Genetically determined atrial fibrillation and risk of stroke: a Mendelian randomization study

Background Based on observational evidence, atrial fibrillation is a well-established risk factor of stroke to be considered for antithrombotic treatment in presence of additional clinical conditions derived from multivariate risk models. Although biologically plausible, it however still is unknown whether this association is causal and confined to the embolic stroke subtype. Purpose Our objective was to explore whether genetically determined manifestation of atrial fibrillation was associated with stroke and its etiologic subtypes by conducting a 2-sample Mendelian randomization (MR) study on publicly available summary statistics from GWAS consortia. Methods Genetic instruments for atrial fibrillation were obtained from the AFGen Consortium comprising 17,931 cases and 115,142 controls. Their associations with stroke and stroke subtypes were evaluated in the MEGASTROKE genome-wide association study data set (67 162 cases; 454 450 controls) applying inverse variance–weighted meta-analysis, weighted-median analysis, Mendelian randomization–Egger regression, and multivariable Mendelian randomization. The dataset of Nielsen et al. comprising a total of 60,620 cases with atrial fibrillation and 970,216 controls of European ancestry from six contributing studies was used as an independent validation sample. Genetic instruments for atrial fibrillation were further tested for association with etiologically related traits by using publicly available genome-wide association study data. Results Genetic predisposition to atrial fibrillation was associated with higher risk of any stroke (beta coefficient [b] ± standard error [se] = 0.22±0.04; P=0.0001), any ischemic stroke (b ± se = 0.24±0.05; P=0.0003), and cardioembolic stroke (b ± se = 0.76±0.10; P<0.0001), but not with small-vessel stroke or large artery stroke, see figure. Analyses in the validation sample showed similar associations (any stroke: b ± se = 0.19±0.04; P<0.0001; any ischemic stroke: b ± se = 0.21±0.04; P<0.0001; cardioembolic stroke: b ± se = 0.82±0.13; P<0.0001). Genetically determined atrial fibrillation was further weakly associated with chronic kidney disease (b ± se = 0.10±0.04; P=0.0261), but not with coronary artery disease and myocardial infarction or any other available phenotype. Conclusions Genetic predisposition to atrial fibrillation is associated with higher risk of any stroke, mainly driven by the ischemic and cardioembolic subtypes. In contrast, large artery and small-vessel strokes did not exhibit a causal relationship with atrial fibrillation. Funding Acknowledgement Type of funding source: Public hospital(s). Main funding source(s): Medical University of Vienna, Austria