Diagnostic performance of inflammatory and nutritional indices in discriminating low-risk and advanced colon adenomas: a retrospective cohort study

Aims: This study aimed to compare the diagnostic performance of cell count-based inflammatory indices and composite inflammatory-nutritional indices in distinguishing advanced from low-risk colon adenomas, hypothesizing that composite indices (particularly the C-reactive protein-to-albumin ratio (CAR)) would demonstrate superior discriminative ability. Methods: This single-center, retrospective cohort study included 406 patients with colon adenomas who underwent colonoscopy during 2024-2025. Patients were classified as low-risk (G2) or advanced adenoma (G3). Cell count based inflammatory indices (neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), monocyte-to-lymphocyte ratio (MLR), Systemic Immune-inflammation Index (SII)) and composite inflammatory-nutritional indices (CAR, Prognostic Nutritional Index (PNI)) were calculated from pre-procedural laboratory data. Group comparisons, ROC analyses and multivariable binary logistic regression incorporating age, sex, total polypectomy count and CAR were performed. Results: Of 406 patients, 202 (49.8%) had low-risk and 204 (50.2%) had advanced adenomas. Advanced adenomas were associated with significantly higher NLR, PLR, MLR, SII and CAR and lower PNI (all p<0.001). Receiver operating characteristic analysis showed CAR had the highest discriminative performance (AUC=0.764, 95% CI: 0.716-0.812), followed by PNI (AUC=0.706, 95% CI: 0.655-0.757). A CAR cut-off of 1.2438 yielded 70.1% sensitivity and 69.8% specificity. In multivariable logistic regression, CAR remained an independent predictor of advanced adenoma (OR=1.647, 95% CI: 1.400-1.938; p<0.001) after adjustment for age, sex and total polypectomy count. Male sex (OR=0.594 for female vs. male, 95% CI: 0.370-0.954; p=0.031) and total polypectomy count (OR=1.431, 95% CI: 1.246-1.642; p<0.001) were also independently associated with advanced adenoma. Conclusion: CAR demonstrated the strongest diagnostic and independent predictive performance in both univariable and multivariable analyses, suggesting that composite inflammatory-nutritional biomarkers may serve as adjunctive tools for adenoma risk stratification. Prospective multicenter validation is warranted before clinical implementation.