Prevalence and prognostic importance of malnutrition, as assessed by four different scoring systems, in elder patients with heart failure

Abstract

Background and aimsThe lack of standard diagnostic criteria in elder patients with heart failure (HF) makes it challenging to diagnose and manage malnutrition. We aimed to explore the prevalence of malnutrition, its associations and prognostic significance among elder patients with HF using four different nutritional scoring systems. Methods and resultsConsecutively presenting patients aged ⩾65 years, diagnosed with HF, and admitted to HF care unit of Fuwai Hospital CAMS&PUMC (Beijing, China) were assessed for nutritional indices. In total, 1371 patients were enrolled (59.4% men; mean age 72 years; median NT-proBNP 2343 ng/L). Using scores for the prognostic nutritional index (PNI) ≤38, controlling nutritional status (CONUT) score >4, geriatric nutritional risk index (GNRI) ≤91, and triglycerides, total cholesterol, and body weight index (TCBI) ≤1109, 10.4%, 18.3%, 9.2%, and 50.0% of patients had moderate or severe malnutrition, respectively. There was a strong association between worse scores and lower body mass index, more severe symptoms, atrial fibrillation, and anemia. The mortality over a median follow-up of 962 days (interquartile range (IQR): 903–1029 days) was 28.3% (n = 388). For those with moderate or severe condition, 1-year mortality was 35.2% for PNI, 28.3% for CONUT, 28.0% for GNRI, and 19.1% for TCBI. Malnutrition, defined by any of the included indices, showed added prognostic value when incorporated into a model and included preexisting prognostic factors (C-statistic: 0.711). However, defining malnutrition by the CONUT score yielded the most significant improvement in the prognostic predictive value (C-statistic: 0.721; p < 0.001). ConclusionMalnutrition is prevalent among elder patients with HF and confers increased mortality risk. Among the nutritional scores studied, the CONUT score was most effective in predicting the mortality risk. Clinical trial registrationURL: ClinicalTrials.gov; Unique Identifier: NCT02664818.