Comparing GLIM and SGA Nutritional Criteria for Malnutrition Assessment and Prognosis in Chronic Heart Failure...

Chronic heart failure (CHF) is a prevalent condition with high morbidity and mortality. Malnutrition is common in CHF patients and is associated with poor prognosis. The Subjective Global Assessment (SGA) and Global Leadership Initiative on Malnutrition (GLIM) criteria are widely used to assess nutritional status, but their prognostic value in CHF remains unclear. This study aimed to compare the effectiveness of SGA and GLIM criteria in assessing malnutrition and predicting adverse outcomes in CHF patients.

This retrospective cohort study included 240 CHF patients admitted between January 2022 and June 2024. Nutritional status was assessed using both SGA and GLIM within 48 hours of admission. The primary outcome was the occurrence of adverse events (worsening heart failure, readmission, or all-cause mortality) within 90 days post-discharge. Statistical analyses included Cohen's kappa for agreement, Receiver Operating Characteristic (ROC) curves for predictive value, and multivariate logistic regression to identify independent risk factors for adverse outcomes.

The agreement between SGA and GLIM criteria was good (Cohen's Kappa = 0.8). ROC analysis showed an AUC of 0.744 for SGA and 0.793 for GLIM in predicting adverse outcomes. The DeLong test revealed that GLIM had a significantly better predictive value (Z = -1.93, p = 0.043). Multivariate analysis identified malnutrition (both SGA and GLIM), smoking, and elevated BNP as independent risk factors for adverse outcomes. Nomograms incorporating these factors showed good predictive accuracy, with the GLIM model yielding a higher AUC of 0.854 compared to 0.816 for SGA.

Malnutrition was identified in 38.8% of patients when assessed by the SGA and in 40.0% when evaluated using the GLIM criteria. GLIM criteria are a reliable and superior tool for predicting adverse outcomes in CHF patients compared to SGA. Incorporating nutritional assessments, BNP, and smoking history into predictive models can enhance risk stratification and guide clinical decision-making in managing CHF patients.