Background/Purpose: The etiology of rheumatoid arthritis (RA) is multi-factorial, with expression of HLA-DRB1 shared epitope (SE), smoking and socioeconomic status (SES) exerting influence. Prior studies have evaluated the associations of air pollutant (AP) exposure with RA incidence, but not RA seropositivity or titers. We evaluated the associations of AP exposure with RA autoantibody status independent of SES, SE expression and tobacco exposure.

Methods: Patients from 7 sites of the Veterans Affairs Rheumatoid Arthritis (VARA) registry were included in the analysis. Socio-demographic, HLADRB1 SE status, tobacco exposure and RA seropositivity parameters (rheumatoid factor [RF], anti-cyclic citrullinated peptide [ACPA]) were available. Mean exposure levels for AP (NO₂, SO₂, Particulate matter {PM_2.5, PM₁₀} and Ozone) were obtained from local air quality monitoring stations based on the patient’s zip code for the year prior to VARA enrollment. Mean standardized scores of individual pollutants were used to calculate a global pollution score and patients were further grouped into pollution quartiles. For SES, the mean standardized score of a composite based on zip code was used, which included percent in labor force, percent unemployed, median household income, and percent below the poverty line, while years of education served as a surrogate. Generalized estimating equations (GEE), using the log of RF and ACPA titers with nested models based on VA location were used to determine independent associations of AP on RF and ACPA positivity and titers.

Results: There were 1078 Veterans (91.1% male, 80.4% Caucasian), with mean age of 69.4 +/- 10.4 years and disease duration of 12.5 +/- 11.7 years. HLADRB1 SE was positive in 73%, while 80.6% and 75.1% were RF and ACPA positive, respectively. Univariate analyses of the individual AP scores revealed no association with either RF or ACPA positivity. In multivariate GEE models, former and current smoking (p<0.0001), HLA-DR SE positivity (p <0.0001), and lower education status (p<0.0001, p=0.027) were predictive of higher RF and ACPA titers, respectively. In the same model, while patients in quartile 2 and 4 of AP exposure had lower ACPA levels (p= 0.03, p< 0.0002), there were inconsistent associations of AP quartiles with RF (Table 1). Analyses of individual pollutants revealed higher PM_2.5 levels to be associated with higher RF and ACPA titers (p=0.0005, p=0.0092).

Conclusion: In a predominantly male RA population with long standing disease, while confirmation of the association of HLADRB1 SE and smoking with autoantibody seropositivity was found, there was an inverse relationship between higher levels of air pollutant exposure and ACPA. Only small particulate matter exposure (PM_2.5) was linked to higher RF and ACPA titers. The overall effect of air pollutants on RA autoantibody status appears to be a varied and complex relationship.