Subtype-associated gut microbial taxa and their clinical correlates in acute ischemic stroke: comparison of large-artery atherosclerosis and small-vessel occlusion.

OBJECTIVE: The gut microbiome gets altered during ischemic stroke (IS); however, whether distinct microbial signatures characterize specific IS subtypes, such as large-artery atherosclerosis (LAA) and small-vessel occlusion (SVO), and their relation to subtype-specific clinical features, remains unclear. This study aimed at comparing gut microbiome profiles among patients with LAA, patients with SVO, and healthy controls (HCs); further, the associations between subtype-enriched microbial taxa and clinical parameters were examined. METHODS: We compared gut microbiome profiles and clinical associations among patients with LAA, patients with SVO, and HCs (n = 50 per group), using 16S rRNA sequencing to analyze microbial diversity and taxonomic composition. Correlation analyses were then performed between clinical parameters and the identified differential taxa. RESULTS: Linear discriminant analysis effect size analysis showed that, compared with HCs, patients with IS had reduced alpha diversity and distinct beta diversity. No significant differences were observed between patients with LAA and SVO in alpha diversity (all indices, p > 0.05) or beta diversity (R 2 = 0.009, p = 0.595). Taxa enriched in HCs (Agathobaculum, Holdemanella, and Prevotella) were negatively associated with age, D-dimer levels, and modified Rankin Scale scores. LAA-enriched taxa (Desulfovibrio and Eubacterium) were positively correlated with Fazekas scale scores. SVO-enriched taxa (Catenibacterium, Collinsella, Megamonas, Sellimonas, and Weissella) were positively correlated with hemoglobin A1c, body mass index, and National Institutes of Health Stroke Scale scores. CONCLUSION: Although overall microbial diversity did not differ between patients with LAA and SVO, subtype-enriched taxa were associated with distinct clinical features. These findings support the presence of subtype-specific microbial patterns that may reflect different pathophysiological processes; however, causal relationships remain to be established.