Abstract P118: County-Level Disparities in the Availability of Farmers Markets in the United States: An Ecological Study

Abstract

Introduction: Farmers markets (FM) have been hypothesized to be a potential community-level obesity prevention strategy for populations at risk for chronic diseases because they provide a mechanism for communities to purchase healthy locally grown produce. This study aimed to identify county-level factors associated with FM availability in an effort to determine if disparities in availability exist in the US. Hypothesis: Increased FM availability will be associated with higher median household income, lower % minority residents, lower % obese residents and a higher number of grocery stores and recreation centers per 100,000 residents. Methods: An ecological study was conducted using 2009 data from the USDA Food Environment Atlas on 3,135 US counties. Crude and multivariable adjusted logistic regression models where used to determine associations between having at least one FM available and county-level variables such as % African American (AA) residents, % Hispanic residents, median household income, % WIC participants, % adults obese, % adults with diabetes, per capita grocery stores, per capita supercenters and per capita recreation centers. All analyses were stratified by metro county status and adjusted to address data clustering at the state-level. Results: There were 1,088 and 2,047 counties labeled metro and non-metro respectively. Metro Results : Median household income (p = 0.002) and per capita recreation centers (p < 0.0001) were positively associated with FM availability while % WIC residents (p = 0.008), per capita grocery stores (p = 0.02) and % adults with diabetes (p < 0.0001) showed a negative association. Non-Metro Results: Median household income (p < 0.0001), per capita recreation centers (p < 0.0001) and per capita supercenters (p < 0.0001) were positively associated with FM availability while % WIC residents (p = 0.02), per capita grocery stores (p < 0.0001) and % adults with diabetes (p < 0.03) showed a negative association. The % AA residents appeared to be negatively associated with FM availability but did not achieve statistical significance. County obesity prevalence was not associated with FM availability in both metro and non-metro counties. Conclusion: Results showed that counties with more recreation centers and a higher median household income have increased FM availability while counties with more WIC participants and residents with diabetes have less availability. More information on the association between FM access, diet and obesity in at risk populations should be collected at the individual level.