Assessment of malnutrition in older patients with acute stroke by different nutritional screening tools and GLIM...

Aim: this study aimed to investigate the frequency of malnutrition and agreement rates in hospitalized older patients with acute stroke using different nutritional screening tools and the Global Leadership Initiative on Malnutrition (GLIM) criteria. Methods: a total of 121 older patients hospitalized for acute stroke were included in this study. All patients underwent screening for nutritional risk at the time of admission by the Malnutrition Universal Screening Tool (MUST), Nutritional Risk Screening-2002 (NRS-2002), and Subjective Global Assessment (SGA) tests. Additionally, the patients were evaluated using GLIM criteria to diagnose malnutrition. Stroke severity was assessed by using the National Institutes of Health Stroke Scale (NIHSS). Results: the mean age of the patients was 76 ± 9 years, and 69 % were female. The median level of NIHSS score was 7 (range, 0-24). According to SGA, 69 % of the patients were at risk for malnutrition (SGA: B or C). All patients (100 %) were at high risk of malnutrition when assessed by MUST, and 86.8 % were at risk of malnutrition according to the NRS-2002 score. On the other hand, 17.4 percent of the patients were diagnosed with malnutrition using GLIM criteria. A fair agreement was observed between NRS-2002 and SGA levels (Cohen's kappa coefficient: 0.263, p < 0.001). All three nutritional screening tools had high area under curve levels when investigated in the ROC curve analysis to detect malnutrition diagnosis determined by the GLIM criteria [MUST (AUC: 732; cut-off > 2; p < 0.001), SGA (AUC: 0.827; cut-off > 0; p < 0.001), and NRS-2002 (AUC: 0.859; cut-off > 3; p < 0.001)]. Conclusion: it was shown in this study that older patients with acute stroke are at high risk of malnutrition, and nearly one in every five patients is malnourished. All screening tools have performed well in detecting malnutrition among this study population.