Background/Purpose: Rheumatoid arthritis (RA) is a complex disease with both genetic and environmental risk factors. A number of susceptibility genes have been identified. Past studies have also observed an association between decreased levels of vitamin D and RA. We aimed to examine the joint effects of genetic markers of RA, and circulating vitamin D levels on RA risk.

Methods: We selected incident Caucasian RA cases (n=133) who contributed blood samples prior to first RA symptom and matched controls (n=134) from the Nurses’ Health Study and Nurses’ Health Study II cohorts. We genotyped 36 RA risk alleles identified by previous GWAS and meta-analysis, and 8 HLA-DBR1 shared epitope (SE) genotypes at 4 digit resolution. We measured 25-hydroxyvitamin D (25(OH)D) levels by chemiluminescence immunoassay. We used conditional logistic regression to examine the marginal genetic and 25(OH)D effects, as well as testing for additive and multiplicative interaction between the genetic markers and 25(OH)D (sufficient vs. insufficient (²20ng/mL)). We ran unconditional logistic regression analysis stratified by seropositive (n=66) and seronegative (n=67) RA phenotypes additionally adjusted for matching factors. We adjusted for multiple comparisons with Bonferroni correction.

Results: We found statistically significant main effects of HLA-DBR1 SE and RA (p=0.0013) and seropositive RA (p=4×10^-5). There was no significant main effect association between vitamin D insufficiency and RA. Among all RA cases we observed a statistically significant additive interaction between FCGR2A (rs7552317) and insufficient 25(OH)D (p=0.004). For subjects with having any FCGR2A alleles and insufficient 25(OH)D there was an >2-fold odds of RA (2.18 (95% CI 1.11, 5.81)) when compared to those without FCGR2A alleles and sufficient 25(OH)D (Table). Among seropositive RA we observed a statistically significant additive interaction between HLA-DRB*0101 and insufficient 25(OH)D (p= 0.0002) where having any HLA-DRB*0101 and insufficient 25(OH)D was associated with an >2-fold increase in odds of RA (2.18 (95% CI 0.82, 5.79)).

Conclusion: We observed significant gene-vitamin D interactions when assessing RA risk and seropositive RA risk. FCGR2A (Fc fragment of IgG, low affinity IIa, receptor) and HLA-DRB1 play important roles in the immune system. FCGR2A has been associated with receptor aggregation, neutrophil activation, SLE and lupus nephritis. These results offer clues into the pathophysiology of RA and support a role for vitamin D in inflammatory pathways in pre-RA.

Table: Gene-vitamin D interactions between FCGR2A and HLA-DRB1*0101 polymorphisms and vitamin D insufficiency in relation to risk of all rheumatoid arthritis (RA) and seropositive RA.

* Categories of vitamin D: defined as sufficient (25(OH)D level >=20ng/mL) and insufficient (<20ng/mL).