Geographic clustering of obesity, diabetes, and hypertension in Nashville, Tennessee.

To better understand how neighborhood environments contribute to health outcomes by encouraging or discouraging healthy lifestyles, we studied the clustering of health outcomes, health behaviors, and environmental characteristics in Nashville, Tenn. Data from 2 large telephone surveys (N = 7606) were geocoded and linked to the long-form census data at the census-tract level. Cluster analysis was used to create 12 scales that measured neighborhood characteristics using census variables. A geographic information system was used to map patterns of health, health behavior, and neighborhood characteristics at the census-tract level. Pearson correlations across the 129 census tracts were in the 0.20 to 0.60 range and provided evidence for spatial clustering of health outcomes, health behaviors, and neighborhood characteristics.