Objective Acute heart failure (AHF) is associated with high mortality and morbidity. Trimethylamine N-oxide (TMAO), a gut-derived metabolite, has reported association with mortality risk in chronic HF but this association in AHF is still unknown. The present study investigated TMAO in patients admitted to hospital with AHF, and association of circulating levels with prognosis.

Methods In total, 972 plasma samples were analysed for TMAO concentration by liquid chromatography-mass spectrometry. Associations with in-hospital mortality (72 events), all-cause mortality (death, 268 events) and a composite of death or rehospitalisation due to HF (death/HF, 384 events) at 1 year were examined.

Results TMAO improved risk stratification for in-hospital mortality in combination with current clinical scorings (OR≥1.13, p≤0.014). TMAO tertile analyses reported a graded risk in adverse outcome within 1 year (OR≥1.61, p≤0.004) and improved outcome prediction when stratified as none, one or both biomarker(s) elevated in combination with N-terminal pro B-type natriuretic peptide (NT-proBNP) (OR≥2.15, p≤0.007). TMAO was independently predictive for death and death/HF when corrected for cardiac risk factors (HR≥1.16, p≤0.037); however, this ability was weakened when indices of renal function were included, possibly due to multicollinearity.

Conclusions TMAO contributed additional information on patient stratification for in-hospital mortality of AHF admissions using available clinical scores that include renal indices. Furthermore, elevated levels were associated with poor prognosis at 1 year and combination of TMAO and NT-proBNP provided additional prognostic information. TMAO was a univariate predictor of death and death/HF, and remained an independent predictor until adjusted for renal confounders.