Atrial fibrillation and kidney function: A bidirectional Mendelian randomization study

Abstract Aims To investigate the causal effects between atrial fibrillation (AF) and kidney function. Methods and Results We performed a bidirectional Mendelian randomization (MR) analysis implementing the results from large-scale genome-wide association study (GWAS) for estimated glomerular filtration rate (eGFR) by the CKDGen (N = 1,046,070) and for AF (N = 588,190) to determine genetic instruments. A bidirectional two-sample MR based on summary-level data was performed. Inverse variance weighted method was the main MR method. For replication, an allele-score based MR was performed by individual-level data within the UK Biobank cohort of white British ancestry with eGFR values (N= 321,260). The genetical predisposition to AF was significantly associated with lower eGFR [beta - 0.002 (standard error 0.0005), P < 0.001] and higher risk of chronic kidney disease [beta 0.051 (0.012), P < 0.001], and the significance remained in various MR sensitivity analyses. The causal estimates were consistent when we limited the analysis to individuals of European ancestry. The genetically predicted eGFR did not show significant association with risk of AF [beta −0.189 (0.184), P = 0.305]. The results were similar in allele-score based MR, as allele-score for AF was significantly associated with lower eGFR [beta −0.069 (0.021), P < 0.001] but allele-score for eGFR did not show significant association with risk of AF [beta −0.001 (0.009), P = 0.907]. Conclusions Our study supports that genetical predisposition to AF is a causal risk factor for kidney function impairment. However, effect from kidney function on AF was not identified in this study.