Association between cerebrospinal fluid metabolites and ischemic stroke: a Mendelian randomization analysis.

Abstract

Ischemic stroke (IS) is one of the leading causes of disability and mortality worldwide. Cerebrospinal fluid (CSF) metabolites can reflect the pathophysiological state of the central nervous system, but their causal relationship with IS remains unclear. This study aims to assess the causal association between CSF metabolites and IS prognosis using Mendelian randomization (MR) methodology. A two-sample MR design was employed, with genetic data on CSF metabolites sourced from the WADRC and WRAP cohorts (a total of 338 metabolites in a European ancestry population), and IS prognosis outcome data obtained from the GISCOME network (6021 IS patients). Single nucleotide polymorphisms (SNPs) significantly associated at P < 5 × 10^-6 were selected as instrumental variables, while those in linkage disequilibrium (r² < 0.001) and weak instruments (F < 10) were excluded. Causal estimates were primarily obtained using the inverse variance weighted (IVW) method, with sensitivity analyses conducted using Cochran's Q test, MR-Egger regression, MR-PRESSO, and leave-one-out analysis. Additionally, metabolomic validation was performed using the MetaboLights database (MTBLS6279), ROC curve analysis was conducted to evaluate predictive performance, and KEGG enrichment analysis was undertaken to explore potential mechanisms. The IVW analysis identified 23 CSF metabolites that have a significant causal association with the prognosis of IS. Among these, 18 are protective metabolites (such as sphingolipids, glucuronic acid, N-acetylserotonin, 2'-deoxyuridine, S-adenosylmethionine, cyclic adenosine monophosphate, guanosine, etc.), while 5 are risk metabolites (cholesterol, phenylalanine, erythritol, glutamine, and methionine). Metabolomics validation revealed that 2'-deoxyuridine, S-adenosylmethionine, cyclic adenosine monophosphate, guanosine, and methionine overlapped with the MR results. The ROC curve indicated that deoxyuridine has a high discriminative power (AUC = 0.944, 95% CI: 0.75-1). Sensitivity analysis showed no detected heterogeneity and pleiotropy, and the leave-one-out analysis supported the robustness of the results. KEGG enrichment analysis suggested that the relevant metabolites may be involved in pathways such as autophagy, AMPK signaling, and vascular smooth muscle contraction. This study is the first to systematically evaluate the causal relationship between CSF metabolites and the prognosis of IS, revealing that 23 metabolites may be involved in the occurrence and development of IS. Among these, risk metabolites such as low-density lipoprotein-related cholesterol and phenylalanine are consistent with the mechanisms of atherosclerosis and neuroinflammation. Deoxyuridine, a key intermediate in DNA synthesis and repair, may influence the progression of IS by regulating nucleotide homeostasis and oxidative stress. The MR design effectively avoids confounding and reverse causation bias, while external metabolomics validation and ROC analysis enhance the reliability of the results. This study confirms the causal relationship between various CSF metabolites and the prognosis of IS, with deoxyuridine emerging as a potential biomarker for predicting IS outcomes. These findings provide new insights into the mechanisms of IS and early intervention strategies.