Rheumatoid Arthritis and Gastroesophageal Reflux Disease: A Bidirectional and Multivariable Two-sample Mendelian Randomization Study

Abstract

Abstract Aims/hypothesis: The association between Gastroesophageal Reflux Disease (GERD) and Rheumatoid Arthritis (RA) has been reported by many observational studies in the Asian population. Here, we examined the bidirectional causal effects between GERD and RA by two sample MR analyses using genetic evidence. Methods: Two-sample Mendelian randomization analyses were performed to determine the causal effect of the GERD (129080 cases vs 602604 control participants) on RA (6236 cases vs 147221 control participants) and RA on GERD respectively. The inverse-variance weighted (IVW) method was used as the primary analysis. Weighted median and MR-Egger regression were taken as supplements. The Cochran Q test evaluated the heterogeneity. Horizontal pleiotropy was detected by estimating the intercept term of MR-Egger regression. Further, multivariable MR analyses were performed to exclude the influence of confounding factors including years of schooling, BMI, and time spent watching television between GERD and RA. Result: Both The univariate MR (UVMR) and multivariable MR (MVMR) provided strong evidence that the RA was causally and positively influenced by GERD (UVMR: OR = 1.49, 95% CI = 1.25–1.76, P = 6.18*10− 6; MVMR: OR = 1.69, 95% CI = 1.24–2.31, P = 8.62*10− 4), while GERD was slightly influenced by RA (UVMR: OR = 1.03, 95% CI = 1.00-1.06, P = 0.042; MVMR: OR = 1.04, 95% CI = 1.00-1.07, P = 0.0271). Conclusion: Our comprehensive bidirectional MR analysis found that for the European population, GERD can induce the occurrence of RA (OR = 1.69, P < 0.00125), while RA only has a slight influence on GERD. Specifically, GERD patients are suffering a 69% increased risk of RA occurrence, which means GERD is a substantial risk factor for RA.