Abstract
BACKGROUND: The triglyceride-glucose (TyG) index and stress hyperglycemia ratio (SHR) are emerging biomarkers in cerebrovascular diseases, but their combined prognostic value for mortality in patients with critically ill stroke remains unexplored. This study, based on a cohort of critically ill patients with stroke, aims to investigate the prognostic value of the combined TyG index and SHR in predicting all-cause mortality at multiple time points in this high-risk population. METHODS: Based on the Medical Information Mart for Intensive Care (MIMIC)-IV database, 2998 critically ill patients with stroke requiring intensive care unit (ICU) admission were included. Patients were stratified into 8 groups based on the median of the TyG index and the quartiles of the SHR. The primary outcomes were 30-day and 365-day all-cause mortality; the secondary outcomes included 90-day and 180-day mortality. Cox proportional hazards regression models, restricted cubic splines (RCS) curves, subgroup analyses, and mediation analyses were employed to assess associations between the combined TyG index and SHR with all-cause mortality. RESULTS: The cohort had a median age of 72.63 years (IQR 61.27-82.69 years), with 51.33% male (1539/2998). Fully adjusted Cox proportional hazards models showed that compared to the reference group (TyG < 8.72 and SHR < 0.86), patients with TyG ≥ 8.72 and SHR ≥ 1.18 had the highest mortality risk (30-day HR 2.481, 95% CI 1.767-3.485; 365-day HR 1.954, 95% CI 1.532-2.493). RCS analysis confirmed linear positive correlations between the TyG index, SHR, and mortality at all time points (all P for non-linearity > 0.05). Subgroup analyses further demonstrated consistent associations between the combined TyG index and SHR on 30-day and 365-day all-cause mortality. Mediation analysis revealed that the highest SHR significantly mediated the association between high TyG and mortality in these patients (30-day mediation proportion: 48.82%, P = 0.024; 365-day: 22.93%, P = 0.004). CONCLUSION: The combination of high TyG (≥ 8.72) and elevated SHR (≥ 1.18) is significantly associated with increased short- and long-term mortality in critically ill patients with stroke. Integrated metabolic monitoring and early intervention targeting at these biomarkers may improve their prognosis.