ARTICLE: Trends in Obesity Prevalence Among Adults Aged 18 Through 25 Years, 1976-2018
AUTHORS: Alejandra Ellison-Barnes, Sara Johnson, Kimberly Gudzune
JOURNAL: JAMA. 2021 Nov 23;326(20):2073-2074. doi: 10.1001/jama.2021.16685.
Abstract
Methods: We used nationally representative data from the National Health and Nutrition Examination Survey (NHANES), a series of cross-sectional surveys including interviews and physical examinations with a stratified, multistage sampling design. We drew from NHANES II (1976-1980), NHANES III (1988-1994), and the continuous NHANES cycles from 1999 through 2018 (response rate range, 48.8%-80%). We limited our study population to nonpregnant emerging adults (aged 18-25 years) of non-Hispanic Black or non-Hispanic White race with complete data for variables of interest (89.8% of defined population had complete data). Given the changes in how the NHANES assessed race and ethnicity over time, we were limited to the aforementioned groups.
Our outcome of interest was body mass index (BMI; calculated as weight in kilograms divided by height in meters squared). We categorized BMI into standard groups of underweight (<18.5), normal weight (18.5-24.9), overweight (25-29.9), and obesity (≥30).5 Covariates included sex (male or female), race and ethnicity (non-Hispanic Black or non-Hispanic White), and household poverty (yes or no). We identified household poverty if the NHANES poverty index (the ratio of family income to the US poverty threshold in NHANES II/III and the ratio of family income to the US poverty guidelines in the continuous NHANES) was at or below the poverty threshold or guidelines for that year. We did not include age, given the narrow range of interest.
For each wave, we estimated mean BMI and prevalence of each BMI group along with 95% CIs using NHANES examination weights to account for the complex sampling design and adjusting for all covariates. To test for trend across all years, we used a nonparametric, Wilcoxon-type test given the time differences between surveys.6 In sensitivity analyses, we combined the continuous NHANES cycles (1999-2018) and conducted a linear or logistic regression, as appropriate, to test for trend over time. Statistical significance was determined by a 2-sided P < .05. All analyses were performed using Stata version 15 (StataCorp). The Johns Hopkins institutional review board reviewed this study and determined that it was not human subjects research.
Results: Across all years, 8015 emerging adults were included. Of these, 3965 were female, 3037 were non-Hispanic Black, and 2386 met criteria for household poverty.
Between 1976 and 2018, mean BMI increased from 23.1 (95% CI, 22.9-23.4) in 1976-1980 to 27.7 (95% CI, 26.2-29.1) in 2017-2018 (P = .006 for trend using a nonparametric test) (Table). The Figure shows the adjusted prevalence of BMI groups over time. Between 1976-1980 and 2017-2018, the prevalence of obesity increased from 6.2% (95% CI, 4.9%-7.9%) to 32.7% (95% CI, 24.7%-41.8%; P = .007 for trend using a nonparametric test), whereas normal weight decreased from 68.7% (95% CI, 66.3%-70.9%) to 37.5% (95% CI, 29.5%-46.4%; P = .005 for trend using a nonparametric test). The sensitivity analyses limited to the continuous NHANES cycles had similar results (Table).
For the full article, click here.
