Genetic correlations and causal inferences in ischemic stroke

Background and purpose Considerable studies have reported inconsistent relationships between ischemic stroke and a large number of factors. These uncertainties may reflect the susceptibility to confounding in observational studies. We aimed to assess genetic correlations and causal relationships between ischemic stroke and diverse phenotypes. Methods Summary-level data for ischemic stroke (34,217 cases and 406,111 controls) from the MEGASTROKE consortium were used as the outcome. Exposures were derived from two GWAS statistics curated databases. We explored the genetic correlations and causalities between hundreds of traits and ischemic stroke, using linkage disequilibrium score regression and Mendelian randomization (MR), respectively. Multiple sensitivity analyses were also performed. Results Genetic correlation analyses reflected genetic overlaps between ischemic stroke and physical activity, cardiometabolic factors, smoking, and lung function. Applying MR, we found suggestive evidence that genetic predisposition to higher concentration of low-density lipoprotein particles (LDL.P) and cholesterol carried in different sizes of LDL.P (LDL.C) were associated with higher risk of ischemic stroke, particular large artery stroke. The strongest effect was observed for small LDL.P in large artery stroke (OR 1.31, 95% CI 1.09-1.56, p = 0.003). The results were overall robust for sensitivity analyses. We further observed significant positive associations of genetically predicted LDL.P and LDL.C with coronary artery disease and myocardial infarction. Conclusions Shared genetic overlaps might exist between ischemic stroke and physical activity, cardiometabolic factors, smoking, and lung function. We provided suggestive evidence for a potential causal role of LDL.P and LDL.C in ischemic stroke, particularly in large artery stroke. Future researches are required to confirm these findings.