Cardiovasc Diabetol. 2026 Jun 28. doi: 10.1186/s12933-026-03270-5. Online ahead of print.
ABSTRACT
BACKGROUND: Hemorrhagic stroke (HS), including non-traumatic intracerebral hemorrhage (ICH) and subarachnoid hemorrhage (SAH), accounts for approximately 30% of all stroke cases and over 40% of stroke-related deaths. With high mortality and disability rates, HS imposes a heavy global health burden. As a core metabolic disorder, insulin resistance (IR) has been proven to be associated with all-cause mortality (ACM) in patients with ischemic stroke in previous studies. Nevertheless, its prognostic value remains unclear in critically ill HS patients admitted to the intensive care unit (ICU). Given the markedly higher mortality and morbidity of HS compared with ischemic stroke, it is essential to explore this association. This study aimed to investigate the correlations between multiple surrogate markers of insulin resistance and all-cause mortality among critically ill HS patients in the ICU setting.
METHODS: Data were extracted from the public eICU-CRD database. Patients with severe HS were identified based on the International Classification of Diseases (ICD)-9/10 diagnostic codes. A total of 1538 ICU-admitted patients with severe HS were enrolled and stratified according to quartiles of various IR surrogate markers. The primary endpoint was in-hospital mortality. Cox regression analysis, Kaplan-Meier survival curves, restricted cubic splines (RCS) and receiver operating characteristic (ROC) curves were adopted for statistical analyses.
RESULTS: Among the 1538 enrolled patients, males accounted for 54.6%, and the overall in-hospital all-cause mortality was 26.59%. Multivariate Cox regression analyses revealed that all IR surrogate markers were significantly correlated with all-cause mortality in severe HS patients. SPISE was negatively correlated with all-cause mortality, while other indicators showed positive correlations. Restricted cubic spline analyses demonstrated non-linear relationships between TyG, SPISE, TG_HDL, METS_IR, TyG_BMI, TyG_RC and mortality. No significant effect modification was observed in interaction analyses. ROC curve analysis indicated that TyG exhibited the highest predictive accuracy.
CONCLUSION: In conclusion, insulin resistance surrogate markers were significantly associated with all-cause mortality in critically ill HS patients. Despite their weak-to-moderate discriminative performance, these indices may serve as auxiliary prognostic references for risk stratification. Clinical application of these indicators is expected to optimize therapeutic strategies and disease progression management. Furthermore, this study enriches current evidence regarding the association between insulin resistance surrogate markers and hemorrhagic stroke, and clarifies their roles in predicting mortality across different stroke subtypes.
PMID:42366337 | DOI:10.1186/s12933-026-03270-5
