Spatial patterns of obesity and associated risk factors in the conterminous u.s.

Abstract
BACKGROUND: The obesogenic environment is hypothesized to increase obesity risk by discouraging physical activity and limiting the availability of healthy food.

PURPOSE: This research reports the prevalence of obesity and risk factors (physical activity and fruit and vegetable consumption) by creating spatially smoothed maps and analyzing local autocorrelation and aims to examine associations of obesity and risk factors at the national level.

METHODS: Data were obtained in 2008 from the Behavioral Risk Factor Surveillance System for the years 2000-2006 aggregated to the county level. A weighted head-banging smoothing algorithm was used that effectively replaced the proportion of obesity and risk factors for each county with a weighted median that incorporates data from neighboring counties. Significant spatial clusters of obesity and risk factors were identified by a local Moran’s I analysis. All analyses were performed in 2008-2009.

RESULTS: A higher prevalence of obesity was generally found in the non-metro counties of the South, whereas lower prevalence was found in the West and the Northeast. A lower prevalence of leisure-time physical activity was generally found in the areas where obesity prevalence was higher and vice versa. A lower prevalence of fruit and vegetable consumption was found mainly in the non-metro counties of the South and the Great Plains.

CONCLUSIONS: The national patterns of obesity and associated risk factors obtained may reflect a unique set of meso-environmental drivers, including climate, land use, population density, and culture. Future research should address this regional variability and explicitly consider the spatial scales at which such environmental factors operate.