Examining the associations between the food environment and dietary intake in British Columbia: A cross-sectional study

This study aimed to characterize neighbourhood food environments in British Columbia (BC) and to determine if food environment characteristics were associated with fruits and vegetables (FV) intake. A cross-sectional study was conducted using geospatial food environment (Can-FED) and neighbourhood environment measures from the BC Generations Project, a cohort of males and females aged 35-69. Food environment measures (chain grocery stores, fast food outlets, convenience stores, and two relative measures of food outlets) were categorized according to density: class 0 (no presence) to class 4 (highest density). Multivariable mixed-effect models adjusted for sociodemographics and neighbourhood characteristics were constructed to assess the associations between food environments and FV intakes as a continuous variable and as <5 or ≥ 5 servings of FV/day. Approximately 50% of participants lived in neighbourhoods without chain grocery stores, fast food outlets or convenience stores within walking distance. An inverse association between neighbourhoods with the highest density of fast-food restaurants and FV consumption was observed (Odds ratio (OR) =0.89, 95% CI: 0.80, 0.98). Associations between chain grocery stores, convenience stores and FV consumption were attenuated with adjustment for neighbourhood characteristics: walkability, material and social deprivation. The findings suggest the presence of food deserts across urban areas in BC. Participants living in neighbourhoods with greater density of fast-food restaurants were less likely to consume ≥5 servings of FV per day. Further studies are needed to understand the lack of associations between other characteristics of the food environment and dietary intake.