Causal association between rheumatoid arthritis and pancreatic cancer: a bidirectional two-sample Mendelian randomization study.

OBJECTIVE: Elucidating the causal relationship between rheumatoid arthritis (RA) and pancreatic cancer (PC) remains methodologically challenging due to limitations inherent in observational studies. This study applied a bidirectional two-sample Mendelian randomization (MR) design to investigate the potential causal interplay between RA and PC, aiming to uncover shared pathogenic pathways and genetic predispositions through comprehensive instrumental variable analysis. METHODS: Utilizing genome-wide association study (GWAS) meta-analysis data, we systematically screened and incorporated independent single nucleotide polymorphisms (SNPs) associated with RA and PC as instrumental variables. The primary analytical approach was the inverse-variance weighted fixed-effects model (IVW-FE). To ensure methodological rigor and validate causal inferences, we complemented this with several approaches: simple median, weighted median, and MR-Egger regression. Heterogeneity was assessed using Cochran's Q test, and pleiotropy was evaluated using MR-Egger intercept analysis. Sensitivity analysis was conducted using the leave-one-out approach. All effect estimates are reported as odds ratios (ORs) with corresponding 95% confidence intervals (CIs). RESULTS: Seventy-seven SNPs met the predefined inclusion criteria. MR analyses revealed that elevated genetic susceptibility to RA was causally associated with an increased risk of PC (OR = 1.187, 95% CI = 1.355-1.039, P = 0.011). MR-Egger regression analysis indicated no significant pleiotropic effects (intercept p = 0.434). Cochran's Q statistics indicated no substantial heterogeneity in the causal estimates for RA (P = 0.064). Consistent with these findings, leave-one-out sensitivity analyses confirmed the absence of influential outlier variants in the instrumental variable sets. However, there was no significant causal association between PC and the risk of RA (OR = 0.991, 95% CI = 0.957-1.027, P = 0.620). CONCLUSIONS: Our results support a unidirectional causal relationship, identifying RA as a potential risk factor for the development of pancreatic cancer, with no evidence for a reverse causal effect.