Blood Endotoxin Level Predicts Poor Stroke Recovery in Patients With Type 2 Diabetes
A gut-derived bacterial toxin in the blood at hospital admission independently predicted worse 90-day outcomes in diabetic stroke patients.
Zusammenfassung
Stroke outcomes vary widely, and people with type 2 diabetes often recover less well. This study looked at endotoxin, or lipopolysaccharide (LPS), a bacterial component that can leak into the blood when the gut barrier is impaired. Researchers measured plasma LPS at admission in 300 diabetic patients with acute ischemic stroke and followed them for 90 days. Patients with poor outcomes (modified Rankin Scale 3-6) had higher LPS levels. LPS tracked with stroke severity, C-reactive protein, and white blood cell count. After adjustment for age, stroke severity, HbA1c, atrial fibrillation, kidney function, and other factors, each standard deviation increase in log-LPS raised the odds of poor outcome by 86%. Adding LPS to a basic prediction model improved discrimination (AUC 0.79 to 0.83). LPS may be a useful, accessible prognostic marker, though the study shows association rather than causation.
Detaillierte Zusammenfassung
Diabetes worsens stroke recovery, yet clinicians have limited tools for predicting which diabetic patients will do poorly. Endotoxemia, meaning circulating bacterial lipopolysaccharide (LPS), often reflects a leaky gut barrier and drives systemic inflammation. Its prognostic value in diabetic stroke had not been defined.
This single-center prospective cohort enrolled 300 patients with acute ischemic stroke and type 2 diabetes. Plasma LPS was measured early after admission. Researchers examined how it related to stroke severity (NIHSS) and inflammatory markers, and whether it independently predicted an unfavorable 90-day functional outcome, defined as modified Rankin Scale 3-6.
Patients with poor outcomes had significantly higher LPS levels. Log-transformed LPS correlated positively with baseline NIHSS, high-sensitivity C-reactive protein, and white blood cell count. After adjustment for age, NIHSS, HbA1c, atrial fibrillation, eGFR, and other covariates, each standard deviation increase in log-LPS was linked to higher odds of poor outcome (OR 1.86, 95% CI 1.39-2.48). Risk of poor outcome and mortality rose across LPS quartiles. Adding LPS to a basic clinical model raised the AUC from 0.79 to 0.83, improved risk reclassification, and offered greater net clinical benefit across decision thresholds.
The findings suggest LPS could help refine early risk stratification in diabetic stroke patients and support interest in the gut-inflammation-brain axis as a target for future study.
Caveats: this analysis rests on the abstract alone. The single-center design limits generalizability, and observational data cannot show that endotoxin causes worse outcomes or that lowering it would help. LPS assay variability, the lack of external validation, and the modest AUC gain also warrant caution.
Wichtigste Erkenntnisse
- Patients with poor 90-day outcomes (mRS 3-6) had significantly higher admission plasma LPS than those with good outcomes.
- Each standard deviation rise in log-LPS independently raised odds of unfavorable outcome (OR 1.86, 95% CI 1.39-2.48).
- Log-LPS correlated positively with NIHSS score, high-sensitivity CRP, and white blood cell count.
- Risk of poor outcome and mortality increased progressively across LPS quartiles, showing a dose-response pattern.
- Adding LPS to a basic clinical model raised AUC from 0.79 to 0.83 and improved reclassification.
Methodik
Single-center prospective cohort of 300 acute ischemic stroke patients with type 2 diabetes. Plasma LPS was measured early after admission, and the primary endpoint was 90-day unfavorable outcome (mRS 3-6). Multivariable logistic regression, quartile analysis, AUC, reclassification, and decision curve analysis assessed predictive value.
Studienlimitierungen
Single-center observational design limits generalizability and cannot establish causality; the gut origin of LPS was not demonstrated in the abstract. The AUC improvement is modest, there was no mention of external validation, and LPS assay standardization and residual confounding remain concerns. This summary is based on the abstract only.
Hat dir diese Zusammenfassung gefallen?
Erhalte die neueste Longevity-Forschung jede Woche in deinen Posteingang.
E-Mail-Adresse zum Abonnieren eingeben: