Evaluating the Diagnosis of Malnutrition Based on Global Leadership Initiative on Malnutrition

• Published in: Nutrients 2024-11, Vol.16 (22)
• Main Authors: Win, Phoo Pyae Sone, Chua, Denise Qian Ling, Gwee, Xinyi, Wee, Shiou Liang, Ng, Tze Pin
• Format: Article
• Language: English
• Subjects: Leadership; Malnutrition; Nutrition; Product/Service Evaluations
• ISSN: 2072-6643; 2072-6643
• DOI: 10.3390/nu16223823

Background: A minority of studies using the GLIM criteria for malnutrition diagnosis have performed formal empirical validation. Objectives: To evaluate the concurrent and predictive validity of GLIM criteria with and without prior screening among community-dwelling older adults in Singapore. Method: In the Singapore Longitudinal Aging Study (SLAS-2, n = 2477), malnutrition was diagnosed using single-step and two-step GLIM procedures using the Mini Nutritional Assessment Short Form (MNA-SF) and Elderly Nutritional Index for Geriatric Malnutrition Assessment (ENIGMA) for initial screening. Criterion validity was evaluated using MNA-Full Form (MNA-FF) as reference malnutrition diagnosis. Prognostic validity was evaluated using logistic and Cox regression analyses with respect to impaired quality of life (QoL) and 10-year mortality. Results: GLIM malnutrition with and without MNA-SF or ENIGMA screening showed significant associations with known clinical correlates; single-step GLIM malnutrition: sensitivity = 80%, specificity = 83%; two-step MNA-SF-GLIM malnutrition: sensitivity = 80%, specificity = 85%; two-step ENIGMA-GLIM malnutrition: sensitivity = 74%, specificity = 88%; positive predictive values of around 20% and negative predictive values above 98%. Cohen’s kappa values of agreement were uniformly low (0.26 to 0.32). All showed significant associations with about 50% increased odds of impaired QoL and 10-year mortality, adjusted for age, sex, ethnicity, education levels, and housing type, with the ENIGMA-GLIM malnutrition showing the highest risk estimates. Compared to MNA-FF malnutrition prevalence of 4.1%, GLIM-based malnutrition increased prevalence (14.6% to 19.7%) estimates. Conclusions: The GLIM criteria showed good construct and criterion validity. It increased the number of individuals diagnosed with malnutrition. The agreement between diagnoses of malnutrition was low. Diagnostic and prognostic accuracy vary with the screening instrument used. Early identification of malnutrition using appropriate tools can provide opportunities to delay or prevent the risk of important adverse outcomes such as impaired QoL and mortality.