Abstract
BACKGROUND: High obesity rates among young black and Hispanic children place them at a higher risk for adult obesity and its comorbidities. Neighborhoods with predominately racial and ethnic minority residents have fewer healthful food options, which may contribute to obesity disparities. Few studies have assessed the relationship between neighborhood food environments and obesity in this population. METHODS: Electronic health records from 2 pediatric primary care clinics serving predominately low-income, black, and Hispanic children were used to create a cohort of 3724 2- to 5-year olds, encompassing 7256 visits from 2007 to 2012 (mean 1.9 visits per patient, range: 1-5 visits per child). Longitudinal regression was used to model the association of mean body mass index z-score (BMI-z) over time and 3 measures of the neighborhood food environment: healthful food availability, availability of stores accepting the Special Supplemental Nutrition Program for Women, Infants and Children (WIC) benefits, and fast food availability. RESULTS: Compared to peers in neighborhoods with no or few stores accepting WIC, children in neighborhoods with many WIC stores had higher BMI-z at age 2 years (average difference of 0.272; 95% confidence interval: 0.041-0.503; P = .021). No relationship was found for healthful food or fast food availability. Although children in neighborhoods with low fast food availability did not have statistically significantly different BMI-z at age 2 as compared to children in areas with high fast food availability, they did have a statistically significantly higher change in average BMI-z over time (0.006 per month, 0.000-0.012, P = .024). CONCLUSIONS: Access to WIC stores was associated with lower obesity rates and more healthful average BMI-z over time and represents a potentially important neighborhood food environment characteristic influencing racial/ethnic disparities in childhood obesity among young black and Hispanic children. More studies are needed to assess what aspects of WIC stores may underlie the observed association.