GLIM Criteria–defined malnutrition and short-term mortality in acute care hospitals: A nationwide claims-based historical cohort study

Abstract

Background & aims

Malnutrition is associated with adverse clinical outcomes in hospitalized patients. The Global Leadership Initiative on Malnutrition (GLIM) criteria have been proposed as an internationally standardized nutritional assessment; however, their adoption and prognostic value in real-world hospital settings remain underexplored. In June 2024, Japan's Diagnosis Procedure Combination (DPC) reimbursement system began encouraging routine GLIM-based assessments, creating a quasi-natural experiment to evaluate its implementation and prognostic utility at scale. This study evaluated the association between GLIM-defined malnutrition and mortality in routine clinical practice under Japan's DPC system.

Methods

This retrospective cohort study utilized Japan's nationwide DPC database provided by JMDC, Inc., which includes patient data from approximately 1,700 hospitals. Adult patients (age ≥ 18 y) hospitalized between June 2024 and August 2024 were included. The primary outcomes were the 30-d and 60-d mortality rates. Cox proportional hazards models were used to estimate the multivariate-adjusted hazard ratios (HR) and 95 % confidence intervals (CI) for mortality.

Results

Of the 174,439 eligible patients, 3.9 % were classified as malnourished, 47.9 % as nonmalnourished, and 48.2 % had GLIM criteria assessments that were not performed or had missing data. GLIM-defined malnutrition was significantly associated with a higher risk of 30-d mortality (HR = 1.46; 95 % CI: 1.33–1.60) and 60-d mortality (HR = 1.46; 95 % CI: 1.34–1.59). Additionally, patients with incomplete GLIM assessment data had a higher mortality risk than those without malnutrition. Among the patients with completed GLIM criteria assessments (n = 90,321), the most frequently missing GLIM sub-items were “reduced food intake/absorption” (2.7 %) and “low muscle mass” (2.6 %).

Conclusions

In this nationwide cohort of 174,439 hospitalized adult patients, both GLIM-defined malnutrition and incomplete or missing GLIM assessment data were associated with substantially higher 30-d and 60-d mortality rates, highlighting the potential utility of the criteria for mortality risk stratification in routine clinical practice and identifying underassessment as a pragmatic risk indicator.