Impact of Gender, Individual/Neighborhood Socioeconomic Status (SES), and Race/Ethnicity on Body Mass Index Trajectories among the U.S. Middle-aged and Older Adults

Introduction: Obesity has become epidemic in the U.S. while adult men and women showed different obesity prevalence growth rates (29% and 7.2% respectively) in 2000-2010. As the aging population grows, obesity and related chronic diseases will increase costs of social welfare programs like Medicare. Literature is inconsistent on the directions/independence of associations between individuals’ and neighborhood sociodemographic characteristics and obesity.

Methods: The datasets including Health and Retirement Study respondents who had lived in the same Census tract in 2000-2010 (N=6,261) and 2000/2010 U.S. Census were analyzed. Three-level hierarchical growth model with individual and state variations was fit to model BMI trajectories using HLM 7 based on the aggregated census tract level demographic variables.

Results: BMI has a significant increase over time following a quadratic decelerating trajectory (p < 0.001). Relative to women, men have higher BMI at baseline (b = 0.43, p = 0.001), and have higher growth rate (b = 0.024, p < 0.001). Non-Hispanic White people have significantly lower initial BMI (b = -1.48, p < 0.001); however, the growth rate is not significantly different from the other race/ethnicity groups. Neighborhood SES has impact on BMI at baseline (b = 0.25, p < 0.01) with no difference of the growth rates among states. The cross-level interaction between neighborhood SES and personal race/ethnicity is not significant while individual variables have stronger main effects.

Discussion: The individuals’ race/ethnicity and neighborhood SES have independent influence on BMI trajectories. The interventions may need to target on middle-aged and older male individuals.