Objective: Delirium is a frequent and serious neurocognitive complication among older and frail patients in intensive care units (ICUs). The prediction of delirium is crucial for early intervention. The triglyceride-glucose index (TGI), a surrogate marker of insulin resistance, has been associated with metabolic and vascular dysfunction, but its relationship with delirium remains unclear. We aimed to investigate the association between TGI and the development of delirium in frail and older ICU patients.

Material and Method: This prospective, double-blind observational study included frail patients aged ≥65 years admitted to tertiary ICUs between April 2024 and March 2025. Frailty was defined as a Rockwood Clinical Frailty Score ≥5. Delirium was diagnosed using the CAM-ICU scale. Fasting triglyceride and glucose levels measured within 24 hours of ICU admission were used to calculate TGI. Categorical variables were compared using Chi-square tests, while continuous variables were analysed using Student’s t-test or Mann-Whitney U-test, depending on distribution. ROC curve analysis was performed to assess the predictive value of TGI and other parameters. Multivariate logistic regression was used to identify independent predictors of delirium. A p-value <0.05 was considered statistically significant.

Results: Of 206 included patients (mean age 72.1±4.6 years), 35.4% developed delirium, predominantly of the hyperactive subtype (84.9%). TGI values were significantly higher in patients with delirium (p <0.001). ROC analysis demonstrated a strong predictive ability for TGI (cut-off >4.85, AUC=0.909, sensitivity 86.3%, specificity 87.2%). Multivariate logistic regression identified high TGI, elevated BMI, and male sex as independent predictors of delirium. TGI was not correlated with APACHE II or SOFA scores. Sedation and physical restraint were significantly more frequent in delirious patients.

Conclusion: TGI is a strong and independent predictor of delirium in frail and older ICU patients. Its ease of calculation from routine laboratory values makes it a promising biomarker for the early identification and targeted prevention of delirium in high-risk populations.