usual level of obesity and its regional distribution, energy expenditure and fitness, respectively.
Repeatability study
Selection of study participants The repeatability of this questionnaire was assessed in par- ticipants in the EPIC-Norfolk study. At baseline, 1993–1997, EPIC-Norfolk recruited 25 633 people aged 45–74 years. In
1997 a follow-up study was commenced which finally resulted in repeat visits for 15 786 members of this cohort. By May 1999, 4853 people had attended for the second health check and had full data entry. They are used in the paper as the representative cohort from which the repeatability sub-sample was selected. At the follow-up visit participants provided a blood and urine sample and completed a detailed health and lifestyle question- naire and the EPAQ2 questionnaire. Beginning in January 1999, 544 volunteers were asked at the end of their visit whether they would complete a second physical activity questionnaire that would be posted to them in 3 months time. This interval was selected in order to have a high degree of overlap between the periods covered without the interval being too short such that participants simply recalled their first response when com- pleting the questionnaire on the second occasion.20 In all, 498 individuals agreed to complete the second questionnaire which was sent to them at three months. In all, 399 individuals returned completed questionnaires.
Statistical analysis
Estimates of energy expenditure at home, work and during recreation from the questionnaire were calculated by multiply- ing participation (h/week) by the metabolic cost of each activity, expressed in metabolic equivalents (MET) obtained from pub- lished tables.18,19 The sum in each category is given in units of MET.h per week. A MET is the ratio of the energy cost of a given activity to resting metabolic rate. In addition we calcu- lated time spent viewing television and videos (h/week) and reported time participating in more vigorous recreational activities i.e. those with a MET score .5.
t-test comparisons between men and women for mean anthro- pometric, energy expenditure and fitness measurements were undertaken. All the questionnaire-derived physical activity variables, apart from hours spent watching television, were skewed and comparison between men and women was per- formed on log-transformed data. The relationship between the questionnaire-derived physical activity variables and the object- ive measures, daytime energy expenditure, VO2max per kg and proportion of time with PAR >5 was described using correlation coefficients with and without adjustment for age and sex. The correlation between repeated assessments of the questionnaire- derived physical activity variables was described using Pearson correlation. In addition, participants were classified into quart- iles of the activity indices. If the raw activity score was equal to zero for more than 25% of the distribution then the baseline category constituted all subjects whose score was zero with the remaining positive scores being split into tertiles. If the number of participants with a zero score was less than 25% of the dis- tribution then the distribution was divided into quartiles. Cohen’s weighted kappa, a measure of agreement, was calculated according to the four resulting categories between the baseline and the repeat questionnaire.20
Results Validity study
The 173 participants in the validation study were middle-aged men and women with BMI and proportions that are typical of those seen in adults of this age in EPIC-Norfolk (Table 1). Average daytime energy expenditure as assessed by heart rate
170 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
monitoring was higher in men than women. Men also had a significantly higher maximum oxygen uptake. Comparison of the mean VO2max data to those seen in the Allied Dunbar National Fitness Survey suggests that the validation population is not any fitter than the general population.13 External comparison of the questionnaire-derived data with other populations is difficult, as limited national data are available. Mean reported television viewing was not significantly different between men and women. Reported energy expenditure at home was significantly greater in women than in men, whereas the reverse was true for activity at work. Total recreational activity was higher in men than in women, but there was no significant difference in time spent in vigorous recreational activity. Overall 58% of men and 54% of women participated in no vigorous activity.
Pearson’s partial correlation coefficients adjusted for age and sex between the objective measures of physical activity and variables derived from the questionnaire are given in Table 2. The questionnaire indices of energy expenditure in recreation and at work showed modest positive correlation with daytime
energy expenditure from heart rate monitoring. However, cor- relation between reported activity at home and daytime energy expenditure was negatively correlated. Reported activity at home was negatively correlated with reported activity at work and recreation (partial correlation coefficients adjusted for age and sex, –0.23 and –0.22, respectively). Thus an overall measure of physical activity summing domestic, occupational and recreational activity was less correlated with daytime energy expenditure than either occupational or recreational activity separately (data not shown). Therefore in constructing a summary index, the self-reported physical activity index, only occupational and recreational activity were summed. This self-reported physical activity index was significantly positively correlated with mean daytime energy expenditure after adjustment for age and sex (r = 0.28, P , 0.001). The unadjusted (r = 0.44, P , 0.001) and age-adjusted data (r = 0.45, P , 0.001) were more strongly cor- related. In sex-specific analyses adjusted for age, the correlation in men was 0.30 (P , 0.01) and 0.23 in women (P , 0.05).
Time spent watching television was negatively associated with our objective measure of cardio-respiratory fitness (VO2max per kg)
EPIC-NORFOLK PHYSICAL ACTIVITY QUESTIONNAIRE 171
Table 1 Validity study: anthropometric characteristics, objective physical activity measures and questionnaire-derived physical activity variables (n = 173)
Men Women (n = 84) (n = 89) P-value
Age (years)a 58.8 (7.9) 55.4 (6.7) 0.003
Body mass index (kg/m2)a 26.2 (2.7) 25.7 (4.4) NS
Per cent body fat by impedance (%)a 25.1 (3.9) 36.4 (6.1) ,0.001
Waist to hip ratioa 0.96 (0.07) 0.78 (0.07) ,0.001
Objective physical activity measures
Mean daytime energy expenditure (kJ/hr)a 754 (146) 556 (106) ,0.001
Mean VO2max (ml/min/kg) a 31.3 (7.2) 26.6 (5.4) ,0.001
Proportion time >5 PAR (%)a 1.7 (0.5) 1.3 (0.5) ,0.001
Questionnaire-derived variables
TV time (h/wk)a 20.2 (9.7) 21.0 (10.3) NS
Activity at home (MET.h/wk)b 14.2 (6.8–23.5) 62.0 (44.0–79.0) ,0.001
Activity at work (MET.h/wk)b 68.9 (0–102.6) 37.7 (0–66.0) 0.013
Recreational activity (MET.h/wk)b 26.2 (9.8–42.0) 14.0 (6.9–28.7) 0.009
Vigorous activity (h/wk)b 0.0 (0.0–1.0) 0.0 (0.0–1.0) NS
Self-reported physical activity index (MET.h/wk)b 95.5 (40.8–135.3) 57.6 (30.8–92.5) 0.001
Values shown are a mean and standard deviation or b median and interquartile range. Comparisons of means is by the t-test, and comparison of medians by the Mann-Whitney test. 58.3% of men and 53.9% of women did no vigorous activity. Mean daytime energy expenditure, mean of 4 days of heart-rate mon- itoring on four separate occasions throughout one year; Mean VO2max, mean of four tests of fitness over one year; Proportion of time >5 PAR, proportion of time when ratio of total energy expenditure to BMR was >5; MET.h, metabolic equivalents.hours.
Table 2 Validity study: partial correlation coefficients between questionnaire-derived physical activity variables and objective physical activity measures (n = 173)
Objective physical activity measures
Mean daytime energy expenditure Mean VO2max Physical activity ratio >5 Questionnaire-derived variables (kJ/h) (ml/min/kg) (%)
TV viewing (h/wk) –0.07 –0.19* 0.02
Activity at home (MET.h/wk) –0.04 –0.09 –0.12
Activity at work (MET.h/wk) 0.17* 0.01 –0.07
Recreational activity (MET.h/wk) 0.13 0.16* 0.20**
Vigorous activity (h/wk) 0.01 0.16* 0.17*
Self-reported physical activity index (MET.h/wk) 0.28*** 0.15* 0.05
Coefficients are adjusted for age and sex.
* P , 0.05, ** P , 0.01, *** P , 0.001.
and the coefficient was unchanged either in the unadjusted analysis or after adjustment for age and sex (r = –0.24 and r = –0.19, respectively). Reported energy expenditure in recrea- tional activity and time spent participating in vigorous activities were both positively correlated with VO2max per kg and also the proportion of time spent with PAR >5, our objective measure of the pattern of vigorous activity. The correlation of these ques- tionnaire parameters of vigorous activity was less strong with mean daytime energy expenditure, as would have been predicted.
Repeatability study
Table 3 shows a comparison of the questionnaire-derived physical activity variables in the larger EPIC cohort compared to the sub-set on whom a second EPAQ2 was collected. There were no statistically significant differences between the main EPIC-Norfolk cohort and the repeatability sub-sample. In men the correlation coefficients for the various indices between the repeat questionnaires were >0.68 for all indices except work activity, which had a correlation coefficient of 0.57 (Table 4). In women, the correlation coefficients were .0.60 for all indices except work activity and amount of time spent participating in vigorous sporting activities which were 0.37 and 0.41, respectively. All of the correlations were statistically significant at the 0.05 level.
Discussion
The questionnaire described in this paper was designed to assess the energy expenditure and fitness enhancing elements of physical activity, a complex and difficult exposure to measure in epidemiological studies. The results suggest that the repeatability of the indices computed from the questionnaire is high but that validity compared to repeated objective estimates of energy expenditure measured throughout the frame of reference of the questionnaire is modest. The 3-month interval between repeats of the questionnaire in the repeatability study, chosen to reduce the probability that individuals recall their previous response, might be considered relatively long. However, the effect of a lengthy re-test interval would be to diminish rather than exag- gerate the correlation between repeats. Thus the high values of repeatability demonstrated represent, at worse, an under- estimate of the true repeatability. The measurement of validity of physical activity questionnaires is difficult, but our study design allows for assessment with objective and independent comparison instruments repeated throughout the frame of reference of the questionnaire. The study group selected for the validation study were chosen at random from a continuing population-based cohort study. As such they tend to be un- selected with regard to physical activity behaviour. By contrast in the closest comparable study, the SAFE study, the participants
172 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
Table 3 Repeatability study: descriptive characteristics and questionnaire-derived physical activity variables in the EPIC cohort (n = 4853) and repeat sample (n = 399)
Men Women
Total cohort Repeat sample Total cohort Repeat sample (n = 2126) (n = 187) (n = 2727) (n = 212)
Age (years)a 64.6 (8.4) 65.0 (8.2) 63.2 (8.5) 63.8 (8.4)
Body mass index (kg/m2)a 26.9 (3.6) 26.9 (3.3) 26.8 (4.9) 26.9 (4.0)
TV viewing (h/wk)a 21.6 (9.8) 24.6 (10.5) 22.5 (9.9) 24.3 (11.4)
Activity at home (MET.h/wk)b 18.1 (9.1,30.4) 20.6 (12.9,32.0) 58.5 (41.2,78.2) 55.2 (41.2,74.0)
Activity at work (MET.h/wk)b 0.0 (0.0,70.4) 0.0 (0.0,15.7) 0.0 (0.0,15.8) 0.0 (0.0,16.8)
Recreational activity (MET.h/wk)b 27.8 (14.7,48.0) 28.4 (14.9,47.0) 17.6 (7.4,31.8) 16.7 (8.4,28.8)
Vigorous activity (h/wk)b 0.0 (0.0,0.8) 0.0 (0.0,0.5) 0.0 (0.0,0.7) 0.0 (0.0,0.7)
Self-reported physical activity index (MET.h/wk)b 54.9 (25.0,108.1) 43.0 (20.1,86.2) 26.7 (11.0,57.4) 24.7 (10.8,49.0)
Values shown are a mean and standard deviation and b median and interquartile range.
Table 4 Repeatability study: Pearson correlation coefficients and Cohen weighted kappa statistics for 3-month repeat of questionnaire-derived variables of physical activity (n = 399)
Correlation coefficient (r) Men Women Weighted kappa (k) (n = 187) (n = 212)
TV time (h/wk) ρ 0.75 0.78 κ 0.71 0.74
Activity at home (MET.h/wk) ρ 0.77 0.74 κ 0.61 0.62
Activity at work (MET.h/wk) ρ 0.57 0.37 κ 0.79 0.82
Recreational activity (MET.h/wk) ρ 0.69 0.64 κ 0.54 0.55
Vigorous activity (h/wk) ρ 0.75 0.41 κ 0.58 0.67
Self-reported physical activity index (MET.h/wk) ρ 0.74 0.72 κ 0.66 0.70
EPIC-NORFOLK PHYSICAL ACTIVITY QUESTIONNAIRE 173
were paid volunteers, mostly graduates employed in administrat- ive or professional positions.21 The participants in our validation study were on average 5 years younger than the population for whom the questionnaire was designed. However, this is un- likely to result in a biased assessment of validity as the physical activity patterns of the two groups are likely to be similar, containing a mixture of working age and retired people. The relatively low validity may be a reflection of the difficulty in capturing the between-individual differences in energy expend- iture in a simple questionnaire. One of the obvious limitations is that a single estimate of the energy cost of an activity taken from a published compendium is applied to all individuals.19
This does not allow for any between-individual variation in energy expenditure for a given activity and so cannot take account of differences in intensity (e.g. through people playing sport at different competitive levels), or through variations in mechanical and metabolic efficiency. The use of the compendium values of energy expenditure is also limited by the validity of the estimates for each activity as they are derived from studies on small groups of selected volunteers. However, the published estimates of energy costs of activities are the only such data that are available.
The validity and reliability results are similar to those pre- viously reported for other questionnaires. In a review of avail- able questionnaires, Kriska and colleagues22 noted how most, but not all, previously published questionnaires were supported by reliability studies and where this was measured, it was almost always high. By contrast, validity was less frequently reported and where it was, it was typically low. One particular problem for the validation of physical activity questionnaires is the choice of appropriate comparison instrument.23 Many physical activity validation studies have used other forms of subjective questionnaire or diary as the validation method. Although this strategy tends to produce higher correlations, the possibility of correlated error is substantial as both the questionnaire under scrutiny and the validation instrument are of the same fundamental type and are subject to the same forms of bias.24–26 It would be preferable therefore to select an ob- jective non-questionnaire based method as the validation instru- ment, ideally one with a high correlation to the true exposure of interest.27 As physical activity behaviour is related to several different underlying physiological constructs including aerobic intensity, overall energy expenditure and load-bearing, researchers should specifically identify which dimension of this complex exposure is being assessed by a questionnaire so that an appropriate validation instrument can be selected.1 If an index is derived from a questionnaire to be a measure of energy expenditure, then comparison to a measure of another under- lying dimension such as cardio-respiratory fitness may not be appropriate because although the different dimensions of activity are correlated, that correlation is not perfect.
Relatively few studies have utilized repeated assessment of an objective measure of energy expenditure throughout the frame of reference of the questionnaire and those that have tend to show lower correlations with the objective measure than with other subjective instruments. Richardson et al. compared the ARIC/Baecke questionnaire with physical activity records and repeated measurements of body fat, peak oxygen consumption and energy expenditure from a movement sensor (Caltrac).28
The one-month test-retest reliability measures for the three aspects of activity assessed by this questionnaire were high (range 0.86–0.92). Although the age-adjusted partial correla- tion coefficients between total leisure time physical activity from the questionnaire and the physical activity record were reasonably high (0.59 in men, 0.33 in women), the correlation with the objective Caltrac assessment of energy expenditure was much lower (men 0.24, women 0.19). A similar disparity between the correlation of a questionnaire with a physical activity record as compared to an objective measure was observed for the Minnesota LTPA questionnaire, for which the comparison with Caltrac was 0.21.12 Similar low correlations were observed in a comparison study of the Godin and Baecke questionnaires with Caltrac.29 In a study comparing eight physical activity questionnaires, the correlation between the questionnaire-derived physical activity index and an estimate of daily physical activity related energy expenditure as defined by total energy intake minus resting energy expenditure, ranged from 0.05 for Health Insurance Plan questionnaire to 0.32 for the Harvard Alumni study questionnaire.30
Previous authors have suggested that the low levels of validity may be related, in part, to the manner in which the questionnaire has been developed and how the questions are framed. Jacobs and colleagues in a study reporting an analysis of 10 physical activity questionnaires noted that ‘the capacity of a questionnaire to perform well against validation measures does not appear to be solely related to its length and attention to detail. More important seems to be the logic with which questions are constructed’.21 In particular they recommended that questions should target specific physical activity domains in the contexts in which people usually perform the target activity. Thus the logical development and construction of a question- naire plays a key role in determining its validity. It is therefore appropriate to describe development alongside validation.31
In summary, the repeatability and validity studies on EPAQ2 suggest that it is a useful tool for assessing physical activity in large epidemiological studies such as EPIC-Norfolk. As with any physical activity questionnaire, the inference about its validity and reliability is limited to the population in whom it was tested and its use is restricted to the purpose for which it was intended i.e. the assessment of past-year usual activity. In other situations where physical activity might be assessed, such as population monitoring, the evaluation of interventions or between- population comparison, other questionnaires may be more suitable.
Acknowledgements The Ely Study was funded by the Medical Research Council and Anglia and Oxford Regional Health Authority. We are grateful to H Shannasy, S Curran, P Murgatroyd, and Drs M Hennings and AM Prentice for their help with the fieldwork for the Ely study and to Suzie Oakes, Robert Luben, Joanna Camus and the EPIC-Norfolk field team for their assistance. NJW is an MRC Clinician Scientist Fellow.
A copy of the EPAQ2 questionnaire can be downloaded from the EPIC-Norfolk website at http://www.srl.cam.ac.uk/epic/ questionnaires/epaq2/
174 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
References 1 Caspersen CJ, Powell KE, Christenson GM. Physical activity, exercise,
and physical fitness: definitions and distinctions for health-related research. Public Health Rep 1985;100:126–31.
2 Wareham N, Rennie K. The assessment of physical activity in individuals and populations: why try to be more precise about how physical activity is assessed? Int J Obesity 1998;22(Suppl.2):S30–S38.
3 LaPorte RE, Montoye HJ, Caspersen CJ. Assessment of physical activity in epidemiologic research: problems and prospects. Public Health Rep 1985;100:131–46.
4 Montoye HJ, Kemper HCG, Saris WHM, Washburn RA. Measuring Physical Activity and Energy Expenditure. Champaign, IL: Human Kinetics, 1996.
5 Drury TF (ed.). Assessing Physical Fitness and Physical Activity in Population-Based Surveys. Hyattsville, MD: National Center for Health Statistics, 1989.
6 Wareham NJ, Wong M-Y, Day NE. Glucose intolerance and physical inactivity: the relative importance of low habitual energy expenditure and cardiorespiratory fitness. Am J Epidemiol 2000;152:132–39.
7 Wareham NJ, Wong M-Y, Hennings S et al. Quantifying the association between habitual energy expenditure and blood pressure. Int J Epidemiol 2000;29:655–60.
8 Day NE, Oakes S, Luben R et al. EPIC-Norfolk: study design and characteristics of the cohort. Br J Cancer 1999;80(Suppl.1):95–103.
9 Pols MA, Peeters PHM, Ocke MC, Slimani N, Bueno-de-Mesquita BH, Collette JA. Estimation of reproducibility and relative validity of the questions included in the EPIC physical activity questionnaire. Int J Epidemiol 1997;26(Suppl.1):181–89.
10 Spurr GB, Prentice AM, Murgatroyd PR, Goldberg GR, Reina JC, Christman NT. Energy expenditure from minute-by-minute heart- rate recording: comparison with indirect calorimetry. Am J Clin Nutr 1988;48:552–59.
11 Taylor HL, Jacobs DR, Schucker B, Knudsen J, Leon AS, Debacker G. A questionnaire for the assessment of leisure time physical activities. J Chron Dis 1978;31:741–55.
12 Richardson MT, Leon AS, Jacobs DR, Ainsworth BE, Serfass R. Comprehensive evaluation of the Minnesota leisure time physical activity questionnaire. J Clin Epidemiol 1994;47:271–81.
13 The Sports Council, The Health Education Authority. Allied Dunbar National Fitness Survey. London: Health Education Authority, 1992.
14 Ainsworth BE, Jacobs DR, Leon AS, Richardson MT, Montoye HJ. Assessment of the accuracy of physical activity questionnaire occu- pational data. J Occup Med 1993;35:1017–27.
15 Wareham NJ, Byrne CD, Williams R, Day NE, Hales CN. Fasting proinsulin concentrations predict the development of type 2 diabetes. Diabetes Care 1999;22:262–70.
16 Wareham NJ, Hennings SJ, Prentice AM, Day NE. Feasibility of heart rate monitoring to estimate total level and pattern of energy
expenditure in a population-based epidemiological study. Br J Nutr 1997;78:889–900.
17 Elia M, Livesey G. Energy expenditure and fuel selection in biological systems: theory and practice of calculations based on indirect calorimetry and tracer methods. In: Simopoulus AP (ed.). Metabolic Control of Eating, Energy Expenditure and the Bioenergetics of Obesity. Basel: Karger, 1992, pp.68–131.
18 James WPT, Schofield EC. Human Energy Requirements. Oxford: Oxford Medical Publications, 1990.
19 Ainsworth BE, Haskell WL, Leon AS et al. Compendium of physical activities: classification of energy costs of human physical activities. Med Sci Sports Exerc 1993;25:71–80.
20 Armstrong BK, White E, Saracci R. Principles of Exposure Measurement in Epidemiology. Oxford: Oxford University Press, 1994.
21 Jacobs DR, Ainsworth BE, Hartman TJ, Leon AS. A simultaneous evaluation of 10 commonly used physical activity questionnaires. Med Sci Sports Exerc 1993;25:81–91.
22 Kriska AM, Caspersen CJ. A collection of Physical Activity Questionnaires for Health-Related Research. Med Sci Sports Exerc 1997; 29(6Suppl.):S1–S205.
23 Rennie KL, Wareham NJ. The validation of physical activity instru- ments for measuring energy expenditure: problems and pitfalls. Public Health Nutr 1998;1:265–71.
24 Weiss TW, Slater CH, Green LW, Kennedy VC, Albright DL, Wun C. The validity of single-item,self assessment questions as measures of adult physical activity. J Clin Epidemiol 1990;43:1123–29.
25 Aaron DJ, Kriska AM, Dearwater SR, Cauley JA, Metz KF, LaPorte RE. Reproducibility and validity of an epidemiologic questionnaire to assess past year physical activity in adolescents. Am J Epidemiol 1995;142:191–201.
26 Wolf AM, Hunter DJ, Colditz GA et al. Reproducibility and validity of a self-administered physical activity questionnaire. Int J Epidemiol 1994;23:991–99.
27 Wong MY, Day NE, Wareham NJ. The design of validation studies II: the multivariate situation. Stat Med 1999;18:2831–45.
28 Richardson MT, Ainsworth BE, Wu H-C, Jacobs DR, Leon AS. Ability of the atherosclerosis risk in communities (ARIC)/Baeke questionnaire to assess leisure-time physical activity. Int J Epidemiol 1995;24:685–93.
29 Miller DJ, Freedson PS, Kline GM. Comparison of activity levels using the Caltrac accelerometer and five questionnaires. Med Sci Sports Exerc 1994;26:376–82.
30 Albanes D, Conway JM, Taylor PR, Moe PW, Judd J. Validation and comparison of eight physical activity questionnaires. Epidemiology 1990;1:65–71.
31 Kriska AM, Knowler WC, LaPorte RE et al. Development of question- naire to examine relationship of physical activity and diabetes in Pima Indians. Diabetes Care 1990;13:401–11.
KEY MESSAGES
• Physical activity is an important but difficult exposure to assess in epidemiological studies.
• The choice of comparison method and selection of study population are key determinants of the inferences that can be drawn from validation studies of physical activity questionnaires.
• The reliability of the European Prospective Investigation into Cancer Study-Norfolk cohort (EPIC-Norfolk) physical activity questionnaire (EPAQ2) has been demonstrated in a sub-study within EPIC itself.
• The validity of EPAQ2 has been demonstrated in a study in a comparable population using repeated objective assessment of energy expenditure and fitness throughout the frame of reference of the questionnaire.
Levine measurement of energy expenditure(1).pdf
Measurement of energy expenditure
James A Levine* Mayo Clinic, Endocrine Research Unit, 5-194 Joseph, 1st Street SW, Rochester, MN 55902, USA
Abstract Measurement of energy expenditure in humans is required to assess metabolic needs, fuel utilisation, and the relative thermic effect of different food, drink, drug and emotional components. Indirect and direct calorimetric and non-calorimetric methods for measuring energy expenditure are reviewed, and their relative value for measurement in the laboratory and field settings is assessed. Where high accuracy is required and sufficient resources are available, an open-circuit indirect calorimeter can be used. Open-circuit indirect calorimeters can employ a mask, hood, canopy or room/chamber for collection of expired air. For short-term measurements, mask, hood or canopy systems suffice. Chamber-based systems are more accurate for the long-term measurement of specified activity patterns but behaviour constraints mean they do not reflect real life. Where resources are limited and/or optimum precision can be sacrificed, flexible total collection systems and non-calorimetric methods are potentially useful if the limitations of these methods are appreciated. The use of the stable isotope technique, doubly labelled water, enables total daily energy expenditure to be measured accurately in free-living subjects. The factorial method for combining activity logs and data on the energy costs of activities can also provide detailed information on free-living subjects.
Keywords Total energy expenditure
Calorimetry Doubly labelled water
Thermogenesis Basal metabolic rate
Accelerometers Hear rate monitors
Introduction
There are three components to total energy expenditure
(TEE) in humans: basal metabolic rate (BMR), thermic
effect of food and the energy expenditure of activity
(activity thermogenesis). BMR is the energy expended
when an individual is lying at complete rest, in the
morning after sleep in the post-absorptive state. In
individuals with sedentary occupations basal metabolic
rate accounts for approximately 60% of the total daily
energy expenditure and is highly predicted by lean body
mass within and across species. Resting energy expen-
diture, in general, is within 10% of the BMR and is
measured in subjects at complete rest in the post-
absorptive state. Thermic effect of food is the increase in
energy expenditure associated with digestion, absorption
and storage of food, and accounts for approximately 10%
of the total daily energy expenditure; many believe there
exist facultative as well as fixed components. Activity
thermogenesis is the thermogenesis that accompanies
physical activities and, therefore, can be divided into
exercise and non-exercise activity thermogenesis (NEAT).
Most individuals do not partake in purposeful sporting
exercise and so their exercise-related activity thermogen-
esis is zero; for those who do exercise regularly, exercise-
related energy expenditure is generally ,10% of the total daily energy expenditure. NEAT or the ‘energy expendi-
ture of spontaneous physical activity’ encompasses the
combined energy costs of the physical activities of daily
living, fidgeting, spontaneous muscle contraction and
maintaining posture when not recumbent, and accounts
for the remainder of the total daily energy expenditure for
most individuals. Other thermogenic variables may also
need to be considered, such as the energetic costs of
altered temperature, medications and emotion.
Each of these components of energy expenditure is
highly variable and the total effect of these variances
determines the variability in daily energy expenditure
between individuals. Also, measurements of energy
expenditure can be used to assess the relative thermic
effects of different foods, nutrient compositions, beve-
rages, medications and psychological components.
Description of methods
Energy expenditure can be measured using one of the
three approaches:
1. In indirect calorimetry, oxygen consumption and/or
carbon dioxide production is measured and converted
to energy expenditure using formulae1,2.
2. In direct calorimetry, the rate of heat loss from the
subject to the calorimeter is measured.
3. A number of non-calorimetric techniques have been
used to predict the energy expenditure by extrapolation
from physiological measurements and observations.
The accuracy, reproducibility and reliability of the
measurements obtained using these various techniques
q The Author 2005*Corresponding author: Email [email protected]
Public Health Nutrition: 8(7A), 1123–1132 DOI: 10.1079/PHN2005800
vary enormously as the complexity and cost of the
techniques themselves. The techniques are summarised in
Table 1.
Indirect calorimetry
There are five principal approaches to the measurement of
energy expenditure using indirect calorimetry.
Total collection systems
Here, expired air is collected in either an airtight rigid
structure or a portable flexible bag.
Rigid total collection system
The Tissot Gasometer is an example of a rigid total
collection system3. It comprises a 100–1000 litre capacity
inverted glass bell fitted with an internal circulation fan
suspended over water. The bell is emptied of air. The
subject then expires through a mouthpiece and a non-
return valve into the bell, which gradually fills with
expired air and progressively rises above the water seal.
The height reached by the bell is recorded every minute
for up to 2 hours and the composition of expired air is
periodically measured from the bell to determine oxygen
consumption and/or carbon dioxide production.
Flexible total collection system
The Douglas bag4–7 is an example of a flexible total
collection system. It comprises a polyvinyl chloride (or
other leak-proof material) bag of typically 100–150 litre
capacity. The top of the bag is connected by tubing to a
three-way valvewhichmaybe rotated to either seal thebag,
admit atmospheric air or admit expired air via tubing
attached to a respiratory valve. To use this approach, the
three-way valve is first rotated to open the circuit to
atmospheric air, the bag is rolled up to expel its contents
and the three-way valve then rotated to seal the bag.
The subject breathes through the mouthpiece and the
three-way valve is rotated to allow the expired air for
10–20 minutes. After the timed collection period, the three-
way valve is turned to seal the bag. An alternative valve
system employs two valves, one at the mouthpiece and the
other proximal to the bag. Several bags can be used to
prolong the total measurement period8. After collection of
the expired air, the volume of the expired air in the bag is
measured (for example using a mass flow meter) and a
sample is analysed to determine oxygen and/or carbon
dioxide concentrations. Under optimal conditions the error
of energy expenditure measurements undertaken with
Douglas bags may be very small (,3%) but can increase
substantially if the equipment is poorly maintained, the
volume, oxygen and/or carbon dioxide measurements
inaccurate and/or the operator untrained or unskilled.
Open-circuit indirect calorimeter systems
Open-circuit systems can be used to record energy
expenditure over several hours or days depending upon
the configuration selected and the experimental require-
ments. In an open-circuit system, the subject inspires air
and the expired gases are then analysed. There are two
types of open-circuit systems: ventilated open-circuit
systems where a subject breathes into a container through
which air is drawn, and expiratory collection systems
where a subject inspires from the atmosphere and expires
via a non-return valve into a measurement unit.
Ventilated open-circuit systems
In general, ventilated open-circuit systems comprise
components to collect and mix expired air, measure flow
rate, analyse gas concentrations and pump air through the
system.
The method of collecting expired air varies consider-
ably. The least complex approach is for expired air to be
collected using a mouthpiece, mask, transparent hood or
canopy9–13. Measurements can be performed for up to
several hours. The most more complex approach is for the
subject to be placed inside a room/chamber of known
volume in which there are often sophisticated sensing
Table 1 Summary of techniques used to measure energy expenditure in humans
Approach Type of calorimeter Basal metabolic
rate Resting energy expenditure
Thermic effect of food
Energy expenditure of specific activities
Total daily energy expenditure
Indirect calorimeter
Room open-circuit Yes Yes Yes Yes Yes (confined subject)
Hood/canopy open-circuit Yes Yes Yes Yes No Open-circuit expiratory
collection Yes* Yes* Yes* Yes Yes*
Doubly labelled water No No No No Yes Total collection Douglas
bag Yes Yes Yes* Yes No
Direct calorimeter Yes Yes Yes Yes Yes (confined subject)
Non-calorimetric methods
No No No No Estimated
‘Yes’ represents where a technique can be used to perform the respective measurement and ‘No’ where it cannot. *Precision may be unreliable.
JA Levine1124
devices to quantify physical activity14. Using a room or
chamber, measurements can be performed for up to
several days15–17.
Regardless of how the expired air is collected, the basic
components of ventilated open-circuit indirect calori-
meters are similar. Expired air is drawn out of the
collection device using a pump; it is critical to measure this
flow rate accurately. The expired air is then mixed using a
fan and/or mixing chamber, and a sample of the expired
air is dried and analysed for oxygen and/or carbon dioxide
concentrations. Oxygen is generally analysed by using
paramagnetic analysers and carbon dioxide by using
infrared analysers; alternatively, a mass spectrometer can
be used to measure the gas concentrations. Burning
known masses of chemical standards, such as butane or
ethanol, within the system and ascertaining what
proportion of the burned mass is detected by the
calorimeter verifies the precision of these calorimeters.
Ventilated open-circuit indirect calorimeters have pre-
cision to within 0.5–2% of actual.
Depending on the software, air mixing and room
volume, response time for a room or chamber system can
vary from ,5 to 30 minutes; for a ventilated hood or canopy, ,2 minutes and for a mask or mouthpiece ,30 seconds. Hood/canopy/mask-based systems are
often configured as a ‘metabolic cart’ whereby the
equipment could be readily moved on a wheeled cart.
Finally, it is important that inspired carbon dioxide should
not exceed 1%12 as concentrations in excess of this may
increase respiratory effort.
Expiratory collection open-circuit systems
There are several expiratory collection open-circuit
systems. The advantage of this approach is that the
calorimeter can be designed as a portable device so that
energy expenditure can be measured in free-living
individuals. In general, these devices comprise a mouth-
piece or a mask connected to a one-way valve whereby
expired air enters the instrument. The flow rate of expired
air through the valve is measured and a small proportion
of the expired air is diverted to a gas storage reservoir that
is analysed at the end of each measurement period.
Measurements can be carried out using such instruments
intermittently for up to 2 days. Various modifications have
been applied to this principle; for example, by having air
drawn through the system at a fixed rate18,19. Technologi-
cal advance20,21 has resulted in the design of more precise,
robust and dependable portable calorimeters that are
likely to provide useful field data in the future.
Confinement system (respiratory chambers)
The subject is placed inside a gas-tight sealed container of
known volume (e.g. 16 m3)22 and oxygen consumption
and carbon dioxide production are estimated from
changes in the concentrations of these gases in chamber
air over time23. The period of observation may be
prolonged by periodically flushing the chamber with fresh
air. A number of confinement systems have been
constructed with errors of ,2% and response times of ,50 minutes. Currently, confinement systems are rarely used.
Closed-circuit systems
Closed-circuit systems consist of a sealed respiratory gas
circuit in which gaseous concentrations are measured over
time24. In one conformation, expired air was drawn from a
sealed 5 m3 chamber. Carbon dioxide and water vapour
were absorbed and then oxygen was reintroduced into the
air stream, which re-entered the chamber. Energy
expenditure was calculated from the quantities of carbon
dioxide absorbed and oxygen re-introduced. A small-scale
application of this approach was to use a spirometer. A
spirometer consists of an oxygen-containing bell from
which the subject inspires; expired carbon dioxide and
water vapour are absorbed before the expired air is
reintroduced to the bell. The bell is suspended over water
so that its height descends at a rate proportional to oxygen
consumption. Closed-circuit systems are rarely used at
present.
Direct calorimetry
An exhaustive description of the different types of direct
calorimeter is beyond the scope of this paper. In general,
the instruments are extremely expensive to build
(.$1 000 000) and run, requiring at least one full-time
technician. They require enormous expertise to establish
and maintain and offer little to the majority of investigators
beyond less expensive and complex indirect calorimeters.
Application of direct calorimetry is in the domain of highly
specialised laboratories where direct heat loss measure-
ments are of specific value.
Direct calorimeters measure the heat lost from the body.
Radiative and convective heat losses account for
approximately 80% of the total heat loss, while
evaporative heat loss accounts for the remainder.
Conductive heat loss is negligible in humans.
There are three principal types of direct calorimeter:
isothermal, heat sink and convection systems. These
approaches have on occasion been used in combination.
Isothermal systems
An isothermal calorimeter consists of a chamber lined with
a layer of insulating material25. The inner aspect of the
layer is in thermal equilibrium with the inside of the
chamber and the outside aspect of the layer is in thermal
equilibrium with the chamber wall, which is maintained at
a constant temperature using circulating fluid. The
temperature gradient across the insulating layer is
proportional to the non-evaporative heat loss from the
subject in the calorimeter. The response time of these
Measurement of energy expenditure 1125
instruments can be ,5 min and measurement error (under
optimal operator application) is 1%.
Heat sink or adiabatic systems
These calorimeters consist of a chamber from which heat
lost by the subject is extracted by a liquid-cooled heat
exchanger26. The rate of heat extraction is regulated so
that the temperatures of the inner and outer chamber walls
are equal, producing a ‘zero temperature gradient wall’.
The response time of these instruments is 10–30 minutes
and measurement error is 1–2%. A ‘suit calorimeter’ was
devised based on this principle. The suit weighed ,10 kg
and could be worn by a subject for up to 48 hours27.
Although evaporative heat loss was impaired by this
device, measurement error was ,3%.
Convection systems
These calorimeters consist of an insulated chamber
ventilated with an air flow at a known rate. Heat lost by
a subject inside the chamber is calculated from the flow
rate, the specific heat capacity of the air and the increase in
temperature of ventilating air leaving the chamber28–30.
The response time of these instruments is 10–20 minutes
and measurement error is 1–2%.
Non-calorimetric methods for measuring TEE
Non-calorimetric methods estimate energy expenditure by
extrapolation from variables that relate to energy
expenditure. These methods are often standardised
against calorimetric methods.
Isotope dilution, doubly labelled water
In the doubly labelled water method, both the hydrogen
and the oxygen of water are labelled or ‘tagged’ using
stable, non-radioactive isotopes (D2O 18)31–36. Elimination
of administered D2O 18 may be used to estimate carbon
dioxide production and energy expenditure.
The principle of this technique is as follows. In body
water, O2 of expired CO2 is in equilibrium with O2:
CO2 þ H2O $ H2CO3 Thus, if O2 in body water is tagged with the tracer O
18, the
label will distribute in not only body water but also
circulating H2CO3 and expired CO2. Over time, the
concentration of O2 label in body water will decrease as
CO2 is expired and body water is lost in urine, perspiration
and respiration. If H2 in body water is tagged with the
tracer D2, the label will distribute solely in the circulating
H2O and H2CO3. Over time, the concentration of H2 label
will decrease as body water is lost (some of the hydrogen
can become portioned into body protein or fat, however).
Thus, if both O2 and H2 in body water are tagged with
known amounts of tracers at the same time, the differences
in the elimination rates of the O2 and H2 tracers will
represent the elimination rate of CO2.
Subjects are usually given doubly labelled water orally
after baseline samples of urine, saliva or blood have been
collected. Time is allowed for complete mixing of isotopes
to occur within the body water space and then samples of
urine, saliva or blood are collected over 7–21 days. These
samples are used for the measurements of D2 and O 18
enrichments using mass spectroscopy. Changes in D2 and
O18 concentrations in body water are then calculated over
time, and CO2 production and energy expenditure are
calculated. Energy expenditure can be measured over
7–21 days using this technique with an error of ,6–8%. This error can be decreased to a small degree by collecting
samples repeatedly over the measurement period rather
than by collecting them only before and after the
measurement period.
Physiological measurements
Heart rate monitoring
In humans, there is a significant relationship between heart
rate and energy expenditure, at least in the absence of
exercise. Heart rate monitors are portable, non-restraining
and unobtrusive and measurements can be carried out over
several days. A number of devices of varying complexity
have been used to record heart rate in free-living
subjects37–45. The conceptual limitation of this approach is
thatenergyexpenditureandheart ratearenot linearly related
for an individual in part because cardiac stroke volume
changes with changing heart rate and even posture. There is
a substantial inter-individual variance for the relationships
betweenheart rate andenergyexpenditure in termsof slope,
intercept and curve characteristics. Furthermore, variance in
covariables thataffectheart rate, suchasemotion,also impact
the ‘heart rate/energy expenditure’ relationship. Hence,
precision of heart rate prediction of energy expenditure is
improvedwhere a separate regression equation is derived to
relate heart rate to energy expenditure for each individual.
Some investigators use multiple regression equations for
each subject. At best, the mean (^95% confidence limits)
error for estimating energy expenditure using heart rate
monitoring is 3 ^ 20% during light activity.
Integrated electromyography
Muscular activity is a component of energy expenditure
and can be measured using integrated electromyography
(EMG). Here, cumulative electrical muscle activity from
several muscle fibres is measured and the data accumulated
over the measurement period. However, strength/force
relationships differ for different muscle groups and fibres,
and multiple muscle groups need to be measured in order
to gain a representative assessment of whole-body activity.
These limitations make this technique impractical46,47.
Pulmonary ventilation volume
Measurement of pulmonary ventilation volume (direct
measurement of the volume of gas exchanged over time)
JA Levine1126
may provide an estimate of energy expenditure but this
technique is impractical for use other than for very short
periods of time48,49.
Thermal imaging
Limited precision and accuracy and the complexity of data
processing complicated the early studies that employed
thermal imaging to detect human heat loss to the
environment. More recent studies have employed auto-
mated, high-resolution, rapid-response thermal imaging50
and offer promise for future studies, particularly in the
area of thermoregulation51.
Physiological observations
Activity recall and time-and-motion studies
Non-specific information about habitual activity and NEAT
can be obtained using questionnaires, interviews or time-
and-motion studies. Predictably, substantial errors are
introduced through inaccurate recall and inadequate data
recording. These approaches can be used, however, for
following trends in certain activities, particularly in
relation to occupational practices52.
Activity logs and the factorial method
This is a frequently used approach for estimating activity
thermogenesis and, in particular, NEAT in free-living
individuals. First, a subject’s physical activities are logged
over the time period of interest (e.g. 1 week). The energy
equivalent of each of these activities is measured or
estimated using a calorimeter or tables18,53,54. The time
spent in each activity is then multiplied by the energy
equivalent for that activity. These values are then summed
to derive an estimate of activity thermogenesis. This
determination of activity thermogenesis is often combined
with information on basal metabolic rate (measured or
calculated) to estimate total daily energy expenditure.
There are two potential sources of error for the factorial
approach for measuring activity thermogenesis. First,
errors may result from inaccurate recording of activities
and, second, from inaccurate determinations of the energy
costs of the activities. To log activity, subjects are often
asked to record in a diary the nature and amount of time
spent performing each of their activities throughout the
day55. This has several limitations: subjects may be
illiterate or innumerate, they may report their activities
inaccurately or incompletely and/or may alter their normal
activity patterns during the period of assessment. To limit
these sources of error, one approach is to have trained
enumerators to follow subjects and objectively record the
subjects’ activities52. This approach is time consuming and
expensive but potentially a valuable source of accurate
and objective data. For this purpose, newer image-
gathering technologies may be useful in future for this
purpose. To determine the energy costs of physical
activities, standard tables are often used. However, these
may introduce substantial (albeit systematic) errors. First,
the tables may not include the precise activity the subjects
perform. Second, the energy cost for a given activity is
highly variable between subjects even independent of
gender. Third, calorimetric methods for measuring the
energy costs of activities have not been standardised
between investigators so that precision and accuracy of
data in the activity tables cannot always be assured. To
limit these errors, the energy costs of each or most of the
activities that the subjects of interest perform can be
measured using calorimeters, as described above. At best,
the energy costs for each subject’s activities would be
measured, but clearly this is rarely practical except for
small studies. In general, population-gender-age specific
group means for the majority of the studied subjects’
activities represents a standard that is worth achieving
where optimum precision is warranted.
Kinematic measurements
In kinematic measurements, a subject’s movements are
quantified and these measurements are usually performed
in conjunction with other measures of energy expenditure.
These tools are used primarily to estimate the energy cost
of NEAT (‘spontaneous physical activity’).
Some techniques are specific for confined spaces such as
radar tracking and cine photography56,57. Other techniques
have been used in free-living individuals and generally
focus on pedometers and accelerometers of varying
sophistication. Pedometers typically detect the displace-
ment of a subject with each stride. However, pedometers
tend to lack sensitivity because they do not quantify stride
length or total body displacement and overall, therefore,
become poor predictors of activity thermogenesis58.
Accelerometers detect body displacement electronically
with varying degrees of sensitivity; uniaxial accelerometers
in one axis and triaxial accelerometers in three axes.
Portable uniaxial accelerometer units have been widely
used to detect physical activity59–61. Careful evaluation
demonstrates that these instruments are not sufficiently
sensitive to quantify the physical activity of a given free-
living subject but rather they are more valuable for
comparing activity levels between groups of subjects.
Greater precision has been obtained using triaxial
accelerometers62–64. In free-living subjects, data from
these devices correlate well with the total daily energy
expenditure, measured using doubly labelled water,
divided by basal metabolic rate65. The utility of motion
tracking using approaches such as Global Positioning
Systems has not been fully defined for human studies.
Practical recommendations for measurement
techniques
The choice of measurement technique is determined by
the objective of the assessment, the resources available
and the ability and willingness of the subjects to partake.
Measurement of energy expenditure 1127
Each investigator needs to define which components of
the energy expenditure are to be measured, in what setting
and with how much resource.
Laboratory setting
To provide precise and accurate short-term (several hours)
measurements of energy expenditure in the laboratory,
well-validated and frequently calibrated indirect or direct
calorimeters should be used. For a relatively modest cost
(US$ 10 000–20 000) a hood/canopy/mask indirect open-
circuit calorimeter canbepurchased. These instruments are
fully automated and simple for a technician with moderate
skill to use, validate and calibrate. The instruments are often
configured as a ‘metabolic cart’ and therefore are
transportable although often not portable. Measurements
to within 1% of chemical standards can be obtained using
these instruments, which can be used for measurements of
basal metabolic rate and resting energy expenditure.
Measurements of the energy expenditure of specific
activities (e.g. hair brushing or walking) and even VO2 max measurements can be made if the calorimeter’s flow
rate is adequate to prevent CO2 accumulation, and if the
accuracy and precision of flow-rate measurements are
maintained at higher oxygen consumptions.
Recently, modestly priced (US$ 1000–20 000), open-
circuit expiratory collection calorimeters have been
produced with adequate precision (,3% error) for basal
metabolic rate, resting energy expenditure measurements
as well as the determination of the energy costs of specific
activities. These instruments may allow widespread
measurements of these variables and their true application
will become apparent over the next several years.
Where investigators wish to make laboratory-based
determinations of the total daily energy expenditure,
greater financial investment is necessary. A room
calorimeter can be built to provide accurate measurements
of energy expenditure over a longer term (1–2 days). It
should be recognised that this is a major undertaking.
Unless the application of the investigators argues against
it, constructing an indirect room calorimeter is simpler and
less costly than a direct calorimeter. Room or chamber
calorimeters require dedicated space, should be equipped
with some means of detecting and quantifying physical
activity within the room or chamber and will necessitate at
least one full-time, highly skilled technician to maintain,
validate, calibrate and use the instrument. Even with
maximum precision (,1% error), it should be noted that
subjects within these chambers are confined and are
unable to perform the activities of daily living.
Regardless of which measurement approach is selected,
it is necessary to adhere to rigorous validation and
calibration protocols. Validation involves burning a
measured mass of a standard of known energy equivalent
within the calorimeter and ascertaining what proportion of
this standard is detected by the calorimeter. Examples of
such standards include ultra-pure butane and ethanol.
These chemical validations should be performed monthly
and have been simplified by the availability of commer-
cially available equipment. Optimum precision for a
calorimeter is to within 3% of predicted and optimum
precision is to within 1% of predicted; this latter goal is
readily achievable with careful attention to technique and
detail.
Calibration should be performed before each measure-
ment and at intervals during the measurement period to
protect against sensor drift. An indirect calorimeter
calibration involves the use of standard gases. It is
recommended that two span gases be used that cover the
spectrum of oxygen and/or carbon dioxide concentrations
that occur during human measurements. With improved
sensor linearity, many systems employ a single-span gas
with 100% nitrogen as the second calibration gas. It is
critical to ensure that the composition of the calibration
gases is guaranteed, that the accepted standard for
calibration gases is to what is termed ‘primary gas
standard’. Flow sensor calibration is more challenging
and some systems allow verification of flow sensor validity
by displacing fixed volumes of gas through the system
using a gas syringe of several litre capacity. Another
approach is to use detachable flow sensors that can be
shipped to the manufacturer for verification of precision.
For direct calorimeters, calibration is performed before
each measurement using a heat emitter of known energy
dissipation. By adopting rigorous standards of validation
and calibration, data will be more reliable and so readily
disseminated and exchanged between laboratories.
Field setting
Calorimeter techniques
There is a paucity of high-quality data pertaining to
activity-related energy expenditure66. Modern tools may
allow this concern to be re-addressed. Several precise,
portable, expiratory collection, open-circuit, indirect
calorimeters have been developed recently (as modifi-
cations of older devices such as the Kofranyi–Michaelis
respirometer19) and these may facilitate field-base
measurements of energy expenditure at rest and with
routine activities. Douglas bags can be used in the field for
measuring basal metabolic rate, resting energy expendi-
ture and, in particular, the energy expenditure of physical
activities. Douglas bags may not necessarily be an
inexpensive option, however. There are the combined
costs of purchasing and maintaining high-quality bags and
high-precision sensors for volume and gas concentration
measurements and, because the technique is highly
operator dependent, skilled technicians are needed.
Experienced technical support, well-maintained Douglas
bags and precise, validated and calibrated instrumentation
are needed to obtain high-quality data using this
approach. Conceptually, metabolic carts can be used in
JA Levine1128
the field but this rarely occurs because of constraints such
as terrain and electricity supply.
Non-calorimetric methods
Field-based measurements of total daily energy expendi-
ture over 7–21 days can be obtained using doubly labelled
water. Administration of the isotopes is straightforward in
that O18 and D2 are weighed, subjects drink the mixture
and collect urine, saliva or blood samples before and after
administration. As discussed above, some investigators
recommend daily sample collection whereas others collect
samples before and after the collection period. Specimens
are readily transportable, they are not radioactive and
stable isotopes do not decay over time. Thus, as long as
the samples are well sealed, measurement of enrichments
can be performed even in another country at any time after
collection. Often, indirect calorimetry is used to measure
BMR in conjunction with the total daily energy expendi-
ture measurements obtained using doubly labelled water.
Activity thermogenesis can thereby be calculated (the
thermic effect of food is generally assumed to equal 10% of
the total daily energy expenditure).
The major advantage of doubly labelled water
measurements is that accurate measurements (error of
,7%) of total daily energy expenditure are obtained in truly free-living individuals. There are important limi-
tations, however. First, no information is obtained
regarding the components of activity thermogenesis.
Second, the thermic effect of food is not measured and
is known to be variable (most believe this to introduce
only a small error). Third, O18 is expensive (,$700/ subject), thereby potentially limiting the number of
subjects that can be studied. Fourth, isotope ratio mass
spectrometers are expensive to purchase and maintain
and skilled staff are needed for their use; hence,
collaboration is encouraged between field investigators
and laboratories where these instruments are in routine
use. It is expected that this approach to field-based
measurement of energy expenditures will become more
widespread despite its costs and limitations.
Logging physical activity and multiplying the nature and
duration of these activities by their metabolic equivalents
have been widely used to estimate activity thermogenesis.
Overall, this approach is potentially valuable, particularly
as the components of activity thermogenesis are detailed.
The precision and accuracy of the approach is highly
variable, however, and depends on how precisely the
subjects’ activities are recorded and how accurately these
are transformed to energy expenditures. It is important to
note that the errors of this approach are additive. Where
maximum precision is needed in small studies, trained
enumerators might be used and measurements are made
of the energy expenditures of typical activities in
representative individuals using calorimeters. Where less
precision is acceptable and where study populations are
larger, activity diaries combined with meaningful (e.g.
gender-specific) tables of energy equivalents for repre-
sentative activities are likely to provide useful group data,
particularly for following population trends.
Techniques have been developed to facilitate more
accurate measurements of body motion and other
components of physical activity in free-living individuals.
There are examples where such measures correlate well
with measured TEE obtained using doubly labelled
water65.
Overall, progress is being made to obtain data on energy
expenditure in free-living individuals. It is expected that
these technological advances can be exploited to provide
improved measures of energy expenditure in the field.
Standardisation of protocols
Once equipment is available to perform measurements of
energy expenditure, adhering to standardised protocols
has several advantages. First, it allows comparisons to be
made between different laboratories. Second, it enables
databases to be generated to explore the variance in the
components of energy expenditure and better characterise
the energy expenditure of physical activities and free-living
physical activity in different individuals and populations.
Basal metabolic rate
BMR should be measured between 06.00 and 09.00 hours
in individuals who slept at the site of measurement
overnight. The individuals should not have consumed
food or energy-containing beverage for 9 hours prior to
the measurement but may have consumed water. The
measurement should be performed with the patient
supine. A single pillow may support the subject’s head
and/or the head of the bed should be at a 108 vertical tilt.
The subject should be in thermal comfort and the room
should not be brightly lit. Subjects should be instructed to
lie motionless and should not be allowed to talk or have
other potentially stimulating distractions during the
measurement. The measurement period should last for
20–40 minutes.
Resting energy expenditure
Resting energy expenditure should be performed in the
post prandial state, at least 6 hours after consumption of
any calories or performing any rigorous activity. Subjects
should be fully rested while supine for 60 minutes prior to
the measurement. The measurement is otherwise as
described for BMR.
Thermic effect of food
Optimally, a measurement of BMR should be performed
first, then subjects should be provided with a meal of food.
The energy content of the food should be known precisely
and should be of 400 kcal or greater. Energy expenditure
should then be measured for 400 minutes or until energy
expenditure falls to within 5% of the BMR. For those using
Measurement of energy expenditure 1129
hood-based systems (with response time ,2 minutes),
energy expenditure can be measured for every 15 minutes
out of 30 minutes to avoid subject agitation. The thermic
effect of food for the meals provided is calculated from the
area under the energy expenditure above basal metabolic
rate versus the time curve. Some would argue that it is of
value to also measure the thermic effect of non-caloric
meals.
Energy expenditure of physical activities
Points of reference are important. Resting energy
expenditure should be measured first, then the energy
expenditure of the posture of reference should be
measured while the subject is motionless. For example,
for measuring the energy expenditure of secretarial work,
sitting energy expenditure should be measured as the
point of reference. For measuring the energy expenditure
of scything, standing energy expenditure should be
measured as the point of reference. Measurement of
energy expenditure during the performance of the activity
of interest should be performed for 10–20 minutes if the
calorimeter has a response time of ,2 minutes. Where
calorimeter response times are longer, the measurement
period needs to be prolonged so that steady-state energy
expenditure is reached. The energy expenditure for the
activity can be calculated as the steady-state energy
expenditure for that activity minus (or divided by) either
the energy expenditure of the posture of reference or the
resting energy expenditure.
Areas for future research
There are a number of areas for future research with
respect to energy expenditure measurements.
. Integration of data on energy expenditure. With standardisation of technical standards and techniques,
it would be advantageous to develop international
databases on BMR, resting energy expenditure, energy
expenditure of specific activities and total daily energy
expenditure measured using doubly labelled water.
This would facilitate validation of energy requirement
recommendations and allow cross-cultural investigation
into metabolic rate variance.
. Physical activity in free-living individuals. Technology is fast improving to provide detailed and accurate
information on NEAT and the types and quantities of
physical activities that individuals perform. It is
recommended to readily employ novel and advanced
technologies once they have been validated in
representative populations.
. Doubly labelled water collaborative agreements. It is recommended to establish a collaborative environment
for the analysis of doubly labelled water determinations.
It is proposed to identify laboratories where isotope
enrichments can be analysed so that studies can be
performed by other investigators lacking necessary
instruments. This would facilitate studies of TEE in
under-developed countries.
. Non-calorimeter methods. The role of newer technol- ogies such as thermal imaging or global positioning
remains to be determined but should be explored.
Acknowledgements
Funded by NIH DK56650, DK63226 and DK66270.
References
1 Cunningham JJ. Calculation of energy expenditure from indirect calorimetry: assessment of the Weir equation. Nutrition 1990; 6(3): 222–3.
2 Weir JB. New methods for calculating metabolic rate with special reference to protein metabolism. Nutrition 1949; 6(3): 213–21.
3 Tissot J. Nouvelle methode de mesure et d’inscription du debit et des movements respiratories de l’homme et des animaux. Journal of Physiology and Pathology General 1904; 6: 688–700.
4 Lum L, Saville A, Venkataraman ST. Accuracy of physiologic deadspace measurement in intubated pediatric patients using a metabolic monitor: comparison with the Douglas bag method. Critical Care Medicine 1998; 26(4): 760–4.
5 de Groot G, Schreurs AW, van Ingen Schenau G. A portable lightweight Douglas bag instrument for use during various types of exercise. International Journal of Sports Medicine 1983; 4(2): 132–4.
6 Yoshida T, Nagata A, Muro M, Takeuchi N, Suda Y. The validity of anaerobic threshold determination by a Douglas bag method compared with arterial blood lactate concen- tration. European Journal of Applied Physiology and Occupational Physiology 1981; 46(4): 423–30.
7 Douglas CG. A method for determining the total respiratory exchange in man. The Journal of Physiology 1911; 42: 17–18.
8 Daniels J. Portable respiratory gas collection equipment. Journal of Applied Physiology 1971; 31(1): 164–7.
9 Levine JA, Schleusner SJ, Jensen MD. Energy expenditure of non-exercise activity. American Journal of Clinical Nutri- tion 2000; 72(6): 1451–4.
10 Sorkin B, Rapoport DM, Falk DB, Goldring RM. Canopy ventilation monitor for quantitative measurement of ventilation during sleep. Journal of Applied Physiology 1980; 48(4): 724–30.
11 Kinney JM. Indirect calorimetry in malnutrition: nutritional assessment or therapeutic reference? Journal of Parenteral and Enteral Nutrition 1987; 11(Suppl 5): S90–S4.
12 Weissman C, Damask MC, Askanazi J, Rosenbaum SH, Kinney JM. Evaluation of a non-invasive method for the measurement of metabolic rate in humans. Clinical Science 1985; 69(2): 135–41.
13 Wilmore JH, Davis JA, Norton AC. An automated system for assessing metabolic and respiratory function during exer- cise. Journal of Applied Physiology 1976; 40(4): 619–24.
14 Sun M, Hill JO. A method for measuring mechanical work and work efficiency during human activities. Journal of Biomechanics 1993; 26(3): 229–41.
15 Sun M, Reed GW, Hill JO. Modification of a whole room indirect calorimeter for measurement of rapid changes in
JA Levine1130
energy expenditure. Journal of Applied Physiology 1994; 76(6): 2686–91.
16 Sestoft L. Metabolic aspects of the calorigenic effect of thyroid hormone in mammals. Clinical Endocrinology 1980; 13(5): 489–506.
17 Dallosso HM, James WP. Whole-body calorimetry studies in adult men. The interaction of exercise and over-feeding on the thermic effect of a meal. British Journal of Nutrition 1984; 52(1): 65–72.
18 Sujatha T, Shatrugna V, Venkataramana Y, Begum N. Energy expenditure on household, childcare and occupational activities of women from urban poor households. British Journal of Nutrition 2000; 83(5): 497–503.
19 Consolazio CF. Energy expenditure studies in military populations using Kofranyi-Michaelis respirometers. Amer- ican Journal of Clinical Nutrition 1971; 24(12): 1431–7.
20 Rietjens GJ, Kuipers H, Kester AD, Keizer HA. Validation of a computerized metabolic measurement system (Oxycon-Pro) during low and high intensity exercise. International Journal of Sports Medicine 2001; 22(4): 291–4.
21 McLaughlin JE, King GA, Howley ET, Bassett DR Jr, Ainsworth BE. Validation of the COSMED K4 b2 portable metabolic system. International Journal of Sports Medicine 2001; 22(4): 280–4.
22 Blaxter KL, Brockway JM, Boyne AW. A new method for estimating the heat production of animals. Quarterly Journal of Experimental Physiology and Cognate Medical Sciences 1972; 57(1): 60–72.
23 Aulick LH, Arnhold H, Hander EH, Mason AD Jr. A new open and closed respiration chamber. Quarterly Journal of Experimental Physiology 1983; 68(3): 351–7.
24 Benedict FG. An apparatus for studying the respiratory exchange. The American Journal of Physiology 1909; 24: 345–74.
25 Spinnler G, Jequier E, Favre R, Dolivo M, Vannotti A. Human calorimeter with a new type of gradient layer. Journal of Applied Physiology 1973; 35(1): 158–65.
26 Webster JD, Welsh G, Pacy P, Garrow JS. Description of a human direct calorimeter, with a note on the energy cost of clerical work. British Journal of Nutrition 1986; 55(1): 1–6.
27 Webb P, Annis JF, Troutman SJ Jr. Energy balance in man measured by direct and indirect calorimetry. American Journal of Clinical Nutrition 1980; 33(6): 1287–98.
28 Tschegg E, Sigmund A, Veitl V, Schmid P, Irsigler K. An isothermic, gradient-free, whole-body calorimeter for long- term investigations of energy balance in man. Metabolism 1979; 28(7): 764–70.
29 Snellen JW. An improved estimation of mean body temperature using combined direct calorimetry and thermo- metry. European Journal of Applied Physiology 2000; 82(3): 188–96.
30 Snellen JW, Chang KS, Smith W. Technical description and performance characteristics of a human whole-body calorimeter. Medical and Biological Engineering and Computing 1983; 21(1): 9–20.
31 Black AE, Coward WA, Cole TJ, Prentice AM. Human energy expenditure in affluent societies: an analysis of 574 doubly- labelled water measurements. European Journal of Clinical Nutrition 1996; 50(2): 72–92.
32 Coward WA, Roberts SB, Cole TJ. Theoretical and practical considerations in the doubly labelled water (2H2(18)O) method for the measurement of carbon dioxide production rate in man. European Journal of Clinical Nutrition 1988; 42(3): 207–12.
33 Coward WA. Contributions of the doubly labelled water method to studies of energy balance in the Third World. American Journal of Clinical Nutrition 1998;68(4): S962–S9.
34 Goran MI, Poehlman ET, Nair KS, Danforth E. Deuterium exchange in humans: effect of gender, body composition and age. Basic Life Sciences 1993; 60(3): 79–81.
35 Kurpad AV, Borgonha S, Shetty PS. Measurement of total energy expenditure by the doubly labelled water technique in free living Indians in Bangalore city. The Indian Journal of Medical Research 1997; 105: 212–19.
36 Schoeller DA, Taylor PB. Precision of the doubly labelled water method using the two-point calculation. Human Nutrition: Clinical Nutrition 1987; 41(3): 215–23.
37 Dauncey MJ, James WP. Assessment of the heart-rate method for determining energy expenditure in man, using a whole- body calorimeter. British Journal of Nutrition 1979; 42(1): 1–13.
38 Baker JA, Humphrey SJ, Wolff HS. Socially acceptable monitoring instruments (SAMI). The Journal of Physiology 1967; 188(2): 4P–5P.
39 Bruce FM, Floyd WF, Ward JS. Oxygen consumption and heart rate during stair climbing. The Journal of Physiology 1967; 191(2): 90P–2P.
40 Rennie KL, Hennings SJ, Mitchell J, Wareham NJ. Estimating energy expenditure by heart-rate monitoring without individual calibration. Medicine and Science in Sports and Exercise 2001; 33(6): 939–45.
41 Beghin L, Budniok T, Vaksman G, Boussard-Delbecque L, Michaud L, Turck D, Gottrand F. Simplification of the method of assessing daily and nightly energy expenditure in children, using heart rate monitoring calibrated against open-circuit indirect calorimetry. Clinical Nutrition 2000; 19(6): 425–35.
42 Kashiwazaki H. Heart rate monitoring as a field method for estimating energy expenditure as evaluated by the doubly labelled water method. Journal of Nutritional Science and Vitaminology 1999; 45(1): 79–94.
43 Maffeis C, Pinelli L, Zaffanello M, Schena F, Iacumin P, Schutz Y. Daily energy expenditure in free-living conditions in obese and non-obese children: comparison of doubly labelled water (2H2(18)O) method and heart-rate monitor- ing. International Journal of Obesity and Related Metabolic Disorders: Journal of the International Association for the Study of Obesity 1995; 19(9): 671–7.
44 Ceesay SM, Prentice AM, Day KC, Murgatroyd PR, Goldberg GR, Scott W, Spurr GB. The use of heart rate monitoring in the estimation of energy expenditure: a validation study using indirect whole-body calorimetry. British Journal of Nutrition 1989; 61(2): 175–86.
45 Kalkwarf HJ, Haas JD, Belko AZ, Roach RC, Roe DA. Accuracy of heart-rate monitoring and activity diaries for estimating energy expenditure. American Journal of Clinical Nutrition 1989; 49(1): 37–43.
46 Seliger V, Dolejs L, Karas V. A dynamometric comparison of maximum eccentric, concentric, and isometric contractions using emg and energy expenditure measurements. Euro- pean Journal of Applied Physiology and Occupational Physiology 1980; 45(2–3): 235–44.
47 deVries HA, Burke RK, Hopper RT, Sloan JH. Relationship of resting EMG level to total body metabolism with reference to the origin of ‘tissue noise’. American Journal of Physical Medicine 1976; 55(3): 139–47.
48 Young BA, Fenton TW, McLean JA. Calibration methods in respiratory calorimetry. Journal of Applied Physiology 1984; 56(4): 1120–5.
49 Malhotra MS, Ramaswamy SS, Joseph NT, Sen Gupta J. Functional capacity and body composition of different classes of Indian athletes. Indian Journal of Physiology and Pharmacology 1972; 16(4): 301–8.
50 Levine JA, Pavlidis I, Cooper M. The face of fear. Lancet 2001; 357: 1757.
51 Shuran M, Nelson RA. Quantitation of energy expenditure by infrared thermography. American Journal of Clinical Nutrition 1991; 53(6): 1361–7.
52 University UN. Research Methods in Nutritional Anthropol- ogy, 1989
Measurement of energy expenditure 1131
53 Morio B, Beaufrere B, Montaurier C, Verdier E, Ritz P, Fellmann N, Boirie Y, Vermorel M. Gender differences in energy expended during activities and in daily energy expenditure of elderly people. American Journal of Physiology Endocrinology and Metabolism 1997; 273: E321–7.
54 Banerjee B, Khew KS, Saha N. A comparative study of energy expenditure in some common daily activities of non- pregnant and pregnant Chinese, Malay and Indian women. The Journal of Obstetrics and Gynaecology of the British Commonwealth 1971; 78(2): 113–16.
55 Ferro-Luzzi A, Scaccini C, Taffese S, Aberra B, Demeke T. Seasonal energy deficiency in Ethiopian rural women. European Journal of Clinical Nutrition 1990; 44(Suppl. 1): 7–18.
56 Schutz Y, Ravussin E, Diethelm R, Jequier E. Spontaneous physical activity measured by radar in obese and control subject studied in a respiration chamber. International Journal of Obesity 1982; 6(1): 23–8.
57 Mayer J. Physical activity and anthropometric measurements of obese adolescents. Federation Proceedings 1966; 25(1): 11–14.
58 Gretebeck RJ, Montoye HJ. Variability of some objective measures of physical activity. Medicine and Science in Sports and Exercise 1992; 24(10): 1167–72.
59 Bassett DR Jr, Ainsworth BE, Swartz AM, Strath SJ, O’Brien WL, King GA. Validity of four motion sensors in measuring moderate intensity physical activity. Medicine and Science in Sports and Exercise 2000; 32(Suppl. 9): S471–80.
60 Melanson EL Jr, Freedson PS. Validity of the computer science and applications, Inc. (CSA) activity monitor. Medicine and Science in Sports and Exercise 1995; 27(6): 934–40.
61 Pambianco G, Wing RR, Robertson R. Accuracy and reliability of the Caltrac accelerometer for estimating energy expenditure. Medicine and Science in Sports and Exercise 1990; 22(6): 858–62.
62 Bouten CV, Westerterp KR, Verduin M, Janssen JD. Assessment of energy expenditure for physical activity using a triaxial accelerometer. Medicine and Science in Sports and Exercise 1994; 26(12): 1516–23.
63 Westerterp KR, Bouten CV. Physical activity assessment: comparison between movement registration and doubly labelled water method. Zeitschrift fur Ernahrungswis- senschaft 1997; 36(4): 263–7.
64 Levine JA, Baukol PA, Westerterp KR. Validation of the Tracmor triaxial accelerometer system for walking. Medicine and Science in Sports and Exercise 2001; 33(9): 1593–7.
65 Bouten CV, Verboeket-van de Venne WP, Westerterp KR, Verduin M, Janssen JD. Daily physical activity assessment: comparison between movement registration and doubly labelled water. Journal of Applied Physiology 1996; 81(2): 1019–26.
66 FAO/WHO. Energy and Protein Requirements. Report of a Joint FAO/WHO Ad Hoc Expert Committee. FAO Nutrition meetings Report Series, No. 52, Rome: FAO, 1973; WHO Technical Report Series, No. 522, Geneva: WHO, 1973.
JA Levine1132
American Society for nutrition 2012 energy balance(1)(1).pdf
CONSENSUS STATEMENT
Energy balance and its components: implications for body weight regulation1–3
Kevin D. Hall,4 Steven B. Heymsfield,5 Joseph W. Kemnitz,6 Samuel Klein,7 Dale A. Schoeller,8 and John R. Speakman9*
4National Institute of Diabetes and Digestive and Kidney Diseases, NIH, Bethesda, MD; 5Pennington Biomedical Research Center, Baton Rouge, LA; 6Institute for Clinical and
Translational Research, University of Wisconsin, Madison, WI; 7Washington University School of Medicine, St Louis, MO; 8Nutritional Sciences, University of Wisconsin,
Madison, WI; and 9Institute of Biological and Environmental Sciences, University of Aberdeen, Aberdeen, United Kingdom
A fundamental principle of nutrition and metabolism is that body weight change is associated with an imbalance between the energy content of food eaten and energy expended by the body to maintain life and to perform physical work. Such an energy balance framework is a potentially powerful tool for investigating the regulation of body weight. However, we need a better un- derstanding of the components of energy balance and their in- teractions over various time scales to explain the natural history of conditions such as obesity and to estimate the magnitude and potential success of therapeutic interventions. Therefore, the ASN and the International Life Sciences Institute convened a panel composed of members with expertise in weight manage- ment, energy metabolism, physical activity, and behavior to review the published scientific literature and to hear presentations from other experts in these fields. The Consensus Panel met 9–12 May 2011 in Chicago, IL, and was charged to provide answers to the following 5 questions:
1. Explain energy balance and imbalance in terms of a biolog- ical system in which energy intake and energy expenditure change over time in response to the environment.
2. What are the interactions between the components of en- ergy balance and how are they regulated?
3. What is the veracity of some of the popular beliefs related to energy balance?
4. What limitations do we face in the study of energy balance and its components?
5. What research would better inform our knowledge of en- ergy balance and its components?
Question 1: Explain energy balance and imbalance in terms of a biological system in which energy intake and energy expenditure change over time in response to the environment
Human physiology complies with the first law of thermody- namics, which states that energy can be transformed from one form to another but cannot be created or destroyed. This law is usually formulated as follows: the rate of change in body ES
10 is equal to the difference between the rates of EI and EO. All of these terms are expressed as energy per unit of time.
EI primarily consists of the chemical energy from the food and fluids we consume. EO includes the radiant, conductive, and convective heat lost; any work performed; and the latent heat of
evaporation. ES is the rate of change in the body’s macronutrient stores. The energy balance equation (ES = EI – EO) is a statement of the principle of energy conservation.
Components of intake
Energy intake includes 3 major macronutrient groups— carbohydrate, protein, and fat—and a smaller component from alcohol. Once ingested, the net absorption of the major macronu- trient groups is variable and incomplete, with fecal losses accounting for ;2–10% of gross EI. The net absorption of dietary energy components varies among individuals and is dependent on the specific foods eaten, how they are prepared, and intestinal factors.
The metabolizable energy (hereafter referred to as EI) of a diet represents the difference between the absolute energy of ingested substrates and the energy losses found in feces and urine. Commonly used energy densities for carbohydrate (4 kcal/g, 17 kJ/g), protein (4 kcal/g, 17 kJ/g), and fat (9 kcal/g, 38 kJ/g) represent population
1 A Consensus Conference entitled “The Role of Energy Balance in
Health and Wellness” was organized and funded by the ASN and the North
American branch of the International Life Sciences Institute (ILSI North
America). The ASN is funded through publication subscriptions, member
dues and contributions, meetings, and foundation and corporate grants. ILSI
North America is funded primarily by the dues of its food industry members.
An independent panel, which consisted of the 6 authors, prepared the resulting
Consensus Statement. This statement represents the panel’s collective analysis,
evaluation, and opinion at the time of publication and does not represent the
official position of the ASN or ILSI North America. The statement was re-
viewed and approved by the ASN’s Reviews, Papers, and Guidelines Commit-
tee and Board of Directors and reviewed by the ILSI North America Energy
Balance and Active Lifestyle Committee. The statement did not undergo edi-
torial peer review by the editors of The American Journal of Clinical Nutrition. 2 This work was supported in part by an educational grant from ILSI North
America; administrative support was provided by the ASN. This research was
supported in part by the Intramural Research Program of the NIH, National
Institute of Diabetes and Digestive and Kidney Diseases (KDH). 3 None of the authors declared any conflicts of interest in relation to the
content of this article. 10 Abbreviations used: AEE, activity energy expenditure; EI, energy in-
take; EO, energy output; ES, energy storage; REE, resting energy expendi-
ture; TEF, thermic effect of food.
*Address correspondence to JR Speakman, Institute of Biological and
Environmental Sciences, University of Aberdeen, Aberdeen AB24 2TZ,
United Kingdom. E-mail: [email protected].
doi: 10.3945/ajcn.112.036350
Am J Clin Nutr 2012;95:989–94. Printed in USA. � 2012 American Society for Nutrition 989
at A M
E R
IC A
N S
O C
IE T
Y F
O R
N U
T R
IT IO
N on A
pril 2, 2012 w
w w
.ajcn.org D
ow nloaded from
averages for metabolizable energy, which is the amount of fuel ac- tually available to cells for conducting biological processes.
Digestibility depends on the composition of the food item and on its content of fiber and other indigestible components. Such components canmechanically limit the access of digestive enzymes to food that would potentially be digestible. For example, nuts and other plant materials have cell walls that cannot be digested by gut enzymes, and they thereby protect the cell contents from digestion if not masticated sufficiently to disrupt the cell structure. These effects can have a large impact on the absorption of ingested macronutrients. The variability in absorptive efficiency depends on many additional factors (eg, gut flora, food preparation, diet composition), which may explain the individual differences in metabolizable EI.
Components of expenditure (EO)
Absorbed carbohydrates, proteins, and fats are transformed in vivo to substrates that can ultimately either be oxidized to produce metabolically useful energy that drives biological processes or they may be stored. The rate of whole-body energy expenditure, or EO, varies within a 24-h period and across the life span. Expended en- ergy reflects fuels metabolized for growth, body maintenance needs, physical activity, pregnancy and lactation, and many other processes.
The main energy expenditure terms are REE, TEF, and AEE. REE is the rate of energy expenditure at rest and comprises ap- proximately two-thirds of EO. REE varies between and within individuals depending on body size, body composition, and recent energy imbalance. Greater total tissue mass increases REE, and the contribution of lean tissue is greater than fat tissue. Moreover, within lean tissue, high metabolic organs such as the brain, heart, kidney, and liver contribute disproportionately to REE. There is also a large variability in REE (;250 kcal/d, ;1000 kJ/d) that is not explained by differences in body composition (1).
The TEF is the obligatory energy expenditure that is associated with digestion and processing of ingested foods. Diet composition has a strong effect on TEF. There is a hierarchy of macronutrient effects on the magnitude of TEF, with isocaloric amounts of protein . carbohydrate . fat. Normally, TEF is assumed to be a fixed percentage of EI, but variation between and within individuals occurs. AEE is the energy expenditure rate during activity and can be further partitioned into exercise energy expenditure and non- exercise activity thermogenesis.
Components of storage
Triglycerides, which are present within adipose tissue, are the body’s major fuel reserve. A lean adult has ;35 billion adipo- cytes, each containing ;0.4–0.6 lg triglyceride and totaling 130,000 kcal stored energy. An extremely obese adult can have 4 times as many adipocytes (140 billion), each containing twice as much lipid (0.8–1.2 lg triglyceride) and totaling ;1 million kcal stored energy (2).
ES reflects net changes in the body mass of carbohydrate, pro- tein, and fat. Carbohydrate is stored mainly in the form of in- tracellular glycogen in skeletal muscle and liver. The total mass of glycogen is relatively small, several hundred grams, and turnover is rapid; maximal amounts are observed in the postmeal state. Water is weakly bonded to glycogen so that glycogen’s synthesis and catabolism also involve alterations in fluid balance. Body protein
takes many specific forms and, as with glycogen, is associated with water but at a lower value per gram. Lipid in the form of tri- glyceride is the largest source of stored energy in most adults and has no water associated with it.
Any imbalance between the intake and utilization of these mac- ronutrients will lead to an alteration in body composition. The energy stored per unit body weight of carbohydrate, fat, and protein varies considerably, especially when accounting for the associated in- tracellular water. Furthermore, dietary carbohydrate intake has an impact on renal sodium excretion, which results in changes in ex- tracellular fluid. Therefore, changes in body weight are expected when the macronutrient composition of the diet is altered, even when the energy content of the diet is held constant.
The long-term stability of body weight is often considered a marker of zero ES, and thus energy balance. However, as described above, changes in body weight also include changes in body water, which may be variable, and therefore weight change may not di- rectly represent energy imbalances, particularly over the short term.
Question 2: What are the interactions between the components of energy balance and how are they regulated?
The 3 main terms of the energy balance equation continuously change over time. Beginning at conception, ES remains positive, on average, throughout growth and development. This positive energy imbalance is reflected by increasing body weight. If adult weight is then maintained over the long term, average ES approaches zero, and an approximate average state of energy balance is present. However, most adults gain fat throughout their lives and in later life lose skeletal muscle; the energy content of body fat change is much higher than that of lean tissue change. Thus, even with weight stability, “perfect” energy balance over the long term does not occur in most older adults.
Over a 24-h period, a typical person eats several meals during the day, and energy balance is strongly positive during and soon after eachmeal. Energy output is continuous but with increases due to episodic physical activity and reduction during sleep. Energy balance is thus highly variable over a 1-d period, and this vari- ability is shown in dynamic changes in ES. Most adults also vary their daily eating and activity patterns; thus, ES also varies from day to day, with energy balance achieved only when averaged over longer time periods.
The development of obesity by necessity requires positive energy imbalance over and above that required for normal growth and development. As in lean individuals, a state of energy balance over the long term with similar short-term fluctuations in intake and expenditure is also approximated in obese individuals, but in obese individuals this is achieved with a higher amount of body fat.
The counterpart of excess weight gain is a negative energy balance leading to weight loss over time. For example, if an acute reduction in EI is maintained over time, then, assuming patterns of behavior remain unchanged, changes in the 3 processes— reduced REE, AEE, and TEF—will gradually also lower EO as weight is lost. Eventually, these passive compensatory effects will lead to a diminishing energy imbalance with ultimate res- toration of a steady state at a lower body weight.
Although it is clear that EI and EO are part of a biologically regulated system, the exact nature of how this system works in humans has not been fully established. Two different system designs have been generally discussed, a “set point” and a “settling point.”
990 HALL ET AL
at A M
E R
IC A
N S
O C
IE T
Y F
O R
N U
T R
IT IO
N on A
pril 2, 2012 w
w w
.ajcn.org D
ow nloaded from
The idea of a set point is borrowed from the field of engineering in which feedback control systems are designed to regulate a particular variable to match a specified target. In contrast, a settling point has traditionally been used to describe a system without active feedback control of food intake and energy expenditure. Models that do not directly specify a set-point value but that include active feedback control have also been called settling-point models. These 2 systems do, in fact, overlap, and there are insufficient data to decide whether one or both are valid. What is clear, however, is that perturbations in the components of energy intake or expenditure result in compen- satory changes in these components. These include passive com- pensatory changes such as an increase in energy expenditure with an increase in body size and active compensation such as changes in food intake after exercise.
The following is a brief review of the interactions among the energy balance components.
Food intake on subsequent food intake
Food intake is temporally variable. We eat meals that reflect the satiation that develops during a meal and satiety between meals. The energy content of a given meal is highly variable between individuals and highly variable between meals in an individual. However, the variation in total caloric intake summed across all meals over a day is far less variable. This suggests that there ismeal- to-meal compensation of intake, which is confirmed by a negative correlation between successive meal energy content. If we over- or underconsume energy in one meal, we partially compensate for that intake in subsequent meals during the same day. In addition to variation in intake between meals on a given day, we also vary the amount of food eaten each day. Energy expenditure rarely shows the same degree of variation across days. Hence, we are almost per- petually in energy imbalance on the time scale of hours or days. When a given day’s intake and expenditure are plotted against each other, there is little association. It is only when they are averaged over much longer periods (weeks) that there begins to be a balance struck between intake and expenditure (3). The panel emphasized that this is a key point that is sometimes overlooked: energy balance as a concept depends on the time domain over which it is con- sidered. We are always in energy imbalance, but the relative im- balance is greater over the short term than over the long term.
Food composition has been suggested to have a large impact on satiety and satiation. It is generally believed that the major macro- nutrientsdiffer intheireffects,withproteinhavingagreatereffect than carbohydrate,whichhasagreatereffectthanfat.However, thedataare not consistent among all studies. In addition, many environmental factors suchas social context, aswell as likingandwanting food, play an important role in the energy consumed at a meal.
Satiety and satiation depend on several physiologic and mo- lecular mechanisms. Satiation mechanisms include distension of the gastrointestinal tract communicated to the brain and the se- cretion of a number of gut peptides that interact with receptors principally in thehind-brain.Afactorpotentially linked tosatiety is the hormone ghrelin, which is produced by the stomach. Ghrelin is unique among known gut peptides in that it is orexegenic. Its production increaseswith timesince the lastmeal, and injectionsof ghrelin promote food intake. The hormonal regulation of food intake has been discussed in greater detail elsewhere (4).
In addition, there are a large number of sensory and cognitive stimuli that affect food intake and physiology. For example,
liking and wanting food can overcome feelings of satiation and satiety and lead to food intake despite feeling full or not being hungry. Also, sensory-specific satiation can affect food intake— although people may feel full after a large main course of savory food, they are still able to eat a sweet dessert.
Food intake on energy expenditure
After the overconsumption of energy there is an increase in body size leading to a passive increase in EO. This is due to the following factors: an increase in REE, mainly as a result of an increase in lean tissuemassand toa lesser extent an increase in fatmass; an increase in AEEassociatedwith the increasedcost ofmovinga largerbodymass; and an increased TEF due to greater EI. Finally, there is an additional energy cost for tissue deposition and increased protein turnover.
There has been a long-standing debate about whether, in addition to these passive effects on EO, there is an active stimulation of expenditure during overfeeding that opposes weight gain; how- ever, there is little evidence for an active effect on REE during overfeeding when one accounts for the additional energy cost of tissue deposition. It has also been suggested that nonexercise ac- tivity thermogenesis may increase to partially offset the effect of overfeeding (5). This effect was reported to be �500 kcal/d (2100 kJ/d), which would be a major compensatory factor for opposing weight gain when caloric consumption is increased, but other studies have failed to find effects of a similar magnitude (6, 7).
During restriction of food intake there is a reduction in whole- body EO, due in part to the reduction in body mass that follows the lowered calorie intake. This can be accounted for by reductions in REE secondary to loss of lean and fat mass, reductions in AEE due to reduced amounts and costs of activity, a decrease in TEF due to lower EI, mostly caused by reductions in protein turnover and its associated energy cost. In addition to the passive compensation described above, there is evidence for an active reduction in REE during calorie restriction whose magnitude is dependent on the degree of calorie restriction (8).
Manystudieshaveaddressed theeffectofmealpatterningonREE during weight stability. On average, almost tripling the number of daily meals but providing the same total amount of energy had a barely detectable effect on REE, which suggests that meal patterning does not elicit a greater or lower compensation in expenditure.
The effect of exercise on EI
If demands for energy are met from food intake then it is often assumed that there must be some mechanism that provides a link between expenditure and intake. However, studies of short duration in which EO is increased by exercise showed no compensatory change in EI over 1 or 2 d. As the duration of the studies increased, evidence for compensation emerged with longer-duration studies showing greater but incomplete compensation.
Data from several studies showed no relation between AEE and subsequent weight change. Therefore, low AEE as measured by doubly labeled water at a single time point was not a predictor of weight gain over a protracted period (9–11). Cross-sectional data on AEE that span the recent increase in the prevalence of obesity showed that during this long period of time, levels of AEE have not declined (12). However, recent modeling work has suggested that declines in occupational activity over the past 5 decades could explain the observed increases in body weight over
ENERGY BALANCE AND ITS COMPONENTS 991
at A M
E R
IC A
N S
O C
IE T
Y F
O R
N U
T R
IT IO
N on A
pril 2, 2012 w
w w
.ajcn.org D
ow nloaded from
the same period (13) but only if such activity changes were not compensated for by nonoccupational activity changes in or modulations of food intake.
Exercise interventionsresult ingreat individualvariation inbody weight response. Part of the variation may be due to adherence. However, even when exercise sessions are closely supervised, and hence the adherence issue is eliminated, there is still a tremendous variation in response, with some individuals losing significant amounts of weight and some actually gaining weight (14). Meas- urements of food intakebefore andafter exercise suggest part of the variability in weight change due to exercise lies in how completely individualscompensatefor theirexerciseprescriptionwithelevated food intake, which corresponds to their hunger after exercise.
The effect of exercise on EO
Apopular ideais thatamajorbenefitofphysicalactivitycomesnot onlyfromtheactualenergy that isexpendedduring theexercise itself but also from an after-effect of physical activity on REE. There are data showing a positive effect of vigorous or moderate physical activity on REE. This follows 2 separate phases: a large effect that lasts;2h and a smaller butmoreprolongedeffect that could take up to 48 h to return to baseline (15). This is called excess postexercise oxygen consumption and accounts for ;6–15% of the energy ex- pended during an exercise session (16), which adds little to TEE.
Another popular belief is that exercise training results in body- composition changes that generate an additional energy benefit of exercise mediated through REE. But such potential effects of exercise training on REE may have been confounded because the post–exercise training REE was measured too soon after the final exercise bout, contaminating it due to excess postexercise oxygen consumption (15). Measurements that are not so confounded suggest that the impact of exercise training on REE is negligible. Whether habitual exercise produces long-term changes in other components of EO is unclear.
Exercise interventions may be counteracted by compensatory reductions in physical activity at other times of the day, although the dataon thispoint aremixed.Somestudies found that exercisehadno overall effect on daily EO because the individuals reduced their normal activities. Other studies reported that there was no activity compensation from the additionof an exercise intervention and thus an increase in EO was observed. Indeed, in some studies there was an increase in EO beyond that accounted for by the exercise alone.
These data emphasize a major point that we would like to reinforce. All of the components of energy balance interact with each other. Consequently, it is absolutely necessary to take all of these interactions into consideration when conducting in- tervention research in the field of obesity. To take a simple ex- ample, it may not be very useful to enhance physical activity but to allow subjects to eat what they wish (and thus compensate for their elevated expenditure).
Question 3: What is the veracity of some of the popular beliefs related to energy balance?
A. “The typically observed weight-loss plateau at 6 to 8 mo after a weight-loss intervention is primarily due to a reduction in energy expenditure, ie, slowed metabolism.”
Although the measurement of EO at the plateau is decreased, it does not decrease to the amount of the prescribed or self-reported
energy intake. Thus, the plateau may well be attributed to failure to comply with the diet (17). Modeling studies support this in- terpretation and suggest that if subjects had complied with the prescribed diet, the plateau due to metabolic change would not have occurred for several years, which would have led to much greater weight loss than that observed (18). These data also em- phasize that, whereas it is possible to cognitively intervene in our food intake amounts, such interventions are extremely difficult to sustain because of the biological and psychological drives to eat.
B. “Obesity is due to low energy expenditure, ie, low metabolism.”
The existence of a lowmetabolic rate in obesity was erroneously reported in early studies in which the REE was inappropriately normalized by dividing it by body weight. A simple division of REE by total weight leads to a lower estimate of the mass-specific metabolic rate because obese people have an increased relative amount of body fat, which has a lower metabolic rate than does lean tissue. This normalization error led to the notion that low metabolism was the cause of the obesity. The error was com- pounded by a misuse of the energy balance concept, which is properly applied only at the level of the entire organism. Thus, it is invalid to consider metabolism per kilogram of body weight, or even per kilogram of fat-free mass, as a component of this system. A balance is not struck between total food intake per individual and expenditure per kilogram but rather between energy intake per individual and energy expenditure per individual. Lower REE per kilogram of body weight therefore cannot be a “cause” of obesity.
In absolute terms, obese people expend more energy than do their lean counterparts. However, this observation should not be overinterpreted to infer that lowREE is not a risk factor for obesity. This is because obese peoplemight have had a lower REE than that predicted for their body size and composition before gaining their excess weight. Therefore, it is unclear the extent to which obesity results from reduced energy expenditure, but it is clear that the maintenance of obesity is not due to reduced energy expenditure.
C. “It takes a reduction of 3500 kcal (15,000 kJ) of energy intake to lose 1 lb of body weight.”
The origin of the “3500 kcal per pound” rule is based on the calculated energy content of body weight change and is often misapplied to predict the weight-change time course after a given intervention (19). This is a fundamental error because no time period is specified for that intervention. The impression is given that even a temporary intervention will therefore result in a per- manent body weight change. Furthermore, the erroneous appli- cation of the rule to predict the impact of a permanent intervention gives the impression that a linear change in body weight is expected over protracted periods of time, which is known to be untrue. Rather, even when perfect adherence to an intervention with no active compensation is assumed, it is generally ac- knowledged that weight change will slow over time due to passive compensatory changes in energy expenditure that occur with the weight change. Therefore, the panel recommended that the 3500 kcal per pound rule should no longer be used.
With the use of a model that accounts for the passive com- pensatory effects on EO, a new rule of thumb representing a best- case scenario has been proposed for the average overweight per- son: every permanent 10-kcal change in energy intake/d will lead
992 HALL ET AL
at A M
E R
IC A
N S
O C
IE T
Y F
O R
N U
T R
IT IO
N on A
pril 2, 2012 w
w w
.ajcn.org D
ow nloaded from
to an eventual weight change of 1 lb when the body weight reaches a new steady state (;100 kJ/d per kg of weight change). It will take nearly 1 y to achieve 50% and ;3 y to achieve 95% of this weight loss (20).
Whereas the above rule of thumb may be useful for approximate estimations and represents a significant theoretical improvement over the 3500 kcal per pound rule, a more accurate assessment of the amount and time course of predicted weight change for a given reduction in EI may be very valuable and informative for an in- dividual patient. Newly developed dynamic energy balance models for weight loss require complex calculations that are simplified for users in web-based programs (http://bwsimulator.niddk.nih.gov; http://www.pbrc.edu/the-research/tools/weight-loss-predictor). Model predictions such as these provide a more realistic guide as to what patients can expect with changes in energy balance.
D. “Small changes in lifestyle can prevent or reverse obesity.”
Small lifestyle changes in either intake or expenditure (ac- tivity) are being increasingly promoted as viable interventions. It is important not to have unreasonable expectations about the impact of such interventions on body weight. Because the 3500 kcal per pound rule has often been used to model the effects of such interventions, unrealistic predictions are frequently made about the likely weight-loss benefits of exercise and dietary interventions that make only minor adjustments to lifestyle. As noted above, it is inappropriate to use the 3500 kcal per pound rule to model the effects of interventions. To illustrate this problem, a 40-kcal/d (170-kJ/d) permanent reduction in energy intake resulting from taxing sweetened beverages has been predicted to result in ;20 lb (9 kg) of weight loss in 5 y ac- cording to the 3500 kcal per pound rule, whereas only 4 lb (2 kg) of weight loss is predicted using the new rule of thumb (20).
The recommendation that an overweight or obese person should expend an additional daily 100 kcal (420 kJ) in walking (ie, walking one mile a day), given the new rule of thumb dis- cussed above, would result in a weight loss of ;10 lb (4.5 kg) over 5 y, as opposed to a loss of 50 lb (23 kg) if the 3500 kcal per pound rule is used. Although a 10-lb weight loss can often produce major health gains, which points to a potentially sig- nificant benefit of small lifestyle changes, it is not nearly the amount of weight loss from this physical activity regimen that the 3500 kcal per pound rule suggests. Moreover, even the re- vised rule is an optimistic assessment of weight change because it does not account for the potential active compensation of EI.
Question 4: What limitations do we face in the study of energy balance and its components?
Our ability to measure precisely individual components of energy expenditure or energy intake is relatively poor in light of the potential impact of small changes described above on body weight, especially over extended time scales in free-living individuals. For example, the doubly labeled water method has a precision of;5%, which translates to an uncertainty of energy expenditure of .100 kcal/d (420 kJ/d). In addition, the accuracy and precision of en- ergy intake measurements by self-report in free-living individuals are much worse. Thus, the combined error of assessing energy imbalance can easily reach 1000 kcal/d (4200 kJ/d) (21). This potential error prevents evaluation of the benefits of interventions that have a small benefit on weight change over time. New
technologies currently in development may be more accurate and precise, but that remains to be seen.
Another limitation that we face is that body weight over a day, and between days, fluctuates unrelated to changes in energy stores because of changes in hydration and alimentary tract content, which are the primary contributors to the typical 1–2-lb day-to-day fluctuations in weight. Yet another limitation we face is that the calculation of the energy deficit generated by a given diet requires knowing the energy requirement to maintain the baseline body weight. As stated above, the imprecision is.100 kcal/d when the most precise methods currently available are used. The un- certainty of baseline energy requirements translates to a consid- erable interindividual variability of weight loss, even if adherence to the prescribed diet is perfect. For example, if the baseline energy requirement of an overweight or obese person is 100–200 kcal/d higher or lower than measured, then perfect adherence to a diet will result in an error of;5–10 lb (2.3–4.5 kg) in predicted weight change over a year because of measurement error alone. This limitation is less of a concern in studies designed to measure average differences between groups.
In inpatient studies, more precise measurement techniques are available, which thereby decreases measurement error. For ex- ample, whole-room calorimeters can measure EO with 1–2% precision (22) and weighed, supervised food intake with measured excreta can provide very accurate and precise measurements of EI. However, such studies do not represent free-living conditions.
Finally, the characteristically long time scale (;1 y half-time) for human body weight and composition changes to occur make it difficult to study comprehensively the dynamics of energy balance because we cannot generally keep humans in metabolic wards for such extended periods. Even in a free-living situation we cannot track EI or EO for prolonged periods using current technologies. We are thus limited to “snapshots” of periods of ;2 wk.
Question 5: What research would better inform our knowledge of energy balance and its components?
It is important to recognize that the energy balance system is interactive and complex: a change in one component can affect one or more other components. The panel identified the following important gaps in our knowledge that deserve future investigation:
1. Although we know much from short-term studies about the major components of energy balance, our knowledge is still deficient regarding their interaction over the long term. Therefore, we need long-term, longitudinal studies to learn the details of the relations between components of energy balance and changes in body composition and weight among children and adults.
2. It has been shown that biological and psychological factors affect the components of energy balance. But generally, these have been studied independently of one another and an integrative approach is required. We need to know the relative importance of preingestive factors (cognitive and sensory effects of food/meals) on energy intake, energy balance, and the physiologic response to a meal.
3. Although our knowledge of the broader implications of physical activity and exercise have been investigated, we need to understand the effects of different doses (volume, intensity, pattern, timing) and types (endurance, resistance)
ENERGY BALANCE AND ITS COMPONENTS 993
at A M
E R
IC A
N S
O C
IE T
Y F
O R
N U
T R
IT IO
N on A
pril 2, 2012 w
w w
.ajcn.org D
ow nloaded from
of exercise on 1) total daily energy expenditure and its components (REE, TEF, AEE), 2) EI and food preferences, and 3) body composition and body weight in children and adults.
4. The individual variation in weight-loss response to energy balance interventions is striking, and therefore we need to know the mechanism or mechanisms responsible for the un- derlying active compensatory differences in energy intake, food preferences, and body weight in children and adults. In particular, we have almost no information from energy bal- ance studies subsequent to weight loss during the difficult period of weight maintenance. How can we identify popu- lation subgroups or even individuals who will respond or not respond to a dietary or exercise intervention?
5. Measurements of energy input and output are neither pre- cise nor accurate enough to allow the calculation of energy balance over the appropriate timeframe needed to under- stand the mechanisms responsible for excess weight gain. Accordingly, we need to develop new methods that can re- liably measure energy balance over extended time periods in free-living people.
The 1-d Consensus Conference included presentations from the following
speakers: David Allison (University of Alabama at Birmingham), John Blundell
(University of Leeds), Myles Faith (University of North Carolina), James Hill
(University of Colorado at Denver), John Jakicic (University of Pittsburgh),
Richard Mattes (Purdue University), John Peters (University of Colorado at
Denver), Eric Ravussin (Pennington Biomedical Research Center), and Susan
Roberts (Jean Mayer USDA Human Nutrition Center on Aging). All authors
read and approved the final manuscript. All authors participated equally in
the development of the statement.
REFERENCES 1. Johnstone AM, Murison SD, Duncan JS, Rance KA, Speakman JR.
Factors influencing variation in basal metabolic rate include fat-free mass, fat mass, age, and circulating thryroxine but not sex, circulating leptin, or triiodothyronine. Am J Clin Nutr 2005;82:941–8.
2. Hirsch J, Knittle JL. Cellularity of obese and non-obese human adipose tissue. Fed Proc 1970;29:1516–21.
3. Edholm OG, Fletcher JG, Widdowson EM, McCance RA. The energy expenditure and food intake of individual men. Br J Nutr 1955;9:286– 300.
4. Morton GJ, Cummings DE, Baskin DG, Barsh GS, Schwartz MW. Central nervous system control of food intake and body weight. Nature 2006;443: 289–95.
5. Levine JA, Eberhardt NL, Jensen MD. Role of nonexercise activity thermogenesis in resistance to fat gain in humans. Science 1999;283: 212–4.
6. Roberts SB, Fuss P, Dallal GE, Atkinson A, Evans WJ, Joseph L, Fiatarone MA, Greenberg AS, Young VR. Effects of age on energy expenditure and substrate oxidation during experimental overfeeding in healthy men. J Gerontol A Biol Sci Med Sci 1996;51:B148–57.
7. Joosen AM, Bakker AHF, Westerterp KR. Metabolic efficiency and energy expenditure during short-term overfeeding. Physiol Behav 2005; 85:593–7.
8. Saltzman E, Roberts SB. The role of energy expenditure in energy reg- ulation: findings from a decade of research. Nutr Rev 1995;53:209–20.
9. Luke A, Dugas LR, Ebersole K, Durazo-Arvizu RA, Cao G, Schoeller DA, Adeyemo A, Brieger WR, Cooper RS. Energy expenditure does not predict weight change in either Nigerian or African American women. Am J Clin Nutr 2009;89:169–76.
10. Tataranni PA, Harper IT, Snitker S, Del Parigi A, Vozarova B, Bunt J, Bogardus C, Ravussin E. Body weight gain in free-living Pima Indians: effect of energy intake vs expenditure. Int J Obes Relat Metab Disord 2003;27:1578–83.
11. Goran MI, Shewchuk R, Gower BA, et al.Longitudinal changes in fatness in white children: no effect of childhood energy expenditure. Am J Clin Nutr 1998;67:309–16.
12. Westerterp KR, Speakman JR. Physical activity energy expenditure has not declined since the 1980s and matches energy expenditures of wild mammals. Int J Obes (Lond) 2008;32:1256–63.
13. Church TS, Thomas DM, Tudor-Locke C, Katzmarzyk PT, Earnest CP, Rodarte RQ, Martin CK, Blair SN, Bouchard C. Trends over 5 decades in U.S. occupation-related physical activity and their associations with obesity. PLoS ONE 2011;6:e19657.
14. Donnelly JE, Hill JO, Jacobsen DJ, Potteiger J, Sullivan DK, Johnson SL, Heelan K, Hise M, Fennessey PV, Sonko B, et al. Effects of a 16-month randomized controlled exercise trial on body weight and composition in young, overweight men and women: the Midwest Ex- ercise Trial. Arch Intern Med 2003;163:1343–50.
15. Speakman JR, Selman C. Physical activity and resting metabolic rate. Proc Nutr Soc 2003;62:621–34.
16. LaForgia J, Withers RT, Gore CJ. Effects of exercise intensity and duration on the excess post-exercise oxygen consumption. J Sports Sci 2006;24:1247–64.
17. Heymsfield SB, Harp JB, Reitman ML, Beetsch JW, Schoeller DA, Erondu N, Pietrobelli A. Why do obese patients not lose more weight when treated with low-calorie diets? A mechanistic perspective. Am J Clin Nutr 2007;85:346–54.
18. Hall KD. Predicting metabolic adaptation, body weight change, and energy intake in humans. Am J Physiol Endocrinol Metab 2010;298: E449–66.
19. Hall KD. What is the required energy deficit per unit weight loss? Int J Obes (Lond) 2008;32:573–6.
20. Hall KD, Sacks G, Chandramohan D, Chow CC, Wang YC, Gortmaker SL, Swinburn BA. Quantifying the effect of energy imbalance on body weight change. Lancet 2011;378:826–37.
21. Speakman JR. Doubly-labelled water: theory and practice. London, United Kingdom: Chapman and Hall, 1997.
22. Melanson EL, Ingebrigtsen JP, Bergouignan A, Ohkawara K, Kohrt WM, Lighton JR. A new approach for flow-through respirometry measurements in humans. Am J Physiol Regul Integr Comp Physiol 2010;298:R1571–9.
994 HALL ET AL
at A M
E R
IC A
N S
O C
IE T
Y F
O R
N U
T R
IT IO
N on A
pril 2, 2012 w
w w
.ajcn.org D
ow nloaded from
Human Energy Requirements(1)(1).pdf
1F A OF O O D A N D N U T R I T I O N T E C H N I C A L R E P O R T S E R I E S
ISSN 1813-3932
Human energy requirements Report of a Joint FAO/WHO/UNU Expert Consultation Rome, 17–24 October 2001
(low) Level of requirement (high)
Energy
Average requirement
Average requirement
P er
ce nt
ag e
o f
in d
iv id
ua ls
(low intake) Usual level of intake (high intake)
1.0 1.0
0 0
P ro
b ab
ili ty
t ha
t st
ip ul
at ed
in ta
ke is
in ad
eq ua
te fo
r a
ra nd
o m
ly s
el ec
te d
in d
iv id
ua l
P ro
b ab
ili ty
t ha
t st
ip ul
at ed
in ta
ke is
e xc
es si
ve fo
r a
ra nd
o m
ly s
el ec
te d
in d
iv id
ua l
iii
FOREWORD
Assessing the calorie and nutrient requirements of human beings, with the greatest possible degree of accuracy, is one of the most important and central mandates of the Food and Agriculture Organization of the United Nations (FAO). Since 1948, FAO has convened numerous expert groups in the field of nutrition to collate, evaluate and interpret current scientific knowledge in order to derive estimates of human energy requirements and use these estimates to provide recommendations to people and policy-makers. The World Health Organization (WHO) began its collaboration with FAO on this important work in the early 1950s, while the United Nations University (UNU) joined the initiative in 1981.
This important publication is the final report of the most recent expert group meeting, the Joint FAO/WHO/UNU Expert Consultation on Human Energy Requirements, convened in October 2001 at FAO headquarters in Rome, Italy. The primary purpose of the expert meetings on nutrient requirements has remained the same throughout the years: to give advice on scientific issues related to food energy and nutrient requirements and to formulate recommendations for action. Various expert groups have contributed principles for determining and applying general requirements, which have been adopted worldwide.
The global scientific community has continued to embrace the advice on requirements that was first published by FAO alone and later in collaboration with WHO. The FAO/WHO recommendations have reflected the state of knowledge at particular points in time, and have also influenced research agendas and methodologies over the years. In fact, the FAO/WHO recommendations are currently utilized in virtually all countries, and nutrient requirement reports are among the most frequently referenced and most sought-after publications in both organizations.
Estimates of human energy requirements are essential for assessing whether food supplies are adequate to meet a population’s nutritional needs. Such estimates are also essential in assessing the proportion and absolute number of undernourished people worldwide. The recommendations derived from these estimates assist governments to monitor nutrition programmes and plan development activities. The recommendations also help with the specific formulation of planning at the national level for agricultural production, food supplies and the mobilization and distribution of emergency food aid. FAO has an ongoing mandate to review these assessments periodically – particularly as new research findings emerge – and to produce estimates using the highest possible degree of accuracy based on recent scientific advances and developments in the field.
FAO publishes this report on behalf of the three United Nations (UN) agencies (FAO/WHO/UNU) that organized the consultation. We would like to express our gratitude to the members of the expert consultation for their contribution to this important report, as well as to the numerous participants of the working groups. The work of these groups preceded the expert consultation and served as the foundation for discussions and exchange during the meeting. Thanks are also due to Dr E. Kennedy, who very skilfully chaired the expert consultation, and to Dr B. Torun for his commitment to the role of rapporteur and for his contribution to early drafts of this report.
We thank all the participants, as well as the non-participating experts who drafted background papers as part of the preparatory process for the expert consultation. These background papers will be published in a special issue of Public Health Nutrition in 2005, thereby providing a more detailed peer-reviewed literature source for many of the ongoing debates on the various topics discussed during the consultation. We would also like to express our special gratitude to the FAO staff members who constituted the Secretariat and completed much of the follow-up work that culminated in this report, in particular the staff of the Nutrition Planning and Evaluation Service (ESNA), P. Shetty, R. Weisell, and B. Burlingame, as well as G. Kennedy, F. Martinez Nocito, T. Ballard and J. Shaw who assisted as consultants both during and after the expert consultation.
Kraisid Tontisirin Hartwig de Haen
Director Assistant Director-General Food and Nutrition Division Economic and Social Department
v
CONTENTS
FOREWORD iii
PREFACE vii
1. INTRODUCTION 1
1.1 What is new in this report? 1
1.2 Intended use of this report 2
1.3 Policy implications 2
References 3
2. PRINCIPLES AND DEFINITIONS 4
2.1 Definitions 4
2.2 Sources of dietary energy 6
2.3 Components of energy requirements 7
2.4 Calculation of energy requirements 7
2.5 Recommendations for physical activity 9
2.6 Glossary and abbreviations 9
References 10
3. ENERGY REQUIREMENTS OF INFANTS FROM BIRTH TO 12 MONTHS 11
3.1 Measurement of total energy expenditure 11
3.2 Equations to predict energy expenditure 11
3.3 Energy needs for growth 13
3.4 Calculation of energy requirements 15
3.5 Catch-up growth 17
References 18
4. ENERGY REQUIREMENTS OF CHILDREN AND ADOLESCENTS 20
4.1 Measurement of total energy expenditure 20
4.2 Equations to predict total energy expenditure 20
4.3 Energy needs for growth 21
4.4 Calculation of energy requirements 21
4.5 Recommendations for regular physical activity 24
4.6 Infections and mild malnutrition 31
References 32
5. ENERGY REQUIREMENTS OF ADULTS 35
5.1 Factorial estimation of total energy expenditure and physical activity level 35
5.2 Estimation of basal metabolic rate 35
5.3 Physical activity level 37
5.4 Energy requirements and dietary energy recommendations 39
5.5 Older adults and the elderly 47
5.6 Recommendations for regular physical activity 49
References 50
6. ENERGY REQUIREMENTS OF PREGNANCY 53
6.1 Gestational weight gain and optimal pregnancy outcome 53
6.2 Determinants of the energy cost of pregnancy 54
6.3 Calculation of energy requirements for pregnancy 56
6.4 Special considerations for malnourished, obese and adolescent pregnant women 60
References 61
vi
7. ENERGY REQUIREMENTS OF LACTATION 63
7.1 Determinants of the energy cost of lactation 63
7.2 Energy requirements for lactation 65
References 66
8. RECOMMENDATIONS FOR FUTURE RESEARCH 67
8.1 Biological questions: conceptual and methodological 67
8.2 Epidemiological and community studies 69
9. CONCLUSIONS 71
References 73
ANNEXES 75
1: Participants – 2001 Joint FAO/WHO/UNU Expert Consultation on Human Energy
Requirements 77
2: Authors and reviewers of papers for expert consultation working groups, meetings and
follow-up 84
3: Update on predictive equations to estimate basal metabolic rate 87
4: Software application for calculating populations’ energy requirements and food needs 89
5: Energy costs of activities 92
vii
PREFACE
The purpose of the expert consultations on human energy requirements convened by FAO, WHO and, more recently, UNU is to advise the Directors-General on scientific issues related to food energy, including requirements, so that appropriate recommendations for action can be formulated. It is hence important that during the process of determining energy requirements the question of “requirements for what?” be constantly borne in mind. While biological scientists are generally concerned with the physiological basis of estimating requirements, it is also necessary to be aware of the practical applications of these recommendations for estimating the energy requirements and food needs of populations worldwide.
The principal objective of expert consultations on human energy requirements is to provide international agencies and their member countries with the necessary tools for addressing practical questions, such as assessing the adequacy of food supplies and the numbers of people who do not attain energy adequacy, drawing up targets for food production and informing national food and nutrition policy and planning. The recommendations and guidelines that result from these consultations will serve to enable governments and organizations to plan, monitor and evaluate nutrition programmes and policies better. They will also help Member Nations to develop estimates of requirements appropriate for local conditions and for direct application in their countries. It is important to remember that while developed countries are able to constitute their own committees of experts who can make recommendations on energy and nutrient requirements for their populations, the majority of humanity in the developing world relies largely on UN agencies such as FAO. Hence, the development of pragmatic recommendations by expert committees convened by UN agencies, which are based on sound scientific evidence and have practical relevance to the conditions prevailing in the developing world, is paramount.
The entire process leading up to the convening of an expert group and the resulting consultation is highly formalized and follows a number of required protocols. For the first time, FAO adopted a two- stage process, which started with convening working groups in those areas where it believed that new scientific knowledge existed that might influence the current recommendations for energy needs. The second stage of the process was the expert consultation itself. The rationale behind convening the working groups was that many of the scientific questions could be dealt with by experts in the areas concerned, even though the participation of those experts at the consultation per se was uncertain owing to the need to provide a globally representative consultative panel. Working groups would also facilitate discussions, as any contentious issues could be debated and settled before the expert meeting, which would benefit from the results of such discussions. Accordingly, working groups met from 27 June to 5 July 2001 at FAO headquarters in Rome, several months before the expert meeting in October 2001. Three of the working groups focused primarily on energy requirements throughout the life cycle and related to two important sub-populations – infants and children, and pregnant and lactating women – for which substantial scientific advances had been made. These working groups were on: 1) energy (and protein) requirements of infants and preschool children; 2) energy (and protein) requirements of pregnancy and lactation; and 3) analytical issues in food energy and composition: energy in food labelling, including regulatory and trade issues, which looked at food energy values. An additional working group was constituted to provide documentation on methodologies for energy balance and energy requirements, but it was felt that – given the nature of the task – there was no need for this group to meet, although their background documents were available to the expert consultation. The chairpersons of all the working groups on energy were invited to the expert consultation to present a summary of the deliberations and recommendations of their groups and to advise the experts. Background papers were commissioned, peer-reviewed and made available to both the pre-consultation working groups and the experts who met for the consultation. The entire process of pre-consultation activities and the consultation itself went smoothly, despite a few hitches that were largely the result of the unhappy events of 11 September 2001, which prevented some of the invited experts from coming to Rome to join the consultative process. Lists of the participants in the various working group sessions, and those invited as experts to the consultation are included as Annex 1 of this report. Annex 2 provides details of the authors and
viii
reviewers of the background documents, which are expected to be published shortly as a supplement to the journal Public Health Nutrition. The wide availability of this publication as a peer-reviewed journal supplement is expected to provide the academic community with an opportunity to examine the collated evidence base that informed the expert panel and influenced their latest recommendations.
As part of the second stage of the process, the members of the expert consultation met in FAO headquarters in Rome from 17 to 21 October 2001. The meeting was chaired by Dr E. Kennedy, with Dr B. Torun serving as rapporteur. The following are the specific tasks outlined in the charge given to this expert consultation on human energy requirements:
1. To review the background documents on the state of the art of the scientific literature in this area of work, assembling the best evidence on the topic and using, where appropriate, the summary, advice and recommendations arising from the deliberations of the working groups that had met earlier in the year.
2. To deliberate on and arrive at recommendations for energy requirements throughout the life cycle, while clearly outlining the approaches used to estimate requirements that may be of benefit to users. This included taking into account physiological states such as growth, pregnancy and lactation and, where relevant, pathological conditions and the additional needs during infections. The recommendations were expected to be reached by consensus, and where differences persisted the reasons for those differences were to be clearly outlined, with all sides presented and appropriately reflected in the report of the expert consultation.
3. To examine the feasibility of arriving at minimum requirements that may be of use in estimating the numbers of individuals in populations who are unable to meet energy adequacy.
4. To comment on the consequences of deficit and excess of energy, and to recommend ways by which the health, social and economic consequences of these can be minimized or avoided.
5. To highlight the main changes to the recommendations of the 1985 report, with particular emphasis on those aspects of the new recommendations that have a significant impact on the way in which nutritional adequacy of population groups is assessed by those involved in policy, planning or analysis of the nutritional status of populations.
6. To suggest areas where further research is needed, either to deal with gaps in the knowledge related to energy requirements in specific groups or situations, or to facilitate the collection of normative data that will aid the process of arriving at future recommendations for energy requirements.
It was the sincere desire of the FAO Secretariat to ensure that the report of the expert consultation on human energy requirements be available within the shortest possible period after the experts met in Rome. The two-year gap before the interim report was available as a downloadable file on the FAO Web site, and a further period before it was available in hard copy were due to a series of post- consultation activities that were deemed essential before the release of the final report. Many of these post-consultation activities were in response to, and out of respect for, the experts who recommended a number of important pieces of work to be followed-up and completed for inclusion in the report.
An important recommendation of the expert group was to update and review the predictive equations for estimating basal metabolic rate (BMR) and to incorporate the updated equations into the new recommendations. These activities proved to be time-consuming, as they involved updating the global database on BMRs that was originally obtained for the 1985 report, reanalysing it with particular emphasis on looking at the influence of methodological biases and ethnic variations, and developing new BMR predictive equations with better predictive performance for international use (Annex 3). The reanalysis was followed by an exercise to test the validity of the new equations, and a further consultation with a sub-group of the expert panel for their final decision. However, after this long exercise the experts concluded that the international equations hitherto used continued to have enhanced precision and robustness. Following reanalysis of the global database, the recommendation to use a seamless single predictive BMR equation was not considered practical, and hence the expert consultation was not persuaded to replace the international equations provided in the 1985 report. These predictive equations have been widely used and are popular with the scientific community and nutritional planners, and the present report’s recommendation is to continue using them.
ix
One of the other recommendations that arose from the deliberations of the working group on analytical issues in food energy and composition, which was subsequently endorsed by the experts, was to convene a meeting to deliberate on food energy values. The objective was to ensure harmony between the expected adoption of new energy requirement values from this consultation, which are based solely on energy expenditure measurements or estimates, and energy requirements based on food intake measurements alone. FAO thus convened a Technical Workshop on Food Energy – Methods of Analysis and Conversion Factors, which was held in Rome from 3 to 6 December 2002. The report of this workshop was published as FAO Food and Nutrition Paper No. 77 in 2003, which complements the present report.
As part of the post-consultation activities in preparation for the release of the expert report, it was decided to produce an updated, Windows-compatible and user-friendly software application for the purpose of calculating population energy requirements and food needs. After the 1981 joint expert consultation report was released (WHO, 1985), FAO sponsored the development of a manual and software package (James and Schofield, 1990), recognizing that less attention had hitherto been paid to the matter of how to apply the requirements to practical food and nutrition planning. The success of this 1990 user’s manual, which was sponsored by FAO and published by Oxford University Press, was constrained because it was a priced publication that was available separately from the 1985 joint expert report. For the 2001 consultation, it was decided to make the new software widely and readily available by releasing it alongside the report. FAO therefore had to find an organization that would assist us in developing such a product to be released at the same time as the expert report in 2004. Early discussions were conducted with the United States Centers for Disease Control and Prevention (CDC) in Atlanta, Georgia, with the objective of developing the software and making it available as a downloadable version alongside CDC’s popular EpiNut software. However, CDC was unable to collaborate in this venture, so other partners had to be sought. The Division of Nutrition, Institute of Population Health and Clinical Research at Bangalore, India and its Dean, Dr A. Kurpad, identified Jenesys Technologies, a software applications firm in India, which collaborated alongside the institute in the development of the software package and accompanying manual (Annex 4). This is now available on CD-ROM. For the first time, the software package is being issued together with the expert report in order to ensure that those interested in the report’s recommendations have the means to investigate and ensure their practical applicability, as well as to benefit from the two outputs’ complementarity. The user’s manual and software application for calculating population energy requirements and food needs thus represent a further milestone in FAO’s continued involvement in both the theoretical and the practical issues related to human energy requirements.
This expert consultation was convened nearly two decades after the last expert group met to deliberate on energy and protein requirements in 1981. In the interim, the International Dietary Energy Consultancy Group (IDECG), sponsored jointly by UNU and the International Union of Nutritional Sciences (IUNS), filled the lacuna by convening meetings to discuss important developments in this area. The IDECG meeting in London in 1994 on Energy and Protein Requirements (whose proceedings were published in European Journal of Clinical Nutrition Vol. 50, Supplement 1 in February 1996) was a seminal meeting that provided much of the preparatory background for this expert consultation. We would like to acknowledge and pay our tribute to the late, Dr Beat Schurch who, as Executive Secretary of IDECG, was the quiet engine behind this invaluable contribution to the advancement and dissemination of nutrition knowledge. FAO and WHO benefited greatly from IDECG’s work and publications, in particular its review of human energy and protein requirements in 1994. While FAO was organizing the 2001 expert consultation, Beat Schurch knew that he was sick but planned to attend both the consultation and the working groups that preceded it. Unfortunately, his illness progressed more quickly than had been anticipated, and he had to decline the invitation. He approached his illness and its culmination with the same equanimity with which he approached most matters and wished the group well. His contribution and friendship will be sorely missed.
Prakash Shetty
Chief Nutrition Planning, Assessment & Evaluation Service (ESNA)
Food & Nutrition Division
Introduction
1
1. INTRODUCTION
Since 1949, the Food and Agriculture Organization of the United Nations (FAO) and, since the early 1950s, the World Health Organization (WHO) have convened groups of experts to evaluate current scientific knowledge in order to define the energy requirements of humans and propose dietary energy recommendations for populations. The purpose of this information is to assist FAO and WHO in implementing their programmes. The United Nations University (UNU) became part of this joint initiative in 1981. The reports of these expert meetings (see the list of References at the end of this chapter) have become important guidelines on energy in human nutrition for academic scientists, nutritionists, physicians and other health workers, as well as for planners and policy-makers in both the agriculture and health sectors throughout the world.
New scientific knowledge generated in the 20 years since the last expert consultation was held prompted FAO, WHO and UNU to assemble a new expert consultation to make recommendations for energy requirements of populations throughout the life cycle (WHO, 1985). This consultation took place from 17 to 24 October 2001 at FAO headquarters in Rome. Its mandate was to revise and update the conclusions and recommendations of the preceding consultation, which was convened in 1981 and whose report was published in 1985. In preparation for the forthcoming expert consultation, well- known scientists with demonstrated expertise in this area of work were asked to examine and write background papers on various topics that required revision and updating. Several of the authors and other leading scientists constituted working groups that met in Rome in June 2001 to discuss and analyse critically the contents of the background papers, which were subsequently modified following the working group suggestions. The modified papers, the working groups’ conclusions and other relevant documents were provided to all members of the expert consultation for analysis and consideration in their deliberations.1
Dr Eileen Kennedy was elected to chair this expert consultation, and Dr Benjamin Torun to be the rapporteur. Several conclusions and recommendations were the immediate results, while a number of topics were identified as requiring further research and analysis before the experts could finalize their recommendations. The rapporteur and other members of the consultation were given the task of pursuing the pending issues with assistance from the FAO Secretariat, and additional working papers were commissioned. This laborious task went on until the end of 2003, when almost all questions had been answered and gaps filled and the rapporteur was able to prepare the final draft for examination and approval by the other experts from the consultation. This report is the final result of those efforts.
1.1 WHAT IS NEW IN THIS REPORT?
Although the basic principles set forth in previous expert meetings have withstood the test of time, several modifications are proposed in this report. Members of the expert consultation and participants in the working groups recognize and accept the responsibility for proposing these modifications, and for the implications that they will have on health, agriculture, the food industry, economic planning, international aid and social programmes related to food and nutrition. It is their belief that the conclusions and recommendations in this report are well grounded, given the current state of the best scientific knowledge. A critical appraisal of their application will be the final proof of their accuracy, applicability and appropriateness.
The new concepts and recommendations set forth in this report include:
calculation of energy requirements for all ages, based on measurements and estimates of total daily energy expenditure and on energy needs for growth, pregnancy and lactation;
in the light of new data, modification of the requirements and dietary energy recommendations for infants and for older children and adolescents, in order to correct previous overestimations for the former and underestimations for the latter;
1 Annex 1 gives the names of participants in the working groups and expert consultation. Annex 2 lists the titles and authors of the background documents.
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
2
proposals for differentiating the requirements for populations with lifestyles that involve different levels of habitual physical activity, starting as early as six years of age;
reassessment of energy requirements for adults, based on energy expenditure estimates expressed as multiples of basal metabolic rates;
classification of physical activity levels based on the degree of habitual activity that is consistent with long-term good health and maintenance of a healthy body weight;
recommendations for physical activity for children and adults to maintain fitness and health and to reduce the risk of developing obesity and co-morbid diseases associated with a sedentary lifestyle;
an experimental approach for factorial estimates of energy needs during pregnancy and lactation;
distribution in the two last trimesters of pregnancy of the recommendations for additional dietary energy needs.
1.2 INTENDED USE OF THIS REPORT
This report is briefer and less detailed than the reports of previous expert meetings and consultations. The commissioned background papers, which will be published in a peer-reviewed journal, complement the report with details on the sources, analysis and interpretation of the scientific information. In addition to a printed version, the report will be placed on the Internet for wider access and faster diffusion.
This report is not meant merely to describe the energy expenditures and requirements of population groups. It intends to be prescriptive, in order to support and maintain health and good nutrition. The recommendations, however, are meant for well-nourished and healthy populations, as the correction of malnutrition – either deficit or excess – involves different energy requirements and dietary recommendations. The report is not meant to be prescriptive for individual subjects, some of whom may be at either extreme of a normal distribution. Although estimates of requirements are derived from measurements of individuals with specific characteristics such as age, gender, body size, presumed body composition and physical activity, the data have been pooled to give the average energy requirements of groups or classes of individuals who have similar characteristics, but on whom measurements have not been made. Consequently, application of these results to any one individual for clinical or other purposes may lead to errors of diagnosis and improper management.
1.3 POLICY IMPLICATIONS
A science-based definition of human energy requirements is crucial for the control and prevention of undernutrition due to insufficient intake of food energy, which remains a major problem for many countries. It is also essential to efforts to curb the excessive intake of food energy that is a major determinant of nutrition-related chronic diseases, at present an important cause of worldwide morbidity and mortality among adults.
Insufficient food energy intake is almost always accompanied by a deficient intake of most nutrients. Awareness of the consequences of insufficient energy intakes in children and adults has influenced health and food and agriculture policies around the world. More recently, the consequences of increasing obesity and nutrition-related chronic diseases have also been recognized as major factors for the health, food and agriculture sectors. These problems are increasing globally as a result of changes in diets and lifestyles that are reflected in changing food cultures and physical activity patterns among all segments of society, and not only among affluent groups or in the richest countries. Undernutrition early in life, followed by an inappropriate diet and low physical activity in childhood and adult life increases vulnerability to chronic non-communicable diseases. Low-income groups in urban areas are especially vulnerable to the risk of obesity owing to a positive energy balance. The current increased incidence of overweight and obesity among children and adults in most countries leads to rapidly rising projections of disability and premature death to nutrition-related chronic diseases.
Prevention is the only feasible approach to control the double burden of under- and overnutrition. The cost of treating and managing the ensuing disabilities and diseases imposes an intolerable economic and health burden, especially for poorer countries. As inappropriate dietary intake and lack
Introduction
3
of physical activity are the main causes of nutritional problems, there is an urgent need for governments, in partnership with all relevant stakeholders, to integrate strategies that promote healthy diets and regular physical activity in all relevant policies and programmes, including those designed to fight undernutrition. Both undernutrition and obesity are preventable, as has been demonstrated by countries with successful programmes. In addition to health promotion, nutrition education and relevant agricultural and food policies, effective food and nutrition programmes must include community action to overcome the environmental, social and economic constraints that limit the improvement of access to food, and to promote better dietary quality and life style practices that encourage a physically active life.
REFERENCES FAO. 1950. Calorie requirements: Report of the Committee on Calorie Requirements. FAO Nutritional Studies No. 5. Rome. FAO. 1957. Calorie requirements: Report of the Second Committee on Calorie Requirements. FAO Nutritional Studies No. 15. Rome. FAO/WHO. 1973. Energy and protein requirements: Report of a joint FAO/WHO ad hoc expert committee. FAO Nutrition Meetings Report Series No. 52. WHO Technical Report Series No. 522. Rome and Geneva. WHO. 1985. Energy and protein requirements: Report of a joint FAO/WHO/UNU expert consultation. WHO Technical Report Series No. 724. Geneva.
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
4
2. PRINCIPLES AND DEFINITIONS
Human energy requirements are estimated from measures of energy expenditure plus the additional energy needs for growth, pregnancy and lactation. Recommendations for dietary energy intake from food must satisfy these requirements for the attainment and maintenance of optimal health, physiological function and well-being. The latter (i.e. well-being) depends not only on health, but also on the ability to satisfy the demands imposed by society and the environment, as well as all the other energy-demanding activities that fulfil individual needs.
Energy balance is achieved when input (i.e. dietary energy intake) is equal to output (i.e. total energy expenditure), plus the energy cost of growth in childhood and pregnancy, or the energy cost to produce milk during lactation. When energy balance is maintained over a prolonged period, an individual is considered to be in a steady state. This can include short periods during which the day- to-day balance between intake and expenditure does not occur. An optimal steady state is achieved when energy intake compensates for total energy expenditure and allows for adequate growth in children, and pregnancy and lactation in women, without imposing metabolic, physiological or behavioural restrictions that limit the full expression of a person’s biological, social and economic potential.
Within certain limits, humans can adapt to transient or enduring changes in energy intake through possible physiological and behavioural responses related to energy expenditure and/or changes in growth. Energy balance is maintained, and a new steady state is then achieved. However, adjustments to low or high energy intakes may sometimes entail biological and behavioural penalties, such as reduced growth velocity, loss of lean body mass, excessive accumulation of body fat, increased risk of disease, forced rest periods, and physical or social limitations in performing certain activities and tasks. Some of these adjustments are important and may even increase the chances of survival in times of food scarcity.
2.1 DEFINITIONS
An adequate, healthy diet must satisfy human needs for energy and all essential nutrients. Furthermore, dietary energy needs and recommendations cannot be considered in isolation of other nutrients in the diet, as the lack of one will influence the others. Thus, the following definitions are based on the assumption that requirements for energy will be fulfilled through the consumption of a diet that satisfies all nutrient needs.
Energy requirement is the amount of food energy needed to balance energy expenditure in order to maintain body size, body composition and a level of necessary and desirable physical activity consistent with long-term good health. This includes the energy needed for the optimal growth and development of children, for the deposition of tissues during pregnancy, and for the secretion of milk during lactation consistent with the good health of mother and child.
The recommended level of dietary energy intake for a population group is the mean energy requirement of the healthy, well-nourished individuals who constitute that group.
Based on these definitions, a main objective for the assessment of energy requirements is the prescription of dietary energy intakes that are compatible with long-term good health. Therefore, the levels of energy intake recommended by this expert consultation are based on estimates of the requirements of healthy, well-nourished individuals. It is recognized that some populations have particular public health characteristics that are part of their usual, “normal” life. Foremost among these are population groups in many developing countries where there are numerous infants and children who suffer from mild to moderate degrees of malnutrition and who experience frequent episodes of infectious diseases, mostly diarrhoeal and respiratory infections. Special considerations are made in this report for such sub-populations.
5
2.1.1 Daily energy requirements and daily energy intakes
Energy requirements and recommended levels of intake are often referred to as daily requirements or
recommended daily intakes. These terms are used as a matter of convention and convenience,
indicating that the requirement represents an average of energy needs over a certain number of days,
and that the recommended energy intake is the amount of energy that should be ingested as a daily
average over a certain period of time. There is no implication that exactly this amount of energy must
be consumed every day, nor that the requirement and recommended intake are constant, day after day.
Neither is there any biological basis for defining the number of days over which the requirement or
intake must be averaged. As a matter of convenience, taking into account that physical activity and
eating habits may vary on some days of the week, periods of seven days are often used when
estimating the average daily energy expenditure and recommended daily intake.
2.1.2 Average requirement and inter-individual variation
Estimates of energy requirements are derived from measurements of individuals. Measurements of a
collection of individuals of the same gender and similar age, body size and physical activity are
grouped together to give the average energy requirement – or recommended level of dietary intake –
for a class of people or a population group. These requirements are then used to predict the
requirements and recommended levels of energy intake for other individuals with similar
characteristics, but on whom measurements have not been made. Although individuals in a given
class have been matched for characteristics that may affect requirements, such as gender, age, body
size, body composition and lifestyle, there remain unknown factors that produce variations among
individuals. Consequently, there is a distribution of requirements within the class or population group
(WHO, 1985) (Figure 2.1).
FIGURE 2.1 Distribution of energy requirements of a population group or class of individuals*
* It is assumed that individual requirements are randomly distributed about the mean requirement for the class of individuals, and that the distribution is Gaussian. Source. WHO, 1985.
For most specific nutrients, a certain excess of intake will not be harmful. Thus, when dietary
recommendations are calculated for these nutrients, the variation among individuals in a class or
population group is taken into account, and the recommended level of intake is an amount that will
meet or exceed the requirements of practically all individuals in the group. For example, the
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation 66
6
recommended safe level of intake for proteins is the average requirement of the population group,
plus 2 standard deviations. This approach cannot be applied to dietary energy recommendations,
because intakes that exceed requirements will produce a positive balance, which may lead to
overweight and obesity in the long term. A high level of energy intake that assures a low probability
of energy deficiency for most people (e.g. the average requirement plus 2 standard deviations) also
implies a high probability of obesity for most people owing to a dietary energy excess (Figure 2.2).
Therefore, in agreement with earlier reports, this expert consultation concluded that the descriptor of
the dietary energy intake that could be safely recommended for a population group is the estimated
average energy requirement of that group.
FIGURE 2.2 Probability that a particular energy intake is inadequate or excessive for an individual*
* Individuals are randomly selected among a class of people or a population group. The two probability curves overlap, so the level of energy intake that assures a low probability of dietary energy deficiency is the same level that implies a high probability of obesity owing to dietary energy excess. Source: WHO, 1985.
2.2 SOURCES OF DIETARY ENERGY Energy for the metabolic and physiological functions of humans is derived from the chemical energy
bound in food and its macronutrient constituents, i.e. carbohydrates, fats, proteins and ethanol, which
act as substrates or fuels. After food is ingested, its chemical energy is released and converted into
thermic, mechanical and other forms of energy.
This report refers to energy requirements that must be satisfied with an adequately balanced diet,
and does not make specific recommendations for carbohydrates, fats or proteins. Reports from other
FAO and WHO expert groups address those topics. Nevertheless, it should be noted that fats and
carbohydrates are the main sources of dietary energy, although proteins also provide important
amounts of energy, especially when total dietary energy intake is limited. Ethanol is not considered
part of a food system, but its contribution to total energy intake cannot be overlooked, particularly
among populations that regularly consume alcoholic beverages. Allowing for the mean intestinal
absorption, and for the nitrogenous portion of proteins that cannot be completely oxidized, the
average values of metabolizable energy provided by substrates in a mixed diet are 16.7 kJ (4 kcal) per
Principles and definitions
7
substrates determined by chemical analysis, or estimated from appropriate food composition tables. A recent related report from a FAO technical workshop provides more information on this topic (FAO, 2003).
2.3 COMPONENTS OF ENERGY REQUIREMENTS
Human beings need energy for the following:
Basal metabolism. This comprises a series of functions that are essential for life, such as cell function and replacement; the synthesis, secretion and metabolism of enzymes and hormones to transport proteins and other substances and molecules; the maintenance of body temperature; uninterrupted work of cardiac and respiratory muscles; and brain function. The amount of energy used for basal metabolism in a period of time is called the basal metabolic rate (BMR), and is measured under standard conditions that include being awake in the supine position after ten to 12 hours of fasting and eight hours of physical rest, and being in a state of mental relaxation in an ambient environmental temperature that does not elicit heat-generating or heat-dissipating processes. Depending on age and lifestyle, BMR represents 45 to 70 percent of daily total energy expenditure, and it is determined mainly by the individual’s age, gender, body size and body composition.
Metabolic response to food. Eating requires energy for the ingestion and digestion of food, and for the absorption, transport, interconversion, oxidation and deposition of nutrients. These metabolic processes increase heat production and oxygen consumption, and are known by terms such as dietary-induced thermogenesis, specific dynamic action of food and thermic effect of feeding. The metabolic response to food increases total energy expenditure by about 10 percent of the BMR over a 24-hour period in individuals eating a mixed diet.
Physical activity. This is the most variable and, after BMR, the second largest component of daily energy expenditure. Humans perform obligatory and discretionary physical activities. Obligatory activities can seldom be avoided within a given setting, and they are imposed on the individual by economic, cultural or societal demands. The term “obligatory” is more comprehensive than the term “occupational” that was used in the 1985 report (WHO, 1985) because, in addition to occupational work, obligatory activities include daily activities such as going to school, tending to the home and family and other demands made on children and adults by their economic, social and cultural environment.
Discretionary activities, although not socially or economically essential, are important for health, well-being and a good quality of life in general. They include the regular practice of physical activity for fitness and health; the performance of optional household tasks that may contribute to family comfort and well-being; and the engagement in individually and socially desirable activities for personal enjoyment, social interaction and community development.
Growth. The energy cost of growth has two components: 1) the energy needed to synthesize growing tissues; and 2) the energy deposited in those tissues. The energy cost of growth is about 35 percent of total energy requirement during the first three months of age, falls rapidly to about 5 percent at 12 months and about 3 percent in the second year, remains at 1 to 2 percent until mid-adolescence, and is negligible in the late teens.
Pregnancy. During pregnancy, extra energy is needed for the growth of the foetus, placenta and various maternal tissues, such as in the uterus, breasts and fat stores, as well as for changes in maternal metabolism and the increase in maternal effort at rest and during physical activity.
Lactation. The energy cost of lactation has two components: 1) the energy content of the milk secreted; and 2) the energy required to produce that milk. Well-nourished lactating women can derive part of this additional requirement from body fat stores accumulated during pregnancy.
2.4 CALCULATION OF ENERGY REQUIREMENTS
The total energy expenditure of free-living persons can be measured using the doubly labelled water technique (DLW) or other methods that give comparable results. Among these, individually calibrated heart rate monitoring has been successfully validated. Using these methods, measurements of total energy expenditure over a 24-hour period include the metabolic response to food and the energy cost
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
8
of tissue synthesis. For adults, this is equivalent to daily energy requirements. Additional energy for deposition in growing tissues is needed to determine energy requirements in infancy, childhood, adolescence and during pregnancy, and for the production and secretion of milk during lactation. It can be estimated from calculations of growth (or weight gain) velocity and the composition of weight gain, and from the average volume and composition of breastmilk.
2.4.1 Factorial estimates of total energy expenditure
When experimental data on total energy expenditure are not available, it can be estimated by factorial calculations based on the time allocated to activities that are performed habitually and the energy cost of those activities. Factorial calculations combine two or more components or “factors”, such as the sum of the energy spent while sleeping, resting, working, doing social or discretionary household activities, and in leisure. Energy spent in each of these components may in turn be calculated by knowing the time allocated to each activity, and its corresponding energy cost.
As discussed in the following sections of this report, the experimental measurement of total energy expenditure and the assessment of growth and tissue composition allow sound predictions to be made regarding energy requirements and dietary recommendations for infants and older children around the world. Special considerations and additional calculations assist the formulation of recommendations for children and adolescents with diverse lifestyles.
Total energy expenditure has also been measured in groups of adults, but this has been primarily in industrialized countries. Variations in body size, body composition and habitual physical activity among populations of different geographical, cultural and economic backgrounds make it difficult to apply the published results on a worldwide basis. Thus, in order to account for differences in body size and composition, energy requirements were initially calculated as multiples of BMR. They were then converted into energy units using a known BMR value for the population, or the mean BMR calculated from the population’s mean body weight. To account for differences in the characteristic physical activity of the associated lifestyles, energy requirements of adults were estimated by factorial calculations that took into account the times allocated to activities demanding different levels of physical effort.
The extra needs for pregnancy and lactation were also calculated using factorial estimates for the growth of maternal and foetal tissues, the metabolic changes associated with pregnancy and the synthesis and secretion of milk during lactation.
2.4.2 Expression of requirements and recommendations
Measurements of energy expenditure and energy requirement recommendations are expressed in units of energy (joules, J), in accordance with the international system of units. Because many people are still used to the customary usage of thermochemical energy units (kilocalories, kcal), both are used in this report, with kilojoules given first and kilocalories second, within parenthesis and in a different font (Arial 9). In tables, values for kilocalories are given in italic type.2
Gender, age and body weight are the main determinants of total energy expenditure. Thus, energy requirements are presented separately for each gender and various age groups, and are expressed both as energy units per day and energy per kilogram of body weight. As body size and composition also influence energy expenditure, and are closely related to basal metabolism, requirements are also expressed as multiples of BMR.
2 1 joule (J) is the amount of mechanical energy required to displace a mass of 1 kg through a distance of 1 m with an acceleration of 1 m per second (1 J = 1 kg × 1 m2 × 1 sec-2). Multiples of 1 000 (kilojoules, kJ) or 1 million (megajoules, MJ) are used in human nutrition. The conversion factors between joules and calories are: 1 kcal = 4.184 kJ, or conversely, 1 kJ = 0.239 kcal.
Principles and definitions
9
2.5 RECOMMENDATIONS FOR PHYSICAL ACTIVITY
A certain amount of activity must be performed regularly in order to maintain overall health and fitness,3 to achieve energy balance and to reduce the risk of developing obesity and associated diseases, most of which are associated with a sedentary lifestyle. This expert consultation therefore endorsed the proposition that recommendations for dietary energy intake must be accompanied by recommendations for an appropriate level of habitual physical activity. This report provides guidelines for desirable physical activity levels, and for the duration, frequency and intensity of physical exercise as recommended by various organizations with expertise in physical activity and health. It also emphasizes that appropriate types and amounts of physical activity can be carried out during the performance of either obligatory or discretionary activities and that recommendations must take into account the cultural, social and environmental characteristics of the target population.
2.6 GLOSSARY AND ABBREVIATIONS
In addition to those defined in the preceding sections, the following terms and abbreviations are used in this report. They are consistent with the definitions used in other related WHO and FAO documents (FAO, 2003; James and Schofield 1990; WHO, 1995).
Basal metabolic rate (BMR): The minimal rate of energy expenditure compatible with life. It is measured in the supine position under standard conditions of rest, fasting, immobility, thermoneutrality and mental relaxation. Depending on its use, the rate is usually expressed per minute, per hour or per 24 hours.
Body mass index (BMI): The indicator of weight adequacy in relation to height of older children, adolescents and adults. It is calculated as weight (in kilograms) divided by height (in meters), squared. The acceptable range for adults is 18.5 to 24.9, and for children it varies with age.
Doubly labelled water (DLW) technique: A method used to measure the average total energy expenditure of free-living individuals over several days (usually 10 to 14), based on the disappearance of a dose of water enriched with the stable isotopes 2H and 18O.
Energy requirement (ER): The amount of food energy needed to balance energy expenditure in order to maintain body size, body composition and a level of necessary and desirable physical activity, and to allow optimal growth and development of children, deposition of tissues during pregnancy, and secretion of milk during lactation, consistent with long-term good health. For healthy, well-nourished adults, it is equivalent to total energy expenditure. There are additional energy needs to support growth in children and in women during pregnancy, and for milk production during lactation.
Heart rate monitoring (HRM): A method to measure the daily energy expenditure of free-living individuals, based on the relationship of heart rate and oxygen consumption and on minute-by-minute monitoring of heart rate.
Total energy expenditure (TEE): The energy spent, on average, in a 24-hour period by an individual or a group of individuals. By definition, it reflects the average amount of energy spent in a typical day, but it is not the exact amount of energy spent each and every day.
Physical activity level (PAL): TEE for 24 hours expressed as a multiple of BMR, and calculated as TEE/BMR for 24 hours. In adult men and non-pregnant, non-lactating women, BMR times PAL is equal to TEE or the daily energy requirement.
3 The term “fitness” encompasses cardiorespiratory health, appropriate body composition (including fat distribution), muscular strength, endurance and flexibility. Fitness can generally be described as the ability to perform moderate to vigorous physical activity without becoming excessively tired.
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
10
Physical activity ratio (PAR): The energy cost of an activity per unit of time (usually a minute or an hour) expressed as a multiple of BMR. It is calculated as energy spent in an activity/BMR, for the selected time unit.
REFERENCES FAO. 2003. Food energy – methods of analysis and conversion factors. Report of a technical workshop. FAO Food and Nutrition Paper No. 77. Rome. James, W.P.T. & Schofield, E.C. 1990. Human energy requirements. A manual for planners and nutritionists. Oxford, UK, Oxford Medical Publications under arrangement with FAO. WHO. 1985. Energy and protein requirements: Report of a joint FAO/WHO/UNU expert consultation. WHO Technical Report Series No. 724. Geneva. WHO. 1995. Physical status: The use and interpretation of anthropometry. Report of a WHO expert committee. WHO Technical Report Series No. 854. Geneva.
Energy requirements of infants from birth to 12 months
11
3. ENERGY REQUIREMENTS OF INFANTS
FROM BIRTH TO 12 MONTHS
The principle of calculating energy requirements from total energy expenditure (TEE) plus the energy needs for growth applies to infants and children of all ages. However, the previous FAO/WHO/UNU expert consultation (WHO, 1985) estimated the energy requirements of infants from the observed intakes of healthy children growing normally, largely owing to the lack of sufficient information on total energy expenditure. For the last report, data on measurements of infants and children were compiled from studies of infants in Canada, Sweden, the United Kingdom and the United States (Whitehead, Paul and Cole, 1981). Results from developing countries were not included in the analysis “to ensure that the intakes represented those of groups of children who, on the average, were growing along the fiftieth percentile of the WHO reference standard”. An additional 5 percent was added to compensate for a possible methodological bias in the calculation of energy intakes.
Scientific information generated in the intervening years has allowed the present consultation to base its estimates and recommendations for infants on energy expenditure plus the energy needs for growth. This assumes that the energy intake of infants is self-regulated and matches energy needs (Fomon, 1974; Dewey and Lönnerdal, 1986). In keeping with the principles followed by preceding expert groups, it was decided to base the analyses, conclusions and recommendations on results of studies carried out on healthy, well-nourished, non-stunted infants born at full term with adequate birth weight, and growing along the trajectory of the WHO reference standards (WHO, 1983). This permits the prescription of dietary recommendations consistent with the optimal growth of healthy, well-nourished infant populations. Special considerations must be made for groups with particular needs, such as undernourished infants and those with low weight or size at birth.
3.1 MEASUREMENT OF TOTAL ENERGY EXPENDITURE
The use of the doubly labelled water (DLW) (2H2 18O) technique to calculate total production of
carbon dioxide (CO2) over several days and, from this, total energy expenditure was originally developed for use in small mammals (Lifson, Gordon and McClintock, 1955), and its application was later validated in humans (Schoeller and van Santen, 1982; Klein et al., 1984; Coward et al., 1984). Although questions have been raised about the appropriateness of the assumptions used for the calculation of TEE, DLW is currently considered the most accurate technique for measuring TEE in free-living individuals. TEE measured by this method includes basal metabolism, the metabolic response to food, thermoregulatory needs, physical activity costs, and the energy cost to synthesize growing tissues. Consequently, energy requirements are calculated as the sum of TEE plus the energy deposited as protein and fat in growing tissues and organs.
This consultation examined an analysis of 13 studies with DLW performed on a total of 417 healthy, well-nourished, non-stunted infants of from 0 to 12 months of age (Butte, 2001). Eleven investigations were carried out in the United Kingdom (Lucas et al., 1987; Roberts et al., 1988; Davies, Ewing and Lucas, 1989; Wells and Davies, 1995; Wells, Cole and Davies, 1996; Davies et al., 1997), the United States (Butte et al., 1990; Stunkard et al., 1999; Butte et al., 2000b) and the Netherlands (de Bruin et al., 1998), one in Chile (Salazar et al., 2000) and one in China (Jiang et al., 1998). Several studies conducted repeated measurements of TEE at intervals of two to three months, increasing the number of TEE data points to 854. One such study showed that the coefficient of variation among individuals was fairly uniform from three to 24 months of age, ranging from 15 to 21 percent for TEE/day (average: 18 percent), and from 13 to 17 percent for TEE/kg/day (average: 15 percent) (Butte et al., 2000b). The average inter-individual variation was similar to that observed among older children (19 percent for TEE/day, and 17 percent for TEE/kg/day; see section 4.1).
3.2 EQUATIONS TO PREDICT ENERGY EXPENDITURE
Longitudinal measurements of TEE with DLW at three-month intervals for the first two years of life on 76 healthy infants (40 breastfed and 36 formula-fed) showed that there is a good linear relationship
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
12
between TEE and body weight (Butte et al., 2000b). TEE was significantly affected by age, gender, weight and length. Age, weight and height were all good predictors of TEE, with a slight advantage for weight. Because the three parameters were highly correlated (r = 0.91 – 0.96), and there were no independent effects of age, gender and length when weight was used as the predictor, the latter was used to develop the following equation (Butte, 2001), which is graphically displayed in Figure 3.1.
TEE (MJ/day) = – 0.416 + 0.371 kg; n = 320, r = 0.85, see = 0.456 MJ/day (109 kcal/day) TEE (kcal/day) = – 99.4 + 88.6 kg (n = number of observations; see = standard error of estimate)
FIGURE 3.1
Linear relationship and 95 percent confidence and prediction intervals of equation to predict TEE from body weight in healthy infants, one to 24 months old
TEE (MJ/d) = – 0.416 + 0.371 kg; n = 320, r = 0.85, see = 0.456 MJ/d (109 kcal/d). TEE (kcal/d) = – 99.4 + 88.6 kg. Source: Butte, 2001.
The relationship between TEE and weight in the 13 studies mentioned in Section 3.1 was explored using the mean values for TEE and body weight. Some studies included longitudinal or cross- sectional data at various ages throughout infancy, or from groups of either breastfed or formula-fed infants. A total of 40 sets of TEE and body weight values, weighted for sample size, gave the following linear regression equation, which does not differ significantly from that shown above:
TEE (MJ/day) = – 0.399 + 0.369 kg; n = 40, r = 0.99, see = 0.527 MJ/day (126 kcal/day) TEE (kcal/day) = – 95.4 + 88.3 kg
As the equation was derived from the mean values of each study, the regression coefficient and standard error of estimate (see) do not reflect individual variation.
3.2.1 Breastfed and formula-fed infants
Four studies with breastfed and formula-fed infants showed that the formula-fed infants had higher TEE during the first year of life (Butte et al., 1990; Butte et al., 2000b; Jiang et al., 1998; Davies et al., 1990). Compared with their breastfed counterparts, formula-fed infants had on average 12, 7, 6 and 3 percent higher TEE at three, six, nine and 12 months of age, respectively. At 18 and 24 months, there was no difference between infants who still received breastmilk and those who did not (Butte, 2001). The equations to predict TEE from body weight are as follows:
J J
J
JJ
J
J J
J JJ
JJ
J
J
J J
J
J J
J J
J
J J J J
J
JJ
J
J
J J
J
J
J
J
J
J JJJ
J
J
J
J
J J
J
J
JJ
J
J
J J
J
J
J
J
J J
J J
J
JJ
J
J J
J
J
J
J J
J
J
J J
J
J
J
J
J
J J
J
J J
J
J
J J
J
J
J
JJ J
J
J
J
J
J
J
JJ
J
J
J
J
J
J
J
J
J
J J
J
J
J
J
J
J
J J
JJ
JJ J
J
J
J
J
J J
J
J
J
J
J JJ
J J
J JJ
J
J J
J
J
J
J
J
J J
J
J
J
J J
J
J
J
J
J
J
J
J
J
J
J JJ
JJ J J J
JJ
J
J J J J
J
J
J
J
J
J J J
J
J
J
J
J
J
J
J
J
J
J
J J
J
J
J
J
JJJ
J
J
J
J
J J
JJ
J
J
J J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J J
J
J
J J
J
J
J J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J J
J
JJ
J
J
J J
J
J
J
J
J
J
J
J
JJ
J
J
J
J
J
J
J
J
J
J J
JJ
J
J
J J
J
J
JJ
J
J
J
J
J
J
J
J
0
1000
2000
3000
4000
5000
6000
0
200
400
600
800
1000
1200
1400
4 6 8 10 12 14
T E
E (
k J /d
)
T E
E (
k c a
l/ d
)
Weight (kg)
Linear regression
95% Confidence interval
95% Prediction interval
Energy requirements of infants from birth to 12 months
13
Breastfed:
TEE (MJ/day) = – 0.635 + 0.388 kg; n = 195, r = 0.87, see = 0.453 MJ/day (108 kcal/day) TEE (kcal/day) = – 152.0 + 92.8 kg
Formula-fed:
TEE (MJ/day) = – 0.122 + 0.346 kg; n = 125, r = 0.85, see = 0.463 MJ/day (110 kcal/day) TEE (kcal/day) = – 29.0 + 82.6 kg
3.3 ENERGY NEEDS FOR GROWTH
Growth is a sensitive indicator of whether an infant’s energy requirements are satisfied. Energy demands for growth constitute about 35 percent of the total energy requirement during the first three months of life (40 percent in the first month), this proportion is halved in the next three months (i.e. to about 17.5 percent), and further reduced to one-third of that during the ensuing six months (i.e. to less than 6 percent) and to only 3 percent at 12 months. Energy for growth falls to less than 2 percent of daily requirements in the second year, remains between 1 and 2 percent until mid-adolescence, and gradually disappears by 20 years of age.
Energy needs for growth have two components: 1) the energy used to synthesize growing tissues, which is part of the total energy expenditure measured with DLW; and 2) the energy deposited in those tissues, basically as fat and protein, because carbohydrate content is insignificant. Hence, energy requirements in infancy can be calculated by adding the energy deposited in growing tissues to TEE.
Much previous knowledge on the energy cost of growth was based on studies in pre-term infants or in children recovering from malnutrition, and used energy balance and the two-component body composition techniques (WHO, 1985; Butte, Wong and Garza, 1989). Methodological advances have allowed a better assessment of body composition changes during infancy through serial measurements of total body electrical conductivity (de Bruin et al., 1998), or with a multi-component body composition model based on measurements of total body water, total body potassium and bone mineral content (Butte et al., 2000a). This permits calculation of the gains in protein and fat, as well as of the corresponding energy deposition assuming that the energy equivalents of protein and fat are 23.6 and 38.7 kJ/g (5.65 and 9.25 kcal/g), respectively. As Table 3.1 shows, energy accrued per gram of weight gain decreased from approximately 26 kJ (6.3 kcal) in the first three months of life to about 10 kJ (2.3 kcal) at nine to 12 months.
TABLE 3.1
Protein, fat and energy deposition during growth in the first year of life Energy accrued in normal growth* Age
months Protein gain
g/d Fat mass gain
g/d Weight gain
g/d kJ/g kcal/g
Boys
0–3 2.6 19.6 32.7 25.1 6.0
3–6 2.3 3.9 17.7 11.6 2.8
6–9 2.3 0.5 11.8 6.2 1.5
9–12 1.6 1.7 9.1 11.4 2.7
Girls
0–3 2.2 19.7 31.1 26.2 6.3
3–6 1.9 5.8 17.3 15.6 3.7
6–9 2.0 0.8 10.6 7.4 1.8
9–12 1.8 1.1 8.7 9.8 2.3 * Energy equivalents: 1 g protein = 23.6 kJ (5.65 kcal); 1 g fat = 38.7 kJ (9.25 kcal).
Source: Butte et al., 2000a.
T A
B L E
3 .2
E n
e rg
y r
e q
u ir
e m
e n
ts o
f in
fa n
ts d
u ri
n g
t h
e f
ir s
t y e
a r
o f
li fe
*
T o
ta l
e n
e rg
y e
x p
e n
d it
u re
a
E n
e rg
y d
e p
o s it
io n
b
D a il
y e
n e rg
y r
e q
u ir
e m
e n
tc A
g e
m o n th
s
W e ig
h t
k g
W e ig
h t
g a in
g /d
M J /d
k c a l/ d
M J /d
k c a
l/ d
M
J /d
k c a l/ d
k J /k
g /d
k c a l/ k g /d
B o
y s
0 –
1
4 .5
8
3 5
.2
1 .2
8 2
3 0
6
0 .8
8 4
2 1
1
2 .1
6 6
5 1
8
4 7 3
1
1 3
1 –
2
5 .5
0
3 0
.4
1 .6
2 3
3 8
8
0 .7
6 4
1 8
3
2 .3
8 7
5 7
0
4 3 4
1
0 4
2 –
3
6 .2
8
2 3
.2
1 .9
1 2
4 5
7
0 .5
8 2
1 3
9
2 .4
9 4
5 9
6
3 9 7
9
5
3 –
4
6 .9
4
1 9
.1
2 .1
5 7
5 1
5
0 .2
2 4
5 3
2 .3
8 0
5 6
9
3 4 3
8
2
4 –
5
7 .4
8
1 6
.1
2 .3
5 7
5 6
3
0 .1
8 9
4 5
2 .5
4 6
6 0
8
3 4 0
8
1
5 –
6
7 .9
3
1 2
.8
2 .5
2 4
6 0
3
0 .1
5 0
3 6
2 .6
7 4
6 3
9
3 3 7
8
1
6 –
7
8 .3
0
1 1
.0
2 .6
6 1
6 3
6
0 .0
6 9
1 7
2 .7
3 0
6 5
3
3 2 9
7
9
7 –
8
8 .6
2
1 0
.4
2 .7
8 0
6 6
4
0 .0
6 5
1 6
2 .8
4 5
6 8
0
3 3 0
7
9
8 –
9
8 .8
9
9 .0
2
.8 8
0
6 8
8
0 .0
5 7
1 4
2 .9
3 6
7 0
2
3 3 0
7
9
9 – 1
0
9 .1
3
7 .9
2
.9 6
9
7 1
0
0 .0
8 9
2 1
3 .0
5 8
7 3
1
3 3 5
8
0
1 0 –
1 1
9 .3
7
7 .7
3
.0 5
8
7 3
1
0 .0
8 7
2 1
3 .1
4 5
7 5
2
3 3 6
8
0
1 1 –
1 2
9 .6
2
8 .2
3
.1 5
0
7 5
3
0 .0
9 3
2 2
3 .2
4 3
7 7
5
3 3 7
8
1
G ir
ls
0 –
1
4 .3
5
2 8
.3
1 .1
9 7
2 8
6
0 .7
4 6
1 7
8
1 .9
4 2
4 6
4
4 4 7
1
0 7
1 –
2
5 .1
4
2 5
.5
1 .4
9 0
3 5
6
0 .6
7 2
1 6
1
2 .1
6 2
5 1
7
4 2 1
1
0 1
2 –
3
5 .8
2
2 1
.2
1 .7
4 2
4 1
6
0 .5
5 9
1 3
4
2 .3
0 1
5 5
0
3 9 5
9
4
3 –
4
6 .4
1
1 8
.4
1 .9
6 0
4 6
9
0 .2
8 5
6 8
2 .2
4 5
5 3
7
3 5 0
8
4
4 –
5
6 .9
2
1 5
.5
2 .1
4 9
5 1
4
0 .2
3 9
5 7
2 .3
8 9
5 7
1
3 4 5
8
3
5 –
6
7 .3
5
1 2
.8
2 .3
0 9
5 5
2
0 .1
9 9
4 7
2 .5
0 7
5 9
9
3 4 1
8
2
6 –
7
7 .7
1
1 1
.0
2 .4
4 2
5 8
4
0 .0
8 3
2 0
2 .5
2 5
6 0
4
3 2 8
7
8
7 –
8
8 .0
3
9 .2
2
.5 6
1
6 1
2
0 .0
6 9
1 7
2 .6
3 0
6 2
9
3 2 8
7
8
8 –
9
8 .3
1
8 .4
2
.6 6
5
6 3
7
0 .0
6 3
1 5
2 .7
2 8
6 5
2
3 2 8
7
8
9 – 1
0
8 .5
5
7 .7
2
.7 5
4
6 5
8
0 .0
7 4
1 8
2 .8
2 8
6 7
6
3 3 1
7
9
1 0 –
1 1
8 .7
8
6 .6
2
.8 3
9
6 7
9
0 .0
6 3
1 5
2 .9
0 2
6 9
4
3 3 1
7
9
1 1 –
1 2
9 .0
0
6 .3
2
.9 2
0
6 9
8
0 .0
6 0
1 4
2 .9
8 1
7 1
2
3 3 1
7
9
* C
a lc
u la
te d f
ro m
l in
e a
r re
g re
s s io
n a
n a
ly s is
o f to
ta l e
n e
rg y e
x p e
n d
it u
re o
n w
e ig
h t,
p lu
s a
llo w
a n
c e
f o
r e
n e
rg y d
e p
o s it io
n i n t
is s u
e s d
u ri n
g g
ro w
th .
a T
E E
( M
J /d
) =
– 0
.4 1 6
+ 0
.3 7
1 k
g (
s e c ti o
n 3
.2 ).
b W
e ig
h t g
a in
× e
n e rg
y a
c c ru
e d
i n
n o
rm a l g
ro w
th (
T a
b le
3 .1
).
c R
e q
u ir
e m
e n t
= t
o ta
l e
n e
rg y e
x p
e n d
it u
re +
e n
e rg
y d
e p o
s it io
n .
S
o u
rc e
s :
B u tt
e ,
2 0 0
1 . W
e ig
h t a
n d
w e
ig h t
g a in
d a
ta f
ro m
W H
O ,
1 9
9 4
.
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
14
Energy requirements of infants from birth to 12 months
15
3.4 CALCULATION OF ENERGY REQUIREMENTS
Table 3.2 shows the average energy requirements of infants from one to 12 months of age, combining the needs of breastfed and formula-fed infants. TEE was calculated with the predictive linear equations described in section 3.2 and the median weight for age of the WHO pooled breastfed data set (WHO, 1994). The rate of median weight gain at monthly intervals was calculated from the same source. Energy deposited in growing tissues was estimated by multiplying the monthly weight gain by the mean energy accrued in each three-month period (Table 3.1). The sum of TEE and energy deposition is the mean daily energy requirement (in MJ or kcal). It is calculated as energy units per kilogram of body weight, dividing the daily requirement by the median weight at each month of age.
Breastmilk is the best food for infants, and exclusive breastfeeding is strongly recommended during the first six months of life, followed by a combination of breastmilk and complementary foods throughout infancy. As TEE is lower among breastfed than formula-fed infants during the first year of life, the energy requirements of breastfed infants are also lower. This is illustrated in Table 3.3, in which requirements are calculated for breastfed and formula-fed infants with the same body weights using the predictive equations described in section 3.2.1. For the purpose of simplicity, the values have been rounded off to the closest 5 kJ/kg/day, or 1 kcal/kg/day. These figures are consistent with the fact that a healthy woman can produce enough milk to provide the energy required by a healthy, exclusively breastfed infant of up to six months of age.
TABLE 3.3
Energy requirements of breastfed, formula-fed and all infants *
Age Breastfed a Formula-fed
b All (breast- and formula-fed)
c
Months Boys Girls Mean Boys Girls Mean Boys Girls Mean
kJ/kg/d
1 445 415 430 510 490 500 475 445 460
2 410 395 405 460 455 460 435 420 430
3 380 375 380 420 420 420 395 395 395
4 330 335 330 360 370 365 345 350 345
5 330 330 330 355 365 360 340 345 345
6 325 330 330 350 355 355 335 340 340
7 320 315 320 340 340 340 330 330 330
8 320 320 320 340 340 340 330 330 330
9 325 320 320 340 340 340 330 330 330
10 330 325 325 340 340 340 335 330 335
11 330 325 325 340 340 340 335 330 335
12 330 325 330 345 340 340 335 330 335
kcal/kg/d
1 106 99 102 122 117 120 113 107 110
2 98 95 97 110 108 109 104 101 102
3 91 90 90 100 101 100 95 94 95
4 79 80 79 86 89 87 82 84 83
5 79 79 79 85 87 86 81 82 82
6 78 79 78 83 85 84 81 81 81
7 76 76 76 81 81 81 79 78 79
8 77 76 76 81 81 81 79 78 79
9 77 76 77 81 81 81 79 78 79
10 79 77 78 82 81 81 80 79 80
11 79 77 78 82 81 81 80 79 80
12 79 77 78 82 81 81 81 79 80
* Numbers rounded to the closest 5 kJ/kg/d, and 1 kcal/kg/d, using the mean body weight and energy deposition in Table 3.1 and the following predictive equations for TEE: a TEE (MJ/kg/d) = (– 0.635 + 0.388 weight) / weight.
b TEE (MJ/kg/d) = (– 0.122 + 0.346 weight) / weight.
c TEE (MJ/kg/d) = (– 0.416 + 0.371 weight) / weight.
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
16
3.4.1 Comparison with previous requirements
Compared with the values in the 1985 FAO/WHO/UNU report, energy requirements proposed by this consultation are about 12 percent lower in the first three months of life, 17 percent lower from three to nine months, and 20 percent lower from nine to 12 months (Table 3.4 and Figure 3.2). The requirements for breastfed infants are 17, 20 and 22 percent lower than the 1985 estimates at ages 0 to three, three to nine and nine to 12 months, respectively. That the 1985 consultation overestimated requirements of this group had already been suggested from an analysis of 3 573 data points of energy intakes of well-nourished infants recorded after 1980 (Butte, 1996).
TABLE 3.4
Comparison of present estimates of energy requirements (kJ/kg/d) of infants with those calculated in the previous (1985) FAO/WHO/UNU report
Age Present estimates % difference from 1985
months All infants Breastfed
1985 estimates All infants Breastfed
0–1 460 430 519 -11 -17
1–2 430 405 485 -11 -16
2–3 395 380 456 -13 -17
3–4 345 330 431 -20 -23
4–5 345 330 414 -17 -20
5–6 340 330 404 -16 -18
6–7 330 320 397 -17 -19
7–8 330 320 395 -16 -19
8–9 330 320 397 -17 -19
9–10 335 325 414 -19 -21
10–11 335 325 418 -20 -22
11–12 335 330 437 -23 -24
3.4.2 Basal metabolic rate and physical activity level
The 1981 FAO/WHO/UNU expert consultation estimated the energy requirements of adults as multiples of BMR (WHO, 1985). This was later called “physical activity level” (PAL) in a manual commissioned by FAO for the calculation of human energy requirements. PAL is defined as the total energy required over 24 hours divided by the basal metabolic rate over 24 hours (James and Schofield, 1990). The 2001 expert consultation upheld this approach to estimating requirements for adults (section 5). However, the approach must be used with caution or avoided altogether in relation to the energy requirements of infants and young children, as PAL values may cause confusion owing to differences in the factors that determine energy requirements among children and among adults. In non-pregnant, non-lactating women energy requirements are equal to TEE. In children, however, energy requirements are equal to TEE plus energy accrued or deposited during growth (Eg). These differences are quantitatively small after two years of age, when Eg represents less than 1 or 2 percent of the total energy requirement of the child; but they are increasingly larger at less than two years of age. For example, Eg is about 40 and 23 percent of the energy requirement in the first and third months of life, respectively. Consequently, energy requirement expressed as a function of BMR is much higher (i.e. > 2.0 at one month of age and > 1.7 at three months) in comparison with a PAL value (which is based on measured total energy expenditure) of 1.2 and 1.3 respectively.
BMR of term infants has been studied extensively producing variable results that range from 180 to 250 kJ/kg/day (43 to 60 kcal/kg/day) (Butte, 2001). This high variability has been attributed to biological differences, mainly in body composition at different stages of infancy, and to differences in methods and experimental conditions. For example, some investigators measured “basal” metabolism in infants who were sleeping spontaneously or under the effect of a sedative, which decreases BMR, and others did the measurements in the fed state, which increases BMR. The 1981 expert consultation endorsed the use of predictive equations to estimate the BMR of children under three years of age, derived from approximately 300 data points obtained by a variety of investigators using different methods and under diverse conditions (Schofield, Schofield and James, 1985). These equations
Energy requirements of infants from birth to 12 months
17
underestimate BMR by about 5 to 12 percent from one to nine months of age (Butte, 1989; Wells et al., 1996). This could partly be owing to the lack of uniformity in the conditions when BMR was measured, and hence could have an important impact on the calculation of PAL.
FIGURE 3.2
Comparison of present estimates of energy requirements of infants (combining breastfed and formula-fed infants) with those in the 1985 FAO/WHO/UNU report
Source: Butte, 2001.
3.5 CATCH-UP GROWTH
Assessment of requirements and dietary recommendations for premature, small for gestational age and malnourished infants is beyond the scope of this report. The consultation recognized, however, that many populations around the world have large numbers of newborns with intrauterine growth retardation, and malnourished children less than one year of age. In addition to proper health, social and emotional support, these infants require special nutritional care for a rapid, catch-up growth that will allow them to attain the expected weight and height of normal children born with adequate size at term, and who have never been malnourished. To this end, high growth velocities can be achieved that, compared with the weight gain of normal, well-nourished children, can be up to 20 times higher among underweight, wasted children and about three to five times higher among short, stunted infants.
Diets for catch-up growth must provide all nutrients and energy sources in amounts that are proportionally higher than those required by well-nourished infants of adequate size. However, it is difficult to generalize about the quantitative energy requirements for catch-up growth, as these must often be assessed on an individual basis. Dietary needs, and hence recommendations, may vary with the extent of and the causes of growth retardation, which include the duration of pregnancy; metabolic, physiological and nutritional alterations during intrauterine development; pre- and postpartum infections; and pre- and postpartum primary or secondary malnutrition. The age of onset and duration of the causes leading to growth retardation must also be considered for appropriate dietary interventions. Because the target body weight and length are not fixed but increase with time in a growing child, the longer the period of growth deficit, the greater the gap to be filled.
There are conflicting reports on whether BMR is depressed in severely malnourished children (Montgomery, 1962; Parra et al., 1973) and rises in the early stages of nutritional rehabilitation. Studies with DLW (Fjeld and Schoeller, 1988) suggest that during the early phases of recovery TEE is about 5 to 10 percent higher than expected in well-nourished children, and this increment disappears in the late stages of nutritional treatment. This is probably a reflection of the accelerated rates of tissue
0
20
40
60
80
100
120
0 1 2 3 4 5 6 7 8 9 10 11 12
E n
e rg
y r
e q
u ir e
m e
n t
(k c a l/ k g /d
)
Age (mo)
1985 FAO/WHO/UNU recommendation
Boys: Proposed energy requirement
Girls: Proposed energy requirement
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
18
synthesis and deposition. There are also reports with varying results on whether the rate of catch-up growth influences the composition of weight gain in children treated for severe malnutrition. Most agree that during the early phases of recovery, 15 to 20 percent of the weight gain seems to be protein, with the rest equally divided between fat and water (Graham et al., 1969; MacLean and Graham, 1980; Fjeld, Schoeller and Brown, 1989).
The influence of malnutrition and the effect of infections on energy requirements are further addressed in section 4.6 of this report, where tentative recommendations are made for populations with high prevalence of infant malnutrition.
REFERENCES Butte, N.F. 1989. Basal metabolism of infants. In B. Schürch and N.S. Scrimshaw, eds. Activity, energy expenditure and energy requirements of infants and children, pp. 117–137. Switzerland, Nestlé Foundation. Butte, N.F. 1996. Energy requirements of infants. Eur. J. Clin. Nutr., 50: 24S–36S. Butte, N.F. 2001. Energy requirements of infants. Background paper prepared for the joint FAO/WHO/UNU Expert Consultation on Energy in Human Nutrition. Butte, N.F., Wong, W.W. & Garza, C. 1989. Energy cost of growth during infancy. Proc. Nutr. Soc., 48: 303– 312. Butte, N.F., Wong, W.W., Ferlic, L. & Smith, E.O. 1990. Energy expenditure and deposition of breast-fed and formula-fed infants during early infancy. Pediatr. Res., 28: 631–640. Butte, N.F., Hopkinson, J.M., Wong, W.W., Smith, E.O. & Ellis, K.J. 2000a. Body composition during the first two years of life: An updated reference. Pediatr. Res., 47: 578–585. Butte, N.F., Wong, W.W., Hopkinson, J.M., Heinz, C.J., Mehta, N.R. & Smith, E.O. 2000b. Energy requirements derived from total energy expenditure and energy deposition during the first two years of life. Am. J. Clin. Nutr., 72: 1558–1569. Coward, W.A., Prentice, A.M., Murgatroyd, P.R. et al. 1984. Measurement of CO2 and water production rate in man using 2H18O-labeled H2O – comparison between calorimeter and isotope values. In Human energy metabolism, physiological activity and energy expenditure in epidemiological research based upon direct and indirect calorimetry. Euronut. Report, 5: 126–128. Davies, P.S.W., Day, J.M.E. & Lucas, A. 1991. Energy expenditure in early infancy and later body fatness. Int. J. Obes., 15: 727–731. Davies, P.S.W., Ewing, G. & Lucas, A. 1989. Energy expenditure in early infancy. Br. J. Nutr., 62: 621–629. Davies, P.S.W., Ewing, G., Coward, W.A. & Lucas, A. 1990. Energy metabolism in breast-fed and formula- fed infants. In S.A. Atkinson, L.Å. Hanson and R.K. Chandra, eds. Breast-feeding, nutrition, infection and infant growth in developed and emerging countries. St. John’s, Newfoundland, Canada. Arts Biomedical, 521. Davies, P.S.W., Wells, J.C.K, Hinds, A., Day, J.M.E. & Laidlaw, A. 1997. Total energy expenditure in 9 month and 12 month infants. Eur. J. Clin. Nutr., 51: 249–252. de Bruin, N.C., Degenhart, H.J., Gàl, S., Westerterp, K.R., Stijnen, T. & Visser, H.K.A. 1998. Energy utilization and growth in breast-fed and formula-fed infants measured prospectively during the first year of life. Am. J. Clin. Nutr., 67: 885–896. Dewey, K.G. & Lönnerdal, B. 1986. Infant self-regulation of breast milk intake. Acta Paediatr. Scand., 75: 893–898. Fjeld, C.R. & Schoeller, D.A. 1988. Energy expenditure of malnourished children during catch-up growth. Proc. Nutr. Soc., 47: 227–231. Fjeld, C.R., Schoeller, D.A. & Brown, K.H. 1989. Body composition of children recovering from severe protein-energy malnutrition at two rates of catch-up growth. Am. J. Clin. Nutr., 50: 1266–1275. Fomon, S.J. 1974. Infant nutrition. Philadelphia, Pennsylvania, USA, W.B. Saunders. Graham, G.G., Cordano, A., Blizzard, R.M. & Cheek, D.B. 1969. Infantile malnutrition. Changes in body composition during rehabilitation. Pediatr. Res., 3: 579–589. James, W.P.T. & Schofield, E.C. 1990. Human energy requirements. A manual for planners and nutritionists. Oxford, UK, Oxford Medical Publications under arrangement with FAO. Jiang, Z., Yan, Q., Su, Y., Heson, K.J., Thélin, A., Piguet-Welsch, C., Ritz, P. & Ho, Z. 1998. Energy expenditure of Chinese infants in Guangdong Province, south China, determined with use of the doubly labelled water method. Am. J. Clin. Nutr., 67: 1256–1264. Klein, P.D., James, W.P.T., Wong, W.W. et al. 1984. Calorimetric validation of the doubly labelled water method for determination of energy expenditure in man. Hum. Nutr. Clin. Nutr., 35C: 95–106. Lifson, B., Gordon, G.B. & McClintock, R. 1955. Measurement of total carbon dioxide production by means of D2
18O. J. Appl. Physiol., 7: 704–710. Lucas, A., Ewing, G., Roberts, S.B. & Coward, W.A. 1987. How much energy does the breast fed infant consume and expend? Br. Med. J., 295: 75–77.
Energy requirements of infants from birth to 12 months
19
MacLean, W.C. Jr. & Graham, G.G. 1980. The effect of energy intake on nitrogen content of weight gained by recovering malnourished infants. Am. J. Clin. Nutr., 33: 903–909. Montgomery, R.D. 1962. Changes in the basal metabolic rate of the malnourished infant and their relation to body composition. J. Clin. Invest., 41: 1653–1663. Parra, A., Garza, C., Garza, Y., Saravia, J.L., Hazlewood, C.F. & Nichols, B.L. 1973. Changes in growth hormone, insulin, and thyroxine values, and in energy metabolism of marasmic infants. J. Pediatr., 82: 133–142. Roberts, S.B., Savage, J., Coward, W.A., Chew, B. & Lucas, A. 1988. Energy expenditure and intake in infants born to lean and overweight mothers. N. Engl. J. Med., 318: 461–466. Salazar, G., Vio, F., Garcia, C., Aguirre, E. & Coward, W.A. 2000. Energy requirements in Chilean infants. Arch. Dis. Child Fetal Neonatal, 83: F120–F123. Schoeller, D.A. & van Santen, E. 1982. Measurement of energy expenditure in humans by doubly labelled water method. J. Appl. Physiol., 53: 955–959. Schofield, W.N., Schofield, C. & James, W.P.T. 1985. Basal metabolic rate – review and prediction, together with an annotated bibliography of source material. Human Nutr. Clin. Nutr., 39C (suppl. 1): 1–96. Stunkard, A.J., Berkowitz, R.I., Stallings, V.A. & Schoeller, D.A. 1999. Energy intake, not energy output, is a determinant of body size in infants. Am. J. Clin. Nutr., 69: 524–530. Wells, J.C.K., Cole, T.J. & Davies, P.S.W. 1996. Total energy expenditure and body composition in early infancy. Arch. Dis. Child., 75: 423–426. Wells, J.C.K. & Davies P.S.W. 1995. Energy cost of physical activity in twelve week old infants. Am. J. Hum. Biol., 7: 85–92. Wells, J.C.K., Joughlin, C., Crisp, J.A., Cole, T.J. & Davies, P.S.W. 1996. Comparison of measured sleeping metabolic rate and predicted basal metabolic rate in the first year of life. Acta Pædiatr., 85: 1013–1018. Whitehead, R.G., Paul, A.A. & Cole, T.J. 1981. A critical analysis of measured food energy intakes during infancy and early childhood in comparison with current international recommendations. J. Hum. Nutr., 35: 339– 348. WHO. 1983. Measuring change in nutritional status. Geneva. WHO. 1985. Energy and protein requirements: Report of a joint FAO/WHO/UNU expert consultation. WHO Technical Report Series No. 724. Geneva. WHO. 1994. WHO Working Group on Infant Growth. An evaluation of infant growth. Geneva, Nutrition Unit, WHO.
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
20
4. ENERGY REQUIREMENTS OF
CHILDREN AND ADOLESCENTS
In the 20 years since the 1981 joint FAO/WHO/UNU consultation (WHO, 1985), significant experimental evidence has been collected on the TEE of children and adolescents. This makes it possible to estimate energy requirements from measurements of TEE and energy needs for growth, rather than from food intake data or from estimates of time allocation and energy costs, as were previously used. The 2001 expert consultation analysed a number of studies on TEE, growth and habitual activity patterns of children and adolescents in different parts of the world (Torun, 2001). As the objective of this report is to make recommendations for healthy, well-nourished populations – thus excluding data from undernourished, overweight and stunted groups – the analysis was restricted to information from groups of healthy, well-nourished individuals.
4.1 MEASUREMENT OF TOTAL ENERGY EXPENDITURE
Studies using the DLW technique were the starting point for estimating energy requirements of children and adolescents. However, most of the existing data on TEE measured with DLW were obtained in industrialized countries, where energy expenditure is influenced by modern technology, school environments, sedentary pastimes, mechanized transportation and social and economic support systems that demand relatively low physical effort (i.e. in developed countries or affluent societies) compared with countries and societies where cultural, economic, social and developmental circumstances require greater physical effort from an early age (i.e. in developing countries or poorer, largely rural societies). On the other hand, several investigations on TEE of healthy, well-nourished children and adolescents have been done in a broader spectrum of countries and societies using minute-by-minute heart rate monitoring (HRM) and individual calibrations of the relationship between heart rate and oxygen consumption. The mean TEE measured with this technique is comparable with the mean value obtained using DLW or whole body calorimetry (Spurr et al., 1988; Ceesay et al., 1989; Livingstone et al., 1990; Livingstone et al., 1992; Emons et al., 1992; Maffeis et al., 1995; van den Berg-Emons et al., 1996; Davidson et al., 1997; Ekelund et al., 2000). Therefore, studies using either DLW or HRM were included in this evaluation in order to encompass data on children and adolescents with a wider variety of lifestyles, and to include age groups with limited information, if based on DLW alone.
The studies reviewed for this consultation involved a total of 801 boys and 808 girls of one to 18 years of age (Torun, 2001). Most (56 percent of the boys, 68 percent of the girls) were from the United States or the United Kingdom; 18 percent of the boys and 18 percent of the girls were from Canada, Denmark, Italy, Sweden or the Netherlands; and 26 percent of the boys and 14 percent of the girls were from Brazil, Chile, Colombia, Guatemala or Mexico. The Latin American children were four to 15 years old, and all lived in urban areas. Inter-individual coefficients of variation ranged from 9 to 34 percent within studies with DLW, and from 9 to 27 percent within studies with HRM. The overall mean coefficient of variation was 19 percent for energy expenditure calculated as TEE/day, and 17 percent when calculated as TEE/kg/day. The inter-individual variability was similar to that observed with DLW among infants (18 percent for TEE/day, and 15 percent for TEE/kg/day; see section 3.1).
4.2 EQUATIONS TO PREDICT TOTAL ENERGY EXPENDITURE
Predictive equations were derived from the studies of TEE. Because many publications did not present results on individual children, the mean values for boys or girls of a specific age, or within a reasonably narrow age range, were used in the calculations, weighting the results of each study on the number of children. Various mathematical models (e.g. linear, multiple, polynomial, etc.) were evaluated, with age and/or body weight as predictors of TEE. Age and weight were highly correlated, with a tolerance of 0.078 among boys and 0.061 among girls. Weight was selected as the single predictor, since it played a greater role than age in predicting TEE, and the exclusion of age from the
Energy requirements of children and adolescents
21
predictive models did not increase the error of the estimate. The lowest errors of estimation were obtained with the following quadratic polynomial regression equations for boys and girls (Figure 4.1) (Torun, 2001):
Boys:
TEE (MJ/day) = 1.298 + 0.265 kg – 0.0011 kg2; nweighted = 801, r = 0.982, r 2 = 0.964, see = 0.518
TEE (kcal/day) = 310.2 + 63.3 kg – 0.263 kg2
Girls:
TEE (MJ/day) = 1.102 + 0.273 kg – 0.0019 kg2; nweighted = 808, r = 0.955, r 2 = 0.913, see = 0.650
TEE (kcal/day) = 263.4 + 65.3 kg – 0.454 kg2
The equations were validated internally by dividing the studies into model-building sub-samples (70 percent of the studies, n = 549–618 boys or girls) and validation sub-samples (30 percent of the studies, n = 183–252 boys or girls). The validation sub-samples were randomly selected for each gender after stratifying the studies on quintiles of mean body weight – the method used to measure TEE (DLW or HRM) – and categorizing according to whether the study was done in an industrialized or a developing country. The correlation coefficients of the quadratic equations derived from the model-building sub-samples ranged from 0.959 to 0.982, with standard errors of the estimate from 0.504 to 0.651 MJ/day. Mean differences between predicted and measured values among boys were within ± 1 percent, with a standard deviation of 6 percent; and among girls, within ± 3 percent, with a standard deviation of 9 percent (Torun, 2001).
4.3 ENERGY NEEDS FOR GROWTH
Energy needs for growth have two components: 1) the energy used to synthesize growing tissues; and 2) the energy deposited in those tissues, basically as fat and protein, because carbohydrate content is negligible. Energy spent in tissue synthesis is part of TEE measured with either DLW or HRM. Hence, only the energy deposited in growing tissues was added to TEE in order to calculate energy requirements.
Table 4.1 shows the mean weight gain of boys and girls calculated from the WHO weight-for-age standards (WHO, 1983). The composition of weight gain was based on measurements at one and two years of age (Butte et al., 2000; Butte, 2001), assuming that the composition of normally growing tissues does not change much between the end of infancy and the onset of puberty. It was estimated as 10 percent fat with an energy content of 38.7 kJ/g (9.25 kcal/g), 20 percent protein of 23.6 kJ/g (5.65 kcal/g) energy content, and 70 percent water, carbohydrate and minerals with negligible content of energy. The average energy deposited in growing tissues was then about 8.6 kJ (2 kcal) per gram of weight gain. Even if this amount of energy were an over- or underestimation as large as 50 percent, it would only produce an error of about 1 percent in the calculations of energy requirements in childhood and adolescence.
4.4 CALCULATION OF ENERGY REQUIREMENTS
TEE was calculated using the predictive quadratic equations and the WHO reference values of weight-for-age (Torun, 2001; WHO, 1983). The median weight at the midpoint of each year of age was used for the ages of between one and 17 years (i.e. median weights at 1.5, 2.5..., 17.5 years). At the lower end of the weight distribution, which corresponds to infants between one and two years of age, predicted values were about 7 percent higher than the actual measurements of TEE. When reduced by that percentage, TEE estimates fell in line with those of 12-month-old infants (Butte, 2001). The small transient increment in TEE/kg/day between one and three years is probably associated with the effort of children starting to walk and run.
Energy deposited in growing tissues was estimated by multiplying the mean daily weight gain at each year of age (Table 4.1), by the average energy deposited in growing tissues (8.6 kJ or 2 kcal per gram of weight gain). The sum of energy deposition and TEE is the mean daily energy requirement (MJ or kcal/day, Tables 4.2 and 4.3). This was then divided by the median weight at each year to express requirements as energy units per kilogram of body weight.
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
22
FIGURE 4.1
Quadratic polynomial regression of total energy expenditure on body weight, weighting each data point by the number of children in the study
BOYS
GIRLS
Boys: y = 1.298 + 0.265x – 0.0011x 2 ; nweighted = 801, r = 0.982, see = 0.518.
Girls: y = 1.102 + 0.273x – 0.0019x 2 ; nweighted = 808, r = 0.955, see = 0.650.
Solid circles: DLW, industrialized countries. Clear circles: DLW, developing countries. Solid triangles: HRM, industrialized countries. Clear triangles: HRM, developing countries. Source: Torun, 2001.
3.5
4.5
5.5
6.5
7.5
8.5
9.5
10.5
11.5
12.5
13.5
14.5
15.5
10 15 20 25 30 35 40 45 50 55 60 65 70 75
Weight (kg)
T E
E (
M J
/d )
3.5
4.5
5.5
6.5
7.5
8.5
9.5
10.5
11.5
12.5
10 15 20 25 30 35 40 45 50 55 60 65
Weight (kg)
T E
E (
M J /d
)
Energy requirements of children and adolescents
23
TABLE 4.1
Mean weight gain of boys and girls, one to 17 years of age Age Boys Girls
years kg/year g/day kg/year g/day
1–2 2.4 6.6 2.4 6.6
2–3 2.0 5.5 2.2 6.0
3–4 2.1 5.8 1.9 5.2
4–5 2.0 5.5 1.7 4.7
5–6 2.0 5.5 1.8 4.9
6–7 2.2 6.0 2.3 6.3
7–8 2.4 6.6 3.0 8.2
8–9 2.8 7.7 3.7 10.1
9–10 3.3 9.0 4.0 11.0
10–11 3.9 10.7 4.5 12.3
11–12 4.5 12.3 4.5 12.3
12–13 5.2 14.2 4.6 12.6
13–14 5.8 15.9 4.2 11.5
14–15 5.9 16.2 3.4 9.3
15–16 5.4 14.8 2.2 6.0
16–17 4.2 11.5 0.8 2.2
17–18 2.6 7.1 0 0
Source: Calculated from WHO references of weight by age (WHO,1983).
BMR was estimated by using the equations endorsed in the report of the 1985 FAO/WHO/UNU expert consultation (Schofield, 1985), and upheld by this consultation (section 5.2, Table 5.2), using the median weight for every year of age. Mean PAL was calculated as a multiple of BMR, dividing total energy expenditure by the estimated BMR. As discussed in section 3.4.2, PAL calculated in this manner is on average 1 percent lower than when daily energy requirement is divided by BMR (James and Schofield, 1990) because growth contributes that proportion to the total energy requirement in childhood and adolescence. Thus, to estimate the energy requirement, the energy accrued during growth must be added, or the PAL value of children and adolescents must be multiplied by 1.01 (i.e. to make it 1 percent higher).
4.4.1 Comparison with previous requirements
In Table 4.4 and Figure 4.2 the new requirements are compared with those of the 1985 report. The cross-over of the curves at ten to 11 years is most probably artificial and the result of the different approaches used by the 1981 consultation to calculate requirements of children under ten years of age (dietary intake) and over ten years (factorial estimate of energy expenditure) (WHO, 1985). Compared with previous estimates, energy requirements proposed by this consultation are on average 18 percent lower for boys and 20 percent lower for girls under seven years of age, and 12 and 5 percent lower, respectively, for boys and girls seven to ten years of age. From 12 years onwards, the proposed requirements are an average of 12 percent higher for both boys and girls.
4.4.2 Influence of habitual physical activity on energy requirements
Energy requirements vary with the level of habitual physical activity. Most studies of TEE were carried out on random or convenient subject samples. Children and adolescents in these samples had different levels of habitual activity, resulting in inter-individual coefficients of variability as high as 34 percent (Torun, 2001). Thus, the values shown in Tables 4.2 and 4.3 may be regarded as the requirements of child and adolescent populations with “average” or “moderate” (i.e. not predominantly sedentary nor vigorous) physical activity.
Children and adolescents in rural, traditional communities in developing countries are more active than their counterparts in urban areas or in developed, industrialized countries. The quantitative differences were assessed from factorial estimates of TEE, calculated from 42 studies with time
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
24
allocation data that involved approximately 4 000 boys and girls in industrialized countries, and 2 400 in rural or urban areas of developing countries (Torun, 2001; 1996). On average, TEE of boys and girls five to nine, ten to 14 and 15 to 19 years of age was, respectively, about 10, 15 and 25 percent higher in rural developing countries than in cities or industrialized countries. Based on these values, on the within-study coefficients of variation of TEE measured with DLW or HRM and on the mean errors of estimation of the predictive equations for TEE, this consultation endorsed the recommendation to reduce or increase by 15 percent the requirement of population groups that are less or more active than average, starting at six years of age (Torun, 2001).
4.4.3 Requirements of populations with different levels of physical activity
Energy requirements were calculated for children over five years of age and for adolescents with lifestyles involving three levels of habitual physical activity, subtracting or adding 15 percent from the requirements shown in Tables 4.2 and 4.3 for children and adolescents with “average” physical activity. Population groups with less, similar or more than average activity were classified as leading “light”, “moderate” or “vigorous” lifestyles, respectively. Their requirements are shown in Tables 4.5 and 4.6. To facilitate recollection, values were rounded to the closest 0.1 MJ (25 kcal)/day, 5 kJ (1 kcal)/kg/day, and 0.05 PAL units.
The following general descriptions may help to decide which level of energy requirement is more appropriate for a specific population group.
Examples of populations with light physical lifestyles, or that are less active than average, are children and adolescents who every day spend several hours at school or in sedentary occupations; do not practise physical sports regularly; generally use motor vehicles for transportation; and spend most leisure time in activities that require little physical effort, such as watching television, reading, using computers or playing without much body displacement.
Examples of populations with vigorous lifestyles, or that are more active than average, are children and adolescents who every day walk long distances or use bicycles for transportation; engage in high energy-demanding occupations, or perform high energy-demanding chores for several hours each day; and/or practise sports or exercise that demand a high level of physical effort for several hours, several days of the week.
Children and adolescents with habitual physical activity that is more strenuous than the examples given for a light lifestyle, but not as demanding as the examples for vigorous lifestyle, would qualify in the category of average or moderate physically active lifestyles.
4.5 RECOMMENDATIONS FOR REGULAR PHYSICAL ACTIVITY
A certain amount of habitual physical activity is desirable for biological and social well-being. The regular performance of physical activity by children, in conjunction with good nutrition, is associated with health, adequate growth and well-being, and probably with lower risk of disease in adult life (Viteri and Torun, 1981; Torun and Viteri, 1994; Boreham and Riddoch, 2001). Children who are physically active explore their environment and interact socially more than their less active counterparts. There may also be a behavioural carry-over into adulthood, whereby active children are more likely to be active as adults, with the ensuing health benefits of exercise (Boreham and Riddoch, 2001).
On the other hand, sedentary lifestyles are increasing in most societies around the world, mainly owing to increased access to effort-saving technology and devices and to structural and social constraints. Examples of these are increased use of automobiles and buses for transportation, piped water and electrical appliances in the household, electronic equipment and computers in the workplace, elevators and escalators in buildings, and television sets and computers for entertainment, as well as a reduction in outdoor playing and walking caused by concerns about crime and the safety of pedestrians and cyclists. Sedentary children often eat amounts of food that exceed their relatively lower energy requirements, go into a positive energy balance and are at risk of becoming overweight or obese (Bar-Or et al., 1998; Goran and Treuth, 2001; Dietz and Gortmaker, 2001).
Energy requirements of children and adolescents
25
FIGURE 4.2
Comparison of proposed energy requirements with FAO/WHO/UNU 1985 requirements
BOYS
GIRLS
Continuous line: proposed energy requirements. Interrupted line: 1985 requirements. Source: Torun, 2001.
160
180
200
220
240
260
280
300
320
340
360
380
400
420
440
460
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Age (years)
E R
( k
J /k
g /d
)
160
180
200
220
240
260
280
300
320
340
360
380
400
420
440
460
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Age (years)
E R
( k
J /k
g /d
)
T A
B L E
4 .2
B o
y ’s
e n
e rg
y r
e q
u ir
e m
e n
ts c
a lc
u la
te d
b y q
u a
d ra
ti c
r e g
re s
s io
n a
n a ly
s is
o f
T E
E o
n w
e ig
h t,
p lu
s a
ll o
w a n
c e
f o
r e
n e
rg y d
e p
o s
it io
n i
n t
is s
u e
s
d u
ri n
g g
ro w
th (
E g )
A g
e
W e ig
h t
T E
E a
E g
b
B M
R e
s tc
D
a il
y e
n e rg
y r
e q
u ir
e m
e n
t P
A L
d
y e a rs
k g
M J /d
k c a l/ d
M J /d
k c a l/ d
M J /d
k c a
l/ d
M
J /d
k c a l/ d
k J /k
g /d
k c a l/ k g /d
T E
E /B
M R
1 –
2 e
1 1
.5
3 .9
0 6
9 3 4
0 .0
5 7
1 4
2
.7 3
7
6 5
4
3 .9
6 3
9 4
8
3 4 5
8
2 .4
1 .4
3
2 –
3
1 3
.5
4 .6
7 5
1 1
1 7
0 .0
4 7
1 1
3
.2 3
5
7 7
3
4 .7
2 2
1
1 2
9
3 5 0
8
3 .6
1 .4
5
3 –
4
1 5
.7
5 .1
8 7
1 2
4 0
0 .0
4 9
1 2
3
.6 0
2
8 6
1
5 .2
3 6
1
2 5
2
3 3 4
7
9 .7
1 .4
4
4 –
5
1 7
.7
5 .6
4 4
1 3
4 9
0 .0
4 7
1 1
3
.7 9
2
9 0
6
5 .6
9 1
1
3 6
0
3 2 2
7
6 .8
1 .4
9
5 –
6
1 9
.7
6 .0
9 2
1 4
5 6
0 .0
4 7
1 1
3
.9 8
2
9 5
2
6 .1
3 9
1
4 6
7
3 1 2
7
4 .5
1 .5
3
6 –
7
2 1
.7
6 .5
3 1
1 5
6 1
0 .0
5 2
1 2
4
.1 7
2
9 9
7
6 .5
8 3
1
5 7
3
3 0 3
7
2 .5
1 .5
7
7 –
8
2 4
.0
7 .0
2 4
1 6
7 9
0 .0
5 7
1 4
4
.3 9
0
1 0
4 9
7 .0
8 1
1
6 9
2
2 9 5
7
0 .5
1 .6
0
8 –
9
2 6
.7
7 .5
8 9
1 8
1 4
0 .0
6 6
1 6
4
.6 4
7
1 1
1 1
7 .6
5 5
1
8 3
0
2 8 7
6
8 .5
1 .6
3
9 –
1 0
2
9 .7
8
.1 9
8
1 9
5 9
0 .0
7 8
1 9
4
.9 3
2
1 1
7 9
8 .2
7 6
1
9 7
8
2 7 9
6
6 .6
1 .6
6
1 0
– 1 1
3
3 .3
8
.9 0
3
2 1
2 8
0 .0
9 2
2 2
5
.2 1
8
1 2
4 7
8 .9
9 5
2
1 5
0
2 7 0
6
4 .6
1 .7
1
1 1
– 1 2
3
7 .5
9
.6 8
9
2 3
1 6
0 .1
0 6
2 5
5
.5 2
9
1 3
2 1
9 .7
9 5
2
3 4
1
2 6 1
6
2 .4
1 .7
5
1 2
– 1 3
4
2 .3
1
0 .5
3 9
2 5
1 9
0 .1
2 3
2 9
5
.8 8
4
1 4
0 6
1 0 .6
6 2
2
5 4
8
2 5 2
6
0 .2
1 .7
9
1 3
– 1 4
4
7 .8
1
1 .4
5 2
2 7
3 7
0 .1
3 7
3 3
6
.2 9
1
1 5
0 4
1 1 .5
8 8
2
7 7
0
2 4 2
5
7 .9
1 .8
2
1 4
– 1 5
5
3 .8
1
2 .3
7 1
2 9
5 7
0 .1
3 9
3 3
6
.7 3
5
1 6
1 0
1 2 .5
1 0
2
9 9
0
2 3 3
5
5 .6
1 .8
4
1 5
– 1 6
5
9 .5
1
3 .1
7 1
3 1
4 8
0 .1
2 7
3 0
7
.1 5
7
1 7
1 1
1 3 .2
9 8
3
1 7
8
2 2 4
5
3 .4
1 .8
4
1 6
– 1 7
6
4 .4
1
3 .8
0 2
3 2
9 9
0 .0
9 9
2 4
7
.5 2
0
1 7
9 7
1 3 .9
0 1
3
3 2
2
2 1 6
5
1 .6
1 .8
4
1 7
– 1 8
6
7 .8
1
4 .2
0 8
3 3
9 6
0 .0
6 1
1 5
7
.7 7
1
1 8
5 7
1 4 .2
7 0
3
4 1
0
2 1 0
5
0 .3
1 .8
3
a T
E E
( M
J /d
) =
1 .2
9 8 +
0 .2
6 5
k g –
0 .0
0 1
1 k
g 2 .
b 8
.6 k
J o
r 2 k
c a
l/ g w
e ig
h t g
a in
.
c B
M R
e s t:
b a s a
l m
e ta
b o
lic r
a te
e s ti m
a te
d w
it h
p re
d ic
ti v e
e q
u a
ti o
n s o
n b
o d
y w
e ig
h t
(S c h
o fi e
ld ,
1 9
8 5
).
d P
A L
e s t:
p h
y s ic
a l a c ti v it y l e
v e
l =
T E
E /B
M R
e s t.
T o
c a lc
u la
te r
e q
u ir
e m
e n ts
, a d
d E
g o
r m
u lt ip
ly b
y 1
.0 1 (
s e e
t e
x t)
. e R
e q
u ir
e m
e n ts
f o r
1 t
o 2
y e a
rs r
e d
u c e
d b
y 7
p e rc
e n
t to
f it w
it h
e n
e rg
y r
e q
u ir
e m
e n ts
o f
in fa
n ts
( s e
e t
e x t)
. S
o u
rc e
: T
o ru
n ,
2 0
0 1
.
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
26
T A
B L E
4 .3
G ir
ls ’
e n
e rg
y r
e q
u ir
e m
e n
ts c
a lc
u la
te d
b y q
u a
d ra
ti c r
e g
re s
s io
n a
n a
ly s
is o
f T
E E
o n
w e
ig h
t, p
lu s
a ll o
w a n
c e
f o
r e
n e
rg y d
e p
o s
it io
n i
n t
is s
u e
s
d u
ri n
g g
ro w
th (
E g )
A g
e
W e ig
h t
T E
E a
E g
b
B M
R e
s tc
D
a il
y e
n e rg
y r
e q
u ir
e m
e n
t P
A L
d
y e a
rs k g
M J /d
k c a l/ d
M J /d
k c a
l/ d
M J /d
k c a
l/ d
M J /d
k c a
l/ d
k J /k
g /d
k c a l/ k g /d
T E
E /B
M R
1 – 2
e
1 0 .8
3 .5
6 1
8
5 1
0
.0 5 7
1 4
2
.5 0
5
5 9
9
3 .6
1 8
8 6
5
3 3
5
8 0 .1
1
.4 2
2 –
3
1 3 .0
4 .3
3 0
1
0 3
5
0 .0
5 2
1 2
3
.0 4
2
7 2
7
4 .3
8 2
1
0 4
7
3 3
7
8 0 .6
1
.4 2
3 –
4
1 5 .1
4 .7
9 1
1
1 4
5
0 .0
4 5
1 1
3
.3 1
7
7 9
3
4 .8
3 6
1
1 5
6
3 2
0
7 6 .5
1
.4 4
4 –
5
1 6 .8
5 .1
5 2
1
2 3
1
0 .0
4 0
1 0
3
.4 6
1
8 2
7
5 .1
9 2
1
2 4
1
3 0
9
7 3 .9
1
.4 9
5 –
6
1 8 .6
5 .5
2 2
1
3 2
0
0 .0
4 2
1 0
3
.6 1
4
8 6
4
5 .5
6 4
1
3 3
0
2 9
9
7 1 .5
1
.5 3
6 –
7
2 0 .6
5 .9
2 0
1
4 1
5
0 .0
5 4
1 3
3
.7 8
4
9 0
4
5 .9
7 4
1
4 2
8
2 9
0
6 9 .3
1
.5 6
7 –
8
2 3 .3
6 .4
3 1
1
5 3
7
0 .0
7 1
1 7
4
.0 1
4
9 5
9
6 .5
0 2
1
5 5
4
2 7
9
6 6 .7
1
.6 0
8 –
9
2 6 .6
7 .0
1 9
1
6 7
8
0 .0
8 7
2 1
4
.2 9
4
1 0
2 6
7
.1 0 6
1
6 9
8
2 6
7
6 3 .8
1
.6 3
9 – 1
0
3 0 .5
7 .6
6 1
1
8 3
1
0 .0
9 4
2 3
4
.6 2
6
1 1
0 5
7
.7 5 5
1
8 5
4
2 5
4
6 0 .8
1
.6 6
1 0 –
1 1
3 4 .7
8 .2
8 7
1
9 8
1
0 .1
0 6
2 5
4
.8 4
1
1 1
5 7
8
.3 9 3
2
0 0
6
2 4
2
5 7 .8
1
.7 1
1 1 –
1 2
3 9 .2
8 .8
8 4
2
1 2
3
0 .1
0 6
2 5
5
.0 9
3
1 2
1 7
8
.9 9 0
2
1 4
9
2 2
9
5 4 .8
1
.7 4
1 2 –
1 3
4 3 .8
9 .4
1 4
2
2 5
0
0 .1
0 8
2 6
5
.3 5
1
1 2
7 9
9
.5 2 3
2
2 7
6
2 1
7
5 2 .0
1
.7 6
1 3 –
1 4
4 8 .3
9 .8
5 5
2
3 5
5
0 .0
9 9
2 4
5
.6 0
3
1 3
3 9
9
.9 5 4
2
3 7
9
2 0
6
4 9 .3
1
.7 6
1 4 –
1 5
5 2 .1
1
0 .1
6 8
2
4 3
0
0 .0
8 0
1 9
5
.8 1
6
1 3
9 0
1 0
.2 4
8
2 4
4 9
1 9
7
4 7 .0
1
.7 5
1 5 –
1 6
5 5 .0
1
0 .3
7 0
2
4 7
8
0 .0
5 2
1 2
5
.9 7
8
1 4
2 9
1 0
.4 2
1
2 4
9 1
1 8
9
4 5 .3
1
.7 3
1 6 –
1 7
5 6 .4
1
0 .4
5 5
2
4 9
9
0 .0
1 9
5
6 .0
5 6
1 4
4 7
1 0
.4 7
4
2 5
0 3
1 8
6
4 4 .4
1
.7 3
1 7 –
1 8
5 6 .7
1
0 .4
7 3
2
5 0
3
0 .0
0 0
0
6 .0
7 3
1 4
5 1
1 0
.4 7
3
2 5
0 3
1 8
5
4 4 .1
1
.7 2
a T
E E
( M
J /d
) =
1 .1
0 2 +
0 .2
7 3
k g –
0 .0
0 1
9 k
g 2 .
b 8
.6 k
J o
r 2 k
c a
l/ g w
e ig
h t g
a in
.
c B
M R
e s t:
b a s a
l m
e ta
b o
lic r
a te
e s ti m
a te
d w
it h
p re
d ic
ti v e
e q
u a
ti o
n s o
n b
o d
y w
e ig
h t
(S c h
o fi e
ld ,
1 9
8 5
).
d P
A L
e s t:
p h
y s ic
a l a c ti v it y l e
v e
l =
T E
E /B
M R
e s t.
T o
c a lc
u la
te r
e q
u ir
e m
e n ts
, a d
d E
g o
r m
u lt ip
ly b
y 1
.0 1 (
s e e
t e
x t)
. e R
e q
u ir
e m
e n ts
f o r
1 t
o 2
y e a
rs r
e d
u c e
d b
y 7
p e rc
e n
t to
f it w
it h
e n
e rg
y r
e q
u ir
e m
e n ts
o f
in fa
n ts
( s e
e t
e x t)
. S
o u
rc e
: T
o ru
n ,
2 0
0 1 .
Energy requirements of children and adolescents
27
T A
B L E
4 .4
C o
m p
a ri
s o
n o
f n
e w
p ro
p o
s a
l fo
r d
a il
y e
n e
rg y r
e q
u ir
e m
e n
ts w
it h
t h
e 1
9 8
5 F
A O
/W H
O /U
N U
r e
p o
rt
B
o y s
G ir
ls
A g
e
N e w
v a lu
e s
F A
O /W
H O
/U N
U ,
1 9 8 5
N e w
v a lu
e s
F
A O
/W H
O /U
N U
, 1 9
8 5
y e
a rs
k J /k
g /d
k c a l/ k g /d
k J /k
g /d
%
d if f
a k J /k
g /d
k c a l/ k g /d
k J /k
g /d
%
d if f
a
1 –
2
3 4
5
8 2 .4
4
3 9
-2
1 .4
3
3 5
8
0 .1
4 3
9
-2 3
.7
2 –
3
3 5
0
8 3 .6
4
1 8
-1
6 .3
3
3 7
8
0 .6
4 1
8
-1 9
.4
3 –
4
3 3
4
7 9 .7
3
9 7
-1
5 .9
3
2 0
7
6 .5
3 9
7
-1 9
.4
4 –
5
3 2
2
7 6 .8
3
9 7
-1
8 .9
3
0 9
7
3 .9
3 9
7
-2 2
.2
5 –
6
3 1
2
7 4 .5
3
7 7
-1
7 .2
2
9 9
7
1 .5
3 5
6
-1 6
.0
6 –
7
3 0
3
7 2 .5
3
7 7
-1
9 .6
2
9 0
6
9 .3
3 5
6
-1 8
.5
7 –
8
2 9
5
7 0 .5
3
2 6
-9
.5
2 7 9
6
6 .7
2 8
0
-0 .4
8 –
9
2 8
7
6 8 .5
3
2 6
-1
2 .0
2
6 7
6
3 .8
2 8
0
-4 .6
9 –
1 0
2 7
9
6 6 .6
3
2 6
-1
4 .4
2
5 4
6
0 .8
2 8
0
-9 .3
1 0 –
1 1
2 7
0
6 4 .6
2
6 7
1
.1
2 4 2
5
7 .8
2 2
7
6 .6
1 1 –
1 2
2 6
1
6 2 .4
2
6 7
-2
.2
2 2 9
5
4 .8
2 2
7
0 .9
1 2 –
1 3
2 5
2
6 0 .2
2
2 8
1 0
.5
2 1 7
5
2 .0
1 8
9
1 4 .8
1 3 –
1 4
2 4
2
5 7 .9
2
2 8
6
.1
2 0 6
4
9 .3
1 8
9
9 .0
1 4 –
1 5
2 3
3
5 5 .7
2
0 0
1 6
.5
1 9 7
4
7 .0
1 7
3
1 3 .9
1 5 –
1 6
2 2
4
5 3 .4
2
0 0
1 2
.0
1 8 9
4
5 .3
1 7
3
9 .2
1 6 –
1 7
2 1
6
5 1 .6
1
8 6
1 6
.1
1 8 6
4
4 .4
1 6
7
1 1 .4
1 7 –
1 8
2 1
0
5 0 .3
1
8 6
1 2
.9
1 8 5
4
4 .1
1 6
7
1 0 .8
a %
d if fe
re n
c e =
n e
w v
a lu
e /F
A O
W H
O U
N U
× 1
0 0
– 1
0 0
. S
o u
rc e
: T
o ru
n ,
2 0
0 1
.
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
28
T A
B L E
4 .5
B o
y s
’ e
n e
rg y r
e q
u ir
e m
e n
ts i
n p
o p
u la
ti o
n s
w it
h t
h re
e l e
v e
ls o
f h
a b
it u
a l p
h y s
ic a l a
c ti
v it
y
L ig
h t
p h
y s
ic a l
a c ti
v it
y
M o
d e ra
te p
h y s ic
a l
a c ti
v it
y
H e a
v y p
h y s ic
a l
a c ti
v it
y
A g
e
W e ig
h t
D a il
y e
n e rg
y r
e q
u ir
e m
e n
t P
A L
D
a il
y e
n e rg
y r
e q
u ir
e m
e n
t P
A L
D
a il
y e
n e rg
y r
e q
u ir
e m
e n
t P
A L
y e a rs
k g
M J /d
k c a l/ d
k J /k
g /d
k c a l/ k g
/d
M
J /d
k c a l/ d
k J /k
g /d
k c a
l/ k g /d
M J /d
k c a
l/ d
k J /k
g /d
k c a l/ k g /d
1 –
2
1 1
.5
4 .0
9
5 0
3
4 5
8 2
1 .4
5
2 –
3
1 3
.5
4 .7
1
1 2
5
3 5
0
8 4
1 .4
5
3 –
4
1 5
.7
5 .2
1
2 5
0
3 3
5
8 0
1 .4
5
4 –
5
1 7
.7
5 .7
1
3 5
0
3 2
0
7 7
1 .5
0
5 –
6
1 9
.7
6 .1
1
4 7
5
3 1
0
7 4
1 .5
5
6 –
7
2 1
.7
5 .6
1 3
5 0
2 6
0
6 2
1
.3 0
6 .6
1
5 7
5
3 0
5
7 3
1 .5
5
7 .6
1
8 0
0
3 5
0
8 4
1 .8
0
7 –
8
2 4
.0
6 .0
1 4
5 0
2 5
0
6 0
1
.3 5
7 .1
1
7 0
0
2 9
5
7 1
1 .6
0
8 .2
1
9 5
0
3 4
0
8 1
1 .8
5
8 –
9
2 6
.7
6 .5
1 5
5 0
2 4
5
5 9
1
.4 0
7 .7
1
8 2
5
2 8
5
6 9
1 .6
5
8 .8
2
1 0
0
3 3
0
7 9
1 .9
0
9 –
1 0
2 9
.7
7 .0
1 6
7 5
2 3
5
5 6
1
.4 0
8 .3
1
9 7
5
2 8
0
6 7
1 .6
5
9 .5
2
2 7
5
3 2
0
7 6
1 .9
0
1 0 –
1 1
3 3
.3
7 .7
1 8
2 5
2 3
0
5 5
1
.4 5
9 .0
2
1 5
0
2 7
0
6 5
1 .7
0
1 0 .4
2
4 7
5
3 1
0
7 4
1 .9
5
1 1 –
1 2
3 7
.5
8 .3
2 0
0 0
2 2
0
5 3
1
.5 0
9 .8
2
3 5
0
2 6
0
6 2
1 .7
5
1 1 .3
2
7 0
0
3 0
0
7 2
2 .0
0
1 2 –
1 3
4 2
.3
9 .1
2 1
7 5
2 1
5
5 1
1
.5 5
1
0 .7
2
5 5
0
2 5
0
6 0
1 .8
0
1 2 .3
2
9 2
5
2 9
0
6 9
2 .0
5
1 3 –
1 4
4 7
.8
9 .8
2 3
5 0
2 0
5
4 9
1
.5 5
1
1 .6
2
7 7
5
2 4
0
5 8
1 .8
0
1 3 .3
3
1 7
5
2 7
5
6 6
2 .0
5
1 4 –
1 5
5 3
.8
1 0
.6
2 5
5 0
2 0
0
4 8
1
.6 0
1
2 .5
3
0 0
0
2 3
5
5 6
1 .8
5
1 4 .4
3
4 5
0
2 7
0
6 5
2 .1
5
1 5 –
1 6
5 9
.5
1 1
.3
2 7
0 0
1 9
0
4 5
1
.6 0
1
3 .3
3
1 7
5
2 2
5
5 3
1 .8
5
1 5 .3
3
6 5
0
2 6
0
6 2
2 .1
5
1 6 –
1 7
6 4
.4
1 1
.8
2 8
2 5
1 8
5
4 4
1
.5 5
1
3 .9
3
3 2
5
2 1
5
5 2
1 .8
5
1 6 .0
3
8 2
5
2 4
5
5 9
2 .1
5
1 7 –
1 8
6 7
.8
1 2
.1
2 9
0 0
1 8
0
4 3
1
.5 5
1
4 .3
3
4 0
0
2 1
0
5 0
1 .8
5
1 6 .4
3
9 2
5
2 4
0
5 7
2 .1
5
N o
te s :
B o d
y w
e ig
h t
a t m
id -p
o in
t o
f a g
e i n te
rv a
l (W
H O
, 1
9 8 3
).
N u
m b
e rs
r o
u n
d e d
t o t
h e
c lo
s e s t
0 .1
M J /d
, 2
5 k
c a l/ d
, 5
k J /k
g /d
, 1
k c a
l/ k g /d
, 0
.0 5
P A
L u
n it .
M
o d
e ra
te p
h y s ic
a l a c ti v it y : M
J /d
= (
1 .2
9 8
+ 0
.2 6
5 k
g –
0 .0
0 1 1
k g
2 )
+ 8
.6 k
J /g
d a
ily w
e ig
h t
g a
in .
L
ig h
t p h
y s ic
a l a
c ti v it y :
1 5
p e
rc e
n t
< m
o d
e ra
te p
h y s ic
a l a
c ti v it y .
V
ig o ro
u s p
h y s ic
a l a
c ti v it y :
1 5
p e
rc e
n t
> m
o d
e ra
te p
h y s ic
a l a
c ti v it y .
P A
L =
T E
E /(
p re
d ic
te d B
M R
/d ).
S
o u
rc e
: T
o ru
n ,
2 0
0 1
.
Energy requirements of children and adolescents
29
T A
B L E
4 .6
G ir
ls ’
e n
e rg
y r
e q
u ir
e m
e n
ts i
n p
o p
u la
ti o
n s
w it
h t
h re
e l e
v e
ls o
f h
a b
it u
a l p
h y s
ic a
l a
c ti
v it
y
L ig
h t
p h
y s ic
a l
a c ti
v it
y
M o
d e ra
te p
h y s ic
a l
a c ti
v it
y
H e a v y p
h y s ic
a l
a c ti
v it
y
A g
e
W e ig
h t
D a il
y e
n e rg
y r
e q
u ir
e m
e n
t P
A L
D
a il
y e
n e
rg y r
e q
u ir
e m
e n
t P
A L
D
a il
y e
n e rg
y r
e q
u ir
e m
e n
t P
A L
Y e
a rs
k g
M J /d
k c a l/ d
k J /k
g /d
k c a l/ k g
/d
M
J /d
k c a l/ d
k J /k
g /d
k c a l/ k g /d
M J /d
k c a l/ d
k J /k
g /d
k c a l/ k g /d
1 –
2
1 0
.8
3
.6
8 5
0
3 3
5
8 0
1 .4
0
2 –
3
1 3
.0
4
.4
1 0
5 0
3 3
5
8 1
1 .4
0
3 –
4
1 5
.1
4
.8
1 1
5 0
3 2
0
7 7
1 .4
5
4 –
5
1 6
.8
5
.2
1 2
5 0
3 1
0
7 4
1 .5
0
5 –
6
1 8
.6
5
.6
1 3
2 5
3 0
0
7 2
1 .5
5
6 –
7
2 0
.6
5 .1
1 2
2 5
2 4
5
5 9
1
.3 0
6
.0
1 4
2 5
2 9
0
6 9
1 .5
5
6 .9
1
6 5 0
3 3
5
8 0
1 .8
0
7 –
8
2 3
.3
5 .5
1 3
2 5
2 3
5
5 7
1
.3 5
6
.5
1 5
5 0
2 8
0
6 7
1 .6
0
7 .5
1
7 7 5
3 2
0
7 7
1 .8
5
8 –
9
2 6
.6
6 .0
1 4
5 0
2 2
5
5 4
1
.4 0
7
.1
1 7
0 0
2 6
5
6 4
1 .6
5
8 .2
1
9 5 0
3 0
5
7 3
1 .9
0
9 –
1 0
3 0
.5
6 .6
1 5
7 5
2 1
5
5 2
1
.4 0
7
.7
1 8
5 0
2 5
5
6 1
1 .6
5
8 .9
2
1 2 5
2 9
5
7 0
1 .9
0
1 0 –
1 1
3 4
.7
7 .1
1 7
0 0
2 0
5
4 9
1
.4 5
8
.4
2 0
0 0
2 4
0
5 8
1 .7
0
9 .6
2
3 0 0
2 7
5
6 6
1 .9
5
1 1 –
1 2
3 9
.2
7 .6
1 8
2 5
1 9
5
4 7
1
.5 0
9
.0
2 1
5 0
2 3
0
5 5
1 .7
5
1 0
.3
2 4
7 5
2 6
5
6 3
2 .0
0
1 2 –
1 3
4 3
.8
8 .1
1 9
2 5
1 8
5
4 4
1
.5 0
9
.5
2 2
7 5
2 1
5
5 2
1 .7
5
1 1
.0
2 6
2 5
2 4
5
6 0
2 .0
0
1 3 –
1 4
4 8
.3
8 .5
2 0
2 5
1 7
5
4 2
1
.5 0
1
0 .0
2 3
7 5
2 0
5
4 9
1 .7
5
1 1
.4
2 7
2 5
2 3
5
5 7
2 .0
0
1 4 –
1 5
5 2
.1
8 .7
2 0
7 5
1 6
5
4 0
1
.5 0
1
0 .2
2 4
5 0
1 9
5
4 7
1 .7
5
1 1
.8
2 8
2 5
2 2
5
5 4
2 .0
0
1 5 –
1 6
5 5
.0
8 .9
2 1
2 5
1 6
0
3 9
1
.5 0
1
0 .4
2 5
0 0
1 9
0
4 5
1 .7
5
1 2
.0
2 8
7 5
2 2
0
5 2
2 .0
0
1 6 –
1 7
5 6
.4
8 .9
2 1
2 5
1 6
0
3 8
1
.5 0
1
0 .5
2 5
0 0
1 8
5
4 4
1 .7
5
1 2
.0
2 8
7 5
2 1
5
5 1
2
.0
1 7 –
1 8
5 6
.7
8 .9
2 1
2 5
1 5
5
3 7
1
.4 5
1
0 .5
2 5
0 0
1 8
5
4 4
1 .7
0
1 2
.0
2 8
7 5
2 1
5
5 1
1 .9
5
N o
te s :
B o d
y w
e ig
h t
a t m
id -p
o in
t o
f a g
e i n te
rv a
l (W
H O
, 1
9 8 3
).
N u
m b
e rs
r o
u n
d e d
t o t
h e
c lo
s e s t
0 .1
M J /d
, 2
5 k
c a l/ d
, 5
k J /k
g /d
, 1
k c a
l/ k g /d
, 0
.0 5
P A
L u
n it .
M
o d
e ra
te p
h y s ic
a l a c ti v it y : M
J /d
= (
1 .1
0 2
+ 0
.2 7
3 k
g –
0 .0
0 1 9
k g
2 )
+ 8
.6 k
J /g
d a
ily w
e ig
h t
g a
in .
L
ig h
t p h
y s ic
a l a
c ti v it y :
1 5
p e
rc e
n t
< m
o d
e ra
te p
h y s ic
a l a
c ti v it y .
V
ig o ro
u s p
h y s ic
a l a
c ti v it y :
1 5
p e
rc e
n t
> m
o d
e ra
te p
h y s ic
a l a
c ti v it y .
P A
L =
T E
E /(
p re
d ic
te d B
M R
/d ).
S
o u
rc e
: T
o ru
n ,
2 0
0 1
.
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
30
Energy requirements of children and adolescents
31
It is therefore important that recommendations for appropriate levels of physical activity accompany recommendations for dietary energy intakes. There is no direct experimental or epidemiological evidence on the minimal or optimal frequency, duration and intensity of exercise that promotes health and well-being in children, but it has been suggested that children should perform a minimum of 60 minutes per day of moderate-intensity physical activity, which may be carried out in cumulative bouts of ten or more minutes, and which should be supplemented by activities that promote flexibility, muscle strength and increase in bone mass (Boreham and Riddoch, 2001). This can be pursued by promoting walking, climbing stairs or cycling as part of everyday activities, and encouraging participation in games and sports that involve body displacement and a certain degree of physical effort. In making such recommendations, local culture, social customs and environmental characteristics must be taken into account.
4.6 INFECTIONS AND MILD MALNUTRITION
Nutritional requirements and dietary energy recommendations for children who are severely malnourished or chronically ill, such as owing to HIV/AIDS, are beyond the scope of this report. It must be recognized, however, that many populations around the world include large proportions of children with some degree of weight deficit and growth retardation as a result of mild to moderate chronic malnutrition and/or repeated bouts of infections (UNICEF, 2001). As was pointed out in the report from the 1985 FAO/WHO/UNU expert consultation (WHO, 1985), when a public health problem is of such magnitude that it affects the energy and protein requirements of a significant part of the population, it may not be ignored in the assessment of such requirements or in the recommendations that are made.
Situations that promote malnutrition also favour a high incidence of infectious diseases, which in turn further contribute to the malnutrition. For many children under five – and particularly those under three – years of age who live in these conditions, being sick or convalescing from diarrhoea or a respiratory infection is part of “normal life”, because they experience this several times a year, with each episode lasting two to 15 days and requiring up to twice that time to achieve full recovery, provided that an intervening new episode of disease does not interrupt the recovery process (Mata, 1978; Black and Lanata, 1995; Steinhoff, 2000). Infections of this nature often result in negative energy balance resulting from poor appetite, decreased absorption of nutrients during diarrhoeal episodes and increased metabolic rate, particularly in febrile processes (Waterlow and Tomkins, 1992; Torun, 2000). This leads to chronic mild wasting (i.e. low weight-for-height) and stunting (i.e. low height-for-age), which may be prevented, ameliorated or corrected if adequate care and food are available, especially in the periods between infectious episodes when appetite has been re-established. If, on the contrary conditions do not improve, the status quo of mild malnutrition is maintained, the possibility for catch-up is reduced and the consequences of malnutrition will continue to prevail in those societies.
Diets for catch-up weight gain must provide all nutrients and energy sources in amounts that surpass the requirements of well-nourished, healthy children. Quantitative estimates of energy requirements for catch-up are difficult to establish for two reasons: 1) the target body weight is not fixed but increases with time in a growing child, so that the longer the period of nutritional deficit, the greater the gap to be filled; and 2) a low weight for a given age may be owing either to a reduction in weight below the acceptable range of weight-for-height (i.e. wasting) or to a low height (i.e. stunting) with a concomitant decrease in weight. In the latter case, if the decrease in weight is proportional to the reduced growth in height, the child will not have a weight deficit as such, and provision of additional dietary energy may lead to overweight, as catch-up in height is much slower and less likely to be achieved than increase in weight.
The extra amounts of energy needed for catch-up growth of a child with actual weight deficit (i.e. low weight-for-height) have been estimated in studies on the rehabilitation of malnourished children as 21 kJ (5 kcal) per gram of tissue to be laid down (Fomon, 1971; Ashworth, 1969; Kerr et al., 1973; Whitehead, 1973; Spady et al., 1976; Krieger and Whitten, 1976). The recommended daily amounts of energy will depend on the rate at which catch-up is expected to occur. Under optimal clinical conditions, children with severe malnutrition can gain weight at rates of up to 20 or more times faster than normal growth. However, at the community level, catch-up rates of free-living children with mild
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
32
to moderate degrees of weight deficit should realistically be expected to be no more than two or three times the normal rate.
In relation to the energy demands imposed by repeated bouts of infection, the paucity of data concerning illness, convalescence and post-convalescence does not allow estimates of energy requirements for infants and children to be based on direct measurements of energy expenditure and growth. This leads to the suggested use of a factorial estimate of theoretical needs during acute illness and/or convalescence. In addition to basal metabolism, the energy costs of normal growth and the energy needs for obligatory and discretionary activities, the factors involved in the estimate also include faecal energy losses owing to malabsorption from diarrhoeal disease, and increased energy needs imposed by fever and other responses to stress. This is no easy task owing to the variability in clinical and metabolic responses to illnesses of different aetiologies and different degrees of severity. A number of studies in different countries have attempted to quantify the proportion of the growth deficit that can be attributed to infections and the extra requirements for recovery from them, but the results have been inconsistent (WHO, 1985).
As was stated by the previous joint FAO/WHO/UNU expert consultation (WHO, 1985), it is still impossible to generalize about the amounts of additional energy needed for catch-up growth in children who have become malnourished, usually as a result of the combined effects of inadequate intake and frequent infection. The relative contributions of these two factors, and their severity, will vary in different communities and at different times. In many countries there are also important seasonal effects on food supply and incidence of infections. This consultation could not give a better recommendation than that previously offered, which was based on theoretical estimates to allow for twice the normal rate of weight gain among infants in countries with high prevalence of infant and childhood malnutrition. As shown in Table 4.7, this ranges from an increase in energy requirements – and intakes – of 14.5 percent at six to nine months of age, to 3.5 percent at 18 to 24 months. To restore growth, diets with a high energy density may be needed during the short anabolic periods following episodes of weight loss.
TABLE 4.7
Increase in energy requirements needed to allow for twice the normal growth rate of children six to 24 months old*
Age months
Average weight gain g/kg/day
% increase over energy requirement
6–9 1.83 14.5
9–12 1.15 8.5
12–18 0.67 5
18–24 0.51 3.5
* It was assumed that the requirements for normal growth were 1.5 times the theoretical estimates based on weight gain. Source: adapted from WHO, 1985.
In practice, children should be fed according to appetite, with food of good overall quality that satisfies the needs for all nutrients. To counteract the effects of anorexia and the metabolic losses that accompany infection, sufficient amounts of food should be available in periods when appetite is restored and the child is recovering from infection. It must also be recognized that supplying the child’s increased requirement is only one of the measures needed to counteract the effects of periodic infectious episodes. The primary need is for prevention through improved sanitation and other public health measures.
REFERENCES Ashworth, A. 1969. Growth rates in children recovering from protein-calorie malnutrition. Brit. J. Nutr., 23: 835–845. Bar-Or, O., Foreyt, J., Bouchard, C., Brownell, K.D., Dietz, W.H., Ravussin, E., Sable, A.D., Schwenger,
S., St Jeor, S. & Torun, B. 1998. Physical activity, genetic and nutritional considerations in childhood weight management. Med. Sci. Sports Exerc., 30: 2–10. Black, R.E. & Lanata, C.F. 1995. Epidemiology of diarrheal diseases in developing countries. In M.J. Blaser, P.D. Smith and I. Ravdin, eds. Infections of the gastrointestinal tract, pp. 13–18. New York, Raven Press.
Energy requirements of children and adolescents
33
Boreham, C. & Riddoch, C. 2001. The physical activity, fitness and health of children. J. Sports Sci., 19: 915– 929. Butte, N.F. 2001. Energy requirements of infants. Background paper prepared for the joint FAO/WHO/UNU Expert Consultation on Energy in Human Nutrition. Butte, N.F., Wong, W.W., Hopkinson, J.M., Heinz, C.J., Mehta, N.R. & Smith, E.O. 2000. Energy requirements derived from total energy expenditure and energy deposition during the first 2 y of life. Am. J. Clin. Nutr., 72: 1558–1569. Ceesay, S.M., Prentice, A.M., Day, K.C., Murgatroyed, P.R., Goldberg, G.R. & Schott, W. 1989. The use of heart rate monitoring in the estimation of energy expenditure: a validation study using indirect whole-body calorimetry. Br. J. Nutr., 61: 175–186. Davidson, L., NcNeill, G., Haggarty, P., Smith, J.S. & Franklin, M.F. 1997. Free-living energy expenditure of adult men assessed by continuous heart-rate monitoring and doubly labelled water. Br. J. Nutr., 78: 695–708. Dietz, W.H. & Gortmaker, S.L. 2001. Preventing obesity in children and adolescents. Annu. Rev. Public Health, 22: 337–353. Ekelund, U., Sjöström, M., Yngve, A. & Nilsson, A. 2000. Total daily energy expenditure and pattern of physical activity measured by minute-by-minute heart rate monitoring in 14–15 year old Swedish adolescents. Eur. J. Clin. Nutr., 54: 195–202. Emons, H.J.G., Groenenboom, D.C., Westerterp, K.R. & Saris, W.H.M. 1992. Comparison of heart rate monitoring combined with indirect calorimetry and the doubly labelled water method (2H2
18O) for the measurement of energy expenditure in children. Eur. J. Appl. Physiol., 65: 99–103. Fomon, S.J. 1971. Food consumption and growth of infants fed milk-based formulas. Acta Paediatr. Scand., 223 (suppl.): 1–36. Goran, M.I. & Treuth, M.S. 2001. Energy expenditure, physical activity, and obesity in children. Pediatr. Clin. North Am., 48: 931–953. James, W.P.T. & Schofield, E.C. 1990. Human energy requirements. A manual for planners and nutritionists. Oxford, UK, Oxford Medical Publications under arrangement with FAO. Kerr, D. et al. 1973. Accelerated recovery from infant malnutrition with high calorie feeding. In L. Gardner and L. Amacher, eds. Endocrine aspects of malnutrition, pp. 467–479. Santa Ynez, California, USA, Kroc Foundation. Krieger, I. & Whitten, C.F. 1976. Nitrogen balance and calorie efficiency in small-for-date dwarfism. Am. J. Clin. Nutr., 29: 38–45. Livingstone, M.B.E., Prentice, A.M., Coward, W.A., Ceesay, S.M., Strain, J.J., McKenna, P.G., Nevin,
G.B., Barker, M.E. & Hickey, R.J. 1990. Simultaneous measurement of free-living energy expenditure by the doubly labelled water method and heart-rate monitoring. Am. J. Clin. Nutr., 52: 59–65. Livingstone, M.B.E., Coward, W.A., Prentice, A.M., Davies, P.S.W., Strain, J.J., McKenna, P.G.,
Mahoney, C.A., White, J.A., Stewart, C.M. & Kerr, M.J. 1992. Daily energy expenditure in free-living children: comparison of heart-rate monitoring with the doubly labelled water (2H2
18O) method. Am. J. Clin. Nutr., 56: 343–352. Maffeis, C., Pinelli, L., Zaffanello, M., Schena, F., Iacumin, P. & Schutz, Y. 1995. Daily energy expenditure in free-living conditions in obese and non-obese children: comparison of doubly labelled water method (2H2 18O) and heart-rate monitoring. Int. J. Obes. Relat. Metab. Disord., 19: 671–677. Mata, L.J. 1978. The children of Santa Maria Cauque: a prospective field study of health and growth. Cambridge, Massachusetts, USA, MIT Press. Schofield, W.N. 1985. Predicting basal metabolic rate, new standards and review of previous work. Hum. Nutr. Clin. Nutr., 39C (suppl. 1): 5–41. Spady, D.W., Payne, P.R., Picou, D. & Waterlow, J.C. 1976. Energy balance during recovery from malnutrition. Am. J. Clin. Nutr., 29: 1073–1078. Spurr, G.B., Prentice, A.M., Murgatroyd, P.R., Goldberg, G.R., Reina, J.C. & Christman, N.T. 1988. Energy expenditure from minute-by-minute heart-rate recording: comparison with indirect calorimetry. Am. J. Clin. Nutr., 48: 552–559. Steinhoff, M.C. 2000. Pulmonary diseases. In G.T. Strickland, ed. Hunter’s tropical medicine and emerging infectious diseases, eighth edition, pp. 1–7. Philadelphia, Pennsylvania, USA, W.B. Saunders. Torun, B. 1996. Energy requirements and dietary energy recommendations for children and adolescents 1 to 18 years old. Eur. J. Clin. Nutr., 50 (suppl. 1): S37–S81. Torun, B. 2000. Protein-energy malnutrition. In G.T. Strickland, ed. Hunter’s tropical medicine and emerging infectious diseases, eighth edition, pp. 927–940. Philadelphia, Pennsylvania, USA, W.B. Saunders. Torun, B. 2001. Energy requirements of children and adolescents. Background paper prepared for the joint FAO/WHO/UNU Expert Consultation on Energy in Human Nutrition, 2001. Torun, B. & Viteri, F.E. 1994. Influence of exercise on linear growth. Eur. J. Clin. Nutr., 48 (suppl. 1): S186– S190.
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
34
UNICEF. 2001. State of the World’s Children 2001. New York. van den Berg-Emons, R.J.G., Saris, W.H.M., Westerterp, K.R. & van Baak, M.A. 1996. Heart rate monitoring to assess energy expenditure in children with reduced physical activity. Med. Sci. Sport Exerc., 28: 496–501. Viteri, F.E. & Torun, B. 1981. Nutrition, physical activity and growth. In M. Ritzén, A. Aperia, K. Hall, A. Larsson, A. Zetterberg and R. Zetterström, eds. The biology of normal human growth, pp. 265–273. New York, Raven Press. Waterlow, J.C. & Tomkins, A.M. 1992. Nutrition and infection. In J.C. Waterlow, ed. Protein energy malnutrition, pp. 290–324. London, Edward Arnold. Whitehead, R.G. 1973. The protein needs of malnourished children. In J.W. Porter and B.A. Rolls, eds. Proteins in human nutrition, pp. 103–117. London, Academic Press. WHO. 1983. Measuring change in nutritional status. Geneva. WHO. 1985. Energy and protein requirements: Report of a joint FAO/WHO/UNU expert consultation. WHO Technical Report Series No. 724. Geneva.
Energy requirements of adults
35
5. ENERGY REQUIREMENTS OF ADULTS
The principles followed by the 1985 FAO/WHO/UNU expert consultation (WHO, 1985) were adhered to, and energy requirements of adults were calculated from factorial estimates of habitual TEE. The use of techniques such as DLW and HRM confirmed the large diversity of TEE – and hence of energy requirements – among adult societies, which were previously reported by time–motion studies. Growth is no longer an energy-demanding factor in adulthood, and BMR is relatively constant among population groups of a given age and gender. Consequently, habitual physical activity and body weight are the main determinants for the diversity in energy requirements of adult populations with different lifestyles (James and Schofield, 1990).
5.1 FACTORIAL ESTIMATION OF TOTAL ENERGY EXPENDITURE AND PHYSICAL ACTIVITY
LEVEL
The diversity in body size, body composition and habitual physical activity among adult populations with different geographic, cultural and economic backgrounds does not allow a universal application of energy requirements based on TEE measured with DLW (or HRM) in groups with a specific lifestyle. Hence, to account for the differences in physical activity, TEE was estimated through factorial calculations that combined the time allocated to habitual activities and the energy cost of those activities. Table 5.1 shows examples of these calculations. To account for differences in body size and composition, the energy cost of activities was calculated as a multiple of BMR per minute, also referred to as the physical activity ratio (PAR), and the 24-hour energy requirement was expressed as a multiple of BMR per 24 hours by using the PAL value (James and Schofield, 1990). Together with BMR of the population, PAL when known or when derived using BMR estimated from age and gender-specific predictive equations based on the average body weight of the population provides an estimate of TEE and hence the mean energy requirement for that population.
To simplify calculations, the previous expert consultation classified the PAL of adult population groups as light, moderate or heavy, depending on their occupational or other work, and multiplied it by the corresponding BMR to arrive at requirements (WHO, 1985). The present consultation considered that the 24-hour PAL should not be based only on the physical effort demanded by occupational work, as there are people with light occupations who perform vigorous physical activity in their spare time, and people with heavy work who are quite sedentary the rest of the day. As discussed in section 5.3, it was decided to base the factorial estimates of energy requirements on the energy expenditure associated with lifestyles that combine occupational and discretionary physical activities.
This consultation also agreed that the average energy cost of activities expressed as a multiple of BMR, or PAR, should be similar for men and women. The effect of gender comes out when the PAR value is converted into energy units, because men have higher BMR for their body weight than women, and this difference is accentuated by the heavier weight of men. Consequently, the energy cost of most activities listed in Table 5.1 as a function of BMR is applicable to both men and women. Notable exceptions are vigorous activities that demand a level of effort proportional to muscle mass and strength, which tend to be greater among men (for example, lifting and carrying heavy loads, cutting wood or working with a sledgehammer).
5.2 ESTIMATION OF BASAL METABOLIC RATE
BMR constitutes about 45 to 70 percent of TEE in adults, and is determined principally by gender, body size, body composition and age. It can be measured accurately with small intra-individual variation by direct or indirect calorimetry under standard conditions, which include being awake in the supine position, ten to 12 hours after a meal, following eight hours of physical rest and no strenuous exercise in the preceding day, and being in a state of mental relaxation and an ambient environmental temperature that does not evoke shivering or sweating. BMR can be measured only under laboratory conditions and in small groups of representative individuals. There is a need to estimate BMR at the population level when using the factorial approach to estimate TEE from the average BMR and PAL value attributable to that population. Hence, the alternative has been to
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
36
estimate a group’s mean BMR using predictive equations based on measurements that are easier to obtain, such as body weight and/or height.
TABLE 5.1
Factorial calculations of total energy expenditure for a population group
Main daily activities Time allocation hours
Energy cost a
PAR Time × energy cost
Mean PAL b
multiple of 24-hour BMR
Sedentary or light activity lifestyle
Sleeping 8 1 8.0
Personal care (dressing, showering) 1 2.3 2.3
Eating 1 1.5 1.5
Cooking 1 2.1 2.1
Sitting (office work, selling produce, tending shop) 8 1.5 12.0
General household work 1 2.8 2.8
Driving car to/from work 1 2.0 2.0
Walking at varying paces without a load 1 3.2 3.2
Light leisure activities (watching TV, chatting) 2 1.4 2.8
Total 24 36.7 36.7/24 = 1.53
Active or moderately active lifestyle
Sleeping 8 1 8.0
Personal care (dressing, showering) 1 2.3 2.3
Eating 1 1.5 1.5
Standing, carrying light loads (waiting on tables, arranging merchandise)
c 8 2.2 17.6
Commuting to/from work on the bus 1 1.2 1.2
Walking at varying paces without a load 1 3.2 3.2
Low intensity aerobic exercise 1 4.2 4.2
Light leisure activities (watching TV, chatting) 3 1.4 4.2
Total 24 42.2 42.2/24 = 1.76
Vigorous or vigorously active lifestyle
Sleeping 8 1 8.0
Personal care (dressing, bathing) 1 2.3 2.3
Eating 1 1.4 1.4
Cooking 1 2.1 2.1
Non-mechanized agricultural work (planting, weeding, gathering) 6 4.1 24.6
Collecting water/wood 1 4.4 4.4
Non-mechanized domestic chores (sweeping, washing clothes and dishes by hand) 1 2.3 2.3
Walking at varying paces without a load 1 3.2 3.2
Miscellaneous light leisure activities 4 1.4 5.6
Total 24 53.9 53.9/24 = 2.25 a Energy costs of activities, expressed as multiples of basal metabolic rate, or PAR, are based on Annex 5 of the previous
consultation’s report (WHO, 1985) (see also Annex 5 of this report). b PAL = physical activity level, or energy requirement expressed as a multiple of 24-hour BMR.
c Composite of the energy cost of standing, walking slowly and serving meals or carrying a light load.
Examples: Sedentary or light activity: If this PAL was from a female population, 30 to 50 years old, with mean weight of 55 kg and mean BMR of 5.40 MJ/day (1 290 kcal/day), TEE = 1.53 × 5.40 = 8.26 MJ (1 975 kcal), or 150 kJ (36 kcal)/kg/d. Active or moderately active: If this PAL was from a female population, 20 to 25 years old, with mean weight of 57 kg and mean BMR of 5.60 MJ/day (1 338 kcal/day), TEE = 1.76 × 5.60 = 9.86 MJ (2 355 kcal), or 173 kJ (41 kcal)/kg/d. Vigorous or vigorously active: If this PAL was from a male population, 20 to 25 years old, with mean weight of 70 kg and mean BMR of 7.30 MJ/day (1 745 kcal/day), TEE = 2.25 × 7.30 = 16.42 MJ (3 925 kcal), or 235 kJ (56 kcal)/kg/d.
Energy requirements of adults
37
The report from the 1985 FAO/WHO/UNU expert consultation used a set of equations derived mostly from studies in Western Europe and North America (Schofield, 1985). Almost half of the data used to generate the equations for adults were from studies carried out in the late 1930s and early 1940s on Italian men with relatively high BMR values, and questions have been raised about the universal applicability of those equations (Soares and Shetty, 1988; de Boer et al., 1988; Henry and Rees, 1991; Arciero et al., 1993; Piers and Shetty, 1993; Soares, Francis and Shetty, 1993; Hayter and Henry, 1993 and 1994; Valencia et al., 1994; Cruz, da Silva and dos Anjos, 1999; Henry, 2001; Ismail et al., 1998). The use of closed-circuit indirect calorimetry in most studies has also been questioned, as this technique might overestimate oxygen consumption and energy expenditure. For the present consultation, predictive equations derived from a database with broader geographical and ethnic representation were evaluated (Henry, 2001; Cole, 2002). The predictive accuracies of the new equations and of those from 1985 were compared with published measurements of BMR in adults from different parts of the world, which were not part of the databases used to generate the predictive equations (Ramirez-Zea, 2002). Although the new equations had some merits, such as small reductions in the error of prediction and the overestimation bias among men, this consultation concluded that these were not robust enough to justify their adoption at present. For the time being, it was decided to retain the equations proposed in 1985 by Schofield (Table 5.2), and to pursue a more thorough analysis of existing information, or to promote a prospective study with broad global geographic and ethnic representation.
TABLE 5.2
Equations for estimating BMR from body weight* Age
Years No. BMR: MJ/day see
a BMR: kcal/day see
a
Males
< 3 162 0.249kg – 0.127 0.292 59.512kg – 30.4 70
3–10 338 0.095kg + 2.110 0.280 22.706kg + 504.3 67
10–18 734 0.074kg + 2.754 0.441 17.686kg + 658.2 105
18–30 2879 0.063kg + 2.896 0.641 15.057kg + 692.2 153
30–60 646 0.048kg + 3.653 0.700 11.472kg + 873.1 167
60 50 0.049kg + 2.459 0.686 11.711kg + 587.7 164
Females
< 3 137 0.244kg – 0.130 0.246 58.317kg – 31.1 59
3–10 413 0.085kg + 2.033 0.292 20.315kg + 485.9 70
10–18 575 0.056kg + 2.898 0.466 13.384kg + 692.6 111
18–30 829 0.062kg + 2.036 0.497 14.818kg + 486.6 119
30–60 372 0.034kg + 3.538 0.465 8.126kg + 845.6 111
60 38 0.038kg + 2.755 0.451 9.082kg + 658.5 108
* Weight is expressed in kg. Predictive equations for children and adolescents are presented for the sake of completeness. Source: Schofield, 1985. a see = standard error of estimate.
5.3 PHYSICAL ACTIVITY LEVEL
The average PAL of healthy, well-nourished adults is a major determinant of their total energy requirement. As growth does not contribute to energy needs in adulthood, PAL can be measured or estimated from the average 24-hour TEE and BMR (i.e. PAL = TEE/BMR). Multiplying the PAL by the BMR gives the actual energy requirements. For example, a male with a PAL of 1.75 and a mean BMR of 7.10 MJ/day (1 697 kcal/day) would have a mean energy requirement of 1.75 × 7.10 = 12.42 MJ/day (2 970 kcal/day).4 Other examples of these calculations are shown at the bottom of each panel in Table 5.1.
PAL has been calculated in several studies from measurements of TEE and measurements or estimates of BMR. Most of the existing data on the TEE of adults are from studies in industrialized societies, although some investigations have been done in developing countries where many people
4 When the averages of the PAL and of the BMR of a population are known, the average energy requirement of the population can be estimated.
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
38
have lifestyles associated with levels of physical activity that differ from those in industrialized countries (Coward, 1998). A meta-analysis of studies that involved a total of 411 men and women from 18 to 64 years of age showed a modal value for PAL of 1.60 (range 1.55 to 1.65) for both men and women (Black et al., 1996). For the most part, subjects were from affluent societies in developed countries. All were healthy, but 13 percent of the women and 9 percent of the men were overweight or obese, with BMI > 30. Typical sub-populations included students, housewives, white-collar or professional workers, and unemployed or retired individuals; only three persons were specifically identified as manual workers. Hence, the authors of the meta-analysis defined the study participants as people with a “predominantly sedentary Western lifestyle”. An expert panel of the International Obesity Task Force (IOTF) suggested a somewhat lower PAL range of 1.50 to 1.55 as being representative of sedentary individuals (Erlichman, Kerbey and James, 2001).
The PAL values that can be sustained for a long period of time by free-living adult populations range from about 1.40 to 2.40. This consultation agreed that a desirable PAL includes the regular practice of physical activity at work or in spare time with an intensity and duration that will reduce the risk of becoming overweight and developing a variety of non-communicable chronic diseases usually associated as co-morbidities with obesity. As discussed in section 5.6, this corresponds to PAL values of 1.75 and higher. On the other hand, a minimum “maintenance” energy requirement was not defined, reaffirming the position of the previous expert consultation which stated that “any figure chosen would reflect a value judgement on what levels of activity above the minimum for survival could be appropriately included in the term “maintenance” (WHO, 1985).
5.3.1 Classification of physical activity levels
Energy requirements are highly dependent on habitual physical activity. This consultation classified the intensity of a population’s habitual physical activity into three categories, as was done by the 1981 FAO/WHO/UNU expert consultation (WHO, 1985). However, in contrast with the 1981 consultation, a range of PAL values, rather than a mean PAL value, was established for each category. Furthermore, the same PAL values were used to assign men and women to a PAL category, for the reasons discussed in section 5.1.
The categories shown in Table 5.3 represent the different levels of activity associated with a population’s lifestyle. These categories indicate the physical activity most often performed by most individuals in the population, over a period of time. Although there is no physiological basis for establishing the duration of that period, it may be defined as one month or longer.
The term “lifestyle” was preferred to “occupational work”, as was used in the 1985 report, because there are groups of people with light or sedentary occupations who perform vigorous discretionary activities regularly, and therefore have a lifestyle that falls more appropriately within the “active” or “vigorously active” categories. It should also be borne in mind that some populations undergo cyclic changes in lifestyle, such as those related to the agricultural cycle among traditional rural societies, or those related to the seasons of the year where hot or mild summers alternate with cold winters. Energy requirements of such populations will change with the energy demands of their cyclical lifestyles.
TABLE 5.3
Classification of lifestyles in relation to the intensity of habitual physical activity, or PAL
Category PAL value
Sedentary or light activity lifestyle 1.40-1.69
Active or moderately active lifestyle 1.70-1.99
Vigorous or vigorously active lifestyle 2.00-2.40*
* PAL values > 2.40 are difficult to maintain over a long period of time.
Energy requirements of adults
39
5.3.2 Examples of lifestyles with different levels of energy demands
Sedentary or light activity lifestyles. These people have occupations that do not demand much physical effort, are not required to walk long distances, generally use motor vehicles for transportation, do not exercise or participate in sports regularly, and spend most of their leisure time sitting or standing, with little body displacement (e.g. talking, reading, watching television, listening to the radio, using computers). One example is male office workers in urban areas, who only occasionally engage in physically demanding activities during or outside working hours. Another example are rural women living in villages that have electricity, piped water and nearby paved roads, who spend most of the time selling produce at home or in the marketplace, or doing light household chores and caring for children in or around their houses.
Active or moderately active lifestyles. These people have occupations that are not strenuous in terms of energy demands, but involve more energy expenditure than that described for sedentary lifestyles. Alternatively, they can be people with sedentary occupations who regularly spend a certain amount of time in moderate to vigorous physical activities, during either the obligatory or the discretionary part of their daily routine. For example, the daily performance of one hour (either continuous or in several bouts during the day) of moderate to vigorous exercise, such as jogging/running, cycling, aerobic dancing or various sports activities, can raise a person’s average PAL from 1.55 (corresponding to the sedentary category) to 1.75 (the moderately active category). Other examples of moderately active lifestyles are associated with occupations such as masons and construction workers, or rural women in less developed traditional villages who participate in agricultural chores or walk long distances to fetch water and fuelwood.
Vigorous or vigorously active lifestyles. These people engage regularly in strenuous work or in strenuous leisure activities for several hours. Examples are women with non-sedentary occupations who swim or dance an average of two hours each day, or non-mechanized agricultural labourers who work with a machete, hoe or axe for several hours daily and walk long distances over rugged terrains, often carrying heavy loads.
Extremes of low and high PALs. Extremely low levels of energy expenditure allow for survival, but they are not compatible with long-term health, moving around freely, or earning a living. Such levels have been reported, for example, in elderly mental patients (Prentice et al., 1989), adolescents with cerebral palsy or myelodysplasia (Bandini et al., 1991) and resting adults confined to a whole body calorimeter (Ravussin et al., 1991; Schulz et al., 1992). The mean PAL of 1.21, which is similar to the baseline energy need of 1.27 estimated in the 1985 report, is suggested for short-term survival of totally inactive dependent people in conditions of crisis (WHO, 1985). The present consultation felt that such a value is too low and should not be used in emergency relief programmes, as people are not completely inactive in situations of crisis and the various stresses that impinge on them may increase their energy demands. The consultation hence suggests that food supplies to satisfy a PAL of 1.40, which represents the lower limit of the sedentary lifestyle range shown in Table 5.3, would be more appropriate for short-term relief interventions.
At the other end of the scale, studies have shown PAL values as high as 4.5 to 4.7 during three weeks of competitive cycling (Westerterp et al., 1986), or hauling sleds across the Arctic (Stroud, Coward and Sawyer, 1993). However, such levels of energy expenditure are not sustainable in the long term.
5.4 ENERGY REQUIREMENTS AND DIETARY ENERGY RECOMMENDATIONS
5.4.1 Calculation of energy requirements
Energy requirements were calculated from the factorial estimates of PAL described in the preceding sections. They were converted into energy units (i.e. joules and calories) by multiplying the PAL value by the BMR. In order to express requirements as energy units per kilogram of body weight, they were divided by the weight used in the equations to predict BMR. The following example to calculate the average energy requirement of a female population 20 to 30 years of age with a moderately active lifestyle and a mean body weight of 55 kg illustrates these calculations:
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
40
BMR (calculated with the predictive equation in Table 5.2): 5.45 MJ/day (1 302 kcal/day). PAL (mid-point of the moderately active lifestyle in Table 5.3): 1.85. Energy requirement: 5.45 × 1.85 = 10.08 MJ/day (2 410 kcal/day), or 10.08/55 = 183 kJ/kg/day (44 kcal/kg/day).
The variation in requirements around the mid-point of the PAL ranges in each lifestyle category in Table 5.3 is between 8 percent and 10 percent (e.g. PAL for moderately active lifestyle = 1.70 to 1.99 = 1.85 8 percent). However, there are individuals with BMR or PAL values at the extremes of a normal distribution around the population mean. Consequently, within each lifestyle category there are people whose individual energy requirement is beyond the limits shown in Table 5.3. This reiterates the fact that the energy requirements and dietary energy recommendations in this report are to be applied to population groups and not to individuals. Requirements of a specific individual must be based on that person’s actual TEE or BMR, or on estimates that take into account the individual’s habitual physical activity and lifestyle characteristics.
Tables 5.4 to 5.9 show the average energy requirements of populations with various levels of habitual physical activity and various mean body weights. Requirements for groups with other weights and/or mean PAL can be calculated easily, as in the following example for men 20 to 25 years of age with an average weight of 68 kg and an estimated PAL of 1.80:
a) Use Table 5.4 for men aged 18 to 30 years. b) Calculate the 24-hour BMR by interpolating between body weights of 65 and 70 kg and multiply the interpolated BMR/kg by the population’s average weight of 68 kg: mean of 108 and 104 = 106 kJ/kg × 68 kg = 7 208 kJ/day. c) Multiply the 24-hour BMR by the estimated average PAL of the population: average daily requirement = 7 208 kJ/day × 1.80 = 12.97 MJ/day (3 100 kcal/day).
Another option for calculations is:
a) Use Table 5.4 for men aged 18 to 30 years. b) Calculate the approximate energy requirement by interpolating between the daily requirements of men weighing 65 and 70 kg, and with a PAL level near that of the population (in this example, 1.75): mean of 12.2 and 12.8 MJ/day = 12.5 MJ/day. c) Multiply the approximate requirement by the ratio between the population’s PAL and the PAL of the column used in Table 5.5: 12.5 × (1.80/1.75) = 12.86 MJ/day (3 074 kcal/day).
5.4.2 Recommendation for daily energy intake
Dietary energy intake of a healthy, well-nourished population should allow for maintaining an adequate BMI at the population’s usual level of energy expenditure. At the individual level, a normal range of 18.5 to 24.9 kg/m2 BMI is generally accepted (WHO 1995 and 2000). At a population level, a median BMI of 21.0 was recently suggested by the joint WHO/FAO Expert Consultation on Diet, Nutrition and the Prevention of Chronic Diseases (WHO/FAO, 2002).
As BMI is a function of weight and height, heights corresponding to a BMI of 18.5, 21.0 and 24.9 were included in Tables 5.4 to 5.9 for each mean weight shown in the first column of each table. This facilitates recommendations for dietary energy intakes aimed at maintaining those values or range of BMI. For example, in the case of a male population 18 to 30 years old with an average height of 1.70 m and an activity lifestyle with a mean PAL of 1.75, the recommended energy intake would be around 11.7 MJ/day or 195 kJ/kg/day, which corresponds to the average requirement of men with a height of 1.69 m and a BMI of 21.0 who have a PAL of 1.75 × BMR (Table 5.4).
In the same example, a range of approximately 11.1 to 12.8 MJ/day or 185 to 200 kJ/kg/day would allow the maintenance of a BMI between 18.5 and 24.9 kg/m2. These figures were obtained from the PAL column of 1.75 × BMR in Table 5.4, between the lowest row with height 1.70 m in the column of 18.5 BMI (in this example, the second row with a height of 1.72 m) and the highest row with height
1.70 m in the column of 24.9 BMI (in this example, the fifth row with a height of 1.68 m).
T A
B L E
5 .4
D a
il y a
v e
ra g
e e
n e
rg y r
e q
u ir
e m
e n
t fo
r m
e n
a g
e d
1 8
t o
2 9
.9 y
e a
rs *
M e a
n
B M
R /k
g a
D a il
y e
n e rg
y r
e q
u ir
e m
e n
t a
c c o
rd in
g t
o B
M R
f a c to
r (o
r P
A L
) a n
d b
o d
y w
e ig
h t
in d
ic a te
d
H e ig
h t
(m )
fo r
w e
ig h
t
1 .4
5 ×
B M
R
1 .6
0 ×
B M
R
1 .7
5 ×
B M
R
1 .9
0 ×
B M
R
2
.0 5 ×
B M
R
2 .2
0 ×
B M
R
B M
I v a lu
e s :b
k g
k J
k c a
l
M J
k J /k
g
k c a l
k c a l/ k g
M
J
k J /k
g k c a l
k c a
l/ k g
M J
k J /k
g k c a l
k c a
l/ k g
M J
k J /k
g k c a l
k c a
l/ k g
M
J
k J /k
g k c a l
k c a l/ k g
M J
k J /k
g k c a l
k c a
l/ k g
2 4
.9
2 1 .0
1 8 .5
5 0
1
2 1
2
9 8
.8
1 7 5
2
1 0
0
4 2
9 .7
1
9 5
2 3
0 0
4 6
1 0
.6 2
1 0
2 5
5 0
5 1
1 1
.5 2
3 0
2 7
5 0
5 5
1 2
.4 2
5 0
2 9
5 0
5 9
1 3
.3 2 6
5 3
2 0 0
6 4
1 .4
2 1 .5
4 1
.6 4
5 5
1
1 6
2
8 9
.2
1 7 0
2
2 0
0
4 0
1
0 .2
1
8 5
2 4
5 0
4 4
1 1
.1 2
0 0
2 6
5 0
4 8
1 2
.1 2
2 0
2 9
0 0
5 3
1 3
.0 2
3 5
3 1
0 0
5 7
1 4
.0 2 5
5 3
3 5 0
6 1
1 .4
9 1 .6
2 1
.7 2
6 0
1
1 1
2
7 9
.7
1 6 0
2
3 0
0
3 9
1
0 .7
1
8 0
2 5
5 0
4 3
1 1
.7 1
9 5
2 8
0 0
4 7
1 2
.7 2
1 0
3 0
5 0
5 1
1 3
.7 2
3 0
3 2
5 0
5 5
1 4
.7 2 4
5 3
5 0 0
5 9
1 .5
5 1 .6
9 1
.8 0
6 5
1
0 8
2
6 1
0 .1
1
5 5
2
4 0
0
3 7
1
1 .2
1
7 0
2 6
5 0
4 1
1 2
.2 1
9 0
2 9
0 0
4 5
1 3
.3 2
0 5
3 1
5 0
4 9
1 4
.3 2
2 0
3 4
5 0
5 3
1 5
.4 2 3
5 3
7 0 0
5 7
1 .6
2 1 .7
6 1
.8 7
7 0
1
0 4
2
5 1
0 .6
1
5 0
2
5 5
0
3 6
1
1 .7
1
6 5
2 8
0 0
4 0
1 2
.8 1
8 5
3 0
5 0
4 4
1 3
.9 2
0 0
3 3
0 0
4 7
1 5
.0 2
1 5
3 6
0 0
5 1
1 6
.1 2 3
0 3
8 5 0
5 5
1 .6
8 1 .8
3 1
.9 5
7 5
1
0 2
2
4 1
1 .1
1
4 5
2
6 5
0
3 5
1
2 .2
1
6 5
2 9
0 0
3 9
1 3
.3 1
8 0
3 2
0 0
4 2
1 4
.5 1
9 5
3 4
5 0
4 6
1 5
.6 2
1 0
3 7
5 0
5 0
1 6
.8 2 2
5 4
0 0 0
5 3
1 .7
4 1 .8
9 2
.0 1
8 0
9
9
2 4
1 1 .5
1
4 5
2
7 5
0
3 4
1
2 .7
1
6 0
3 0
5 0
3 8
1 3
.9 1
7 5
3 3
0 0
4 1
1 5
.1 1
9 0
3 6
0 0
4 5
1 6
.3 2
0 5
3 9
0 0
4 9
1 7
.5 2 2
0 4
1 5 0
5 2
1 .7
9 1 .9
5 2
.0 8
8 5
9
7
2 3
1 2 .0
1
4 0
2
8 5
0
3 4
1
3 .2
1
5 5
3 1
5 0
3 7
1 4
.4 1
7 0
3 4
5 0
4 1
1 5
.7 1
8 5
3 7
5 0
4 4
1 6
.9 2
0 0
4 0
5 0
4 8
1 8
.2 2 1
5 4
3 5 0
5 1
1 .8
5 2 .0
1 2
.1 4
9 0
9
5
2 3
1 2 .4
1
4 0
2
9 5
0
3 3
1
3 .7
1
5 0
3 3
0 0
3 6
1 5
.0 1
6 5
3 6
0 0
4 0
1 6
.3 1
8 0
3 9
0 0
4 3
1 7
.6 1
9 5
4 2
0 0
4 7
1 8
.8 2 1
0 4
5 0 0
5 0
1 .9
0 2 .0
7 2
.2 1
* V
a lu
e s r
o u
n d e
d t o
c lo
s e
s t
0 .1
M J /d
, 5
0 k
c a
l/ d
, 5
k J /k
g /d
, 1
k c a l/ k g
/d .
a B
M R
c a
lc u
la te
d f o
r e
a c h w
e ig
h t
fr o
m t
h e
e q
u a ti o
n s i n T
a b
le 5
.2 .
V a lu
e s o
f B
M R
/k g
a re
p re
s e n
te d
f o
r e
a s e o
f c a
lc u
la ti o n
s f
o r
th o s e
w h
o w
is h
t o
u s e
d if fe
re n t
P A
L v
a lu
e s o
r d if fe
re n t
w e
ig h ts
. b H
e ig
h t ra
n g e
s a
re p
re s e n
te d
f o r
e a
c h m
e a
n w
e ig
h t fo
r e a
s e
o f
m a
k in
g d
ie ta
ry e
n e rg
y r
e c o
m m
e n
d a
ti o
n s t
o m
a in
ta in
a n
a d
e q
u a
te B
M I
b a s e
d o
n a
p o p
u la
ti o
n 's
m e
a n
h e ig
h t a
n d
P A
L .
F o r
e x a
m p
le ,
th e
r e
c o
m m
e n
d e
d m
e a
n e
n e rg
y i n
ta k e
f o r
a m
a le
p o
p u
la ti o
n o
f th
is a
g e
g ro
u p w
it h
a m
e a
n h
e ig
h t
o f
1 .7
0 m
a n
d a
l if e
s ty
le w
it h
a m
e a
n P
A L o
f 1
.7 5 ,
is a
b o
u t
1 1 .7
M J (
2 8
0 0 k
c a
l) /d
a y o
r 1
9 5
k J (
4 7
k c a
l) /k
g /d
a y t
o m
a in
ta in
a n o
p ti m
u m
p o
p u
la ti o
n m
e d
ia n
B M
I o f
2 1 .0
( W
H O
/F A
O ,
2 0
0 2 ),
w it h a
n i n
d iv
id u a
l ra
n g
e o
f a b
o u t
1 1
.1 t
o 1
2 .8
M J (
2 6
5 0
t o
3 0
5 0
k c a
l) /d
a y o
r 1
8 5
t o
2 0
0 k
J (
4 4
t o 4
8
k c a
l) /k
g /d
a y t
o m
a in
ta in
t h e
i n
d iv
id u
a l B
M I
lim it s o
f 1 8
.5 t
o 2
4 .9
( W
H O
, 2
0 0 0
).
Energy requirements of adults
41
T A
B L E
5 .5
D a
il y a
v e
ra g
e e
n e
rg y r
e q
u ir
e m
e n
t fo
r m
e n
a g
e d
3 0
t o
5 9
.9 y
e a
rs *
M e
a n
B
M R
/k g
a
D a il
y e
n e rg
y r
e q
u ir
e m
e n
t a
c c o
rd in
g t
o B
M R
f a c to
r (o
r P
A L
) a n
d b
o d
y w
e ig
h t
in d
ic a te
d
H e ig
h t
(m )
fo r
w e
ig h
t
1 .4
5 ×
B M
R
1 .6
0 ×
B M
R
1 .7
5 ×
B M
R
1 .9
0 ×
B M
R
2
.0 5 ×
B M
R
2 .2
0 ×
B M
R
B M
I v a lu
e s :b
k g
k J
k c a
l
M J
k J /k
g
k c a l
k c a l/ k g
M
J
k J /k
g k c a l
k c a
l/ k g
M J
k J /k
g k c a l
k c a
l/ k g
M J
k J /k
g k c a l
k c a
l/ k g
M
J
k J /k
g k c a l
k c a l/ k g
M J
k J /k
g k c a l
k c a
l/ k g
2 4
.9
2 1 .0
1 8 .5
5 0
1 2 1
2
9 8
.8
1 7 5
2
1 0
0
4 2
9
.7
1 9 5
2 3
0 0
4 6
1 0
.6 2
1 0
2 5
5 0
5 1
1 1
.5 2
3 0
2 7
5 0
5 5
1 2
.4 2
5 0
2 9
5 0
5 9
1 3
.3 2 6
5 3
2 0 0
6 4
1 .4
2 1 .5
4 1
.6 4
5 5
1 1 4
2
7 9
.1
1 6 5
2
2 0
0
4 0
1
0 .1
1
8 5
2 4
0 0
4 4
1 1
.0 2
0 0
2 6
5 0
4 8
1 2
.0 2
1 5
2 8
5 0
5 2
1 2
.9 2
3 5
3 1
0 0
5 6
1 3
.8 2 5
0 3
3 0 0
6 0
1 .4
9 1 .6
2 1
.7 2
6 0
1 0 9
2
6 9
.5
1 6 0
2
2 5
0
3 8
1
0 .5
1
7 5
2 5
0 0
4 2
1 1
.4 1
9 0
2 7
5 0
4 6
1 2
.4 2
0 5
2 9
5 0
4 9
1 3
.4 2
2 5
3 2
0 0
5 3
1 4
.4 2 4
0 3
4 5 0
5 7
1 .5
5 1 .6
9 1
.8 0
6 5
1 0 4
2
5 9
.8
1 5 0
2
3 5
0
3 6
1
0 .8
1
6 5
2 6
0 0
4 0
1 1
.9 1
8 0
2 8
5 0
4 4
1 2
.9 2
0 0
3 1
0 0
4 7
1 3
.9 2
1 5
3 3
0 0
5 1
1 4
.9 2 3
0 3
5 5 0
5 5
1 .6
2 1 .7
6 1
.8 7
7 0
1 0 0
2
4 1
0 .2
1
4 5
2
4 5
0
3 5
1
1 .2
1
6 0
2 7
0 0
3 8
1 2
.3 1
7 5
2 9
5 0
4 2
1 3
.3 1
9 0
3 2
0 0
4 5
1 4
.4 2
0 5
3 4
5 0
4 9
1 5
.4 2 2
0 3
7 0 0
5 3
1 .6
8 1 .8
3 1
.9 5
7 5
9 7
2 3
1 0 .5
1
4 0
2
5 0
0
3 4
1
1 .6
1
5 5
2 7
5 0
3 7
1 2
.7 1
7 0
3 0
5 0
4 0
1 3
.8 1
8 5
3 3
0 0
4 4
1 4
.9 2
0 0
3 5
5 0
4 7
1 6
.0 2 1
5 3
8 0 0
5 1
1 .7
4 1 .8
9 2
.0 1
8 0
9 4
2 2
1 0 .9
1
3 5
2
6 0
0
3 2
1
2 .0
1
5 0
2 8
5 0
3 6
1 3
.1 1
6 5
3 1
5 0
3 9
1 4
.2 1
8 0
3 4
0 0
4 3
1 5
.4 1
9 0
3 6
5 0
4 6
1 6
.5 2 0
5 3
9 5 0
4 9
1 .7
9 1 .9
5 2
.0 8
8 5
9 1
2 2
1 1 .2
1
3 0
2
7 0
0
3 2
1
2 .4
1
4 5
2 9
5 0
3 5
1 3
.5 1
6 0
3 2
5 0
3 8
1 4
.7 1
7 5
3 5
0 0
4 1
1 5
.9 1
8 5
3 8
0 0
4 5
1 7
.0 2 0
0 4
0 5 0
4 8
1 .8
5 2 .0
1 2
.1 4
9 0
8 9
2 1
1 1 .6
1
3 0
2
7 5
0
3 1
1
2 .8
1
4 0
3 0
5 0
3 4
1 4
.0 1
5 5
3 3
5 0
3 7
1 5
.1 1
7 0
3 6
0 0
4 0
1 6
.3 1
8 0
3 9
0 0
4 3
1 7
.5 1 9
5 4
2 0 0
4 7
1 .9
0 2 .0
7 2
.2 1
* V
a lu
e s r
o u
n d e
d t o
c lo
s e
s t
0 .1
M J /d
, 5
0 k
c a
l/ d
, 5
k J /k
g /d
, 1
k c a l/ k g
/d .
a B
M R
c a
lc u
la te
d f o
r e
a c h w
e ig
h t
fr o
m t
h e
e q
u a ti o
n s i n T
a b
le 5
.2 .
V a lu
e s o
f B
M R
/k g
a re
p re
s e n
te d
f o
r e
a s e o
f c a
lc u
la ti o n
s f
o r
th o s e
w h
o w
is h
t o
u s e
d if fe
re n t
P A
L v
a lu
e s o
r d if fe
re n t
w e
ig h ts
. b H
e ig
h t ra
n g e
s a
re p
re s e n
te d
f o r
e a
c h m
e a
n w
e ig
h t fo
r e a
s e
o f
m a
k in
g d
ie ta
ry e
n e rg
y r
e c o
m m
e n
d a
ti o
n s t
o m
a in
ta in
a n
a d
e q
u a
te B
M I
b a s e
d o
n a
p o p
u la
ti o
n 's
m e
a n
h e ig
h t a
n d
P A
L .
F o r
e x a
m p
le ,
th e
r e
c o
m m
e n
d e
d m
e a
n e
n e rg
y i n
ta k e
f o r
a m
a le
p o
p u
la ti o
n o
f th
is a
g e
g ro
u p w
it h
a m
e a
n h
e ig
h t
o f
1 .7
0 m
a n
d a
l if e
s ty
le w
it h
a m
e a
n P
A L o
f 1
.7 5 ,
is a
b o
u t
1 1 .4
M J (
2 7
5 0 k
c a
l) /d
a y o
r 1
9 0
k J (
4 6
k c a
l) /k
g /d
a y t
o m
a in
ta in
a n o
p ti m
u m
p o
p u
la ti o
n m
e d
ia n
B M
I o f
2 1 .0
( W
H O
/F A
O ,
2 0
0 2 ),
w it h a
n i n
d iv
id u a
l ra
n g
e o
f a b
o u t
1 1
.0 t
o 1
2 .3
M J (
2 6
5 0
t o
2 9
5 0
k c a
l) /d
a y o
r 1
7 5
t o
2 0
0 k
J (
4 2
t o 4
8
k c a
l) /k
g /d
a y t
o m
a in
ta in
t h e
i n
d iv
id u
a l B
M I
lim it s o
f 1 8
.5 t
o 2
4 .9
( W
H O
, 2
0 0 0
).
42
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
T A
B L E
5 .6
D a
il y a
v e
ra g
e e
n e
rg y r
e q
u ir
e m
e n
t fo
r m
e n
a g
e d
6
0 y
e a
rs *
M e a
n
B M
R /k
g a
D a il
y e
n e rg
y r
e q
u ir
e m
e n
t a
c c o
rd in
g t
o B
M R
f a c to
r (o
r P
A L
) a n
d b
o d
y w
e ig
h t
in d
ic a te
d
H e ig
h t
(m )
fo r
w e
ig h
t
1 .4
5 ×
B M
R
1 .6
0 ×
B M
R
1 .7
5 ×
B M
R
1 .9
0 ×
B M
R
2
.0 5 ×
B M
R
2 .2
0 ×
B M
R
B M
I v a lu
e s :b
k g
k J
k c a
l
M J
k J /k
g
k c a l
k c a l/ k g
M
J
k J /k
g k c a l
k c a
l/ k g
M J
k J /k
g k c a l
k c a
l/ k g
M J
k J /k
g k c a l
k c a
l/ k g
M
J
k J /k
g k c a l
k c a l/ k g
M J
k J /k
g k c a l
k c a
l/ k g
2 4
.9
2 1 .0
1 8 .5
5 0
9 8
2 3
7 .1
1
4 0
1
7 0
0
3 4
7 .9
1
5 5
1 9
0 0
3 8
8 .6
1 7
0 2 0
5 0
4 1
9 .3
1 8
5 2 2
5 0
4 5
1 0
.1 2
0 0
2 4
0 0
4 8
1 0
.8 2 1
5 2
6 0 0
5 2
1 .4
2 1 .5
4 1
.6 4
5 5
9 4
2 2
7 .5
1
3 5
1
8 0
0
3 3
8 .2
1
5 0
1 9
5 0
3 5
9 .0
1 6
5 2 1
5 0
3 9
9 .8
1 8
0 2 3
5 0
4 3
1 0
.6 1
9 0
2 5
5 0
4 6
1 1
.3 2 0
5 2
7 0 0
4 9
1 .4
9 1 .6
2 1
.7 2
6 0
9 0
2 2
7 .8
1
3 0
1
8 5
0
3 1
8 .6
1
4 5
2 0
5 0
3 4
9 .4
1 5
5 2 2
5 0
3 8
1 0
.3 1
7 0
2 4
5 0
4 1
1 1
.1 1
8 5
2 6
5 0
4 4
1 1
.9 2 0
0 2
8 5 0
4 8
1 .5
5 1 .6
9 1
.8 0
6 5
8 7
2 1
8 .2
1
2 5
1
9 5
0
3 0
9 .0
1
4 0
2 1
5 0
3 3
9 .9
1 5
0 2 3
5 0
3 6
1 0
.7 1
6 5
2 5
5 0
3 9
1 1
.6 1
8 0
2 7
5 0
4 2
1 2
.4 1 9
0 2
9 5 0
4 5
1 .6
2 1 .7
6 1
.8 7
7 0
8 4
2 0
8 .5
1
2 0
2
0 5
0
2 9
9 .4
1
3 5
2 2
5 0
3 2
1 0
.3 1
4 5
2 4
5 0
3 5
1 1
.2 1
6 0
2 6
5 0
3 8
1 2
.1 1
7 0
2 9
0 0
4 1
1 3
.0 1 8
5 3
1 0 0
4 4
1 .6
8 1 .8
3 1
.9 5
7 5
8 2
2 0
8 .9
1
2 0
2
1 5
0
2 9
9 .8
1
3 0
2 3
5 0
3 1
1 0
.7 1
4 5
2 5
5 0
3 4
1 1
.7 1
5 5
2 8
0 0
3 7
1 2
.6 1
7 0
3 0
0 0
4 0
1 3
.5 1 8
0 3
2 5 0
4 3
1 .7
4 1 .8
9 2
.0 1
8 0
8 0
1 9
9 .2
1
1 5
2
2 0
0
2 8
1
0 .2
1
3 0
2 4
5 0
3 1
1 1
.2 1
4 0
2 6
5 0
3 3
1 2
.1 1
5 0
2 9
0 0
3 6
1 3
.1 1
6 5
3 1
5 0
3 9
1 4
.0 1 7
5 3
3 5 0
4 2
1 .7
9 1 .9
5 2
.0 8
8 5
7 8
1 9
9 .6
1
1 5
2
3 0
0
2 7
1
0 .6
1
2 5
2 5
5 0
3 0
1 1
.6 1
3 5
2 7
5 0
3 2
1 2
.6 1
5 0
3 0
0 0
3 5
1 3
.6 1
6 0
3 2
5 0
3 8
1 4
.6 1 7
0 3
5 0 0
4 1
1 .8
5 2 .0
1 2
.1 4
9 0
7 6
1 8
1 0 .0
1
1 0
2
4 0
0
2 7
1
1 .0
1
2 0
2 6
5 0
2 9
1 2
.0 1
3 5
2 8
5 0
3 2
1 3
.1 1
4 5
3 1
0 0
3 4
1 4
.1 1
5 5
3 3
5 0
3 7
1 5
.1 1 7
0 3
6 0 0
4 0
1 .9
0 2 .0
7 2
.2 1
* V
a lu
e s r
o u
n d e
d t o
c lo
s e
s t
0 .1
M J /d
, 5
0 k
c a
l/ d
, 5
k J /k
g /d
, 1
k c a l/ k g
/d .
a B
M R
c a
lc u
la te
d f o
r e
a c h w
e ig
h t
fr o
m t
h e
e q
u a ti o
n s i n T
a b
le 5
.2 .
V a lu
e s o
f B
M R
/k g
a re
p re
s e n
te d
f o
r e
a s e o
f c a
lc u
la ti o n
s f
o r
th o s e
w h
o w
is h
t o
u s e
d if fe
re n t
P A
L v
a lu
e s o
r d if fe
re n t
w e
ig h ts
. b H
e ig
h t ra
n g e
s a
re p
re s e n
te d
f o r
e a
c h m
e a
n w
e ig
h t fo
r e a
s e
o f
m a
k in
g d
ie ta
ry e
n e rg
y r
e c o
m m
e n
d a
ti o
n s t
o m
a in
ta in
a n
a d
e q
u a
te B
M I
b a s e
d o
n a
p o p
u la
ti o
n 's
m e
a n
h e ig
h t a
n d
P A
L .
F o r
e x a
m p
le ,
th e
r e
c o
m m
e n
d e
d m
e a
n e
n e rg
y i n
ta k e
f o r
a m
a le
p o
p u
la ti o
n o
f th
is a
g e
g ro
u p w
it h
a m
e a
n h
e ig
h t
o f
1 .7
0 m
a n
d a
l if e
s ty
le w
it h a
m e a
n P
A L
o f
1 .7
5 ,
is a
b o
u t
9 .4
M J (
2 2
5 0
k c a l)
/d a
y o
r 1 5
5 k
J (
3 8
k c a
l) /k
g /d
a y t
o m
a in
ta in
a n o
p ti m
u m
p o
p u
la ti o
n m
e d
ia n
B M
I o f
2 1 .0
( W
H O
/F A
O ,
2 0
0 2 ),
w it h a
n i n
d iv
id u a
l ra
n g
e o
f a b
o u
t 9 .0
t o
1 0
.3 M
J (
2 1
5 0 t
o 2
4 5
0 k
c a
l) /d
a y o
r 1
4 5 t
o 1
6 0
k J (
3 5 t
o 3
9
k c a
l) /k
g /d
a y t
o m
a in
ta in
t h e
i n
d iv
id u
a l B
M I
lim it s o
f 1 8
.5 t
o 2
4 .9
( W
H O
, 2
0 0 0
).
43
Energy requirements of adults
T A
B L E
5 .7
D a
il y a
v e
ra g
e e
n e
rg y r
e q
u ir
e m
e n
t fo
r w
o m
e n
a g
e d
1 8 t
o 2
9 .9
y e
a rs
*
M e a
n
B M
R /k
g a
D a il
y e
n e rg
y r
e q
u ir
e m
e n
t a
c c o
rd in
g t
o B
M R
f a c to
r (o
r P
A L
) a n
d b
o d
y w
e ig
h t
in d
ic a te
d
H e ig
h t
(m )
fo r
w e
ig h
t
1 .4
5 ×
B M
R
1 .6
0 ×
B M
R
1 .7
5 ×
B M
R
1 .9
0 ×
B M
R
2
.0 5 ×
B M
R
2 .2
0 ×
B M
R
B M
I v a lu
e s :b
k g
k J
k c a
l
M J
k J /k
g
k c a l
k c a l/ k g
M
J
k J /k
g k c a l
k c a
l/ k g
M J
k J /k
g k c a l
k c a
l/ k g
M J
k J /k
g k c a l
k c a
l/ k g
M
J
k J /k
g k c a l
k c a l/ k g
M J
k J /k
g k c a l
k c a
l/ k g
2 4
.9
2 1 .0
1 8 .5
4 5
1
0 7
2
6 7
.0
1 5 5
1
6 5
0
3 7
7 .7
1
7 0
1 8
5 0
4 1
8 .4
1 9
0 2 0
0 0
4 4
9 .2
2 0
5 2 2
0 0
4 9
9
.9 2
2 0
2 3
5 0
5 2
1 0
.6 2 3
5 2
5 5 0
5 7
1 .3
4 1 .4
6 1
.5 6
5 0
1
0 3
2
5 7
.4
1 5 0
1
8 0
0
3 6
8 .2
1
6 5
1 9
5 0
3 9
9 .0
1 8
0 2 1
5 0
4 3
9 .8
1 9
5 2 3
5 0
4 7
1 0
.5 2
1 0
2 5
0 0
5 0
1 1
.3 2 2
5 2
7 0 0
5 4
1 .4
2 1 .5
4 1
.6 4
5 5
9
9
2 4
7 .9
1
4 5
1
9 0
0
3 5
8 .7
1
6 0
2 1
0 0
3 8
9 .5
1 7
5 2 3
0 0
4 2
1 0
.3 1
9 0
2 4
5 0
4 5
1 1
.2 2
0 5
2 6
5 0
4 8
1 2
.0 2 2
0 2
8 5 0
5 2
1 .4
9 1 .6
2 1
.7 2
6 0
9
6
2 3
8 .3
1
4 0
2
0 0
0
3 3
9 .2
1
5 5
2 2
0 0
3 7
1 0
.1 1
7 0
2 4
0 0
4 0
1 0
.9 1
8 0
2 6
0 0
4 3
1 1
.8 1
9 5
2 8
0 0
4 7
1 2
.7 2 1
0 3
0 5 0
5 1
1 .5
5 1 .6
9 1
.8 0
6 5
9
3
2 2
8 .8
1
3 5
2
1 0
0
3 2
9 .7
1
5 0
2 3
0 0
3 5
1 0
.6 1
6 5
2 5
5 0
3 9
1 1
.5 1
7 5
2 7
5 0
4 2
1 2
.4 1
9 0
2 9
5 0
4 5
1 3
.3 2 0
5 3
2 0 0
4 9
1 .6
2 1 .7
6 1
.8 7
7 0
9
1
2 2
9 .2
1
3 0
2
2 0
0
3 1
1
0 .2
1
4 5
2 4
5 0
3 5
1 1
.2 1
6 0
2 6
5 0
3 8
1 2
.1 1
7 5
2 9
0 0
4 1
1 3
.1 1
8 5
3 1
0 0
4 4
1 4
.0 2 0
0 3
3 5 0
4 8
1 .6
8 1 .8
3 1
.9 5
7 5
8
9
2 1
9 .7
1
3 0
2
3 0
0
3 1
1
0 .7
1
4 5
2 5
5 0
3 4
1 1
.7 1
5 5
2 8
0 0
3 7
1 2
.7 1
7 0
3 0
5 0
4 1
1 3
.7 1
8 5
3 3
0 0
4 4
1 4
.7 1 9
5 3
5 0 0
4 7
1 .7
4 1 .8
9 2
.0 1
8 0
8
7
2 1
1 0 .1
1
2 5
2
4 0
0
3 0
1
1 .2
1
4 0
2 7
0 0
3 4
1 2
.2 1
5 5
2 9
5 0
3 7
1 3
.3 1
6 5
3 2
0 0
4 0
1 4
.3 1
8 0
3 4
5 0
4 3
1 5
.4 1 9
0 3
7 0 0
4 6
1 .7
9 1 .9
5 2
.0 8
8 5
8
6
2 1
1 0 .6
1
2 5
2
5 5
0
3 0
1
1 .7
1
4 0
2 8
0 0
3 3
1 2
.8 1
5 0
3 0
5 0
3 6
1 3
.9 1
6 5
3 3
0 0
3 9
1 5
.0 1
7 5
3 6
0 0
4 2
1 6
.1 1 9
0 3
8 5 0
4 5
1 .8
5 2 .0
1 2
.1 4
* V
a lu
e s r
o u
n d e
d t o
c lo
s e
s t
0 .1
M J /d
, 5
0 k
c a
l/ d
, 5
k J /k
g /d
, 1
k c a l/ k g
/d .
a B
M R
c a
lc u
la te
d f o
r e
a c h w
e ig
h t
fr o
m t
h e
e q
u a ti o
n s i n T
a b
le 5
.2 .
V a lu
e s o
f B
M R
/k g
a re
p re
s e n
te d
f o
r e
a s e o
f c a
lc u
la ti o n
s f
o r
th o s e
w h
o w
is h
t o
u s e
d if fe
re n t
P A
L v
a lu
e s o
r d if fe
re n t
w e
ig h ts
. b H
e ig
h t ra
n g e
s a
re p
re s e n
te d
f o r
e a
c h m
e a
n w
e ig
h t fo
r e a
s e
o f
m a
k in
g d
ie ta
ry e
n e rg
y r
e c o
m m
e n
d a
ti o
n s t
o m
a in
ta in
a n
a d
e q
u a
te B
M I
b a s e
d o
n a
p o p
u la
ti o
n 's
m e
a n
h e ig
h t a
n d
P A
L .
F o r
e x a
m p
le ,
th e
r e
c o
m m
e n
d e
d m
e a
n e
n e rg
y i n
ta k e
f o r
a f
e m
a le
p o p
u la
ti o
n o
f th
is a
g e
g ro
u p
w it h
a m
e a n
h e
ig h t
o f 1
.7 0
m a
n d
a l if e
s ty
le w
it h a
m e
a n
P A
L o
f 1
.7 5
, is
a b o
u t 1
0 .1
M J (
2 4
0 0
k c a l)
/d a
y o
r 1
7 0 k
J
(4 0
k c a
l) /k
g /d
a y t
o m
a in
ta in
a n
o p
ti m
u m
p o p
u la
ti o
n m
e d
ia n
B M
I o
f 2
1 .0
( W
H O
/F A
O ,
2 0 0
2 ),
w it h
a n
i n d
iv id
u a l ra
n g e
o f a
b o
u t 9
.5 t
o 1
1 .2
M J (
2 3
0 0
t o
2 6
5 0
k c a l)
/d a
y o
r 1
6 0
t o
1 7
5 k
J (
3 8
t o
4 2
k c a
l) /k
g /d
a y t
o m
a in
ta in
t h e
i n
d iv
id u
a l B
M I
lim it s o
f 1 8
.5 t
o 2
4 .9
( W
H O
, 2
0 0 0
).
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
44
T A
B L E
5 .8
D a
il y a
v e
ra g
e e
n e
rg y r
e q
u ir
e m
e n
t fo
r w
o m
e n
a g
e d
3 0 t
o 5
9 .9
y e
a rs
*
M e a
n
B M
R /k
g a
D a il
y e
n e rg
y r
e q
u ir
e m
e n
t a
c c o
rd in
g t
o B
M R
f a c to
r (o
r P
A L
) a n
d b
o d
y w
e ig
h t
in d
ic a te
d
H e ig
h t
(m )
fo r
w e
ig h
t
1 .4
5 ×
B M
R
1 .6
0 ×
B M
R
1 .7
5 ×
B M
R
1 .9
0 ×
B M
R
2
.0 5 ×
B M
R
2 .2
0 ×
B M
R
B M
I v a lu
e s :b
k g
k J
k c a
l
M J
k J /k
g
k c a l
k c a l/ k g
M
J
k J /k
g k c a l
k c a
l/ k g
M J
k J /k
g k c a l
k c a
l/ k g
M J
k J /k
g k c a l
k c a
l/ k g
M
J
k J /k
g k c a l
k c a l/ k g
M J
k J /k
g k c a l
k c a
l/ k g
2 4
.9
2 1 .0
1 8 .5
4 5
1
1 3
2
7
7
.3
1 6 5
1
7 5
0
3 9
8 .1
1
8 0
1 9
5 0
4 3
8 .9
1 9
5 2 1
0 0
4 7
9 .6
2 1
5 2 3
0 0
5 1
1 0
.4 2
3 0
2 5
0 0
5 6
1 1
.1 2 5
0 2
6 5 0
5 9
1 .3
4 1 .4
6 1
.5 6
5 0
1
0 5
2
5
7
.6
1 5 0
1
8 0
0
3 6
8 .4
1
7 0
2 0
0 0
4 0
9 .2
1 8
5 2 2
0 0
4 4
1 0
.0 2
0 0
2 4
0 0
4 8
1 0
.7 2
1 5
2 5
5 0
5 1
1 1
.5 2 3
0 2
7 5 0
5 5
1 .4
2 1 .5
4 1
.6 4
5 5
9
8
2 4
7 .8
1
4 5
1
8 5
0
3 4
8 .7
1
5 5
2 0
5 0
3 7
9 .5
1 7
0 2 2
5 0
4 1
1 0
.3 1
8 5
2 4
5 0
4 5
1 1
.1 2
0 0
2 6
5 0
4 8
1 1
.9 2 1
5 2
8 5 0
5 2
1 .4
9 1 .6
2 1
.7 2
6 0
9
3
2 2
8 .1
1
3 5
1
9 5
0
3 3
8 .9
1
5 0
2 1
5 0
3 6
9 .8
1 6
5 2 3
5 0
3 9
1 0
.6 1
7 5
2 5
5 0
4 3
1 1
.4 1
9 0
2 7
5 0
4 6
1 2
.3 2 0
5 2
9 5 0
4 9
1 .5
5 1 .6
9 1
.8 0
6 5
8
8
2 1
8 .3
1
3 0
2
0 0
0
3 1
9 .2
1
4 0
2 2
0 0
3 4
1 0
.1 1
5 5
2 4
0 0
3 7
1 0
.9 1
7 0
2 6
0 0
4 0
1 1
.8 1
8 0
2 8
0 0
4 3
1 2
.6 1 9
5 3
0 0 0
4 6
1 .6
2 1 .7
6 1
.8 7
7 0
8
5
2 0
8 .6
1
2 5
2
0 5
0
2 9
9 .5
1
3 5
2 2
5 0
3 2
1 0
.4 1
5 0
2 5
0 0
3 6
1 1
.2 1
6 0
2 7
0 0
3 9
1 2
.1 1
7 5
2 9
0 0
4 1
1 3
.0 1 8
5 3
1 0 0
4 4
1 .6
8 1 .8
3 1
.9 5
7 5
8
1
1 9
8 .8
1
2 0
2
1 0
0
2 8
9 .7
1
3 0
2 3
5 0
3 1
1 0
.7 1
4 0
2 5
5 0
3 4
1 1
.6 1
5 5
2 7
5 0
3 7
1 2
.5 1
6 5
3 0
0 0
4 0
1 3
.4 1 8
0 3
2 0 0
4 3
1 .7
4 1 .8
9 2
.0 1
8 0
7
8
1 9
9 .1
1
1 5
2
1 5
0
2 7
1
0 .0
1
2 5
2 4
0 0
3 0
1 1
.0 1
3 5
2 6
0 0
3 3
1 1
.9 1
5 0
2 8
5 0
3 6
1 2
.8 1
6 0
3 0
5 0
3 8
1 3
.8 1 7
0 3
3 0 0
4 1
1 .7
9 1 .9
5 2
.0 8
8 5
7
6
1 8
9 .3
1
1 0
2
2 5
0
2 6
1
0 .3
1
2 0
2 4
5 0
2 9
1 1
.2 1
3 0
2 7
0 0
3 2
1 2
.2 1
4 5
2 9
0 0
3 4
1 3
.2 1
5 5
3 1
5 0
3 7
1 4
.1 1 6
5 3
4 0 0
4 0
1 .8
5 2 .0
1 2
.1 4
* V
a lu
e s r
o u
n d e
d t o
c lo
s e
s t
0 .1
M J /d
, 5
0 k
c a
l/ d
, 5
k J /k
g /d
, 1
k c a l/ k g
/d .
a B
M R
c a
lc u
la te
d f o
r e
a c h w
e ig
h t
fr o
m t
h e
e q
u a ti o
n s i n T
a b
le 5
.2 .
V a lu
e s o
f B
M R
/k g
a re
p re
s e n
te d
f o
r e
a s e o
f c a
lc u
la ti o n
s f
o r
th o s e w
h o
w is
h t o
u s e
d if fe
re n t
P A
L v
a lu
e s o
r d if fe
re n t
w e
ig h ts
. b H
e ig
h t ra
n g e
s a
re p
re s e n
te d
f o r
e a
c h m
e a
n w
e ig
h t fo
r e a
s e
o f
m a
k in
g d
ie ta
ry e
n e rg
y r
e c o
m m
e n
d a
ti o
n s t
o m
a in
ta in
a n
a d
e q
u a te
B M
I b
a s e
d o
n a
p o p
u la
ti o
n 's
m e
a n
h e ig
h t a
n d
P A
L .
F o r
e x a
m p
le ,
th e
r e
c o
m m
e n
d e
d m
e a
n e
n e rg
y i n
ta k e
f o r
a f
e m
a le
p o p
u la
ti o
n o
f th
is a
g e
g ro
u p
w it h
a m
e a n
h e
ig h t
o f 1
.7 0
m a
n d
a l if e
s ty
le w
it h a
m e
a n
P A
L o
f 1
.7 5
, is
a b o
u t 9
.8 M
J (
2 3
5 0 k
c a
l) /d
a y o
r 1
6 5
k J (
3 9
k c a
l) /k
g /d
a y t
o m
a in
ta in
a n o
p ti m
u m
p o
p u
la ti o
n m
e d
ia n
B M
I o f
2 1 .0
( W
H O
/F A
O ,
2 0
0 2 ),
w it h a
n i n
d iv
id u a
l ra
n g
e o
f a b
o u t
9 .5
t o
1 0 .4
M J (
2 2
5 0 t
o 2
5 0
0 k
c a
l) /d
a y o
r 1
5 0 t
o 1
7 0
k J (
3 6 t
o 4
1
k c a
l) /k
g /d
a y t
o m
a in
ta in
t h e
i n
d iv
id u
a l B
M I
lim it s o
f 1 8
.5 t
o 2
4 .9
( W
H O
, 2
0 0 0
).
45
Energy requirements of adults
T A
B L E
5 .9
D a
il y a
v e
ra g
e e
n e
rg y r
e q
u ir
e m
e n
t fo
r w
o m
e n
a g
e d
6
0 y
e a
rs *
M e a
n
B M
R /k
g a
D a il
y e
n e rg
y r
e q
u ir
e m
e n
t a
c c o
rd in
g t
o B
M R
f a c to
r (o
r P
A L
) a n
d b
o d
y w
e ig
h t
in d
ic a te
d
H e ig
h t
(m )
fo r
w e
ig h
t
1 .4
5 ×
B M
R
1 .6
0 ×
B M
R
1 .7
5 ×
B M
R
1 .9
0 ×
B M
R
2
.0 5 ×
B M
R
2 .2
0 ×
B M
R
B M
I v a lu
e s :b
k g
k J
k c a
l
M J
k J /k
g
k c a l
k c a l/ k g
M
J
k J /k
g k c a l
k c a
l/ k g
M J
k J /k
g k c a l
k c a
l/ k g
M J
k J /k
g k c a l
k c a
l/ k g
M
J
k J /k
g k c a l
k c a l/ k g
M J
k J /k
g k c a l
k c a
l/ k g
2 4
.9
2 1 .0
1 8 .5
4 5
9
9
2 4
6 .5
1
4 5
1
5 5
0
3 4
7 .1
1
6 0
1 7
0 0
3 8
7 .8
1 7
5 1 8
5 0
4 1
8 .5
1 9
0 2 0
5 0
4 5
9
.2 2
0 5
2 2
0 0
4 9
9 .8
2 2
0 2
3 5 0
5 2
1 .3
4 1 .4
6 1
.5 6
5 0
9
3
2 2
6 .7
1
3 5
1
6 0
0
3 2
7 .4
1
5 0
1 8
0 0
3 6
8 .1
1 6
5 1 9
5 0
3 9
8 .8
1 7
5 2 1
0 0
4 2
9
.5 1
9 0
2 3
0 0
4 6
1 0
.2 2 0
5 2
4 5 0
4 9
1 .4
2 1 .5
4 1
.6 4
5 5
8
8
2 1
7 .0
1
3 0
1
7 0
0
3 1
7 .8
1
4 0
1 8
5 0
3 4
8 .5
1 5
5 2 0
5 0
3 7
9 .2
1 6
5 2 2
0 0
4 0
9
.9 1
8 0
2 3
5 0
4 3
1 0
.7 1 9
5 2
5 5 0
4 6
1 .4
9 1 .6
2 1
.7 2
6 0
8
4
2 0
7 .3
1
2 0
1
7 5
0
2 9
8 .1
1
3 5
1 9
5 0
3 2
8 .8
1 4
5 2 1
0 0
3 5
9 .6
1 6
0 2 3
0 0
3 8
1 0
.3 1
7 0
2 4
5 0
4 1
1 1
.1 1 8
5 2
6 5 0
4 4
1 .5
5 1 .6
9 1
.8 0
6 5
8
0
1 9
7 .6
1
1 5
1
8 0
0
2 8
8 .4
1
3 0
2 0
0 0
3 1
9 .1
1 4
0 2 2
0 0
3 4
9 .9
1 5
5 2 3
5 0
3 7
1 0
.7 1
6 5
2 5
5 0
3 9
1 1
.5 1 7
5 2
7 5 0
4 2
1 .6
2 1 .7
6 1
.8 7
7 0
7
7
1 8
7 .9
1
1 0
1
9 0
0
2 7
8 .7
1
2 5
2 0
5 0
3 0
9 .5
1 3
5 2 2
5 0
3 2
1 0
.3 1
4 5
2 4
5 0
3 5
1 1
.1 1
6 0
2 6
5 0
3 8
1 1
.9 1 7
0 2
8 5 0
4 1
1 .6
8 1 .8
3 1
.9 5
7 5
7
5
1 8
8 .1
1
1 0
1
9 5
0
2 6
9 .0
1
2 0
2 1
5 0
2 9
9 .8
1 3
0 2 3
5 0
3 1
1 0
.6 1
4 0
2 5
5 0
3 4
1 1
.5 1
5 5
2 7
5 0
3 7
1 2
.3 1 6
5 2
9 5 0
3 9
1 .7
4 1 .8
9 2
.0 1
8 0
7
2
1 7
8 .4
1
0 5
2
0 0
0
2 5
9 .3
1
1 5
2 2
0 0
2 8
1 0
.1 1
2 5
2 4
0 0
3 0
1 1
.0 1
4 0
2 6
5 0
3 3
1 1
.9 1
5 0
2 8
5 0
3 5
1 2
.7 1 6
0 3
0 5 0
3 8
1 .7
9 1 .9
5 2
.0 8
8 5
7
0
1 7
8 .7
1
0 0
2
0 5
0
2 4
9 .6
1
1 5
2 3
0 0
2 7
1 0
.5 1
2 5
2 5
0 0
2 9
1 1
.4 1
3 5
2 7
0 0
3 2
1 2
.3 1
4 5
2 9
5 0
3 4
1 3
.2 1 5
5 3
1 5 0
3 7
1 .8
5 2 .0
1 2
.1 4
* V
a lu
e s r
o u
n d e
d t o
c lo
s e
s t
0 .1
M J /d
, 5
0 k
c a
l/ d
, 5
k J /k
g /d
, 1
k c a l/ k g
/d .
a B
M R
c a
lc u
la te
d f o
r e
a c h w
e ig
h t
fr o
m t
h e
e q u
a ti o
n s i n T
a b
le 5
.2 .
V a lu
e s o
f B
M R
/k g
a re
p re
s e n
te d
f o r
e a
s e o
f c a
lc u
la ti o n
s f
o r
th o s e w
h o
w is
h t o
u s e
d if fe
re n t
P A
L v
a lu
e s o
r d if fe
re n t
w e
ig h ts
. b H
e ig
h t ra
n g e
s a
re p
re s e n
te d
f o r
e a
c h m
e a
n w
e ig
h t fo
r e a
s e
o f
m a
k in
g d
ie ta
ry e
n e rg
y r
e c o
m m
e n
d a
ti o
n s t
o m
a in
ta in
a n
a d
e q
u a te
B M
I b
a s e
d o
n a
p o p
u la
ti o
n 's
m e
a n
h e ig
h t a
n d
P A
L .
F o r
e x a
m p
le ,
th e
r e
c o
m m
e n
d e
d m
e a
n e
n e rg
y i n
ta k e
f o r
a f
e m
a le
p o p
u la
ti o
n o
f th
is a
g e
g ro
u p
w it h
a m
e a n
h e
ig h t
o f 1
.7 0
m a
n d
a l if e
s ty
le w
it h a
m e
a n
P A
L o
f 1
.7 5
, is
a b o
u t 8
.8 M
J (
2 1
0 0 k
c a
l) /d
a y o
r 1
4 5
k J (
3 5
k c a
l) /k
g /d
a y t
o m
a in
ta in
a n o
p ti m
u m
p o
p u
la ti o
n m
e d
ia n
B M
I o f
2 1 .0
( W
H O
/F A
O ,
2 0
0 2 ),
w it h a
n i n
d iv
id u a
l ra
n g
e o
f a b
o u t
8 .5
t o
9 .5
M J (
2 0
5 0
t o
2 2
5 0
k c a l)
/d a
y o
r 1 3
5 t
o 1
5 5 k
J (
3 2
t o
3 7
k c a
l) /k
g /d
a y t
o m
a in
ta in
t h e
i n
d iv
id u
a l B
M I
lim it s o
f 1 8
.5 t
o 2
4 .9
( W
H O
, 2
0 0 0
).
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
46
Energy requirements of adults
47
5.5 OLDER ADULTS AND THE ELDERLY
Many age-related changes that influence energy requirements occur continually throughout the adult life cycle. A decline in BMR with age has been recognized since the studies of Keys, Taylor and Grande (1973), who estimated it at a rate of 1 to 2 percent per decade. Average decreases of 2.9 and 2.0 percent per decade, respectively, for men and women of normal weight (BMI of 18.5 to 25.0 kg/m2) were calculated more recently (Roberts and Dalall, 2001; Food and Nutrition Board/Institute of Medicine, 2002). The decreases were 3.1 and 1.9 percent per decade among overweight men and women, respectively (Roberts and Dalall, 2001). The decline is not linear, and has a suggested breakpoint at about 40 years of age in men and 50 years in women (Poehlman, 1992; Poehlman et al., 1993). The organization of the data in Table 5.10 by decades points to a breakpoint at about 50 years of age for both genders. This decrease in BMR has been explained partly by the reduction in fat-free mass that occurs with ageing, and by changes in the composition of that fat-free mass (Piers et al., 1998). Several studies, however, suggest that even after adjusting for changes in fat-free mass, BMR is 5 percent lower in older persons compared with young adults (Roberts and Dalall, 2001).
On the other hand, body weight tends to increase with age in many societies. For example, there are more overweight men and women (defined as BMI > 25) than people with BMI 25 in the database of the United States Academy of Sciences (Table 5.10) (Food and Nutrition Board/Institute of Medicine, 2002). The largest series currently available of TEE and BMR measurements in people 70 to 79 years old involved 150 men and 150 women randomly recruited in two large cities of the United States. Average weight and BMI were 82.4 kg and 27.4 among men, and 70.6 kg and 27.3 among women (Blanc et al., 2001). Overweight and obesity increase BMR and TEE owing to the increase in the fat-free mass needed to carry the extra weight and to the increased energy cost of activities. However, BMR per unit of body weight is reduced in overweight and obese subjects owing to the larger gain in fat mass relative to metabolically active fat-free mass.
Habitual physical activity, and hence TEE, decrease after a given age (Black et al., 1996; Roberts, 1996). However, studies with standardized activity protocols in a whole body calorimeter did not show differences in TEE between young and old adults (Vaughan, Zurlo and Ravussin, 1991; Pannemans and Westerterp, 1995). Furthermore, although maximal oxygen consumption decreases progressively with age (Suominen et al., 1980), some elderly individuals who have remained physically active are able to maintain high levels of energy expenditure, with PAL values as high as 2.48 (Reilly et al., 1993; Withers et al., 1998). This indicates that the age at which TEE and energy requirements start decreasing depends on individual, social and cultural features that promote or limit habitual physical activity among older adults.
Calculation of energy requirements for the elderly based on PAL is highly dependent on the accuracy with which BMR is measured or estimated. For example, the preliminary TEE results – and therefore energy requirements – of 70 to 79 year-old people in a United States study on health, ageing and body composition (Blanc et al., 2001) were 10.1 1.8 MJ/day for men, and 8.0 1.5 MJ/day for women. Based on actual measurements of BMR (men: 5.9 0.1 MJ/day BMR; women: 4.8 0.1 MJ/day BMR), mean PAL was 1.72 among men and 1.68 among women, but using the predictive equations in Table 5.2, PAL would be 1.55 for men and 1.47 for women. The error in prediction may have been associated with the excessive weight of this population group.
In conclusion, energy requirements for older adults and the elderly should be calculated on the basis of PALs, just as they are calculated for younger adults. Therefore, the accuracy with which BMR of older adults can be estimated becomes of primary importance. As more reliable information on BMR of older adults with differing lifestyles, body composition and physical activity becomes available, it may be necessary to revise the predictive equations for this age group in order to make better estimations of their energy requirements. Allowances must be made for population groups who are more or less active at an advanced age, rather than using age as the single cut-off point to define energy requirements for the elderly.
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
48
TABLE 5.10
Daily energy expenditure, basal metabolic rate and physical activity level measured in United States adults
Age No. Weight TEE measured with DLW BMR measured individually PAL
years kg MJ kJ/kg kcal kcal/kg MJ kJ/kg kcal kcal/kg
Men, BMI 18.5–25.0
20–30 48 70.7 12.7 180 3 047 43 7.4 105 1 770 25 1.75
30–40 47 71.7 12.4 173 2 964 41 7.0 98 1 676 23 1.78
40–50 22 70.6 12.8 181 3 048 43 7.0 100 1 683 24 1.84
50–60 8 73.1 10.5 144 2 513 34 6.7 91 1 590 22 1.60
60–70 14 67.8 10.0 148 2 397 35 6.2 92 1 487 22 1.61
70–80 30 70.0 10.1 144 2 407 34 6.3 89 1 497 21 1.62
80–90 4 67.1 7.1 106 1 700 25 6.1 91 1 457 22 1.17
>90 6 65.6 8.1 123 1 935 29 5.9 90 1 415 22 1.38
Women, BMI 18.5–25.0
20–30 76 59.4 10.2 171 2 428 41 5.7 96 1 361 23 1.79
30–40 59 58.7 10.1 172 2 412 41 5.6 95 1 328 23 1.83
40–50 8 58.2 10.2 175 2 441 42 5.4 93 1 300 22 1.89
50–60 18 59.8 9.1 153 2 182 36 5.2 87 1 241 21 1.75
60–70 48 59.0 8.5 145 2 042 35 5.1 86 1 219 21 1.69
70–80 14 59.0 7.9 134 1 888 32 5.1 87 1 229 21 1.55
80–90 6 51.9 5.8 111 1 382 27 4.8 92 1 143 22 1.21
>90 9 52.2 5.7 109 1 356 26 4.9 94 1 168 22 1.17
Overweight men
20–30 10 89.9 13.5 150 3 224 36 7.8 86 1 858 21 1.90
30–40 53 102.4 15.5 151 3 703 36 8.6 84 2 046 20 1.81
40–50 37 94.6 14.5 153 3 465 37 7.9 83 1 878 20 1.88
50–60 17 100.3 14.5 144 3 458 34 7.8 77 1 857 19 1.88
60–70 30 87.8 11.9 136 2 851 32 7.1 80 1 687 19 1.71
70–80 34 84.8 11.0 129 2 624 31 7.2 85 1 713 20 1.55
80–90 7 78.1 9.6 123 2 294 29 6.5 83 1 558 20 1.47
>90 2 77.5 7.8 101 1 863 24 6.5 84 1 550 20 1.29
Overweight women
20–30 33 83.4 11.4 136 2 713 33 6.4 77 1 536 18 1.78
30–40 41 83.9 11.7 139 2 794 33 6.6 79 1 587 19 1.78
40–50 14 96.9 12.7 131 3 032 31 7.1 73 1 696 18 1.80
50–60 29 83.3 9.8 118 2 349 28 5.9 71 1 409 17 1.68
60–70 46 78.2 8.6 110 2 061 26 5.7 74 1 374 18 1.52
70–80 19 69.3 7.8 113 1 868 27 5.2 75 1 234 18 1.51
80–90 6 62.8 7.3 116 1 748 28 5.2 82 1 233 20 1.42
>90 7 74.8 7.4 99 1 766 24 5.6 75 1 332 18 1.33
Sources: Roberts and Dallal, 2001; Food and Nutrition Board/Institute of Medicine, 2002.
Energy requirements of adults
49
5.6 RECOMMENDATIONS FOR REGULAR PHYSICAL ACTIVITY
The practice of regular physical activity is associated with the maintenance of adequate body weight, cardiovascular and respiratory health, and fitness,5 and a lower risk of developing chronic non- communicable diseases associated with diet and lifestyle (Erlichman, Kerbey and James, 2001; WHO, 2000; WHO/FAO, 2002; Pollock et al., 1998; Ferro-Luzzi and Martino, 1996; Schoeller, 1998; WHO, 2002; Erlichman, Kerbey and James, no date; American Heart Association, 2002; IARC, 2002; CDC, 1996; World Cancer Research Fund/American Institute for Cancer Research, 1997; Saris et al., 2003). Consequently, dietary energy recommendations to satisfy requirements should be accompanied by recommendations to perform adequate amounts of physical activity regularly.
There is consensus among experts that a habitual PAL of 1.70 or higher is associated with a lower risk of overweight and obesity, cardiovascular disease, diabetes and several types of cancer (Black et al., 1996; Erlichman, Kerbey and James, 2001; Pollock et al., 1998; Ferro-Luzzi and Martino, 1996; Schoeller, 1998; Erlichman, Kerbey and James, no date; World Cancer Research Fund/American Institute for Cancer Research, 1997; Saris et al., 2003). Therefore, it is particularly important to recommend regular physical activity to populations and individuals with a sedentary lifestyle or one of light activity. Those with moderately or vigorously physically active lifestyles already have a habitual physical activity close to, or higher than, the health-associated PAL threshold of 1.70 times BMR. Recommendations for these individuals should be aimed at the maintenance of that activity level.
5.6.1 Frequency, duration and intensity of physical activity
Table 5.11 summarizes the minimum amounts of exercise, expressed in terms of frequency, duration and intensity, advocated by several organizations to maintain and promote health among adults (WHO/FAO, 2002; Pollock et al., 1998; WHO, 2002; American Heart Association, 2002; IARC, 2002; CDC, 1996; World Cancer Research Fund/American Institute for Cancer Research, 1997; Saris et al., 2003). The conclusions reached in the cited publications can be summarized as follows:
There is consensus that, in order to promote general health, at least 30 minutes of moderate to vigorous activity should be performed, three or more days per week.
Sedentary people and those with low physical fitness levels will benefit from the lower amounts of exercise prescribed in Table 5.11 (i.e. 30 minutes of moderate activity, three days per week). To obtain increased benefits, those with better conditions should exercise longer and/or at a higher intensity (e.g. 60 minutes of moderate or vigorous activity, five or more days per week).
Longer periods of exercise are required to maintain a healthy body weight and to reduce the risk of obesity than are needed to help reduce the risk of chronic diseases such as coronary heart disease and diabetes mellitus.
Sixty minutes of daily exercise have been advocated to increase the PAL of sedentary people to a value of 1.75 or greater, assist in weight maintenance and play a part in the prevention of some types of cancer, especially colorectal and breast cancer.
Any activity that is rhythmic and aerobic in nature, uses large muscle groups and can be maintained continuously is recommended for general health and fitness. Activities that can be practised as part of everyday life are particularly useful. Examples include brisk walking, climbing stairs, cycling, dancing, jogging/running, hiking, low-impact aerobic exercises, swimming, skipping rope, ice/roller skating and various endurance games and sport activities.
Duration is related to intensity. Thus, when ranges are given in Table 5.11 (e.g. 30 to 60 minutes at 50 to 80 percent capacity), lower-intensity activity should be conducted over a longer period of time (i.e. 60 minutes at 50 percent capacity [moderate activity], or 30 minutes at 80 percent aerobic capacity [vigorous activity]).
5 The term “fitness” encompasses cardiorespiratory health, appropriate body composition (including fat distribution), muscular strength, endurance and flexibility, and it can generally be described as the ability to perform moderate to vigorous physical activity without becoming excessively tired (Pollock et al., 1998).
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
50
TABLE 5.11
Minimum frequency, duration and intensity of physical activity advocated by selected organizations
Organization Recommendation
World Health Organization (WHO, 2002) 30 minutes of moderate activity every day.
World Cancer Research Fund/American Institute for Cancer Research (1997)
30 minutes of vigorous or 60 minutes of moderate activity daily, plus additional 30 to 60 minutes of vigorous activity once a week.
United States Centers for Disease Control and Prevention (CDC, 1996)
30 minutes of moderate activity on all or most days of the week.
American Heart Association (2002) 30 to 60 minutes of exercise at 50 to 80% aerobic capacity, at least 3 to 4 days per week.
American College of Sports Medicine (Pollock et al., 1998)
For cardio-respiratory fitness and body composition: 20 to 60 minutes of continuous or intermittent (bouts of at least 10 minutes) aerobic activity at 55 to 90% maximum heart rate, or at 40 to 85% maximum oxygen uptake, 3 to 5 days per week.
For muscular strength and endurance, body composition and flexibility: One set of 8 to 10 exercises, with 8 to 12 repetitions of each exercise, 2 to 3 days per week.
International Agency for Research on Cancer (IARC, 2002)
To maintain healthy body weight: 60 minutes moderate activity on all or most days of the week.
a
For cancer prevention: Substitute moderate for vigorous activity several times per week.
International Association for the Study of Obesity (Saris et al., 2002)
To prevent weight regain in formerly obese individuals: 60 to 90 minutes of moderate activity daily, or shorter periods of vigorous activity.
To prevent transition to overweight or obesity: 45 to 60 minutes of moderate activity daily, or 1.7 PAL. For children, more activity time is recommended.
a Endorsed by the joint WHO/FAO Expert Consultation on Diet, Nutrition and the Prevention of Chronic Diseases (FAO/WHO,
2002).
An activity regimen of moderate- rather than high-intensity exercise is recommended, and total fitness is more readily attained with longer sessions. Consequently, it may be better to suggest activity of moderate intensity for 60 minutes than of high intensity for 30 minutes.
Although longer sessions are generally preferable, their duration may hinder compliance among some people. In these cases it may be appropriate to recommend accumulated bouts of activity for shorter durations throughout the day (e.g. 15 minutes two or four times daily, instead of 30 or 60 minutes once daily).
For general populations, particularly those with sedentary occupations, a joint WHO/FAO Expert Consultation on Diet, Nutrition and the Prevention of Chronic Diseases (WHO/FAO, 2002) recently advocated the performance of moderate-intensity activity, such as brisk walking, for a total of one hour per day on most days of the week to help maintain a healthy body weight and reduce the risk of co-morbid diseases associated with overweight. This level of exercise may be considered part of the daily routine of people with occupations entailing moderate or vigorous, energy-demanding physical activity for one or more hours, five or more days per week.
REFERENCES American Heart Association. 2002. Physical activity and cardiovascular health: How much physical activity is enough? www.americanheart.org/presenter.jhtml?identifier=830. Arciero, P.J., Goran, M.I., Gardner, A.M., Ades, P.A., Tyzbir, R.S. & Poehlman, E.T. 1993. A practical equation to predict resting metabolic rate in older females. J. Am. Geriatr. Soc., 41: 389–395. Bandini, L.G., Schoeller, D.A., Fukagawa, N.K., Wykes, L.J. & Dietz, W.H. 1991. Body composition and energy expenditure in adolescents with cerebral palsy or myelodysplasia. Pediatr. Res., 29: 70–77. Black, A., Coward, W., Cole, T. & Prentice, A. 1996. Human energy expenditure in affluent societies: an analysis of 574 doubly labelled water measurements. Eur. J. Clin. Nutr., 50: 72–92. Blanc, S., Schoeller, D., Bauer, D., Danielson, M.E., Harris, T., Kritchevsky, S.B., Taaffe, D. & Everhart,
J. 2001. Free-living energy requirements of the well-functioning elderly: The Health, Aging and Body Composition Study. Paris, First meeting of the International Academy on Nutrition and Aging, June 2001.
Energy requirements of adults
51
CDC. 1996. Physical activity and health. A report of the surgeon general. United States Centers for Disease Control and Prevention (CDC). www.cdc.gov/nccdphp/sgr/chapcon.htm. Cole. T.J. 2002. The Oxford Brookes BMR database – a reanalysis. Report commissioned by FAO for the joint FAO/WHO/UNU Expert Consultation on Energy in Human Nutrition. Coward, A. 1998. Contributions of the doubly labelled water method to studies of energy balance in the Third World. Am. J. Clin. Nutr., 68 (suppl.): 962S–969S. Cruz, C.M., da Silva, A.F. & dos Anjos, L.A. 1999. A taxa metabolica basal e superestimade pelas equacoes preditivas em universitaras do Rio de Janeiro, Brasil. Arch. Latinoam. Nutr., 49: 232–237. de Boer, J.O., van Es, A.J.H., Voorrips, L.E., Blockstra, F. & Vogt, J.E. 1988. Energy metabolism and requirements in different ethnic groups. Eur. J. Clin. Nutr., 42: 983–997. Erlichman, J., Kerbey, A. & James, P. 2001. Are current physical activity guidelines adequate to prevent unhealthy weight gain? A scientific appraisal for consideration by an Expert Panel of the International Obesity
Task Force (IOTF). London, IOTF. 113 pp. Erlichman, J., Kerbey, A. & James P. (no date). Physical activity and its impact on health outcomes, Paper II: prevention of unhealthy weight gain and obesity by physical activity: An analysis of the evidence. Ferro-Luzzi, A. & Martino, L. 1996. Obesity and physical activity. In The origins and consequences of obesity, pp. 207–227. Wiley Chichester, Ciba Foundation Symposium 201. Food and Nutrition Board/Institute of Medicine. 2002. Dietary reference intakes for energy, carbohydrate, fiber, fat, fatty acids, cholesterol, protein, and amino acids. Institute of Medicine of the National Academies. Washington, DC, National Academy Press. Hayter, J. & Henry, C.J.K. 1993. Basal metabolic rate in human subjects migrating between tropical and temperate regions – a longitudinal study and review of previous work. Eur. J. Clin. Nutr., 47: 724–734. Hayter, J.E. & Henry, C.J.K. 1994. A re-examination of basal metabolic rate predictive equations: the importance of geographic origin of subjects in sample selection. Eur. J. Clin. Nutr., 48: 702–707. Henry, C.J.K. 2001. Basal metabolic rate studies in humans: measurement and application. Background document prepared for the joint FAO/WHO/UNU Expert Consultation on Energy in Human Nutrition, 2001. Henry, C.J.K. & Rees, D.G. 1991. New predictive equations for the estimation of basal metabolic rate in tropical peoples. Eur. J. Clin. Nutr., 45: 177–185. IARC. 2002. Handbook of cancer prevention. Volume 6: Weight control and physical activity. Lyons, France, International Agency for Research on Cancer (IARC) Press. Ismail, M.N., Ng, K.K., Chee, S.S., Roslee, R. & Zawiah, H. 1998. Predictive equations for the estimations of basal metabolic rate in Malaysian adults. Mal. J. Nutr., 4: 81–90. James, W.P.T. & Schofield, E.C. 1990. Human energy requirements. A manual for planners and nutritionists. Oxford, UK, Oxford Medical Publications under arrangement with FAO. Keys, A., Taylor, H.L. & Grande, F. 1973. Basal metabolism and age of adult man. Metab., 22: 579–587. Pannemans, D.L.E. & Westerterp, K.R. 1995. Energy expenditure, physical activity and basal metabolic rate of elderly subjects. Br. J. Nutr., 73: 571–581. Piers, L.S. & Shetty, P.S. 1993. Basal metabolic rates of Indian women. Eur. J. Clin. Nutr., 47: 586–591. Piers, L.S., Soares, M.J., McCormack, L.M. & O’Dea, K. 1998. Is there evidence for an age-related reduction in metabolic rate? J. Appl. Physiol., 85: 2196–2204. Poehlman, E.T. 1992. Energy expenditure and requirements in aging humans. J. Nutr., 122: 2057–2065. Poehlman, E.T., Goran, M.J., Gardner, A.W., Ades, P.A., Arciero, P.J., Katzman-Rooks, S.M.,
Montgomery, S.M., Toth, M.J. & Sutherland, P.T. 1993. Determinants of decline in resting metabolic rate in aging females. Am. J. Physiol., 264: E450–E455. Pollock, M.L., Gaesser, G.A., Butcher, J.D., Despres, J.P., Dishman, R.K., Franklin, B.A. & Ewing-
Garber, C. 1998. American College of Sports Medicine Position Stand on the recommended quantity and quality of exercise for developing and maintaining cardiorespiratory and muscular fitness, and flexibility in healthy adults. Med. Sci. Sports Exerc., 30: 975–991. Prentice, A.M., Leavelesley, K., Murgatroyd, P.R., Coward, W.A., Schorah, C.J., Bladon, P. &
Whitehead, R.G. 1989. Is severe wasting in elderly mental patients caused by an excessive energy requirement? Age Aging, 18: 158–167. Ramirez-Zea, M. 2002. Validation of three predictive equations for basal metabolic rate. Report commissioned by FAO for the joint FAO/WHO/UNU Expert Consultation on Energy in Human Nutrition, 2002. Ravussin, E., Harper, I.T., Rising, R. & Bogardus, C. 1991. Energy expenditure by doubly labelled water: Validation in lean and obese subjects. Am. J. Physiol., 261: E402–E409. Reilly, J.J., Lord, A., Bunker, V.W., Prentice, A.M., Coward, W.A., Thomas, A.J. & Briggs, R.S. 1993. Energy balance in healthy elderly women. Br. J. Nutr., 69: 21–27. Roberts, S.B. 1996. Energy requirements of older individuals. Eur. J. Clin. Nutr., 50 (suppl. 1): S112–S118. Roberts, S. & Dallal, D.E. 2001. Energy requirements and aging. Energy working paper No. 8R prepared for the joint FAO/WHO/UNU Expert Consultation on Energy in Human Nutrition, 2001.
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
52
Saris, W.H., Blair, S.N., van Baak, M.A., Eaton, S.B., Davies, P.S., Di Pietro, L., Fogelholm, M., Rissanen,
A., Schoeller, D., Swinburn, B., Tremblay, A., Westerterp, K.R. & Wyatt, H. 2003. How much physical activity is enough to prevent unhealthy weight gain? Outcome of the IASO 1st Stock Conference and consensus statement. Obes. Rev., 4: 101–114. Schoeller, D. 1998. Balancing energy expenditure and body weight. Am. J. Clin. Nutr., 68 (suppl.): 956S–961S. Schofield, W.N. 1985. Predicting basal metabolic rate, new standards and review of previous work. Hum. Nutr. Clin. Nutr., 39C (suppl. 1): 5–41. Schulz, L.O., Alger, S., Harper, I., Wilmore, J.H. & Ravussin, E. 1992. Energy expenditure of elite female runners measured by respiratory chamber and doubly labelled water. J. Appl. Physiol., 72: 23–28. Soares, M.J., Francis, D.G. & Shetty, P.S. 1993. Predictive equations for basal metabolic rates of Indian males. Eur. J. Clin. Nutr., 47: 389–394. Soares, M.J. & Shetty, P.S. 1988. Validity of Schofield’s predictive equations for basal metabolic rates of Indians. Indian J. Med. Res., 88: 253–260. Stroud, M.A., Coward, W.A. & Sawyer, M.B. 1993. Measurements of energy expenditure using isotope- labelled water (2H2
18O) during an Arctic expedition. Eur. J. Appl. Physiol., 67: 375–379. Suominen, H., Heikkinen, E., Parkatti, T., Frosberg, S. & Kiiskinen, A. 1980. Effects of lifelong physical training on functional aging in men. J. Appl. Physiol., 68: 302–308. Valencia, M.E., Moya, S.Y., McNeill, G. & Haggarty, P. 1994. Basal metabolic rate and body fatness of adult men in northern Mexico. Eur. J. Clin. Nutr., 48: 205–211. Vaughan, L., Zurlo, F. & Ravussin, E. 1986. Aging and energy expenditure. Am. J. Clin. Nutr., 53: 821–825. Westerterp, K.R., Saris, W.H.M., Van Es, M. & ten Hoor, F. 1986. Use of the doubly labelled water technique in man during heavy sustained exercise. J. Appl. Physiol., 61(6): 2162–2167. WHO. 1985. Energy and protein requirements: Report of a joint FAO/WHO/UNU expert consultation. WHO Technical Report Series No. 724. Geneva. WHO. 1995. Physical status: The use and interpretation of anthropometry. Report of a WHO expert committee. WHO Technical Report Series No. 854. Geneva. WHO. 2000. Obesity: preventing and managing the global epidemic. Report of a WHO Consultation. WHO Technical Report Series No. 894. Geneva. WHO. 2002. World Health Day 2002 “Move for Health”. Questions and Answers. www.who.int/world-health- day/q_and_a.en.shtml. WHO/FAO. 2002. Joint WHO/FAO Expert Consultation on Diet, Nutrition and the Prevention of Chronic Diseases. Draft 28 March 2002. Geneva. Withers, R.T., Smith, D.A., Tucker, R.C., Brinkman, M. & Clark, D.G. 1998. Energy metabolism in sedentary and active 49- to 70-yr-old women. J. Appl. Physiol., 84: 1333–1340. World Cancer Research Fund/American Institute for Cancer Research. 1997. Food, nutrition and the prevention of cancer: a global perspective. Washington, DC, American Institute for Cancer Research.
Energy requirements of pregnancy
53
6. ENERGY REQUIREMENTS OF
PREGNANCY
Dietary intake during pregnancy must provide the energy that will ensure the full-term delivery of a healthy newborn baby of adequate size and appropriate body composition by a woman whose weight, body composition and PAL are consistent with long-term good health and well-being. The ideal situation is for a woman to enter pregnancy at a normal weight and with good nutritional status. Therefore, the energy requirements of pregnancy are those needed for adequate maternal gain to ensure the growth of the foetus, placenta and associated maternal tissues, and to provide for the increased metabolic demands of pregnancy, in addition to the energy needed to maintain adequate maternal weight, body composition and physical activity throughout the gestational period, as well as for sufficient energy stores to assist in proper lactation after delivery. Special considerations must be made for women who are under- or overweight when they enter pregnancy.
This consultation reviewed recent information on the association of maternal weight gain and body composition with the newborn birth weight, on the influence of birth weight on infant mortality, and on the associated metabolic demands of pregnancy (WHO, 1995a; Kelly et al., 1996; Butte and King, 2002), in order to perform factorial calculations of the extra energy required during this period. It was acknowledged that estimates of energy requirements and recommendations for energy intake of pregnant women should be population-specific, because of differences in body size, lifestyle and underlying nutritional status. Well-nourished women raised in affluent or economically developed societies may have different energy needs in pregnancy than women from low-income developing societies; pregnancy energy requirements of stunted or undernourished women may differ from those of overweight and obese women; and physical activity patterns may change during pregnancy to an extent that is determined by socio-economic and cultural factors. Even within a particular society, high variability is seen in the rates of gestational weight gain and energy expenditure of pregnant women, and therefore in their energy requirements.
6.1 GESTATIONAL WEIGHT GAIN AND OPTIMAL PREGNANCY OUTCOME
The WHO Collaborative Study on Maternal Anthropometry and Pregnancy Outcomes (WHO, 1995a; Kelly et al., 1996) reviewed information on 110 000 births from 20 countries to determine anthropometric indicators associated with poor foetal outcomes, such as low birth weight (LBW), intrauterine growth retardation (IUGR) and pre-term birth, and with poor maternal outcomes, such as pre-eclampsia, eclampsia, need for assisted delivery, and postpartum haemorrhage. Attained maternal weight (pre-pregnancy weight plus weight gain) was the most significant predictor of LBW and IUGR (with odds ratios of 2.5 and 3.1, respectively). Low pre-pregnancy weight and BMI, and weight gain between 20 and 28 weeks of gestation were moderate predictors of pre-term delivery (odds ratios of 1.3 and 1.4, respectively), and low maternal height (e.g. 146 compared with 160 cm) was a moderate predictor of caesarean delivery (odds ratio: 1.6) (Merchant, Villar and Kestler, 2001).
Women with short stature, especially in developing countries with inadequate health care systems and high prevalence of impaired growth during childhood, are also at high risk of LBW and pre-term delivery, and of obstetric complications during labour and delivery (WHO, 1995a; Martorell et al., 1981). A study of healthy women with uncomplicated pregnancies in the United States showed a positive association between maternal height and birth weight among white, black and Asian women, but not Hispanic women (Picket, Abrams and Selvin, 2000).
6.1.1 Desirable birth weight and gestational weight gain
Weight gain during pregnancy comprises the products of conception (foetus, placenta, amniotic fluid), the growth of various maternal tissues (uterus, breasts) and the increase in blood, extracellular fluid and maternal fat stores. The desirable amount of weight to be gained is that which is associated with optimal outcome for the mother, in terms of preventing maternal mortality and complications of pregnancy, labour and delivery, and allowing adequate postpartum body weight and lactation
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
54
performance; and with optimal outcome for the infant, in terms of allowing adequate foetal growth and maturation, and in the prevention of gestational and perinatal morbidity and mortality. The WHO Collaborative Study on Maternal Anthropometry and Pregnancy Outcomes showed that birth weights between 3.1 and 3.6 kg, with a mean of 3.3 kg, were associated with the optimal ratio of good foetal and maternal outcomes (WHO, 1995a; Kelly et al., 1996). The range of maternal gestational weight gains associated with such birth weights was between 10 and 14 kg, with a mean of 12 kg. This is in agreement with earlier estimates that healthy women in developing countries, who eat in accordance with appetite, gain 10 to 12 kg (Institute of Medicine, 1992). An analysis of gestational weight gains associated with optimal outcomes and full-term delivery of 3- to 4-kg infants in the United States gave a similar although somewhat higher range (11.5 to 16.0 kg) for women with pre-pregnancy BMI between 19.8 and 26.0 (Institute of Medicine/Food and Nutrition Board, 1990; Abrams, Altman and Pickett, 2000).
This consultation endorsed the WHO recommendation that healthy, well-nourished women should gain 10 to 14 kg during pregnancy, with an average of 12 kg, in order to increase the probability of delivering full-term infants with an average birth weight of 3.3 kg, and to reduce the risk of foetal and maternal complications.
6.2 DETERMINANTS OF THE ENERGY COST OF PREGNANCY
The energy cost of pregnancy is determined by the energy needed for maternal gestational weight gain, which is associated with protein and fat accretion in maternal, foetal and placental tissues, and by the increase in energy expenditure associated with basal metabolism and physical activity. It was estimated by previous FAO/WHO/UNU expert committees and consultations (FAO/WHO, 1973; WHO, 1985) through factorial calculations based on a theoretical model that assumed an average gestational weight gain of 12.5 kg, an average infant birth weight of 3.4 kg, cumulative deposition of 925 g protein and 3 825 g fat, an efficiency of energy utilization of 90 percent, and a cumulative increment of 150 MJ in BMR (Hytten, 1980; Hytten and Chamberlain, 1991). Since then, several longitudinal studies in developed and developing countries have allowed for the revision of these theoretical estimates.
6.2.1 Protein and fat deposition during pregnancy
Protein is deposited predominantly in the foetus (42 percent), but also in the uterus (17 percent), blood (14 percent), placenta (10 percent) and breasts (8 percent) (Hytten, 1980; Hytten and Chamberlain, 1991). Total protein deposition has been estimated indirectly from calculations of total body potassium accretion, measured by whole body counting in a number of studies of pregnant women (Butte and King, 2002). Based on results of the most reliable longitudinal studies, which involved 93 women in Sweden (Forsum, Sadurskis and Wager, 1988), the United Kingdom (Pipe et al., 1979) and the United States (King, Calloway and Margen, 1973; Butte et al., 2003), and assuming a potassium to nitrogen (K:N) ratio of 2.15 meq K/g N in foetal tissues, protein deposition was estimated at 686 g, in association with a gestational weight gain of 13.8 kg (Butte and King, 2002). The corresponding protein gain associated with the mean weight gain of 12 kg (range 10 to 14 kg) observed in the WHO collaborative study would be 597 g (range 497 to 696 g).
Cumulative fat deposition in foetal and maternal tissues contributes substantially to the overall energy cost of pregnancy. Therefore, methodological errors in the estimation of fat accretion can affect significantly the calculation of energy requirements. Calculations based on skin-fold measurements lack the precision for an accurate estimate of changes in fat mass during pregnancy, because fat accumulation is not distributed evenly in all parts of the body. Two-component body composition models based on measurement of total body water, body density or total body potassium are acceptable only if they include appropriate corrections to account for pregnancy-related changes in the hydration, density and potassium content of fat-free mass (Butte and King, 2002). Three- and four-component models where the hydration or density of fat-free mass is measured are acceptable to calculate body fat at various stages of pregnancy.
Fat accretion was calculated from the results of 11 longitudinal studies that used three- and four- component body composition models, or two-component models with corrected constants, in 273 well-nourished pregnant women from the Netherlands (van Raaij et al., 1988; Spaaij, 1993; de Groot
Energy requirements of pregnancy
55
et al., 1994), Sweden (Forsum, Sadurskis and Wager, 1988; Sohlström and Forsum, 1997), the United Kingdom (Pipe et al., 1979; Goldberg et al., 1993) and the United States (Butte et al., 2003; Lederman et al., 1997; Lindsay et al., 1997; Kopp-Hoolihan et al., 1999a). Mean fat accretion measured up to 36 weeks of gestation was 3.7 kg, associated with a mean weight gain of 11.9 kg. Extrapolating the calculations to 40 weeks of gestation increased mean fat accretion to 4.3 kg, associated with a mean weight gain of 13.8 kg (Butte and King, 2002). The fat gain associated with the mean weight gain of 12 kg (range 10 to 14 kg) observed in the WHO collaborative study would be 3.7 kg (range 3.1 to 4.4 kg).
Rates of fat accretion during the first, second and third trimesters of pregnancy were available in a subset of the studies mentioned (Forsum, Sadurskis and Wager, 1988; Pipe et al., 1979; Butte et al., 2003; de Groot et al., 1994; Goldberg et al., 1993; Kopp-Hoolihan et al., 1999a). These were, on average, 8 g/day in the first trimester, and 26 g/day in the second trimester. Results varied markedly in the third trimester, from –7 to 23 g/day (average: 8 g/day), but if the three studies with very low mean values (–7.0, –1.4 and 4.8 g/day) are excluded from calculations, the average accretion rate would be 18 g fat/day in the third trimester.
6.2.2 Basal metabolism in pregnancy
Basal metabolism increases in pregnancy as a result of accelerated tissue synthesis, increased active tissue mass, and increased cardiovascular and respiratory work. Several studies have measured basal or resting metabolic rate at several stages of pregnancy. As energy requirements should be based on healthy populations with favourable pregnancy outcomes, this consultation only considered the results of studies that involved healthy, well-nourished groups of women with adequate weight gains during pregnancy, who gave birth to infants with adequate weights (Table 6.1) (Forsum, Sadurskis and Wager, 1988; de Groot et al., 1994; Goldberg et al., 1993; Durnin et al., 1987; van Raaij et al., 1987; Spaaij et al., 1994; Piers et al., 1995; Muthayya, 1998; Kopp-Hoolihan et al., 1999b; Cikrikci, Gokbel and Bediz, 1999).
As Table 6.1 shows, the cumulative increment in BMR calculated in relation to pre-pregnancy values, or to early pregnancy values when pre-pregnancy BMR was not available, ranged from 124 to 200 MJ, with an average increase of 154 MJ for the entire gestational period. The average increases in BMR over pre-pregnancy values were in the order of 5, 10 and 25 percent for the first, second and third trimesters, respectively. The coefficient of variability of the cumulative increase in BMR was 16 percent between studies, but the variability between women in each study was higher, with a cumulative variability of 45 to 70 percent in many cases. This demonstrates once again that the application of mean population requirements to specific individuals may lead to large errors. The variation in BMR during pregnancy, which is further illustrated by a striking reduction well into the third trimester of pregnancy found among undernourished Gambian women (Lawrence et al., 1987), a depression in BMR up to 24 weeks of gestation reported in groups of well-nourished United Kingdom (Prentice et al., 1989) and Netherlands (Spaaij, 1993) women, and a cumulative reduction or low increase in BMR during pregnancy among some United States women (Kopp-Hoolihan et al., 1999b).
Cumulative increases in BMR are significantly correlated with gestational weight gain (r = 0.79; p < 0.001) and pre-pregnancy percentage fat mass (r = 0.72; p < 0.001) (Prentice et al., 1996). Hence, the cumulative increase of 154 MJ associated with an average gestational weight gain of 12.5 kg (Table 6.1) would correspond to 148 MJ for a weight gain of 12 kg. These values are remarkably close to the 150 MJ estimated from changes in oxygen consumption of individual organs (Hytten, 1980), which was used by previous expert consultations (FAO/WHO, 1973; WHO, 1985).
6.2.3 Total energy expenditure during pregnancy
A review of 122 studies on practices related to work and pregnancy indicated that in most societies women were expected to continue with partial or full household and other duties throughout most of pregnancy (Institute of Medicine, 1992). Similarly, a review and summary of time–motion studies in Scotland, the Netherlands, Thailand, the Philippines, the Gambia and Nepal did not find conclusive evidence that women engaged in less activity during pregnancy and thus reduced their energy expenditure (Prentice et al., 1996). But these studies did not give information about changes in the intensity of the effort associated with habitual tasks. However, there was a suggestion of increased
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
56
efficiency in energy utilization for physical activity during pregnancy, as the energy cost of weight- bearing activities remained fairly constant during the first two trimesters of pregnancy, even though body weight had increased by 5 to 8 kg by the end of the second trimester (Prentice et al., 1996).
Longitudinal measurements with DLW in free-living, well-nourished women in Sweden (Forsum et al., 1992), the United Kingdom (Goldberg et al., 1993 and 1991) and the United States (Butte et al., 2003; Kopp-Hoolihan et al., 1999b) showed a mean increase of 16.5 percent in TEE by the third trimester of pregnancy, compared with non-pregnant values (Table 6.2). Some of these studies provided information at each trimester of pregnancy and in the non-pregnant state, suggesting that TEE increased by about 1, 6 and 17 percent in the first, second and third trimesters of pregnancy, respectively. This was proportional to recorded increments in weight gain of 2, 8 and 18 percent during the same periods ((Butte and King, 2002). The relationship between TEE and weight gain is reflected in the lack of difference between non-pregnant and pregnant women when TEE is expressed per kilogram of body weight (Table 6.2). The estimated increments in TEE were 100, 400 and 1 500 kJ/day (25, 95 and 360 kcal/day) in the first, second and third trimesters of pregnancy, respectively, in association with an average weight gain of 13.8 kg (Butte and King, 2002). For an average gain of 12 kg, the corresponding values would be 85, 350 and 1 300 kJ/day (20, 85 and 310 kcal/day).
Because of the larger increment in BMR, especially in the second and third trimesters of pregnancy (Table 6.1), PAL declined from 1.74 prior to pregnancy to 1.60 in late gestation (Table 6.2). Compared with non-pregnant values, total energy expenditure to activity (activity energy expenditure [AEE]) near the end of gestation ranged from a decrease of 22 percent to an increase of 17 percent, but on average did not differ significantly between non-pregnant women and women in the third trimester of pregnancy (3 percent 15 percent, Table 6.2). However, when expressed per unit of body weight, there was a tendency towards lower AEE/kg/day in the last trimester of pregnancy.
Cross-sectional studies with DLW, HRM or time–motion techniques in Colombia (Dufour, Reina and Spurr, 1999), Nepal (Panter-Brick, 1993), and two (Heini et al., 1991; Lawrence and Whitehead, 1988) of three (including Singh et al., 1989) studies in the Gambia, showed a slight decrease in TEE, ranging from 1 to 7 percent, and larger reductions, from 10 to 38 percent, in AEE by the third trimester of pregnancy, relative to non-pregnant controls (Butte and King, 2002). This was consistent with observations that many women perform less arduous tasks as they approach the end of pregnancy.
6.3 CALCULATION OF ENERGY REQUIREMENTS FOR PREGNANCY
The extra amount of energy required during pregnancy was calculated in association with a mean gestational weight gain of 12 kg by two factorial approaches, using either the cumulative increment in BMR during pregnancy (section 6.2.2) or the cumulative increment in TEE (section 6.2.3), plus the energy deposited as protein and fat (section 6.2.1). In the calculations using the increment in BMR, it was assumed that the efficiency in energy utilization to synthesize protein and fat was 90 percent. Adjustments for efficiency of energy utilization were not necessary in the calculations that used the increment in TEE, as TEE measured with DLW includes the energy cost of synthesis. As Table 6.3 shows, the estimates of the additional energy required during pregnancy were very similar using either BMR or TEE for the calculations: 323 MJ (77 100 kcal) and 320 MJ (76 500 kcal), respectively. These values, which were based on experimental data, differ by only 4 percent from the theoretical estimate of 335 MJ (80 000 kcal) made by the 1981 FAO/WHO/UNU expert consultation (WHO, 1985).
The energy cost of pregnancy is not distributed equally throughout the gestational period. The deposition of protein occurs primarily in the second (20 percent) and third trimesters (80 percent). Assuming that the rate of fat deposition follows the same pattern as the rate of gestational weight gain, 11, 47 and 42 percent of fat is deposited in the first, second and third trimesters, respectively (Institute of Medicine/Food and Nutrition Board, 1990). The increments in BMR in these trimesters are about 5, 10 and 25 percent, respectively (section 6.2.2 and Table 6.2), whereas the increase in TEE for women gaining 12 kg in pregnancy was estimated at about 85, 350 kcal/day and 1 300 kJ/day per trimester (section 6.2.3).
T A
B L E
6 .1
C u
m u
la ti
v e
i n
c re
a s e
i n
b a
s a
l m
e ta
b o
li c
r a
te o
f w
e ll
-n o
u ri
s h
e d
w o
m e
n d
u ri
n g
p re
g n
a n
c y
M e
a n
B M
R M
J /d
P e rc
e n
ta g
e (
% )
c h
a n
g e
i n
B M
R r
e la
ti v e t
o :
P re
-p re
g n
a n
c y
1 s t
tr im
e s te
r C
o u
n tr
y (
re fe
re n
c e )
N o
. W
e ig
h t
g a in
in
4 0 w
e e k s
k g
a P
re -
p re
g n
a n
c y
1 s
t tr
im .
2 n
d t
ri m
. 3
rd t
ri m
.
C u
m u
la ti
v e
in c re
a s e i
n
B M
R a
t 4
0
w e e k s
M J
b 1 s t
tr im
. 2 n
d t
ri m
. 3
rd t
ri m
. 2 n
d t
ri m
. 3 rd
t ri
m .
U K
( D
u rn
in e
t a l. ,
1 9 8
7 )
8 8
1 2 .4
6
.0
6 .3
6
.5
7 .3
1
2 6
5
8
2 2
3
1 6
N e
th e
rl a
n d
s (
v a
n R
a a
ij e
t a
l. ,
1 9
8 7 )
5 7
1 1 .6
1
4 4
S w
e d
e n (
F o
rs u
m , S
a d
u rs
k is
a n d W
a g
e r,
1 9
8 8
) 2
2
1 3 .4
5
.6
6
.0
7 .3
2
0 0
7
3 0
U K
( G
o ld
b e rg
e t a
l. ,
1 9
9 3 )
1 2
1 3 .7
6
.0
6 .3
6
.4
7 .2
1
2 4
5
7
2 0
2
1 4
N e
th e
rl a
n d
s (
v a
n R
a a
ij e
t a
l. ,
1 9
8 7 )
2 6
1 3 .7
5
.4
5 .7
6
.2
6 .6
1
8 9
6
1 5
2 2
9
1 6
N e
th e
rl a
n d
s (
d e G
ro o
t e
t a l. ,
1 9 9
4 )
1 2
1 1 .6
5
.8
6 .3
6
.5
7 .2
1
4 9
9
1 2
2 4
3
1 4
I n
d ia
( P
ie rs
e t
a l. ,
1 9 9
5 )
1 8
1 2 .0
5 .1
5
.6
6 .2
1
4 3
1 0
2 2
I n
d ia
( M
u th
a y y a
, 1
9 9
8 )
2 6
1 1 .3
4
.6
5 .0
5
.3
6 .0
1
5 1
9
1 5
3 0
6
2 0
U S
A (
K o
p p -H
o o
lih a
n e
t a
l. ,
1 9
9 9 )
1 0
1 3 .2
5
.5
5 .4
6
.4
7 .1
1
5 1
-2
1
6
2 9
1 9
3 1
T u rk
e y (
C ik
ri k c i, G
o k b
e l a n
d B
e d iz
, 1
9 9
9 )
2 4
1 2 .3
5 .2
5
.8
6 .4
1
6 2
1 2
2 3
A v e
ra g
e c
1 2 .5
5
.6
5 .7
6
.1
6 .9
1
5 4
5
.3
1 1 .4
2
5 .3
8
.0
1 9 .5
s d
c 0 .9
0
.5
0 .6
0
.4
0 .4
2 4
4
.0
4 .0
4
.3
5 .8
5 .8
a W
e ig
h t g
a in
w a
s e
x tr
a p
o la
te d
t o
4 0
w e
e k s o
f g
e s ta
ti o
n ,
a s s u m
in g t
h a
t th
e a
v e
ra g e
w e
ig h t
g a
in d
u ri
n g
t h
e f ir
s t te
n t
o 1
2 w
e e k s o
f p
re g n
a n c y i s 0
.6 5
k g ,
a n d
t h a
t w
e ig
h t
g a
in
in c re
a s e s i n
t h e
l a s t fo
u r
to e
ig h
t w
e e
k s b
y 0
.4 0
k g /w
e e
k (
H y tt
e n a
n d
C h
a m
b e
rl a
in ,
1 9
9 1
).
b C
a lc
u la
te d
a s c
u m
u la
ti v e i n
c re
a s e t
h ro
u g
h o
u t p
re g n
a n c y i n r
e la
ti o n
t o p
re -p
re g n
a n
c y o
r e a
rl y p
re g
n a
n c y v
a lu
e s o
f B
M R
. c N
o n
-w e ig
h te
d a
v e
ra g e
s a
n d
s ta
n d a
rd d
e v ia
ti o n
s o
f th
e m
e a
n r
e s u
lt s i n t
h e s
tu d ie
s s
h o
w n i n
t h
is t
a b le
.
Energy requirements of pregnancy
57
T A
B L E
6 .2
T o
ta l e
n e
rg y e
x p
e n
d it
u re
m e a
s u
re d
w it
h D
L W
i n
w e
ll -n
o u
ri s
h e
d n
o n
-p re
g n
a n
t a
n d
p re
g n
a n
t w
o m
e n
C o
u n
tr y ,
(r e fe
re n
c e
) N
o .
M e a
s u
re m
e n
t, w
e e k
o f
g e s ta
ti o
n
W e ig
h t
k g
T E
E
M J /d
B
M R
M
J /d
A
E E
M
J /d
P
A L
P
re g
T E
E /
N P
T E
E a
%
P re
g A
E E
/ N
P A
E b
% T
E E
k J /k
g /d
A
E E
k J /k
g /d
1 0
N
P
5 7 .1
9
.8
5 .9
3
.9
1 .6
7
1 7
1
6 9
U
K (
G o ld
b e rg
e t a
l. ,
1 9
9 1 )
1 0
3
6
6 9
.0 b
1 0 .3
7
.3
3 .0
1
.4 2
5
.6
-2 2 .4
1 5
0 c
4 4
b
1 9
N
P
6 0 .7
1
0 .1
5
.6
4 .5
1
.8 0
1
6 6
7 4
S
w e
d e
n (
F o
rs u
m e
t a l. ,
1 9 9
2 )
1 9
3
6
7 2 .7
1
2 .2
7
.3
4 .9
1
.6 7
2
0 .8
8
.9
1 6
8
6 7
2 2
N
P
6 1 .0
1
0 .4
5
.6
4 .8
1
.8 6
1
7 0
7 9
2 2
3
0
7 0 .2
1
2 .5
6
.9
5 .6
1
.8 1
2
0 .2
1
6 .7
1
7 8
8 0
1 2
N
P
6 1 .7
9
.5
6 .1
3
.5
1 .5
7
1 5
4
5 6
U
K (
G o ld
b e rg
e t a
l. ,
1 9
9 3 )
1 2
3
6
7 3 .6
1
1 .3
7
.6
3 .7
1
.4 9
1
8 .2
6
.6
1 5
3
5 0
1 0
N
P
6 3 .5
9
.2
5 .5
3
.7
1 .6
8
1 4
7
5 8
U
S A
( K
o p
p -H
o o
lih a
n e
t a
l. ,
1 9
9 9 b
)
1 0
3
4 -3
6
7 5 .1
1
1 .4
7
.1
4 .4
1
.6 1
2
3 .7
1
6 .6
1
5 3
5 9
3 4
N
P
5 9 .3
1
0 .2
5
.5
4 .7
1
.8 4
1
7 2
7 8
U
S A
( B
u tt
e e
t a
l. , 2
0 0 3
)
3 4
3
6
7 2 .2
1
1 .3
7
.0
4 .3
1
.6 1
1
0 .7
-8
.2
1 5
6
5 9
M e
a n
n o
n -p
re g
n a
n t
6 0 .6
9
.9
5 .7
4
.2
1 .7
4
1 6
4
6 9
s d
d
2 .2
0
.4
0 .2
0
.5
0 .1
1
1 1
1 0
M e
a n
3 0
– 3
6 w
e e k
s
7 2 .1
1
1 .5
7
.2
4 .3
1
.6 0
1
6 .5
3
.0
1 6
0
6 0
s d
2 .2
0
.8
0 .2
0
.9
0 .1
4
6 .9
1
5 .4
1
1
1 3
a P
re g =
p re
g n a
n t.
b N
P =
n o n
-p re
g n a
n t.
c B
a s e d
o n
e s ti m
a te
d m
e a n
b o
d y w
e ig
h t.
d s d
= s
ta n
d a
rd d
e v ia
ti o
n o
f th
e m
e a
n .
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
58
Energy requirements of pregnancy
59
TABLE 6.3
Additional energy cost of pregnancy in women with an average gestational weight gain of 12 kg
*
A. Rates of tissue deposition
1st trimester 2nd trimester 3rd trimester Total deposition g/d g/d g/d g/280 d
Weight gain 17 60 54 12 000
Protein deposition a 0 1.3 5.1 597
Fat deposition a 5.2 18.9 16.9 3 741
B. Energy cost of pregnancy estimated from the increment in BMR and energy deposition
1st trimester 2nd trimester 3rd trimester Total energy cost
kJ/d kJ/d kJ/d MJ kcal
Protein deposition a 0 30 121 14.1 3 370
Fat deposition a 202 732 654 144.8 34 600
Efficiency of energy utilization b 20 76 77 15.9 3 800
Basal metabolic rate 199 397 993 147.8 35 130
Total energy cost of pregnancy (kJ/d) 421 1 235 1 845 322.6 77 100
C. Energy cost of pregnancy estimated from the increment in TEE and energy deposition
1st trimester 2nd trimester 3rd trimester Total energy cost
kJ/d kJ/d kJ/d MJ kcal
Protein deposition a 0 30 121 14.1 3 370
Fat deposition a 202 732 654 144.8 34 600
Total energy expenditure c 85 350 1 300 161.4 38 560
Total energy cost of pregnancy (kJ/d) 287 1,112 2 075 320.2 76 530
* Calculated as suggested by Butte and King (2002). Weight gain and tissue deposition in first trimester computed from last menstrual period (i.e. an interval of 79 days). Second and third trimesters computed as 280/3 = 93 days each. a Protein and fat deposition estimated from longitudinal studies of body composition during pregnancy, and an energy value of
23.6 kJ (5.65 kcal)/g protein deposited, and 38.7 kJ (9.25 kcal)/g fat deposited. b Efficiency of food energy utilization for protein and fat deposition taken as 0.90 (Hytten, 1990).
c Efficiency of energy utilization not included in this calculation, as the energy cost of synthesis is included in the measurement
of TEE by DLW.
Based on these considerations and averaging the two factorial calculations shown in Table 6.3, the extra energy cost of pregnancy is 321 MJ (77 000 kcal) divided into approximately 0.35 MJ/day, 1.2 MJ/day and 2.0 MJ/day (85 kcal/day, 285 kcal/day and 475 kcal/day) during the first, second and third trimesters, respectively. There are many societies with a high proportion of non-obese women who do not seek prenatal advice before the second or third month of pregnancy. Under these circumstances a practical option to achieve the total additional intake of 321 MJ (77 000 kcal) during pregnancy is to add the extra 0.35 MJ/day required in the first trimester to the 1.2 MJ/day required in the second trimester. Rounding numbers for ease of calculation, this consultation recommends that in such societies pregnant women increase their food intake by 1.5 MJ/day (360 kcal/day) in the second trimester, and by 2.0 MJ/day (475 kcal/day) in the third.
The preceding joint FAO/WHO/UNU expert consultation suggested that the additional energy allowance could be lowered in cases where women reduce their activity level during pregnancy. When such a reduction occurred among the women who participated in the studies listed in Table 6.2, it was built into the 24-hour TEE used to calculate the energy cost of pregnancy in Table 6.3. On the other hand, not all women have the option to reduce physical activity during pregnancy. In particularly, low-income women from developing countries must often continue a strenuous work pattern until shortly before delivery. Furthermore, women who are sedentary prior to pregnancy have
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
60
little flexibility to reduce their level of physical activity. Consequently, this consultation does not recommend a reduction in the additional energy allowance for pregnancy.
6.4 SPECIAL CONSIDERATIONS FOR MALNOURISHED, OBESE AND ADOLESCENT
PREGNANT WOMEN
Undernutrition, whether manifested as underweight or as stunting, and obesity increase the risk of poor maternal and foetal outcomes. Ideally, women should begin pregnancy at a healthy weight, defined as a BMI between 18.5 and 24.9 (WHO, 1995a; March of Dimes, 2002). Adolescent girls who are pregnant must fulfil the dietary requirements imposed by growth associated with their age, in addition to the extra demands of pregnancy.
6.4.1 Pregnancy and undernutrition
A large number of women in many parts of the world enter pregnancy at suboptimal weight and/or height. An analysis of studies in 20 countries (Kelly et al., 1996) showed that in ten countries many women had pre-pregnancy weights of < 50 kg and heights of < 150 cm. These cut-off points were associated with increased risks of maternal complications. In addition, weight below 45 kg or height below 148 cm were associated with poor foetal outcomes. The linear relationship between gestational weight gain and birth weight is influenced by maternal pre-pregnancy BMI, such that women with a BMI < 18.5 must gain more weight than those with a normal BMI in order to have babies with adequate birth weight. It is then particularly important that underweight women increase their energy intake to gain the prescribed 10 to 14 kg during pregnancy, depending on their height (e.g. taller women should strive for a weight gain of 14 kg). Gestational weight gains as high as 18 kg have been suggested for undernourished women (Institute of Medicine/Food and Nutrition Board, 1992).
The association of short stature with increased risk of either delivering a low birth weight infant or requiring special assistance during delivery owing to cephalo-pelvic disproportion (Merchant, Villar and Kestler, 2001) indicates the importance for such women to have adequate prenatal attention and access to appropriate care during labour and delivery. This also reinforces the recommendations for good nutrition and measures to prevent repeated infections during childhood, which may result in stunting and in pregnancy-related problems at a later age.
6.4.2 Pregnancy and obesity
Maternal obesity is also associated with a higher risk of maternal and foetal complications. As for undernutrition, the relative risks of neural tube defects, congenital malformations and pre-term delivery are higher in overweight and obese women (March of Dimes, 2002). Incidences of hypertension, gestational diabetes and the need for caesarean section operations are also higher than in women with normal weight.
Women with a pre-pregnancy BMI > 25 tend to have babies with high birth weights, even when the women have relatively low gestational weight gains (Institute of Medicine/Food and Nutrition Board, 1992; Shapiro, Sutija and Bush, 2000). As this may lead to problems during delivery, it is likely that such women will be better off gaining weight at, or somewhat below, the lower limit of the 10 to 14 kg range recommended for women with normal BMI. It has been suggested that weight gain should be as low as 7 kg for women who enter pregnancy with BMI > 26 (Institute of Medicine/Food and Nutrition Board, 1992).
6.4.3 Pregnancy in adolescence
It is important to satisfy the energy needs of adolescence, when as much as 20 percent of total growth in stature can occur (WHO, 1995b). These needs increase during gestation and must be satisfied by appropriate dietary intakes to satisfy the requirements of both adolescence and pregnancy, in order to allow adequate maternal and foetal growth.
Compared with older women, those under 18 years of age have an increased risk of pre-term delivery, giving birth to infants with low birth weight or small size for gestational age, and requiring special obstetrical assistance (Kumbi and Isehak, 1999; Larsson and Svanberg, 1983; Bwibo, 1985; Gortzak-Uzan et al., 2001). The risks increase with decreasing age (Bwibo, 1985; Bhalerao et al.,
Energy requirements of pregnancy
61
1990). Owing to the high incidence of complications associated with an immature body and small size, it is essential that, in addition to a suitable diet, adolescent pregnant girls receive adequate prenatal care and have access to appropriate medical facilities during labour and delivery.
REFERENCES Abrams, B., Altman, S.L. & Pickett, K.E. 2000. Pregnancy weight gain: still controversial. Am. J. Clin. Nutr., 71: 1233S–1241S. Bhalerao, A., Desai, S., Dastur, N. & Daftary, S. 1990. Outcome of teenage pregnancy. J. Postgrad. Med., 36: 136–139. Butte, N. & King, J.C. 2002. Energy requirements during pregnancy and lactation. Energy background paper prepared for the joint FAO/WHO/UNU Consultation on Energy in Human Nutrition. Butte, N.F., Ellis, K.J., Wong, W.W., Hopkinson, J.M. & Smith, E.O. 2003. Compostion of weight gain impacts maternal fat retention and infant birth weight. Am. J.Obstet. Gynecol., 189: 1423–1432. Bwibo, N. 1985. Birthweights of infants of teenage mothers in Nairobi. Acta Paediatr. Scand., 319 (suppl.): 89– 94. Cikrikci, E., Gokbel, H. & Bediz, C. 1999. Basal metabolic rates of Turkish women during pregnancy. Ann. Nutr. Metab., 43: 80–85. de Groot, L.C.P.G.M., Boekholt, H.A., Spaaij, C.J.K., van Raaij, J.M.A., Drijvers, J.J.M.M., van der
Heijden, L.J.M. & Hautvast, J.G.A.J. 1994. Energy balances of Dutch women before and during pregnancy: limited scope for metabolic adaptations in pregnancy. Am. J. Clin. Nutr., 59: 827–832. Dufour, D.L., Reina, J.C. & Spurr, G. 1990. Energy intake and expenditure of free-living, pregnant Colombian women in an urban setting. Am. J. Clin. Nutr., 70: 269–276. Durnin, J.V.G.A., McKillop, F.M., Grant, S. & Fitzgerald, G. 1987. Energy requirements of pregnancy in Scotland. Lancet, 2: 897–900. FAO/WHO. 1973. Energy and protein requirements: Report of a joint FAO/WHO ad hoc expert committee. FAO Nutrition Meetings Report Series No. 52. WHO Technical Report Series No. 522. Rome and Geneva. Forsum, E., Sadurskis, A. & Wager, J. 1988. Resting metabolic rate and body composition of healthy Swedish women during pregnancy. Am. J. Clin. Nutr., 47: 942–947. Forsum, E., Kabir, N., Sadurskis, A., Westerterp, K. 1992. Total energy expenditure of healthy Swedish women during pregnancy and lactation. Am. J. Clin. Nutr., 56: 334–342. Goldberg, G.R., Prentice, A.M., Coward, W.A., Davies, H.L., Murgatroyd, P.R., Sawyer, M.B., Ashford,
J. & Black, A.E. 1991. Longitudinal assessment of the components of energy balance in well-nourished lactating women. Am. J. Clin. Nutr., 54: 788–798. Goldberg, G.R., Prentice, A.M., Coward, W.A., Davies, H.L., Murgatroyd, P.R., Wensing, C., Black, A.E.,
Harding, M. & Sawyer, M. 1993. Longitudinal assessment of energy expenditure in pregnancy by the doubly labelled water method. Am. J. Clin. Nutr., 57: 494–505. Gortzak-Uzan, L., Hallak, M., Press, F., Katz, M. & Shoham-Vardi, I. 2001. Teenage pregnancy: risk factors for adverse perinatal outcome. J. Matern. Fetal Med., 10: 393–397. Heini, A., Schutz, Y., Diaz, E., Prentice, A.M., Whitehead, R.G. & Jequier, E. 1991. Free-living energy expenditure measured by two independent techniques in pregnant and non-pregnant Gambian women. Am. J. Physiol., 261: E9–E17. Hytten, F.E. 1980. Nutrition. In F.E. Hytten and G. Chamberlain, eds. Clinical physiology in obstetrics. Part 2. Nutrition and metabolism. Oxford, UK, Blackwell Scientific Publications. Hytten, F.E. & Chamberlain, G. 1991. Clinical physiology in obstetrics. Oxford, UK, Blackwell Scientific Publications. Institute of Medicine. 1992. Nutrition issues in developing countries. Washington, D,C., National Academy Press. Institute of Medicine/Food and Nutrition Board. 1990. Nutrition during pregnancy. Washington DC, National Academy Press. Kelly, A., Kevany, J., de Onis, M. & Shah, P.M. 1996. A WHO collaborative study of maternal anthropometry and pregnancy outcomes. Int. J. Gynecol. Obst., 53: 219–233. King, J.C., Calloway, D.H. & Margen, S. 1973. Nitrogen retention, total body 40K and weight gain in teenage pregnant girls. J. Nutr., 103: 772–785. Kopp-Hoolihan, L.E., Van Loan, M.D., Wong, W.W. & King, J.C. 1999a. Fat mass deposition during pregnancy using a four-component model. J. Appl. Physiol., 87: 196–202. Kopp-Hoolihan, L.E., Van Loan, M.D., Wong, W.W. & King, J.C. 1999b. Longitudinal assessment of energy balance in well-nourished, pregnant women. Am. J. Clin. Nutr., 69: 697–704. Kumbi, S. & Isehak, A. 1999. Obstetric outcome of teenage pregnancy in northwest Ethiopia. East African Medical Journal, 76: 138–140.
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
62
Larsson, J. & Svanberg, L. 1983. Teenage deliveries in a Swedish population in the 1970s. Acta Obstet. Gynecol. Scand., 62: 467–472. Lawrence, M. & Whitehead, R.G. 1988. Physical activity and total energy expenditure of child-bearing Gambian village women. Eur. J. Clin. Nutr., 42: 145–160. Lawrence, M., Coward, W.A., Lawrence, F., Cole, T.J. & Whitehead, R.G. 1987. Fat gain during pregnancy in rural African women: the effect of season and dietary status. Am. J. Clin. Nutr., 45: 1442–1450. Lederman, S.A., Paxton, A., Heymsfield, S.B., Wang, J., Thornton, J. & Pierson, R.N. Jr. 1997. Body fat and water changes during pregnancy in women with different body weight and weight gain. Obstetr. Gynecol., 90: 483–488. Lindsay, C.A., Huston, L., Amini, S.B. & Catalano, P.M. 1997. Longitudinal changes in the relationship between body mass index and percent body fat in pregnancy. Obstetr. Gynecol., 89: 377–382. March of Dimes. 2002. Nutrition today matters tomorrow. A report from the March of Dimes Task Force on Nutrition and Optimal Human Development. Martorell, R., Delgado, H.L., Valverde, V. & Klein, R.E. 1981. Maternal stature, fertility and infant mortality. Hum. Biol., 53: 303–312. Merchant, K.M., Villar, J. & Kestler, E. 2001. Maternal height and newborn size relative to risk of intrapartum caesarean delivery and prenatal distress. Br. J. Obst. Gynecol., 108: 689–696. Muthayya, S. 1998. Maternal energy requirements and nutritional status in well nourished pregnant and lactating women. Bangalore University, India. (Ph.D. thesis) Panter-Brick, C. 1993. Seasonality of energy expenditure during pregnancy and lactation for rural Nepali women. Am. J. Clin. Nutr., 57: 620–628. Pickett, K.E., Abrams, B. & Selvin, S. 2000. Maternal height, pregnancy weight gain, and birthweight. Am. J. Hum. Biol., 12: 682–687. Piers, L., Diggavi, S., Thangam, S., van Raaij, J.M.A., Shetty, P.S. & Hautvast, J. 1995. Changes in energy expenditure, anthropometry and energy intake during the course of pregnancy and lactation in well-nourished Indian women. Am. J. Clin. Nutr., 61: 501–513. Pipe, N.G.J., Smith, T., Halliday, D., Edmonds, C.J., Williams, C. & Coltart, T.M. 1979. Changes in fat, fat-free mass and body water in normal human pregnancy. Br. J. Obstet. Gynaecol., 86: 929–940. Prentice, A.M., Goldberg, G.R., Davies, H.L., Murgatroyd, P.R. & Scott, W. 1989. Energy-sparing adaptations in human pregnancy assessed by whole-body calorimetry. Br. J. Nutr., 62: 5–22. Prentice, A.M., Spaaij, C.J.K., Goldberg, G.R., Poppitt, S.D., van Raaij, J.M.A., Totton, M., Swann, D. &
Black, A.E. 1996. Energy requirements of pregnant and lactating women. Eur. J. Clin. Nutr., 50 (suppl. 1): S82–S111. Shapiro, J., Sutija, V.G. & Bush, J. 2000. Effect of maternal weight on infant birth weight. Perinat. Med., 28: 428–431. Singh, J., Prentice, A.M., Diaz, E., Coward, W.A., Ashford, J., Sawyer, M. & Whitehead, R.G. 1989. Energy expenditure of Gambian women during peak agricultural activity measured by the doubly labeled water method. Br. J. Nutr., 62: 315–329. Sohlström, A. & Forsum, E. 1997. Changes in total body fat during the human reproductive cycle as assessed by magnetic resonance imaging, body water dilution, and skinfold thickness: a comparison of methods. Am. J. Clin. Nutr., 66: 1315–1322. Spaaij, C.J.K. 1993. The efficiency of energy metabolism during pregnancy and lactation in well-nourished Dutch women. Wageningen, Netherlands, University of Wageningen. Spaaij, C.J.K., van Raaij, J.M.A., van der Heijden, L.J.M., Schouten, F.J.M., Drijvers, J.J.M.M., de
Groot, L.C.P.G.M., Boekholt, H.A. & Hautvast, J.G.A.J. 1994. No substantial reduction of the thermic effect of a meal during pregnancy in well-nourished Dutch women. Br. J. Nutr., 71: 335–344. van Raaij, J.M.A., Vermaat-Miedema, S.H., Schonk, C.M., Peek, M.E.M. & Hautvast, J.G.A.J. 1987. Energy requirements of pregnancy in the Netherlands. Lancet, 2: 953–955. van Raaij, J.M.A., Peek, M.E.M., Vermaat-Miedema, S.H., Schonk, C.M. & Hautvast, J.G.A.J. 1988. New equations for estimating body fat mass in pregnancy from body density or total body water. Am. J. Clin. Nutr., 48: 24–29. WHO. 1985. Energy and protein requirements: Report of a joint FAO/WHO/UNU expert consultation. WHO Technical Report Series No. 724. Geneva. WHO. 1995a Maternal anthropometry and pregnancy outcomes. A WHO Collaborative Study. Wrld Hlth Org. Bull., 73 (suppl.): 1–98. WHO. 1995b. Physical status: The use and interpretation of anthropometry. Report of a WHO expert committee. WHO Technical Report Series No. 854. Geneva.
Energy requirements of lactation
63
7. ENERGY REQUIREMENTS OF
LACTATION
Exclusive breastfeeding is recommended during the six months after delivery, with introduction of complementary foods and continued breastfeeding thereafter (WHO, 2001). The energy requirement of a lactating woman is defined as the level of energy intake from food that will balance the energy expenditure needed to maintain a body weight and body composition, a level of physical activity and breastmilk production that are consistent with good health for the woman and her child, and that will allow economically necessary and socially desirable activities to be performed. To operationalize this definition, the energy needed to produce an appropriate volume of milk must be added to the woman’s habitual energy requirement, assuming that she resumes her usual level of physical activity soon after giving birth.
The mean amount of breastmilk produced daily is similar among population groups with different cultural and socio-economic settings (Prentice et al., 1986; Butte, Lopez-Alarcon and Garza, 2002) (Table 7.1). There may be some variation in milk composition related to maternal nutrition, but the main factors that influence the energy needs of lactating women are the duration of breastfeeding practices and the extent of exclusive breastfeeding. As these vary significantly in different societies, dietary energy recommendations for lactating women should be population-specific. Regardless of the cultural and social environment, the ideal situation is that women be well nourished from the beginning of pregnancy and that they maintain adequate nutritional intake with appropriate weight gain throughout gestation. This will allow them to attain body fat reserves that may act as an energy substrate to cover part of the additional energy needs in preparation for and during lactation.
TABLE 7.1
Average milk production rates (g/d) Postpartum period (months) 1 2 3 4 5 6 7 8 9 10 11 12
Exclusive breastfeeding
Industrialized countries 699 731 751 780 796 854
Traditional countries 562 634 582 768 778 804
Partial breastfeeding
Industrialized countries 611 697 730 704 710 612 569 417 497 691 516 497
Traditional countries 568 636 574 634 714 611 688 635 516 565 511
Source: Butte, Lopez-Alarcon and Garza, 2002.
7.1 DETERMINANTS OF THE ENERGY COST OF LACTATION
The energy cost of lactation is determined by the amount of milk that is produced and secreted, its energy content, and the efficiency with which dietary energy is converted to milk energy.
7.1.1 Human milk production
The mean amount of milk ingested by exclusively breastfed infants is similar in industrialized and more traditional societies, according to a WHO-sponsored comprehensive review (Butte, Lopez- Alarcon and Garza, 2002). After six months, variation among individuals and populations increases, owing to the nature and amount of complementary foods provided to the growing infant. From the age of six months onwards, when infants are partially breastfed, milk production is estimated at 550 g/day.
7.1.2 Energy content of human milk
The energy content of human milk depends primarily on milk fat concentration, which shows complex diurnal, within-feed and between-breast fluctuations. Twenty-four-hour milk sampling schemes have been developed that interfere minimally with the secretion of milk flow and capture the
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
64
diurnal and within-feed variation (Garza and Butte, 1986). Measurements of the gross energy content of representative 24-hour milk samples determined by adiabatic bomb calorimetry or macronutrient analysis in a number of studies of well-nourished women gave a mean value of 2.8 kJ/g (0.67 kcal/g) from 1 to 24 months of lactation (Garza and Butte, 1986; Prentice and Prentice, 1988; Butte and King, 2002; WHO, 1985; Institute of Medicine, 1991; Goldberg et al., 1991; Panter-Brick, 1993).
7.1.3 Efficiency of energy conversion
The efficiency with which food energy and body energy reserves are converted into milk energy has been calculated from theoretical estimates of the biochemical efficiency associated with the synthesis of milk lactose, protein and fat, and from metabolic balance studies (Prentice and Prentice, 1988). Taking into account the energy costs of digestion, absorption, conversion and transport, biochemical efficiency has been estimated at 80 to 85 percent (Butte and King, 2002). Based on that estimate, on the theoretical efficiency used in the 1985 FAO/WHO/UNU report and on the suggestion of the United States Institute of Medicine (1991), an efficiency factor of 80 percent was applied to calculate the energy cost of human milk production.
7.1.4 Energy cost of milk production
Table 7.2 shows the energy cost to produce the mean amounts of milk needed for exclusively breastfed infants. Monthly milk volumes are those reported for well-nourished women with healthy babies in the WHO-sponsored review (Butte, Lopez-Alarcon and Garza, 2002), and gross energy contents are those described in section 7.1.2.
The results were compared with the energy requirements of exclusively breastfed infants from one to six months of age, calculated as described in chapter 3. To do so, human milk intakes that are measured by the test-weighing technique must be corrected for insensible water loss during the course of feeding (5 percent correction factor) and for the digestibility of human milk. The metabolizable energy in human milk was assumed to be 5.3 percent lower than its gross energy content based on proximate analyses and energy factors of 23.6 kJ (5.65 kcal) per gram of protein and free amino acids, 38.7 kJ (9.25 kcal) per gram of fat and 16.5 kJ (3.95 kcal) per gram of lactose. From months one to six the figures are on average within 5 percent, which is remarkable considering that energy requirements of infants were calculated from quite different information (i.e. from predictive equations based on DLW measurements of TEE, plus estimates of growth accretion based on growth velocity and body composition).
TABLE 7.2
Energy cost of human milk production by women who practise exclusive breastfeeding
Months postpartum
Mean milk intake
g/day a
Human milk intake, corrected for
insensible water losses g/day
b
Gross energy content
kJ/g c
Daily gross energy secreted
kJ/day
Energy cost of milk production
kJ/day d
1 699 734 2.8 2 055 2 569
2 731 768 2.8 2 149 2 686
3 751 789 2.8 2 208 2 760
4 780 819 2.8 2 293 2 867
5 796 836 2.8 2 340 2 925
6 854 897 2.8 2 511 3 138
Mean 769 807 2.8 2 259 2 824 a From Butte, Lopez-Alarcon and Garza, 2002.
b Insensible water losses assumed to be equal to 5 percent milk intake.
c Gross energy content measured by adiabatic bomb calorimetry or macronutrient analysis.
d Based on energetic efficiency of 80 percent.
Energy requirements of lactation
65
TABLE 7.3
Comparison of the energy cost of human milk production and energy requirements of exclusively breastfed infants
Months postpartum
Mean milk
intake
g/day a
Human milk intake,
corrected for insensible
water losses g/day
b
Gross energy content
kJ/g c
Daily gross energy
secreted
kJ/day
Metabolizable energy intake
kJ/day d
Infant energy requirement
kJ/day e
Requirement/ME intake
1 699 734 2.8 2 055 1 946 1 922 0.99
2 731 768 2.8 2 149 2 035 2 143 1.05
3 751 789 2.8 2 208 2 091 2 284 1.09
4 780 819 2.8 2 293 2 172 2 219 1.02
5 796 836 2.8 2 340 2 216 2 376 1.07
6 854 897 2.8 2 511 2 378 2 501 1.05
Mean 769 807 2.8 2 259 2 140 2 241 1.05 a From Butte, Lopez-Alarcon and Garza, 2002.
b Insensible water losses assumed to be equal to 5 percent milk intake.
c Gross energy content measured by adiabatic bomb calorimetry or macronutrient analysis.
d Metabolizable energy values based on proximate analysis of milk are 5.3 percent lower than bomb calorimetry values.
e Mean values of boys and girls, calculated as described in chapter 3 of this report.
7.2 ENERGY REQUIREMENTS FOR LACTATION
Compared with non-pregnant, non-lactating women, during lactation there are no significant changes in BMR, efficiency of work performance, or TEE (Butte and King, 2002), and in most societies women resume their usual level of physical activity in the first month postpartum or shortly thereafter (Goldberg et al., 1991; Panter-Brick, 1993; Roberts et al., 1982; Tuazon et al., 1987; van Raaij et al., 1990). It could be argued that where exclusive breastfeeding is prevalent, lactating mothers may have a lower TEE than non-pregnant, non-lactating women owing to the frequency of breastfeeding, which involves periods of little maternal activity. On the other hand, lactating women often carry their infants while moving around, and this additional workload might balance the lower physical activity associated with breastfeeding. Thus, total energy requirements during lactation are equal to those of the pre-pregnancy period, plus the additional demands imposed by the need for adequate milk production and secretion.
These additional demands correspond to the energy cost of milk production. For women who feed their infants exclusively with breastmilk during the first six months of life, the mean energy cost over the six-month period is: 807 g milk/day × 2.8 kJ/g / 0.80 efficiency = 2.8 MJ/day (675 kcal/day) (Table 7.2). From the age of six months onwards, when infants are partially breastfed and milk production is on average 550 g/day (Table 7.1), the energy cost imposed by lactation is 1.925 MJ/day (460kcal/day).
Fat stores accumulated during pregnancy may cover part of the additional energy needs in the first few months of lactation. Postpartum loss of body weight is usually highest in the first three months, and generally greater among women who practise exclusive breastfeeding, but the extent to which the energy mobilized supports lactation depends on the gestational weight gain and the nutritional status of the mother. A review of 17 studies indicated that, on average, well-nourished women lost 0.8 kg/month, whereas undernourished mothers lost only an average of 0.1 kg/month (Butte and Hopkinson, 1998). Assuming an energy factor of 27.2 MJ/kg (Butte and King, 2002; Butte and Hopkinson, 1998), the rate of weight loss in well-nourished women would correspond to the mobilization of 27.2 × 0.8 kg/month = 21.8 MJ/month, or 0.72 MJ/day (170 kcal/day) from body energy stores. This amount of energy can be deducted from the 2.8 MJ/day (675 kcal)/day needed during the first six months of lactation. The result, 2.1 MJ/day (505 kcal/day), is similar to the additional energy required when infants are partially breastfed after six months of lactation.
On the other hand, undernourished women and those who did not gain adequate body weight during pregnancy must conserve as much energy as possible for their own well-being and that of their infants. Hence, in these women the full energy demands of lactation must be provided by an increment in dietary intake.
In conclusion, well-nourished women with adequate gestational weight gain should increase their food intake by 2.1 MJ/day (505 kcal/day) for the first six months of lactation, while undernourished
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
66
women and those with insufficient gestational weight gain should add to their personal energy demands 2.8 MJ/day (675 kcal/day) during the first semester of lactation. Energy requirements for milk production in the second six months are dependent on rates of milk production, which are highly variable among women and populations.
REFERENCES Brown, K., Dewey, K.G. & Allen, L. 1998. Complementary feeding of young children in developing countries: A review of current scientific knowledge. Geneva, WHO. Butte, N.F. & Hopkinson, J.M. 1998. Body composition changes during lactation are highly variable among women. J. Nutr., 128: 381S–385S. Butte, N. & King, J.C. 2002. Energy requirements during pregnancy and lactation. Energy background paper prepared for the joint FAO/WHO/UNU Consultation on Energy in Human Nutrition. Butte, N.F., Lopez-Alarcon, M.D. & Garza, C. 2002. Nutrient adequacy of exclusive breastfeeding for the term infant during the first six months of life. Geneva, WHO. Garza, C. & Butte, N.F. 1986. Energy concentration of human milk estimated from 24-h pools and various abbreviated sampling schemes. J. Pediatr. Gastroenterol. Nutr., 5: 943–948. Goldberg, G.R., Prentice, A.M., Coward, W.A., Davies, H.L., Murgatroyd, P.R., Sawyer, M.B., Ashford,
J. & Black, A.E. 1991. Longitudinal assessment of the components of energy balance in well-nourished lactating women. Am. J. Clin. Nutr., 54: 788–798. Institute of Medicine. 1991. Nutrition during lactation. Washington, DC, National Academy Press. Panter-Brick, C. 1993. Seasonality of energy expenditure during pregnancy and lactation for rural Nepali women. Am. J. Clin. Nutr., 57: 620–628. Prentice, A.M. & Prentice, A. 1988. Energy costs of lactation. Ann. Rev. Nutr., 8: 63–79. Prentice, A., Paul, A., Black, A., Cole, T. & Whitehead, R. 1986. Cross-cultural differences in lactational performance. In M. Hamosh and A.S. Goldman, eds. Human lactation 2: Maternal and environmental factors, pp. 13–44. New York, Plenum Press. Roberts, S.B., Paul, A.A., Cole, T.J. & Whitehead, R.G. 1982. Seasonal changes in activity, birth weight and lactational performance in rural Gambian women. Trans. R. Soc. Trop. Med. Hyg., 76: 668–678. Tuazon, M.A., van Raaij, J.M., Hautvast, J.G. & Barba, C.V. 1987. Energy requirements of pregnancy in the Philippines. Lancet, 2: 1129–1131. van Raaij, J.M.A., Schonk, C.M., Vermaat-Miedema, S.H., Peek, M.E.M. & Hautvast, J.G.A.J. 1990. Energy cost of walking at a fixed pace and self-selected pace before, during and after pregnancy. Am. J. Clin. Nutr., 51: 158–161. WHO. 1985. Energy and protein requirements: Report of a joint FAO/WHO/UNU expert consultation. WHO Technical Report Series No. 724. Geneva. WHO. 2001. Expert consultation on the optimal duration of exclusive breastfeeding. Conclusions and recommendations. Geneva.
Recommendations for future research
67
8. RECOMMENDATIONS FOR FUTURE
RESEARCH
Expert consultation meetings that are convened to make recommendations on nutrient requirements are sometimes faced with situations in which adequate information is lacking and questions need to be answered before evidence-based recommendations that are applicable to population groups worldwide can be provided. Recognition of the lacunae in the existing knowledge base helps the identification of potential areas for future research and investigation by the wider scientific and academic communities. The deliberations and recommendations of experts in this important sphere carry much weight within the academic community, as well as with research funding bodies, international agencies and bilateral donors.
The following recommendations for future research are based on the topics and issues that were identified during the more focused discussions at the preliminary working group sessions, and that were fed into the expert consultation, as well as on those identified by the experts themselves during the consultation. However, as the present expert consultation acknowledged, it is not enough to come up with a wish list of research topics without prioritizing what needs to be done. With resources becoming increasingly limited, the experts recognized that it would be futile either to outline research needs too broadly or to attempt to include every conceivable topic that may be relevant to the issues raised during their deliberations. The expert consultation recognized the need to make judgements on priorities when they stated: “We need to prioritize our recommendations so as not to dilute the strength of our requests.”
The questions and topics that the 2001 expert consultation identified as being in need of further investigation are categorized into two broad groups. The first group consists of those biological questions whose answers will provide better numerical estimates of human requirements. These include conceptual, methodological and data-gathering components. The second group includes epidemiological and community studies aimed at testing the validity of the estimates in populations living in different environmental and social conditions, for which more general, and indeed more realistic, criteria of health and function are required than those that are used in metabolic or clinical investigations. This second group also includes questions that relate to the use of the recommended nutrient requirement estimates and their implications for planners and policy-makers at the national, regional and global levels.
8.1 BIOLOGICAL QUESTIONS: CONCEPTUAL AND METHODOLOGICAL
8.1.1 Basal metabolic rate (BMR) and total energy expenditure (TEE)
1. BMR predictive equations are to be revisited, reviewed and reformulated, if necessary, based on access to a larger, more comprehensive global database that should be expanded and collated with strict and transparent quality and inclusion criteria.
2. There are insufficient data to judge whether either ethnicity or habitation in a tropical environment influences BMR. It is possible that aspects that are attributed to ethnicity may well be responses to early life exposure to suboptimal nutritional environments. It is therefore recommended that when ethnicity is researched and reported, additional information on history of nutritional status and/or environmental exposure in early life also be measured and reported. This phenomenon needs to be better understood, the physiological basis needs to be established and the plausible mechanisms involved need to be clarified.
3. Prospective studies to measure daily TEE by DLW and/or other methods (such as the flex- heart rate method) need to be undertaken in order to provide comparisons for the same subjects with estimates based on the factorial method. Measured BMR and the energy cost of sitting, standing, etc. may be used with the PAL values presented here for calculating daily energy expenditure in an effort to enhance the application of these PAL values. There is accumulating evidence from various laboratories of discrepancies between estimates of TEE by the factorial method using published PAL values compared with estimates using other
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
68
methods, such as the flex-heart rate method. This discrepancy becomes more apparent when the intensity of an activity is increased. Studies comparing the two methods using different sources of PAL values suggest that it may be possible to use these data more appropriately and to reconcile the data generated. Given the shortage and inordinate expense of stable isotopes, it is necessary to invest more in accepted methodologies in order to broaden the database. There is also a need for more data on PAL values.
4. There is an urgent need for more TEE and measured BMR studies, coupled with time–motion studies from developing countries that cover prevailing and changing life styles. The use of DLW studies will be essential for the purpose of validating existing methodologies and developing new ones. Support from the International Atomic Energy Agency (IAEA) in making available more isotopes at less cost, in improving the ability to analyse these and in general capacity building to cater to developing country needs is crucial in this area.
5. There is shortage of information on BMR and TEE from elderly groups because this subpopulation is increasing owing to changes in longevity and demographics in developing and developed societies.
6. Further development and validation of techniques for measuring TEE and BMR, as well as energy cost of activity expenditures and patterns, need to be supported. New techniques should be accurate, precise, portable, cheap and appropriate for field-based studies worldwide. Ideally, all new techniques need to be validated against both indirect calorimetry and DLW methods.
7. There is a need to update and expand the data bank on the energy cost of a range of activities undertaken in real-life conditions by children and adults, distinguishing weight-bearing from non-weight-bearing activities, and specifying whether energy cost refers to “net” activity or is integrated over tasks.
8. The number of available DLW studies on infants (and young children) from developing countries is limited and needs to be expanded in normal birth weight infants.
9. Studies with DLW (or other methods) need to be carried out in order to determine TEE of school-going children and adolescents in urban and rural areas of developing countries.
10. The DLW method provides a means of determining the amount of energy expended in physical activity. PALs consistent with normal health and the development of infants and children should be described qualitatively and ethnographically across cultures.
11. Further studies are needed to confirm whether the increased TEE observed in some settings is caused solely by differences in size and body composition or whether other mitigating factors are involved.
12. More information is needed about the influence of habitual physical activity on the growth and development of all children and adolescents, and on the duration, intensity and frequency of the physical activity that is necessary to achieve optimal effects.
8.1.2 Nutritional anthropometry and body composition
1. The use of United States-based reference data for assessing adolescent growth worldwide is a matter of concern, and it is recommended that research be conducted in order to evaluate their universal applicability, specifically the upper percentile elevations and skewness of the NCHS value, especially as they apply to developing countries.
2. More data are needed on variations in body composition of individuals in different population groups. There is a further need to develop methodologies for body size normalization when estimating the energy cost of different activities.
8.1.3 Studies in undernourished subpopulations
1. The effect of the quality of dietary protein, carbohydrate and fat on rates of weight gain, particularly during the recovery period from malnutrition, needs to be understood better. Biological (and behavioural) studies are needed to help establish appropriate levels of energy intake during convalescence from such episodes.
2. Nutrient needs for the rehabilitation of stunted children are also poorly understood. Information is needed on the energy intake and expenditure requirements for catch-up growth
Recommendations for future research
69
in body mass and stature of stunted and undernourished children. Special nutrient requirements for catch-up growth of bones require further research. Physiological adjustments in physical activity and growth in response to undernutrition should be investigated with newer methodologies such as DLW.
3. There is a need for estimates of BMR and TEE using DLW methodology in undernourished children and adults. This should include an investigation of intra-uterine growth retarded (IUGR) infants, and stunted and undernourished groups of children, compared with children with adequate growth.
8.1.4 Food energy
1. Factors affecting the dietary intake that is necessary to satisfy energy requirements should be explored, including diet digestibility, viscosity, and energy and nutrient density.
2. The validity of metabolizable energy (ME) food energy conversion factors as quantitative equivalents of biologically useful, energy and their relationship to net metabolizable energy (NME) need to be reviewed. It may be necessary to investigate how best NME and energy requirement estimations can be integrated and reconciled.
3. The AOAC (Prosky) method of dietary fibre analysis is now widely used in food analysis. Further research is needed to develop reliable analytical methods for resistant starches.
8.2 EPIDEMIOLOGICAL AND COMMUNITY STUDIES
1. Large numbers of children in developing countries have experienced repeated episodes of infections, which are often accompanied by a negative energy balance owing to decreased appetite and/or increased metabolic activity. Studies on the effects of infection on energy requirements of infants are limited, and should be expanded to cover a broad range of infectious agents of varying severity and duration.
2. More qualitative and quantitative information is needed on the habitual physical activity of children and adolescents in developing societies. This includes physiological, anthropological and behavioural studies. Anthropologists and other social scientists must be invited to participate in this endeavour, as information already exists in reports and monographs in the social sciences literature, and this should be analysed.
3. Currently, there are major gaps in the knowledge regarding estimating the survival level of energy expenditure, and consequently the lower limits of emergency rations and food aid baskets, particularly in refugee settings. This is in need of urgent evaluation. The support of FAO/WHO is essential if the academic community is to obtain funds for such investigations from research organizations.
4. There is a general consensus that the most crucial aspects in understanding the energy requirements in pregnancy and lactation are now known. However, more research with respect to public health-related issues (e.g. low birth weight) should be carried out. There is a need for longitudinal studies on pregnant woman, in order to relate the associated physiological parameters with birth outcomes and risks. More research is needed on the range of issues that affect obese and underweight women during pregnancy and lactation.
5. There is a need to establish the nature, duration, frequency and intensity of physical exercise required to maintain generally good health and to prevent specific pathologies, such as obesity and its related co-morbidities.
6. There is a need to understand better the health risks of people with BMI less than 18.5. 7. Overweight and obesity are closely linked to a positive (i.e. surplus) energy balance.
Biological and behavioural investigations are needed to develop and test methods that will guide children and adolescents towards an energy balance that reduces the risk of becoming overweight.
8. Techniques must be developed to stimulate children’s and adolescents’ interest in performing an appropriate level of physical activity in the context of different geographic, cultural and socio-economic environments.
9. Reliable documentation on life-styles and time use needs to be collected in order to improve the existing energy expenditure estimates using adults, children and the elderly in diverse
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
70
contexts, with special efforts to include information from developing country and transitional society contexts.
Conclusions
71
9. CONCLUSIONS
The primary aim of the expert consultations on nutrient requirements has remained the same: to provide advice on scientific issues related to nutrient requirements and to formulate appropriate recommendations for action. An examination of the historical precedents in this endeavour reveals how the various expert groups have contributed to the principles for determining human energy requirements and their practical applications, which have been adopted worldwide. The recommendations from the resulting reports have not only reflected the state of knowledge at particular points in time, but have also been embraced by the global scientific community, thereby influencing research agendas and methodologies over the years.
Many of the points made by the first committee on calorie requirements, which met in 1949, are still pertinent today. The requirements set by the experts were intended for groups of people rather than individuals, and the committee established the principle, which is often misunderstood, that “an average requirement can never be compared directly with an individual (requirement)” (FAO, 1950). The first committee noted that its recommendations should be adjusted depending on how and for whom they are used, and it cautioned that nutrition and health experts within countries should take into account local conditions in applying the requirements. There is always the need to exercise judgement in interpreting and using requirement values, yet advice in this area is most difficult to impart to users. The first committee on calories offered the very practical rule of thumb that if the person “is in good health and calorie balance, that is, neither over- nor underweight, then he or she is consuming food according to his or her calorie requirements” (FAO, 1950). Subsequent committees also recognized the importance of maintaining an adequate level of energy expenditure, thus acknowledging that non-occupational activities were just as important as occupational ones to the overall health status of many people and that energy requirements did not refer to a minimum level (FAO, 1973).
The 1973 Report of the Joint FAO/WHO Ad Hoc Expert Committee on Energy and Protein Requirements reiterated statements that had been made in past reports that the recommendations for nutrient requirements should be applied to groups and not to individuals. However, the 1973 report also made two additional important points: 1) that estimates of requirements are derived from individuals rather than groups; and 2) that the nutrient requirements of comparable individuals often vary.
The report of the Joint FAO/WHO/UNU Expert Consultation on Energy and Protein Requirements held in 1981 (WHO, 1985) was very clear in its statement that estimates of energy requirements should, as far as possible, be based on estimates of energy expenditure, as the prevailing method of determination – from observed intakes of food energy – was becoming unreliable and served to support a circular argument that access to food determined energy needs. The rationale for this conclusion was that in both developing and developed countries actual energy intakes are not necessarily those that either maintain a desirable body weight or provide for optimal levels of physical activity, and hence health in its broadest sense. The experts at the 1981 consultation were aware of the limited data on energy expenditures, particularly among children. They were also conscious of the fact that no reliable and widely useable method was available to the scientific and academic community for collecting such data from a range of population groups worldwide.
The 1981 expert consultation felt that, except for children, sufficient information was available to approach this issue using data on BMR at the centre of a new conceptual framework to estimate total energy expenditure. Thus the use of BMR became important in determining energy requirements. The experts identified a new methodology for calculating energy requirements, and substantial research that needed to be carried out after the expert consultation. One significant departure of the 1981 expert group from that of the 1971 experts was the rejection of the concept of a single reference man or woman. The 1971 group defined such people as “arbitrarily selected convenient starting points for extrapolation ... and ... not intended to suggest ideal standards. They were originally chosen as being representative of groups of men and women whose food consumption and energy expenditure had been carefully studied” (FAO/WHO, 1973). The 1981 group found this concept too restrictive and not reflective of the wide ranges of both body size and patterns of physical activity.
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
72
In keeping with each of its predecessors, the present report has attempted to build on these efforts, while also moving forward in breaking new ground. The 2001 Joint FAO/WHO/UNU Expert Consultation on Human Energy Requirements met after a lapse of nearly 20 years and deliberated only the requirements of energy in the diet, leaving deliberations and debates on protein and amino acid requirements to a separate expert group, which met at WHO in Geneva in 2002. Following up on the recommendations arising from the Energy Consultation, FAO also convened a group of experts to discuss the issue of “food energy”, because recommendations for optimal energy requirements become practical only when they are related to foods that provide the energy to meet those requirements. Gains in understanding of the digestion and metabolism of food and the increasing sophistication of analytical techniques meant that the various options available to express the energy value of foods needed to be standardized and harmonized. The recommendations of this group have since been published (FAO, 2002) to complement this report.
When the Expert Consultation on Human Energy Requirements met in 2001, the situation regarding lack of data to arrive at realistic and evidence-based recommendations had changed dramatically. Major technological advances using stable (i.e. non-radioactive) isotopes had by then had a dramatic impact on the measurement of energy expenditures of free-living individuals in real- life situations. Estimates based on these measurements have, to a large extent, replaced estimates using both direct and indirect calorimetry and the associated dependent methodologies such as heart rate monitoring, activity monitoring, pedometers and actometers. It is important to reiterate that these conventional methods continue to be important because the stable isotope technique measures cumulative total energy expenditure over a period but provides no accurate estimate of day-to-day variations, or information relating to the nature and pattern of daily activities. Thus, in this report, almost all of the recommendations made are based on reliable measurements of TEE obtained from infants, children, adolescents, adults and the elderly, as well as from women in special physiological states such as pregnancy and lactation.
A summary of the new concepts and changes in this 2004 expert report include the following:
The calculation of energy requirements for all ages should be based on measurements and estimates of total daily energy expenditure, including the energy needs for growth.
New values for energy requirements of infants, children and adolescents were proposed because existing values had been overestimated for children under ten years of age, and underestimated for children over 11 years of age and for adolescents.
Different requirement levels were proposed for populations with various lifestyles and levels of habitual physical activity, starting at six years of age.
A comparison and testing of the different BMR databases with varying degrees of ethnic and geographical coverage was carried out to determine whether new equations for estimating BMR from mean age and body weight of population groups were needed (Annex 3).
New factorial estimates of the additional energy needs imposed by pregnancy and lactation were applied.
Recommendations for the levels of physical activity required to maintain fitness and health and reduce the risk of developing obesity and diseases associated with sedentary lifestyles were made, and PALs based on the degree of habitual activity recommended for long-term good health were classified.
From the start, most expert groups have sought to address the practical application of the requirements. Time and events have shown that this aspect is as complex as determining the requirements themselves. As in the 1985 report, the section on issues regarding the application of requirements has been omitted from this report. However, in keeping with the recommendation made by the experts, the FAO Secretariat has spent time and effort in developing both a user’s manual and a software application (on CD-ROM, see Annex 4), which are released alongside this report so that they might complement each other. Both the scientific content and the recommendations that ensue from this evidence base, as well as the usefulness and appropriateness of the accompanying applications manual and software, will await the judgement of the community of users, who are the best arbiters of the importance of this ongoing exercise by international agencies.
Conclusions
73
REFERENCES FAO. 1950. Calorie requirements: report of the Committee on Calorie Requirements. FAO Nutritional Studies No. 5. Washington, DC. FAO. 2002. Food energy – methods of analysis and conversion factors. FAO Food and Nutrition Paper No. 77. Rome. FAO/WHO. 1973. Energy and protein requirements: Report of a joint FAO/WHO ad hoc expert committee. FAO Nutrition Meetings Report Series No. 52. WHO Technical Report Series No. 522. Rome and Geneva. WHO. 1985. Energy and protein requirements: Report of a joint FAO/WHO/UNU expert consultation. WHO Technical Report Series No. 724. Geneva.
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
74
75
ANNEXES
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
76
Annexes
77
ANNEX 1: PARTICIPANTS – JOINT FAO/WHO/UNU EXPERT CONSULTATION ON
HUMAN ENERGY REQUIREMENTS, 17 TO 24 OCTOBER 2001, FAO HEADQUARTERS,
ROME, ITALY
EXPERTS
Eric Ategbo
Universite Nationale du Benin Cotonou, Benin
Stephanie Atkinson (Unable to attend) Mc Master University Hamilton, Canada
Nahla Houalla Baba
Universite Americaine de Beyrouth Riad el Solh, Lebanon
Anna Ferro-Luzzi
National Institute of Research for Food and Nutrition Rome, Italy
Zhi Qian He (Unable to attend) Sun Yat-sen University of Medical Sciences Guangzhou, China
Eric Jequier
Instutut de Physiologie Lausanne, Switzerland
Joyce Kanyangwa-Luma
World Food Programme Country Office Islamabad, Pakistan
Eileen Kennedy
167 Yarnick Road Great Falls, United States
Sook He Kim (Unable to attend) Ewha Woman’s University Seoul, Republic of Korea
Janet King
University of California – Davis Davis, United States
Surat Komindr
Ramathibodi Hospital Bangkok, Thailand
Ismail Noor
University of Kebangsaan Kuala Lumpur, Malaysia
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
78
Narsinga Rao
National Institute of Nutrition Hyderabad, India
Patric Ritz
Universite d’Angers Angers, France
Irwin Rosenberg
Tufts University Boston, United States
Benjamin Torun
Instituto de Nutrición de Centro America y Panama (INCAP) Guatemala City, Guatemala
Ricardo Uauy
Instituto de Nutrición y Tecnologia de los Alimentos (INTA) Santiago, Chile
Este Vorster
Potchefstroomse Universiteit Potchefstroom, South Africa
Klaas Westerterp
University of Maastricht Maastricht, Netherlands
Roger Whitehead
Dunn Nutrition Centre Cambridge, United Kingdom
TECHNICAL ADVISORS TO SECRETARIAT
Nancy Butte
Baylor College of Medicine Houston, United States
Joop Van Raaij
Agricultural University of Waginengen Waginengen, Netherlands
OBSERVERS
Venkatesh Iyengar
International Atomic Energy Agency Vienna, Austria
Hiroshi Kashiwazaki
National Institute of Health and Nutrition Tokyo, Japan
Annexes
79
Sonya Rabeneck
Standing Committee on Nutrition Geneva, Switzerland
Beat Schurch (Unable to attend) Nestle Foundation Lausanne, Switzerland
SECRETARIAT
Barbara Burlingame
FAO-ESNA Rome, Italy
Graeme Clugston (Unable to attend) World Health Organization Geneva, Switzerland
Mercedes de Onis
World Health Organization Geneva, Switzerland
Gina Kennedy
FAO-ESNA, Consultant Rome, Italy
Prakash Shetty
FAO-ESNA Rome, Italy
Kraisid Tontisirin
FAO-ESND Rome, Italy
Robert Weisell
FAO-ESNA Rome, Italy
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
80
MEMBERS OF WORKING GROUPS, 27 JUNE TO 5 JULY 2001, FAO HEADQUARTERS,
ROME, ITALY
1. WORKING GROUP ON ENERGY (AND PROTEIN) REQUIREMENTS OF INFANTS
AND PRESCHOOL CHILDREN
Chairperson
Nancy Butte
USDA/ARS Children’s Nutrition Research Center Houston, United States
Members
Kathryn Dewey
University of California – Davis Davis, United States
Erik Diaz
Instituto de Nutrición y Tecnología de los Alimentos (INTA) Santiago, Chile
Mercedes de Onis
World Health Organization Geneva, Switzerland
Cutberto Garza
United Nations University Ithaca, United States
Michael Goran
University of Southern California Los Angeles, United States
Paul Pencharz
The Hospital for Sick Children Toronto, Canada
Benjamin Torun
Instituto de Nutrición de Centro America y Panama (INCAP) Guatemala City, Guatemala
2. WORKING GROUP ON ENERGY AND PROTEIN REQUIREMENTS OF PREGNANCY
AND LACTATION
Chairperson
Joop Van Raaij
Agricultural University of Waginengen Waginengen, Netherlands
Annexes
81
Members
Lindsay Allen
University of California – Davis Davis United States
Corazon Barba
Food and Nutrition Research Institute (FNRI) Metro Manila, Phillipines
Elisabet Forsum
University of Linkoping Linkoping, Sweden
Andrew Prentice
Dunn Clinical Nutrition Centre Cambridge, United Kingdom
3. WORKING GROUP ON METHODOLOGY FOR ENERGY BALANCE AND ENERGY
REQUIREMENTS
Chairperson
Anna Ferro-Luzzi
Instituto Nazionale della Nutrizione Rome, Italy
Members
Steve Heymsfield
St. Lukes Roosevelt Hospital New York, United States
Prakash Shetty
FAO-ESNA Rome, Italy
Robert Weisell
FAO-ESNA Rome, Italy
Klaas Westerterp
Maastricht University Maastricht, Netherlands
4. WORKING GROUP ON PROTEIN AND AMINO ACID REQUIREMENTS
Chairperson
Denis Bier
Baylor College of Medicine Houston, United States
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
82
Members
Peter Furst
Universitat Hohenheim Stuttgart, Germany
Peter Garlick
Health Sciences Center Stony Brook, United States
Alan Jackson
University of Southampton Southampton, United Kingdom
Anura Kurpad
St. John’s Medical College Bangalore, India
Joe Millward
University of Surrey Guildford, United Kingdom
Paul Pencharz
The Hospital for Sick Children Toronto, Canada
Niels Raiha
Lund University Malmo, Sweden
William Rand
Tufts University Boston, United States
Peter Reeds (Unable to attend) University of Illinois Urbana, United States
Daniel Tomé (Observer) Institut National Agronomique Paris-Grignon Paris, France
Benjamin Torun
Instituto de Nutrición de Centro America y Panama (INCAP) Guatemala City, Guatemala
Ricardo Uauy
Instituto de Nutrición Technologia de Alimentos (INTA) Santiago, Chile
John Waterlow
15 Hillgate Street London, United Kingdom
Annexes
83
Vernon Young
Massachusetts Institute of Technology Cambridge, United States
5. WORKING GROUP ON ANALYTICAL ISSUES IN FOOD ENERGY AND
COMPOSITION. ENERGY IN FOOD LABELLING, INCLUDING REGULATORY AND
TRADE ISSUES
Chairperson
Ghulam Sarwar Gilani
Health Canada Ottowa, Canada
Members
Barbara Burlingame
FAO-ESNA Rome, Italy
Malcolm Fuller
107 Quaker Path Stony Brook, United States
Peter Jones
McGill University Montreal, Canada
Geoff Livesey
Independent Nutrition Logic Wymondham, United Kingdom
Paul Moughan
Massey University Palmerston North, New Zealand
Peter Pellett
University of Massachusetts Amherst, United States
Daniel Tomé (Observer) Institut National Agronomique Paris-Grignon Paris, France
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
84
ANNEX 2: AUTHORS AND REVIEWERS OF PAPERS FOR EXPERT CONSULTATION
WORKING GROUPS, MEETINGS AND FOLLOW-UP
Energy Background Paper No. 1
Andrew Prentice
Macronutrients as sources of food energy
Reviewer: Jean Pierre Flatt
Energy Background Paper No. 2
Geoff Livesey
Analytical issues related to food energy and food composition, energy in food labeling – including
regulatory and trade issues
Reviewers: Janis Baines, Janine Lewis, Penelope Warwick
Energy Background Paper No. 3
Anna Ferro-Luzzi
The conceptual framework for estimating food energy requirements
Reviewers: Paul Haggarty, Andreu Palou, Robert Weisell
Energy Background Paper No. 4
Nancy Butte
Energy requirements of infants
Reviewers: Peter Sauer, Jonathan Wells
Energy Background Paper No. 5
Benjamin Torun
Energy requirements of children and adolescents (with addendum) Reviewers: Margaret Livingstone, Virginia Stallings
Energy Background Paper No. 6
Prakash Shetty
Energy requirements of adults
Reviewers: Michael Goran, Dale Schoeller, Yves Schutz
Energy Background Paper No. 7
7a: Joop van Raaij (Initial draft) Energy requirements during pregnancy and lactation
Reviewers: Nancy Butte, Janet King 7b: Nancy Butte and Janet King (Final version) Reviewers: Kathryn Dewey, Elisabet Forsum
Energy Background Paper No. 8
Susan Roberts and Gerry Dallal
Energy requirements and ageing
Reviewers: Elisabet Rothenberg, Jane Wuu
Energy Background Paper No. 9
Marinos Elia
Insights into energy requirements in disease
Reviewers: Bruce Bistrian, Eileen Gibney
Annexes
85
Energy Background Paper No. 10
Anura Kurpad, Sumithra Muthayya, Mario Vaz
Consequences of inadequate food energy and negative energy balance in adults
Reviewer: Nick Norgan
Energy Background Paper No. 11
Ricardo Uauy and Erik Diaz
Consequences of food energy excess and positive energy balance
Reviewers: William Dietz, James Hill
Energy Background Paper No. 12
Mario Vaz, Nadine Karaolis, Alizon Draper, Prakash Shetty
A compilation of energy costs of physical activities
Reviewers: Gerald Spurr, K. Satyanarayana, Angela Polito
Energy Background Paper No. 13
Jeya Henry
Basal metabolic rate studies in humans: measurement and application
Reviewers: Anna Ferro-Luzzi, Kevin Acheson, Philip James, George Bray
Energy Background Paper No. 14
Nick Norgan
Laboratory and field measures of body composition
Reviewers: Paul Deurenberg, Tim Lohman, Cameron Chumelea
Energy Background Paper No. 15
Ingrid Coles-Rutishauser
Laboratory and field measures of dietary intake
Reviewers: Elisabet Wirfaelt, Wija van Staveren
Energy Background Paper No. 16
James Levine
Measurement of energy expenditure
Reviewers: Anna Ferro-Luzzi, Klaas Westerterp, Eric Ravussin, Catherine Geissler
Opinion Paper No. 1
Michael Goran
Estimating energy requirements: regression based prediction equations or multiples of resting
metabolic rate
Opinion Paper No. 2
Penelope Warwick and Janis Baines
Point of view: Energy factors for food labelling and other purposes should be derived in a consistent
fashion for all food components
Opinion Paper No. 3
Cutberto Garza
Effect of infection on energy requirements of infants and children
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
86
POST-CONSULTATION SOLICITED DOCUMENTS
Document No. 1
Sally Grantham-McGregor and Helen Heningham-Baker
Review of the evidence linking protein and energy to mental development
Document No. 2
Tim Cole and Jeya Henry
The Oxford Brookes BMR database—a reanalysis
Document No. 3
Manuel Ramirez-Zea
Validation of three predictive equations for basal metabolic rate
Document No. 4
Anna Ferro-Luzzi
A review of Tim Cole and Manuel Ramirez-Zea’s reports on BMR predictive equations
Annexes
87
ANNEX 3: UPDATE ON PREDICTIVE EQUATIONS TO ESTIMATE BASAL METABOLIC
RATE
The Joint FAO/WHO/UNU Expert Consultation on Energy and Protein Requirements, which met at FAO in Rome in 1981, concluded that – wherever possible – estimates of energy requirements should be based on measurements of energy expenditures rather than on energy intakes. It also decided that there would be many advantages in expressing the various components of total energy expenditure (TEE) as multiples of the basal metabolic rate (BMR). BMR is the most dominant component of TEE, and this is the primary reason for expressing the energy requirement (primarily BMR plus energy requirements for physical activity) as a multiple of the BMR. As a result, measurements of BMR and the methods to predict BMR have gained increased significance in estimating human energy requirements.
Dr J. Durnin of Glasgow University, the United Kingdom, made an extensive examination of the scientific literature and produced a background document for the 1981 expert consultation, which laid the foundations for the use of the BMR factorial approach to estimate TEE and energy requirements. It was noted that, while attempts had been made to carry out post hoc analysis of the existing data on BMRs in the past at the express request of FAO (Quenouille et al., 1951), subsequent FAO and FAO/WHO committees had not followed up on this approach of using BMR as the starting point to assess human energy requirements. On the recommendation of the 1981 expert consultation, it was decided to undertake a more comprehensive analysis of the available data on BMRs worldwide in order to generate predictive equations that could be used in the report. In a relatively short period following the 1981 expert consultation, FAO initiated a thorough research of the available literature for robust BMR data in order to construct a series of regression equations by sex and age groups. These equations and the related scientific papers appeared as a supplement of Human Nutrition: Clinical Nutrition (Volume 39C, Supplement 1, 1985). The new database contained 7 173 data points drawn from 114 published studies. Shortcomings in the data sets were duly noted, predominantly the over- or underprediction of BMR, and these were viewed primarily as the result of a lack of ethnic and geographical representation in the data.
The BMR predictive equations were used for the first time in the 1985 Joint FAO/WHO/UNU Expert Report on Energy and Protein Requirements and have gained considerable popularity since then. They were also used by several national expert groups that deliberated on energy requirements. Since then, however, questions have frequently been raised in the literature about the adequacy and accuracy of these predictive equations for universal use. In the 1990s, based on the recommendations made at a workshop organized by the International Dietary Energy Consultative Group (IDECG) in London (Scrimshaw, Waterlow and Schurch, 1996) and supported by FAO and the Nestle Foundation, Dr C.J.K. Henry (in collaboration with Dr Durnin) was commissioned to conduct a review of the literature of BMR data, in order to expand and refine the earlier database and to derive new equations based on selective criteria using more geographically representative data. This work resulted in the creation of a new database that has been referred to as the “Oxford database”. It also resulted in an increased number of data points and included additional data from several developing countries. The findings of this review were presented at an IDECG meeting in December 1997 in Rome. The results were reviewed and found to be inconclusive in furthering the need to produce new, representative and internationally useable BMR predictive equations for future use.
In preparation for the 2001 Joint FAO/WHO/UNU Expert Consultation on Human Energy Requirements, Dr Henry prepared a background paper that constituted the final analysis of this complete data set, taking into account the feedback provided on his original findings. Concurrently, a subcommittee was formed to guide the expert consultation regarding the appropriateness of the
Created in 1986, IDECG studies the effects of varying levels of dietary energy intake on the health and welfare
of individuals and societies. Its objectives include the compilation and interpretation of relevant research data on functional and other consequences of deficiency, change or excess of dietary energy; the identification and promotion of related research needs and priorities; the publication of scientific and policy statements and other information on the significance of chronic deficiencies and excesses of dietary energy; and the identification and promotion of appropriate and practical means of corrective action.
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
88
methods used to measure BMR and its related issues. The final conclusion of the subcommittee was to conduct a more in-depth analysis of the Oxford database.
In December 2001, following the 2001 Joint Expert Consultation on Human Energy Requirements, FAO assigned Dr T. Cole the task of reanalysing both the earlier Schofield database and the more recent Oxford data sets on BMR. This analysis was expected to provide information on the influence of ethnicity on BMR, to reveal any possible methodological biases, and to help develop new BMR predictive equations that would replace the currently used international equations in the 1985 report, if their predictive performance was better. Dr Cole carried out an elegant and sophisticated analysis in two stages using the Oxford database consisting of 13 910 BMR data points. The thrust of the first analysis was to develop a single unified and seamless predictive equation that would apply to all ages, i.e. through the life cycle from infancy to old age. The expectation was that a seamless, continuous equation could eliminate the split of the predictive curves at the joining points of the various age groups. This analysis showed that cleaning the data sets by exclusion increased their inefficiency, and hence the inclusion of all data irrespective of the methodology used, the date of publication or the geographical region was the favoured approach. Two major factors that seemed to affect BMR were age and weight, with height having far less effect than weight. The inclusion of both weight and height in the model ensured that variations in body composition were adjusted for. Following a presentation of the first analysis of the data sets, Dr Cole was persuaded to carry out a supplementary analysis to develop predictive equations from subsets of the original database. These were meant to consider adults only, within normal ranges of body mass index (18.5 to 25.0) and were further separated to observe the effect of the time period when the data were collected (i.e. pre- and post- 1950), as well as the effect of excluding any data based on close circuit calorimetry. This analysis showed that all four models examined among adults produced very similar results, and comparison with the earlier predictive equations indicated the absence of any significant improvement from the equations generated for the 1985 report.
Following the reanalysis, Dr M. Ramirez-Zea was asked to validate the equations developed by Dr Cole and compare their predictive performance with that of other BMR predictive equations generally used. Dr Ramirez-Zea then researched the recent literature for additional data sets that had not been included in the 1980s and in the Oxford data sets, but that fulfilled the established criteria for selecting data for those databases, and then tested the various equations against these data. Following this process, a careful review of all findings to date was undertaken by Dr A. Ferro-Luzzi, who suggested that this lengthy post-consultation exercise may not result in providing the experts with a new set of BMR predictive equations to be presented in the report.
In conclusion, the enhanced precision and robustness of the earlier equations, many of them in the published literature since the 1985 report, and the seamlessness of the Cole equation proved to be inadequate to persuade the expert consultation to warrant discontinuing the use of the international equations presented in the1985 report and widely used since then. Thus, for the current energy report, the experts decided to follow the advice of the FAO Secretariat and continue to use the Schofield BMR predictive equations. However, it was agreed that it was necessary to pursue an aggressive review of all the work that had been done to see whether the BMR equation question could be resolved more satisfactorily, both as a follow-up and in preparation for the next energy review, which it is hoped will take place within the next five years. Additional details as to how this decision was reached, along with Dr Henry’s background document on this topic, Dr Cole’s analysis and the review by Drs Ramirez-Zea and Ferro-Luzzi, will be published alongside all the background documentation related to this expert consultation as a supplement to the Public Health Nutrition journal in 2005.
REFERENCES Quenouille, M.H., Boyne, A.W., Fisher, W.B. & Leitch, I. 1951. Statistical studies of recorded energy expenditure of man. Part I. Basal metabolism related to sex, stature, age, climate and race. Commonwealth Bureau of Animal Nutrition Technical Communication No 17. Aberdeen, UK, Commonwealth Agricultural Bureau. Scrimshaw, S., Waterlow, J.C. & Schurch, B. (eds). 1996. Energy and protein requirements. Proceedings of an IDECG Workshop 31 October to 4 November 1994. Eur. J. Clin. Nutr., 50: S1–S197.
Annexes
89
ANNEX 4: SOFTWARE APPLICATION FOR CALCULATING POPULATIONS’ ENERGY
REQUIREMENTS AND FOOD NEEDS
This software application is an interactive program that allows the user great flexibility in customizing input parameters and population statistics. It is composed of a series of modules with the following functions:
calculation of average daily energy requirements for populations; estimation of corresponding quantities of food commodities (cereals, pulses, roots and tubers,
fish, meat, and fruits and vegetables) needed to meet the energy needs of the population; display of results in report and graphical formats.
The software allows users to base calculations on default data provided for countries and UN- defined regions, or to customize data for sub-country-level populations.
A. CONTENT OF THE CD-ROM: CALCULATING POPULATION ENERGY
REQUIREMENTS AND FOOD NEEDS
Installation instructions. Setup.exe. Read me first.txt. Calculating population energy requirements and food needs: user’s manual. This is a PDF
file of the user’s manual, which includes background information for the estimation of population energy requirements, a description of the data needed for the calculations, a description of the calculations for single age groups, an explanation of the values obtained from the software application, advice on maximizing information obtained from the software application, and annexes that demonstrate how the calculations are made. The user is advised to read this manual to get the most out of the software application.
Manual annexes with formulas and application databases in MS Excel format. Application files.
B. CALCULATION OF ENERGY REQUIREMENTS USED BY THE SOFTWARE
APPLICATION
There are four options available for calculating energy requirements, as summarized in Annex Table 1.
ANNEX TABLE 1
Options in the energy requirement module Default data, average daily energy requirements for populations
Customized data, average daily energy requirements for populations
Default data, daily energy requirements for special groups
Customized data, daily energy requirements for special groups
The default data option is chosen when the user wishes to calculate energy requirements for a UN- defined region or a country using default data provided with the application. The data provided include population structure by age and sex, crude birth rate and percent urban population for five- year periods from 2000 to 2025, as well as average body weight by age and sex. When choosing the default data option, a screen appears with the default data, any of which may be modified on screen if more recent information is available. This profile can be saved with a unique name and used again at a later date.
A N
N E
X T
A B
L E
2
S u
m m
a ry
o f
a g
e -s
p e c
if ic
c a lc
u la
ti o
n s
o f
a v e
ra g
e e
n e
rg y r
e q
u ir
e m
e n
ts
A g
e g
ro u
p
D e fa
u lt
b o
d y w
e ig
h ta
A
d d
it io
n a l
e n
e rg
y a
ll o
w a n
c e s
A v e ra
g e d
a il
y e
n e rg
y r
e q
u ir
e m
e n
ts b
A d
ju s tm
e n
ts
In fa
n ts
: 0 –
5 .9
9
m o
n th
s
R e
fe re
n c e w
e ig
h t-
fo r-
a g e
b y
s e x , 6
-m o n
th a
v e
ra g
e
E n
e rg
y d
e p o
s it io
n i n
n o
rm a l g
ro w
th
T E
E f
o r
b re
a s tf e
d i n
fa n
ts +
e n e
rg y
n e e
d e d
f o r
g ro
w th
F
o r
c a
lc u
la ti o n
o f th
e a
v e ra
g e
p e
r c a p
it a
e n
e rg
y
re q
u ir
e m
e n t
fo r
a c
o u n
tr y ,
th e
s u m
o f
a v e ra
g e
d a
ily e
n e
rg y r
e q u
ir e
m e
n ts
a c ro
s s a
g e
g ro
u p
s
d o
e s n
o t
in c lu
d e
b re
a s tf
e d
i n
fa n
ts a
g e
d 0
.0 – 5
.9
m o n
th s ,
w h
e re
a s t
h e a
d d
it io
n a
l a
llo w
a n c e
f o r
b re
a s tm
ilk p
ro d u
c ti o
n i s i n c lu
d e d
f o r
th e a
d u
lt
a g
e g
ro u p
o f
1 8 –
2 9
y e
a rs
In fa
n ts
: 6 –
1 1 .9
9
m o
n th
s
R e
fe re
n c e w
e ig
h t-
fo r-
a g e
b y
s e x , 6
m o
n th
a v e
ra g e
E
n e
rg y d
e p o
s it io
n i n
n o
rm a l g
ro w
th
T E
E f
o r
m ix
e d d
ie t
in fa
n ts
+ e
n e
rg y
n e e
d e d
f o r
g ro
w th
C h
ild re
n :
1 – 4
.9 9
y e a
rs
R e
fe re
n c e w
e ig
h t-
fo r-
a g e
, 1
y e
a r
a v e ra
g e
E n
e rg
y d
e p o
s it io
n i n
n o
rm a l g
ro w
th
T E
E +
e n e
rg y n
e e d
e d f
o r
g ro
w th
T
E E
i s r
e d
u c e
d b
y 7
% f
o r
a g
e 1
.0 –
1 .9
y e
a rs
c
C h
ild re
n :
5 – 1
7 .9
9
y e a
rs
W e ig
h t
fo r
g iv
e n
h e
ig h t
c o rr
e s p o
n d in
g t o
m e
d ia
n
re fe
re n
c e B
M I
fo r
a g
e -s
e x
E n
e rg
y d
e p o
s it io
n i n
n o
rm a l g
ro w
th
T E
E +
e n e
rg y n
e e d
e d f
o r
g ro
w th
A d u
lt s : 1
8 – 2
9 .9
9
y e a
rs
W e ig
h t
fo r
g iv
e n
h e
ig h t
c o rr
e s p o
n d in
g t o
B M
I o f
2 2
.0
(b o d
y w
e ig
h t
is h
e ld
c o
n s ta
n t
a c ro
s s a
d u lt a
g e
g ro
u p
s )
P re
g n a
n c y : in
c re
a s e
o f
1 .1
6 8
M J p
e r
d a
y a
v e
ra g e
d o
v e
r e
n ti re
p re
g n
a n c y ,
o r
0 .8
7 0
M J p
e r
d a
y a
v e ra
g e
d o
v e r
a
y e a
r
L a
c ta
ti o
n :
in c re
a s e o
f 2
.1 0
0 M
J p
e r
d a
y a
v e
ra g e
d o
v e
r 6
m o n
th s a
ft e r
b ir
th o
f c h
ild , o r
1 .0
5 0
M J p
e r
d a
y
a v e
ra g e
d o
v e
r a y
e a
r
B M
R ×
P A
L
(l o c a
ti o
n -s
p e
c if ic
P A
L a
s d
e te
rm in
e d
b y m
ix o
f ru
ra l a n
d u
rb a n
p e
rc e
n t
a n d
p h
y s ic
a l a
c ti v it y ,
o r
fi x e d
i n
c a
s e
o
f m
in im
u m
r e q
u ir
e m
e n t)
A d u
lt s : 3
0 – 5
9 .9
9
y e a
rs
W e ig
h t
fo r
g iv
e n
h e
ig h t
c o rr
e s p o
n d in
g t o
B M
I o f
2 2
.0
B M
R ×
P A
L
(a s a
b o
v e
)
A d u
lt s : 6
0 +
y e
a rs
W
e ig
h t
fo r
g iv
e n
h e
ig h t
c o rr
e s p o
n d in
g t o
B M
I o f
2 2
.0
B M
R ×
P A
L
(a s a
b o
v e
) a M
e d
ia n w
e ig
h t-
fo r-
a g
e (
in fa
n ts
a n
d c
h ild
re n 1
– 4
.9 9
y e
a rs
) a n
d m
e d
ia n
w e ig
h t-
fo r-
h e ig
h t fr
o m
N C
H S
/W H
O i n te
rn a ti o
n a
l re
fe re
n c e
p o
p u
la ti o
n g
ro w
th f
o r
in fa
n ts
a n
d c
h ild
re n (
W H
O ,
1 9
8 3 )
(A
n n e
x 1
o f
u s e
r’ s m
a n
u a l)
. T
h e
g iv
e n
h e
ig h ts
u s e
d t o
c a lc
u la
te a
v e
ra g e
a g
e -s
p e c if ic
b o
d y w
e ig
h ts
c o
m e
f ro
m t
h e
g ro
w th
c u rv
e s p
ro v id
e d
i n
t h
e 1
9 8 5
J a m
e s a
n d
S c h o
fi e
ld m
a n
u a l (J
a m
e s ,
W
.P .T
. &
S c h
o fi e
ld , E
.C .
1 9 9
0 .
H u
m a
n e
n e rg
y r
e q
u ir
e m
e n ts
. A
m a n
u a
l fo
r p la
n n e
rs a
n d n
u tr
it io
n is
ts . O
x fo
rd ,
U K
, O
x fo
rd M
e d
ic a
l P
u b
lic a ti o
n s u
n d e
r a
rr a n
g e
m e
n t
w it h
F A
O ).
b E
q u
a ti o n
s f
o r
T E
E a
n d
B M
R u
s e
d i n t
h e c
a lc
u la
ti o
n s a
re g
iv e
n i n
A n
n e
x 3
o f
th e
u s e
r’ s m
a n
u a l.
c F
o r
a g
e 1
.0 –
1 .9
y e a
rs ,
th e
p re
d ic
te d v
a lu
e s o
f T
E E
w e
re a
b o
u t
7 %
h ig
h e
r th
a n a
c tu
a l T
E E
m e
a s u
re m
e n ts
, s o
i n t
h e c
u rr
e n
t p
ro g ra
m t h
e T
E E
f o r
th e
y e
a r
1 .0
– 1 .9
i s a
d ju
s te
d d
o w
n b
y 7
%
to a
c c o
u n
t fo
r th
e d
is c re
p a
n c y (
s e
e C
h a
p te
r 4 o
f th
is r
e p
o rt
).
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
90
Annexes
91
The customized data option allows the user to supply data for a subnational area for which default data are not available, or for a country for which more recent information is available, using a template provided by the application. This profile can be saved with a unique name and used again at a later date.
The average daily energy requirements for populations option calculates energy requirements for healthy populations with a full range of physical activity lifestyles among adults and a mix of urban and rural residence. This option is indicated for food and nutrition planning under normal conditions. Average requirements may be calculated for both default and customized data. The user will be asked to make an educated guess of the PAL based on lifestyle patterns of urban and rural populations in order to calculate the location-specific PAL.
The daily energy requirements for special groups option calculates energy requirements for groups of people with more homogeneous activity lifestyles and residence (i.e. either urban or rural) using a fixed PAL value. Examples of special groups include settlements for refugees or internally displaced persons. This option is indicated for users who have a good knowledge of appropriate PAL values and who are concerned with food planning for special groups, such as rations for emergency use.
Annex Table 2 summarizes the protocols used by the software application to calculate age-specific average daily energy requirements.
C. PRESENTATION OF THE RESULTS
The energy requirement results are presented in two forms. First, per capita requirements for specific age and sex groups and the entire population are calculated, i.e. the average daily energy requirement for an average person of that group. Per capita requirements are then converted to population daily energy needs, i.e. the total number of joules or kilocalories needed to meet the daily energy needs of everyone in that group or in the population as a whole.
Food quantities corresponding to the percentages of the national food supply accounted for by six commodities that meet the population energy needs are reported in metric tonnes (1 000 kg) on a daily, monthly, semi-annual or annual basis for the population under consideration.
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
92
ANNEX 5: ENERGY COSTS OF ACTIVITIES
ACTIVITY MALES FEMALES
Average PAR
PAR Range Average PAR
PAR Range
General personal activities
Sleeping a 1.0 1.0
Lying a 1.2 1.2
Sitting quietly a 1.2 1.2
Standing a 1.4 1.5
Dressing 2.4 1.6–3.3 3.3
Washing hands/ face and hair 2.3
Plaiting hair 1.8
Eating and drinking 1.4 1.6
Means of transport
Walking around/ strolling 2.1 2.0–2.2 2.5 2.1–2.9
Walking slowly 2.8 2.6–3.0 3.0
Walking quickly 3.8
Walking uphill 7.1 5.5–8.6 5.4 4.8–6.1
Walking downhill 3.5 3.1–4.0 3.2
Climbing stairs 5.0
Sitting on a bus/train 1.2
Cycling 5.6 3.8–8.6 3.6
Cycling on a dirt road 7.0 5.0–9.0
Driving a motor cycle 2.7 2.4–3.0
Driving a car/truck 2.0
Paddling a canoe 3.0
Pulling a rickshaw (one person/no load) 5.3 4.0–6.6
Pulling a rickshaw (2 persons) 7.2 6.7–7.8
Horseback riding (slow) 3.6
Horseback riding (trotting) 5.2 4.8–5.5
Activities involving weight bearing
Walking with 15–20 kg load 3.5 3.4–3.5
Walking with 25–30 kg load 3.9 3.8–4.1
Carrying 20–30 kg load on head 3.5 2.4–4.2
Carrying 35–60 kg load on head 5.8 5.0–7.0
Carrying 27 kg load with shoulder straps – varying gradients 5.0 2.3–7.7
Carrying 27 kg load with forehead strap – varying gradients 5.32 2.4–8.0
Loading 9 kg sack on to a truck 5.78
Loading 16 kg sack on to a truck 9.65
Pulling hand cart – unloaded 4.82
Pulling hand cart with 185–370 kg load 8.3 7.0–9.6
Domestic chores
Cooking/preparing food
Collecting wood (for fuel) 3.3
Collecting water (from well) 4.5
Chopping wood (for fuel) 4.2 2.3–6.5
Kneading dough 3.4
Making tortillas 2.4
Peeling vegetables 1.9 1.3–2.4 1.5
Pounding grain 5.6 5.0–6.3
Annexes
93
ACTIVITY MALES FEMALES
Average PAR
PAR Range Average PAR
PAR Range
Cooking/preparing food (cont.)
Shopping 4.6
Squeezing coconut 2.4
Washing dishes 1.7 1.6–1.9
Child care
Child care (unspecified) 2.5
Bathing child (standing) 3.5
Carrying child 1.9
House cleaning
Housework (unspecified) 2.8 2.5–3.0
Beating mats/carpets 6.2 5.1–7.4
Bed making (tropical climate) 3.4
Bed making (cold climate) 4.9 4.6–5.1
Mopping/washing floor 4.4 3.4–6.5
Polishing floor 4.4
Sweeping 2.3 2.0–2.5
Vacuuming 3.9
Window cleaning 3.0 2.8–3.3
Laundry
Washing clothes (sitting/squatting) 2.8 2.6–3.0
Hanging washing out to dry 4.4 4.3–4.6
Ironing clothes 3.5 1.7
Sewing/knitting 1.6 1.5 1.3–1.8
Care of the yard/garden
Cleaning/sweeping yard 3.7 2.9–4.5 3.6
Weeding garden 3.3 2.4–5.1 2.9 2.7–3.6
Shovelling snow from driveway 7.9
Agricultural activities
General activities
Digging 5.6 5.7
Driving a tractor 2.1 1.9–2.3
Fertilizing (spreading manure) 5.2 4.9–5.4
Gleaning 4.5
Grinding grain using a mill stone 4.6
Hoeing 4.2 3.6–4.6 5.3 4.7–6.5
Loading sacks on to a truck 6.6
Ploughing with horse 4.8
Ploughing with tractor 3.4
Ploughing with buffalo 3.6
Spraying crops 4.3
Weeding 4.0 2.6–4.7 3.7 3.7–3.8
Cocoa crop
Collecting cocoa 2.9
Pruning 2.4
Splitting cocoa 2.0
Activities for coconut crop
Collecting (climbing trees) 4.2
Husking 5.6
Bagging and splitting 3.9
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
94
ACTIVITY MALES FEMALES
Average PAR
PAR Range Average PAR
PAR Range
Fruit crops (apple, orange)
Picking (with pole) 3.8
Picking by hand 3.4
Pruning trees 3.6
Groundnut crop
Harvesting 4.7
Planting 3.1
Shelling 1.6
Sorting 1.9
Weeding 3.2
Maize crop
Harvesting 5.1
Planting 4.1
Rice crop
Bundling rice 3.7 3.0
Fertilizing 3.1
Harvesting 3.5 2.4–4.2 3.8 3.5–4.4
Planting 3.7 3.5–4.0 3.6 2.6–4.7
Spraying 5.2
Threshing 5.4 4.6–5.0 5.1 4.8–5.4
Transplanting seedlings 3.3 3.1–3.4 3.7 3.5–4.0
Winnowing 2.9 2.3–3.6 2.7 2.5–2.9
Sugar cane crop
Cutting 7.0 6.6–7.9
Loading on to wagon 5.6
Tying cane 3.0
Tuber crops
Harvesting 4.4 3.5–5.7 3.0 2.8–3.4
Planting 5.0 3.9 3.6–5.0
Sorting (kneeling) 2.2 1.6–2.7
Animal husbandry
Carrying straw 3.1
Cleaning equipment 4.0
Cutting straw 5.0
Feeding animals 3.6
Grooming horses 5.5 3.8–7.1
Milking by hand 3.6 3.1–4.1
Milking by machine 3.2
Tending animals (feeding, watering, cleaning stable) 4.6
Hunting/fishing
Crabbing 4.51
Fishing with a line b 1.9
Fishing with a spear 2.3
Fishing with hands 3.94
Hunting (bats, birds, pigs) b 3.2
Occupational categories
Bakery work 2.5
Brewery work 2.9
Annexes
95
ACTIVITY MALES FEMALES
Average PAR
PAR Range Average PAR
PAR Range
Brickmaker
Earth cutting 5.6 5.5–5.7
Making mud bricks (squatting) 3.0
Builder
Carrying wood 6.6
Cement mixing with shovel 5.3
Chipping cement walls 3.3
Chiselling 5.0
Nailing 3.0
Planing softwood 5.7 4.4–7.1
Planing hardwood 8.0
Roofing 2.9
Sandpapering 2.9
Sawing softwood 5.3 5.0–5.6
Sawing hardwood 6.6
Painting 3.6
Firefighter
Dragging fire hose 9.8
Climbing steps with full gear 12.2
Flight attendant (serving food, beverages and galley work) b 3.0 3.1
Forester
Tree cutting 6.9 5.4–8.0
Sawing 5.7
Planting trees 4.1
Nursery work 3.6
Military training
Digging trenches 6.4 4.6–7.9
Drill 4.5 4.1–4.8
March (slow) 3.18
March 2–4 m/h (3.2–6.4 km/h) with 27 kg load b 4.9
Obstacle course 5.7 5.0–6.3
Miner
Drilling with jackhammer 3.9
Loading operations 3.2
Shovelling 4.6
Office worker
Filing 1.3 1.5
Reading 1.3 1.5
Sitting at desk a 1.3
Standing/moving around a 1.6
Typing 1.8 1.8
Writing 1.4 1.4
Postal worker
Climbing stairs 8.9 7.7–10.7
Sorting parcels (habitual) 5.4
Shoemaker 2.6 2.2
Tailor b
2.5
Textile factory worker (average of spinning, weaving, dyeing) b
3.1 2.2
Human energy requirements: Report of a Joint FAO/WHO/UNU Expert Consultation
96
ACTIVITY MALES FEMALES
Average PAR
PAR Range Average PAR
PAR Range
Sports activities
Aerobic dancing – low-intensity 3.51 4.24
Aerobic dancing – high-intensity 7.93 8.31
Basketball 6.95 7.74
Batting 4.85
Bowling 4.21
Callisthenics 5.44
Circuit training 6.96 6.29
Football 8.0 7.5–8.5
Golf 4.38
Rowing 6.7 5.34
Running – long distance b 6.34 6.55
Running – sprinting 8.21 8.28
Sailing 1.42 1.54
Swimming 9 8.5–9.4
Tennis 5.8 5.92
Volleyball 6.06 6.06
Miscellaneous recreational activities
Dancing 5.0 5.09
Listening to radio/music b 1.57 1.45–1.9 1.43
Painting 1.25 1.27
Playing cards/board games b 1.5 1.4–1.8 1.75
Playing the drums 3.71
Playing the piano 2.25
Playing the trumpet 1.77
Reading 1.22 1.25
Watching TV 1.64 1.72
Notes: This annex has been compiled from the background document provided to the expert consultation by M. Vaz et al. and also from the values referred in WHO. 1985. Energy and protein requirements: report of a joint FAO/WHO/UNU expert consultation. WHO Technical Report Series No. 724. Geneva. The average PAR is the average PAR reported from multiple studies, when such data exist. PAR range refers to the minimum and maximum PAR reported across studies for a particular activity. a These entries come from the WHO, 1985 report.
b These activities are averages of two or more categories.
- Human energy requirements
- FOREWORD
- CONTENTS
- PREFACE
- 1. INTRODUCTION
- 2. PRINCIPLES AND DEFINITIONS
- 3. ENERGY REQUIREMENTS OF INFANTS FROM BIRTH TO 12 MONTHS
- 4. ENERGY REQUIREMENTS OF CHILDREN AND ADOLESCENTS
- 5. ENERGY REQUIREMENTS OF ADULTS
- 6. ENERGY REQUIREMENTS OF PREGNANCY
- 7. ENERGY REQUIREMENTS OF LACTATION
- 8. RECOMMENDATIONS FOR FUTURE RESEARCH
- 9. CONCLUSIONS
- ANNEXES
Factorial approach(1).pdf
SHORT COMMUNICATION
Factorial estimation of daily energy expenditure using a simplified method was improved by adjustment for excess post-exercise oxygen consumption and thermic effect of food
PM Warwick
School of Biological Biomedical and Molecular Sciences, University of New England, Armidale, New South Wales, Australia
This study validated a simplified factorial method for measuring energy expenditure (EE) against EE measured by doubly labelled water (DEE), and investigated whether adjustment for excess post-exercise oxygen consumption (EPOC) and thermic effect of food (TEF) as described by Institute of Medicine (2002) improved the factorial measure. Seventeen normal weight subjects (10 females, seven males) were recruited from among university staff and students. Factorial EE was measured using a 16-activity- category method, published energy costs of activities and measured basal metabolic rate, before (FEE) and after (FEEadj) adjustment for EPOC and TEF, and by DEE. FEE underestimated daily EE by 14.6% relative to DEE (P¼ 0.000), but this underestimate was improved to 5.1% (P¼0.071) when FEE was adjusted for EPOC and TEF. Individual differences between DEE and FEEadj ranged from �20.2 to 17.6%, with 88% of subjects showing differences of less than 712%. European Journal of Clinical Nutrition (2006) 60, 1337–1340. doi:10.1038/sj.ejcn.1602460; published online 14 June 2006
Keywords: energy expenditure; doubly labelled water; factorial method
Introduction
Food energy requirements are best determined from estimates
of energy expenditure (EE), and factorial methods based on
activity levels expressed as multiples of basal metabolic rate
(BMR) have been proposed to do this (IOM, 2002; WHO/FAO/
UNU, 2004). However, factorial methods may underestimate
usual energy needs (IOM, 2002), and this may be partly owing
to the failure to account for (1) greater increases in daily EE
than estimated from measured energy costs of activities as a
result of the induction by exercise of an increase in EE
for some time after the exertion has been completed (excess
post-exercise oxygen consumption, or EPOC) and (2) the
thermic effect (TEF) of the extra food needed to cover the
energy cost of the activity (IOM, 2002). Thus, a modified
factorial approach that incorporates factors to account for
EPOC and TEF has been recommended (IOM, 2002). Regard-
less of the method used, it is difficult to accurately describe or
record the myriad of activities performed throughout the day
and simple approaches to diary recording are needed for
subject acceptance (Bratteby et al., 1997).
The aim of this study was to validate a simplified factorial
method of estimating daily EE against EE measured by
doubly labelled water (DEE), and to investigate whether
adjustment for EPOC and TEF as suggested by IOM (2002)
improved the factorial estimate.
Subjects and methods
Seventeen, normal weight non-smokers (10 females, seven
males) with sedentary occupations but variable levels of
leisure activity were recruited from among university staff
and students. Time spent in 16 activity categories (see
Table 1) was recorded for 28 days to the nearest 1–5 min on
time sheets. Daily EE was measured using a factorial method
before (FEE) and after (FEEadj) adjustment for EPOC and TEF, Received 17 October 2005; revised 13 January 2006; accepted 9 March 2006;
published online 14 June 2006
Correspondence: Dr PM Warwick, School of Biological Biomedical and
Molecular Sciences, University of New England, Armidale, New South Wales
2351, Australia.
E-mail: [email protected]
Guarantor: PM Warwick.
Contributor: PM Warwick.
European Journal of Clinical Nutrition (2006) 60, 1337–1340 & 2006 Nature Publishing Group All rights reserved 0954-3007/06 $30.00
www.nature.com/ejcn
and by using DEE. BMR was measured using a ventilated
hood as described previously (Warwick and Baines, 1996).
FEE was calculated from the time spent in each activity
category and its energy cost, as described previously
(Warwick, 1989), except that the energy cost of the first
on-foot category (F1 in Table 1) was increased from 2.0 to 2.5
because subjects had difficulty in differentiating between F1
and F2 activities and because good agreement between FEE
and 24 h EE in a respiration chamber was obtained using the
value of 2.5 (Warwick et al., 1988; Warwick and Busby, 1993).
The energy costs in Table 1 are compatible with those listed
by IOM (2002) and FAO/WHO/UNU (2004). FEE was
adjusted to account for EPOC and TEF exactly described by
IOM (2002), that is, by multiplying the energy expended in
standing, walking and exercise activities over and above
those necessary for a sedentary lifestyle by 1.278. This value
assumes 15% energy expenditure for EPOC and 9% for TEF
(1.15/0.9¼1.278). In the present study, the first 120 min of on foot (F1þ F2) activities were not included in the adjustment as they were assumed to be part of a sedentary
lifestyle. This amount of time was chosen because Table 12.2
of IOM (2002) lists 112 min of active-type activities as part of
a sedentary lifestyle, and because studies from this laboratory
have shown that 24 h EE measured in a whole-body
respirometer is about 1.3–1.4�BMR when subjects spend 90–120 min/day ‘on their feet’ but not exercising (unpub-
lished data from several studies).
DEE was measured in the middle of the study exactly as
described previously (Warwick and Baines, 1996), except that
DEE was calculated over 8 days in all subjects and equation
(2) of Speakman et al. (1993) was used after modification to
incorporate the measured 2H2-to- 18O dilution space ratios
(1.051470.020 in males and 1.028070.018 in females). All data manipulation and statistical analyses were carried out
using the MINITAB package (Minitab Inc., State College, PA,
USA). Statistical differences were assessed with paired
Student’s t-tests. Differences between measures of EE were
also assessed with the method of Bland and Altman (1986).
The study was approved by the Human Ethics Committee of
this university.
Results and discussion
The physical characteristics of the subjects, and daily EE by
FEE and DEE methods (mean7s.d.), and differences between the methods are shown in Table 2. FEEadj was 9% higher than
FEE, with individual differences ranging from 3.5 to 13.3%.
FEE was underestimated by 14.6% relative to DEE, but this
was improved to 5.1% (P¼0.071) when FEE was adjusted for EPOC and TEF. Some studies have also found factorial
underestimation of daily EE (Haggarty et al., 1994; Leonard
et al., 1997), but others have reported overestimation
(Alfonzo-Gonzalez et al., 2004; Walsh et al., 2004), or no
difference between FEE and DEE measures (Bratteby et al.,
1997; Jones et al., 1997; Morio et al., 1997; Withers et al.,
1998; Seale et al., 2002). Such discrepancies are very likely
owing to differences in methods used to determine BMR and
energy costs of activities, the number of activity categories
used and the accuracy of estimates or records of time spent
in activities, all of which affect the precision of factorial
methods. Indeed, IOM (2002) states that errors in the
calculation of physical activity levels using the energy costs
of activities in their tables are of minor importance
Table 1 Activity categories and their average energy costs expressed as multiples of basal metabolic rate
Activity code Energy cost of activity Example activities
LA 1.0 Lying asleep LQ 1.2 Lying quietly: for example, reading, thinking, listening) SQ 1.2 Sitting quietly: for example, reading, thinking, listening, watching TV, traveling in vehicle) SB 1.5 Sitting busy: for example, talking, eating, playing cards, typing, driving a vehicle, sewing, handcrafts, writing) F1 2.5 Low level ‘on-foot’ activities: for example, talking, ironing, washing, light bench work, very light handyman work,
food preparation, showering, dressing, very light housework, slow shopping F2 2.5 Light ‘on-foot’ activities: for example, sweeping, hanging clothes, light gardening activities, shopping,
lab/bench/handyman work, light household activities, production line assembly, fruit picking, pruning F3 4.0 Moderate ‘on-foot’ activities: for example, scrubbing, vacuuming, clearing garden, pushing wheelbarrow,
average building/lifting, bricklaying, carpentry, some agricultural work F4 7.0 Strenuous ‘on-foot activities’: for example, digging, chopping wood, very heavy building or lifting W1 2.0 Walking slowly (strolling), B2–3 km/h W2 3.0 Walking average pace, B4–6 km/h W3 5.0 Walking fast, B6 km/h, or uphill, or mow lawn with push mower E1 2.0 Very light exercise: for example, billiards, lawn bowls E2 3.0 Light exercise: for example, table tennis, ten pin bowling, golf, tennis (leisurely pace) E3 4.0 Moderate exercise: for example, cycling (slowly), volleyball, cricket, horse-riding, sailing, dancing, badminton,
tennis (moderate pace), swimming (slowly), aerobics (stretching) E4 7.0 Strenuous exercise; for example, tennis (fast pace), ice/roller skating, swimming (moderate pace),
cycling (moderate pace), rowing (moderate pace), fencing, gymnastics/aerobics, basketball, football, jogging/running B7–9 km/h, squash (moderate pace), weight training
E5 10.0 Very strenuous exercise: for example, swimming (fast pace), rowing (fast pace), cycling (fast pace), squash (fast pace), running B10–15 km/h
Factorial estimation of energy expenditure PM Warwick
1338
European Journal of Clinical Nutrition
compared to the large uncertainties in measuring the
duration and intensity of physical activities. However,
while adjustment for EPOC and TEF as described by IOM
(2002) improved our estimate of FEE, it would not be
appropriate to adjust other factorial methods that do not
underestimate EE.
One limitation of the IOM (2002) adjustment process is
that the 15% EPOC factor was derived from a single report,
which examined activities at 70% of maximal heart rate
(Bahr et al., 1987), and may not be applicable to all activities.
Thus, it might be more appropriate to apply the EPOC factor
only to those activities with energy costs of greater than
about 4�BMR, and then to apply the TEF adjustment. However, in the present study simply increasing the energy
costs of all our active activities (standing, walking and
exercise) by 30% would have resulted in very close agree-
ment between FEE and DEE (0.6% higher, P¼0.808). Some studies have reported that the degree of under-
estimation of EE by factorial methods increased with
increasing daily EE (Haggarty et al., 1994; Leonard et al.,
1997). Using a Bland–Altman analysis (Bland and Altman,
1986) where DEE minus FEE (or FEEadj) is plotted against the
average of the two EEs, this relationship was also found in
the present study before adjusting for EPOC and TEF
(r¼0.663, P¼0.004), but not after this adjustment (r¼�0.378, P¼0.134), or after increasing the cost of active activities by 30% (r¼0.163, P¼0.532).
In the present study, individual differences between DEE
and FEEadj ranged from �20.2 to 17.6%, with differences of less than 712% in 88% of subjects. Had we applied the 30% increase in energy cost to all active activities, 94% of subjects
would have shown differences of less than 712.8%. Many other studies have reported acceptable factorial estimates for
groups but less acceptable estimates for individuals (Bratteby
et al., 1997; Morio et al., 1997; Withers et al., 1998).
In conclusion, FEE was underestimated in a group of
normal weight subjects and this difference was improved
after adjustment for EPOC and TEF as outlined by IOM
(2002). Increasing the energy costs of all active activities by
30% also resulted in close agreement between FEE and DEE.
References
Alfonzo-Gonzalez G, Doucet E, Almeras N, Bouchard C, Tremblay A (2004). Estimation of daily energy needs with the FAO/WHO/UNU 1985 procedures in adults: comparison to whole-body indirect calorimetry measurements. Eur J Clin Nutr 58, 1125–1131.
Bahr R, Ingnes I, Vaage O, Sejersted OM, Newsholme EA (1987). Effect of duration of exercise on excess post-exercise oxygen consumption. J Appl Physiol 62, 485–490.
Bland JM, Altman DG (1986). Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1, 307–310.
Bratteby LE, Sandhagen B, Fan H, Samuelson G (1997). A 7-day activity diary for assessment of daily energy expenditure validated by the doubly labelled water method in adolescents. Eur J Clin Nutr 51, 585–591.
FAO/WHO/UNU (2004). Human Energy Requirements. FAO: Rome. Haggarty P, McNeill G, Abumanneh MK, Davidson L, Milne E,
Duncan G et al. (1994). The influence of exercise on the energy- requirements of adult males in the UK. Br J Nutr 72, 799–813.
IOM (2002). Physical activity. In: Institute of Medicine/Food and Nutrition Board (ed). Dietary Reference Intakes for Energy, Carbohy- drate, Fiber, Fat, Fatty Acids, Cholesterol, Protein and Amino Acids (Macronutrients). National Academy Press: Washington, DC, pp 697–736.
Jones PJH, Martin LJ, Su WF, Boyd NF (1997). Canadian recom- mended nutrient intakes underestimate true energy requirements in middle-aged women. Can J Public Health 88, 314–319.
Leonard WR, Galloway V, Ivakine E (1997). Underestimation of daily energy expenditure with the factorial method: Implications for anthropological research. Am J Phys Anthropol 103, 443–454.
Morio B, Ritz P, Verdier E, Montaurier C, Beaufrere B, Vermorel M (1997). Critical evaluation of the factorial and heart-rate recording methods for the determination of energy expenditure of free-living elderly people. Br J Nutr 78, 709–722.
Seale JL, Klein G, Friedmann J, Jensen GL, Mitchell DC, Smiciklas- Wright H (2002). Energy expenditure measured by doubly labeled water, activity recall, and diet records in the rural elderly. Nutrition 18, 568–573.
Speakman JR, Nair KS, Goran MI (1993). Revised equations for calculating CO2 production from doubly labeled water in humans. Am J Physiol 264, E912–E917.
Table 2 Physical characteristics of subjects, EE and PAL by FEE and FEEadj and DEE methods (mean7s.d.), and differences between measures of EE
All subjects (n¼17)
Males (n¼7)
Females (n¼10)
Age (years) 27.875.8 29.674.2 26.676.7 BMI (kg/m2) 22.271.6 22.971.7 21.771.7 Body weight (kg) 66.078.7 73.278.8 61.073.9 Body fat (%) 23.378.3 15.775.9 28.674.9 FEE (MJ/day) 10.4172.46 12.4872.40 8.9771.15 FEEadj (MJ/day) 11.3572.90 13.6872.89 9.7371.50 DEE (MJ/day) 11.9373.36 14.2974.09 10.2871.23 PAL (FEE) 1.6370.13 1.6770.15 1.6070.12 PAL (FEEadj) 1.7770.19 1.8370.21 1.7470.18 PAL (DEE) 1.8670.24 1.9070.34 1.8470.17
Difference: FEEadj�FEE MJ/day (%) 0.94 (9.0%) 1.20 (9.6%) 0.76 (8.5%) P-valuea P¼0.000 P¼0.001 P¼0.000 95% CIb 0.70, 1.12 0.71, 1.69 0.48, 1.03 Limitsb �0.03, 1.91 0.15, 2.25 0.00, 1.51
Difference: DEE�FEE MJ/day (%) 1.52 (14.6%) 1.81 (14.5%) 1.31 (14.6%) P-valuea P¼0.000 P¼0.057 P¼0.000 95% CIb 0.81, 2.22 �0.08, 3.71 0.83, 1.78 Limitsb �1.23, 4.26 �2.28, 5.90 0.02, 2.64
Difference: DEE�FEEadj MJ/day (%) 0.58 (5.1%) 0.61 (4.5%) 0.55 (5.7%) P–valuea P¼0.071 P¼0.388 P¼0.062 95% CIb �0.06, 1.21 �1.00, 2.23 �0.04, 1.14 Limitsb 1.89, 3.05 2.87, 4.10 �1.09, 2.20
Abbreviations: CI, confidence interval; DEE, doubly labelled water; EE, energy
expenditure; FEE, factorial energy expenditure; PAL, physical activity levels. aStatistical significance of differences using paired t-tests. bData from Bland–Altman analysis (Bland and Altman, 1986) where 95% CI is
for the bias, and limits is the limits of agreement in MJ/day (the mean
difference in daily EE as estimated by the two methods 72 s.d.).
Factorial estimation of energy expenditure PM Warwick
1339
European Journal of Clinical Nutrition
Walsh MC, Hunter GR, Sirikul B, Gower BA (2004). Comparison of self-reported with objectively assessed energy expenditure in black and white women before and after weight loss. Am J Clin Nutr 79, 1013–1019.
Warwick PM (1989). Predicting food energy requirements from estimates of energy expenditure. Aust J Nutr Diet 46 (Suppl), S1–S28.
Warwick PM, Baines J (1996). Energy expenditure in free-living smokers and nonsmokers: comparison between factorial, intake-balance, and doubly labeled water measures. Am J Clin Nutr 63, 15–21.
Warwick PM, Busby R (1993). Prediction of twenty-four-hour energy expenditure in a respiration chamber in smokers and non- smokers. Eur J Clin Nutr 47, 600–603.
Warwick PM, Edmundson HM, Thomson ES (1988). Prediction of energy expenditure: simplified FAO/WHO/UNU factorial method vs continuous respirometry and habitual energy intake. Am J Clin Nutr 48, 1188–1196.
Withers RT, Smith DA, Tucker RC, Brinkman M, Clark DG (1998). Energy metabolism in sedentary and active 49- to 70-yr-old women. J Appl Physiol 84, 1333–1340.
Factorial estimation of energy expenditure PM Warwick
1340
European Journal of Clinical Nutrition
Am J Clin Nutr-1988-Romieu-406-12(1)(1).pdf
Energy intake and other determinants of relative weight13
406 Am J C/in Nuir l988;47:406-l2. Printed in USA. © 1988 American Society for Clinical Nutrition
Isabelle Romieu, MD; Walter C Willett, MD; Meir J Stampfer, MD; Graham A Colditz, MBBS; Laura Sampson, RD; Bernard Rosner, PhD; Charles H Hennekens, MD; and Frank E Speizer, MD
ABSTRACT The relationships ofrelative weight to energy intake and to physical activity were studied among 141 females aged 34-59 y. As observed in previous studies Quetelet index
(wt/ht2) was inversely related to energy intake (r = -0. 1 1). However, obese women tended to
be older (r = 0.16), exercise less (r = -0.30), and drink less alcohol (r = -0. 16) than nonobese women. Older women had lower energy intake (r = -0.23) and exercised less (r = -0. 12) than younger women. Energy intake and physical activity were positively related (r = 0.23). After adjustment for age, physical activity, alcohol, and smoking, the inverse correlation between relative weight and energy intake was significantly reduced (p = 0.04) from r = -0. 1 1 to r = -0.02. Obese women reported higher intakes oftotal fat, and relative weight was significantly
correlated with intakes of total fat (r = 0.20) and saturated fatty acids (r = 0. 16). These data highlight the importance of considering factors that may confound the relationship between
energy intake and obesity, and they suggest that fat intake may play a role in obesity that is
independent of total energy intake. Am J C/in Nutr 1988;47:406-12.
KEY WORDS Caloric intake, energy, physical activity, alcohol, cigarettes, fat, obesity
Introduction
Weight gain leading to obesity occurs when energy in- take chronically exceeds energy expenditure. This dis- equilibrium in energy balance has been widely studied in experimental settings (1-9). J#{233}quierand Schutz (2), us- ing a respiratory chamber, clearly demonstrated that en- ergy expenditure was a linear function of body weight. Similar results reported by Prentice and Black (10) lead to the conclusion that, given a uniform level of physical activity, to gain weight and to maintain their weight obese subjects eat more than lean subjects because their 24-h energy expenditure is higher (1 1). In contrast, an inverse relationship between body fatness and energy in- take has been observed in most (12-19) but not all (20, 2 1) cross-sectional studies. One explanation for this in- verse association may be that obese people systematically
tend to underestimate their food intake (1 1, 14, 15, 17). Another possibility is that obese people tend to have lower levels ofphysical activity.
Additional explanations for the inverse relationship between body fatness and energy intake must also be considered because energy intake has a complex mean- ing. Body size, metabolic efficiency, physical activity,
and net energy balance (ie, change in body energy stores) (22) are all related to energy intake, therefore any study that simply examines the relationship between obesity and total energy intake is difficult to interpret. Age, smoking, and alcohol intake may influence metabolic
efficiency and distort the association between energy in- take and obesity. Obesity tends to increase while energy intake tends to decrease with age (19). Cigarette smoking
and relative weight are inversely related (12, 22-25) de- spite the fact that on the average smokers consume more calories per day than nonsmokers (1 8). Evidence sug- gests that drinkers also weigh less on average than non- drinkers despite consuming more calories, perhaps be-
cause metabolism of energy from alcohol calories is in- efficient (26).
Another possible explanation of the inverse associa- tion between obesity and energy intake is that, although obese subjects do not eat more, they follow different di- etary patterns and consume different nutrients. In ani- mal studies dietary fat is more efficiently converted to fat than is energy in other forms (27, 28). Similar findings were reported among humans in carbohydrate balance
I From the Channing Laboratory(allauthors), Department of Medi-
cine, Harvard Medical School and Brigham and Women’s Hospital, Boston, MA; the Department of Epidemiology (IR, WCW), Harvard School ofPublic Health, Boston, MA; and the Department of Preven- tive Medicine and Clinical Epidemiology (BR, CHH), Harvard Medi-
cal School, Boston, MA. 2 Supported by research grants HL 24074, HL 34594, AM 36798,
CA 40935, and CA 40356 from the National Institutes of Health. 3 Address reprint requests to Dr Walter C Willett, The Channing
Laboratory, 180 Longwood Avenue, Boston, MA 02 1 15. Received December 1, 1986.
Accepted for publication May 21, 1987.
at LO N
D O
N M
E T
R O
P O
LIT A
N U
N IV
E R
S IT
Y on M
arch 18, 2015 ajcn.nutrition.org
D ow
nloaded from
DETERMINANTS OF RELATIVE WEIGHT 407
(29-32). Individual differences in metabolic efficiency
(absorption and utilization of nutrients) may also play a role, because under carefully controlled conditions some
subjects gain weight more rapidly than others who have similar energy intake (33). The mechanisms and deter- minants of these individual differences are poorly de- fined in humans, however, and are difficult to measure in an epidemiologic setting (22).
To further investigate the relationship between obesity and energy intake, we analyzed data collected during the validation of a dietary questionnaire from the Nurses’ Health Study (34). This provided an unusually detailed assessment of dietary intake over 1 y.
Materials and methods
Population and collection of diet records
As part ofa study conducted to validate a self-administered semiquantitative food-frequency questionnaire (34) an age-
stratified sample of 225 female participants in the Nurses’ Health study, aged 34-59 y, residing in the greater Boston area, was randomly selected and asked to complete four l-wk weighed records of food intake. Of the original sample, 194 women (86%) agreed to participate and were instructed at home by a research dietitian to record all food and beverages consumed for a l-wk period in a specially designed booklet. Women were provided with a dietetic scale and taught a method of common household measures to use when away from home. The subjects provided detailed recipes and lists of ingredients for complex food items. Participants completed 1- wk records at 3-mo intervals for a total of 4 wk throughout a full year. To minimize variability in the interpretation, all the diet records were reviewed by one research dietitian. Nutrient intake from the diet records were computed using the Univer- sity of Massachussetts data-base system. Alcohol intake was similarly calculated from the consumption of beer, wine, and liquor recorded during the 28 d. Information on weight, height, and smoking habits had been obtained previously by question- naire. The research protocol was approved by the Human Re- search Committee ofthe Brigham and Women’s Hospital.
Physical activity
A questionnaire to assess physical activity was mailed to the 194 subjects at the 4th week ofdiet recording. This self-admin- istered form, based on the Harvard Alumni Activity Survey questionnaire (35), was designed to obtain information on the
types and duration ofactivities engaged in during the previous week and year. Questions included how many flights of stairs were climbed per day, how many city blocks were walked per day, how much time was spent in specific sports or recreational activities, and what was the average time spent on sports at each of four levels of intensity of activities (vigorous, moderate, light, and sleeping time) over the previous 7 d, divided into weekdays and weekend.
A physical-activity index was devised to provide a composite estimate of energy expenditure; weights were assigned accord- ing to the intensity of the activity, average time per episode (in minutes), and frequency. Activities were then expressed in kilocalories ofenergy expenditure and summed (36).
Bodyfat index
Quetelet index (wt/ht2) was calculated from the weights and heights reported on the questionnaires. This body-mass index
was independent of height (r -0.03, p 0.65) and strongly related to weight (r = 0.86, p = 0.0001). To validate the self- reported weight, the 194 participants were weighed in their homes on a digital bathroom scale after they completed the questionnaire. Although the measured weight averaged 1 .5 kg (3.3 lb) higher than the self-reported weight, the weights were highly correlated (r = 0.96) (25). We restricted our analysis to the 141 women who did not report a weight change > 2.3 kg (5 lb) during the year of observation and therefore were consid- ered to have a stable energy balance.
St atistical analysis
Quetelet index and energy intake were used as either categor- ical (tertiles) or continuous variables. As an alternate measure-
ment of energy intake the residual from the regression model, with height as the independent variable and total energy intake as the dependent variable, was used to remove the variability in energy intake attributable to height. Data on physical activity,
expressed in kilocalories per day, were combined into two mdi- ces after the preliminary analysis. The first index (Exercise I) was based on blocks walked and stairs climbed per day and on sports during the previous week (“list any sports or recreations you have participated in during the past week as actual play- time, bicycling time . . .“); the second index (Exercise II) re- lated more specifically to vigorous activity (“on a usual week-
day and a weekend day during the past week how much time did you spend on vigorous activities such as digging in garden, jogging, bicycling on hills . . .“). To better control for physical activity both indices were used although they are not indepen- dent of each other. Nutrient intake scores were adjusted for total caloric intake and height by computing residuals from re- gression models, with total energy intake and height as inde- pendent variables and nutrient index scores as the dependent variable. This was done to explore the relationship between those nutrients and Quetelet index after the variability due to energy intake and height had been removed (22).
Data were analyzed using multivariate linear regression models (37). The log transformation was used as necessary to normalize skewed variables such as Exercise I, Exercise II, and most nutrients. Product moment (Pearson) correlation co- efficients were used (37) to examine the crude relationship be-
tween Quetelet index and energy intake after controlling for potential confounders. With a sample of 141 subjects a correla-
tion coefficient of � 0. 16 is statistically significant (p < 0.05, two-sided test). We compared the group means by analysis of variance (ANOVA component of SAS [37]), and when appro- priate we calculated the test for trend across tertiles of obesity
or energy intake (37). Correlation coefficients were compared
by Wolfe’s method (38).
Results
In our sample of 14 1 women, the Quetelet index ranged from 1 7.6 to 39.2 kg/m2 with a mean of 24.0 ± 9.7 (SD) kg/m2. Total energy intake ranged from 909 to 25 1 8 kcal/d with a mean of 1 622 ± 332.3 kcal/d.
Age was positively correlated with obesity (r = 0. 16, p = 0.05) and negatively correlated with Exercise II (r = -0. 12, p = 0. 10). There was a weak inverse correlation
(both crudely and after adjustment for height) between energy intake (kcal/d) and Quetelet index (Table 1). A strong inverse correlation was observed between Exer- cisc I (including blocks walked and stairs climbed per day
at LO N
D O
N M
E T
R O
P O
LIT A
N U
N IV
E R
S IT
Y on M
arch 18, 2015 ajcn.nutrition.org
D ow
nloaded from
408 ROMIEU ET AL
TABLE 1 Mean energy intake and other variable s according to tertile ofQ uetelet index for a sample o f 141 women aged 34-59 yr#{174}
Variables
Lowest tertile < 22.2 kg/m2
(n = 45)
Middle tertile
22.2-24.4 kg/rn2 (n = 49)
Highest tertile
> 24.4 kg/m2 (n = 47)
Pearson correlation with
Quetelet indext
r p
Energyintake(kcal/d) 1684±315 1639±274 1635±340 -0.13 0.13
Energy intake, height-adjusted (kcal/d) 17 12 ± 3 10 1624 ± 267 1624 ± 337 -0. 1 1 0.18
Total fat (g/d) 70 ± 17 68 ± 14 70 ± 17 -0.05 0.51 Saturated fauy acid (g/d) 25 ± 6 24 ± 6 26 ± 7 -0.02 0.81 Polyunsaturated fatty acids (g/d) 1 1 ± 4 1 1 ± 3 1 1 ± 3 -0.09 0.30 Age (y) 45 ± 7 47 ± 6 48 ± 8 0. 16 0.05 Exercise I (kcal/d)t 247 ± 200 209 ± 195 179 ± 190 -0.3 1 0.0002 Exercise II (kcal/d)� 26.5 ± 37.9 24.0 ± 43.4 15 ± 23 -0.00 0.34 Alcohol(g/d) 13.37±11.13 10.58±10.18 7.58±8.53 -0.30 0.006 Current cigarette smokers (%)lI 23.8 32.6 31.9
6 Mean ± SD.
t All correlation coefficients were calculated with Quetelet index as a continuous variable and log-transformed nutrient and exercise data. :1:Exercise I: blocks walked and stairs climbed per day plus sports over the previous week. § Exercise II: vigorous activity over the previous week. II Three smoking-status values missing in lowest tertile ofQuetelet index. Chi-square for trend not significant.
and physical activity during previous week) and Quetelet index (r = -0.3 1, p = 0.0002). Adjustment for age did not affect the results. The mean Exercise I score of the highest tertile of Quetelet index was significantly lower
than that ofthe lowest tertile (p = 0.03 by ANOVA). Alcohol intake was also inversely related to obesity (r
= -0.30, p = 0.006). The proportion of smokers was lower in the first tertile of relative weight but not signifi-
cantly different from the other tertiles (Table 1). We then examined the distribution of the different
variables by tertile of energy intake (Table 2). Younger
women tend to eat more calories (r = -0.23), and energy intake was positively related to exercise (Exercise I, r
= 0. 18; Exercise II, r = 0.23). Results remained similar after adjustment for age. The mean level of Exercise II was significantly higher in the highest tertile of energy
intake compared with the lower two tertiles. As expected, fatty acid intake (total, saturated, and polyunsaturated) was correlated with total energy intake, but alcohol in-
take was not. We next examined these covariates within strata of
smoking status (Table 3). An inverse relationship existed
TABLE 2 Mean height, age, and other variables according to tertile ofenergy intake among a sample of 141 women aged 34-59 y#{174}
Variables
Lowest tertile < l48Okcal/d
(n = 40)
Middle tertile l480-l7l8kcal/d
(n = 48)
Highest tertile > l7l8kcal/d
(n = 53)
Pearson correlation withenergyintake�
r p
Height (m) 1.64 ± 0.06 1.63 ± 0.05 1.65 ± 0.05 0. 17 0.04
Quetelet index (kg/m2) 24.8 ± 4.5 23.5 ± 2.7 23.5 ± 2.9 -0. 13 0.13 Age (y) 49. 1 ± 7.2 46.8 ± 7.6 45. 1 ± 7.6 -0.23 0.006 Total fat(g/d) 54.1 ± 10.2 66.3 ± 9.3 85.4 ± 13 0.88 0.0001 Saturatedfattyacid(g/d) 19.6±3.9 23.5±4.6 31.6±0.7 0.83 0.0001
Polyunsaturated fatty acids(g/d) 8.9 ± 2.7 11 ± 2.4 13.3 ± 3.5 0.64 0.0001
Exercisel(kcal/d4 159.7±185 236.6±179 230.1±212.5 0.18 0.02
Exercise II (kcal/d)� 1 1 .5 ± 18.9 28 ± 48.4 23.8 ± 30.7 0.23 0.007 Alcohol (g/d) 6.8 ± 7 7.8 ± 8 12.6 ± 12.3 0.09 0.19
Current cigarette smokers (%)II 36.5 25 25
6 Mean ± SD.
t All correlation coefficients were calculated with Quetelet index as a continuous variable and log-transformed nutrient and exercise data. t Exercise I: blocks walked and stairs climbed per day plus over the previous week. § Exercise II: vigorous activity over the previous week. II Three values on smoking status missing in lowest tertile.
at LO N
D O
N M
E T
R O
P O
LIT A
N U
N IV
E R
S IT
Y on M
arch 18, 2015 ajcn.nutrition.org
D ow
nloaded from
DETERMINANTS OF RELATIVE WEIGHT 409
TABLE 3 Mean energy intake and other variables according to tertile ofQuetelet index, by smoking status#{174}t
Lowest tertile Middi e tertile Highes t tertile
Smokers NonsmokersSmokers Nonsmokers Smokers Nonsmokers
Variables (n=lO) (n=32) (n=l6) (n=33) (n=l5) (n=32)
Energy intake (kcal/d) 1540 ± 350 1697 ± 288 1564 ± 309 1676 ± 252 1541 ± 379 1681 ± 316
Energy intake, height-adjusted (kcal/d) 1439±355 1734±282 1480±307 1694±244 1395±383 1693 ±308
Totalfat(g/d) 64.1 ± 16.6 69.9± 16.5 66.9± 16.4 68.8± 13.4 66.9± 17.9 72.4± 16.9 Saturated fatty acids(g/d) 23.5 ± 6.3 24.9 ± 6.8 25.3 ± 7.4 24.5 ± 5.6 24.5 ± 6.8 26.5 ± 7.4 Polyunsaturated fattyacids(g/d) 10.7 ± 3.7 1 1.5 ± 3.8 10.2 ± 3 1 1.6 ± 3.1 10.2 ± 3.5 1 1.3 ± 3.2 Age (y) 45.8 ± 10 45 ± 7. 1 47.6 ± 6.8 46.7 ± 6.5 49.4 ± 7.5 47.2 ± 8.4 Exercisel(kcal/d4 178± 147 270±217 146± 120 239±218 161 ±234 137± 170 Exercise II(kcal/d).� 20.7 ± 22.9 24.9 ± 35.8 13.7 ± 29.7 28.9 ± 48.4 14.5 ± 21.6 15.2 ± 24.1 Alcohol (g/d) 16.66 ± 18.66 12.46 ± 8.2 14.0 ± 1 1.2 8.92 ± 9.4 7.18 ± 8.03 7.77 ± 8.8
* Mean ± SD.
t Three values on smoking status missing in lowest tertile. :1:Exercise I: blocks walked and stairs climbed per day plus sports over the previous week. § Exercise II: vigorous activity over the previous week.
between smoking and energy intake. Nonsmokers tended to be younger, to consume more energy and fat, and to report more vigorous exercise (Exercise II). In the first two tertiles ofQuetelet index, nonsmokers also consumed less alcohol; however, none of the comparisons of group means for alcohol intake were statistically significant.
Because factors associated with both energy intake and obesity may distort the association between these two variables, we calculated correlation coefficients adjusted
for these potentially confounding factors using multivar- iate linear regression models (Table 4). After accounting for age and exercise the correlation between energy in- take and Quetelet index was reduced from r = -0. 1 1 to r = -0.02, a statistically significant difference (p = 0.04) (38). Further adjustment for exercise and fat and alcohol
consumption did not modify this result. Subsequent analysis within strata of smokers and nonsmokers showed that the inverse correlation between Quetelet in- dex and energy intake was present only among current smokers (Table 4). Among nonsmokers the crude corre- lation was only r = -0.02 and subsequent adjustment modified this to r = 0.02. Because Donato et al (39) sug- gested that the energy contribution of dietary fat should be 1 1 . 1 kcal/g (ie, 24% higher than the typical value of 9 kcal/g), we used this value in an alternative analysis. This
did not materially alter the correlation between energy intake and obesity.
Different distributions of macronutrient intake for a given level of total energy intake may play a role in obe-
sity. We therefore calculated the partial correlation co-
efficients of nutrients (after adjustment for age and total energy intake) with relative weight (Table 5). Calorie-ad- justed total fat intake was positively correlated with obe- sity (r = 0.20, p = 0.02). When split into its components, this positive association was principally due to saturated fatty acids (r = 0.22, p = 0.01). After we controlled for smoking, the coefficients remained similar (nonsmokers,
r = 0.20; smokers, r = 0. 16), but the statistical signifi- cance was lost because of the small sample size in this stratum (97 nonsmokers and 41 smokers).
Discussion
These findings indicate that there is a weak inverse as-
sociation between relative weight and total energy in- take. They are consistent with findings from earlier epi-
TABLE 4 Correlation between energy intake and Quetelet index adjusted for height and potential confounders#{174}
Pearson correlation
Non- Total population Smokers smokers
Covariates (n= 138) (n=41) (n=97)
Height -0. 1 1 -0. 12 -0.02 + age -0.06 p = 0.04t -0.07 -0.01 + age + exercise Ij -0.02 -0.03 0.02 + age + exercise I
+ exercise II� -0.03 -0.04 0.01 + age + exercise I
+ exercise II + total fat� (g/d) -0.02 -0.03 0.01
+ age + exercise I + exercise II + total fat (g/d) + alcohol (g/d) -0.03 -0.05 0.02
6 Smokers and nonsmokers were analyzed separately (smoking sta-
tus was missing for three women).
t Test for difference in correlation coefficients(model with height vs model with height, age, and exercise I).
t Exercise and dietary variables were log-transformed to improve normality.
§ Total fat and alcohol intake were calorie-adjusted.
at LO N
D O
N M
E T
R O
P O
LIT A
N U
N IV
E R
S IT
Y on M
arch 18, 2015 ajcn.nutrition.org
D ow
nloaded from
410 ROMIEU ET AL
TABLES Partialcorrelation coefficients between Quetelet index and different
nutrients after adjustment for age and total caloric intake
Dietary variables#{174}
Pearson correlationt
r p
g/d
Total fatty acids Saturated fatty acids Polyunsaturated fatty acids Carbohydrate
Protein
Alcohol
0.20 0.16
0.05 -0.12
0.10 -0.29
0.02 0.05
0.5 0.2
0.3 0.00
6 All dietary variables were calorie-adjusted.
t All correlation coefficients were calculated with Quetelet index as
a continuous variable and log-transformed nutrient and exercise data.
demiologic studies (12-19). In the Zutphen study (16) the authors reported an inverse relationship between bo- dy-fat index and energy intake. A similar association has been observed by Baecke et al (14) in young Dutch adults and in the Framingham study population (19). Lincoln (2 1) reported no difference between obese and control groups with respect to energy intake and physical activ- ity, and similar results are reported in the National Health and Nutrition Survey of 197 1 (HANES I) (15). However, when we controlled for age and exercise in our data the inverse association between diet and obesity was reduced to essentially null.
Lack ofprecision in the assessment ofenergy intake is unlikely to account for the finding that there is no associ- ation between diet and obesity, because it was based on 28 d of weighed-diet recording and participants were highly motivated and professionally experienced in quantitative measurements. However, underestimation ofenergy intake by overweight women, suggested by sev-
eral authors (1 1 , 14, 1 5, 17), cannot be ruled out. maccu- racy in self-report of weight may have led to some ran- dom misclassification in Quetelet index, but that is un- likely to explain the lack ofa positive association. Recall of body weight was substantiated by measurement, and we limited the study to women with stable energy bal- ance. The use of Quetelet index as a measure of body fatness was debated recently (40, 4 1); however, it re- mains a widely used standard and was appropriate in this population because of its independence of height and high correlation with weight.
Physical activity was negatively correlated with obe- sity, which is consistent with many previous results (15, 19, 42), although other studies failed to demonstrate any difference between obese subjects and lean controls with respect to physical activity (2 1, 43, 44). Washburn and Montoye (45) assessed the reliability of the Harvard Alumni Activity Survey questionnaire in 59 postmeno- pausal women, aged 44-77 y. They found test-retest cor- relations for kilocalories of energy expenditure, blocks walked, stairs climbed, and sweat episodes per week of
0.73, 0.42, 0.54, and 0.46, respectively. The reliability of the part of our questionnaire that addressed intensity of activities was assessed on two occasions 9 mo apart and yielded a Pearson correlation of0.3 (unpublished obser- vations). This is consistent with other epidemiologic
measures ofexercise that are strong predictors of disease, such as the Framingham questionnaire (46, 47). How- ever, the validity of our questionnaire has not been tested, and one potentially important source of misclassi- fication concerns the accuracy of ascribing intensity to physical activity (48). A random misclassification of phy- sical-activity level would reduce the ability to control for its confounding effect as a covariate (49) and hence would tend to cause an underestimation of any positive correlation (50) between energy intake and obesity, be- cause physical activity is inversely correlated with obe- sity.
Adjustment for potential confounding variables, such as age, physical activity, smoking, and alcohol intake, re- duced the weak inverse relation between energy intake and obesity observed in our crude data. Braitman et al (15) reported a similar result in an adult female popula- tion after controlling for age and physical activity. In contrast, an inverse relationship between energy intake and relative weight persisted in young Dutch adults (14), but the age range in that population (20-32 y) was different from ours, and food intake was measured on two randomly selected days. Cigarette consumption is of particular interest because the inverse association was observed only among current smokers. Through its effects on the sympathetic nervous system, smoking may tend to increase the activity of energy-expending sub- strate cycles. Ceasing cigarette smoking would presum- ably return sympathetic activity and catecholamine 1ev- els to normal and facilitate more efficient energy storage and weight gain (24). Hofsteller et al (23) observed a 10% increase in 24-h energy expenditure among smokers of 25 cigarettes/d vs nonsmokers. In our study smokers tended to report lower energy intake than nonsmokers, and the lowest tertile of Quetelet index had the fewest smokers. By contrast, most other studies have found that smokers consume more calories than nonsmokers (51). The lack ofan inverse association between smoking and Quetelet index in this sample may well be due to random error because, as in most populations, in the Nurses’ Health Study (25) smokers tended to weigh less than nonsmokers.
The higher intake of saturated fatty acids observed among obese women is consistent with results from ani- mal studies. Obese animals gain weight in proportion to the fat concentration in the diet (27, 28). The weight gain is particularly enhanced by diets containing long- rather than medium-chain triglycerides and by those contain- ing solid fats rather than vegetable oil (52). In humans short-term experiments suggest that the obese subjects select foods with much more attention to palatability than do normal-weight individuals (53), and a preference for high-fat mixture (a bland liquid formula with high fat concentration) has been reported (54). Under carbohy-
at LO N
D O
N M
E T
R O
P O
LIT A
N U
N IV
E R
S IT
Y on M
arch 18, 2015 ajcn.nutrition.org
D ow
nloaded from
DETERMINANTS OF RELATIVE WEIGHT 411
drate balance a diet with high lipid content causes fat to accumulate until the expansion of the adipose-tissue mass ultimately raises the contribution offat to the meta- bolic fuel mix (29) by increasing free-fatty-acid levels (30, 3 1). Dietary fat can be stored as triglycerides in adipose
tissue at a metabolic cost of 3% of ingested calories; by comparison, the metabolic cost of storing dietary carbo- hydrates in fat requires 23% ofthe ingested calories (55). Furthermore, carbohydrates are preferentially oxidized and used to replenish glycogen stores and may play a spe- cial role in regulating energy expenditure and balance (56, 57). These metabolically expensive processes would have the tendency to reduce the contribution of energy from carbohydrate sources to fat deposition, whereas a high fat intake would be metabolically efficient and tend to expand the growth of the fat stores (58). Berry (32) used monounsaturated fatty acids as a marker of animal- fat intake and reported a positive correlation between this marker and body-mass index in a sample of4l 3 free- living males. This finding is consistent with the possible association between saturated fatty acids and obesity. In the Zutphen study (16), the authors reported a positive association between body-fat index and intake of all forms of fat, polysaccharides, and alcohol. However, to- tal energy intake was not accounted for in their analysis. Although the association we observed between fat intake and obesity is consistent with previous results, the inter- pretation of our findings is constrained by the cross-see- tional nature ofthe study.
Our findings demonstrate the need to control for a number of potential confounding factors, particularly physical activity and alcohol and cigarette use, in the study of diet and obesity. Furthermore, the results sug- gest that increased intake of fat may be associated with obesity independent oftotal energy intake. B
References
1 . J#{233}quierE. Long-term measurement ofenergy expenditure in man:
direct or indirect calorimetry. ln: Bjorntorp P. Cairella M, Howard AN, eds. Recent advances in obesity research III. Proceeding of
the 3rd International Congress on Obesity. London: John Libbey, 1982:130-5.
2. Webb P, Annis JF, Troutman SJ. Energy balance in man measured
by direct and indirect calorimetry. Am J Clin Nutr l980;33:1287- 98.
3. J#{233}quierE, Schutz Y. Long-term measurement of energy expendi- ture in humans using a respiration chamber. Am J Clin Nutr l983;38:989-98.
4. Schoeller DA. Energy expenditure from doubly labelled water: some fundamental considerations in humans. Am J Clin Nutr 1983;38:999-lOOS.
S. Woo R, Daniels-Kush R, Horton ES. Regulation of energy bal-
ance. Annu Rev Nutr l985;S:4l 1-33.
6. Rothwell NJ, Stock Mi. A role ofthe brown adipose tissue in diet- induced thermogenesis. Nature l979;28l:31-5.
7. Acheson KJ, Schutz Y, Bessard T, Ravussin E, J#{233}quierE. Nutri- tional influences on lipogenesis and thermogenesis after a carbohy- drate meal. Am J Physiol 1984;246:E62-70.
8. Newsholme EA. A possible metabolic basis for the control of body weight. N Engl J Med l980;302:400-S.
9. J#{233}quierE, Schutz Y. Does a defect in energy metabolism contrib-
ute to human obesity? In: Hirsch J, Van Itallie TB, eds. Recent
advances in obesity research IV. Proceeding of the 4th Interna- tional Congress on Obesity. London: John Libbey, 1985:76-81.
10. Prentice AM, Black AE, Coward WA, et al. High level of energy
expenditure in obese women. Br Med J 1986;292:983-7. 1 1. Sjostrom L. A review ofweight maintenance and weight changes
in relation to energy metabolism and body composition. In: Hirsch J, Van Itallie TB, eds. Recent advances in obesity research IV. Pro-
ceeding of the 4th International Congress on Obesity. London: John Libbey, 1985:82-94.
12. Ravussin E, Burnaud B, Schutz Y, J#{233}quierE. Twenty-four energy expenditure and resting metabolic rate in obese, moderately obese, and control subjects. Am J Can Nutr l982;35:566-73.
13. Keen H, Thomas RI, Jarrett Ri, Fuller JH. Nutrient intake, adi- posity and diabetes. Br Med J l979;l :665-8.
14. Baecke JAH, Van Staveren WA, Burema J. Food consumption,
habitual physical activity, and body fatness in young Dutch adults. Am J Gin Nutr l983;37:278-86.
IS. Braitman LE, Adlin EV, Stanton JL Jr. Obesity and caloric intake: the National Health and Nutrition Survey of 197 1-75 (HANES I).
J Chronic Dis l985;9:727-32. 16. Kromhout D. Energy and macronutrient intake in lean and obese
middle-aged men (the Zutphen study). Am J Gin Nutr l983;37:
295-9.
17. Baudoin R, Mayer J. Food intake ofobese and nonobese women.
J Am Diet Assoc l953;29:29-33. 18. Wack JJ, Rodin J. Smoking and its effect on body weight and the
systems ofcaloric regulation. Am J Clin Nutr 1982;35:366-80.
19. Kannel WB, Gordon 1. Some determinants ofobesity and its im- pact as a cardiovascular risk factor. In: Howard A, ed. Recent ad- vances in obesity research I. Proceeding of the 1st International Congress on Obesity. London: Newman, 1975:14-27.
20. Birkbeck JA. Obesity socioeconomic variables and eating habits in New Zealand. J Biosoc Sci 198 1;l 3:299-307.
21. Lincoln JE. Caloric intake obesity and physical activity. Am J Gin
Nutr 1972;25:390-4.
22. Willett W, Stampfer Mi. Total energy intake: implications for epi- demiologic analysis. Am J Epidemiol 1986;l24:l7-27.
23. Hofstetter A, Schutz Y, J#{233}quierE, Wahren J. Increased 24-hour energy expenditure in cigarette smokers. N Engl J Med l986;2:79-
82.
24. Newsholme EA. Metabolic aspects ofenzyme activity regulation. Symp Soc Exp Biol l976;41:6l-l 10.
25. Willett W, Stampfer MJ, Bain C, et al. Cigarette smoking, relative weight and menopause. Am J Epidemiol 1983;l 17:651-8.
26. Pirola RC, Lieber CS. Hypothesis: energy wastage in alcoholism and drug abuse: possible role ofhepatic microsomal enzymes. Am
J Gin Nutr l976;29:90-3. 27. Schemmel R, Mickelsen 0, Motawi K. Conversion of dietary to
body energy in rats as affected by strain, sex, and ration. J Nutr l972;l02:l 187-97.
28. Salmon DMW, Flatt JP. Effect ofdietary fat content on the mci- dence of obesity among ad libitum fed mice. mt J Obes l985;9: 443-9.
29. Flatt JP. Dietary fat, carbohydrate balance, and weight mainte-
nance: effects ofexercise. Am J Clin Nutr l987;45:296-306.
30. Golay A, Felber JP, Meyer HU, Curchod B, Maeder E, J#{233}quierE.
Study on lipid metabolism in obesity diabetes. Metabolism l984;23:l 1 1-6.
31. Bjorntorp P, Bergman H, Varnauskas E, Lindholm B. Lipid mobi- lization in relation to body composition. Metabolism l969;18: 841-51.
at LO N
D O
N M
E T
R O
P O
LIT A
N U
N IV
E R
S IT
Y on M
arch 18, 2015 ajcn.nutrition.org
D ow
nloaded from
412 ROMIEU ET AL
32. Berry EM, Hirsch J, Most J, Thorton J. The role ofdietary fat in human obesity. Int J Obes 1986; 10:123-31.
33. Sims EAH, Danforth E, Horton ES, et al. Endocrine and metabolic effects of experimental obesity in man. Recent Prog Horm Res l973;29:457-87.
34. Willett WC, Sampson L, Stampfer MJ, et al. Reproducibility and validity ofa semiquantitative food frequency questionnaire. Am J Epidemiol 1985;l22:5 1-65.
35. Paffenbarger RS Jr, Wing AL, Hyde RT. Physical activity as an index of heart attack risk in college alumni. Am J Epidemiol 1978;108:l61-75.
36. Sallis JF, Haskell WL, Wood P, et al. Physical activity assessment methodology in the five-city project. Am J Epidemiol 1985;l21: 91-106.
37. SAS Institute, Inc. SAS User’s Guide to Statistics. Cary, NC: SAS Institute, Inc, 1982.
38. Wolfe DA. On testing equality of related correlation coefficients. Biometrika 1976;63:214-5.
39. Donato K, Hegsted DM. Efficiency ofutilization of various sources ofenergy forgrowth. Proc NatI Acad Sci USA 1985;82:4866-70.
40. Roche AF, Siervogel RM, Chumlea WC, Webb P. Grading body fatness from limited anthropometric data. Am J Gin Nutr 198 l;34:283 1-8.
41. Micozzi MS, Albanes D, Jones DY, Chumlea WC. Correlations of body mass indices with weight, stature, and body composition in men and women in NHANES I and II. Am J Clin Nutr l986;44: 725-31.
42. Hutson EM, Cohen NL, Kunkel ND, Steinkamp RC, Rourke MH, Walsh HE. Measures of body fat and related factors in normal adults. J Am Diet Assoc 1965;47: 179-86.
43. Maxfield E, Konishi F. Pattern offood intake and physical activity in obesity. J Am Diet Assoc l966;49:406-8.
44. McCarthy MC. Dietary and activity patterns of obese women in Trinidad. J Am Diet Assoc 1966;48:33-7.
45. Washburn RA, Montoye HJ. The assessment of physical activity by questionnaire. Am J Epidemiol 1986;l23:563-76.
46. Kannel WB, Sorlie P. Some health benefits ofphysical activity: the Framingham study. Arch Intern Med l979;1 39:857-61.
47. Garcia PMR, Costas R, Cruzvidal M, et a!. Increased physical ac- tivity: a protective factor against heart attacks in Puerto Rico. Am
J Cardiol l982;S0:749-55. 48. Laporte RE, Montoye HJ, Caspersen Cl. Assessment of physical
activity in epidemiological research: problems and prospects. Pub- licHealthRep 1985;l00:131-45.
49. Greenland S. The effect of misclassification in the presence of co-
variates. Am J Epidemiol 1980;122:564-669.
50. Greenland 5, Robins JM. Confounding and misclassification. Am J Epidemiol l985;122:495-506.
51. Fisher M, Gordon T. The relation ofdrinking and smoking habits to diet: the Lipid Research Clinics Prevalence Study. Am J Gin Nutr 1985;41 :623-30.
52. Sclafani A. Dietary obesity. In: Bray G, ed. Recent advances in obesity research II. Proceedings ofthe 2nd International Congress on Obesity. London: Newman, 1978:123-31.
53. Nisbett RE. Taste deprivation and weight determinants of eating behavior. J Pers Soc Psychol. 1968;lO: 107-16.
54. Hashim 5, Van Itaffie T. Studies in normal and obese subjects with a monitored food-dispensing device. Ann NY Acad Sci l965;131: 654-61.
55. Raft JP. The biochemistry ofenergy expenditure. In: Bray G, ed. Recent advances in obesity research II. Proceedings ofthe 2nd In- ternational Congress on Obesity. London: Newman, 1978:211-28.
56. Acheson K!, Flatt JP, J#{233}quierE. Glycogen synthesis versus lipo- genesis after 500-g carbohydrate meal in man. Metabolism 1982;3l:l234-40.
57. FlaIl JP, Acheson KJ, Ravussin E, J#{233}quierE. Dietary fat substrate utilization and short-term body composition changes in man. mt J Vitam Nutr Res 1984;54:279(abstr).
58. Durnin JVGA, Rahaman MM. The assessment ofthe amount of fat in the human body from measurement of skinfold thickness. BrJ Nutr 1967;21:68l-9.
at LO N
D O
N M
E T
R O
P O
LIT A
N U
N IV
E R
S IT
Y on M
arch 18, 2015 ajcn.nutrition.org
D ow
nloaded from
