BackgroundSymptomatic severe acute respiratory syndrome-coronavirus-2 (SARS-CoV-2) infection incidence is higher in the elderly patients. Pre-existing geriatric conditions such as comorbidity and frailty seem related to worse hospital outcomes.AimsTo assess the role of nutritional status as an independent prognostic factor for in-hospital death in elderly patients.MethodsConsecutive elderly patients (age > 65 years) hospitalized for novel coronavirus disease (COVID-19) were enrolled. Demographics, laboratory and comorbidity data were collected. Nutritional status was evaluated using the Geriatric Nutritional Risk Index (GNRI). Uni- and multivariate Cox regression analyses to evaluate predictors for in-hospital death were performed.ResultsOne hundred and nine hospitalized elderly patients (54 male) were consecutively enrolled. At univariate analysis, age (HR 1.045 [CI 1.008–1.082]), cognitive impairment (HR 1.949 [CI 1.045–3.364]), C-reactive protein (HR 1.004 [CI 1.011–1.078]), lactate dehydrogenases (HR 1.003 [CI 1.001–1.004]) and GNRI moderate–severe risk category (HR 8.571 [CI 1.096–67.031]) were risk factors for in-hospital death, while albumin (HR 0.809 [CI 0.822–0.964]), PaO2/FiO2 ratio (HR 0.996 [CI 0.993–0.999]) and body mass index (HR 0.875 [CI 0.782–0.979]) were protective factors. Kaplan–Meier survival curves showed a significative higher survival in patients without GNRI moderate or severe risk category (p = 0.0013).At multivariate analysis, PaO2/FiO2 ratio (HR 0.993 [CI 0.987–0.999], p = 0.046) and GNRI moderate–severe risk category (HR 9.285 [1.183–72.879], p = 0.034) were independently associated with in-hospital death.ConclusionNutritional status assessed by GNRI is a significative predictor of survival in elderly patients hospitalized for COVID-19. The association between GNRI and PaO2/FiO2 ratio is a good prognostic model these patients.