Background Previous studies suggest links between air pollution (i.e. fine particulate matter, PM2.5) and serum antibodies related to rheumatic diseases. No one has yet examined associations between anti-nuclear antibodies (ANA) and ultrafine particles (UFP) or ozone (O3), both of which can enter through the lungs and may have the potential to trigger systemic effects.

Methods Our analyses were based within the CARTaGENE general population cohort (n=20,000) based in the province of Quebec, Canada. We determined baseline ANA (HEp-2000, Immuno Concepts) on a random sample. Air pollutant exposures were assigned by linking subjects’ residential postal codes with estimated levels (determined by hybrid approaches including satellite imagery and modelling). We performed multivariable logistic regression models for the outcome of positive ANA, assessing for independent effects of UFP (available for Montreal only) and O3 in separate models, adjusting for age, sex, smoking, and self-reported ancestry.

Results ANA positivity of at least 1:160 occurred in 713 (20%) of 3578 randomly selected patients tested. The ANA positive subjects were more likely than ANA negative subjects to be female (63% vs 49%) while the average age (55.4 vs 54.0) and percent never-smokers (37% vs 40%) were similar. There was missing information on covariates for 232 subjects, which therefore were not included in the model estimates.

There was a trend for higher average UFP: exposure in ANA positive subjects (24 606 particles/cm³, standard deviation, SD 4979) versus ANA negative (24328, SD 5078), while average ozone levels were very similar (22.5 vs 22.6 µg/m). The multivariable model (table 1) for UFP showed a trend to higher levels in the ANA positive group (1.008, 95% CI 0.982 to 1.034) while in the multivariable model for O3 the OR was very close to the null value (0.996, 95% CI 0.965 to 1.029). In all models, risk factors for ANA positivity included older age and female sex, with trends for lower ANA positivity in French Canadians.

Conclusions We saw a non-significant trend towards higher UFP levels in ANA positive versus negative subjects, while O3 levels seemed very similar in the two groups. Expected trends for more ANA positivity with older age and female sex was seen. Further study of UFP levels with a larger sample size is in progress. If confirmed, these results may strengthen the hypothesis that air pollution is an environmental trigger of immune system activation.

Acknowledgements This research was supported by the CIHR. CARTaGENE is a member of the Canadian Project for Tomorrow cohort.