Effects of Preoperative Nutritional Status on Postoperative Delirium in Patients Undergoing Elective Non-cardiac Surgery: Implications for Preoperative Nursing Practice.

PURPOSE: To investigate the effect of preoperative nutritional status on postoperative delirium (POD) and develop a preliminary predictive model for patients undergoing elective non-cardiac surgery.

DESIGN

Prospective case-control study

METHODS

This study enrolled 315 adults undergoing elective non-cardiac surgery under general anesthesia (2023 to 2024). Nutritional risk was screened using the Nutritional Risk Screening 2002, and nutritional status was assessed according to the Global Leadership Initiative on Malnutrition criteria. Anthropometric measurements-body mass index, triceps skinfold thickness (TSF), mid-arm circumference, calf circumference, and waist circumference (WC)-and nutrition-related laboratory markers were recorded. POD was evaluated using the Confusion Assessment Method and its intensive care unit version. Patients were stratified into POD and non-POD groups. Binary logistic regression identified independent POD risk factors, and a preliminary predictive model was developed, with its performance assessed via receiver operating characteristic curve analysis and the DeLong test.

FINDINGS: A total of 67 patients (21.3%) developed POD, of which 60 cases (19%) occurred in the post-anesthesia care unit (PACU). The Mini-Mental State Examination score (odds ratio [OR] = 0.567, P = .003), TSF (OR = 0.233, P = .047), and WC (OR = 0.860, P = .003) were independent protective factors for POD, whereas major surgery (OR = 4.434, P = .041) and a PACU numerical rating scale score ≥4 (OR = 4.328, P = .017) were independent risk factors. The preliminary prediction model achieved an area under the curve of 0.889, with 77.6% sensitivity and 87.9% specificity.

CONCLUSIONS

A simple preoperative nutritional assessment incorporating anthropometric indicators (eg, TSF and WC) may enable early identification of patients at high risk for POD and support timely risk stratification and prevention in the PACU. Our combined predictive model demonstrated strong predictive performance for POD, supporting integration of these practical indicators into routine perioperative assessment.