Objective and Self-Reported Factors Associated With Food-Environment Perceptions and Fruit-And-Vegetable Consumption: A Multilevel Analysis

INTRODUCTION: Few studies have assessed how people’s perceptions of their neighborhood environment compare with objective measures or how self-reported and objective neighborhood measures relate to consumption of fruits and vegetables. METHODS: A telephone survey of 4,399 residents of Philadelphia, Pennsylvania, provided data on individuals, their households, their neighborhoods (self-defined), their food-environment perceptions, and their fruit-and-vegetable consumption. Other data on neighborhoods (census tracts) or “extended neighborhoods” (census tracts plus 1-quarter–mile buffers) came from the US Census Bureau, the Philadelphia Police Department, the Southeastern Pennsylvania Transportation Authority, and the federal Supplemental Nutrition Assistance Program. Mixed-effects multilevel logistic regression models examined associations between food-environment perceptions, fruit-and-vegetable consumption, and individual, household, and neighborhood characteristics. RESULTS: Perceptions of neighborhood food environments (supermarket accessibility, produce availability, and grocery quality) were strongly associated with each other but not consistently or significantly associated with objective neighborhood measures or self-reported fruit-and-vegetable consumption. We found racial and educational disparities in fruit-and-vegetable consumption, even after adjusting for food-environment perceptions and individual, household, and neighborhood characteristics. Having a supermarket in the extended neighborhood was associated with better perceived supermarket access (adjusted odds ratio for having a conventional supermarket, 2.04 [95% CI, 1.68–2.46]; adjusted odds ratio for having a limited-assortment supermarket, 1.28 [95% CI, 1.02–1.59]) but not increased fruit-and-vegetable consumption. Models showed some counterintuitive associations with neighborhood crime and public transportation. CONCLUSION: We found limited association between objective and self-reported neighborhood measures. Sociodemographic differences in individual fruit-and-vegetable consumption were evident regardless of neighborhood environment. Adding supermarkets to urban neighborhoods might improve residents’ perceptions of supermarket accessibility but might not increase their fruit-and-vegetable consumption.