Causal association between metformin and arrhythmias: A Mendelian randomization study

There is ongoing discussion in the medical community on the effect of metformin in cardiac arrhythmias. Investigating the causal relationship between metformin and arrhythmias was the target of the research. We utilized the Mendelian randomization (MR)-based platform to collect genome-wide association data linking arrhythmias and metformin. We utilized SNPs connected to metformin as instrumental variables (IVs) to evaluate the causative link between metformin and arrhythmia through two-sample MR analysis. Four statistical techniques were employed: MR-Egger regression, weighted median estimator, weighted mode method, and inverse-variance weighted (IVW) method. The impact of individual SNPs on the outcomes of IVW analyses was investigated using the leave-one-out method, and the study's possible bias was examined using a funnel plot to guarantee the results' robustness. We identified 40 independent single nucleotide polymorphisms (SNPs) of genome-wide significance from GWAS data for metformin as IVs. The IVW method supported a causal relationship between metformin and reduced incidence of arrhythmia (OR = 0.0093; 95% CI = 0.002-0.055; P = 2.72E-07). MR-Egger regression indicated that there is no need to consider the effect of gene pleiotropy on the results of the study (intercept = 0.0014; P = .874). Although the MR-Egger method didn't pinpoint a causal link, it did reveal a similar directionality in the β values (OR = 0.0056; 95% CI = 9.384E-06 to 3.369; P = .1207). Fortunately, the weighted median and weighted mode techniques demonstrated a causal link between metformin and a decreased arrhythmia risk (OR = 0.004; 95% CI = 0.0003-0.052, P = .000023; OR = 0.0025; 95% CI = 7.418E-05 to 0.085, P = .001915). Neither Cochran Q test nor the funnel plot showed signs of directional pleiotropy, heterogeneity, or asymmetry. The sensitivity of the method was analyzed by leave-one-out method, and the results suggested that the method was stable. With the MR method, we've substantiated a potential cause-and-effect link between metformin usage and cardiac arrhythmia.