Association of human plasma and cerebrospinal fluid metabolomes with vascular dementia and its subtypes: A Mendelian...

Vascular dementia (VaD) is one of the most common subtypes of dementia after Alzheimer's disease. Investigating body fluid metabolites is critical for understanding VaD pathophysiology and identifying potential therapeutic targets. This study employs Mendelian randomization (MR) analysis to explore the causal relationship between body fluid metabolites and VaD.

Data for VaD were retrieved from the FinnGen database. 1,400 plasma metabolites were collected from the GWAS Catalog. 338 cerebrospinal fluid (CSF) metabolites data were obtained from a subset of participants in the WADRC and WRAP studies. The inverse-variance weighted (IVW) method was used to explore causal relationships between plasma/CSF metabolites and VaD, with supplementary analyses using Weighted mode, MR-Egger, and Weighted median methods. Multiple sensitivity analyses were conducted for robustness.

Following strict validation and FDR correction, significant associations (p_fdr < 0.05) were identified exclusively in plasma metabolites. The most significant metabolite was N-acetyl-aspartyl-glutamate (NAAG), with higher NAAG levels linked to reduced risks of VaD of acute onset and SVaD. Metabolonic lactone sulfate also showed significant associations across multiple disease groups, with elevated levels associated with lower disease risk, supported by FDR correction and sensitivity analyses. No significant CSF metabolites were identified after FDR correction. Disparities between CSF and plasma metabolites in disease-risk expression were observed, with only partial overlap in causal relationships (IVW, p < 0.05).

This study identified fluid metabolite biomarkers associated with VaD through Mendelian randomization, offering new insights and strategies for the prediction and treatment of VaD.