Association of triglyceride-glucose index with all-cause mortality in critically ill non-diabetic ischemic stroke: a retrospective cohort study based on the MIMIC-IV database

Despite the established association between the triglyceride-glucose (TyG) index and outcomes in some cardiovascular and cerebrovascular diseases, its prognostic role for all-cause mortality (ACM) in non-diabetic patients with ischemic stroke (IS) is not well-defined. Given the significant burden of IS, identifying such a marker in this specific subgroup is important. This study aimed to investigate the relationship between the TyG index and ACM in non-diabetic patients with IS using data from the MIMIC database. This retrospective cohort study analyzed data from the Medical Information Mart for Intensive Care-IV (MIMIC-IV, version 3.1) database to investigate the association between the TyG index and ACM in non-diabetic adult patients with IS. Patients were categorized into quartiles based on their TyG index, calculated from fasting triglyceride and glucose levels. ACM was assessed at the intensive care unit (ICU), during hospitalization, and at 30, 90, 180 days, and 1 year. Cox proportional hazards regression analysis, Kaplan-Meier survival curves, restricted cubic spline (RCS) analysis, and subgroup analyses were utilized to evaluate this association and explore its characteristics. This retrospective study included 1073 non-diabetic IS patients (51.2% female) from the MIMIC-IV database. The ACM rates at ICU, in-hospital, 30, 90, 180 days, and 1 year were 5.7%, 9.7%, 17.1%, 21.7%, 24.1%, and 26.6%, respectively. Multivariate Cox regression analysis showed that the highest TyG index quartile was significantly associated with increased HR for ACM compared to the lowest quartile: in-hospital (HR: 2.00; 95% CI: 1.09–3.65; P = 0.025), 30-day (HR: 1.59; 95% CI: 1.02–2.47; P = 0.041), 90-day (HR: 1.61; 95% CI: 1.09–2.38; P = 0.017), 180-day (HR: 1.65; 95% CI: 1.13–2.39; P = 0.009), and 1-year (HR: 1.53; 95% CI: 1.07–2.18; P = 0.020). RCS analysis showed a linear relationship between TyG index and mortality risk across all time points (p for non-linearity > 0.05). Subgroup interactions were significant for hypertension (ICU and 90-day mortality) and sex (30-day, 180-day, and 1-year mortality). Higher TyG index is an independent predictor of increased ACM in non-diabetic patients with IS. These findings suggest that TyG index may be a readily available marker for risk stratification in this patient population.