Abstract MP080: Contributions of Social, Behavioral, and Biological Factors to Racial Disparities in Incident Obesity: The Coronary Artery Risk Development in Young Adults (CARDIA) Study

Abstract

Background: A combination of social, behavioral, and biological factors are hypothesized to account in large part for racial disparities in obesity, but few studies have shown this empirically. Thus, the goal of this study was to determine the extent to which these factors accounted for racial differences in the development of obesity in men and women over a 30-year period (mean follow-up 21.7 years). Methods: This study includes the 1,617 Black (743 men and 874 women) and 1,950 White (928 men and 1,022 women) CARDIA participants with complete data on all baseline covariates who were not obese at baseline (1985-1986; ages 18-30). A series of gender-stratified Cox proportional hazards models were run adjusting separately for socioeconomic, psychosocial, neighborhood, lifestyle, and physiological factors to determine the extent to which they explained race differences in incident obesity. Results: Among men, 36.9% of Whites and 43.7% of Blacks became obese over follow-up. In women, 33.7% of Whites and 60.3% of Blacks became obese. After adjusting for age and baseline BMI, Black men were 1.39 (95% confidence interval (CI): 1.19, 1.62) times more likely to become obese than White men (Table). Covariate adjustment reduced race differences by 34%; neighborhood factors were the strongest contributors. Black women were 1.63 (95% CI: 1.42, 1.88) times more likely to become obese than White women in the base model. This association was reduced by 57% in fully adjusted models. Lifestyle factors explained the largest proportion of the race difference, followed by individual-level socioeconomic factors and neighborhood conditions. Conclusions: Racial disparities in incident obesity, and the extent to which these disparities were explained by baseline covariates, differ for men and women. Whereas substantial attenuation occurred after adjustment for a wide variety of baseline covariates, race differences in obesity incidence remained significant.