Work

Global leadership initiative on malnutrition-based nutritional phenotype and factors associated with the outcomes of hospitalized patients managed by a nutrition support team: A retrospective study

Year: 2025

Type: article

Abstract: BACKGROUND AND AIMS: What are the Global Leadership Initiative on Malnutrition (GLIM) criteria leading to the diagnosis of malnutrition at hospital, specifically in diabetic patients, is not well known. The factors associated with the clinical outcomes of nutritionally at-risk hospitalized patients have been scarcely investigated. AIMS: To characterize the GLIM-based nutritional phenotype of hospitalized patients managed by a Nutrition Support Teams (NST) (primary), and to compare the GLIM malnutrition diagnostic criteria between diabetic and non-diabetic patients (secondary); to identify the factors associated with hospital readmission one month after discharge and one-month mortality (secondary). METHODS: We conducted a retrospective observational study including all the consecutive adult patients managed by the NST of a tertiary university hospital. The exclusion criteria were age under 18 yr and absence of assessment of body composition by bioimpedance analysis (BIA). Malnutrition was diagnosed according to the GLIM criteria: the etiologic criterion was the hospitalization for an acute or chronic disease, and three phenotypic criteria were: low body mass index (BMI), weight loss, and/or low fat-free mass index (FFMI) derived from BIA. Food intake was assessed by the Simple Evaluation of Food Intake® (SEFI®). Handgrip strength was measured using a dynamometer. Hospital readmission and death one month after discharge were collected from the electronic medical records (EMR). Multivariable logistic regression analyses using a stepwise top-down selection identified factors associated with one-month hospital readmission and mortality. RESULTS: 231 patients were included: 43 % of females, age (mean ± SD), 62.2 ± 15.9 yr, BMI, 23.5 ± 6.4, 73.6 % (170/231) were malnourished, of whom 66.5 % (113/170) had severe malnutrition, and 20.3 % (47/231) had diabetes. Malnutrition was diagnosed according to the three phenotypic criteria in 59/170 patients (34.7 %), two in 56/170 patients (33.1 %), and only one in 54/170 patients (31.8 %). The diabetic patients were less likely to have a low FFMI (29.8 % (14/47) vs 60.9 % (109/179), P < 0.001) or be malnourished (61.7 % (29/47) vs 76.6 % (141/184), P = 0.04), but more likely to have a low handgrip strength (64.3 % (27/42) vs 45.6 % (68/149), P = 0.03) than non-diabetic patients. Multivariable analyses showed that cancer (odds ratio (OR) = 2.51 [1.17-5.37], P = 0.02) and SEFI® score (OR = 0.89 [95 % confidence interval, 0.79-0.99], P = 0.04) were associated with higher risk of hospital readmission, whereas weight loss (OR = 3.20 [1.26-8.17], P = 0.01) and diabetes (OR = 2.76 [1.10-6.91], P = 0.03) were associated with mortality. CONCLUSION: One third of patients were diagnosed malnutrition with one phenotypic criterion, suggesting that all the GLIM phenotypic criteria should be used to ensure the diagnosis of malnutrition. The diabetic patients had higher FFMI and lower handgrip strength than non-diabetic patients. In patients managed by NST, low food intake and cancer are associated with hospital readmission, while diabetes and weight loss with early mortality.

Source: Clinical Nutrition ESPEN

DOI: 10.1016/j.clnesp.2025.102889

Authors: Mathilde Barrallier, Emeric Rageul, Charlotte Nesseler, Quentin Le Cornu, Ronan Thibault