Causal association between rheumatoid arthritis and risk of stroke: A Mendelian randomization study

Abstract

Background

Active rheumatoid arthritis (RA) may damage vascular endothelial cells, thereby increasing the likelihood of adverse cardiovascular events. However, it is not yet clearly established whether RA also increases the risk of adverse cerebrovascular events, particularly stroke.

Objective

This study was designed to evaluate the likelihood of a causal association between RA and stroke.

Method

A two-sample Mendelian randomization (MR) analysis was performed using the inverse variance-weighted (IVW) average, weighted median, and MR-Egger regression methods. The analysis utilized publicly available summary statistics datasets from Genome-wide association studies (GWAS) meta-analyses for RA in individuals of European descent (total n = 484,598; case = 5427, control = 479,171) as the exposure cohort, and from GWAS meta-analyses for "vascular/heart problems diagnosed by doctor: stroke" in individuals included in the UK Biobank (total n = 461,880; case = 7055, control = 454,825, MRC-IEU consortium) as the outcome cohort.

Results

Eight single-nucleotide polymorphisms with genome-wide significance were selected from the GWASs on RA as the instrumental variables. The results of the MR-Egger and weighted median analyses showed no causal association between RA and stroke (OR = 1.081, 95 % CI [0.943–1.240], P = 0.304) vs. OR = 1.079, 95 % CI [0.988–1.179], P = 0.091), respectively. However, the inverse variance-weighted (IVW) analysis results revealed a causal association between RA and stroke (OR = 1.115, 95 % CI [1.040–1.194], P = 0.002). Cochran's Q test and MR-Egger regression revealed no evidence of heterogeneity and horizontal pleiotropy.

Conclusion

The MR analysis results indicated that rheumatoid arthritis (RA) may be causally associated with an increased risk of stroke.