Ecotoxicology and Environmental Safety · May 11, 2026
Associations of short-term exposure to PM2.5 and its metal constituents with fasting plasma glucose in older adults: The mediating role of systemic inflammation
By Wu M, Yao H, Zhao D et al.
Fine particulate matter (PM2.5) exposure is linked to elevated fasting plasma glucose (FPG), but roles of metal constituents and mechanisms remain unclear. We examined cumulative lagged effects of short-term exposure to PM2.5 and its metal constituents on FPG in older adults, identified metal sources, and assessed whether systemic inflammation, measured by the systemic inflammation response index (SIRI), statistically mediated these associations. This 2024 longitudinal cohort enrolled 109 non-diabetic older adults from four communities in Hefei city, Anhui province, with baseline and three follow-ups measuring FPG and SIRI. PM2.5 and 19 metal constituents were estimated for the 1-7 days prior to each visit. Linear mixed-effects (LME) models were used to assess single-pollutant associations, while weighted quantile sum (WQS) regression and Bayesian kernel machine regression (BKMR) were applied to evaluate mixture effects. Positive matrix factorization (PMF) was used to apportion metal sources, and mediation modeling was employed to examine the potential mediating role of SIRI. At lag0-3 days, each 10 μg/m³ increase in PM2.5 was associated with a 4.921% rise in FPG (95% CI: 2.138%, 7.781%). Additionally, 14 metals showed positive associations, with uranium (U), selenium (Se), and manganese (Mn) exhibiting the largest effects. WQS and BKMR indicated that Mn and cesium (Cs) were the primary contributors to the positive mixture effect. PMF attributed Mn to fuel oil combustion and Cs to coal combustion. SIRI statistically mediated the associations between PM2.5, several metals (aluminum (Al), arsenic (As), barium (Ba), gallium (Ga), manganese (Mn)), and FPG. Short-term exposure to PM2.5 and its metal constituents, especially Mn, is associated with higher FPG in the elderly, and these associations are partially explained by systemic inflammation levels.
doi.org/10.1016/j.ecoenv.2026.120249Open accessPubMed 42114430Populations