Aditi Krishna1, Fahad Razak1, Alexandre Lebel1, George Davey Smith1, S V Subramanian1. 1. From the Department of Social and Behavioral Sciences, Harvard School of Public Health, Boston, MA (AK and SVS); the Department of Medicine, University of Toronto, Toronto, Canada (FR); the Li Ka Shing Knowledge Institute and Division of General Internal Medicine, St. Michael's Hospital, Toronto, Canada (FR); the Harvard Center for Population and Development Studies. Boston, MA (FR); Laval University, Quebec, Canada (AL); the Quebec Heart and Lung Institute Research Center, Quebec, Canada (AL); and the School of Social and Community Medicine, University of Bristol, Bristol, United Kingdom (GDS).
Abstract
BACKGROUND: Marked increases in mean body mass index (BMI) and prevalence of obesity and overweight in the United States are well known. However, whether these average increases were accompanied by changing dispersion (or SD) remains understudied. OBJECTIVE: We investigated population-level changes in the BMI distribution over time to understand how changes in dispersion reflect between-group compared with within-group inequalities in weight gain in the United States. DESIGN: Using data from the Behavioral Risk Factor Surveillance System survey (1993-2012), we analyzed associations between mean, SD, and median BMI and BMI at the 5th and 95th percentiles for 3,050,992 non-Hispanic white, non-Hispanic black, and Hispanic men and women aged 25-64 y. RESULTS: Overall, an increase of 1.0 in mean BMI (in kg/m²) was associated with an increase of 0.70 (95% CI: 0.67, 0.73) in the SD of BMI. A change of 1.0 in median BMI was associated with a change of 0.18 (95% CI: 0.14, 0.21) in the BMI value at the 5th percentile compared with a change of 2.94 (95% CI: 2.81, 3.07) at the 95th percentile. Quantile-quantile plots showed unequal changes in the BMI distribution, with pronounced changes at higher percentiles. Similar patterns were observed in subgroups stratified by sex, race-ethnicity, and education with non-Hispanic black women and women with less than a high school education having highest mean BMI, SD of BMI, and BMI values at the 5th and 95th percentiles. CONCLUSIONS: Mean BMI and the percentage of overweight and obese individuals do not fully describe population changes in BMI. Increases in within-group inequality in BMI represent an underrecognized characteristic of population-level weight gain. Crucially, similar increases in dispersion within groups suggest that growing inequalities in BMI at the population level are not driven by these socioeconomic and demographic factors. Future research should focus on understanding factors driving inequalities in weight gain between individuals.
BACKGROUND: Marked increases in mean body mass index (BMI) and prevalence of obesity and overweight in the United States are well known. However, whether these average increases were accompanied by changing dispersion (or SD) remains understudied. OBJECTIVE: We investigated population-level changes in the BMI distribution over time to understand how changes in dispersion reflect between-group compared with within-group inequalities in weight gain in the United States. DESIGN: Using data from the Behavioral Risk Factor Surveillance System survey (1993-2012), we analyzed associations between mean, SD, and median BMI and BMI at the 5th and 95th percentiles for 3,050,992 non-Hispanic white, non-Hispanic black, and Hispanic men and women aged 25-64 y. RESULTS: Overall, an increase of 1.0 in mean BMI (in kg/m²) was associated with an increase of 0.70 (95% CI: 0.67, 0.73) in the SD of BMI. A change of 1.0 in median BMI was associated with a change of 0.18 (95% CI: 0.14, 0.21) in the BMI value at the 5th percentile compared with a change of 2.94 (95% CI: 2.81, 3.07) at the 95th percentile. Quantile-quantile plots showed unequal changes in the BMI distribution, with pronounced changes at higher percentiles. Similar patterns were observed in subgroups stratified by sex, race-ethnicity, and education with non-Hispanic black women and women with less than a high school education having highest mean BMI, SD of BMI, and BMI values at the 5th and 95th percentiles. CONCLUSIONS: Mean BMI and the percentage of overweight and obese individuals do not fully describe population changes in BMI. Increases in within-group inequality in BMI represent an underrecognized characteristic of population-level weight gain. Crucially, similar increases in dispersion within groups suggest that growing inequalities in BMI at the population level are not driven by these socioeconomic and demographic factors. Future research should focus on understanding factors driving inequalities in weight gain between individuals.
Keywords:
BRFSS; Behavioral Risk Factor Surveillance System; body mass index; distributional change; health inequalities; obesity; socioeconomic status
Authors: L Lenchik; T C Register; G B Russell; J Xu; S C Smith; D W Bowden; J Divers; B I Freedman Journal: Osteoporos Int Date: 2018-05-31 Impact factor: 4.507
Authors: Arjumand Siddiqi; Faraz Vahid Shahidi; Vincent Hildebrand; Anthony Hong; Sanjay Basu Journal: Ann Epidemiol Date: 2018-02-15 Impact factor: 3.797
Authors: Milagros C Rosal; Christina F Haughton; Barbara B Estabrook; Monica L Wang; Germán Chiriboga; Oahn H T Nguyen; Sharina D Person; Stephenie C Lemon Journal: BMC Public Health Date: 2016-09-09 Impact factor: 3.295