Causal relationship between cerebral atherosclerosis and cerebrospinal fluid metabolites: A bidirectional Mendelian randomization study.

Cerebral atherosclerosis has been linked in observational studies to changes in cerebrospinal fluid (CSF) metabolites, yet whether these associations reflect causal relationships is uncertain. Understanding such links could advance biomarker discovery and prevention strategies. We conducted a bidirectional Mendelian randomization analysis to evaluate potential causal associations between 338 CSF metabolites and the risk of cerebral atherosclerosis, using summary statistics from genome-wide association studies. The inverse variance weighted method was the primary approach, complemented by Mendelian randomization-Egger regression, Mendelian randomization-PRESSO outlier tests, Cochran's Q test, and linkage disequilibrium score regression. Effect sizes were expressed as odds ratios (ORs) with 95% confidence intervals (CIs) and corresponding P-values. Twelve metabolites showed suggestive causal associations (P < .05) with cerebral atherosclerosis. After accounting for pleiotropy and heterogeneity, 11 remained significant. Protective associations (OR < 1) included acetoacetate (OR: 0.78, 95% CI: 0.63-0.98, P = .030), 5-oxoproline (OR: 0.41, 95% CI: 0.17-0.98, P = .044), 3-hydroxyoctanoate (OR: 0.78, 95% CI: 0.63-0.98, P = .033), tartarate (OR: 0.64, 95% CI: 0.42-0.98, P = .039), and X-23739 (uncharacterized; OR: 0.74, 95% CI: 0.58-0.93, P = .011). Risk associations (OR > 1) included 3-methoxytyramine sulfate (OR: 1.38, 95% CI: 1.03-1.86, P = .032), carnitine (OR: 1.38, 95% CI: 1.03-1.86, P = .032), homoarginine (OR: 1.74, 95% CI: 1.02-2.98, P = .042), isovalerate (i5:0; OR: 1.21, 95% CI: 1.06-1.39, P = .006), ascorbic acid 3-sulfate (OR: 1.11, 95% CI: 1.01-1.22, P = .037), and creatinine (OR: 1.01, 95% CI: 1.0005-1.0185, P = .037). Sensitivity analyses showed no significant pleiotropy or heterogeneity, and no reverse causality for most associations. Linkage disequilibrium score regression identified a significant genetic correlation only for 3-hydroxyoctanoate. This study identifies several CSF metabolites that may be associated with the risk of cerebral atherosclerosis. While the results are robust across sensitivity analyses, they should be interpreted cautiously, and further validation in larger, longitudinal cohorts is warranted before inferring definitive causal roles.