The nutritional marker pre-albumin is strongly associated with adverse outcome in heart failure

Abstract

Background: Pre-albumin is the best indicator of nutritional status. We have previously shown that pre-albumin is an independent predictor of in-hospital mortality in heart failure (HF). Still, its medium-term prognostic value in HF is yet unknown. We aimed to evaluate if a low discharge pre-albumin would predict morbidity and mortality in HF. Methods: We conducted a prospective observational study. All patients discharged after admission with a primary diagnosis of HF were eligible for study entry. We excluded those with: acute coronary syndromes, no echocardiographic structural or functional abnormalities, and no discharge pre-albumin available. Fasting venous blood samples were collected to all patients on the discharge day. Patients were followed up for 6 months. Endpoints were: all-cause and HF-death; all-cause readmission and readmission due to HF. Patients with pre-albumin ≤ 15.0 mg/dl (1st quartile) and those with pre-albumin >15mg/dL were compared. A time-dependent Cox-regression analysis was used to evaluate the prognostic impact of low pre-albumin. Results: A total of 514 patients were analysed. Mean age was 78 years and 45.7% were male. HF was ischaemic in 41.4% and 43.8% of the patients had HF with preserved ejection fraction. During the 6-month follow up 101 patients died, 78 due to HF; 209 patients were hospital readmitted and 140 due to worsening HF. Median (interquartile range) pre-albumin was 20.1 (15.3-25.3)mg/dL. Patients with lower pre-albumin were more often women, older aged and with non-ischaemic HF; they also had lower albumin, haemoglobin and total cholesterol and higher C-reactive protein (CRP) and B-type natriurectic peptide (BNP). Lower pre-albumin was also associated with lower creatinine and less beta-blocker (BB) use. Patients with discharge pre-albumin ≤15mg/dl had a HR of 6 months all-cause death of 2.49 (95% CI: 1.67-3.70) and of HF death of 2.66 (95% CI: 1.70-4.17); the HR for all cause hospitalization was 1.87 (95% CI: 1.39-2.50) and for HF readmission of 1.92 (95% CI 1.35-2.74). Association with morbidity and mortality was independent of age, sex, albumin, total cholesterol, anaemia, renal dysfunction, CRP, ischaemic aetiology, BNP, BB use and left ventricular systolic dysfunction. Multivariate adjusted HR for all-cause death and HF death were 1.78 (1.02-3.09) and 2.21 (1.18-4.13); and for all-cause readmission and HF readmission were 1.68 (1.15-2.47) and 1.94 (1.23-3.08). Conclusions: Patients with discharge pre-albumin ≤15mg/dl have a significant higher risk of 6 months morbidity and mortality. Malnutrition predicts ominous outcome after HF hospitalization.