Optimizing Nutritional Assessment in Gastrointestinal Oncosurgery: A Comparative Analysis of GLIM and PG-SGA

Objective This prospective cohort study evaluates the diagnostic efficacy of the Global Leadership Initiative on Malnutrition (GLIM) criteria and the Patient-Generated Subjective Global Assessment (PG-SGA) in identifying malnutrition among patients undergoing elective gastrointestinal tumor resection. It further examines associations between malnutrition, clinical biomarkers, body composition, and postoperative quality of life (QoL). Methods Nutritional risk screening (NRS2002) and subsequent malnutrition diagnosis (GLIM/PG-SGA) were performed in 216 patients from a tertiary care center. Correlations between malnutrition severity, biochemical indices, and QoL were analyzed. Multivariate regression identified risk factors for malnutrition. Results Malnutrition prevalence was 74.5% (GLIM) and 87.5% (PG-SGA). Both tools inversely correlated with albumin, prealbumin, hemoglobin, and 6-month QoL scores. GLIM-associated risk factors included advanced age ( β : 0.181, p  = 0.013), combined with cardiovascular and metabolic diseases (CVMD; β : -0.253, p <0.01), and tumor location ( β : -0.234, p <0.01). PG-SGA correlated with elevated neutrophil-to-lymphocyte ratio (NLR; β : 0.181, p  = 0.007) and TNM stage > II ( β : 0.212, p  = 0.002). Inter-method agreement was negligible (κ = 0.07). Conclusions Selection of nutritional assessment tools could align with clinical objectives: GLIM prioritizes current physiological metrics and tumor-specific factors (e.g., sarcopenia and gastric cancer), while PG-SGA to some extent reflects tumor burden (e.g., TNM staging and NLR).