Cross-sectional Study Article

J

RESEARCH

Original Research: Brief

Sociodemographic and Behavioral Factors Associated with Added Sugars Intake among US Adults

Sohyun Park, PhD; Frances E. Thompson, PhD, MPH; Lisa C. McGuire, PhD; Liping Pan, MD, MHP; Deborah A. Galuska, PhD; Heidi M. Blanck, PhD

ARTICLE INFORMATION

Article history: Submitted 24 November 2015 Accepted 19 April 2016 Available online 26 May 2016

Keywords: Added sugars Sugar-sweetened beverages Sweet food Sociodemographic characteristics Behaviors

2212-2672/Published by Elsevier Inc. on behalf of the Academy of Nutrition and Dietetics. http://dx.doi.org/10.1016/j.jand.2016.04.012

ABSTRACT Background Reducing added sugars intake is one of the Healthy People 2020 objec- tives. High added sugars intake may be associated with adverse health consequences. Objective This cross-sectional study identified sociodemographic and behavioral characteristics associated with added sugars intake among US adults (18 years and older) using the 2010 National Health Interview Survey data (n¼24,967). Methods The outcome variable was added sugars intake from foods and beverages using scoring algorithms to convert dietary screener frequency responses on nine items to estimates of individual dietary intake of added sugars in teaspoons per day. Added sugars intake was categorized into tertiles (lowest, middle, highest) stratified by sex. The explanatory variables were sociodemographic and behavioral characteristics. Multino- mial logistic regression was used to estimate the adjusted odds ratios for the highest and middle tertile added sugars intake groups as compared with the lowest tertile group. Results Estimated median added sugars intake was 17.6 tsp/d for men and 11.7 tsp/d for women. For men and women, those who had significantly greater odds for being in the highest tertile of added sugars intake (men: �22.0 tsp/d; women: �14.6 tsp/d) were younger, less educated, had lower income, were less physically active, were current smokers, and were former or current infrequent/light drinkers, whereas non-Hispanic other/multiracial and those living in the West had significantly lower odds for being in the highest tertile of added sugars intake. Different patterns were found by sex. Non- Hispanic black men had lower odds for being in the highest tertile of added sugars intake, whereas non-Hispanic black women had greater odds for being in the highest tertile. Conclusions One in three men consumed �22.0 tsp added sugars and one in three women consumed �14.6 tsp added sugars daily. Higher added sugars intake was associated with various sociodemographic and behavioral characteristics; this infor- mation can inform efforts to design programs and policies specific to high-intake populations. J Acad Nutr Diet. 2016;116:1589-1598.

T HE 2015-2020 DIETARY GUIDELINES FOR AMERICANS defines added sugars as “syrups and other caloric sweeteners used as a sweetener in other food prod- ucts. Specific examples of added sugars that can be

listed as an ingredient include brown sugar, corn sweetener, corn syrup, dextrose, fructose, glucose, high-fructose corn syrup, honey, invert sugar, lactose, malt syrup, maltose, molasses, raw sugar, sucrose, trehalose, and turbinado sugar.”1 Naturally occurring sugars, such as fructose in fruits and lactose in milk and dairy products, are not considered added sugars.1 During food processing or preparation, added sugars are incorporated into foods and beverages to improve palatability and to extend the shelf life of products.1

The upper limits of calories from added sugars intake recommended by the American Heart Association are

<150 kcal/d (approximately 9 tsp) for most American men and <100 kcal/d (approximately 6 tsp) for most American women.2 The 2015-2020 Dietary Guidelines for Americans recommends that Americans should limit their consumption of added sugars and that they consume <10% of total calories from added sugars per day.1 In the 2005-2010 National Health and Nutrition Examination Surveys (NHANES), the mean percentage of total calories from added sugars was 12.7% (335 kcal/d) for men and 13.2% (239 kcal/d) for women.3 Non-Hispanic blacks and adults with lower incomes consumed a larger percentage of total calories from added sugars compared with other adults in the population.3 In addition, 67% of calories from added sugars came from foods and 33% came from beverages among US adults.3 Many studies have shown that high consumption of added sugars

OURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS 1589

RESEARCH

(eg, >10% of total daily calories from added sugars) is asso- ciated with adverse health outcomes in adults, including obesity/weight gain,4,5 type 2 diabetes,6 cardiovascular dis- ease,7,8 and periodontal disease.9

Although a few studies have investigated the associations between added sugars and sociodemographic characteristics among US adults, these studies did not examine behavioral factors related to added sugars intake.3,10 Certain unhealthy nondietary behaviors might be associated with unhealthy dietary behaviors.11,12 Identifying behavioral characteristics that coexist with high added sugars intake is important because this information can be used to inform intervention efforts to reach high-risk groups. The objective of this cross- sectional analysis was to examine the relationship between sociodemographic and behavioral factors and total added sugars intake for men and women using a nationally repre- sentative sample of civilian noninstitutionalized US adults. In addition, the relative contributions of foods and beverages to total added sugars intake are described for subgroups defined by demographic and behavioral variables.

METHODS Study Sample and Survey Administration Since 1957, the Centers for Disease Control and Prevention’s National Center for Health Statistics has conducted annually the National Health Interview Survey (NHIS). The NHIS uses a multistage sampling design to survey a representative sample of civilian noninstitutionalized US households. This cross- sectional analysis used data from the 2010 NHIS.13 Informa- tion on health and other characteristics of individual family members within the household was gathered using in- person interviews, and supplementary data were obtained from one randomly selected child (“Sample Child”) or one randomly selected adult (“Sample Adult”) from the house- hold. The Sample Adult Module collected information on sociodemographic characteristics, occupations, certain med- ical conditions, illness, health status, functional limitations, health behaviors, and health care access and utilization. The Cancer Control Supplement, funded by the National Cancer Institute (NCI) and Centers for Disease Control and Preven- tion, collects additional behavioral information on sample adults (aged 18 years and older) and is administered every 5 years. The 2010 Cancer Control Supplement included a 26-item Dietary Screener Questionnaire, which was admin- istered via computer-assisted personal interviewing. Adults reported the number of times per day, per week, or per month during the past month that they consumed selected foods and beverages.13,14 The 2010 NHIS was approved by the Research Ethics Review Board at the National Center for Health Statistics. The advance letter that is sent to potential respondents of NHIS before the interview is the consent statement. Each year, the Division of Health Interview Sta- tistics, which conducts the NHIS, requests and is granted a waiver of documentation of signed informed consent for the NHIS.

Outcome Variables Total calorie intake was not assessed in the NHIS and so the outcome variable was total added sugars intake estimated from reported frequency of foods and beverages during the past month by using NCI’s scoring algorithms. These

1590 JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS

algorithms convert dietary screener frequency responses on nine items (ie, regular soda, coffee/tea drinks, sport and en- ergy drink, sweetened fruit drinks, chocolate/candy, dough- nuts/sweet rolls/Danish/muffins/toaster pastries, cookies/ cake/pie/brownies, ice cream/frozen desserts, and hot/cold cereals) to estimates of individual dietary intake of added sugars in teaspoons per day.10,13,15 A separate variable for sugar-sweetened beverages (SSB) was created from the four beverage intake items. For the development of the scoring algorithms, NCI used 24-hour dietary recall data from the 2003-2006 NHANES. More information on the scoring algo- rithms can be found at the NCI website.15 Because estimated total added sugars intake was not normally distributed, me- dian intake values for total added sugars, added sugars from foods, and added sugars from SSBs are presented. As a cate- gorical variable, three groups were created based on tertile distributions of total added sugars intake separately for men and women, because previous studies suggested a dose� response of added sugars intake on health outcomes, such as cardiovascular disease mortality7 and poor oral health.9

Explanatory Variables The explanatory variables were sociodemographic and behavioral characteristics. Mutually exclusive response cate- gories for each explanatory variable were created. Socio- demographic variables included age (18 to 24, 25 to 39, 40 to 59, or 60 years and older), race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, or non-Hispanic other/ multiracial), marital status (currently married/domestic partnership or not married), education (less than high school graduate, high school graduate or recipient of a general education development certificate, some college, or college graduate), annual family income (<$35,000, $35,000 to $74,999, $75,000 to $99,999, or �$100,000), and census re- gion (Northeast, Midwest, South, and West). Behavioral variables included reported participation in

leisure-time physical activity for at least 10 minutes (0, >0 to <3, 3 to 5, or >5 times/wk), smoking status (never smokers, former smokers, or current smokers), and alcohol consump- tion (lifetime abstainer, former drinker, current infrequent/ light drinker, or current moderate/heavier drinker).16 Lifetime abstainers were defined as <12 drinks in lifetime; former drinkers were defined as �12 drinks in lifetime but none in past year; current infrequent/light drinkers were defined as �12 drinks in lifetime and �3 drinks/wk in past year; and current moderate/heavier drinkers were defined as �12 drinks in lifetime and >3 drinks/wk in past year. Using self- reported weight and height data, weight status was catego- rized as underweight/normal weight (body mass index [BMI; calculated as kg/m2] <25), overweight (BMI 25 to <30), or obese (BMI �30).17

Statistical Analysis In the 2010 NHIS, 27,157 adults aged 18 years and older completed the Sample Adult Module and the Cancer Control Supplement (response rate for the Sample Adult Module: 60.8%). Of these, a total of 2,190 (8%) were excluded from the study because of missing data on any of the nine foods used to estimate added sugars intake (n¼2,073) or having extreme added sugars intake values (defined as >75th percentile valueþ2 times the interquartile range using the cube-root

October 2016 Volume 116 Number 10

RESEARCH

transformed variable) (n¼117).10 Adults with missing re- sponses on covariates were excluded from analyses when that variable was used in the analysis. The final analytic sample consisted of 24,967 adults (10,962 men and 14,005 women). When comparing the analytic sample and the sample of respondents who were excluded, there were no differences in age. However, the analytic sample contained a slightly higher proportion of females and non-Hispanic whites (c2 test, P<0.05) than the excluded sample. All analyses were stratified by sex. c2 tests were used to

examine the bivariate associations of sociodemographic and behavioral characteristics with total added sugars intake and P<0.05 to evaluate statistical significance. Multinomial logistic regression analysis was used to calculate adjusted odds ratio (aOR) and 95% CI for variables associated with the highest and middle tertiles of added sugars intake groups as compared with the lowest tertile of added sugars intake group. The multinomial logistic regression models included all explanatory variables in one model. Added sugars intake from foods was calculated by subtracting added sugars from SSB from total added sugars. SAS software (version 9.3, 2011, SAS Institute Inc) was used to perform all statistical analyses and account for the complex sampling design.

RESULTS The study population distributions by sociodemographic and behavioral characteristics are shown in Table 1. Overall, 56% were women, 38% were aged 18 to 39 years, and 68% were non-Hispanic whites. The majority of men and women were in a married/domestic partnership, and about 60% of men and women had at least some college education. Based on c2

tests, all sociodemographic and behavioral characteristics analyzed were significantly associated with the tertile dis- tribution of estimated added sugars intake for both men and women (P�0.01, Table 1). For both men and women, those who had significantly

greater odds for being in the highest tertile of added sugars intake (men: �22.0 tsp/d; women: �14.6 tsp/d) were younger (aOR range¼2.52 to 18.52 for 18 to 59 years vs 60 years and older), less educated (aOR range¼1.36 to 2.06 for less than high school, high school, or some college vs college graduate or beyond), had lower family income (aOR range¼1.22 to 1.33; $35,000 to $74,999 for men and <$75,000 for women vs �$100,000), reported no physical activity (aOR¼1.22 for men, aOR¼1.64 for women vs >5 times/wk), were current smokers (aOR¼1.96 for men, aOR¼2.29 for women vs nonsmokers), and were former or current infrequent/light drinkers (aOR range¼1.24 to 1.61 vs lifetime abstainers) (Table 2). For both men and women, lower odds for being in the

highest tertile of added sugars intake was significantly associated with being non-Hispanic other/multiracial (aOR¼0.55 for men and aOR¼0.78 for women vs non- Hispanic white) and living in the West (aOR¼0.66 for men and aOR¼0.64 for women vs South). For men, lower odds for being in the highest tertile of added sugars intake (�22.0 tsp/d) was also associated with being non-Hispanic black (aOR¼0.80 vs non-Hispanic white) and obese (aOR¼0.76, BMI �30 vs BMI <25). For women, being non-Hispanic black (aOR¼1.22 vs non-Hispanic white) had significantly greater odds for being in the highest tertile of added sugars intake

October 2016 Volume 116 Number 10 JO

(�14.6 tsp/d), but those living in the Midwest (aOR¼0.77 vs South) had lower odds for being in the highest tertile of added sugars intake. For both men and women, the odds for being in the highest tertile of added sugars intake did not differ between Hispanics and non-Hispanic whites and between current moderate/heavy drinkers and lifetime ab- stainers of alcohol (Table 2). When only sociodemographic variables were included in the multinomial logistic regres- sion model without behavioral variables, the findings remained the same for both men and women (data not shown). Although findings for the middle tertile of added sugars intake (tertile 2) were somewhat similar to those findings for the highest tertile of added sugars (tertile 3), the magnitude of association was attenuated for tertile 2 (Table 2). For descriptive statistics, the contribution of foods and SSBs

to added sugars for men and women, as well as by age and race/ethnicity, are shown in Figure 1. Estimated median intake of added sugars was 17.6 tsp/d for men (Figure 1A) and 11.7 tsp/d for women (Figure 1B). For men, the overall median intake of added sugars from foods was similar to that for SSBs (7.9 and 7.7 tsp/d, respectively). However, the relative con- tributions of added sugars from foods and SSBs to total added sugars varied by the age and race/ethnicity of the men. For example, the median intake of added sugars from SSBs was higher than those from foods among younger men and His- panic men. For women, overall median intake of added sugars from foods was higher than that for SSBs (6.5 and 3.8 tsp/d, respectively), and the same pattern was observed in older women and in non-Hispanic white and non-Hispanic other/multiracial women. For descriptive statistics, the con- tributions of foods and SSBs to total added sugars by behavioral characteristics are shown in Figure 2. This relative contribution varied by smoking status for men (Figure 2A) and women (Figure 2B); median intake of added sugars from SSBs was higher than those from foods among current smokers in both men and women. Patterns seen across sub- groups defined by physical activity and alcohol consumption were similar.

DISCUSSION The NHIS offers a valuable resource for examining associa- tions among a wide array of both demographic and behav- ioral variables of interest in public health policy and practice. Although NHIS lacks the 24-hour recall data collected in NHANES, the NHIS has particular advantages. It surveys a larger sample than NHANES and allows us to stratify by sex and still consider a wide range of sociodemographic and behavioral variables. The current study focused on the particular dietary factor of added sugars, a variable that has been related to a variety of conditions and diseases. In this analysis of NHIS data, 1 in 3 men consumed at least 22.0 tsp (358 kcal) of added sugars and 1 in 3 women consumed at least 14.6 tsp (237 kcal) of added sugars daily, based on the tertile distribution cutoffs. In addition, high consumption of added sugars was associated with a number of sociodemo- graphic and behavioral characteristics among US men and women. In the 2010 NHIS, added sugars was assessed by nine

questions on a screener rather than by a more accurate and thorough diet assessment tool, such as the 24-hour recall.15

URNAL OF THE ACADEMY OF NUTRITION AND DIETETICS 1591

Table 1. Tertile distribution of estimated total added sugars intake from the dietary screener by sociodemographic and behavioral characteristics among US men and women aged 18 years and older—National Health Interview Survey, 2010

Characteristic

Tertile Distribution of Total Added Sugars Intake During the Past Month by Men and Womena

Men Women

n (%)b Tertile 1 (<13.5 tsp/d)

Tertile 2 (13.5 to <22.0 tsp/d)

Tertile 3 (‡22.0 tsp/d) n (%)b

Tertile 1 (<9.2 tsp/d)

Tertile 2 (9.2 to <14.6 tsp/d)

Tertile 3 (‡14.6 tsp/d)

�������������%�standard error��������! ����������%�standard error����������! All respondents 10,962 33.0�0.6 33.1�0.5 33.9�0.6 14,005 32.7�0.5 33.2�0.5 34.2�0.5 Age, y 10,962 14,005

18 to 24 1,189 (13.3) 14.0�1.3 28.5�1.6 57.5�1.9 1,387 (12.4) 15.8�1.3 32.1�1.7 52.1�1.9 25 to 39 3,082 (27.2) 21.0�0.8 34.7�1.0 44.3�1.0 3,725 (25.6) 22.2�0.8 31.5�0.9 46.3�1.0 40 to 59 3,898 (36.8) 35.7�0.9 33.6�0.8 30.7�0.9 4,775 (36.0) 34.9�0.8 34.4�0.8 30.7�0.8 60 and older 2,793 (22.7) 54.0�1.2 33.2�1.1 12.8�0.8 4,118 (26.0) 47.8�0.9 33.7�0.8 18.5�0.7 Race/ethnicity 10,962 14,005

White, non-Hispanic 6,354 (68.4) 33.8�0.7 34.0�0.6 32.2�0.7 7,895 (68.1) 33.6�0.6 34.4�0.6 32.0�0.7 Black, non-Hispanic 1,597 (10.6) 31.8�1.5 31.4�1.3 36.8�1.4 2,356 (12.2) 28.5�1.1 28.2�1.2 43.4�1.2 Hispanic 2,117 (15.1) 26.3�1.2 30.8�1.2 42.2�2.1 2,627 (13.1) 28.7�1.1 31.5�1.0 39.8�1.2 Other/multiracial, non-Hispanic 894 (5.9) 42.2�2.1 32.1�2.0 25.6�1.9 1,127 (6.7) 38.4�1.8 33.3�1.7 28.2�1.6 Marital status 10,949 13,980

Married/domestic partnership 6,032 (64.1) 35.5�0.7 33.8�0.7 30.7�0.7 6,595 (59.2) 33.1�0.7 33.8�0.6 33.1�0.7 Not married 4,917 (35.9) 28.3�0.8 31.9�0.8 39.8�1.0 7,385 (40.8) 32.0�0.7 32.2�0.7 35.8�0.8 Education level 10,917 13,958

Less than high school 1,871 (14.8) 30.3�1.2 28.0�1.3 41.7�1.4 2,345 (13.6) 30.6�1.2 26.8�1.0 42.6�1.3 High school/GEDc 2,901 (26.9) 28.0�1.0 31.9�1.0 40.1�1.1 3,615 (26.1) 31.2�0.9 30.9�1.1 37.9�1.1 Some college 3,075 (28.9) 31.1�1.0 32.2�1.0 36.7�1.1 4,402 (32.4) 31.0�0.9 33.6�0.9 35.5�0.9 College graduate and beyond 3,070 (29.4) 40.5�1.0 37.9�1.0 21.7�0.9 3,596 (27.9) 37.0�1.0 38.0�0.9 25.1�0.8 Annual family income 10,481 13,314

<$35,000 4,192 (31.6) 30.7�0.9 30.2�0.8 39.1�1.0 6,253 (35.9) 30.2�0.7 29.4�0.7 40.4�0.8 $35,000 to $74,999 3,413 (33.2) 30.8�0.9 34.0�1.0 35.2�1.0 3,994 (32.3) 31.6�0.9 33.3�0.8 35.1�1.0 $75,000 to $99,999 1,106 (12.7) 32.3�1.7 32.1�1.6 35.6�1.8 1,227 (11.8) 32.6�1.5 36.1�1.5 31.3�1.6 �$100,000 1,770 (22.4) 38.4�1.3 36.6�1.3 25.0�1.4 1,840 (19.9) 36.4�1.2 37.9�1.3 25.7�1.1

(continued on next page)

R E S E A R C H

1 5 9 2

JO U R N A L O F T H E A C A D E M Y O F N U T R IT IO N

A N D

D IE T E T IC S

O cto

b er

2016 Vo

lu m e 116

N u m b er

10

Table 1. Tertile distribution of estimated total added sugars intake from the dietary screener by sociodemographic and behavioral characteristics among US men and women aged 18 years and older—National Health Interview Survey, 2010 (continued)

Characteristic

Tertile Distribution of Total Added Sugars Intake During the Past Month by Men and Womena

Men Women

n (%)b Tertile 1 (<13.5 tsp/d)

Tertile 2 (13.5 to <22.0 tsp/d)

Tertile 3 (‡22.0 tsp/d) n (%)b

Tertile 1 (<9.2 tsp/d)

Tertile 2 (9.2 to <14.6 tsp/d)

Tertile 3 (‡14.6 tsp/d)

�������������%�standard error��������! ����������%�standard error����������! Census region 10,962 14,005

Northeast 1,730 (17.7) 32.3�1.1 35.3�1.1 32.3�1.4 2,241 (17.4) 32.0�1.2 35.5�1.3 32.5�1.2 Midwest 2,401 (23.0) 32.6�1.2 32.4�1.1 35.1�1.2 3,100 (23.3) 33.6�1.1 31.8�1.0 34.5�1.2 South 3,975 (35.2) 32.2�1.0 30.9�0.8 36.9�1.1 5,169 (35.7) 30.2�0.8 30.8�0.8 39.0�1.0 West 2,856 (24.1) 34.8�1.1 35.5�1.2 29.7�1.2 3,495 (23.6) 35.9�1.0 36.3�0.9 27.7�0.9 Any leisure-time physical activity for at least 10 min

10,811 13,840

0 times/wk 3,392 (29.0) 33.2�1.0 31.1�1.0 35.7�0.9 5,310 (35.2) 31.0�0.8 29.0�0.8 40.0�0.9 >0 to 3 times/wk 2,448 (23.1) 32.0�1.2 32.4�1.1 35.5�1.2 2,961 (22.2) 30.2�1.0 36.9�1.2 32.9�1.0 >3 to 5 times/wk 1,332 (13.1) 34.2�1.5 37.5�1.5 28.4�1.5 1,543 (11.5) 36.8�1.5 36.0�1.4 27.2�1.4 >5 times/wk 3,639 (34.8) 33.0�0.9 33.6�0.9 33.4�1.0 4,026 (31.1) 34.7�1.0 34.4�0.9 30.8�0.9 Smoking status 10,946 13,993

Never smoked 5,757 (53.2) 32.6�0.7 34.6�0.7 32.8�0.8 9,129 (64.6) 34.1�0.6 34.6�0.6 31.4�0.6 Former smoker 2,822 (25.7) 43.0�1.1 32.3�1.0 24.6�0.9 2,510 (18.2) 38.1�1.3 35.6�1.2 26.4�1.0 Current smoker 2,367 (21.1) 21.3�1.0 30.4�1.1 48.3�1.2 2,354 (17.2) 21.7�1.0 25.3�1.1 53.1�1.2 Alcohol intake 10,829 13,923

Lifetime abstainer 1,589 (14.8) 34.4�1.5 30.5�1.4 35.1�1.6 3,901 (25.8) 36.3�1.0 30.2�0.9 33.5�1.1 Former drinker 1,706 (14.7) 36.2�1.4 32.8�1.3 31.0�1.3 2,075 (14.0) 35.5�1.3 31.2�1.3 33.3�1.3 Current infrequent/light drinker 4,479 (42.2) 32.0�0.8 32.9�0.8 35.1�0.9 6,223 (46.6) 29.7�0.7 34.9�0.7 35.4�0.7 Current moderate/heavy drinker 3,055 (28.3) 31.9�1.0 35.2�1.0 32.9�1.1 1,724 (13.7) 32.6�1.4 35.0�1.3 32.5�1.3 Weight statusd 10,883 13,476

Underweight/ normal weight 3,354 (30.3) 29.3�0.9 31.6�1.0 39.1�1.1 5,605 (43.8) 31.1�0.8 34.8�0.8 34.2�0.8 Overweight 4,541 (41.4) 33.7�0.9 34.6�0.9 31.7�0.9 3,918 (28.2) 34.1�0.9 32.6�0.9 33.3�1.0 Obese 2,988 (28.3) 35.8�1.1 32.6�1.0 31.6�1.0 3,953 (28.0) 32.7�0.9 31.2�1.0 36.1�1.0 aBased on c2 tests across categories. All variables had P values �0.01 for both men and women. bUnweighted sample size and weighted percent are presented. Weighted percent may not add up 100% due to rounding. cGED¼general education development certificate. dBased on body mass index (BMI; calculated as kg/m2). BMI <25 for underweight/normal weight, BMI 25 to <30 for overweight, and BMI �30 for obese.

R E S E A R C H

O cto

b er

2016 Vo

lu m e 116

N u m b er

10 JO

U R N A L O F T H E A C A D E M Y O F N U T R IT IO N

A N D

D IE T E T IC S

1 5 9 3

Table 2. Adjusted odds ratios and 95% CI for estimated total added sugars intake from the dietary screener associated with sociodemographic and behavioral characteristics among US men and women aged 18 years and older—National Health Interview Survey, 2010

Characteristic

Total Added Sugars Intake

Men (n[10,122)

Women (n[12,564)

Tertile 2 (13.5 to <22.0 tsp/d)

Tertile 3 (‡22.0 tsp/d)

Tertile 2 (9.2 to <14.6 tsp/d)

Tertile 3 (‡14.6 tsp/d)

���������������������������������adjusted odds ratio (95% CI)a���������������������������������! Age, y

18 to 24 3.38 (2.52-4.54)b 18.52 (13.41-25.58)b 2.92 (2.26-3.75)b 9.91 (7.54-13.01)b

25 to 39 2.82 (2.38-3.35)b 10.46 (8.42-12.99)b 2.01 (1.75-2.32)b 6.48 (5.47-7.67)b

40 to 59 1.51 (1.30-1.75)b 3.84 (3.16-4.67)b 1.38 (1.21-1.56)b 2.52 (2.17-2.93)b

60 and older Reference Reference Reference Reference

Race/ethnicity

White, non-Hispanic Reference Reference Reference Reference

Black, non-Hispanic 0.90 (0.74-1.08) 0.80 (0.66-0.97)b 0.99 (0.83-1.17) 1.22 (1.03-1.45)b

Hispanic 0.97 (0.80-1.18) 1.11 (0.91-1.35) 0.99 (0.85-1.16) 1.03 (0.86-1.23)

Other/multiracial, non-Hispanic 0.64 (0.51-0.81)b 0.55 (0.43-0.71)b 0.85 (0.69-1.03) 0.78 (0.63-0.96)b

Marital status

Married/domestic partnership Reference Reference Reference Reference

Not married 0.93 (0.81-1.07) 0.89 (0.78-1.03) 0.94 (0.84-1.05) 0.89 (0.79-1.01)

Education level

Less than high school 1.06 (0.86-1.31) 2.06 (1.63-2.59)b 1.01 (0.83-1.23) 1.82 (1.47-2.26)b

High school/GEDc 1.22 (1.02-1.45)b 2.03 (1.68-2.46)b 1.08 (0.91-1.27) 1.63 (1.37-1.95)b

Some college 1.02 (0.87-1.19) 1.53 (1.27-1.84)b 1.06 (0.92-1.22) 1.36 (1.16-1.59)b

College graduate or beyond Reference Reference Reference Reference

Annual family income

<$35,000 0.98 (0.81-1.20) 1.23 (0.97-1.55) 1.04 (0.87-1.25) 1.33 (1.09-1.62)b

$35,000 to $74,999 1.11 (0.93-1.33) 1.29 (1.06-1.58)b 1.08 (0.92-1.27) 1.22 (1.02-1.46)b

$75,000 to $99,999 0.96 (0.77-1.20) 1.30 (1.00-1.69) 1.09 (0.89-1.34) 1.15 (0.92-1.44)

�$100,000 Reference Reference Reference Reference

(continued on next page)

R E S E A R C H

1 5 9 4

JO U R N A L O F T H E A C A D E M Y O F N U T R IT IO N

A N D

D IE T E T IC S

O cto

b er

2016 Vo

lu m e 116

N u m b er

10

Table 2. Adjusted odds ratios and 95% CI for estimated total added sugars intake from the dietary screener associated with sociodemographic and behavioral characteristics among US men and women aged 18 years and older—National Health Interview Survey, 2010 (continued)

Characteristic

Total Added Sugars Intake

Men (n[10,122)

Women (n[12,564)

Tertile 2 (13.5 to <22.0 tsp/d)

Tertile 3 (‡22.0 tsp/d)

Tertile 2 (9.2 to <14.6 tsp/d)

Tertile 3 (‡14.6 tsp/d)

���������������������������������adjusted odds ratio (95% CI)a���������������������������������! Census region

Northeast 1.17 (1.01-1.36)b 1.00 (0.85-1.18) 1.10 (0.94-1.27) 0.90 (0.76-1.05)

Midwest 0.99 (0.85-1.16) 0.89 (0.74-1.07) 0.87 (0.76-1.00) 0.77 (0.65-0.90)b

South Reference Reference Reference Reference

West 1.03 (0.88-1.21) 0.66 (0.55-0.79)b 0.98 (0.85-1.12) 0.64 (0.55-0.75)b

Any leisure-time physical activity for at least 10 min

0 times/wk 1.06 (0.90-1.24) 1.22 (1.04-1.44)b 1.14 (0.99-1.32) 1.64 (1.42-1.89)b

>0 to 3 times/wk 0.98 (0.84-1.14) 1.13 (0.95-1.34) 1.36 (1.16-1.59)b 1.36 (1.17-1.57)b

>3 to 5 times/wk 1.03 (0.85-1.24) 0.80 (0.65-1.00) 0.96 (0.79-1.16) 0.80 (0.66-0.98)b

>5 times/wk Reference Reference Reference Reference

Smoking status

Never smoked Reference Reference Reference Reference

Former smoker 0.87 (0.75-1.00) 0.93 (0.80-1.08) 0.99 (0.86-1.14) 1.04 (0.89-1.22)

Current smoker 1.29 (1.09-1.52)b 1.96 (1.66-2.30)b 1.06 (0.90-1.25) 2.29 (1.98-2.65)b

Alcohol intake

Lifetime abstainer Reference Reference Reference Reference

Former drinker 1.36 (1.09-1.70)b 1.61 (1.25-2.07)b 1.23 (1.03-1.47)b 1.24 (1.03-1.49)b

Current infrequent/light drinker 1.23 (1.03-1.47)b 1.39 (1.13-1.70)b 1.41 (1.24-1.61)b 1.37 (1.18-1.59)b

Current moderate/heavy drinker 1.28 (1.06-1.56)b 1.19 (0.96-1.48) 1.26 (1.06-1.49)b 1.07 (0.87-1.32)

Weight statusd

Underweight/normal weight Reference Reference Reference Reference

Overweight 1.04 (0.91-1.20) 0.94 (0.81-1.10) 0.90 (0.79-1.03) 0.95 (0.82-1.10)

Obese 0.91 (0.78-1.07) 0.76 (0.64-0.91)b 0.88 (0.77-1.01) 0.97 (0.84-1.11)

aMultivariate logistic regression models included all variables in one model separately by sex. The comparison group for multivariate logistic regression analysis was tertile 1 category for men (<13.5 tsp/d) and women (<9.2 tsp/d). bConsidered statistically significant based on the 95% CI. cGED¼general education development certificate. dBased on body mass index (BMI; calculated as kg/m2). BMI <25 for underweight/normal weight, BMI 25 to <30 for overweight, and BMI �30 for obese.

R E S E A R C H

O cto

b er

2016 Vo

lu m e 116

N u m b er

10 JO

U R N A L O F T H E A C A D E M Y O F N U T R IT IO N

A N D

D IE T E T IC S

1 5 9 5

17.6

23.9 20.6

16.7

17.0

17.0 18.7

20.2

15.0

7.9 8.2 8.0 7.8 7.9 8.1 7.4 7.5 7.37.7

14.1 11.3

7.2

7.0

7.0 9.3

10.8

6.0

0

5

10

15

20

25

Overall 18-24 y 25-39 y 40-59 y ≥60 y Non- Hispanic

White

Non- Hispanic

Black

Hispanic Non- Hispanic

Other

M en

's a

dd ed

s ug

ar in

ta ke

in

te as

po on

s pe

r da

y

Total Foods SSBs

Age Race/Ethnicity

A

11.7

15.1 14.0

11.2 9.4

11.5 13.2 12.7

10.9

6.5 7.1 6.8 6.3 6.2 6.7 6.0 6.4 6.1

3.8

6.7 5.7

3.5 2.1

3.4

5.8 4.7

3.3

0 2 4 6 8

10 12 14 16

Overall 18-24 y 25-39 y 40-59 y ≥60 y Non- Hispanic

White

Non- Hispanic

Black

Hispanic Non- Hispanic

Other

W om

en 's

a dd

ed s

ug ar

in ta

ke in

te

as po

on s

pe r

da y

Total Foods SSBs

Age Race/Ethnicity

B

Figure 1. Estimated median intake of total added sugars from the dietary screener and the contribution of foods and sugar- sweetened beverages to total added sugars for men (A) and women (B) aged 18 years and older by age and race/ethnicity— National Health Interview Survey, 2010.

RESEARCH

Despite this, for adults, the screener estimates of median added sugars intake in the 2010 NHIS were comparable to 24- hour recall estimates in the 2007-2010 NHANES: 17.6 tsp/d vs 17.1 tsp/d among men and 11.7 tsp/d vs 12.7 tsp/d among

17.9 17.8 16.6 17.5 17.5

1

7.6 7.7 8.3 8.1 8.18.1 8.5 6.8 7.5 7.5

0

5

10

15

20

25

M en

's a

dd ed

s ug

ar in

ta ke

in

te as

po on

s pe

r da

y

Physical Ac vity Smokin

A

12.5 11.9 10.8 11.3 11.4 1

6.3 6.6 6.7 6.7 6.7

4.5 3.8 3.1 3.3 3.4

0 2 4 6 8

10 12 14 16

W om

en 's

a dd

ed s

ug ar

in ta

ke in

te

as po

on s

pe r

da y

Physical Ac vity Smokin

B

Figure 2. Estimated median intake of total added sugars from t sweetened beverages to total added sugars for men (A) and wom activity for at least 10 minutes, smoking status, and alcohol intake

1596 JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS

women.18 In addition, the present study found that re- lationships between added sugars and demographic variables in the NHIS mirrored results from NHANES: added sugars intake was higher among younger, less educated, and adults

5.1

21.4

17.6 16.6 17.7 17.8

7.9 7.3 7.9 8.0 8.0 7.7 5.0

12.6

7.7 6.8 7.7 8.2

Total Foods SSBs

g Status Alcohol Intake

0.7

15.3

11.4 11.3 12.0 11.7

6.4 5.9 6.4 6.3 6.7 6.3

3.1

8.3

3.4 3.4 4.1 4.0

Total Foods SSBs

g Status Alcohol Intake

he dietary screener and the contribution of foods and sugar- en (B) aged 18 years and older by any leisure-time physical

—National Health Interview Survey, 2010.

October 2016 Volume 116 Number 10

RESEARCH

with lower incomes than their counterparts.3,10,18 Consistent with earlier analyses by Park and colleagues,19 the South was the region with the highest intakes. Although there is limited information on the regional variation of total added sugars intake among the US adult population, previous studies re- ported that SSB intake varied by states and geographical re- gions.12,19-22 Potential reasons for this geographical variation may be differences in the food environment, such as the availability/access of foods and beverages, advertising, cul- ture, and regulations.23-26

These analyses add to existing knowledge by examining important behavioral variables associated with added sugars consumption controlling for demographics. For example, the present study found that physically inactive men and women had 22% and 64%, respectively, higher odds for being in the highest tertile of added sugars intake compared to men and women who were physically active. Consistent with findings in the present study, a previous study12 reported that daily SSB intake was associated with physical inactivity among US adults. The inverse association of physical activity with higher added sugars intake is also consistent with studies that have found that physically inactive adults have poorer diet quality in other dimensions than physically active adults.27,28

Regardless of sociodemographic factors, smokers consumed the highest amounts of added sugars in the pre- sent study. This is consistent with other studies that have found smokers have other less healthy dietary7,11,12 and health behaviors.11 The clustering may reflect overall atti- tudes about health and engaging in healthy behaviors. The finding might also reflect differences in taste perception caused by smoking. Previous studies showed that smokers had decreased sensitivity (or increased thresholds) to sucrose,29,30 indicating smokers might be less likely to perceive a food as sweet and thus eat more. In addition, some research indicates that addictions and addictive-like behaviors can be shared across domains, for example, smoking and food, in this case, added sugars.31-33

In the present analyses, added sugars consumption varied by the pattern of alcohol consumption; added sugars intake was higher among former alcohol drinkers or current infrequent/light alcohol drinkers than lifetime abstainers of alcohol, but there was no difference in added sugars intake between current moderate/heavy drinker and lifetime abstainers of alcohol after controlling for other factors. A previous study reported that alcohol drinkers had lower odds of consuming SSB at least once per day than non�alcohol drinkers among US adults.12 The reason for this finding in the present study is not clear. Unlike the findings from research on smokers, evidence that those who drink alcohol have a differential preference for sweets is inconsistent.34-36 Poten- tially, those who drink lightly are more likely to mix their alcohol with SSBs. Research that tests this and other hypotheses could improve the design of interventions to address excessive added sugars intake. In addition, the present study found that obese men had

24% lower odds for consuming �22.0 tsp/d of total added sugars than underweight/normal-weight men, and weight status was not associated with added sugars intake among women after controlling for other factors. It is possible that obese men might decrease their added sugars intake as a strategy for weight control.

October 2016 Volume 116 Number 10 JO

While the present study analyzed data from a large, na- tionally representative sample of US adults and included various behavioral characteristics, there are at least two limitations. First, the outcome variable (ie, total added sugars) was estimated from questions about usual intake of nine items on a screener rather than from a less-biased in- strument that collects complete information on daily food and beverage intake, such as a 24-hour recall.15 However, the fact that the relationships found among important de- mographic variables are consistent with the findings based on NHANES data using 24-hour recall is reassuring. Second, food environmental variables (eg, access to supermarkets and food outlets) were not included in the study because the 2010 NHIS did not collect such information.

CONCLUSIONS Based on the tertile distribution cutoff points, 1 in 3 men consumed at least 22.0 tsp of added sugars and 1 in 3 women consumed at least 14.6 tsp of added sugars daily. Findings in the present study showed that higher added sugars intake was associated independently with various sociodemo- graphic and behavioral characteristics, such as younger age, lower education, lower income, physical inactivity, current smoking status, and alcohol intake. This information may inform intervention efforts by characterizing high-intake populations and aiding efforts to reduce their added sugars intake.

References 1. US Department of Agriculture, US Department of Health and Human

Services. Dietary Guidelines for Americans, 2015-2020, 8th ed. Washington, DC: US Government Printing Office; 2015. http:// health.gov/dietaryguidelines/2015/guidelines/?linkId¼20169028. Accessed March 8, 2016.

2. Johnson RK, Appel LJ, Brands M, et al. Dietary sugars intake and cardiovascular health: A scientific statement from the American Heart Association. Circulation. 2009;120(11):1011-1020.

3. Ervin RB, Ogden CL. Consumption of added sugars among US adults, 2005-2010. NCHS Data Brief. 2013;(122):1-8.

4. Te Morenga L, Mallard S, Mann J. Dietary sugars and body weight: Systematic review and meta-analyses of randomised controlled tri- als and cohort studies. BMJ. 2013;346:e7492.

5. Bray GA, Nielsen SJ, Popkin BM. Consumption of high-fructose corn syrup in beverages may play a role in the epidemic of obesity. Am J Clin Nutr. 2004;79(4):537-543.

6. Sonestedt E, Overby NC, Laaksonen DE, Birgisdottir BE. Does high sugar consumption exacerbate cardiometabolic risk factors and in- crease the risk of type 2 diabetes and cardiovascular disease? Food Nutr Res. 2012;56.

7. Yang Q, Zhang Z, Gregg EW, Flanders WD, Merritt R, Hu FB. Added sugar intake and cardiovascular diseases mortality among US adults. JAMA Intern Med. 2014;174(4):516-524.

8. DiNicolantonio JJ, Lucan SC. The wrong white crystals: Not salt but sugar as aetiological in hypertension and cardiometabolic disease. Open Heart. 2014;1(1):e000167.

9. Lula EC, Ribeiro CC, Hugo FN, Alves CM, Silva AA. Added sugars and periodontal disease in young adults: An analysis of NHANES III data. Am J Clin Nutr. 2014;100(4):1182-1187.

10. Thompson FE, McNeel TS, Dowling EC, Midthune D, Morrissette M, Zeruto CA. Interrelationships of added sugars intake, socioeconomic status, and race/ethnicity in adults in the United States: National Health Interview Survey, 2005. J Am Diet Assoc. 2009;109(8):1376- 1383.

11. Lee CG, Seo DC, Middlestadt SE, Lin HC. Does the relationship be- tween cigarette smoking and other key health behaviors vary by geographic area among US young adults? A multilevel analysis. Int J Behav Med. 2015;22(4):481-488.

URNAL OF THE ACADEMY OF NUTRITION AND DIETETICS 1597

RESEARCH

12. Park S, Pan L, Sherry B, Blanck HM. Consumption of sugar-sweetened beverages among US adults in 6 states: Behavioral Risk Factor Sur- veillance System, 2011. Prev Chronic Dis. 2014;11:E65.

13. Centers for Disease Control and Prevention. 2010 National Health Interview Survey (NHIS) Public Use Data Release. NHIS Survey Description. Hyattsville, MD: National Center for Health Statis- tics; 2011. ftp://ftp.cdc.gov/pub/Health_Statistics/NCHS/Dataset_ Documentation/NHIS/2010/srvydesc.pdf. Accessed March 8, 2016.

14. National Cancer Institute. Dietary Screener Questionnaire in the National Health Interview Survey Cancer Control Supplement 2010: Overview. http://epi.grants.cancer.gov/nhis/2010-screener/. Revised November 2, 2015. Accessed March 8, 2016.

15. National Cancer Institute. Dietary Screener Questionnaire (DSQ) in the NHANES 2009-10: Data Processing & Scoring Procedures. http:// epi.grants.cancer.gov/nhanes/dietscreen/scoring/. Revised October 15, 2015. Accessed March 8, 2016.

16. Schiller JS, Lucas JW, Ward BW, Peregoy JA. Summary health statis- tics for US adults: National Health Interview Survey, 2010. Vital Health Stat 10. 2012;(252):1-207.

17. National Heart, Lung, and Blood Institute. Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults. The evidence report. NIH Publication No. 98-4083. Bethesda, MD: National Institutes of Health; 1998. http://www. nhlbi.nih.gov/guidelines/obesity/ob_gdlns.pdf. Accessed March 8, 2016.

18. National Cancer Institute. Usual intake of added sugars. Table A40. Added sugars: Means, percentiles and standard errors of usual intake, 2007-2010. http://appliedresearch.cancer.gov/diet/ usualintakes/pop/2007-10/table_a40.html. Revised June 2, 2015. Accessed March 8, 2016.

19. Park S, Onufrak S, Sherry B, Blanck HM. The relationship between health-related knowledge and sugar-sweetened beverage intake among US adults. J Acad Nutr Diet. 2014;114(7):1059-1066.

20. Kumar GS, Pan L, Park S, Lee-Kwan SH, Onufrak S, Blanck HM. Sugar- sweetened beverage consumption among adults—18 states, 2012. MMWR Morb Mortal Wkly Rep. 2014;63(32):686-690.

21. Park S, McGuire LC, Galuska DA. Regional differences in sugar- sweetened beverage intake among US adults. J Acad Nutr Diet. 2015;115(12):1996-2002.

22. Park S, Xu F, Town M, Blanck HM. Prevalence of sugar-sweetened beverage intake among adults—23 states and the District of Columbia, 2013. MMWR Morb Mortal Wkly Rep. 2016;65(7): 169-174.

1598 JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS

23. Martin-Biggers J, Yorkin M, Aljallad C, et al. What foods are US su- permarkets promoting? A content analysis of supermarket sales circulars. Appetite. 2013;62:160-165.

24. Tucker KL, Maras J, Champagne C, et al. A regional food-frequency questionnaire for the US Mississippi Delta. Public Health Nutr. 2005;8(1):87-96.

25. Pomeranz JL. Advanced policy options to regulate sugar-sweetened beverages to support public health. J Public Health Policy. 2012;33(1):75-88.

26. Studdert DM, Flanders J, Mello MM. Searching for public health law’s sweet spot: The regulation of sugar-sweetened beverages. PLoS Med. 2015;12(7):e1001848.

27. Erinosho TO, Thompson OM, Moser RP, Yaroch AL. Fruit and vege- table intake of US adults: Comparing intake by mode of survey administration. J Am Diet Assoc. 2011;111(3):408-413.

28. Wang Y, Chen X. Between-group differences in nutrition- and health-related psychosocial factors among US adults and their associations with diet, exercise, and weight status. J Acad Nutr Diet. 2012;112(4):486-498 e483.

29. Pepino MY, Mennella JA. Effects of cigarette smoking and family history of alcoholism on sweet taste perception and food cravings in women. Alcohol Clin Exp Res. 2007;31(11):1891-1899.

30. Sato K, Endo S, Tomita H. Sensitivity of three loci on the tongue and soft palate to four basic tastes in smokers and non-smokers. Acta Otolaryngol Suppl. 2002;546:74-82.

31. Mahler SV, de Wit H. Cue-reactors: Individual differences in cue- induced craving after food or smoking abstinence. PLoS One. 2010;5(11):e15475.

32. Thorgeirsson TE, Gudbjartsson DF, Sulem P, et al. A common bio- logical basis of obesity and nicotine addiction. Transl Psychiatry. 2013;3:e308.

33. Fortuna JL. Sweet preference, sugar addiction and the familial history of alcohol dependence: Shared neural pathways and genes. J Psychoactive Drugs. 2010;42(2):147-151.

34. Tremblay KA, Bona JM, Kranzler HR. Effects of a diagnosis or family history of alcoholism on the taste intensity and hedonic value of sucrose. Am J Addict. 2009;18(6):494-499.

35. Kampov-Polevoy AB, Garbutt JC, Khalitov E. Family history of alco- holism and response to sweets. Alcohol Clin Exp Res. 2003;27(11): 1743-1749.

36. Fischer ME, Cruickshanks KJ, Schubert CR, et al. Taste intensity in the Beaver Dam Offspring Study. Laryngoscope. 2013;123(6):1399-1404.

AUTHOR INFORMATION S. Park and L. Pan are epidemiologists, D. A. Galuska is an associate director for science, and H. M. Blanck is a branch chief, Division of Nutrition, Physical Activity and Obesity, and L. C. McGuire is a lead epidemiologist, Division of Population Health, all at the National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, Atlanta, GA. F. E. Thompson is a program director, Risk Factor Assessment Branch, Epidemiology and Genomics Research Program, Division of Cancer Control and Population Sciences, National Cancer Institute, National Institutes of Health, Bethesda, MD.

Address correspondence to: Sohyun Park, PhD, Division of Nutrition, Physical Activity, and Obesity, National Center for Chronic Disease Pre- vention and Health Promotion, Centers for Disease Control and Prevention, 4770 Buford Hwy NE, Mailstop F77, Atlanta, GA 30341. E-mail: [email protected]

STATEMENT OF POTENTIAL CONFLICT OF INTEREST No potential conflict of interest was reported by the authors.

FUNDING/SUPPORT There is no funding to disclose.

ACKNOWLEDGEMENTS The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.

October 2016 Volume 116 Number 10