Publication Date



Higher serum bilirubin has been associated with poorer prognosis in patients with heart failure (HF). We examined the association between serum bilirubin and clinical outcomes in patients with clinical signs of HF and/or left ventricular systolic dysfunction after acute myocardial infarction (MI). A total of 7,467 patients from the High-Risk Myocardial Infarction Database Initiative with an available baseline total bilirubin concentration were studied. The association between baseline bilirubin concentrations and the composite outcome of cardiovascular mortality (CVM), nonfatal stroke, nonfatal MI or hospitalization for HF, CVM, and all-cause mortality were assessed using Cox proportional hazards models. An interaction with time was observed with associations present only in the first 90 days after randomization. The median (percentile25-75) baseline total bilirubin concentration was 11 (8 to 14) µmol/L and was above the “normal” range (>17.1 µmol/L) in 1,053 (14.1%) patients. In multivariable analysis, with adjustment for baseline characteristics (demographic, co-morbidities, Killip score, left ventricular ejection fraction, and laboratory variables), patients with a bilirubin concentration of >17.1 µmol/L had a significantly higher risk of all the studied outcomes at 90 days (e.g., CVM: adjusted hazard ratio 1.45, 95% confidence interval 1.14 to 1.86, p = 0.003). The addition of bilirubin to a validated survival model modestly improved the risk reclassification to predict 90-day events (continuous net reclassification improvement for CVM 6.4%, 95% confidence interval 0.7% to 9.6%, p = 0.04). In patients with MI complicated with HF and/or systolic dysfunction, bilirubin concentration is an independent predictor of mortality and may improve risk stratification.


Mary MacKillop Institute for Health Research

Document Type

Journal Article

Access Rights

ERA Access

Access may be restricted.