Currently, $77 million is invested in a federal program, the Healthy Food Financing Initiative, in order to subsidize grocery store development in low-income areas without healthy food retailers. Despite significant investment, research examining dietary changes at the community level from the introduction of a grocery store is limited and mixed. Our objective is to evaluate whether a new supermarket with healthy food retail funding opening in January 2015 in the city of Philadelphia, Pennsylvania, an urban food desert, favorably influences dietary intake of residents. We predict that the new supermarket will improve eating and shopping behaviors, perceived access to food, and increase the total availability and affordability of foods in the community. Several methods, including multiple pretests, propensity score stratification, and use of an instrumental variable, improve causal inference over prior research. The study design consists of a prospective comparative interrupted time-series design based on the natural experiment of a new supermarket which will be built in a low-income, predominantly minority city. In order to obtain an adequate final sample size for both intervention (n=300) and comparison groups (n=300), 1310 men and women aged 18+ who identify themselves as the primary food shopper of the household and live within a 1 mile radius from the catchment area will be recruited for participation. Primary outcome measures and covariates include daily caloric intake (ASA-24), daily fruit and vegetable consumption (ASA-24), Healthy Eating Index score (HEI), shopping behaviors and customer demographics (NEMS-P), shopping residents'store distance and perceived access to fresh food (NEMS-P), availability and quality of foods (NEMS-S), and body mass index (BMI). To collect data, surveys will be administered and completed in waves at 4 periods: 9 months prior to store opening, 1 month before opening, 9 months after opening, and 21 months after opening. Each wave entails the completion of the ASA-24 twice, two weeks apart, as well as one administration of the NEMS-P. The NEMS-S will be administered in supermarkets and a random sample of 20 convenience stores at each wave. Data analysis of within/between-subjects repeated measures analysis and logistic regression for ASA-24 and NEMS surveys will be conducted. Uniquely positioned to inform upcoming policy decisions related to the future of HFFI, this study represents a unique and time sensitive opportunity to connect research to policy and practice.
The proposed study will provide important insight into ongoing national policy efforts, such as the Healthy Food Financing Initiative. The main purpose of this study is to evaluate whether a new supermarket opening in an urban food desert in January 2015, favorably influences dietary intake of residents.