Abstract A014: Gastrointestinal Microbiome and Stroke Outcomes Network: Initial Results

Background: The gut microbiome is increasingly implicated in neurological disease, yet its relationship with stroke severity and recovery in humans remains unclear. We investigated the associations between gut microbial profiles and clinical outcomes in patients with acute ischemic stroke. Methods: This was a two-center cohort of 204 adult acute ischemic stroke patients. Stool samples were collected at baseline (n = 144) and 3 months (n = 65) and profiled via 16S rRNA gene sequencing. Stroke severity and disability were assessed using the NIH Stroke Scale (NIHSS) and the modified Rankin Scale (mRS) at both time points. Global associations between community composition and outcomes were tested with PERMANOVA. Taxon-level associations used multivariable regression and mixed-effect models, which adjusted for age, sequencing batch, and premorbid mRS. α was set to 0.05 for the global/community tests, and at FDR-adjusted <0.1 for the taxon tests. Change in NIHSS was defined as the 3-month score minus the baseline score. Results: Alpha-diversity (richness, evenness, Shannon) was not associated with any outcome. Baseline gut microbiome composition was associated with baseline NIHSS (p = 0.00019) and premorbid mRS (p = 0.000016), as well as with the change in NIHSS over time (p = 0.019). These findings remained significant after adjusting for age and sequencing batch, and the change in NIHSS association remained significant with additional adjustment for premorbid mRS (p = 0.020). No significant associations were observed between the baseline microbiota and 90-day outcomes, likely due to the smaller follow-up sample size (n = 65). At the taxonomic level, several baseline taxa were associated with baseline severity and longitudinal change; notably, higher abundance of Lachnospiraceae-ND3007 was associated with greater improvement in NIHSS (FDR < 0.10). Longitudinal mixed-effects models also identified multiple taxa correlated with NIHSS and mRS measures over time. Conclusion: Baseline microbiome composition at stroke onset was associated with premorbid disability, initial stroke severity, and change in NIHSS over time. Specific taxa (e.g., Lachnospiraceae-ND3007 ) may serve as biomarkers of recovery trajectories. These findings support a potential role for the microbiome in modulating stroke outcomes and warrant further study of microbiota-targeted interventions.