Associations between racialized economic segregation and retail food environments in the United States: an observational, cross-sectional study

Abstract

Historic policies segregating US cities have shaped inequities in contemporary retail food environments (RFEs). This study investigated associations between segregation and RFEs. Using publicly available data from the American Community Survey, we calculated Index of Concentration at the Extremes (ICE) measures for racial, economic, and racialized economic segregation (i.e., ICE (Race), ICE (Income), ICE (Race + Income)) at the census tract (CT) and ZIP code tabulation area (ZCTA) levels. At the CT level, we used zero-inflated beta models with random effects to characterize associations between segregation and the modified retail food environment index (mRFEI). At the ZCTA level, we used negative binomial hurdle models with random effects to characterize associations between segregation and numbers of grocery retailers (overall, large and small), full-service restaurants, and limited-service restaurants. We observed evidence suggestive of a positive dose-response relationship between all ICE measures and the mRFEI, such that CTs with more privileged populations tended to have healthier RFEs. We found associations with ICE (Race + Income) were null for grocery retailers (overall, large and small), but a positive relationship emerged with density of full-service restaurants (i.e., more privilege, more outlets), and the most privileged areas had significantly more limited-service restaurants. Greater ICE (Race) (i.e., greater racial privilege) was associated with greater numbers of grocery retailers, full-service restaurants, and limited-service restaurants. ICE (Income) was not associated with the number of food retailers, though the least privileged quintile consistently had the most retailers. Overall, these findings highlight a potential pathway between US segregation and inequitable health outcomes via neighborhood RFEs.