A national study of the association between food environments and county-level health outcomes.

PURPOSE: This national, county-level study examines the relationship between food availability and access, and health outcomes (mortality, diabetes, and obesity rates) in both metro and non-metro areas.
METHODS: This is a secondary, cross-sectional analysis using Food Environment Atlas and CDC data. Linear regression models estimate relationships between food availability and access variables (direct-to-consumer farm sales, per capita grocery stores, full-service restaurants, fast food restaurants, and convenience stores) with health outcomes. Controls include smoking, race/ethnicity, gender, age, education, poverty, primary care availability, recreational facility availability, and mobility/distance-from-grocery-store.
FINDINGS: Non-metro findings: Lower adjusted mortality rates were associated with more per capita full-service restaurants and grocery stores, and greater per capita direct farm sales. Lower adjusted diabetes rates were associated with a lower per capita supply of fast food restaurants and convenience stores, and more per capita full-service restaurants and grocery stores. Lower adjusted obesity rates were associated with more per capita full-service restaurants and grocery stores. Unexpectedly, obesity rates were positively associated with per capita grocery stores and negatively associated with fast food restaurants. Metro findings: More per capita full-service restaurants, grocery stores, and direct farm sales are associated with positive health outcomes; fast food restaurants and convenience stores are associated with negative health outcomes.
CONCLUSIONS: The food access/availability environment is an important determinant of health outcomes in metro and non-metro areas. Future research should focus on more refined specifications that capture variability across non-metro settings.