Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Oct 9.
Published in final edited form as: J Community Health. 2019 Jun;44(3):507–518. doi: 10.1007/s10900-018-00613-6

Income, Race and its Association with Obesogenic Behaviors of U.S. Children and Adolescents, NHANES 2003–2006

Ethan T Hunt 1, Keith Brazendale 1, Caroline Dunn 1, Alycia K Boutté 3, Jihong Liu 2, James Hardin 2, Michael W Beets 1, R Glenn Weaver 1
PMCID: PMC11462129  NIHMSID: NIHMS2025177  PMID: 30659412

Abstract

Objectives

To describe the associations of income and race with obesogenic behaviors and % body fat among a large sample of U.S. children and adolescents.

Design

Data were obtained from the 2003–2004 and 2005–2006 National Health and Nutritional Examination Survey. Multiple linear regression models and interactions were used to examine the associations of moderate-to-vigorous physical activity (MVPA), sedentary time, diet quality, and screen-time with income-to-poverty ratio and race. Separate stratified analyses explored associations among individual obesogenic behaviors within race and income groups.

Results

This study included children and adolescents (n = 3551, mean = 13.1 years, SD = 3.9 years) who were 37% Hispanic, 27% White, and 35% Black. Overall, Hispanic children/adolescents had significantly higher levels of adiposity (3.6, 95 CI = 0.9, 6.3) than white children and adolescents. Medium-income children and adolescents engaged in less MVPA (−3.3 min, 95 CI = −5.1, −1.5), had poorer diet quality (−1.1, 95 CI = −1.9, −0.2), and used screens less (−33.9 min, 95 CI = −45.4, −22.4) than children and adolescents from low-income households. High-income children and adolescents also engaged in less MVPA (−3.1 min, 95 CI = −5.5, −0.7) and used screens less (−62.9 min, 95 CI = −78.3, −47.4) than children and adolescents from low-income households. However, there were significant race/ethnicity-by-income interactions for high-income Hispanic children and adolescents with diet quality (−3.5 HEI-2010 score, 95 CI = −6.6, −0.4) and screen time (66.9 min, 95 CI = 24.7, 109.0). There was also a significant race/ethnicity-by-income interaction for the screen-time of Black children and adolescents from medium (33.8 min, 95% CI 0.2, 67.3) and high (75.8 min, 95% CI 34.7, 117.0) income households.

Conclusions

There appears to be a complex relationship that varies by race/ethnicity between income, obesogenic behaviors, and adiposity levels among children and adolescents. More work is needed to identify the behavioral mechanisms that are driving disparate rates of overweight and obesity among minority children and those from low-income households.

Keywords: Obesity, Youth, Physical activity, Diet, Screen time, Adiposity

Introduction

Childhood obesity is an international “public health epidemic” [37, 43, 55]. The prevalence of overweight and obesity among children and adolescents (6–19 years) is at an all-time high in the United States (U.S.) [45]. Obesity has been linked to a variety of non-communicable diseases including abnormal fasting glucose, insulin resistance, type 2 diabetes, hypertension, sleep apnea, and asthma [4, 40]. A mounting body of evidence has examined the intersections between race and socioeconomic position (i.e., commonly measured by household income) on a myriad of health outcomes [1, 5, 11, 20, 28, 34, 58]. Studies have identified longitudinal trends in childhood and adolescent overweight or obesity and the relationship to income from birth to adulthood [26, 31]. In the U.S., children and adolescents from low-income families are more likely to be overweight or obese by age 5 than their high-income counterparts [10, 14, 25, 31, 32, 51] and the relationship between income and obesity may vary between racial/ethnic groups [18, 19, 31, 49].

Obesity is a multifactorial condition influenced by social, environmental, and genetic determinates [34]. Modifiable obesogenic behaviors such as moderate-to-vigorous physical activity (MVPA), sedentary behaviors, diet quality, and screen time play a large role in determining weight status. Examining differing levels of engagement in obesogenic behaviors by race and income-level might uncover important targets for intervention with the potential to reduce or eliminate income and race/ethnicity related disparities in overweight and obesity. The objective of this study was to examine the association between income, race, and four obesogenic behaviors using 2 waves (2003–2004 and 2005–2006) from the National Health and Nutrition Examination Survey (NHANES).

Methods

Sample

NHANES is a repeated cross-sectional study conducted in 2-year cycles and collects data on a nationally representative sample of the U.S. noninstitutionalized population. Trained staff members conduct both interviews and examinations throughout the U.S. A more detailed description of NHANES is available elsewhere [8, 30]. This study includes data from the 2003–2004 and 2005–2006 waves of NHANES. These years of data collection were selected because they include objective measures of physical activity, sedentary behaviors, and adiposity measured via dual X-ray absorptiometry (DXA) in addition to measuring diet intake and screen-time. Following standard protocols using NHANES data, participants were excluded if they were younger than 6 or older than 19 years. Participants were included if they had valid data from all four obesogenic behavior measures (i.e. MVPA, sedentary time, dietary intake, and screen-time). Descriptive characteristics of the sample can be found in Table 1.

Table 1.

Sample characteristics of children and adolescent aged 6–19, NHANES 2003–2006

Variable All Low income (PIR < 1.00) Medium income (PIR 1.00–4.00) High income (PIR > 4.00) Missing income
Total, n (%) 3551 (100) 1042 (29.3) 1797 (51.0) 567 (15.9) 145 (4.1)
Age, years (±SD) 13.1 ± 3.86 13.3 ± 4.0 12.9 ± 3.8 13.1 ± 3.7 13.8 ± 3.8
Sex, n (%)
 Boys 1778 (50.1) 493 (47.3) 918 (51.1) 286 (50.4) 81 (55.9)
 Girls 1773 (49.9) 549 (52.7) 879 (48.9) 281 (49.6) 64 (44.1)
Race/ethnicity, n (%)
 Hispanic 1346 (37.9) 468 (44.9) 693 (38.6) 123 (21.7) 62 (42.8)
 White 965 (27.2) 140 (13.4) 499 (27.8) 301 (53.1) 25 (17.2)
 Black 1240 (34.9) 434 (41.7) 605 (33.7) 143 (25.2) 58 (40.0)
Obesogenic behaviors
 MVPA (mins) ± SD 40.4 ± 26.5 42.0 ± 27.1 40.0 ± 26.5 39.7 ± 26.0 38.9 ± 23.7
 Sedentary (mins) ± SD 498.2 ± 162.2 500.7 ± 168.8 493.6 ± 161.1 504.3 ± 148.8 515.0 ± 178.4
 HEI-2010 ± SD 44.1 ± 11.1 44.5 ± 10.8 43.6 ± 11.1 44.8 ± 11.8 44.1 ± 11.2
 Screen (mins) ± SD 264.8 ± 152.9 292.6 ± 155.5 257.9 ± 150.2 228.3 ± 148.1 296.8 ± 145.6
 % body fat ± SD 29.4 ± 8.8 30.0 (8.9) 29.4 (8.9) 28.0 (7.9) 29.2 (9.5)
Open in a new tab

Participants included in all obesogenic analysis had full data including valid accelerometer, dietary intake, and screen time measurements (N = 3551)

% body fat measured by the dual energy X-ray (DXA) included (3,133) children with valid accelerometer, dietary intake, and screen time measurements

Measures

Physical Activity and Sedentary Time

The physical activity monitor, Actigraph (Actigraph, LLC; Ft. Walton Beach, FL) model 7164 accelerometer was placed on the right hip attached to an elastic belt for a 7-day wear period. Participants were asked to wear the device while they were awake and to take it off only for swimming and bathing. Monitors were then returned by express mail to the NHANES contractor, where data were downloaded, and the device was checked to determine whether it was still within the manufacture’s calibration specifications. Participants whose monitors were not calibrated when returned were excluded. For this study, consistent with previously published protocols [53], a valid day was defined as having 10 or more, hours of monitor wear. Wear time was defined by subtracting non-wear time (defined as 60 or more minutes of consecutive accelerometer counts of zero) from 24 (hours in a day) [53]. Participants were included if they had 4 or more valid days of wear, with at least one valid day occurring on a weekend. Cut points associated with sedentary behaviors [38] and MVPA [16] in children and adolescents were applied to the data to distill the amount of minutes per day spent in each activity level.

Screen-Time

Screen time included television, video, and computer game use, and was ascertained as part of the physical activity questionnaire [52], which was completed during a home interview. Participants reported separately, the number of hours during a typical day in the past 30 days that they watched television and/or used a computer (including video gaming consoles). Responses included ‘none’, ‘< 1 h’, ‘1 h’, ‘2 h’, ‘3 h’, ‘4 h’, or ‘5 or more hours’. Response categories for hours of computer use included ‘< 1 h’, ‘1 h’, ‘2 h’, ‘3 h’, ‘4 h’, or ‘5 h’. Hours of television and computer use were summed to ascertain an estimation of overall screen use. Self-reported screen time has been established as a valid and reliable predictor of cardiometabolic risk factors, such as obesity [21, 50]. Values that exceeded 480 min (8 h of daily screen use) were considered implausible and were converted to 480 min.

Dietary Intake

Dietary intake was obtained using two multiple pass 24-h dietary recall interviews conducted by trained research staff on 2 days in 1 week [6]. All 24-h recalls underwent a four-step quality assurance process to evaluate the overall completeness, acceptability, and plausibility of reported dietary intakes. Dietary recalls that met the minimum criteria for intakes collected with the Automated Multiple Pass Method were analyzed [2]. Responses from the dietary recall were then converted to Healthy Eating Index-2010 (HEI) scores via codes provided by the National Cancer Institute for participants with two nonconsecutive days of complete dietary data [22]. Additionally, children and adolescents who were under the age of 8 used a parent proxy when recalling their dietary intake. The HEI is a measure of diet quality that assesses conformance to the Dietary Guidelines for Americans, which is the basis of nutrition policy for the U.S. government and the foundation of federal nutrition guidance [22, 39].

The HEI is comprised of 12 components including: (1) total fruit; (2) whole fruit; (3) total vegetables; (4) greens and beans; (5) whole grains; (6) dairy; (7) total protein; (8) seafood and plant protein; and (9) ratio of unsaturated to saturated fatty acids; (10)refined grains; (11) sodium; and (12) empty calories (i.e., energy from solid fats, alcohol, and added sugars) [23]. For all components, higher scores reflect better diet quality. Component scores are summed and HEI total scores range from 0 to 100 [23].

Adiposity

Whole body (DXA) scans were performed during NHANES 1999–2006. Scans were acquired using a Hologic QDR 4500A fan-beam densitometer (Hologic, Inc., Bedford Massachusetts) among participants at least 8 years of age. Previous studies have reported the reliability and precision estimates using this device [7, 27, 35]. DXA scans are reviewed for quality using standardized protocols, and invalid DXA data were reclassified as missing values. To account for missingness that was associated with participant age, weight, and height, sequential regression multivariate imputation (SRMI) was performed. Detailed procedures and protocols regarding imputations have been described elsewhere [8, 48]. For the purposes of this study, we focused on total % body fat.

Income-to-Poverty-Ratio

The income-to-poverty ratio is a ratio that represents the ratio of household income to poverty and is calculated by dividing the total reported household income by the Department of Health and Human Services’ poverty level [13]. For example, an income-to-poverty ratio of 0.5, indicates that a household is earning an equivalent of 50% the amount of income as the federally established poverty level, while an income-to-poverty ratio of 1.5 indicates that a household is earning 150% of the federally established poverty level. For this study, each participant was classified as low, medium, or high based upon their household income-to-poverty ratio. Classifications align with previous literature [18] and were created as follows. If a child’s household received an income-to-poverty ratio of 0.00–1.00, the child was assigned to the low-income group. Likewise, children and adolescents living in a household with an income-to-poverty ratio of 1.01–4.00 were assigned to the medium-income group and children and adolescents living in households above 4.00 on the income-to-poverty ratio were classified as high-income.

Race/Ethnicity

Race/ethnicity of children and adolescents was collected via parent proxy report. In order to align with previous work using the NHANES datasets [9], a stepwise process for designating children and adolescents into race/ethnicity categories was employed. First, children and adolescents whose parent/guardian designated them as Hispanic or Mexican American were categorized as Hispanic regardless of race. Next, children and adolescents who were designated as Black and had not been assigned to Hispanic category were assigned to the Black category. Next, children and adolescents who were designated as White and had not been assigned to another category were assigned to the non-Hispanic White category. Finally, children and adolescents that were not assigned to any of the previous categories were assigned to the Other race/ethnicity category.

Statistical Analyses

Analyses were performed in Stata (v14.2 StataCorp College Station, Texas). The relationship between race/ethnicity, income, the four obesogenic behaviors of interest (i.e., MVPA, sedentary time, screen time, and diet quality) and % body fat was examined using a multi-step analysis. First, to examine the relationship of the income-to-poverty (i.e. low, medium, and high), four obesogenic behaviors, and% body fat, separate linear and regression models were conducted using income as the independent variable. This model allowed for the evaluation of the association of income with obesogenic behaviors and % body fat for all children and adolescents in the sample before adding the additional covariates. Second, to examine the relationship of, income, the four obesogenic behaviors, and % body fat whilst adding race (i.e. White, Black, and Hispanic), separate linear regression models were estimated for each behavior and adiposity outcome. In these models, either the behavior of interest or % body fat served as the dependent variable and income-to-poverty ratio was the independent variable. Age, sex, race/ethnicity, and all possible race/ethnicity-by-income-level interaction terms were included as covariates. Low-income and White were used as the reference categories. Henceforth this approach is referred to as “model 2”. Third, an additional stratified analysis was conducted to expand upon and/or confirm the findings of the full model. In this analysis a separate linear regression model was estimated for each race/ethnicity group, obesogenic behavior, and adiposity status. In these models, either the behavior of interest or % body fat was the dependent variable and income-to-poverty ratio was the independent variable with age and sex included as covariates. Finally, a fourth stratified additional analysis was conducted with separate linear regression models estimated for each income level (i.e., low, medium, high), obesogenic behavior, and adiposity status. In these models the behavior of interest or weight outcome was the dependent variable and race/ethnicity was the independent variable. Age and sex were included as covariates.

Results

Descriptive and sociodemographic information of the final sample are shown in Table 1. The final sample included (N = 3551) children and adolescents. Children and adolescents in the Other race/ethnicity category were excluded from analyses due to inadequate sample size with valid data (n = 162). A subset of children and adolescents with valid DXA data and measures of all obesogenic behaviors included (n = 3132) participants. The mean age of the participants in the final sample was 13.1 years (SD = 3.9) and 50.1% were male. Overall, 17.8% and 21.2% of the final sample were either overweight or obese respectively.

The results of the linear regression models analyzing the association of income, race, and the interaction of race and income with adiposity and obesogenic behaviors can be found in Table 2. The initial linear model (model 1) with no covariates indicated high-income children and adolescents had significantly less adiposity when compared to their low-income peers (−2.2 95% CI −3.5, −0.8). The linear regression models indicated that low-income children and adolescents engaged in significantly more MVPA than their medium (−3.3 min 95% CI −5.1, −1.5) and high-income (−3.1 min 95% CI −5.5, −0.7) counterparts. Medium-income children and adolescents had significantly poorer diet quality (−1.1 HEI-2010 score 95% CI −1.9, −0.2) than their low-income counterparts. Low-income children and adolescents also engaged in more screen-time than their medium (−33.9 min 95% CI −45.4, −22.4) and high-income (−62.9 min 95% CI −78.3, −47.4) counterparts. There were no statistically significant associations between income-to-poverty-ratio and sedentary time.

Table 2.

Multiple linear regressions examining race and income’s association with % body fat and four obesogenic behaviors

Variable Ref % Body Fat 95% CI (MVPA) mins 95% CI Sedentary (mins) 95% CI HEI-2010 95% CI Screen Time (mins) 95% CI
Model 1
 Constant 30.0 29.1 30.8 42.7 41.3 44.1 497 487.8 506.2 44.6 43.9 45.3 293 283.9 302.2
 Medium Low-income −0.7 −1.8 0.3 −3.3 −5.1 −1.5 −0.4 −11.9 11.2 −1.1 −1.9 −0.2 −33.9 −45.4 −22.4
 High Low-income −2.2 −3.5 −0.8 −3.1 −5.5 −0.7 6.7 −8.7 22.2 0.3 −0.9 1.4 −62.9 −78.3 −47.4
Model 2
 Constant 28.4 25.9 30.8 37.1 33.3 40.9 502.7 477.6 527.8 42.9 41.1 44.7 287.2 262.4 311.9
 Medium Low-income white 0.4 −2.2 3.1 −1 −5.3 3.4 −11.7 −40.3 16.8 −0.7 −2.8 1.3 −55.1 −83.1 −27.1
 High Low-income white −1.1 −3.9 1.7 0.7 −3.9 5.4 −3.8 −34.3 26.7 2.2 0 4.4 −94.1 −124 −64.1
 High Medium-income white −1.5 −3.4 0.4 1.7 −1.6 5 8.5 −13.1 30.1 2.9 1.3 4.5 −39 −60.3 −17.6
 Hispanic White 3.6 0.9 6.3 2.8 −1.6 7.1 −4.1 −32.7 24.4 4.4 2.7 5.8 2 −26.2 30.2
 Black White −0.2 −2.9 2.6 10.8 6 15.5 −9.2 −38 19.7 −0.7 −2.8 1.3 12 −16.4 40.5
 Black Hispanic −3.8 −2.9 2.6 7.6 4.6 10.6 −5 −24.8 14.7 −5.2 −6.6 −3.7 10 −9.5 29.6
Hispanic-x-medium-incomea Low-income white −1.5 −4.5 1.5 0.5 −4.6 5.6 1.2 −32.3 34.6 −1.4 −3.8 1 28.2 −4.9 61.2
Hispanic-x-high-incomea Low-income white −0.4 −0.4 3.3 −0.1 −6.6 6.4 10.4 −32.3 53.1 −3.5 −6.6 −0.4 66.9 24.7 109
Black-x-medium-incomea Low-income white −0.7 −3.8 2.4 −4.5 −9.7 0.7 25.3 −8.6 59.3 1.2 −1.3 3.6 33.8 0.2 67.3
Black-x-high-incomea Low-income white −1.1 −4.8 2.6 −5.8 12.15 0.5 15.47 26.19 57.1 −1 −4 2 75.8 34.7 117
Open in a new tab

Bold values indicate significant findings (p < 0.05)

Models controlled for age and sex

a

Interaction between income level and racial/ethnic group

Model 2 used the income-to-poverty ratio as the independent variable while adding race, race/ethnicity-by-income interactions, age, and sex as covariates. Results of the linear model indicated that the proportion of body fat for Hispanic children and adolescents were significantly less (3.6 95% CI 0.9, 6.3) when compared to their White peers. No race/ethnicity-by-income interactions reached statistical significance in the MVPA regression model. For HEI, there was a statistically significant race/ethnicity-by-income interaction for high-income Hispanic children and adolescents (−3.5 HEI-2010 95% CI −6.6, −0.4). The Hispanic by high-income (66.9 min 95% CI 24.7, 109.0), Black by medium-income (33.8 min 95% CI 0.2, 67.3), and Black by high-income (75.8 min 95% CI 34.7, 117.0) interactions were all statistically significant for screen time.

In the third model (Table 3), stratifying groups by race, results indicated that income was not associated with body adiposity. Within the logistic regression models, no other significant associations were found. There were also no significant associations between income and MVPA for both Hispanics and White children and adolescents. Medium-income Black children and adolescents engaged in significantly less MVPA (−5.9 min 95% CI −8.9, −2.8) than their low-income counterparts. High-income Black children and adolescents also engaged in significantly less MVPA (−5.4 min 95% CI −10.1, −0.7) than their low-income counterparts. Across all racial/ethnic groups, there were no significant associations between income and sedentary time. For the HEI, medium-income Hispanic children and adolescents had poorer diet quality (−2.1 HEI-score 95% CI −3.4, −0.8) than their low-income counterparts. High-income White children and adolescents had significantly healthier diet quality (2.9 HEI-2010 score 95% CI 1.2, 4.6) than their low-income White counterparts. There were no significant associations between income and diet when restricting the model to just Black children and adolescents. Medium-income Hispanic children and adolescents used screens significantly less (−26.8 min 95% CI −44.9, −8.6) than their low-income counterparts. Both medium and high-income White children and adolescents also used screens significantly less than their low-income counterparts (−54.9 min 95% CI −82.4, −27.5) and (−94.3 min 95% CI 123.6, −65.0) respectively. Medium-income White children and adolescents also used screens significantly less (−39.3 min 95% CI −60.2, −18.5) than their low-income counterparts. Medium-income Black children and adolescents used screens significantly less (−20.9 min 95% CI −39.0, −2.9) than their low-income Black counterparts.

Table 3.

Additions linear regressions examining race and income’s association with % body fat and four obesogenic behaviors stratified by race/ethnicity

Model 3 Ref Hispanic
 White
 Black
Coef 95% CI Coef 95% CI Coef 95% CI
% body fat
 Constant 31.8 30.6 33 28.5 26.1 30.9 28.2 26.8 29.7
 Medium-income Low-income −0.9 −2.4 0.6 0.3 −2.4 2.9 −0.3 −2.0 1.5
 High-income Low-income −1.3 −3.6 1.0 −1.2 −4.0 1.5 −2.2 −4.8 0.4
 High-income Medium-income −0.4 −2.6 1.8 −1.5 −3.2 0.2 −1.9 −4.4 0.5
MVPA (mins)
 Constant 39.8 37.9 41.9 36.7 33.1 40.4 47.8 45.4 50.1
 Medium-income Low-income −0.4 −2.9 2.2 −0.6 −4.7 3.6 −5.9 −8.9 −2.8
 High-income Low-income 0.6 −3.7 5.0 1.2 −3.3 5.6 −5.4 −10.1 −0.7
 High-income Medium-income 1.0 −3.2 5.2 1.8 −1.4 4.9 0.5 −4.0 5.0
Sedentary time (mins)
 Constant 498.4 485.2 511.6 503.1 476.3 529.9 493.5 479.4 507.6
 Medium-income Low-income −9.1 −26.2 7.9 −10.9 −41.3 19.4 14.5 −4 33.0
 High-income Low-income 8.6 −20.3 37.4 −2.3 −34.8 30.1 13.4 −14.9 41.7
 High-income Medium-income 17.7 −10.2 45.6 8.6 −14.5 31.7 −1.2 −28.5 26.1
HEI-2010 score
 Constant 47.3 46.3 48.3 42.9 40.9 44.8 42.2 41.2 43.1
 Medium-income Low-income −2.1 −3.4 −0.8 −0.7 −2.9 1.5 0.4 −0.9 1.7
 High-income Low-income −1.3 −3.5 0.9 2.2 −0.1 4.5 1.1 −0.8 3.1
 High-income Medium-income 0.8 −1.4 2.9 2.9 1.2 4.6 0.7 −1.1 2.6
Screen time (mins)
 Constant 289.1 275 303.2 287.5 263.2 311.7 299.5 285.8 313.3
 Medium-income Low-income −26.8 −44.9 −8.6 −54.9 −82.4 −27.5 −20.9 −39 −2.9
 High-income Low-income −27.3 −58.1 3.5 −94.3 −123.6 −65 −18.1 −45.8 9.5
 High-income Medium-income −0.5 −30.3 29.3 −39.3 −60.2 −18.5 2.8 −23.8 29.4
Open in a new tab

Bold values indicate significant findings (p < 0.05)

Models controlled for age and sex

In the final model (Table 4.), stratifying groups by each income-to-poverty-ratio group, regression results indicated that the proportion of body adiposity for low-income Black children and adolescents were significantly less (−3.6 95% CI −5.4, 1.8) when compared to low-Income Hispanic children and adolescents. The proportion of body adiposity for medium-income Black children and adolescents was significantly less (−3.1 95% CI −4.5, −1.7) when compared to medium-income Hispanic children and adolescents. The proportion of body adiposity for high-income Hispanic children and adolescents was significantly more (3.2 95% CI 1.0, 5.5) when compared to high-income white children and adolescents. The proportion of body adiposity for high-income Black children was significantly less (OR 0.6 95% CI 0.5, 0.8) when compared to high-income Hispanic children and adolescents. Low-income Black children and adolescents engaged in significantly more MVPA than low-income White (10.5 min 95% CI 5.9, 15.9) and Hispanic (7.5 min 95% CI 4.4, 10.6) children and adolescents. When restricting the model to medium-income children and adolescents, Hispanics engaged in significantly more MVPA (3.3 min 95% CI 0.7, 5.9) than White children and adolescents. Medium-income Black children and adolescents also engaged in more MVPA than medium-income White (5.9 min 95% CI −3.3, 8.7) and Hispanic (2.7 min 95% CI 0.2, 5.2) children and adolescents. When restricting the model to high-income, Black children and adolescents engaged in significantly more MVPA (4.6 min 95% CI 0.2, 9.0) than high-income White children and adolescents. Within low-income children and adolescents, there were no significant association between race and sedentary time. Medium-income Black children and adolescents engaged in significantly more sedentary time (19.8 min 95% CI −3.3, 36.2) than their medium-income Hispanic counterparts. There were no significant associations between race and sedentary time within the high-income group. When analyzing diet quality for low-income children and adolescents only, Hispanics had significantly healthier diet quality (4.4 HEI-2010 score 95% CI 2.4, 6.4) than their low-income White counterparts. Low-income Black children and adolescents had significantly poorer diet quality (−5.2 HEI-2010 score 95% CI −6.6, −3.8) compared to low-income Hispanic children and adolescents. Medium-income Hispanic children and adolescents had significantly healthier diet quality (3.0 min 95% CI 1.8, 4.3) than medium-income White children and adolescents. Medium-income Black children and adolescents had significantly poorer diet quality (−2.6 min 95% CI −3.8, −1.4) than their medium-income Hispanic counterparts. There were no significant associations between race and diet quality within the high-income group. There were also no significant associations between race and screen time within the low-income group. Medium-income Hispanic children and adolescents used screens significantly more (30.1 min 95% CI −13.1, 47.0) than their medium-income White counterparts. When compared to medium-income White children and adolescents, medium-income Black children and adolescents used screens significantly more (44.9 min 95% CI 27.3, 62.5). High-income Hispanic children and adolescents used screens significantly more (68.9 min 95% CI 38.8, 99.1) than high-income White children and adolescents. High-income Black children and adolescents also used screens significantly more (88.1 min 95% CI 59.5, 116.7) than high-income White children and adolescents.

Table 4.

Additions linear regressions examining race and income’s association with % body fat and four obesogenic behaviors stratified by income level

Model 4 Ref Low-income
 Medium-income
 High-income
Coef 95% CI Coef 95% CI Coef 95% CI
% body fat
 Constant 28.5 25.8 31.1 28.8 27.5 30.0 27.3 26.0 28.6
 Hispanic White 3.3 0.3 6.2 2.1 0.6 3.7 3.2 1.0 5.5
 Black White −0.3 −3.3 2.7 −0.9 −2.5 0.7 −1.4 −3.6 0.9
 Black Hispanic −3.6 −5.4 1.8 −3.1 −4.5 −1.7 −4.6 −7.2 −2.0
MVPA (mins)
 Constant 36.9 32.9 40.8 36.1 34.1 38.1 37.9 35.3 40.4
 Hispanic White 3 −1.5 7.5 3.3 0.7 5.9 2.7 −1.9 7.4
 Black White 10.5 5.9 15 5.9 3.2 8.7 4.6 0.2 9.0
 Black Hispanic 7.5 4.4 10.6 2.7 0.2 5.2 1.9 −3.4 7.3
Sedentary time (mins)
 Constant 504.4 477.9 530.9 492.3 479 505.5 500.9 485.3 516.5
 Hispanic White −5.5 −35.7 24.5 −2.9 −20.3 14.4 6.1 −22.9 35
 Black White −10.2 −40.5 20.2 16.8 −1.1 34.8 5.7 −21.8 33.2
 Black Hispanic −4.6 −25.4 16.2 19.8 3.3 36.2 −0.3 −33.6 32.9
HEI-2010 score
 Constant 42.9 41.2 44.7 42.9 41.2 43.1 45.1 43.8 46.4
 Hispanic White 4.4 2.4 6.4 3 1.8 4.3 0.9 −1.5 3.4
 Black White −0.8 −2.8 1.2 0.5 −0.8 1.8 −1.7 −4 0.6
 Black Hispanic −5.2 −6.6 −3.8 −2.6 −3.8 −1.4 −2.7 −5.5 0.1
Screen time (mins)
 Constant 287.5 261.7 313.4 232.5 219.5 245.5 193.1 176.9 209.3
 Hispanic White 1.5 −27.9 30.9 30.1 13.1 47 68.9 38.8 99.1
 Black White 12.1 −17.6 41.8 44.9 27.3 62.5 88.1 59.5 116.7
 Black Hispanic 10.5 −9.9 30.9 14.9 −1.3 31 19.2 −15.4 53.8
Open in a new tab

Bold values indicate significant findings (p < 0.05)

Models controlled for age and sex

Discussion

This study examined the association between income, race, four obesogenic behaviors (MVPA, sedentary time, diet quality, and screen time), and the proportion of adiposity using two waves of national data. With regards to adiposity, this study found, consistent with past work utilizing portions the same data [46], income was protective against the odds of being overweight/obese. We also found that, regardless of race/ethnicity, screen time was lower in middle and high-income households. Interestingly, this study also found that there was an inverse association between income and MVPA. No clear patterns were evident for diet quality or sedentary time by income level. Income appears to have a protective effect on screen use among this population. However, significant race by income interactions demonstrate the associations between income and the obesogenic behaviors measured varied by race/ethnic group. Decreasing sedentary behavior, such as screen use, is a potential intervention lever to reduce overweight and obesity in low-income populations.

When stratifying models by race/ethnicity group, increasing income was associated with a statistically significant decrease in screen time for White children and adolescents. Both children and adolescents who were Hispanic and Black were trending in a similar direction, but the relationships did not reach statistical significance across all income levels. It appears that the association between income and screen time in model 1 were largely driven by White children and adolescents. Table 3 also highlights that Black children appear to be decreasing their time spent in MVPA as income level increases. This association would also explain the findings in model 1 that MVPA was inversely associated with income. When restricting the models to each race/ethnicity group, there were not consistent patterns by racial/ethnic group and income with any other obesogenic behavior. With the exception of decreasing sedentary behaviors such as screen time in White children and adolescents, no patterns of relationships were evident within racial/ethnic groups and across income levels.

When stratifying models by income group, Black children and adolescents engaged in statistically significantly more MVPA than their White counterparts across all income levels. Black children and adolescents also engaged in statistically significantly more MVPA than Hispanics in both low- and medium-income groups. It appears here that the inverse associations between income and MVPA seen in model 1 were largely driven by Black children and adolescents. Across both low- and medium-income levels, Hispanic children and adolescents had healthier diet quality when compared to their White and Black counterparts. Both Black and Hispanic children and adolescents used screens significantly more as income increased when compared to Whites. This finding informs us that the protective effects of income and screen seen in model 1 were largely driven by White children and adolescents. These findings suggest that future interventions need to be tailored to individual racial/ethnic groups and children’s household income level income level of the children.

The findings presented herein do not clarify the behavioral mechanisms that appear to be driving the disparities in overweight/obesity or adiposity levels by race/ethnicity and income level that have been previously noted [18, 19, 49]. According to nationally representative samples of U.S. children and adolescents drawn from the same datasets that this study is based upon, Hispanic and Black children and adolescents have the highest prevalence rates of overweight/obesity measured by BMI [44]. Contradictory to the studies previously mentioned, we found that Black children and adolescents have the lowest proportions of adiposity when compared to both White and Hispanic children and adolescents. The present study’s findings do not help to explain these earlier findings, as children and adolescents who are Hispanic had the healthiest diet quality according to HEI-2010. In addition to diet, the obesogenic behaviors of Hispanic children and adolescents did not consistently differ from other race/ethnic groups. Similar to previous findings [53], this study found that Black children and adolescents were the most active. Thus, it is unclear why overweight and obesity is more prevalent in these groups when their behaviors and adiposity levels do not seem to indicate increased risk.

Explanations for these unclear relationships between obesogenic behaviors and overweight and obesity risk are broad and potentially complex. First, the disparities in overweight/obesity may be limited by the measurement tool. BMI is an indicator of excess weight, not excess adiposity. Thus, differences in BMI-determined prevalence of overweight and obesity may be due to differences in lean mass rather than fat mass. Several studies have shown that BMI may be a poor measure of obesity in minority children and adolescents [12, 15, 17, 42, 57]. If this is the case, future studies of obesity by race/ethnicity should employ an alternative measure of obesity in order to better link obesogenic behaviors to unhealthy weight in children and adolescents. Alternatively, the current measures of obesogenic behaviors may not be precise enough to elucidate differences that are driving disparate rates of obesity. In this study children and adolescents reported on their diets for only 2 days. Further, MVPA and sedentary time estimates were based on 7 days of accelerometer wear, and screen time was recalled for 30 days prior. The brevity and complexity of some of these time frames for measurements may not fully capture the long-term behaviors of these children and adolescents or how these behaviors change over time. Recently reliability studies of youth physical activity suggest a minimum of 10 valid days are required to assess youth PA [3]. Finally, sleep was not measured in the current study; however, sleep duration and quality have been associated with obesity in youth [29, 47]. Thus, disparate rates of obesity may be driven by behaviors that are not included in this study.

There are several strengths of this study including the large representative sample and the analysis of four obesogenic behaviors by income and race. This study also employed DXA scan as a measure of adiposity. DXA has been identified as a benchmark measure of adiposity in children and adolescents [24] Appropriate statistical analyses were conducted with follow up stratified analysis. However, this study is not without limitation. The cross-sectional nature of this study limits the ability to make casual inferences about the relationship between income, race, and the obesogenic behaviors of children and adolescents. Further, the limited number of valid cases in other racial/ethnic groups (e.g., Asian and Native American children and adolescents), this study was only able to explore the relationship of income and race/ethnicity in children and adolescents who are White, Black, and Hispanic. Finally, this study used a single indicator of social disparity (i.e., family income) which may incorrectly classify certain families’ socioeconomic status. Future studies should aim to collect multiple indicators of social disparity. Future studies should also work to understand if there are differences between race, income, and locality. Liu and colleagues found that rural children and adolescents were persistently more likely to be overweight and obese, but in some cases engaged in more healthful behaviors than their urban peers [36]. Seasonality also needs to be considered in future studies given the recent finding that summer break is a period of accelerated weight gain [41, 54, 56].

Conclusion

The findings presented herein did not clearly identify patterns of obesogenic behaviors by race/ethnicity and income among children and adolescents. Additional research is needed using alternative measures of obesity and increased behavioral measurement protocols. Studies employing longitudinal designs need to examine developmental trajectories in obesogenic behaviors and their relationship with risk for obesity by race/ethnicity and income.

Funding

The first author reported in this publication was supported by the Eunice Kennedy Shriver National Institute of Child Health & Human Development of the National Institutes of Health under Award Number R21HD090647. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Footnotes

Conflict of interest The authors declare that they have no conflict of interest.

References

  • 1.Albano JD, Ward E, Jemal A, Anderson R, Cokkinides VE, Murray T, Henley J, Liff J, & Thun MJ (2007). Cancer mortality in the United States by education level and race. Journal of the National Cancer Institute, 99(18), 1384–1394. [DOI] [PubMed] [Google Scholar]
  • 2.Anand J, Raper NR, & Tong A (2006). Quality assurance during data processing of food and nutrient intakes. Journal of Food Composition and Analysis, 19, S86–S90. [Google Scholar]
  • 3.Barreira TV, Schuna J, Tudor-Locke C, Chaput J-P, Church TS, Fogelholm M, Hu G, Kuriyan R, Kurpad A, & Lambert EV (2015). Reliability of accelerometer-determined physical activity and sedentary behavior in school-aged children: A 12-country study. International Journal of Obesity Supplements, 5(S2), S29. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Barriuso L, Miqueleiz E, Albaladejo R, Villanueva R, Santos JM, & Regidor E (2015). Socioeconomic position and childhood-adolescent weight status in rich countries: A systematic review, 1990–2013. BMC Pediatrics, 15(1), 129. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Bucholz EM, Ma S, Normand S-LT, & Krumholz HM (2015). Race, socioeconomic status, and life expectancy after acute myocardial infarction. Circulation, 10.1161/CIRCULATIONAHA.115.017009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Campbell CR, Turner-Lomperis AM, Gillespie KN, & Arrington B (2006). Competency-based healthcare management education: The Saint Louis University experience. Journal of Health Administrative Education, 23(2), 135–168. [PubMed] [Google Scholar]
  • 7.Cavalcanti RB, Cheung AM, Raboud J, & Walmsley S (2005). Reproducibility of DXA estimations of body fat in HIV lipodystrophy: Implications for clinical research. Journal of Clinical Densitometry, 8(3), 293–297. [DOI] [PubMed] [Google Scholar]
  • 8.CDC National Center for Health Statistics. National Health and Nutrition Examination Survey. [Google Scholar]
  • 9.Chang VW, & Lauderdale DS (2005). Income disparities in body mass index and obesity in the United States, 1971–2002. Archives of Internal Medicine, 165(18), 2122–2128. [DOI] [PubMed] [Google Scholar]
  • 10.Costa-Font J, & Gil J (2013). Intergenerational and socioeconomic gradients of child obesity. Social Science & Medicine, 93, 29–37. [DOI] [PubMed] [Google Scholar]
  • 11.Dailey AB, Kasl SV, Holford TR, Lewis TT, & Jones BA (2010). Neighborhood- and individual-level socioeconomic variation in perceptions of racial discrimination. Ethnicity & Health, 15(2), 145–163. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Daniels SR, Khoury PR, & Morrison JA (1997). The utility of body mass index as a measure of body fatness in children and adolescents: Differences by race and gender. Pediatrics, 99(6), 804–807. [DOI] [PubMed] [Google Scholar]
  • 13.Department of Health and Human Services. Department of Health and Human Services Poverty Guidelines for 2017. (2018). [Google Scholar]
  • 14.El-Sayed AM, Scarborough P, & Galea S (2012). Socioeconomic inequalities in childhood obesity in the United Kingdom: A systematic review of the literature. Obesity Facts, 5(5), 671–692. [DOI] [PubMed] [Google Scholar]
  • 15.Ellis KJ, Abrams SA, & Wong WW (1999). Monitoring childhood obesity: Assessment of the weight/height2 index. American Journal of Epidemiology, 150(9), 939–946. [DOI] [PubMed] [Google Scholar]
  • 16.Evenson KR, Catellier DJ, Gill K, Ondrak KS, & McMurray RG (2008). Calibration of two objective measures of physical activity for children. Journal of Sports Sciences, 26(14), 1557–1565. [DOI] [PubMed] [Google Scholar]
  • 17.Flegal KM, Ogden CL, Yanovski JA, Freedman DS, Shepherd JA, Graubard BI, & Borrud LG (2010). High adiposity and high body mass index–for-age in US children and adolescents overall and by race-ethnic group. The American Journal of Clinical Nutrition, 91(4), 1020–1026. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Freedman DS, Ogden CL, Flegal KM, Khan LK, Serdula MK, & Dietz WH (2007). Childhood overweight and family income. Medscape General Medicine, 9(2), 26. [PMC free article] [PubMed] [Google Scholar]
  • 19.Freedman DS, Wang J, Thornton JC, Mei Z, Pierson RN, Dietz WH, & Horlick M (2008). Racial/ethnic differences in body fatness among children and adolescents. Obesity, 16(5), 1105–1111. [DOI] [PubMed] [Google Scholar]
  • 20.Geronimus AT, Hicken M, Keene D, & Bound J (2006). “Weathering” and age patterns of allostatic load scores among blacks and whites in the United States. American Journal of Public Health, 96(5), 826–833. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Grøntved A, Ried-Larsen M, Møller NC, Kristensen PL, Wedderkopp N, Froberg K, Hu FB, Ekelund U, & Andersen LB (2014). Youth screen-time behaviour is associated with cardiovascular risk in young adulthood: The European Youth Heart Study. European Journal of Preventive Cardiology, 21(1), 49–56. [DOI] [PubMed] [Google Scholar]
  • 22.Guenther PM, Casavale KO, Reedy J, Kirkpatrick SI, Hiza HA, Kuczynski KJ, Kahle LL, & Krebs-Smith SM (2013). Update of the healthy eating index: HEI-2010. Journal of the Academy of Nutrition and Dietetics, 113(4), 569–580. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Guenther PM, Kirkpatrick SI, Reedy J, Krebs-Smith SM, Buckman DW, Dodd KW, Casavale KO, & Carroll RJ (2013). The Healthy Eating Index-2010 is a valid and reliable measure of diet quality according to the 2010 Dietary Guidelines for Americans1–3. The Journal of Nutrition, 144(3), 399–407. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Gutin B, Litaker M, Islam S, Manos T, Smith C, & Treiber F (1996). Body-composition measurement in 9–11-y-old children by dual-energy X-ray absorptiometry, skinfold-thickness measurements, and bioimpedance analysis. The American Journal of Clinical Nutrition, 63(3), 287–292. [DOI] [PubMed] [Google Scholar]
  • 25.Hansen K, Joshi H, & Dex S (2010). Children of the 21st century: The first five years (Vol. 2). New York: Policy Press. [Google Scholar]
  • 26.Hernandez DC, & Pressler E (2014). Accumulation of childhood poverty on young adult overweight or obese status: race/ethnicity and gender disparities. Journal of Epidemiology Community Health, 10.1136/jech-2013-203062. [DOI] [PubMed] [Google Scholar]
  • 27.Hsu FC, Lenchik L, Nicklas BJ, Lohman K, Register TC, Mychaleckyj J, Langefeld CD, Freedman BI, Bowden DW, & Carr JJ (2005). Heritability of body composition measured by DXA in the diabetes heart study. Obesity Research, 13(2), 312–319. [DOI] [PubMed] [Google Scholar]
  • 28.Jackson JW, Williams DR, & VanderWeele TJ (2016). Disparities at the intersection of marginalized groups. Social Psychiatry and Psychiatric Epidemiology, 51(10), 1349–1359. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Jarrin DC, McGrath JJ, & Drake CL (2013). Beyond sleep duration: Distinct sleep dimensions are associated with obesity in children and adolescents. International Journal of Obesity, 37(4), 552. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Johnson CL, Paulose-Ram R, Ogden CL, Carroll MD, Kruszan-Moran D, Dohrmann SM, & Curtin LR (2013). National health and nutrition examination survey. Analytic guidelines, 1999–2010. [PubMed] [Google Scholar]
  • 31.Jones-Smith JC, Dieckmann MG, Gottlieb L, Chow J, & Fernald LC (2014). Socioeconomic status and trajectory of overweight from birth to mid-childhood: The early childhood longitudinal study-birth cohort. PLoS One. 9(6), e100181. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Kinra S, Nelder RP, & Lewendon GJ (2000). Deprivation and childhood obesity: a cross sectional study of 20 973 children in Plymouth, United Kingdom. Journal of Epidemiology & Community Health, 54(6), 456–460. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Koplan JP, Liverman CT, & Kraak VI (2005). Preventing childhood obesity: Health in the balance: executive summary. Journal of the American Dietetic Association, 105(1), 131–138. [DOI] [PubMed] [Google Scholar]
  • 34.Krieger N, & Sidney S (1996). Racial discrimination and blood pressure: The CARDIA Study of young black and white adults. American Journal of Public Health, 86(10), 1370–1378. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Leonard CM, Roza MA, Barr RD, & Webber CE (2009). Reproducibility of DXA measurements of bone mineral density and body composition in children. Pediatric Radiology, 39(2), 148–154. [DOI] [PubMed] [Google Scholar]
  • 36.Liu J-H, Jones SJ, Sun H, Probst JC, Merchant AT, & Cavicchia P (2012). Diet, physical activity, and sedentary behaviors as risk factors for childhood obesity: An urban and rural comparison. Childhood Obesity (Formerly Obesity and Weight Management), 8(5), 440–448. [DOI] [PubMed] [Google Scholar]
  • 37.Lobstein T, Jackson-Leach R, Moodie ML, Hall KD, Gortmaker SL, Swinburn BA, James WP, Wang Y, & McPherson K (2015). Child and adolescent obesity: Part of a bigger picture. The Lancet, 385(9986), 2510–2520. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Matthews CE, Chen KY, Freedson PS, Buchowski MS, Beech BM, Pate RR, & Troiano RP (2008). Amount of time spent in sedentary behaviors in the United States, 2003–2004. American Journal of Epidemiology, 167(7), 875–881. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.McGuire S. (2011). US department of agriculture and US department of health and human services, dietary guidelines for Americans, 2010. Washington, DC: US government printing office, January 2011. Advances in Nutrition: An International Review Journal, 2(3), 293–294. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Miech RA, Kumanyika SK, Stettler N, Link BG, Phelan JC, & Chang VW (2006). Trends in the association of poverty with overweight among US adolescents, 1971–2004. JAMA, 295(20), 2385–2393. [DOI] [PubMed] [Google Scholar]
  • 41.Moreno JP, Johnston CA, & Woehler D (2013). Changes in weight over the school year and summer vacation: results of a 5-year longitudinal study. Journal of School Health, 83(7), 473–477. [DOI] [PubMed] [Google Scholar]
  • 42.Nelson DA, & Barondess DA (1997). Whole body bone, fat and lean mass in children: comparison of three ethnic groups. American Journal of Physical Anthropology: The Official Publication of the American Association of Physical Anthropologists, 103(2), 157–162. [DOI] [PubMed] [Google Scholar]
  • 43.Ng M, Fleming T, Robinson M, Thomson B, Graetz N, Margono C, Mullany EC, Biryukov S, Abbafati C, & Abera SF (2014). Global, regional, and national prevalence of overweight and obesity in children and adults during 1980–2013: A systematic analysis for the Global Burden of Disease Study 2013. The Lancet, 384(9945), 766–781. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Ogden CL, Carroll MD, & Flegal KM (2008). High body mass index for age among US children and adolescents, 2003–2006. Jama, 299(20), 2401–2405. [DOI] [PubMed] [Google Scholar]
  • 45.Ogden CL, Carroll MD, Lawman HG, Fryar CD, Kruszon-Moran D, Kit BK, & Flegal KM (2016). Trends in obesity prevalence among children and adolescents in the United States, 1988–1994 through 2013–2014. Jama, 315(21), 2292–2299. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Ogden CL, Lamb MM, Carroll MD, & Flegal KM (2010). Obesity and Socioeconomic Status in Children and Adolescents: United States, 2005–2008. NCHS Data Brief. vol. 51. USA: National Center for Health Statistics. [PubMed] [Google Scholar]
  • 47.Olds TS, Maher CA, & Matricciani L (2011). Sleep duration or bedtime? Exploring the relationship between sleep habits and weight status and activity patterns. Sleep, 34(10), 1299–1307. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Rubin DB (2004). Multiple imputation for nonresponse in surveys (Vol. 81). London: Wiley. [Google Scholar]
  • 49.Sobal J, & Stunkard AJ (1989). Socioeconomic status and obesity: A review of the literature. Psychological Bulletin, 105(2), 260–275. 10.1037/0033-2909.105.2.260. [DOI] [PubMed] [Google Scholar]
  • 50.Stamatakis E, Hamer M, Tilling K, & Lawlor DA (2012). Sedentary time in relation to cardio-metabolic risk factors: Differential associations for self-report vs accelerometry in working age adults. International Journal of Epidemiology, 41(5), 1328–1337. [DOI] [PubMed] [Google Scholar]
  • 51.Stamatakis E, Wardle J, & Cole TJ (2010). Childhood obesity and overweight prevalence trends in England: evidence for growing socioeconomic disparities. International journal of obesity, 34(1), 41. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Survey NH Physical activity and physical fitness-PAQ-C. [Google Scholar]
  • 53.Troiano RP, Berrigan D, Dodd KW, Mâsse LC, Tilert T, & McDowell M (2008). Physical activity in the United States measured by accelerometer. Medicine and Science in Sports and Exercise, 40(1), 181. [DOI] [PubMed] [Google Scholar]
  • 54.von Hippel PT, & Workman J (2016). From kindergarten through second grade, US children’s obesity prevalence grows only during summer vacations. Obesity, 24(11), 2296–2300. [DOI] [PubMed] [Google Scholar]
  • 55.Wang Y, & Lobstein T (2006). Worldwide trends in childhood overweight and obesity. International Journal of Pediatric Obesity, 1(1), 11–25. [DOI] [PubMed] [Google Scholar]
  • 56.Wang YC, Vine S, Hsiao A, Rundle A, & Goldsmith J (2015). Weight-related behaviors when children are in school versus on summer breaks: Does income matter? Journal of School Health, 85(7), 458–466. [DOI] [PubMed] [Google Scholar]
  • 57.Weber DR, Moore RH, Leonard MB, & Zemel BS (2013). Fat and lean BMI reference curves in children and adolescents and their utility in identifying excess adiposity compared with BMI and percentage body fat. The American Journal of Clinical Nutrition, 98(1), 49–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Williams DR, Mohammed SA, Leavell J, & Collins C (2010). Race, socioeconomic status, and health: Complexities, ongoing challenges, and research opportunities. Annals of the New York Academy of Sciences, 1186(1), 69–101. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES