Dose-response relationships between early postoperative nutrition and subsequent complications in gastrointestinal cancer: optimal intake ranges for energy and protein.

BACKGROUND AND AIM: Postoperative complications remain common after major gastrointestinal (GI) cancer surgery. Although early nutritional support is a core component of enhanced recovery protocols, optimal postoperative targets for energy and protein intake remain uncertain. This study aimed to characterize dose-response relationships between early postoperative nutritional intake and postoperative complications.

METHODS

This prospective observational cohort included adults undergoing elective gastrectomy or partial colectomy for GI malignancy. Postoperative energy and protein intake were quantified during postoperative days 1-2 (POD 1-2). Overall postoperative complications occurring after POD 2 and within 30 days after surgery were the primary outcome; infectious complications were secondary. Nonlinear associations were examined using multivariable logistic regression with restricted cubic splines (RCS) as the primary method, supported by generalized additive and quadratic models. Analyses were adjusted for demographic, surgical, oncologic, and nutritional covariates, including GLIM-defined malnutrition. Optimal intake ranges were defined using a CI-overlap criterion with bootstrap validation.

RESULTS

Among 642 patients, early postoperative energy intake demonstrated a statistically significant nonlinear association with complication risk (p-for nonlinearity = 0.018). The nadir occurred at 14.4 kcal/kg/day, with an optimal intake range of 12.8-20.1 kcal/kg/day. Protein intake showed a less consistent association, without statistically significant evidence of nonlinearity (p = 0.166). The nadir occurred at 0.81 g/kg/day, with an optimal intake range of 0.55-1.51 g/kg/day, with no consistent upper harm threshold observed within routine clinical ranges.

CONCLUSION

Early postoperative energy intake after GI cancer surgery demonstrated a robust nonlinear association with postoperative complications, with lowest predicted risk observed at moderate caloric provision. In contrast, protein intake showed a less consistent association and a wider estimated optimal range. These findings warrant further investigation of individualized, precision-oriented postoperative nutrition strategies in surgical oncology.