Background: Rheumatoid arthritis (RA) is associated with diverse extra-articular manifestations, including interstitial lung disease (ILD). No previous studies have examined the bidirectional relationship between RA and ILD using the Mendelian randomization (MR) analyses. Therefore, we aimed to investigate this subject using a two-sample bidirectional MR method. Methods: We performed bidirectional two-sample MR using summary statistics from genome-wide association studies (GWASs). The data are publicly available, de-identified, and from European (EUR) and East Asian (EAS) ancestries. Results: A total of 474,450 EUR participants and 351,653 EAS participants were included for either forward or reverse MR analysis. In our primary analysis, we found significant evidence of an increased risk of ILD associated with RA among individuals of EUR ancestry (ORMR-cML = 1.08; 95% confidence interval [CI] = 1.03-1.14; p = 0.003) and EAS ancestry (ORMR-cML = 1.37; 95% CI = 1.23-1.54; p < 0.001). Additionally, the reverse MR showed significant evidence of an increased risk of RA associated with ILD among those of EUR ancestry (ORMR-cML = 1.12; 95% CI = 1.05-1.19; p < 0.001). However, only one instrumental variable was selected in the EAS ILD GWAS, and there was no increased risk of RA associated with ILD in those of EAS ancestry (ORMR-cML = 1.02; 95% CI = 0.91-1.14; p = 0.740). Conclusions: Our findings indicate that RA and ILD have a bidirectional causal inference when using the MR analysis of GWAS datasets. The findings are only relevant for genetic predisposition; thus, further research is needed to determine the impact of non-genetic predispositions.
Keywords: Mendelian randomization; bidirectional; causal effect; interstitial lung disease; rheumatoid arthritis.