Ozone and Fine Particulate Matter Components of Air Pollution Are Associated with Systemic Lupus Erythematosus Risk

Objectives Concerns are growing regarding the relationship of air pollution (especially fine particulate matter, PM2.5) in diseases like systemic lupus erythematosus (SLE). It is unclear which chemical components of ambient PM2.5 may be most harmful, and whether other air pollutants play additional roles. We aimed to evaluate the association between the mixture of PM2.5 components and SLE onset, quantifying their relative contributions to SLE risk, and potential effect modification by ambient ozone levels.

Methods Using MarketScan® administrative health data, we assembled an urban open cohort of all enrollees ≥18-year-old (without prior SLE) with residential core-based statistical area (CBSA) information. Each year after 2013, eligible individuals entered the cohort and were followed until SLE onset, death, insurance disenrollment, or study end (Dec. 2023). SLE incident cases were identified by ≥1 hospitalization or ≥2 physician billing diagnostic codes. From the cohort, all SLE cases and a 20% random baseline sub-cohort were combined into a case-cohort sample. Concentrations of PM2.5 components (ammonium, black carbon, mineral dust, sulfate, nitrate, organic matter, sea salt) and ambient ozone for 2 years before cohort entry were estimated by satellite- and ground-based models and assigned based on CBSAs at cohort entry. Extended quantile g-computation models assessed potential associations of SLE onset with the mixture of PM2.5components, ozone and their interaction, adjusting for sex, age, baseline chronic obstructive pulmonary disease (as a proxy for smoking), geographic region, and year of cohort entry. Index weights estimated by quantile g-computation models quantified the relative contributions of individual PM2.5 components to SLE risk.

Results Our case-cohort sample numbered 8,345,067 individuals including 21,485 new SLE cases. At the median ozone referent level (ie, 36.1 parts per billion), the adjusted hazard ratio for SLE onset was 1.142 (95% confidence interval, CI 1.107-1.179) per every quartile increase in all PM2.5 components (Table 1). Ozone was also associated with increased risk of SLE (HR 1.009, 95% CI 1.001-1.017). There was effect modification such that the HR for PM2.5 was highest when ozone level was lowest (Table 1). Similar results were seen in sub-groups stratified by sex or age. Mineral dust consistently had the largest index weight across different sub-groups and ozone levels.

Conclusion PM2.5 and ozone were associated with SLE onset; mineral dust was an important contributor. Mineral dust triggers pulmonary inflammation and is a plausible trigger of autoimmunity and SLE onset. Addressing sources of ambient mineral dust (road traffic, construction, farming) may help reduce SLE incidence.