Comparison of the adequacy of geriatric nutritional risk index with that of the mini nutritional assessment-short form and global leadership initiative on malnutrition criteria in assessing nutritional status to predict the 1-year prognosis of hospitalized Japanese older adults: a single-institutional cohort study.

BACKGROUND: The global leadership initiative on malnutrition (GLIM) proposed the first international standards (GLIM criteria) for malnutrition diagnosis. Early screening using nutritional tools is recommended to improve the prognosis of older patients. The association between Mini Nutritional Assessment-Short Form (MNAⓇ-SF) and Geriatric Nutritional Risk Index (GNRI) and prognosis has been reported, but there is insufficient evidence to develop the GLIM criteria for older inpatients. We aimed to evaluate the MNAⓇ-SF, GNRI, and GLIM criteria to determine their contribution to the prognosis prediction of hospitalized older patients at 1 year after discharge. METHODS: This study included 386 patients hospitalized between September 2014 and October 2015, and May and December 2019. After excluding 17 patients who died at the time of initial hospitalization, 23 who were lost to follow-up after 1 year, and 28 who had missing data on admission, only 318 were included in the final analysis. The primary outcome was death within 1 year after discharge, assessed using the MNA®-SF, GNRI, and GLIM criteria, and survival analysis was conducted. Multivariate Cox proportional hazards analysis was performed to identify the nutritional assessment tools that contributed to the prognosis prediction. RESULTS: A total of 43 patients died within 1 year. Of them, 58.1% had malnutrition and 37.2% were at risk of malnutrition, assessed using the MNAⓇ-SF; 27.9% had severely malnourished assessed using the GNRI; and 58.1% had severely malnourished assessed using the GLIM criteria. The proportions of malnourished and severely malnourished patients were significantly higher in the mortality group than in the survival group. Multivariate Cox proportional hazards analysis showed hazard ratios of 1.06 (95% confidence interval [CI]: 0.24-4.71) for at risk and 2.17 (95% CI: 0.48-9.84) for malnutrition (MNAⓇ-SF); 5.68 (95% CI: 2.74-11.80) for moderately malnourished and 7.69 (95% CI: 3.13-18.91) for severely malnourished (GNRI); and 1.47 (95% CI: 0.48-4.50) for moderately malnourished and 2.45 (95% CI: 1.22-4.93) for severely malnourished (GLIM criteria); GNRI had the most significant contribution to prognosis prediction. CONCLUSIONS: GNRI significantly contributed to the prognosis prediction 1 year after hospital discharge of older patients.