Improving Healthy Food Access Spatial Analyses With Community Input and Retailer and Food Pantry Surveys

Abstract

ObjectivesHealthy food access can be analyzed spatially, but availability, affordability, desirability, and transport mode are among many potential impediments that have not been adequately measured in previous geographic information system (GIS) based food access studies. We aimed to identify and measure non-spatial factors to inform and improve the GIS/spatial analysis of healthy food access in three Health Equity Zone designated municipalities in Rhode Island. MethodsHere, healthy food was defined as produce. Three instruments were deployed: a community questionnaire; a survey of fresh and canned produce availability, general pricing, and produce quality at food retailers; and a survey of produce availability at food pantries. Retailer affordability was ranked using the collected pricing data and verified against community perceptions of affordable stores. Community input defined travel time parameters for mapping provider service areas for car, bus, and walking transportation. A novel bus access analysis incorporated walk, wait, and weighted drive times. Accessibility was mapped separately for regular-cost supermarkets, low-cost supermarkets, and food pantries. Results were compared to poverty, car ownership, and older households at the block group level to evaluate how neighborhoods most in need were being served. ResultsPricing data corresponded closely to community perceptions of store affordability, allowing us to categorize two of 16 supermarkets serving our study communities as expensive and to be excluded from spatial analyses. Five were categorized low-cost. Four out of 10 food pantries were excluded because they did not offer fresh produce. We found no convenience, corner, or dollar store met a priori criteria for a healthy food provider, with the threshold for fresh produce met in only one out of 48 of these stores. Two communities had no low-cost supermarkets accessible by car, bus, or foot. A limited number of people had bus or foot access to healthy food. ConclusionsHealthy food access analyses can be more accurate and useful when (1) travel time parameters are set by the community and the providers assessed for affordability, availability of fresh and canned produce, and quality, and (2), refined GIS techniques are used, including multi-step transit analysis and multiple layers of retailer and pantry service areas. Funding SourcesNone.