Multilevel Analysis of the Food and Physical Activity Environment and Adult Obesity Across U.S. Counties and States

Adult obesity rates have risen steadily across the United States over the past decade with more than 40% adults affected [1]. There exist persistent geographic and demographic disparities in obesity rates across the nation. While prior research has examined individual or environmental predictors of obesity, limited studies have addressed both physical activity and food environment across the nation using multilevel approaches. This study used a two-level random intercept model to assess the association between county- and state-level predictors and adult obesity prevalence across over 3,000 U.S. counties nested within 51 states. County-level predictors included food insecurity, poverty, unemployment, median household income, limited access to stores, and the density of various food outlets (grocery stores, convenience stores, supercenters, fast-food restaurants, SNAP-authorized retailers, and farmers’ markets), along with access to recreational facilities. State-level predictors included SNAP benefits per capita and the presence of soda and chip taxes. Variables were group-mean or grand-mean centered to distinguish within- and between-state effects. Results showed that food insecurity, poverty, unemployment, limited food access to stores, and the number of fast-food and convenience stores were positively associated with higher adult obesity rates. While higher recreational facility access, supercenter availability, median household income, and SNAP benefits per capita were associated with lower adult obesity prevalence, these associations varied in strength across counties and states. These results emphasize the need for place-based strategies that address both the physical activity and food environment in shaping obesity disparities.