Differential implications of gut-related metabolites on outcomes between heart failure and myocardial infarction

Gut metabolites, through their role in atherosclerotic plaque formation,1 myocardial fibrosis,2 and myocardial function suppression,3 have been implicated in the pathophysiology of various cardiovascular (CV) diseases.4,5 Clinical studies have shown the association of gut metabolites with adverse outcomes, severity, and risk stratification in several CV diseases [e.g. heart failure (HF)6 and myocardial infarction (MI)7]. Despite most of the past studies having focused mainly on a single metabolite (i.e. trimethylamine N-oxide—TMAO), it has been suggested that additional metabolites of this pathway, involving choline and carnitine (e.g. acetyl-L-carnitine, L-carnitine, betaine, and γ-butyrobetaine), may have an additional role to play.8,9 However, to date, the proper contribution of the different metabolites in the spectrum of the CV diseases still remains unclear; therefore, the aim of the present study is to investigate whether there is a differential contribution of metabolite biomarkers of the choline/carnitine–metabolic pathway in association with the adverse outcomes of MI and HF. The six investigated gut-related metabolites (acetyl-L-carnitine, betaine, choline, γ-butyrobetaine, L-carnitine, and TMAO) were measured in a historical, well-characterized cohort of post-MI patients collected at the University Hospitals of Leicester, UK, between August 2004 and April 2007.10 As a primary endpoint, the composite of mortality and reinfarction due to MI (death/MI) or mortality and rehospitalization due to HF (death/HF), as well as all-cause mortality at 2 years, was used. The association of outcomes at 2 years was investigated using log-transformed metabolite concentrations. In addition, the role of the combination of different metabolites has been investigated. The baseline demographics of 525 patients are given in Table 1. During the 2-year follow-up, the overall mortality rate was 12% (n = 63); the death/MI rate was 25% (n = 129), death/HF 21% (n = 109), MI 11% (n = 59), and HF 9% (n = 45). Patient demographics and characteristics Data are presented as median (interquartile range) for continuous variables and % for categorical values. For outcomes of death, death/MI, and death/HF, patients who died had higher levels of acetyl-L-carnitine, L-carnitine, and TMAO (P ≤ 0.024), whereas betaine levels were elevated in death/MI and death/HF (P ≤ 0.042) but showed no difference for outcomes of death. Both choline and γ-butyrobetaine were elevated in patients with an event of death and death/HF, but no difference was observed for death/MI (P ≤ 0.037). TMAO levels were also elevated in patients with HF (HF rehospitalization) (P = 0.040; Table 2). Backward Cox regression for outcomes at 2 years for the gut metabolites showed that models for best predicting death included acetyl-L-carnitine and TMAO, for death/MI, they included acetyl-L-carnitine, betaine, and TMAO, while acetyl-L-carnitine, γ-butyrobetaine, and TMAO best represented death/HF. Outcomes of MI were best represented by betaine alone, whereas HF outcomes were represented by γ-butyrobetaine and TMAO (Table 3). An analysis of univariate associations with outcomes showed an association of acetyl-L-carnitine, choline, γ-butyrobetaine, L-carnitine, and TMAO with death (P ≤ 0.026), acetyl-L-carnitine, betaine, L-carnitine, and TMAO with death/MI (P ≤ 0.025), and all six metabolites with death/HF (P ≤ 0.042). After adjustment for risk factors and variables traditionally used for CV diseases [i.e. age, sex, past history of MI/angina, systolic blood pressure (BP), heart rate, revascularization, beta-blocker at discharge, angiotensin-converting enzyme inhibitors at discharge, and N-terminal pro-hormone B-type natriuretic peptide (NT-proBNP)], a multivariate analysis showed an association of acetyl-L-carnitine, L-carnitine, and TMAO for death [hazard ratio (HR) 1.46–1.72 (95% confidence interval, CI 1.05–2.33) P ≤ 0.025], acetyl-L-carnitine, betaine, and TMAO for death/MI [HR 1.28–1.33 (95% CI 1.02–1.65) P ≤ 0.032], and γ-butyrobetaine, L-carnitine, and TMAO for death/HF at 2 years [HR 1.27–1.34 (95% CI 1.02–1.75) P ≤ 0.034]. Betaine was also associated with MI outcomes [HR 1.30 (95% CI 1.00–1.68) P = 0.048] (Table 4). With further adjustment for renal function estimated glomerular filtration rate (eGFR), past history of smoking and diabetes, statins at discharge, and cholesterol levels, betaine was associated with death/MI [HR 1.28–1.35 (95% CI 1.04–1.69) P ≤ 0.019], whereas L-carnitine and TMAO were associated with death and death/MI [HR 1.25–1.91 (95% CI 1.01–2.65) P ≤ 0.040] (see Supplementary material online, Table S1). Gut metabolite concentrations according to no event vs. all-cause mortality (death), composite outcomes of all-cause mortality or reinfarction (death/myocardial infarction), mortality, and rehospitalization due to heart failure (death/heart failure), myocardial infarction, and heart failure at 2 years Data are presented as median (interquartile range). Backward Coxregression for the gut-related markers for outcomes of death, death/myocardial infarction, death/heart failure, myocardial infarction, and heart failure at 2 years Independent prediction abilities of gut-related metabolites for death, death/myocardial infarction, death/heart failure, myocardial infarction, and heart failure at 2 years Data are displayed as hazard ratio (95% confidence intervals) P-value. Metabolite levels were log-transformed and normalized to 1 SD so that HRs refer to the z-transformed variables. Adjusted—model adjusted for—age, sex, past medical history of MI/angina, systolic BP, heart rate, revascularization, STEMI, beta-blocker at discharge, angiotensin-converting enzyme inhibitor at discharge, logNT-proBNP. The present study, based on a cohort of acute MI patients, investigates for the first time the separate and specific association of gut dysfunction biomarkers (i.e. the choline/carnitine–TMAO pathway metabolites) with the possible clinical evolution after an acute MI (i.e. MI and HF outcomes). The main finding of this study is the novel concept that, from the same pathway, different metabolites are specifically associated with different outcomes in different CV diseases; specifically, the entire pathway (i.e. choline/carnitine–TMAO) is associated with death/MI (i.e. MI), whereas only the carnitine–TMAO pathway is involved in death/HF (i.e. HF). Interestingly, betaine is a standalone marker with its association with MI outcomes (i.e. reinfarction due to MI). Specifically, findings from this study further validate the association with long-term MI mortality for TMAO.11–13 A recent study investigating TMAO in MI and HF patients showed that elevated TMAO levels were associated with poor outcomes, especially in patients with elevated C-reactive protein, suggesting the role of TMAO in inflammation.14 With regard to betaine, mixed results are available in the literature.8 Reduced betaine levels have shown associations with secondary MI risk;8 on the other hand, elevated betaine levels are associated with a poor prognosis of major adverse cardiovascular event with related increased TMAO.15 Similarly, both reduced16 and increased17 carnitine levels have shown associations with acute MI. Previous studies have demonstrated a role for the carnitine pathway metabolites with associations and risk stratification in HF patients, with both betaine and choline not associated with HF outcomes, which is similar to our findings for death/HF outcomes.6,10 Our findings further support the hypothesis that choline, being involved in lipid metabolism, might contribute more to MI than to carnitine, which is associated with cardiac metabolism and therefore more associated with HF.18 From a clinical point of view, our findings shed light on the possibility for clinicians to predict the possible evolution of a patient after an acute coronary syndrome. To date, it is not possible to predict, after the index event, which patients are at risk of reinfarction and those at risk of reoccurrence of acute HF. Our findings would help in risk stratification of these patients and allow for a more tailored therapy (e.g. a more aggressive anti-ischaemic treatment vs. referral to an HF specialist service after discharge). This study has some limitations. First, it enrolled patients from a single centre. Second, the cohort is a historical cohort; however, it has been well validated. Furthermore, treatment strategies have evolved since study recruitment. Equally, the natural course of disease after an MI is currently difficult to quantify with immediate revascularization and therefore historical cohorts have value to identify the contribution of pathophysiological factors such as the gut microbiome to outcome. Data that can influence metabolite levels, such as diet and physical activity, were not available to adjust for these confounding factors; however, this is in line with previous findings. In conclusion, for the first time, it has been shown, in the same population, that different metabolites are associated with different outcomes in different CV diseases; specifically, metabolites of the choline, carnitine, and TMAO pathway are associated with outcomes of death/MI, whereas only carnitine–TMAO metabolites, but not choline, are associated with death/HF. Betaine is associated with outcomes of reinfarction caused by MI. All authors listed in this manuscript have substantially contributed to the study’s conception, design, analysis, draft, review, and performance in accordance with the journal guidelines. M.Z.I. and A.S. contributed to the conception, design, analysis, draft, and review. S.S. contributed to design, analysis, and review. L.L.N. contributed to the conception, design, and review. T.S. contributed to the conception, design, analysis, drafting, review, and provision of funds. All gave final approval and agreed to be accountable for all aspects of work, ensuring integrity and accuracy. Supplementary material is available at European Journal of Preventive Cardiology. Japan Heart Foundation, National Institute for Health Research (Leicester Biomedical Research Centre), the British Heart Foundation (BHF), the Medical Research Council (MRC) UK Consortium on MetAbolic Phenotyping (MAP/UK), and the LeDucq Foundation. The data underlying this article will be shared on reasonable request to the corresponding author.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.