Comparison of objective nutritional indexes for the prediction of in-hospital mortality among elderly patients with acute heart failure.

Abstract

The association between objective nutritional indexes and prognosis in patients with acute heart failure have not been well studied. Therefore, we aimed to compare the prognostic value of modified Glasgow prognostic score, prognostic nutritional index, controlling nutritional status score, and geriatric nutritional risk index for the prediction of in-hospital mortality in patients with acute heart failure. All consecutive elderly patients (aged ≥65 years) who had tests for C-reactive protein, total lymphocyte count, total cholesterol, and albumin levels at admission, and hospitalized due to acute heart failure were retrospectively included. The primary endpoint of the study was in-hospital mortality. We used a base model for the prediction of in-hospital mortality, including age, gender, log N-terminal pro-B-type natriuretic peptide, and the presence of coronary artery disease. We added each of the malnutrition scores, in turn, to the base model and used C-statistics to evaluate model discrimination in survival analysis. A total of 628 patients were included, and 80 (12.7%) of the patients died during the hospital stay. Multivariate analysis showed that older age, prognostic nutritional index < 41.2, controlling nutritional status score > 5, geriatric nutritional risk index <92, and modified Glasgow prognostic score were independent predictors of in-hospital mortality. Among the malnutrition scores, geriatric nutritional risk index increased model performance most compared with base model. Though all objective nutritional indexes were associated with prognosis in elderly patients with acute heart failure, geriatric nutritional risk index was superior to other scores in predicting in-hospital mortality.