Abstract
This cross-sectional study uses data from the National Health and Nutrition Examination Surveys to evaluate obesity trends among US adolescents from 1999 to 2018, stratified by household income and head of household education level.
The high prevalence of obesity among US adolescents is a public health concern,1 as obesity is a risk factor for cardiovascular disease.2 An association between obesity and low socioeconomic status (SES) was observed in a 2021 study of trends among US adults.3 However, obesity trends by SES among adolescents have not been fully described since 2008.4 We evaluated obesity trends among US adolescents from 1999 to 2018, stratified by household income and head of household education level.
Methods
The Kyoto University Institutional Review Board approved this cross-sectional study. Data were extracted from the 1999-2018 National Health and Nutrition Examination Surveys (NHANES). Written informed consent was obtained from adults (aged ≥18 years), and parental consent and child assent were obtained for adolescents (aged 10-17 years) for NHANES participation. This study followed the STROBE reporting guideline.
Obesity among adolescents (aged 10-19 years) was defined as age- and sex-specific body mass index greater than or equal to the 95th percentile based on the 2000 Centers for Disease Control and Prevention growth charts.2 Individuals were stratified by household income (≤138% vs >138% federal poverty level)5 and head of household education level. Race and ethnicity data were included owing to their association with obesity prevalence among US children and adolescents.
We first described 1999-2018 trends in obesity prevalence in 4-year increments stratified by income and education. We then investigated the association of SES and obesity prevalence using ordinary least-squares regressions with robust SEs,6 adjusting for race and ethnicity, height, and marital status of the head of household. After fitting the models, we estimated obesity prevalence under hypothetical exposure levels (income and education) with the observed distribution of covariates at each 4-year cycle. We included additive interaction terms between SES and each cycle (as a categorical variable) to evaluate changes over time in socioeconomic differences in obesity prevalence using Wald tests. We also examined linear trends in socioeconomic differences in obesity prevalence using the 4-year cycle as an ordinal variable. Sampling weights were used to account for the NHANES design in all analyses. Analyses were conducted using R, version 4.1.1.
Results
Of 21 296 individuals in this study, the mean (SD) age was 14.5 (2.8) years; 49.3% were female. Information on household income and head of household education levels was available for 19 465 (91.0%) and 20 302 (95.0%), respectively (Table). Adolescents from low-SES households were more likely to be non-Hispanic Black (21.7%), have obesity (22.8%), or have an unmarried parent (45.5%) (Table). The trend in adjusted obesity prevalence increased over 20 years, particularly among adolescents from low-SES households (Figure). Living in a low-income household was associated with a 4.2–percentage point increase in obesity prevalence (95% CI, 2.4-5.9), and lower head of household education level was associated with a 9.0–percentage point increase (95% CI, 7.2-10.7). The gap in obesity prevalence between adolescents from low-income households vs others was 6.4 percentage points greater (95% CI, 1.5-11.4) in 2015-2018 vs 1999-2002.
Table. Characteristics of 1999 to 2018 National Health and Nutrition Examination Survey Participants Aged 10 to 19 Years in This Study .
| Variable | Household income level | Head of household education level | ||
|---|---|---|---|---|
| >138% FPL (n = 10 734) | ≤138% FPL (n = 8731) | Graduated college or above (n = 3351) | Did not graduate college (n = 16 951) | |
| Age, mean (SD), y | 14.3 (2.7) | 14.6 (2.9) | 14.1 (2.7) | 14.5 (2.8) |
| Sex, % | ||||
| Female | 48.4 | 50.9 | 49.6 | 49.3 |
| Male | 51.6 | 49.1 | 50.4 | 50.7 |
| Household income, %a | ||||
| ≤138% FPL | NA | NA | 33.5 | 5.9 |
| >138% FPL | NA | NA | 66.5 | 94.1 |
| Head of household education level, % | ||||
| Graduated college or above | 33.5 | 5.9 | NA | NA |
| Did not graduate college | 66.5 | 94.1 | NA | NA |
| Race and ethnicity, %b | ||||
| Hispanic | 13.4 | 30.6 | 7.2 | 23.7 |
| Non-Hispanic Black | 10.4 | 21.7 | 8.4 | 16.5 |
| Non-Hispanic White | 68.4 | 39.6 | 74.0 | 52.7 |
| Otherc | 7.8 | 8.0 | 10.3 | 7.1 |
| Head of household marital status, % | ||||
| Married | 77.0 | 45.5 | 79.9 | 62.4 |
| Unmarried | 23.0 | 54.5 | 20.1 | 37.6 |
| Body mass index, mean (SD)d | 22.6 (5.6) | 23.7 (6.3) | 21.6 (4.9) | 23.4 (6.1) |
| Height, mean (SD), cm | 162.6 (12.5) | 161.2 (12.1) | 162.7 (12.6) | 161.9 (12.2) |
| Obesity, % | 17.3 | 22.8 | 11.6 | 21.8 |
Abbreviations: FPL, federal poverty level; NA, not applicable.
Low income was defined as household income below 138% of the FPL. Among 21 296 adolescents included in the survey, 19 465 (91.0%) and 20 302 (95.0%) had information on household income and head of household education levels, respectively.
Some percentages do not total 100% because of rounding.
Indicates individuals who reported their race and ethnicity as “other race–including multi-racial” on the National Health and Nutrition Examination Surveys.
Body mass index was calculated as weight in kilograms divided by height in meters squared.
Figure. Trends in Obesity Prevalence Among US Adolescents Aged 10 to 19 Years by Household Income and Head of Household Education Level, 1999 to 2018.
Obesity prevalence was adjusted for race and ethnicity, height, and marital status of the head of household. Obesity was defined as age- and sex-specific body mass index in the 95th percentile or greater based on the 2000 Centers for Disease Control and Prevention growth charts. Plots are adjusted differences in obesity prevalence between groups and changes in the differences over time (4-year cycles) relative to 1999 to 2002. Sampling weights were used to account for the National Health and Nutrition Examination Survey design and produce data representative of the general US population. FPL indicates federal poverty level.
We found a similar trend for education, with 4.2–percentage point greater prevalence (95% CI, –0.8 to 9.3) for individuals with lower head of household education levels in 2015-2018 vs 1999-2002. When we assessed linear trends, the gap in obesity prevalence by income and education increased by an average of 1.5 (95% CI, 0.4-2.6) and 1.1 (95% CI, 0.0-2.3) percentage points every 4 years, respectively.
Discussion
The findings of this cross-sectional study suggest that socioeconomic disparities existed in obesity prevalence among US adolescents during 1999-2018, building on a previous study using NHANES 1999-2008 data,4 and suggest that socioeconomic disparities in obesity have widened during the last 2 decades. Obesity during adolescence can have immediate health consequences and long-term outcomes in adulthood.2 Accordingly, the larger obesity prevalence among adolescents from lower-SES households may exacerbate socioeconomic disparities in chronic diseases into adulthood.
Study limitations include potential unmeasured confounding and misclassification due to self-reported SES. Future studies should assess strategies to reduce socioeconomic disparities in obesity among US adolescents and evaluate their long-term health consequences.
References
- 1.Skinner AC, Ravanbakht SN, Skelton JA, Perrin EM, Armstrong SC. Prevalence of obesity and severe obesity in US children, 1999-2016. Pediatrics. 2018;141(3):e20173459. doi: 10.1542/peds.2017-3459 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Grossman DC, Bibbins-Domingo K, Curry SJ, et al. ; US Preventive Services Task Force . Screening for obesity in children and adolescents: US Preventive Services Task Force recommendation statement. JAMA. 2017;317(23):2417-2426. doi: 10.1001/jama.2017.6803 [DOI] [PubMed] [Google Scholar]
- 3.He J, Zhu Z, Bundy JD, Dorans KS, Chen J, Hamm LL. Trends in cardiovascular risk factors in US adults by race and ethnicity and socioeconomic status, 1999-2018. JAMA. 2021;326(13):1286-1298. doi: 10.1001/jama.2021.15187 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Ali MK, Bullard KM, Beckles GL, et al. Household income and cardiovascular disease risks in U.S. children and young adults: analyses from NHANES 1999-2008. Diabetes Care. 2011;34(9):1998-2004. doi: 10.2337/dc11-0792 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Decker SL, Kostova D, Kenney GM, Long SK. Health status, risk factors, and medical conditions among persons enrolled in Medicaid vs uninsured low-income adults potentially eligible for Medicaid under the Affordable Care Act. JAMA. 2013;309(24):2579-2586. doi: 10.1001/jama.2013.7106 [DOI] [PubMed] [Google Scholar]
- 6.Hellevik O. Linear versus logistic regression when the dependent variable is a dichotomy. Qual Quan. 2009;43(1):59-74. doi: 10.1007/s11135-007-9077-3 [DOI] [Google Scholar]