Evaluation of the Geriatric Nutritional Risk Index in predicting mortality in older patients with COVID-19 in the AgeBru cohort

OBJECTIVE: To determine whether the Geriatric Nutritional Risk Index (GNRI) on hospital admission was associated to an increased 14-day and 12-month mortality-risk in older inpatients with COVID-19. METHODS: Cohort study of consecutive inpatients admitted with COVID-19 in a university hospital (20/03/2020-11/05/2021). Inclusion criteria: age over 65 years and positive polymerase chain reaction test. Exclusion criteria: missing data for weight, height, and/or albumin, hospital-acquired COVID-19, or patients transferred to other health facilities. Outcome: all-cause mortality at 14-day and 12-month follow-up. GNRI [1.489 x albumin (g/L)] + [41.7 (weight/ideal body weight)] was assessed at admission; scores ≤98 indicated risk of malnutrition. Cox-proportional hazards models assessed the association between the admission GNRI and 14-day and 12-month mortality-risk, after adjusting by demographic and clinical variables, including inflammation (C-reactive protein). RESULTS: Of the 570 eligible patients, 224 (mean age 78 years; 52.2% women) met inclusion criteria and 151 (67.4%) were classified at risk of malnutrition. Twenty patients died during the 14-day and 42 during the 12-month follow-up. The risk of 14-day mortality was nearly 10 times higher in patients with GNRI scores ≤98 (HR=9.6 [95%CI 1.3-71.6], P=0.028); this association was marginally significant in the adjusted model (HR=6.73 [95%CI 0.89-51.11], P=0.065)]. No association between GNRI and the 12-month mortality-risk was found. CONCLUSIONS: The GNRI may play a role in the short-term prognosis of older inpatients with COVID-19. Further studies are required to confirm the short-term predictive validity of the GNRI within this population (Clinicaltrials.gov_NCT05276752).