Abstract
BACKGROUND: Ischemic stroke (IS) is a major cause of disability and death worldwide. However, previous observational studies failed to establish the causality between various exposures and IS occurrence. Currently, an increasing number of Mendelian randomization (MR) studies have been performed to investigate the causal effects of exposure factors on IS risk, but the results remain inconsistent and inconclusive. Thus, our systematic review summarized all published MR studies focusing on IS and its subtypes, and further identified causal relationships with robust evidence. METHODS: We retrieved PubMed and Embase databases for MR analyses exploring the correlation of genetically determined exposures with the risk of IS and its subtypes. For publications selected in the main analysis, we summed up these MR results and classified the strength of evidence into Grade 1-4 based on the significance and concordance of the findings from the primary and complementary MR methods, as well as the robustness of sensitivity analyses. RESULTS: After conducting a comprehensive search, a total of 207 published studies with 1199 MR associations were included. Among these causal correlations, 8 (Grade 1) and 81 (Grade 2) were graded as robust evidence, while 226 (Grade 3) and 884 (Grade 4) causal relationships were considered to have insufficient reliability. Specifically, with regard to IS susceptibility, genetically predicted diastolic blood pressure, type 2 diabetes, Parkinson's disease, and primary aldosteronism were evaluated as the prominent robust causal determinants. As for IS subtypes, genetically determined higher vitamin K1 levels and primary aldosteronism were related to greater occurrence of large artery stroke and small vessel stroke, respectively, whereas increased adult height was associated with a lower risk of small vessel stroke but a higher risk of cardioembolic stroke. CONCLUSIONS: Our field synopsis systematically summarized and evaluated the evidence strength for causality between various exposures and the occurrence of IS and its subtypes, and ultimately identified 89 causal relationships as convincing findings. We hope that our results can provide constructive and impressive insights for developing effective IS prevention strategies.