Abstract
BACKGROUND: Malnutrition has numerous adverse effects on the treatment and prognosis of diabetic foot (DF) patients, making it essential to determine the nutritional state to recognize malnutrition as early as possible. However, there is currently no acknowledged nutritional screening instrument for DF patients. This research aimed to identify the most appropriate nutritional assessment tool for this population. METHODS: We conducted a cross-sectional study with 247 DF patients. Nutritional assessments were performed using Nutritional Risk Screening 2002 (NRS2002) and the Mini-Nutritional Assessment short form (MNA-SF). The comparisons between scales were carried out based on the Global Leadership Initiative on Malnutrition (GLIM) criteria. The Cohen's kappa (k) and the area under the receiver operating characteristic curve (AUC) were analyzed to measure the diagnostic agreement of malnutrition among the screening tools and the GLIM criteria. RESULTS: Ninety-eight patients (39.68%) were diagnosed with malnutrition according to the GLIM criteria. The detection rates of MNA-SF and NRS2002 were 48.18 and 42.51%, respectively. MNA-SF was better correlated with the GLIM criteria, with a higher Kappa value (0.665 vs. 0.535) and a greater area under the receiver operating characteristic curve (0.860 vs. 0.792) than NRS2002. Additionally, MNA-SF and NRS2002 had similar specificity (79.2 vs. 85.2%), but MNA-SF demonstrated higher sensitivity (89.8 vs. 67.4%). CONCLUSIONS: This study is the first to describe the malnutrition diagnostic capacity of nutritional screening tools (MNA-SF and NRS2002) compared with the GLIM criteria. Our results indicate that the incidence of malnutrition is relatively high among DF patients, and the MNA-SF showed better sensitivity and correlation with the GLIM diagnostic criteria for malnutrition than NRS2002. Therefore, MNA-SF is more recommended for screening malnutrition in the DF population.