Abstract
Objectives: Rheumatoid arthritis (RA) is a complex disease that is associated with genetic and environmental factors. We have investigated geospatial variation in the risk of developing RA within Stockholm County, Sweden, with respect to established environmental risk factors for RA, as well as serologically defined subgroups of RA.
Method: Information regarding geographical location for 1432 cases and 2529 controls from the Epidemiological Investigation of Rheumatoid Arthritis (EIRA) study, living in Stockholm County at RA symptom onset, or matched date for controls, was used to estimate geospatial variation in risk. We used generalized additive models (GAMs) to create a risk surface, calculate odds ratios (ORs), and adjust for potential confounding by smoking, education level, and RA within family. We performed a stratified analysis based on the presence/absence of anti-citrullinated peptide antibodies (ACPA).
Results: We found significant spatial variation in the odds of developing RA in Stockholm County. After adjustment for smoking, education level, and family history of RA, this geospatial variation remained. The stratified analysis showed areas with higher ORs for ACPA-positive RA and ACPA-negative RA, after adjusting for smoking, education level, and having a family history of RA. Living in the city of Stockholm was associated with decreased risk of RA.
Conclusions: The risk of developing RA in Stockholm County is not distributed evenly and there are areas of increased risk that could not be explained by known factors. Further investigations of local exposures or social factors are warranted.
Acknowledgements
We thank Marie-Louise Serra and Lena Nise for administrative support. Au: Please check acknowledgements.This work was supported by National Institutes of Health (NIH) grants AR49880 and AR047782, COFAS (D.no. 2009-1905), the Swedish Council for Working Life and Social Research (FAS), the insurance company AFA and 2P42ES00 7381-16A1.