Stroke-associated pneumonia (SAP) is a major complication of acute ischemic stroke. The brain-gut-lung axis suggests that post-stroke gut dysbiosis may offer independent risk information beyond clinical scores like A2DS2. The aim of this study was to evaluate the incremental predictive value of early post-stroke gut microbiome biomarkers for SAP and to construct a simplified microbiome risk score (M-score). In a prospective nested case-control study, we identified 63 SAP cases from a cohort of 551 patients with acute ischemic stroke and stool samples were collected 24-72 h post-admission. Cases were matched 1:2 (age, sex) to 126 controls. Gut microbiota was profiled via 16S rRNA sequencing. Three differentially abundant genera (prevalence ≥20%, FDR q < 0.10) were used to construct an M-score via bootstrapped regression. Independent associations were assessed with conditional logistic regression. Incremental value over the A2DS2 score was evaluated using AUC, continuous net reclassification improvement (NRI), and the Brier score. SAP cases exhibited reduced α-diversity and distinct β-diversity (both P < 0.01). Cases had higher Enterococcus and Streptococcus and lower Faecalibacterium levels (all q < 0.05). The M-score was independently associated with SAP (adjusted OR per 1-SD: 1.76, 95% CI: 1.30-2.39, P = 0.001). Adding the M-score to A2DS2 significantly improved prediction: AUC increased from 0.76 to 0.84 (ΔAUC = 0.08, P = 0.009), NRI was 0.31 (95% CI: 0.12-0.51), and the Brier score decreased from 0.18 to 0.16. Results were robust in sensitivity analyses. A gut microbiome score based on three genera provides significant independent and incremental predictive value for SAP over the A2DS2 score, enabling more precise early risk stratification after stroke.
Zhang et al. (Tue,) studied this question.