Abstract

Background: Previous studies have reported an interaction between ever cigarette smoking and the presence of the human leukocyte antigen (HLA)-DRB1 shared epitope (SE) genotype and rheumatoid arthritis (RA) risk. To address the effect of dosage, a case-control study nested within two prospective cohorts to determine the interaction between heavy smoking and the HLA-SE was conducted.

Methods: Blood was obtained from 32 826 women in the Nurses’ Health Study and 29 611 women in the Nurses’ Health Study II. Incident RA diagnoses were validated by chart review. Controls were matched for age, menopausal status and postmenopausal hormone use. High-resolution HLA-DRB1 genotyping was performed for SE alleles. HLA-SE, smoking, HLA-SE* smoking interactions and RA risk, were assessed using conditional logistic regression models, adjusted for age and reproductive factors. Additive and multiplicative interactions were tested.

Results: In all, 439 Caucasian matched pairs were included. Mean age at RA diagnosis was 55.2 years; 62% of cases were seropositive. A modest additive interaction was observed between ever smoking and HLA-SE in seropositive RA risk. A strong additive interaction (attributable proportion due to interaction (AP) = 0.50; p<0.001) and significant multiplicative interaction (p = 0.05) were found between heavy smoking (>10 pack-years) and any HLA-SE in seropositive RA risk. The highest risk was in heavy smokers with double copy HLA-SE (odds ratio (OR) 7.47, 95% CI 2.77 to 20.11).

Conclusions: A strong gene–environment interaction was observed between HLA-SE and smoking when stratifying by pack-years of smoking rather than by ever smoking. Future studies should assess cumulative exposure to cigarette smoke when testing for gene–smoking interactions.