Triglyceride-Glucose Index as a Predictor of Short-Term Outcomes in Acute Ischemic Stroke: A Prospective Observational Study From Bangladesh

Background Acute ischemic stroke (AIS) is a primary cause of death and disability globally, imposing a significant burden in Bangladesh. Preliminary prognostic evaluation continues to be difficult, especially in resource-constrained environments. The triglyceride-glucose (TyG) index, an indicator of insulin resistance (IR), has surfaced as a possible predictor of poor outcomes in stroke. This study aimed to evaluate the association and predictive performance of the TyG index for short-term outcomes among patients with AIS. The primary objective was to assess its association and predictive ability for in-hospital mortality, while the secondary objective was to evaluate its association with 90-day functional outcome. Methods This prospective observational study was carried out in the Department of Neurology, National Institute of Neurosciences and Hospital (NINS), Dhaka, from January 2024 to June 2025. A total of 120 patients with radiologically confirmed first-ever AIS were included, of whom 118 completed the 90-day follow-up. Demographic data, clinical status, comorbidities, admission stroke severity, and laboratory values were gathered. Patients were observed for in-hospital mortality, and 90-day functional outcomes were evaluated using the modified Rankin Scale (mRS). The TyG index was calculated from fasting triglyceride and fasting plasma glucose levels measured within 24 hours of admission. Results The average age of participants was 63.38 ± 14.49 years, with 61.9% (n=73) being male. The in-hospital mortality rate was 28.8% (n=34). A higher TyG index was strongly linked with in-hospital mortality (median 9.49 vs. 8.85, p-value: 0.001). In multivariate logistic regression analysis, the TyG index remained an independent predictor of death (OR: 2.84; 95% CI: 1.43-5.64; p-value: 0.003). The ROC analysis showed moderate predictive performance (AUC: 0.680, p-value: 0.002), with an appropriate cutoff value of ≥8.87 (sensitivity 79.41%, specificity 51.19%). At 90 days, poor functional outcome was observed in 105 (89.0%) patients, which may reflect the tertiary-care hospital setting and inclusion of clinically more severe admitted stroke cases. A higher TyG index was significantly associated with poor functional outcome (OR: 2.21, 95% CI: 1.023-4.798; p=0.044). Patients with a higher TyG index had significantly lower survival rates, according to Kaplan-Meier analysis (log-rank p-value: <0.001). Conclusion A higher TyG index was significantly associated with increased mortality and poor functional outcomes among patients with AIS. However, its discriminative ability was moderate; therefore, the TyG index may be considered only as a simple, low-cost adjunct marker for early risk assessment rather than a standalone prognostic tool, particularly in resource-limited settings.