Association between preoperative nutritional status and anastomotic leakage in patients undergoing rectal cancer surgery

Background Anastomotic leakage is a clinically important complication after rectal cancer surgery, and preoperative nutritional and inflammatory status may be associated with anastomotic healing. This study aimed to evaluate the association between preoperative nutritional status, particularly the prognostic nutritional index (PNI), and postoperative anastomotic leakage in patients undergoing rectal cancer surgery. Methods This retrospective study included consecutive eligible patients who underwent rectal cancer surgery between January 2023 and December 2025. Anastomotic leakage within 30 days was identified according to International Study Group of Rectal Cancer criteria. Preoperative nutritional status was primarily assessed using PNI, with Nutritional Risk Screening 2002 (NRS-2002), serum nutritional markers, and C-reactive protein (CRP) evaluated as complementary measures. Propensity score matching (PSM), multivariable logistic regression, unmatched-cohort analysis, sequential adjustment, subgroup analyses, and receiver operating characteristic analysis were performed. Results Among 2,468 patients, 308 developed anastomotic leakage. After 1:1 PSM, 308 patients with leakage and 308 without leakage were analyzed. Mean PNI was lower in patients with leakage (43.2 ± 5.1 vs. 46.8 ± 4.9; p < 0.001). Each 1-point increase in PNI was associated with lower adjusted odds of leakage (odds ratio [OR], 0.898; 95% confidence interval [CI], 0.861–0.936; p < 0.001), with a similar estimate in the unmatched cohort (OR, 0.904; 95% CI, 0.882–0.927). NRS-2002 ≥ 3 was associated with higher odds of leakage (OR, 1.72; 95% CI, 1.12–2.65). Additional adjustment for CRP attenuated but did not eliminate the PNI association (OR, 0.907; 95% CI, 0.870–0.945). The optimal PNI cutoff was 45.5, with an area under the receiver operating characteristic curve of 0.646, sensitivity of 62.7%, and specificity of 60.8%. Conclusion Lower preoperative PNI was independently associated with higher odds of postoperative anastomotic leakage after rectal cancer surgery. PNI may provide a readily available adjunct to multidimensional preoperative risk assessment but should not be considered a stand-alone predictor.