Differential impacts of short-term PM2.5 constituent exposure on type 2 diabetes mellitus: a nationwide case-crossover study with advanced statistical insights
摘要
The associations between short-term exposure to PM2.5 constituents and diabetes remain inconsistent and understudied. We conducted a nationwide case-crossover study across 153 hospitals in Chinese mainland. Daily average concentrations of five major PM2.5 constituents, including black carbon (BC), organic matter (OM), sulfate (SO42−), nitrate (NO3−), and ammonium (NH4+), were assessed by satellite-based models. We applied single-constituent analysis and inverse-probability-of-treatment weighting (IPTW) to assess each constituent’s independent effect. In addition, we developed a weighted quantile sum (WQS) regression for paired data with multiple controls to examine the joint impacts of the constituents and quantify each one’s contribution. The single-constituent analysis suggested that an interquartile range increase in each of the five constituents was significantly associated with elevated risk of T2DM hospitalization, the effects ranging from 1.49%–2.14%. After adjusting for other constituents using IPTW methods, only OM and SO42− remained significant associations with T2DM. The WQS analysis further confirmed SO42− and OM as the dominant contributors. Our findings highlight the importance of targeted air pollution control measures for T2DM prevention.