The Retail Food Environment in Relation to Socioeconomic Characteristics, Weight Status and Diabete

The food environment is drawing increasing attention as an important population-level determinant of diet, obesity and related health outcomes. Using retail food data sourced from a commercial database, we examined several dimensions of the local retail food environment in relation to area socio-economic characteristics, and weight status and diabetes risk among adult residents of three urban regions in southern Ontario.;The first study showed that local access to different types of food retail was patterned by level of neighbourhood material deprivation. More deprived neighbourhoods generally provided better access to stores and restaurants of all types, including those selling more or less healthful foods. These patterns were partially explained by urban form factors. Relative access to unhealthy food retailers (as a proportion of all outlets) showed little variation by level of neighbourhood material deprivation.;In the second study, we found that measures of absolute and relative availability of different types of restaurants (volume and proportion) within walking distance of residential areas had differential effects on the weight status of local residents. We also identified a novel interaction between absolute and relative restaurant measures, whereby exposure to a higher proportion of fast-food relative to all restaurants was directly related to excess weight, particularly in areas with high volumes of fast-food restaurants (e.g. odds ratio for obesity = 2.55 in areas with 5+ fast-food restaurants, 95% confidence interval: 1.55--4.17, across the interquartile range).;The third study showed a similar synergistic effect between absolute and relative dimensions of fast-food restaurant exposure in relation to the development of diabetes. Among younger adults (20--65 years), a greater proportion of fast-food restaurants was directly associated with incident diabetes after adjustment for individual- and area-level covariates, but only in areas with high volumes of fast-food outlets (hazard ratio = 1.79, 95% confidence interval: 1.03--3.12, across the interquartile range). No significant associations were observed in areas with low volumes of fast food or among older adults.;This dissertation contributes to a better understanding of how different aspects of the retail food environment relate to health, which may help to guide the design of programs and policies to create healthier and more equitable food environments.