Proteomic signatures of ambient air pollution and risk of type 2 diabetes.
TL;DR
It is suggested that air pollution might contribute to T2D risk through inflammation-related proteins, highlighting the potential of proteomics as a tool for precision public health.
Abstract
Despite accumulating evidence linking air pollution to type 2 diabetes (T2D), the underlying mechanisms remain largely unexplored. We aimed to profile proteomic signatures associated with air pollution and examine their relationship to T2D. We conducted proteome-wide association studies on 2911 plasma proteins among 49,134 UK Biobank participants. Exposures to fine particulate matter (PM2.5), nitrogen dioxide (NO2), sulfur dioxide (SO2), and benzene were estimated based on residential addresses. Proteomic signatures and their corresponding scores for each air pollutant were identified using linear and elastic net regression models, comprising 368 proteins for PM2.5, 207 for NO2, 206 for SO2, and 236 for benzene. Cox proportional hazards regression models were subsequently used to examine the effect of air pollution and proteomic signature scores on the risk of incident T2D. In both the time-independent and time-dependent Cox models, all four air pollutants were significantly associated with higher T2D risk. In the time-dependent Cox models, the hazard ratios (HRs) and 95% confidence intervals (CIs) were 1.02 (1.00, 1.05) for PM2.5, 1.02 (1.01, 1.02) for NO2, 1.12 (1.06, 1.18) for SO2, and 1.88 (1.38, 2.57) for benzene, respectively. Higher proteomic signature scores of PM2.5, NO2, SO2, and benzene were also associated with an elevated risk of T2D, with HRs (95% CIs) of 1.05 (1.00, 1.09), 1.17 (1.12, 1.23), 1.11 (1.06, 1.16), and 1.11 (1.06, 1.16) for a per-standard-deviation increase, respectively. Moderate mediation effects of the proteomic signature scores were observed. Pathway analyses further implicated systemic inflammation as a potential underlying mechanism. Our findings suggested that air pollution might contribute to T2D risk through inflammation-related proteins, highlighting the potential of proteomics as a tool for precision public health. Building on this, our study also offered a framework to explore molecular pathways connecting modifiable risk factors to diseases.