BackgroundMalnutrition is a common condition that is associated with adverse prognosis in patients with heart failure (HF). The Prognostic Nutritional Index (PNI), Geriatric Nutritional Risk Index (GNRI) and controlling nutritional status (CONUT) have all been used as objective indices for evaluating nutritional status. We aimed to clarify the relationship between these nutritional indices and the parameters of inflammatory markers, cardiac function and exercise capacity, as well as to compare the ability of these indexes for predicting mortality.MethodsWe evaluated PNI, GNRI and CONUT in consecutive 1307 patients with HF.ResultsFirst, there were significant correlations between nutritional indices and the following: C reactive protein; tumour necrosis factor-α; adiponectin; B-type natriuretic peptide; troponin I; inferior vena cava diameter and peak VO2 (P<0.05, respectively). Second, in the Kaplan-Meier analysis (follow-up 1146 days), all-cause mortality progressively increased from normal to mild, moderate and severe disturbance groups in the indices (log-rank, P<0.01, respectively). In the Cox proportional hazard analysis, each index was an independent predictor of all-cause mortality in patients with HF (P<0.001, respectively). Third, receiver operating curve demonstrated that the areas under the curve of PNI and GNRI were larger than that of CONUT score (P<0.05, respectively).ConclusionPatients with HF being malnourished had higher mortality accompanied by higher levels of C reactive protein, tumour necrosis factor-α, adiponectin, B-type natriuretic peptide, troponin I, right-sided volume overload and impaired exercise capacity, rather than left ventricular systolic function. Additionally, PNI and GNRI were superior to CONUT score in predicting mortality in patients with HF.