Author + information
- Katharina Schuett, MD,
- Marcus E. Kleber, PhD,
- Hubert Scharnagl, MD,
- Stefan Lorkowski, PhD,
- Winfried März, MD,
- Alexander Niessner, MD,
- Nikolaus Marx, MD∗ ( and )
- Andreas Meinitzer, PhD
- ↵∗Department of Internal Medicine I, University Hospital Aachen, RWTH Aachen University, Pauwelsstrasse 30, 52074 Aachen, Germany
Trimethylamine-N-oxide (TMAO), a microbiota-associated metabolite, is associated with cardiovascular (CV) events in various populations (1), including patients with chronic heart failure (HF) (2,3). It has remained unclear whether the predictive value of TMAO is different in HF patients with reduced ejection fraction (HFrEF) and those with HF with preserved ejection fraction (HFpEF).
We measured TMAO in 2,490 patients referred for coronary angiography participating in the LURIC (Ludwigshafen Risk and Cardiovascular Health) study (4). The study has been approved by the ethics committee at the Ärztekammer Rheinland-Pfalz and is conducted in accordance with the Declaration of Helsinki. HFrEF and HFpEF were diagnosed according to European guidelines (5). Among the 2,490 patients, 728 deaths occurred during a mean follow-up time of 9.7 years including 446 deaths due to CV causes. Compared with subjects in the lowest tertile, the age- and sex-adjusted hazard ratios (HRs) in the highest tertile were 1.70 (95% confidence interval [CI]: 1.41 to 2.04) and 1.87 (95% CI: 1.48 to 2.38) for death and CV death, respectively. Additional adjustments for body mass index, New York Heart Association functional class, N-terminal pro-brain natriuretic peptide (NT-proBNP), estimated glomerular filtration rate (eGFR), smoking, hypertension, coronary artery disease, diabetes mellitus, and atrial fibrillation (model 2) and further adjustment for high-sensitivity C-reactive protein (model 3) did not affect the significant association of TMAO with mortality.
Patients with HFrEF (n = 428) and HFpEF (n = 395) exhibited significantly higher TMAO than patients without HF (quartile 1 to quartile 3: median 4.73 μmol/l, range 3.40 to 6.82 μmol/l; and median 4.74 μmol/l, range 3.22 to 6.85 μmol/l, respectively, vs. median 3.98 μmol/l, range 2.83 to 5.74 μmol/l; p < 0.001). Among the 823 HF patients, 393 (47.8%) died during follow-up with 273 deaths from CV causes (33.2%). Age- and sex-adjusted HRs for death and CV death were 1.84 (95% CI: 1.44 to 2.36) and 1.98 (95% CI: 1.47 to 2.68), respectively, in subjects in the highest versus the lowest TMAO tertile. TMAO remained predictive in models 2 and 3. TMAO was not significantly associated with overall or CV mortality in HFpEF patients. In contrast, in HFrEF patients age- and sex-adjusted HRs for death and CV death were 2.33 (95% CI: 1.67 to 3.24) and 2.27 (96% CI: 1.52 to 3.37) in the highest versus the lowest TMAO tertile (Figure 1A). Adjustment in models 2 and 3 did not affect these associations. In the HFrEF group, NT-proBNP, an established risk marker in HF not known to be directly influenced by TMAO, was significantly associated with mortality and CV mortality in the age- and sex- adjusted model for the highest versus the lowest tertile of NT-proBNP. This association lost significance after additional adjustment (model 3) suggesting that TMAO might be a better predictor of mortality than NT-proBNP in HFrEF patients.
To analyze an additive prognostic value of TMAO in relation to NT-pro-BNP in HFrEF patients, we stratified patients into 9 groups according to tertiles of TMAO and NT-proBNP. Compared with the reference group (patients in the lowest tertile of NT-proBNP and TMAO), patients in the highest tertiles of NT-proBNP and TMAO had HRs of 5.20 (p > 0.001) for overall and of 10.25 (p > 0.001) for CV mortality (Figure 1B). Category-free net reclassification improvement analyses confirmed a significant improvement in individual risk prediction of TMAO in addition to NT-proBNP with 37.7% (p < 0.001) for all-cause mortality. We analyzed the influence of diabetes mellitus and renal function on the predictive value of TMAO. TMAO was more predictive for total mortality in patients with than without diabetes mellitus, but equally predictive for CV mortality in both groups. TMAO was a stronger predictor of both total and CV mortality in the lowest eGFR tertile (<72.2 ml/min/1.73 m2) than in the intermediate or the highest eGFR tertile (>72.3 ml/min/1.73 m2). These findings may reflect the high coincidence of HFrEF, diabetes mellitus, and chronic kidney disease (41.9% of HFrEF had diabetes mellitus; 49.2% were in the lowest eGFR tertile).
In summary, elevated levels of TMAO are predictive for mortality and CV mortality in HFrEF but not HFpEF patients, and TMAO has a predictive value in HFrEF patients above and beyond NT-proBNP.
The authors thank the participants of the LURIC study; without their collaboration this article would not have been written. The authors also thank the LURIC study team members who were either temporarily or permanently involved in patient recruitment and sample and data handling; the laboratory staff at the Ludwigshafen General Hospital; and the Universities of Freiburg, Ulm, and Graz.
Please note: The LURIC study received funding from the 7th Framework Program of the European Union, integrated projects Atheroremo (grant number 201668), and RiskyCAD (grant number 305739), and through e:AtheroSysMed (Systems Medicine of Coronary Heart Disease and Stroke, German Ministry of Education and Research [grant number 01ZX1313A-K]). Dr. Schuett received funding from the Deutsche Forschungsgemeinschaft (grant number HE 5666/1-2). Drs. Schuett and Marx have received funding from the “Corona-Stiftung,” Germany. Dr. Lorkowski is a member of the BfR committee for nutrition, dietetic products, novel food and allergies (Federal Institute for Risk Assessment; Bundesinstitut für Risikobewertung, BfR); and has received personal fees from AMGEN, Berlin-Chemie, Merck, Sharp & Dohme, OmniaMed, Synlab, and Sanofi, Dr. März has received funding from Synlab Holding Deutschland, during the conduct of the study; and has received grants and personal fees outside the submitted work from Siemens Diagnostics, Aegerion Pharmaceuticals, AMGEN, AstraZeneca, Danone Research, Hoffmann LaRoche, Merck, Sharp & Dohme, Pfizer, Sanofi, Synageva, BASF, Abbott Diagnostics, Numares, and Berlin-Chemie. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- 2017 American College of Cardiology Foundation
- Lever M.,
- George P.M.,
- Slow S.,
- et al.
- McMurray J.J.,
- Adamopoulos S.,
- Anker S.D.,
- et al.