Ambient particulate air pollution (PM2.5) is associated with the ratio of type 2 diabetes to obesity.

We used county level data for T2D prevalence across the mainland USA and matched this to county level ambient PM2.5. Multiple linear regression was used to determine the relation between prevalence of T2D with PM2.5 after adjustment for confounding factors. PM2.5 explained 6.3% of the spatial variation in obesity, and 17.9% of the spatial variation in T2D. After correcting the T2D prevalence for obesity, race, poverty, education and temperature, PM2.5 still explained 8.3% of the residual variation in males (P < 0.0001) and 11.5% in females (P < 0.0001). The effect on obesity prevalence corrected for poverty, race education and temperature was much lower and hence the ratio of T2D to obesity prevalence was significantly associated with PM2.5 in males (R(2) = 11.1%, P < 0.0001) and females (R(2) = 16.8%, P < 0.0001). This association was repeated across non-African countries (R(2) = 14.9%, P < 0.0001). High levels of PM2.5 probably contribute to increased T2D prevalence in the USA, but have a more minor effect on the obesity. Exposure to high environmental levels of PM2.5 (relative to the USA) may explain the disproportional risk of T2D in relation to obesity in Asian populations.