Ambient air pollution may be associated with diabetes mellitus. However, evidence from developing countries is limited although the concentrations of air pollution are disproportionably higher in these countries. We collected daily data on diabetes mortality, air pollution, and weather conditions from 16 Chinese provincial cities during 2007-2013. A quasi-Poisson regression combined with a distributed lag model was used to quantify the city-specific mortality risk of PM10 (particulate matter with aerodynamic diameter < 10 μm). Then, a random-effect meta-analysis was conducted to pool effect estimates from 16 cities. We also calculated the attributable fraction and attributable number of diabetes mortality due to PM10. Effects of PM10 were found to be acute and limited to 3 days. Harvesting effect of PM10 was found during lag 4-10 days on diabetes mortality. An increase of 0.17% (95%CI: 0.01-0.34), 0.48% (95%CI: 0.22-0.73), and 0.53% (95%CI: 0.27-0.80) in diabetes mortality was associated with per 10 μg/m3 increase in PM10 at lag 0, 0-4 and 0-10 days, respectively. Totally, 5.76% (95%CI: 2.59-8.00%) and 5878 (95%CI: 2639-8163) deaths due to diabetes could be attributable to PM10. If the concentration of PM10 attained the Chinese government and WHO targets, the reduction in number of PM2.5-attributed diabetes deaths was 2016 and 5528, respectively. Higher effect estimates of PM10 were observed among females and those aged 0-64 years old at lag 0 day, while greater cumulative effects of PM10 were among males, the elderly aged 75 or over, and the illiterate at lag 0-10 days. However, the between-group differences were not statistically significant. It is one of the few studies on examining the attributable burden of diabetes mortality caused by particulate matter. Our findings indicated that effective efforts on controlling air pollution could reduce a prominent number of air pollution-related diabetes deaths.
Keywords: Air pollution; Attributable fraction; China; Diabetes mortality; Particulate matter.