Causal Associations Between Plasma Metabolites and Stroke Risk Identified by Mendelian Randomization

Stroke is a leading cause of global mortality and morbidity, yet its underlying metabolic mechanisms remain largely undefined. Elucidating the causal role of blood metabolites in stroke pathogenesis is crucial for developing effective interventions. This study employed a two-sample Mendelian randomization (MR) framework to investigate the causal relationships between 1400 plasma metabolites and the risk of stroke and its subtypes. We integrated genome-wide association study (GWAS) data on metabolites from 3441 to 8299 European-ancestry individuals in the Canadian Longitudinal Study on Aging (CLSA) with European-ancestry stroke GWAS data from the MEGASTROKE consortium. IVW associations were interpreted together with Benjamini-Hochberg false-discovery-rate (FDR) correction, sensitivity analyses, and mRnd power estimates. The IVW screening analysis identified 235 nominal associations (P_IVW < 0.05) across 124 metabolites and ratios. Among the final prioritized findings, plasma free asparagine (OR = 0.852, 95% CI: 0.782-0.929, FDR = 0.033) and the glutamine-to-asparagine ratio (OR = 1.148, 95% CI: 1.068-1.235, FDR = 0.033) showed FDR-significant associations with cardioembolic stroke. Additional associations for steroid sulfates, bile-acid conjugates, and octadecenedioylcarnitine were nominally significant and supported by directionally consistent sensitivity analyses but did not pass FDR correction. Our MR analysis prioritizes amino-acid metabolism in cardioembolic stroke and suggests additional lipid-related metabolic pathways for follow-up. Because several associations were exploratory after FDR correction and all GWAS datasets were restricted to European ancestry, replication in larger and more diverse populations is warranted before clinical translation.