Plasma metabolomic signatures of air pollution and rheumatoid arthritis risk

Ambient air pollutants have been increasingly implicated in immune-mediated disorders, yet the metabolic perturbations through which long-term exposure may contribute to rheumatoid arthritis (RA) remain insufficiently characterized.

We used data from 403,332 RA-free participants in the UK Biobank cohort to investigate the associations among long-term ambient air pollution exposure, plasma metabolic signatures, and incident RA risk. Participants were followed for a median of 14.01 years. Residential annual average concentrations of PM_2.5, PM₁₀, NO₂, and NO_x were linked to participants' addresses, and a composite air pollution score (APS) was calculated to reflect overall exposure burden. Plasma metabolomics data were used to characterize metabolic perturbations related to each pollutant. A machine learning model was applied to identify air pollutant-associated metabolites and construct corresponding metabolic signatures. We then examined the relationships of air pollutants and derived metabolic signatures with RA risk using Cox proportional hazards models. Generalized propensity score analyses were performed to assess whether the findings persisted after covariate balancing, and mediation analyses were conducted to estimate the contribution of metabolic signatures to the relationship of air pollutants and RA.

Each standard deviation increase in the ${\text{PM}}_{2.5^{-}}$ , ${\text{NO}}_{2^{-}}$ , ${\text{NO}}_{{\text{x}}^{-}}$ , and APS-related metabolic signatures was associated with elevated RA risk, with hazard ratios of 1.208 (95% CI: 1.180, 1.236), 1.141 (95% CI: 1.109, 1.173), 1.174 (95% CI: 1.148, 1.201), and 1.184 (95% CI: 1.156, 1.212), respectively. Similar estimates were observed in generalized propensity score analyses. Metabolic signatures accounted for 15.7% to 21.1% of the associations between air pollution exposure and RA risk.

These findings suggest that long-term air pollution exposure is associated with incident RA risk and that systemic metabolic perturbations may represent potential biological pathways underlying this association.