Physical activity and the risk of rheumatoid arthritis: evidence from meta-analysis and Mendelian randomization

Background: There is very little information about the association between physical activity (PA) and the risk of rheumatoid arthritis (RA). The purpose of this study is to understand the effect of PA on subsequent risk of developing RA. Methods: A literature search was performed in PubMed and Web of Science up to 19 September 2020. Observational studies examining associations between PA and the RA development were identified. Categorical and dose-response meta-analyses were both performed. Then two-sample Mendelian randomization (MR) analysis was conducted to interrogate the causal relationship by utilizing genetic instruments identified from a genome-wide association study of self-reported and accelerometer-based PA traits. Results: Four eligible studies were included in the meta-analyses, involving 4213 RA cases among 255 365 participants. The summary relative risk (RR) of RA risk was 0.79 [95% confidence interval (CI): 0.72, 0.87] for the highest vs the lowest PA, and 0.85 (95% CI: 0.79, 0.92) for PA vs inactivity/occasional PA. However, we found no convincing evidence supporting a causal role of genetically predicted accelerometer-measured PA [odds ratio (OR): 0.97; 95% CI: 0.88, 1.08 per 1-SD unit increment], genetically predicted moderate-to-vigorous PA (OR: 1.08; 95% CI: 0.49, 2.39 per 1-SD unit increment) or genetically predicted vigorous PA >= 3 days/week (OR: 2.63; 95% CI: 0.05, 130.96) with RA risk. Conclusions: The meta-analyses of the observational studies indicated that higher PA levels correlate with reduced risk of RA. In contrast to meta-analyses, the MR analyses reported here suggested PA may not help to prevent RA.