Obesity. where individual and environmental context converge

Background: More information is needed to examine the relationship between individual and environmental factors contributing to obesity. Objective: To analyze the interactions between individual level obesity and county level obesity prevalence in the U.S. Methods: Cross-sectional study of 94693 from the Adventist Health Study-2. Public domain obesity data at the county level were merged with individual-level data from the AHS-2. Individual-level data on obesity, socio-demographics and life style factors of interest (e.g., diet and physical activity) were extracted from the AHS-2 cohort. Three types of obesity prevalence estimates were employed in the analysis: (1) county-level estimates produced by the CDC for the general population; (2) county-level estimates based on the AHS-2 subject-level BMI data; and (3) individual-level estimates for the AHS-2 subjects based on individual BMI. Descriptive mapping and spatial analyses were implemented to explore associations between (2) and (1). Logistic regression was used to assess the association between (3) and (1). Analysis was adjusted for socio-demographics and other relevant factors. Results: Spatial analyses showed that obesity prevalence in the AHS-2 followed geographically that of the background population (P<0.001). Covariate adjusted logistic regression analysis confirmed these results and additionally showed that individual level physical activity >45 min/day (OR 0.7, P<0.001) and plant rich dietary patterns (OR 0.5, P<0.001) had an independent effect in lowered risk of obesity on an individual level. Conclusions: Individuals are more likely to be obese in environments with a higher obesity prevalence. However, individual lifestyle practices such as physical activity and dietary choices can modulate obesity outcomes in high risk environments.