Ó³»­´«Ã½

BACKGROUND: Malnutrition influences prognosis in patients with heart failure, but current nutritional evaluation methods are excessively complex for routine clinical use. The role of the serum albumin concentration, a widely used surrogate marker, in acute heart failure remains unclear. We assessed whether changes in serum albumin during heart failure hospitalisation can serve as a prognostic marker.METHODS: Among 14 847 patients in a nationwide heart failure registry in 2013, 5836 eligible patients (mean age: 77.7 years, 51.2% men) were categorised into two groups: the albumin-increase group (n=2048) and the no-albumin-increase group (n=3788). The incidences of primary (all-cause mortality) and secondary (a composite of cardiovascular death and rehospitalisation for worsening heart failure) outcomes were compared between the groups. The prognostic impact of serum albumin changes was also evaluated.RESULTS: The median follow-up period was 1545 days (95% CI 1535 to 1557 days). Event-free survival rates for clinical outcomes were higher in the albumin-increase group than in the no-albumin-increase group (primary outcome: 54.5% vs 45.4%, 95% CI 51.4% to 57.6% vs 42.9% to 47.8%, p<0.005; secondary outcome: 40.5% vs 36.9%, 95% CI 36.4% to 44% vs 34.1% to 39.6%, p<0.005). A generalised additive model revealed a linear relationship between serum albumin changes and prognosis; declines had a stronger negative effect than the positive impact of increases, but the change rate better reflected the relationship with prognosis. Survival classification and regression tree analysis indicated that a 25% decrease in the serum albumin concentration identified the most vulnerable population; even a 3% decrease was associated with worse outcomes.CONCLUSIONS: Serum albumin changes during hospitalisation can predict prognosis in patients with heart failure, indicating potential as a target for interventions.