Agreement of nutritional screening tools with GLIM-defined malnutrition in older hospitalized people

Malnutrition remains a major public health challenge, particularly among hospitalized older adults, where it is associated with increased morbidity and mortality. This study examined the agreement of four commonly used nutrition screening tools-Mini Nutritional Assessment Short-Form (MNA-SF), Nutritional Risk Screening-2002 (NRS-2002), Graz Malnutrition Screening (GMS), and Malnutrition Screening Tool (MST)-with Global Leadership Initiative on Malnutrition (GLIM)-defined malnutrition in hospitalized older adults. In this cross-sectional study, 842 participants aged ≥65 years admitted to Bandırma Training and Research Hospital between June and October 2024 were assessed. Anthropometric and nutritional data were collected using standardized procedures, and sensitivity, specificity, receiver operating characteristic (ROC) analyses, and agreement measures were calculated in relation to GLIM classification. The prevalence of GLIM-defined malnutrition was 24.5%. MNA-SF showed strong overall agreement with GLIM-defined malnutrition and high observed specificity (91.2%), whereas GMS showed high observed sensitivity (94.2%) in this sample. MST and NRS-2002 also demonstrated agreement with GLIM classification, with Cohen's kappa indicating moderate and fair agreement, respectively. Compared with non-malnourished participants, malnourished participants had longer hospital stays, lower body mass index, and smaller mid-upper arm and calf circumferences. Among the evaluated tools, MNA-SF showed close alignment with GLIM-defined malnutrition in hospitalized older adults. These findings should be interpreted as agreement with GLIM-based classification rather than evidence of definitive diagnostic superiority. Tailored screening approaches may support earlier recognition of nutritional problems in this vulnerable population. Further research is needed to identify the most appropriate screening strategies for older adults across different clinical settings.