Genetically Confirmed Optimal Causal Association of Cerebrospinal Fluid Metabolites With Hemorrhagic Stroke.

Abstract

Hemorrhagic stroke (HS) mainly includes intracerebral hemorrhage (ICH) and subarachnoid hemorrhage (SAH), both of which seriously affect the patient's prognosis. Cerebrospinal fluid (CSF) metabolites and HS showed a link in observational studies. However, the causal association between them is not clear. We aimed to establish the optimal causality of CSF metabolites with HS. Mendelian randomization (MR) was employed to identify associations between CSF metabolites and different sources of HS. Univariable MR and false discovery rates (FDR) were used to identify initial causal associations. Linkage disequilibrium score regression determined genetic correlations. Multiple sensitive analyses ensured the reliability of the results. Multivariable MR and MR Bayesian Model Averaging were used to identify the optimal causal associations. The combined effects of metabolites and HS were assessed by meta-analyses. Pathway analyses were performed to identify potential pathways of action. Reverse MR was also conducted to identify reverse causal associations. Finally, Corresponding blood metabolites were used to explore the multiple roles of metabolites. We identified 20 CSF metabolites and six metabolic pathways associated with ICH; 15 CSF metabolites and three metabolic pathways associated with SAH. Nineteen and seven metabolites were causally associated with deep and lobar ICH, respectively. CSF levels of mannose (OR 0.63; 95% CI 0.45-0.88; P_combined = 0.0059) and N-acetyltaurine (OR 0.68; 95% CI 0.47-0.98; P_combined = 0.0395) may serve as the optimal exposures for ICH and SAH, respectively. Additionally, CSF ascorbic acid 3-sulfate levels significantly decrease the risk of deep ICH (OR 0.79; 95% CI 0.66-0.94; p = 0.0065; P_FDR = 0.091). Supplemental analysis of blood metabolites suggested multiple roles for CSF and blood N-formylanthranilic acid and hippurate. There are significant causal associations between CSF metabolites and HS, which provides a further rationale for the prevention and monitoring of ICH and SAH.