Anastomotic leaks (ALs) are among the most serious complications following colorectal surgery, contributing to significant morbidity, reoperation, and prolonged hospitalisation. Poor preoperative nutritional status has been proposed as a modifiable risk factor; however, evidence from studies remains inconsistent. This systematic review, conducted according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines and registered in the International Prospective Register of Systematic Reviews (PROSPERO) (CRD420251034523), synthesised findings from 32 eligible studies, including randomised controlled trials, cohort studies, and case-control designs, published between 2005 and 2025. Across 474 initially identified articles, nutritional status was assessed using serum albumin, body mass index (BMI), Subjective Global Assessment (SGA), Nutritional Risk Index (NRI), or Prognostic Nutritional Index (PNI). AL incidence ranged from 2.8% to 11.3%, with hypoalbuminaemia and low nutritional indices consistently associated with increased risk. Several studies have suggested that nutritional optimisation, particularly immunonutrition and enteral support initiated 7-14 days preoperatively, improves secondary outcomes, such as wound infection rates, hospital stay, and overall morbidity. However, reductions in leak incidence are less consistent. The certainty of evidence linking poor nutritional status to leak risk was rated as moderate, while the evidence for nutritional interventions was rated as low due to heterogeneity and small sample sizes. Perioperative factors, including operative time, intraoperative blood loss, transfusion, and steroid use, were also significant contributors to leak risk. Overall, nutritional status is a key, modifiable predictor of AL; however, integration with surgical and perioperative optimisation is essential. High-quality multicentre trials are needed to define optimal nutritional strategies and establish standardised risk assessment tools for clinical practice.
Elfadil et al. (2026) studied this question.