Children Health Issues - Library Research Assignment

Eating Patterns and Obesity in Children The Bogalusa Heart Study Theresa A. Nicklas, DrPH, LN, Su-Jau Yang, MS, Tom Baranowski, PhD, Issa Zakeri, PhD, Gerald Berenson, MD

Background: Childhood obesity is a growing public health problem. This study examined the association between eating patterns and overweight status in children who participated in the Bogalusa Heart Study.

Methods: A single 24-hour dietary recall was collected on a cross-sectional sample of 1562 children aged 10 years (65% Euro-American [EA], 35% African American [AA]) over a 21-year period. Overweight was defined as body mass index greater than the 85th percentile using Centers for Disease Control and Prevention reference standards. Multivariate logistic regression was used to investigate the association between eating patterns and overweight.

Results: Consumption of sweetened beverages (58% soft drinks, 20% fruit flavor drinks, 19% tea, and 3% coffee) (p�0.001); sweets (desserts, candy, and sweetened beverages) (p�0.001); meats (mixed meats, poultry, seafood, eggs, pork, and beef) (p�0.051); and total consumption of low-quality foods (p�0.01) were positively associated with overweight status. Total amount of food consumed, specifically from snacks, was positively associated with overweight status (p�0.05). There was a lack of congruency in the types of eating patterns associated with overweight status across four ethnic– gender groups. The percent variance explained from the eating pattern– overweight models was very small. The interaction of ethnicity and gender was significantly associated with overweight status (p�0.001). The odds of being overweight for EA males were 1.2 times higher than for AA females.

Conclusions: These results demonstrate that numerous eating patterns were associated with overweight status, yet the odds of being overweight were very small. Additional studies are needed to confirm these findings in a longitudinal sample having multiple days of assessment. (Am J Prev Med 2003;25(1):9 –16) © 2003 American Journal of Preventive Medicine

Introduction

Obesity among children has increased dramati-cally over the past 3 decades.1,2 Today anestimated one in four children in the United States is at risk of overweight (body mass index [BMI] �85th percentile), while 11% are overweight (BMI �95th percentile). Obese children tend to become obese adults.3–5 Further, obesity in early life is associ- ated with several risk factors for coronary heart dis- ease6,7 and is predictive of coronary heart disease,8

hypertension,9 and diabetes9 in adulthood. U.S. society has been increasingly characterized as

“obesogenic.”10 Although obesity has a strong genetic background,11 environmental factors are commonly considered to be the underlying cause of the increase

in obesity by promoting or exacerbating the prob- lem.11,12 The dietary causes of obesity are complex and poorly understood.13 While individual nutrients have been implicated in obesity,14 –18 few attempts have been made to identify eating patterns in this regard. Several studies have shown an association between BMI and restaurant food consumption,19 –21 soft drink consump- tion,22,23 increased portion sizes,24 meal patterns and meal frequency,25–28 diet quality,29 and diet diversity.20

However, most of these studies were conducted with adults, with very little reported on the eating pattern– obesity relationship in children. Moreover, several of the adult findings have yet to be replicated with other populations.

Secular increases in relative weight and adiposity have been documented among children in Bogalusa, Louisiana, over 2 decades.3,30,31 The prevalence of overweight among these children in 1973–1974 in- creased approximately twofold by 1994, with the largest increases observed among 19- to 24-year-olds. Further- more, the annual increases in relative weight and obesity from 1983 through 1994 were 50% greater than those between 1973 and 1982.30 During the 1970s and

From the Children’s Nutrition Research Center, Department of Pediatrics, Baylor College of Medicine (Nicklas, Yang, Baranowski, Zakeri), Houston, Texas; and Tulane Center for Cardiovascular Health, Tulane School of Public Health and Tropical Medicine (Berenson), New Orleans, Louisiana

Address correspondence to: Theresa A. Nicklas, DrPH, LN, Chil- dren’s Nutrition Research Center, Department of Pediatrics, Baylor College of Medicine, 1100 Bates Street, Houston TX 77030. E-mail: [email protected].

9Am J Prev Med 2003;25(1) 0749-3797/03/$–see front matter © 2003 American Journal of Preventive Medicine • Published by Elsevier Inc. doi:10.1016/S0749-3797(03)00098-9

1980s, the average increase in the body weight of children was 2.5 kg, without a significant increase in height. During the 1980s and 1990s, the increase was 5 kg. In this article, we assess the extent to which eating patterns (e.g., food consumption and meal patterns) are related to childhood obesity in the Bogalusa Heart Study.

Materials and Methods Population

The Bogalusa Heart Study, which began in 1973, is a long-term epidemiologic study designed to examine the early natural history of heart disease in a well- defined biracial (African-American [AA] and Euro- American [EA]) pediatric population. Dietary intake and anthropometric data were obtained on 1562 fifth- grade students (10-year-olds) in the Bogalusa, Louisi- ana, school system between 1973 and 1994, and then the seven cross-sectional surveys were combined. Sam- ple sizes varied slightly across the surveys; ethnicity (65% EA, 35% AA) and gender (51% female, 49% male) distributions were similar for each survey year of 10-year-olds and reflective of the total population.32,33

Dietary Methodology and Nutrient Database

The 24-hour dietary recall method was used in inter- viewing the children.34,35 Quality controls included: (1) a standardized protocol that specified exact tech- niques for interviewing, recording, and calculating results; (2) standardized graduated food models for quantifica- tion of foods and beverages consumed; (3) a product identification notebook for probing of snack consump- tion; (4) school lunch assessment to identify all school lunch recipes, preparation methods, and average por- tion sizes of menu items reflected in each 24-hour recall36; (5) follow-up telephone calls to parents to obtain information on brand names, recipes, and prep- aration methods of meals served at home; (6) products researched in the field to obtain updated information on their ingredients and weights, primarily snack foods, candy, and fast foods; and (7) the Moore Extended Nutrients (MENu), formerly known as the Extended Table of Nutrient Values, for nutrient composition.37

All interviewers participated in rigorous training ses- sions and pilot studies before the field surveys to minimize interviewer effects. One 24-hour dietary recall was collected on each study participant. Duplicate recalls were obtained from 10% random subsamples of each study population to assess interviewer variabili- ty.38,39 The timeframe of the 24-hour recall period included everything the child consumed from the time he/she woke up until the time of the interview and everything after the interview time on the previous day until the time the child went to bed. The same dietary- recall interview protocol was followed for all surveys.

Nutrient Database

MENu is a nutrient database that includes more than 5000 core foods and recipes, with values for 97 dietary components.37 The data bank is a flexible system permitting continuous updates of existing values and additions of new single or composite foods. Periodic updates are made to MENu to reflect nutrient changes in food products. Nutrient values were obtained from U.S. Department of Agriculture data, other published references, manufacturers’ information, and recipe cal- culation by ingredients. The database includes brand names of foods, school and family recipes, and foods commonly consumed by children.

For each survey period, the 24-hour recalls were analyzed with MENu. The version of MENu used for that analysis was saved on a tape. Analyses reflected data retrieved from stored information files specific to each time period.

Food Groups

The food-grouping scheme was designed for all foods or entries (core and recipe) appearing in MENu. Food types were identified for groups (e.g., cheese, as a major ingredient, was included in a food group list). Twenty-one major food groups were established, based on similar source characteristics (e.g., “fruit and fruit juices” formed one major group; “rice, biscuits, and cereals” were included in the breads and grains catego- ry). Composite food items, such as recipes, were as- signed to food groups according to primary ingredi- ents. If no single type of food (other than water) accounted for at least 60% of the weight, the item was classified as a mixed food. Examples of foods included in the food groups have been documented previously.40

Four food groups were deleted from the analyses due to small sample sizes (i.e., formula, vitamins, veal/lamb, alcohol), resulting in 17 food groups.

Four larger food categories were created and used in the analyses: FJV (fruit, fruit juices, vegetables); meats (mixed meats, poultry, seafood, eggs, pork, and beef); sweets (desserts, candy, and sweetened beverages); and dairy (milk and cheese). High- and low-quality food groups were also created. The high-quality food group reflected foods consumed at least once from meats, dairy, breads/grains, fruits/fruit juices, and vegetable groups. The low-quality food groups reflected foods that were consumed at least once from salty snacks, candy, desserts, fats/oils, and sweetened beverage groups.

The eating patterns selected for this study included food consumption patterns, total gram amount of food/beverages consumed by meal period, total eating episodes, number of meals and snacks consumed, and total gram amount of high- and low-quality foods. These eating patterns were selected based on an exten-

10 American Journal of Preventive Medicine, Volume 25, Number 1

sive review of the literature41 and the eating patterns that could actually be extracted from the 24-hour dietary recall.

Measure of Adiposity

Trained examiners followed rigid protocols that changed little over time.32 Briefly, height was measured twice to the nearest 0.1 cm on a standard board, and weight was measured twice to the nearest 0.1 kg by using a balance-beam metric scale. For both weight and height, the two readings were averaged. The children were clothed in only a hospital gown, underpants, and socks. BMI (kg/m2) was used as a measure of adiposity. The age- and gender-specific Centers for Disease Con- trol and Prevention (CDC) reference standards42 were used to classify children who were normal weight (BMI �85th percentile); at risk of overweight (BMI �85th and �95th percentile); and overweight (BMI �95th percentile).

Statistical Analysis Prevalence of Overweight

In the analysis, children at risk for overweight (BMI �85th percentile to �95th percentile) and those who were overweight (BMI �95th percentile) were com- bined to reflect the overweight group. The number and percentage of 10-year-olds who were overweight were identified by study year, ethnicity, gender, or ethnicity by gender groups. The Cochran–Armitage trend test was applied to examine the trend of the proportions of overweight 10-year-olds over a 21-year period. Trends in height among normal-weight and overweight 10-year- olds were examined separately by ANOVA over the same period.

Association Between Eating Patterns and Overweight Status

Data from the seven surveys were analyzed together to investigate the association between eating patterns and being overweight. The association was evaluated by logistic regression analysis via the PROC LOGISTIC procedure of SAS (version 8.0, SAS Institute Inc., Cary, North Carolina, 1999). In each logistic regression model, being overweight was used as a dependent variable and eating patterns were used as independent variables. The collinearity among independent vari- ables was checked first, using the PROC REG proce- dure of SAS with options VIF and COLLINOINT, before logistic regression was carried out. Multivariate models were conducted for overall (N�1562) as well as separately for each ethnicity– gender group (n�497 for EA male; n�513, EA female; n�273, AA male; and n�279, AA female) because ethnicity– gender interac- tion had significant effects on being overweight. Each

model for overall effects included total calorie intake, age, study year, ethnicity, gender, and ethnicity � gender interaction to control for their effects on being overweight. Likewise, the models for each ethnicity– gender group were adjusted for total calorie intake, age, and study year. An association was defined if the unity was not in the 95% confidence interval (CI) of an odds ratio (OR). The OR presented in Tables 1 and 2 were calculated depending on the type of eating pat- tern construct. If the eating pattern was measured as consumption in grams, then the OR was calculated as the ratio of odds of being overweight for participants with the mean amount of increased consumption in grams, compared to participants with mean gram con- sumption. The mean gram consumption was the aver- age amount of food/beverage consumed for each of the food groups. This average amount consumed re- flected the average “serving size” of each food group. For example, the odds of being overweight for a 10-year-old who consumed two average servings of a sweetened beverage (2�399 g) would be 1.33 times higher than a 10-year-old who consumed only an aver- age serving (399 g). If the eating pattern was not measured in grams, such as the number of eating episodes, then the OR was calculated as the ratio of odds of being overweight for a 10-year-old who had “n�1” eating episodes, compared to that for a 10-year- old with n eating episodes “(n�0).” For example, the odds of being overweight for a 10-year-old, African- American girl who had three meals was 0.56 times lower than for those who had only two meals.

Results Trends in Obesity Status

The percentage of children with a BMI in the �50th percentile significantly decreased (p�0.0001) from 55% in 1973 to 34% in 1994. In contrast, there was a twofold increase (p�0.0001) in the percentage of chil- dren with a BMI �85th and �95th percentile and a five-fold increase (p�0.0001) in the percentage of children with a BMI �95th percentile (4% to 20%) over 2 decades. Mean height (cm) significantly in- creased (p�0.05) for 10-year-old children with a BMI �50th percentile. In all survey years for children with a BMI �85th percentile, mean height was significantly (p�0.0001) higher than those children with a BMI �85th percentile.

Percentage of Overweight 10-Year-Olds by Ethnicity and Gender

All surveys were combined and divided into two weight- status groups (based on BMI): normal weight (�85th percentile), and overweight (�85th percentile) (Table

Am J Prev Med 2003;25(1) 11

3).43 The overall percentage of overweight among 10-year-olds was 24%, with 76% being normal weight. The percentage of overweight children by ethnicity (approximately 24%) and gender (approximately 24%) was equally distributed. There was a significant (p�0.0001) difference in the ethnic � gender distri- bution of overweight status. EA males had the highest prevalence of overweight (27%) and AA males had the lowest prevalence (19%). The interaction of

ethnicity and gender was significantly associated with overweight status (p�0.01). Among the four ethnici- ty– gender groups, the likelihood to be overweight for EA males was 1.2 times higher than AA females at 10 years of age (data not shown). The percentage of 10-year-olds overweight in Bogalusa (24%) was slightly higher than among 6- to 11-year-olds in the National Health and Nutrition Examination Survey (22%), partic- ularly for EA males.

Table 1. The association between eating-pattern variables and overweight status

Eating-pattern variable OR (95% CI) Mean g Example of food weight (g)

Food groups consumption Ia,b (R2�0.05) Fats/oils 0.98 (0.91–1.00) 22.0 1/8 c. oil (25.8) Fruits/fruit juices 0.97 (0.87–1.07) 135.0 1 apple (129.0) Vegetables 0.98 (0.86–1.12) 161.0 1/8 c. broccoli (23.3) Breads/grains 1.03 (0.85–1.24) 187.0 1 c. cereal (37.6) Mixed meats 1.05 (0.96–1.14) 60.0 1 slice of 12� pizza (66.4) Desserts 0.98 (0.87–1.10) 70.0 1/8 c. ice cream (16.6) Candy 0.93 (0.82–1.06) 40.0 Snickers Fun Bar (23.7) Sweetened beverages 1.33 (1.12–1.57)**** 399.0 8 oz. Coke (235.2) Poultry 1.01 (0.95–1.07) 31.0 1 fried chicken wing (32.0) Salty snacks 0.95 (0.88–1.02) 12.0 1 bag potato chips (32.0) Seafood 1.00 (0.97–1.02) 9.0 1 fish stick (6.0) Condiments 1.00 (0.95–1.05) 7.0 1 Tb mayonnaise (13.8) Eggs 1.00 (0.95–1.04) 11.0 1 medium egg (44.0) Milk 1.08 (0.87–1.33) 409.0 1 c. whole milk (244.0) Pork 1.03 (0.98–1.09) 23.0 1 slice bacon (6.3) Cheese 1.01 (0.97–1.06) 22.0 1 slice cheese (18.9) Beef 1.06 (0.99–1.15) 47.0 1 hotdog (22.0) Food groups consumption IIb,c (R2�0.05) Fats/oils 0.99 (0.92–1.06) FJV 0.96 (0.81–1.13) Breads/grains 1.05 (0.87–1.25) Meats 1.21 (1.00–1.46)* Sweets 1.38 (1.12–1.71)* Salty snacks 0.96 (0.89–1.03) Condiments 1.00 (0.95–1.05) Dairy 1.12 (0.91–1.39) Grams from high- and low-quality foods (R2�0.04) Gram of high-quality foodsd 1.19 (0.78–1.80) Gram of low-quality foodse 1.35 (1.08–1.68)* Gram amountc

Total (R2�0.04) 1.77 (1.02–3.08)* From breakfast (R2�0.04) 0.96 (0.79–1.16) From lunch (R2�0.04) 1.14 (0.87–1.49) From dinner (R2�0.04) 1.26 (0.96–1.63) From snacks (R2�0.04) 1.24 (1.02–1.50)* Eating episodef

Total (R2 � 0.04) 0.97 (0.90–1.05) No. of meals (R2 � 0.04) 0.91 (0.72–1.15) No. of snacks (R2 � 0.04) 0.98 (0.90–1.05)

*p�0.051; *p�0.05; *p�0.01; ****p�0.001. aFood group consumption I includes all individual food groups. bOdds ratio � risk of being overweight if increasing mean gram consumption. cFood group consumption II includes the four larger food group categories (FIV, dairy, meats, and sweets) and four individual food groups (fats/oils, breads/grains, salty snacks, condiments). dHigh-quality foods: fruits/fruit juices, vegetables, breads/grains, meats, dairy. eLow-quality foods: fats/oils, sweets, salty snacks. fOdds ratio � risk of being overweight if having one more eating episode. CI, confidence interval; dairy, milk and cheese; FJV, fruits/fruit juices and vegetables; g, grams; meats, mixed meats, poultry, seafood, eggs, pork, and beef; OR, odds ratio; sweets, desserts, candy, and sweetened beverages.

12 American Journal of Preventive Medicine, Volume 25, Number 1

Association Between Eating Patterns and Overweight Status Total gram amount of food/beverage consumed, par- ticularly from snacks (p�0.05), and total gram con- sumption of low-quality foods (p�0.01) were positively associated with overweight status (Table 1). Consump- tion of sweets (p�0.001), specifically sweetened bever-

ages (p�0.001), and meats (p�0.051) was positively associated with overweight status.

Despite these significant associations, the percentage of variance explained by the model was very small (Table 1). In the food group consumption model, only 5% of the variance was explained, of which sweetened beverages alone explained 1%. The combined food

Table 2. The association between eating-pattern variables and overweight status by ethnicity– gender groups

Eating pattern EA male OR (95% CI)

EA female OR (95% CI)

AA male OR (95% CI)

AA female OR (95% CI)

Food groups consumption Ia,b R2�0.08 R2�0.10 R2�0.17 R2�0.13 Fats/oils 0.97 (0.85–1.10) 1.00 (0.83–1.19) 0.93 (0.71–1.22) 1.06 (0.86–1.32) Fruits/fruit juices 1.03 (0.88–1.20) 1.10 (0.92–1.31) 0.97 (0.69–1.41) 0.55 (0.38–0.79)* Vegetables 0.98 (0.77–1.24) 1.09 (0.87–1.36) 1.05 (0.74–1.49) 0.75 (0.51–1.09) Breads/grains 1.20 (0.86–1.67) 0.90 (0.62–1.30) 0.62 (0.33–1.16) 1.03 (0.60–1.79) Mixed meats 1.12 (0.95–1.31) 0.93 (0.78–1.12) 1.06 (0.82–1.37) 0.97 (0.78–1.19) Desserts 0.89 (0.73–1.09) 1.08 (0.86–1.35) 0.89 (0.65–1.22) 0.89 (0.66–1.21) Candy 0.94 (0.76–1.18) 0.78 (0.60–1.01) 0.79 (0.51–1.23) 1.00 (0.73–1.35) Sweetened beverages 1.68 (1.21–2.33)* 1.53 (1.05–2.22)* 1.02 (0.72–1.46) 0.92 (0.65–1.30) Poultry 0.99 (0.89–1.09) 1.04 (0.94–1.16) 0.97 (0.76–1.23) 0.99 (0.84–1.16) Salty snacks 0.98 (0.88–1.09) 0.92 (0.80–1.05) 1.15 (0.94–1.42) 0.84 (0.66–1.06) Seafood 0.97 (0.92–1.02) 1.07 (1.01–1.13)* 0.73 (0.48–1.11) 1.03 (0.95–1.11) Condiments 1.02 (0.93–1.12) 0.99 (0.90–1.08) 1.02 (0.86–1.22) 0.89 (0.74–1.07) Eggs 0.97 (0.88–1.06) 1.07 (0.99–1.17) 0.99 (0.85–1.14) 0.91 (0.79–1.05) Milk 0.96 (0.64–1.46) 1.19 (0.82–1.73) 1.25 (0.68–2.30) 0.93 (0.55–1.59) Pork 1.04 (0.94–1.14) 1.00 (0.89–1.12) 1.16 (0.99–1.35) 0.98 (0.84–1.13) Cheese 1.02 (0.95–1.11) 1.04 (0.97–1.13) 0.95 (0.82–1.09) 0.90 (0.77–1.06) Beef 1.08 (0.92–1.25) 1.06 (0.91–1.24) 1.11 (0.89–1.37) 1.02 (0.88–1.18) Food groups consumption IIb,c R2�0.05 R2�0.08 R2�0.12 R2�0.11 Fats/oils 0.98 (0.86–1.11) 1.00 (0.84–1.19) 0.94 (0.75–1.17) 1.08 (0.88–1.32) FJVd 1.05 (0.81–1.36) 1.23 (0.93–1.62) 1.05 (0.64–1.75) 0.43 (0.26–0.73)* Breads/grains 1.25 (0.92–1.70) 0.93 (0.66–1.31) 0.70 (0.41–1.19) 1.15 (0.48–2.74) Meatsd 1.26 (0.89–1.78) 1.25 (0.88–1.78) 1.49 (0.87–2.56) 0.98 (0.63–1.52) Sweetsd 1.65 (1.12–2.44)* 1.65 (1.06–2.57)* 1.01 (0.63–1.59) 0.93 (0.59–1.46) Salty snacks 0.99 (0.89–1.10) 0.91 (0.80–1.05) 1.16 (0.96–1.42) 0.85 (0.67–1.07) Condiments 1.01 (0.92–1.11) 1.00 (0.92–1.08) 0.99 (0.84–1.18) 0.93 (0.79–1.10) Dairyd 1.08 (0.73–1.62) 1.30 (0.90–1.89) 1.27 (0.72–2.26) 0.91 (0.55–1.53) Gram from high-and

low-quality foods R2�0.05 R2 � 0.07 R2�0.09 R2�0.06

High-quality foodsd 1.44 (0.70–2.95) 1.89 (0.90–3.96) 1.06 (0.34–3.33) 0.34 (0.12–1.04) Low-quality foodse 1.69 (1.12–2.53)* 1.60 (1.01–2.53)* 1.00 (0.62–1.63) 0.89 (0.56–1.42) Gram amount2 R2�0.05 R2�0.07 R2�0.09 R2�0.06 Total 3.17 (1.20–8.41)*

R2�0.05 2.97 (1.03–8.52)* R2�0.07

1.03 (0.25–4.30) R2�0.11

0.33 (0.09–1.22) R2�0.07

From breakfast 1.14 (0.81–1.60) 1.18 (0.83–1.71) 0.67 (0.40–1.10) 0.62 (0.42–0.93)* From lunch 1.20 (0.75–1.93) 1.38 (0.86–2.23) 1.03 (0.48–2.23) 0.91 (0.47–1.77) From dinner 1.82 (1.13–2.91)* 1.24 (0.72–2.14) 0.81 (0.43–1.54) 0.95 (0.54–1.67) From snacks 1.39 (1.00–1.95) 1.43 (0.96–2.14) 1.34 (0.81–2.20) 0.68 (0.41–1.13) Eating episodef R2�0.04 R2�0.06 R2�0.09 R2�0.06 Total 0.90 (0.79–1.03) 1.07 (0.93–1.24) 1.04 (0.85–1.28) 0.91 (0.76–1.08)

R2�0.04 R2�0.06 R2�01.4 R2�0.07 No. of meals 0.97 (0.63–1.50) 1.23 (0.80–1.89) 0.70 (0.38–1.33) 0.56 (0.33–0.95)* No. of snacks 0.90 (0.78–1.03) 1.06 (0.91–1.23) 1.07 (0.87–1.31) 0.93 (0.78–1.12)

*p�0.05; *p�0.01. aFood group consumption I: individual food group consumption as eating pattern variables. bOdds ratio � risk of being overweight if increasing mean gram consumption. cFood group consumption II includes the four larger food group categories (FIV, dairy, meats, sweets) and four individual food groups (fats/oils, breads/grains, salty snacks, condiments). dHigh-quality foods: fruits/fruit juices, vegetables, breads/grains, meats, dairy. eLow-quality foods: fats/oils, sweets, salty snacks. fOdds ratio � risk of being overweight if having one more eating episode. AA, African American; CI, confidence interval; EA, Euro-American; FJV, fruits/fruit juices and vegetables; meats, mixed meats, poultry, seafood, eggs, pork, and beef; OR, odds ratio; sweets, desserts, candy, and sweetened beverages; dairy, milk and cheese.

Am J Prev Med 2003;25(1) 13

ategories of meats and sweets (1%) explained very little of the variance in obesity. Similarly, the percentage of vari- ance explained by total gram amount of low-quality foods consumed and the total gram amount of foods/beverages consumed (particularly from lunch and dinner), the number of snacking episodes was equally low.

Association Between Eating Patterns and Overweight Status by Ethnicity and Gender

The association of eating patterns and obesity varied by ethnicity and gender (Table 2). For EA males, con- sumption of sweets (p�0.05), especially sweetened bev- erages (p�0.01); total gram amount of food/beverages consumed (p�0.05), particularly from the dinner meal (p�0.05); and total gram amount of low-quality foods consumed (p�0.05) were positively associated with obesity. For EA females, consumption of seafood (p�0.05) and sweets (p�0.05), particularly sweetened beverages (p�0.05), were positively associated with obesity. Total gram amount of food consumed (p�0.05), particularly low-quality foods (p �0.05), was positively associated with obesity. For AA females, con- sumption of fruits/fruit juices (p �0.01); FJV (p�0.01); total gram amount of breakfast consumed (p�0.05); and total number of meals consumed (p�0.05) were negatively associated with obesity. Overall, the percent of variance explained (range 4% to 8%) by the model was low for all ethnic � gender groups.

Discussion

The present investigation adds to previous studies, using dietary intake data to establish the association of eating patterns with overweight status in 10-year-old children. The prevalence of overweight among the children was 24%, which is comparable to the national average,1,2 with the highest prevalence shown among

EA males. Overall, several eating patterns were posi- tively associated with overweight status: consumption of sweetened beverages, sweets, and meats, and total gram consumption of low-quality foods. Total gram amount of food/beverages consumed, particularly from snacks, was also positively associated. Other studies have shown an association between soft drink consumption,22,23

low-quality foods,29 and overweight status, while other studies have not.44,45 In the present study, soft drink consumption accounted for 50% of the sweetened beverages consumed, with the remaining percentage reflecting consumption of tea or coffee with sugar and fruit drinks. When only soft drink consumption is considered, the positive association with overweight status still existed but the percent of variance explained was very low (1%). The positive association between sweets and overweight status resulted from consump- tion of sweetened beverages and not desserts or candy.

The association between total gram amount of bev- erage/food consumed and overweight status was not surprising, particularly since the amount of time chil- dren are spending in physical activity has de- creased.46,47 A previous study of these same Bogalusa children showed that the mean total gram amount of food/beverages consumed increased from 1973 to 1994 (first author, unpublished observations, 2003). If phys- ical activity is decreasing48 –51 and the total gram amount of food consumed has increased, it is reason- able to find an association between total gram amount consumed and overweight status.

A particularly interesting finding was the lack of congruency in the types of eating patterns associated with overweight status across the four ethnic– gender groups. Data suggest that the eating patterns associated with obesity may in fact vary by ethnicity and gender. This initial finding needs to be confirmed in other

Table 3. Percentage of overweight 10-year-olds by ethnicity and gender

Weight status Bogalusa NHANES II and NHES III

Normal n (%)

Overweight n (%)

Overweight n (%)

Ethnicity Euro-American 768 (76) 242 (24) — African American 427 (77) 125 (23) —

Gender Male 584 (76) 186 (24) 467 (23) Female 611 (77) 181 (23) 458 (21)

Ethnicity and gendera

Euro-American, male 364 (73) 133 (27) 267 (21) Euro-American, female 404 (79) 109 (21) 270 (22) African American, male 220 (81) 53 (19) 257 (27) African American, female 207 (74) 72 (26) 224 (31)

Total 1195 (76) 367 (24) 1817 (22) a�2�7.75; p�0.0515. NHANES II, National Health and Nutrition Examination Survey43; NHES, National Health Examination Survey.43

14 American Journal of Preventive Medicine, Volume 25, Number 1

national surveys with a larger geographic representa- tion and larger sample size.

An important consideration when interpreting the results is that the percent variance explained from the eating pattern– overweight models was extremely small, suggesting that substantial variance in obesity was un- explained. Since the eating patterns studied were not mutually exclusive, it was impossible to determine the cumulative effect of the significant eating patterns on overweight status. One can hypothesize that the associ- ation between eating patterns and overweight status is not a result of a single eating pattern but from a combination of eating patterns that are interrelated and cumulative in their effect on overweight status.

Limitations

The present study was a cross-sectional analysis, and thus causal inferences cannot be made. Longitudinal studies are needed to confirm these preliminary find- ings. A single 24-hour dietary recall was collected on each participant. One 24-hour dietary recall is not adequate for characterizing the usual eating patterns of an individual, but is sufficient for charactering the eating patterns of large groups of children.52 Despite this limitation, several eating pattern correlates of obesity in 10-year-old children were detected. The sample size in this study was 1562 children, which is larger than most regionally specific studies but smaller than national surveys. These findings are specific to Bogalusa 10-year-old children and may not be represen- tative of the nation as a whole. However, the dietary intakes of Bogalusa children are comparable to na- tional surveys.53–55 The overweight group defined in this study was those children at risk for overweight or already overweight. Although this approach has been used in other published studies, the findings need to be interpreted with some caution and replicated in larger studies. The percentage of children who were over- weight in this study (11%) is comparable to national averages (11%).56 Another major limitation is the lack of physical activity data on this cohort of children. Others have found an association between physical activity and overweight status.57–59 Energy expenditure from physical activity directly influences the overall energy balance equation.58 This may explain, in part, the low percentage of variance of the single eating pattern variables investigated in this study.

Application and Implications

These results have important implications for obesity- prevention research targeting children. These associa- tions were poorly explained by a single eating pattern, and the pattern of significant relationships varied by ethnicity and gender. Further research with multiple days of assessment is needed to better understand the

associations among eating patterns and overweight status among children.

We are grateful to Margaret Moore for development and Catherine Champagne, PhD, for maintenance of the Moore Extended Nutrients (MENu) database and Pamelia Harris for help in preparing the manuscript. We also extend a special thanks to the children and young adults of Bogalusa without whom this work could not be accomplished.

This research was supported by the National Institutes of Health, Evolution of Cardiovascular Risk with Normal Aging National Institute on Aging (AG 16592) and the U.S. Depart- ment of Agriculture, Food Assistance and Nutrition Research Program. Partial support was received from the Kellogg’s Company, the Sugar Association, and MARS Inc.

This work is a product of the U.S. Department of Agricul- ture (USDA/ARS) Children’s Nutrition Research Center, Department of Pediatrics, Baylor College of Medicine, Hous- ton TX, and had been funded in part with federal funds from the USDA/ARS (Cooperative Agreement 58-6250-6001). The contents of this publication do not necessarily reflect the views or policies of the USDA, nor does mention of trade names, commercial products, or organizations imply endorse- ment from the U.S. government.

References 1. Flegal KM. The obesity epidemic in children and adults: current evidence

and research issues. Med Sci Sports Exerc 1999;31:509 –14. 2. Rippe J. The obesity epidemic: challenges and opportunities. J Am Diet

Assn 1998;98:5. 3. Webber LS, Wattigney WA, Srinivasan SR, Berenson GS. Obesity studies in

Bogalusa. Am J Med Sci 1995;310(suppl):S53–61. 4. Guo SS, Roche AF, Chumlea WC, Gardner JD, Siervogel RM. The predic-

tive value of childhood body mass index values for overweight at age 35 y. Am J Clin Nutr 1994;59:810 –9.

5. Webber LS, Cresanta JL, Croft JB, Srinivasan SR, Berenson GS. Transitions of cardiovascular risk from adolescence to young adulthood—the Bogalusa Heart Study: II. Alterations in anthropometric blood pressure and serum lipoprotein variables. J Chronic Dis 1986;39:91–103.

6. Johnston FE. Health implications of childhood obesity. Ann Intern Med 1985;103:1068 –72.

7. Aristimuno GG, Foster TA, Voors AW, Srinivasan SR, Berenson GS. Influence of persistent obesity in children on cardiovascular risk factors: the Bogalusa Heart Study. Circulation 1984;69:895–904.

8. Hoffmans MD, Kromhout D, Coulander CD. Body mass index at the age of 18 and its effects on 32-year mortality from coronary heart disease and cancer. A nested case-control study among the entire 1932 Dutch male birth cohort. J Clin Epidemiol 1989;42:513–20.

9. Abraham S, Collins G, Nordsieck M. Relationship of childhood weight status to morbidity in adults. HSMHA Rep 1971;86:273–84.

10. Swinburn B, Egger G, Raza F. Dissecting obesogenic environments: the development and application of a framework for identifying and prioritiz- ing environmental interventions for obesity. Prev Med 1999;29:563–70.

11. Hebebrand J, Wulftange H, Goerg T, et al. Epidemic obesity: are genetic factors involved via increased rates of assortative mating? Int J Obes Relat Metab Disord 2000;24:345–53.

12. Hill JO, Melanson EL, Wyatt HT. Dietary fat intake and regulation of energy balance: implications for obesity. J Nutr 2000;130(suppl):S284 –88.

13. U.S. Department of Agriculture. Childhood obesity: causes and prevention. Proceedings of a symposium sponsored by the Center for Nutrition Policy and Promotion. Washington DC, October 27, 1998. Available at: www.nal. usda.gov:8001/Training/obesity.pdf. Accessed July 12, 2002.

14. Zemel MB, Shi H, Greer B, Dirienzo D, Zemel PC. Regulation of adiposity by dietary calcium. FASEB J 2000;14:1132–8.

15. Roberts SB, McCrory MA, Saltzman E. The influence of dietary composi- tion on energy intake and body weight. J Am Coll Nutr 2002;21(suppl): S140 –45.

Am J Prev Med 2003;25(1) 15

16. Gibson SA. Are high-fat, high-sugar foods and diets conducive to obesity? Int J Food Sci Nutr 1996;47:405–15.

17. Macdiarmid JI, Vail A, Cade JE, Blundell JE. The sugar–fat relationship revisited: differences in consumption between men and women of varying BMI. Int J Obes Relat Metab Disord 1998;22:1053–61.

18. Bray GA, Popkin BM. Dietary fat intake does affect obesity! Am J Clin Nutr 1998;68:1157–73.

19. Clemens LH, Slawson DL, Klesges RC. The effect of eating out on quality of diet in premenopausal women. J Am Diet Assn 1999;99:442–4.

20. McCrory MA, Fuss PJ, McCallum JE, et al. Dietary variety within food groups: association with energy intake and body fatness in men and women. Am J Clin Nutr 1999;69:440 –7.

21. Coll M, Meyer A, Stunkard AJ. Obesity and food choices in public places. Arch Gen Psychiatry 1979;36:795–7.

22. Ludwig DS, Peterson KE, Gortmaker SL. Relation between consumption of sugar-sweetened drinks and childhood obesity: a prospective, observational analysis. Lancet 2001;357:505–8.

23. Dennison BA, Rockwell HL, Baker SL. Excess fruit juice consumption by preschool-aged children is associated with short stature and obesity. Pediatrics 1997;99:15–22.

24. Hill JO, Peters JC. Environmental contributions to the obesity epidemic. Science 1998;280:1371–4.

25. Ceru-Bjork C, Andersson I, Rossner S. Night eating and nocturnal eating— two different or similar syndromes among obese patients? Int J Obes Relat Metab Disord 2001;25:365–72.

26. Speechly DP, Rogers GG, Buffenstein R. Acute appetite reduction associ- ated with an increased frequency of eating in obese males. Int J Obes 1999;23:1151–9.

27. Fabry P, Fodo J, Hejl Z, Braun T, Zvolankova K. The frequency of meals: its relation to overweight, hypercholesterolaemia, and decreased glucose tolerance. Lancet 1964;2:614 –5.

28. Kant A. Frequency of eating occasions and weight change in NHANES I Epidemiologic Follow-up Study. Int J Obes 1995;19:468 –74.

29. Basiotis PP, Lino M, Anand RS. Report card on the diet quality of African Americans. Fam Econ Nutr Rev 1998;11:61–3.

30. Freedman DS, Srinivasan SR, Valdez RA, et al. Secular increases in relative weight and adiposity among children over two decades: the Bogalusa Heart Study. Pediatrics 1997;99:420 –6.

31. Gidding SS, Bao W, Srinivasan SR, Berenson GS. Effects of secular trends in obesity on coronary risk factors in children: the Bogalusa Heart Study. J Pediatr 1995;127:868 –74.

32. Berenson GS. Cardiovascular risk factors in children: the early natural history of atherosclerosis and essential hypertension. New York: Oxford University Press, 1980.

33. Harsha DW. Body composition and obesity in a population of children and their influence on cardiovascular risk. In: Berenson GS, ed. Causation of cardiovascular risk factors in children: perspectives on cardiovascular risk in early life. New York: Raven Press, 1986:190 –222.

34. Frank GC, Berenson GS, Schilling PE, Moore MC. Adapting the 24-hr. recall for epidemiologic studies of school children. J Am Diet Assn 1977;71:26 –31.

35. Farris RP, Nicklas TA. Characterizing children’s eating behavior. In: Suskind RM, Suskind LL, eds. Textbook of pediatric nutrition. 2nd ed. New York: Raven Press, 1993;505–16.

36. Nicklas TA, Forcier JE, Webber LS, Berenson GS. School lunch assessment to improve accuracy of 24-hour dietary recall for children. J Am Diet Assn 1991;91:711–3.

37. Moore MC, Goodloe MH. Extended table of nutrient values (ETNV), 2nd rev. ext. ed. New Orleans LA: International Dietary Information Founda- tion, Louisiana State University Medical Center, 1982.

38. Farris RP, Frank GC, Webber LS, Berenson GS. Quality controls for processing dietary data. Diet J Software 1984;2:3–4.

39. Frank GC, Hollatz AT, Webber LS, Berenson GS. Effect of interviewer recording practices on nutrient intake—Bogalusa Heart Study. J Am Diet Assn 1984;84:1432–36, 1439.

40. Nicklas TA, Farris RP, Johnson CC, Webber LS, Berenson GS. Food sources of nutrients: a tool for dietary management and health. The Bogalusa Heart Study. 1973–1983. Tulane Center for Cardiovascular Health, New Orleans LA, 1990.

41. Nicklas TA, Baranowski T, Cullen KW, Berenson G. Eating patterns, dietary quality, and obesity. J Am Coll Nutr 2001;20:598 –608.

42. National Center for Health Statistics. National Health and Nutrition Examination Survey, CDC growth charts: United States. Available at: www.cdc.gov/nchs/about/major/nhanes/growthcharts/datafiles.htm. Ac- cessed on July 12, 2002.

43. Troiano RP, Flegal KM, Kuczmarski RJ, Campbell SM, Johnson CL. Overweight prevalence and trends for children and adolescents: the National Health and Nutrition Examination Surveys, 1963–1991. Arch Pediatr Adolesc Med 1995;149:1085–91.

44. Bouchard C. Can obesity be prevented? Nutr Rev 1996;54(suppl):S125– 130.

45. Seidell JC. Dietary fat and obesity: an epidemiologic perspective. Am J Clin Nutr 1998;67:546 –50.

46. Pratt M, Macera CA, Blanton C. Levels of physical activity and inactivity in children and adults in the United States: current evidence and research issues. Med Sci Sports Exerc 1999;31:526 –33.

47. Kann L, Kinchen SA, Williams BI, et al. State and local YRBSS coordinators. Youth Risk Behavior Surveillance—United States. MMWR Surveill Summ 1998;47:1–89.

48. Porkka KV, Raitakari OT, Leino A, et al. Trends in serum lipid levels during 1980 –1992 in children and young adults. The Cardiovascular Risk in Young Finns Study. Am J Epidemiol 1997;146:64 –77.

49. Kann L, Warren CW, Harris WA, et al. Youth risk behavior surveillance— United States, 1993. J Sch Health 1995;65:163–71.

50. Grunbaum JA, Kann L, Kinchen SA, et al. Youth risk behavior surveil- lance—United States, 2001. J Sch Health 2002;72:313–28.

51. Lowry R, Wechsler H, Kann L, Collins JL. Recent trends in participation in physical education among US high school students. J Sch Health 2001;71: 145–52.

52. Thompson FE, Byers T. Dietary assessment resource manual. J Nutr 1994;124(suppl):S2245–317.

53. Nicklas TA. Dietary studies of children and young adults (1973–1988): the Bogalusa Heart Study. Am J Med Sci 1995;310(suppl):S101–8.

54. Nicklas TA, Farris RP, Srinivasan SR, Webber LS, Berenson GS. Nutritional studies in children and implications for change: the Bogalusa Heart Study. J Adv Nurs 1989;2:451–74.

55. Nicklas TA, Johnson CC, Myers L, Webber L, Berenson GS. Eating patterns, nutrient intakes and alcohol consumption patterns of young adults: the Bogalusa Heart Study. Med Exerc Nutr Health 1995;4:316 –24.

56. Ogden CL, Flegal KM, Carroll MD, Johnson CL. Prevalence and trends in overweight among children and adolescents, 1999 –2000. JAMA 2002;288: 1728 –32.

57. National Center for Chronic Disease Prevention and Health Promotion. Physical activity and health: a report of the Surgeon General. Atlanta, GA: Centers for Disease Control and Prevention, 1996.

58. Barlow SE, Dietz WH. Obesity evaluation and treatment: expert committee recommendations. The Maternal and Child Health Bureau, Health Re- sources and Services Administration and the Department of Health and Human Services. Pediatrics 1998;102:E29.

59. Dietz W. Physical activity and childhood obesity. Nutrition 1991;7:295–6.

16 American Journal of Preventive Medicine, Volume 25, Number 1