Objective To evaluate the causal relationship between serum uric acid (SUA) level and cerebrovascular disease (CVD) by a two-sample bidirectional Mendelian randomization (MR) study.

Methods Single nucleotide polymorphisms (SNPs) from Genome-Wide Association Studies (GWAS) were obtained as instrumental variables for both samples. Inverse-variance weighted (IVW) method was primarily adopted, with weighted median method, weighted mode method, and MR-Egger regression serving as supplementary approaches for sensitivity analyses to verify the robustness of the results.

Results The forward IVW analysis results showed that increased SUA was a risk factor for stroke (OR=1.183, 95% CI, 1.081 to 1.295, P=2.51×10^-4), ischemic stroke (OR=1.196, 95% CI, 1.084 to 1.320, P=3.81×10^-4), and large artery atherosclerotic stroke (OR=1.466, 95% CI, 1.134 to 1.897, P=0.004), and also a protective factor for vascular dementia (OR=0.451, 95% CI, 0.273 to 0.745, P=0.002) and multi-infarct dementia (OR=0.372, 95% CI, 0.144 to 0.959, P=0.041). The reverse IVW analysis results did not support a causal effect of genetically predicted CVD risk on SUA level. All significant results were corrected by Bonferroni with P value less than 0.005. Sensitivity analyses further confirmed the reliability of the study findings.

Conclusion The MR analysis reveals positive correlations between increased SUA level and the risk of stroke, ischemic stroke and large artery atherosclerotic stroke, and negative correlations of SUA level with the risk of vascular dementia and multi-infarct dementia.