Impact of ankylosing spondylitis on stroke limited to specific subtypes: Evidence from Mendelian randomization study.

Background: The relationship between Ankylosing Spondylitis (AS) and the risk of stroke is complex. Therefore, we utilized Two-Sample Mendelian randomization to examine the probable causal link between these two features. Methods: The genetic instruments linked to AS were chosen from a summary-level genetic data set from the FinnGen consortium in people of European ancestry (1462 cases and 164,682 controls). Stroke and its subtypes were selected as outcomes, and the MEGASTROKE consortium population was used to identify the genetic associations of AS on stroke (40,585 cases and 406,111 controls), ischemic stroke (IS) (34,217 cases and 406,111 controls), and its subtypes including large artery stroke (LAS) (4373 cases and 146,392 controls), small vessel stroke (SVS) (5386 cases and 192,662 controls), and cardioembolic stroke (CES) (7193 cases and 204,570 controls). Intracerebral hemorrhage (ICH) (1687 cases and 201,146 controls) data set from the FinnGen consortium was also used. To obtain the casual estimates, the inverse variant weighted (IVW) method was mainly used. By examining the heterogeneity and pleiotropy of particular single nucleotide polymorphisms (SNPs), the robustness of the results was also examined. Results: There was no evidence found to prove the correlation between genetically predicted AS and stroke (odds ratio [OR] 1.014; 95% confidence interval [CI] 0.999-1.031; P = 0.063), ICH (OR 1.030; 95% CI 0.995-1.067; P = 0.090), and IS (OR 1.013; 95% CI 0. 998-1.030; P = 0.090). In terms of the different subtypes of IS, there was strong evidence of positive causal inferences on CES (OR 1.051; 95% CI 1.022-1.081; P = 0.001), and suggestive evidence of positive causal inferences on LAS (OR 1.042; 95% CI 1.003-1.082; P = 0.033), while it was not significant for SVS (OR 1.010; 95% CI 0.975-1.047; P = 0.563). Conclusion: This study suggests that the possible causative impact of genetically predicted AS on stroke may be restricted to the CES and LAS subtypes.