Assessing the bidirectional causal relationship between malnutrition, self-reported fatigue, and stroke: A 2-sample Mendelian randomization study.

This study aims to investigate the causal relationships between malnutrition, fatigue, and stroke (including its subtypes) using a 2-sample bidirectional Mendelian randomization (MR) analysis. This study utilized publicly available Genome-Wide Association Study summary statistics from the Genome-Wide Association Study Catalog and the MEGASTROKE consortium. A 2-sample MR approach was employed to explore the causal effects of malnutrition and fatigue on stroke risk. Instrumental variables were rigorously selected based on strict criteria. The primary analysis used the inverse-variance weighted method, supplemented by weighted median, MR Egger regression, and other MR methods. Sensitivity analyses, including Cochran Q test, MR Egger intercept test, and MR-PRESSO, were conducted to assess heterogeneity and pleiotropy. Gene Ontology enrichment analysis was performed to explore biological pathways. All analyses were conducted using R packages, including "TwoSampleMR," "mr.raps," and "MR-PRESSO." Self-reported fatigue showed a significant causal association with any stroke, with an odds ratio of 1.43 (95% confidence interval: 1.04-1.96, P = .028) using the inverse-variance weighted method, supported by the weighted median method. However, no significant causal relationships were observed between fatigue and stroke subtypes (any ischemic stroke, large-artery atherosclerotic stroke, cardioembolic stroke, and small-vessel disease). Malnutrition was not causally associated with any type of stroke or its subtypes. Additionally, no significant causal effects of stroke or its subtypes on fatigue or malnutrition were detected. Sensitivity analyses confirmed the absence of heterogeneity or pleiotropy. Gene Ontology enrichment analysis revealed that biological pathways related to fatigue and any stroke primarily involved positive regulation of CREB transcription factor activity, synaptic function, and plasma membrane-related processes. Fatigue is a risk factor for any stroke, and its mediating biological pathways may include CREB transcription factor activity regulation, synaptic function, and plasma membrane-related processes. No causal relationships were identified between malnutrition and stroke or between stroke and fatigue.