Impact of rheumatoid arthritis on Alzheimer's disease: A two-sample bidirectional Mendelian randomization study

Controversial relationship of rheumatoid arthritis (RA) with Alzheimer disease (AD) risk has been reported in previous research. However, epidemiological studies are susceptible to confounding factors and reverse causality. This study aimed to explore the causal relationship between RA and AD by using a 2-sample bidirectional Mendelian randomization (MR) study. Genetic data for RA and AD were extracted from published genome-wide association study databases and the FinnGen project. The primary statistical method was inverse variance weighted, which was supplemented by MR Egger regression, weighted median, simple mode, and weighted mode methods. In addition, Cochran Q test, MR Egger intercept, and MR-PRESSO global test were used to detect heterogeneity and pleiotropy. Further sensitivity analyses were conducted using leave-one-out method and funnel plots, to evaluate the robustness of the results. Genetically predicted RA had a positive casual effect on the risk of AD development (inverse variance weighted odds ratio [OR] = 1.062, 95% confidence interval [CI] = 1.019-1.107, P = .004; weighted median OR = 1.073, 95% CI = 1.021-1.126, P = .005; weighted mode OR = 1.077, 95% CI = 1.025-1.131, P = .007). Notably, reverse MR analysis indicated no significant effect of AD on RA (all P > .05). No pleiotropy or heterogeneity was identified in the bidirectional MR analysis. And leave-one-out analysis and funnel plots confirmed the robustness and reliability of the findings. This study provides new evidence for the causal relationship between RA and the increased risk of AD. Early screening of cognitive function in patients with RA may be beneficial in preventing future AD progression.