Extreme obesity is understood to have an outsized effect on health and mortality, yet little is known about the distribution of classes of obesity among populations with obesity. This study identifies counties in the United States that have a high proportion of individuals with extreme obesity among the general obese population and uses exploratory spatial data analysis to identify spatial clusters of high burden.
County-level prevalence estimates of moderate and extreme obesity created from the 2012 Behavioral Risk Factor Surveillance System and the Census Bureau were used for this analysis. The proportion of extreme obesity among the obese was calculated for each county. The Global Moran’s test assessed spatial dependence and local Moran’s I statistics were used for identifying clusters of higher and lower extreme obesity burden. A parallel coordinate plot was created to visually identify counties displaying unique patterns of class of obesity.
County-level proportion of extreme obesity among the obese ranged from 0.06 to 0.42 and the Moran’s I score (0.18) indicated significant but low spatial clustering. There were significant burden clusters in several regions including the Mississippi Delta region, the Carolinas, and Ohio’s rural northwest. Two notable patterns by obesity class were identified; counties showing both high prevalence of extreme obesity and obesity overall (‘double burdened’), and those with high prevalence of extreme obesity relative to lower prevalence of obesity overall (‘hidden burden’).
County-level extreme obesity burden in the United States demonstrated spatial dependence. Hot spots were identified indicating clusters of counties with a high proportion of extreme obesity among those with obesity. Results of this research indicate that current means of identifying counties with high rates of obesity may be missing high-risk counties with differential patterns of moderate and extreme obesity.