Continuously elevated plasma ketone bodies are linked to adverse outcomes in patients with heart failure

Background Elevated plasma ketone bodies (KBs) are linked to poor outcomes in heart failure (HF), yet the prognostic value of repeated KB measurements is uncertain. Objective To study temporal changes in plasma KBs in HF patients in relation to prognosis. Methods Plasma KBs were measured by nuclear magnetic resonance spectroscopy in 467 patients from the TRIUMPH study. Blood was obtained up to seven times during 1-year follow-up after HF hospitalization. Associations between serial KB measurements and the endpoint of death and HF rehospitalization were investigated by joint modeling. Similar analyses were conducted in the Bio-SHiFT study, including 397 chronic HF patients with up to 10 repeated blood samples during a median follow-up of 2.1 years. The endpoint in Bio-SHiFT was a composite of cardiovascular death, HF hospitalization, left ventricular device implantation and heart transplantation. Results In TRIUMPH, median age was 74 years, 37% of patients were female and 82% were in NYHA class III/IV. Estimated KBs were continuously higher in the 185 (40%) patients reaching the combined endpoint. Hazard ratios (HR) (95% confidence interval [CI]) for the endpoint per doubling of total KBs (TKBs), acetone, and β-hydroxybutyrate were 2.10 (1.27–3.71), 1.62 (1.08–2.44) and 1.62 (1.22–2.23), respectively, at any time during follow-up after adjustment for clinical characteristics and NT-proBNP at baseline. In Bio-SHiFT, median age was 64 years, 28.5% of patients were female, and 124 (31.2%) patients reached the primary endpoint. KBs were higher in patients reaching the endpoint. The HR per doubling of TKB was 5.36 (95% CI 2.18–14.0) and 4.24 (31.4–5.95) for acetone. Conclusion Serially measured KB levels were associated with adverse events in two independent HF cohorts. Patients with adverse events displayed consistently higher KBs, potentially reflecting ongoing metabolic stress.