Longitudinal Changes in Circulating Ketone Body Levels in Patients With Acute Heart Failure: A Post Hoc Analysis of the EMPA-Response-AHF Trial

Research output: Contribution to journalArticleAcademicpeer-review

4 Citations (Scopus)
21 Downloads (Pure)


Background: Ketone bodies are endogenous fuels produced by the liver under conditions of metabolic or neurohormonal stress. Circulating ketone bodies are increased in patients with chronic heart failure (HF), yet little is known about the effect of acute HF on ketosis. We tested the hypothesis that ketogenesis is increased in patients with acute decompensated HF. Methods and results: This was a post hoc analysis of 79 patients with acute HF included in the EMPA-RESPONSE-AHF trial, which compared sodium-dependent glucose-cotransporter protein 2 inhibitor treatment with empagliflozin for 30 days with placebo in patients with acute HF [NCT03200860]. Plasma concentrations of ketone bodies acetone, β-hydroxybutyrate, and acetoacetate were measured at baseline and 5 different timepoints. Changes in ketone bodies over time were monitored using repeated measures analysis of variance. In the total cohort, median total ketone body concentration was 251 µmol/L (interquartile range, 178–377 µmol/L) at baseline, which gradually decreased to 202 µmol/L (interquartile range, 156–240 µmol/L) at day 30 (P = .041). Acetone decreased from 60 µmol/L (interquartile range, 34–94 µmol/L) at baseline to 30 µmol/L (interquartile range, 21–42 µmol/L) ( P < .001), whereas β-hydroxybutyrate and acetoacetate remained stable over time. Higher acetone concentrations were correlated with higher N-terminal pro brain natriuretic peptide levels (r = 0.234; P = .039). Circulating ketone bodies did not differ between patients treated with empagliflozin or placebo throughout the study period. A higher acetone concentration at baseline was univariately associated with a greater risk of the composite end point, including in-hospital worsening HF, HF rehospitalizations, and all-cause mortality after 30 days. However, after adjustment for age and sex, acetone did not remain an independent predictor for the combined end point. Conclusions: Circulating ketone body concentrations, and acetone in particular, were significantly higher during an episode of acute decompensated HF compared with after stabilization. Treatment with empagliflozin did not affect ketone body concentrations in patients with acute HF.

Original languageEnglish
Pages (from-to)33-41
Number of pages9
Issue number1
Publication statusPublished - Jan-2023


  • acetone
  • Acute heart failure
  • empagliflozin
  • ketone bodies
  • NT-pro BNP
  • SGLT2 inhibitors

Cite this