Prevalence and prognostic implications of malnutrition as defined by GLIM criteria in elderly patients with heart failure

Abstract

Although the Global Leadership Initiative on Malnutrition (GLIM) proposed a consensus scheme for diagnosing malnutrition in adults in clinical settings globally, the prevalence and prognostic value of malnutrition defined by GLIM criteria have yet to be evaluated in elderly patients with heart failure. This study aimed to determine the prevalence and prognostic implication of malnutrition as defined by GLIM criteria in comparison to those for a pre-existing definition of malnutrition, the geriatric nutritional risk index (GNRI), in elderly patients with heart failure METHODS: We evaluated malnutrition by two metrics, the GLIM criteria and geriatric nutritional risk index (GNRI), in 890 hospitalized patients with decompensation of heart failure aged ≥65 years, able to ambulate at discharge. The primary outcome was all-cause death within 1 year of discharge. According to GLIM criteria and GNRI <92, 42.4% and 46.5% of participants, respectively, had malnutrition, with moderate agreement (Cohen's kappa coefficient: 0.46 [95% confidence interval: 0.40-0.51]). During 1 year of follow-up, 101 (11.4%) deaths were observed, and malnutrition defined by either the GLIM criteria or GNRI was associated with a higher mortality rate, independent of other prognostic factors (GNRI: hazard ratio, 1.45, P=0.031; GLIM: hazard ratio, 1.57, P=0.016). Although malnutrition defined by either criterion showed additive prognostic value when added to a model incorporating pre-existing prognostic factors, defining malnutrition by GLIM criteria instead of the GNRI yielded a statistically significant improvement in model prognostic predictive ability (net-reclassification improvement, 0.44, P<0.001; integrated discrimination index, 0.013, P<0.001). In elderly patients with heart failure, 42.4% are malnourished according to the GLIM criteria, which is associated with a poor prognosis, independent of known prognostic factors. University Hospital Medical Information Network (UMIN-CTR, https://www.umin.ac.jp/ctr/index.htm, study unique identifier: UMIN000023929).