Abstract
Background: Accumulating evidence from observational studies suggested that circulating adiponectin levels are associated with the risk of rheumatoid arthritis (RA), but the causality remains unknown. We aimed to assess the causal relationship of adiponectin with RA risk. Methods: Based on summary statistics from large-scale genome-wide association studies (GWAS), we quantified the genetic correlation between adiponectin and RA. Then bidirectional Mendelian randomization (MR) analysis was performed to assess the causal relationship. Twenty single-nucleotide polymorphisms (SNPs) associated with adiponectin were selected as instrumental variables from a recent GWAS (n = 67,739). We applied theses SNPs to a large-scale GWAS for RA (14,361 cases and 43,923 controls) with replication using RA data from the FinnGen consortium (6,236 cases and 147,221 controls) and the UK Biobank (5,201 cases and 457,732 controls). The inverse-variance weighted (IVW) and multiple pleiotropy-robust methods were used for two-sample MR analyses. Results: Our analyses showed no significant genetic correlation between circulating adiponectin levels and RA [rG = 0.127, 95% confidence interval (CI): -0.012 to 0.266, P = 0.074]. In MR analyses, genetically predicted adiponectin levels were not significantly associated with the RA risk (odds ratio: 0.98, 95% CI: 0.88-1.09, P = 0.669). In the reverse direction analysis, there is little evidence supporting an association of genetic susceptibility to RA with adiponectin (β: 0.007, 95% CI: -0.003 to 0.018, P = 0.177). Replication analyses and sensitivity analyses using different models yielded consistent results. Conclusions: Our findings provided no evidence to support the causal effect of adiponectin levels on RA risk and of RA on circulating adiponectin levels.