Different combinations of the GLIM criteria for patients awaiting a liver transplant: Poor performance for malnutrition diagnosis but a potentially useful prognostic tool

• Published in: Clinical nutrition (Edinburgh, Scotland) Scotland), 2022-01, Vol.41 (1), p.97-104
• Main Authors: Santos, Bárbara Chaves, Fonseca, Ana Luisa Ferreira, Ferreira, Lívia Garcia, Ribeiro, Helem Sena, Correia, Maria Isabel Toulson Davisson, Lima, Agnaldo Soares, Penna, Francisco Guilherme Cancela e, Anastácio, Lucilene Rezende
• Format: Article
• Language: English
• Subjects: Adult; Cirrhosis; Female; GLIM criteria; Hand Strength; Humans; Liver Cirrhosis - complications; Liver Cirrhosis - physiopathology; Liver Cirrhosis - surgery; Liver transplantation; Liver Transplantation - mortality; Male; Malnutrition; Malnutrition - diagnosis; Malnutrition - etiology; Middle Aged; Mortality; Nutrition Assessment; Nutritional Status; Prevalence; Prognosis; Reproducibility of Results; Retrospective Studies; Risk Assessment - methods; Waiting Lists - mortality
• ISSN: 0261-5614; 1532-1983
• DOI: 10.1016/j.clnu.2021.11.008

Studies using the Global Leadership Initiative on Malnutrition (GLIM) criteria for patients with liver cirrhosis are limited. This study aimed to assess the impact of malnutrition according to the GLIM criteria on the outcomes of patients awaiting a liver transplant (LTx) and compare these criteria with Subjective Global Assessment (SGA). This retrospective observational study included adult patients awaiting LTx. Patient clinical data, nutritional status according to various tools including SGA, and resting energy expenditure were assessed. The distinct phenotypic and etiologic criteria provided 36 different GLIM combinations. The GLIM criteria and SGA were compared using the kappa coefficient. The variables associated with mortality before and after the LTx and with a longer length of stay (LOS) after LTx (≥18 days) were assessed by Cox regression and logistic regression analyses, respectively. A total of 152 patients were included [median age 52.0 (interquartile range: 46.5–59.5) years; 66.4% men; 63.2% malnourished according to SGA]. The prevalence of malnutrition according to the GLIM criteria ranged from 0.7% to 30.9%. The majority of the GLIM combinations exhibited poor agreement with SGA. Independent predictors of mortality before and after LTx were presence of ascites or edema (p = 0.011; HR:2.58; CI95%:1.24–5.36), GLIM 32 (PA-phase angle + MELD) (p = 0.026; HR:2.08; CI95%:1.09–3.97), GLIM 33 (PA + MELD-Na≥12) (p = 0.018; HR:2.17; CI95%:1.14–4.13), and GLIM 34 (PA + Child–Pugh) (p = 0.043; HR:1.96; CI95%:1.02–3.77). Malnutrition according to GLIM 28 (handgrip strength + Child-Pugh) was independently associated with a longer LOS (p = 0.029; OR:7.21; CI95%:1.22–42.50). The majority of GLIM combinations had poor agreement with SGA, and 4 of the 36 GLIM combinations were independently associated with adverse outcomes.