Fengwei Yao, Linquan Wu, Enliang Li
The combined assessment of TyG and SHR may provide a novel approach for integrating chronic metabolic vulnerability and acute stress hyperglycemia, thereby improving delirium risk stratification in critically ill surgical and trauma patients.
BACKGROUND: Delirium is a frequent acute neuropsychiatric complication in critically ill surgical patients and is associated with poor clinical outcomes. Nutritional-metabolic disturbances, including insulin resistance and stress-related glycemic dysregulation, may contribute to delirium susceptibility. This study investigated the independent and joint associations of the triglyceride-glucose index (TyG) and stress hyperglycemia ratio (SHR) with delirium in critically ill surgical patients.
METHODS: Adult patients admitted to surgical or trauma intensive care units were included. Delirium was identified using the Confusion Assessment Method for the Intensive Care Unit. Multivariable logistic regression, restricted cubic spline analysis, receiver operating characteristic curve analysis, subgroup analysis, Boruta feature selection, machine learning models, and SHapley Additive exPlanations were applied.
RESULTS: Among 1,370 eligible patients, 198 (14.45%) developed delirium. After full adjustment, the highest SHR tertile was associated with a higher odd of delirium compared with the lowest tertile [odds ratio (OR) 3.18, 95% confidence interval (CI) 1.87-5.42; p < 0.001]. The highest TyG tertile was also associated with increased delirium odds (OR 1.87, 95% CI 1.13-3.11; p = 0.015). When analyzed as continuous variables, both SHR (OR 1.72, 95% CI 1.20-2.45; p < 0.001) and TyG (OR 1.93, 95% CI 1.39-2.67; p < 0.001) remained independently associated with delirium. Patients with concomitantly high TyG and high SHR had the greatest delirium odds (OR 3.57, 95% CI 2.00-6.36; p < 0.001). TyG showed a linear association with delirium, whereas SHR showed a nonlinear relationship. The combined TyG-SHR indicator showed moderate discrimination [area under the curve (AUC) 0.716]. The GBM model achieved an AUC of 0.853 (95% CI: 0.804-0.902) on the test set.
CONCLUSION: The combined assessment of TyG and SHR may provide a novel approach for integrating chronic metabolic vulnerability and acute stress hyperglycemia, thereby improving delirium risk stratification in critically ill surgical and trauma patients.