Prevalence of Malnutrition and 1-Year All-Cause Mortality in Institutionalized Elderly Patients Comparing Different Combinations of the GLIM Criteria.

Abstract

The Global Leadership Initiative on Malnutrition (GLIM) provided a new approach for the diagnosis of malnutrition based on a number of phenotypic and etiologic criteria. However, different diagnostic criteria could potentially lead to heterogeneity in the diagnosis. We identified different subsets of criteria to define malnutrition in a cohort of elder nursing-home residents and investigated their clinical utility in terms of 1-year survival. Our study included all residents (n = 485) from 3 nursing homes. We proposed 12 GLIM models based on different phenotypic and etiologic criteria. The main outcome was 1-year all-cause mortality. We also assessed the sensitivity and specificity for different phenotypic criteria using time-dependent receiver operating characteristic curves, and cutoff values were calculated. During 1 year of follow-ups, 94 deaths occurred. The prevalence of malnutrition was 13.5% for models based on reduced food intake (RFI) and 10.45% for models based on inflammation associated with acute disease (IAD). Unadjusted Cox regression analyses showed that malnutrition was associated with a 2.31- to 4.64-fold increase in mortality when diagnosed using IAD-dependent models or with a 1.37- to 1.78-fold increase in mortality using RFI-dependent models. Cutoffs associated with mortality for the phenotypic criteria were lower than those recommended in the GLIM criteria. This study describes the association of several variations of the GLIM model with 1-year mortality in nursing-home residents. However, our data suggest a high heterogeneity to fulfill the GLIM criteria and the necessity of finding specific, tailored cutoff points for the studied populations.