Preoperative geriatric nutritional risk index as a predictor of intraoperative hypothermia in esophageal cancer surgery

Esophageal cancer is life-threatening, and esophagectomy patients have a high incidence of intraoperative hypothermia (linked to adverse postoperative outcomes), with no stable predictive indicator available. The Geriatric Nutritional Risk Index (GNRI), a validated tool for predicting cancer prognosis/complications, has not been studied for its association with intraoperative hypothermia in esophageal cancer surgery. This study evaluated their association and GNRI's predictive efficacy.

A retrospective analysis of 676 esophageal cancer surgery patients (Fujian Provincial Cancer Hospital, July 2022-February 2025) was conducted. Logistic regression assessed GNRI-hypothermia association, restricted cubic splines (RCS) explored nonlinear relationships, and receiver operating characteristic (ROC) curves evaluated predictive efficacy. Intraoperative hypothermia was defined as core temperature ≤ 36°C.

Among the 676 patients, the incidence of intraoperative hypothermia was 41.4%. Multivariate logistic regression analysis indicated that patients in the nutritional risk group (GNRI < 98) had a significantly higher risk of intraoperative hypothermia compared to those without nutritional risk (GNRI≥ 98). The adjusted odds ratio (OR) was 1.95 (95% confidence interval [CI]: 1.34-2.82, P < 0.05) in the fully adjusted model. The RCS analysis confirmed a nonlinear relationship between GNRI and intraoperative hypothermia (P < 0.0001), identifying a threshold at 93, which indicated a progressive increase in hypothermia risk when GNRI was below this point. Furthermore, GNRI demonstrated a sensitivity of 0.511 and a specificity of 0.852 for predicting intraoperative hypothermia; its incorporation significantly improved the area under the curve (AUC) of the prediction model (P = 0.048).

Preoperative GNRI is an independent predictor of intraoperative hypothermia in patients undergoing esophageal cancer surgery. The thresholds of 93 and 98 can serve as clinical reference points for nutritional interventions, offering a novel assessment tool for perioperative temperature management in esophageal cancer patients.