Abstract
BACKGROUND: State-issued identification cards are a promising data source for neighborhood-level obesity estimates. METHODS: We used information from three million Oregon state-issued identification cards to compute age-adjusted estimates of average adult body mass index (BMI) for each census tract in the state. We used multivariate linear regression to identify associations between weight status and population characteristics, food access, commuting behavior, and geography. RESULTS: Together, home values, education, race, ethnicity, car commuting, and rural-urban commuting area (RUCA) explained 86% of the variation in BMI among tracts. BMI was lower in areas with higher home values and greater educational attainment, and higher in areas with more workers commuting by car. DISCUSSION: Our findings are consistent with other research on socioeconomic disparities in obesity. This demonstrates state-issued identification cards are a promising data source for BMI surveillance and may offer new insight into the association between weight status and economic and environmental factors. Public health agencies should explore options for developing their own obesity estimates from identification card data.