Elevated plasma TMAO levels associated with all-cause mortality in CAD patients in South China

Abstract

Background

Coronary artery disease (CAD) is a prevalent cardiovascular condition worldwide. Trimethylamine N-oxide (TMAO), a metabolite produced by gut microbiota, plays a crucial role in the pathogenesis and progression of CAD. However, the association between plasma TMAO and all‑cause mortality in Chinese CAD patients remains to be fully explored.

Methods

In this observational cohort study, 389 hospitalized CAD patients, confirmed via coronary angiography at Xiangya Hospital in 2022, were enrolled. Plasma TMAO levels were measured using liquid chromatography-tandem mass spectrometry. All-cause mortality events were identified through telephone interviews, hospital outpatient visits, and official hospital records, conducted semi-annually. Kaplan-Meier analysis and Cox regression analysis were employed to investigate the relationship between TMAO levels and all-cause mortality.

Results

Among 364 CAD patients who completed the median follow-up period of 39 months (IQR: 37–42 months), 40 patients (11.0%) experienced all-cause mortality. Patients with elevated TMAO levels, based on the optimal cutoff value of 317.62 ng/mL, had a significantly higher mortality rate compared to those with lower levels (P < 0.0001). After adjusting for conventional risk factors, including diabetes, elevated TMAO levels showed a significant association of all-cause mortality (hazard ratio [HR] 2.657; 95% CI: 1.389 to 5.084; P = 0.003).

Conclusions

Elevated plasma TMAO levels are significantly associated with increased all-cause mortality over a median follow-up of 39 months in CAD patients from South China.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12872-026-06214-9.

Introduction

Coronary artery disease (CAD) remains a leading cause of morbidity worldwide, with a continuously rising prevalence in China [1]. Despite advances in medical management, the risk of death among CAD patients remains substantial. Recently, growing attention has focused on trimethylamine N-oxide (TMAO), a gut microbiota-dependent metabolite, which has been implicated in the pathogenesis and progression of CAD [2–4]. TMAO is produced when dietary nutrients such as choline, L-carnitine, and betaine, commonly found in red meat, eggs, sausage, and processed foods, are metabolized by intestinal flora into trimethylamine (TMA). This TMA is subsequently oxidized to TMAO by flavin-containing monooxygenase 3 (FMO3) in the liver [5, 6].

Accumulating evidence has demonstrated that TMAO is closely associated with several pathophysiological processes that underlie CAD. These include disturbances in cholesterol and bile acid metabolism [7–9], foam cell formation [7], inflammation, endothelial dysfunction [10, 11], platelet activation [12–14], atherosclerosis [7, 15], fibrosis and vascular aging [7, 16].

Several studies have further suggested that elevated circulating TMAO levels are associated with an increased risk of major adverse cardiovascular events (MACEs) in CAD patients [17–19]. For instance, Tang et al. reported that higher plasma TMAO levels predicted an increased risk of MACEs, including all-cause death, in 4007 patients with stable CAD during a three-year follow-up [17]. Similarly, Toru et al. demonstrated that plasma TMAO was a superior independent predictor of all-cause mortality or myocardial infarction compared to other biomarkers (adrenomedullin, oxidized LDL, and natriuretic peptides) in 1079 patients with acute myocardial infarction over a two-year period [18]. Recently, Yu et al. found that elevated TMAO levels were significantly associated with MACEs only in CAD patients with diabetes in northern China, but not in those without diabetes [19], which clearly contrasted with previous studies.

Despite the potential of TMAO as a biomarker for CAD prognosis and as a target for therapeutic intervention, the current evidence remains limited and somewhat inconsistent. Particularly, the association between plasma TMAO levels and all-cause mortality among Chinese CAD patients, after adjusting for traditional risk factors (including diabetes), warrants further investigation. Therefore, this study aimed to investigate the relationship between circulating TMAO levels and all-cause mortality in a cohort of CAD patients from South China.

Methods

Study population and clinical outcomes

This observational cohort study enrolled 389 patients hospitalized with CAD between January 1, 2022, and December 31, 2022, at the Department of Geriatrics, Xiangya Hospital, Central South University. The diagnosis of CAD was confirmed based on coronary angiography, and defined as ≥ 50% stenosis in the luminal diameter of one or more major epicardial coronary arteries. The exclusion criteria for patients were as follows: (1) aged < 18 years old; (2) a diagnosis of malignant tumor requiring advanced medical or surgical therapy; (3) with clinically significant infectious diseases; (4) patients with severe hepatic dysfunction [alanine amino transferase (ALT) level > 135 U/L].

All patients underwent comprehensive clinical evaluations, including physical examinations and biochemical assessments such as complete blood count, blood glucose, blood lipid profiles, hepatic and renal function tests, high-sensitivity troponin T (hs-TnT), and N-terminal pro-brain natriuretic peptide (NT-proBNP). Left ventricular ejection fraction (LVEF) was assessed using transthoracic echocardiography. Hypertension was defined as systolic blood pressure ≥ 140 mmHg or diastolic blood pressure ≥ 90 mmHg, consistent with standard guidelines [20]. Diabetes was diagnosed based on a fasting glucose level of ≥ 7.0 mmol/L or a 2-hour oral glucose tolerance test (OGTT) value of ≥ 11.1 mmol/L [21]. Smoking status included both current and former smokers.

Clinical outcomes focused on all-cause mortality. These outcomes were tracked and verified semi-annually through a combination of telephone interviews, hospital outpatient visits, and official hospital records.

The study was conducted in compliance with the Declaration of Helsinki and received approval from the Ethics Committee of Xiangya Hospital, Central South University (Approval number: 2024040396). Written informed consent was obtained from 389 participants or their legal guardians, as appropriate (age range: 38–95 years, 271 males, and 118 females).

Sample preparation and TMAO analysis

Blood samples were collected from the peripheral veins of patients in the morning of the second day of hospitalization, following an overnight fast. The samples were drawn into ethylenediaminetetraacetic acid (EDTA) anticoagulant tubes, centrifuged at 3000 rpm for 10 min, and the plasma fraction was separated and stored at − 80 °C until further analysis.

Plasma TMAO levels were measured using high-performance liquid chromatography-tandem mass spectrometry (HPLC-MS) with a d9-(trimethyl)-labeled internal standard, following previously established protocols [22]. Absolute quantification was performed by calculating the peak area ratio of TMAO to the internal standard TMAO-d9, with the concentration of TMAO standards used as the independent variable. The actual TMAO concentration was determined based on a standard calibration curve generated during each analytical run. All sample recovery rates were within the range of 85%–115%, and the coefficients of variation (CV) for precision and interbatch variability were both ≤ 15%, meeting the established acceptance criteria.

Other laboratory parameters, including renal and liver function, lipid profiles, and cardiac biomarkers, were assessed in the clinical laboratory of Xiangya Hospital. Estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) formula.

Management of missing data and outliers

To minimize bias, variables with missing values less than 5% (triglyceride, total cholesterol, high-density lipoprotein cholesterol, eGFR, NT-proBNP, hs-TnT, and LVEF) were replaced with the mean of that variable. Variables with abnormal values were handled by the winsorization method, with the 1% and 99% as cut-off points. All missing and abnormal data were processed using the Statistical Package for Social Sciences (SPSS; version 26.0) software.

Statistical analysis

Continuous variables were presented as mean ± standard deviation (SD) or median (interquartile range, IQR), depending on their distribution. If the data followed a normal distribution and exhibited homogeneity of variance, Student’s t-test was used for inter-group comparisons; otherwise, the Mann–Whitney U-test was employed. Categorical variables were expressed as numbers (percentages) and compared between groups using the chi-square test or Fisher’s exact test, as appropriate.

The receiver operating characteristic (ROC) curve analysis was used to assess the association of selected variable TMAO. The optimal cutoff value for identifying increased mortality risk was calculated using the Youden index. The area under the ROC curve and its statistical significance were also calculated. The Kaplan–Meier (KM) survival curve was plotted, and the two groups were compared using the log-rank test to assess statistical significance. We used both univariate and multivariate Cox proportional hazard models to investigate the connection between TMAO and all-cause mortality. To assess potential effect modification by subgroup variables on the TMAO association, the “TMAO grouping × subgroup variable” interaction term in the Cox models was statistically tested using the R package “jstable” on the CNSknowall Platform (https://cnsknowall.com), and a forest plot was generated.

Considering the limited number of deaths in this study, we utilized bootstrapping (with 2000 bootstrap iterations) to reduce the risk of model overfitting. By repeatedly sampling the data for model validation. All statistical analyses were performed using SPSS (version 26.0; IBM Corp., Armonk, NY, USA), the CNSknowall Platform (a comprehensive web service for biomedical data analysis and visualization) or SPlot (a free online platform for data visualization and graphing) [23]. P < 0.05 was considered statistically significant.

Results

Baseline characteristic

Among the 389 CAD patients, 25 were lost to follow-up, leaving 364 patients for outcome analysis. During a median follow-up of 39 months (IQR: 37, 42 months), 40 deaths (11.0%) were recorded.

The baseline characteristic of the 364 enrolled CAD patients are summarized in (Table 1). The mean age of the cohort was 68.1 ± 10.4 years, and 30.8% of the participants were female. Patients were categorized into two groups based on survival outcomes (Live vs. Death) during follow-up.

Table 1.

Baseline characteristic stratified by survival outcomes in CAD patients

Variables	All N = 364	Live N = 324	Death N = 40	P-value
Age (yrs)	68.1 ± 10.4	67.1 ± 9.95	75.9 ± 11.0	< 0.001
Female (N, %)	112 (30.8)	96 (29.6)	16 (40.0)	0.180
Diabetes (%)	37.9	36.1	52.5	0.044
Smoke (%)	47.5	47.8	45.0	0.734
Hypertension (%)	71.2	69.4	85.0	0.040
Chronic lung disease (%)	12.9	12.0	20.0	0.194
CKD (%)	13.7	11.1	35.0	0.001
Postoperative malignancy (%)	0.8	0.3	5.0	0.021
CAD type				0.003
SA(%)	6.3	7.1	0.0
UA(%)	54.9	54.6	57.5
MI(%)	30.2	30.9	25.0
ICM(%)	8.5	7.4	17.5
Number of coronary diseased vessels				0.181
1 vessel (%)	17.3	16.0	27.5
2 vessels (%)	16.8	17.3	12.5
3 vessels/left main(%)	65.9	66.7	60.0
TG (mmol/L)	1.32 (0.96, 1.93)	1.32 (0.96, 1.95)	1.35 (0.99, 1.82)	0.721
TC (mmol/L)	3.77 (3.12, 4.48)	3.76 (3.13, 4.50)	3.82 (2.90, 4.34)	0.441
HDL-C (mmol/L)	1.02 (0.86, 1.17)	1.03 (0.87, 1.18)	0.94 (0.77, 1.10)	0.070
LDL-C (mmol/L)	2.37 (1.78, 2.99)	2.35 (1.79, 3.02)	2.60 (1.70, 2.93)	0.875
eGFR (mL/min/1.73 m2)	80.15 (61.95, 91.40)	83.15 (65.05, 92.55)	52.00 (25.63, 71.93)	< 0.001
NT-proBNP (pg/mL)	208.42 (62.20, 782.33)	157.85 (56.94, 757.87)	1713.79 (355.20, 3189.39)	< 0.001
hs-TnT (pg/mL)	9.0 (4.0, 32.0)	8.5 (3.0, 26.0)	30.5 (9.0, 91.3)	0.238
LVEF (%)	54.0 (40.3, 62.0)	55.0 (42.0, 62.0)	43.5 (29.0, 61.0)	0.020
TMAO (ng/mL)	201.82 (120.91, 299.80)	193.31 (113.67, 280.26)	324.81 (167.06, 649.19)	0.001
Antiplatelet drugs (%)	98.6	99.4	92.5	0.011
ACEI/ARB (%)	69.0	69.4	65.0	0.567
β-blocker (%)	81.3	81.5	80.0	0.821
Statins (%)	98.9	99.4	95.0	0.088

Values expressed as mean ± SD, percentage or median (interquartile range)

Abbreviations: CKD Chronic kindney disease, CAD Coronary Artery Disease, SA stable angina, UA unstable angina, MI myocardial infarction, ICM ischemic cardiomyopathy, TG Triglyceride, TC total cholesterol, HDL-C high density lipoprotein cholesterol, LDL-C low-density lipoprotein cholesterol, eGFR estimated glomerular filtration rate, NT-proBNP N-terminal pro-brain natriuretic peptide, hs-TnT high‐sensitivity troponin T,LVEF left ventricular ejection fraction, TMAO trimethylamine N-oxide, ACEI Angiotensin-Converting Enzyme Inhibitor, ARB Angiotensin Receptor Blocker

Patients who experienced all-cause mortality were generally older, had lower eGFR and LVEF, and lower usage of antiplatelet drugs, and showed a higher prevalence of diabetes and hypertension. They also exhibited significantly higher levels of plasma TMAO and NT-proBNP compared to those who survived (all P<0.05).

Baseline characteristic of participants with missing data were compared with those with complete data (including age, sex, diabetes, smoking status, hypertension, eGFR, and TMAO), with all P-values >0.05, indicating no systematic differences (S1 Table).

Baseline fasting plasma TMAO and all-cause mortality

In the ROC curve analysis of the entire population, TMAO levels above 317.62 ng/mL identified patients at greater risk of death, with a sensitivity of 55%, specificity of 81%, area under the ROC curve (AUC) 0.68; 95% CI: 0.579–0.780; P< 0.0001, and Youden index 0.37 (Fig. 1).

Fig. 1.

The ROC curve of plasma TMAO level correlates with the prognosis of CAD patients

Using this cutoff (317.62 ng/mL) determined in the ROC curve analysis, patients were classified into two groups: Group 1 (n = 283) with lower TMAO levels and Group 2 (n = 81) with higher TMAO levels. Kaplan–Meier survival analysis demonstrated that patients with elevated TMAO levels (> 317.62 ng/mL) had a significantly higher incidence of all-cause mortality compared to those with lower TMAO levels (≤ 317.62 ng/mL), with an unadjusted HR of 4.546 (95% CI: 2.087–9.903; P < 0.0001) (Fig. 2).

Fig. 2.

Kaplan–Meier analysis of all-cause mortality based on plasma TMAO levels in CAD patients. Patients with higher TMAO levels (> 317.62 ng/mL) exhibited a significantly increased risk of all-cause mortality compared to those with lower levels (TMAO ≤ 317.62 ng/mL), with an unadjusted HR of 4.546 (95% CI: 2.087–9.903; P < 0.0001)

Association between TMAO levels and all-cause mortality

Univariate COX regression analyses stratified by survival outcomes are presented in S2 Table. Multifactorial COX regression analyses were presented in (Table 2). Model 1 adjusted for traditional risk factors including age, hs-TnT, NT-proBNP, LVEF and eGFR. In Model 1, elevated plasma TMAO levels were independently associated with an increased risk of all-cause mortality (HR 2.657; 95% CI: 1.389 to 5.084; P = 0.003). Given the importance of lifelong antiplatelet therapy for secondary prevention in CAD and its link to improved survival, we included antiplatelet use as a covariate. After adjustment for antiplatelet use, age, hs-TnT, NT-proBNP, LVEF, and eGFR, elevated TMAO remained significantly associated with all-cause mortality (HR = 2.644, 95% CI: 1.370–5.105, P = 0.004).

Table 2.

Relationship between all-cause mortality and plasma TMAO levels based on multivariate Cox regression analysis and bootstrapping analysis

Model	Multivariate analysis				Bootstrapping analysis
	P	HR	CI lower	CI upper	P	HR	CI lower	CI upper
Unadjusted	< 0.001	4.766	2.556	8.889	< 0.001	4.768	2.609	9.143
Model 1	0.003	2.657	1.389	5.084	0.003	2.656	1.343	5.501

Model 1: adjusted for conventional risk factors including age, hs-TnT, NT-proBNP, LVEF and eGFR

Through bootstrap validation, the association between elevated TMAO levels and increased mortality risk remained statistically significant across 2000 bootstrap samples (HR 2.625; 95% CI: 1.330 to 6.190; P = 0.004). To assess the robustness of our findings to potential confounding factors and model specification, we log-transformed TMAO and included it as a continuous variable in the Cox model (HR = 1.509, 95% CI: 1.002–2.272, P = 0.049).

Because severe illness or proximity to death may alter TMAO levels, we performed landmark analyses excluding patients who died within 30 days (n = 2) and within 90 days (n = 5) after enrollment. The association between TMAO and all-cause mortality remained significant in both sensitivity analyses: after excluding 30-day deaths, HR = 2.743 (95% CI: 1.536–5.239), P = 0.002; after excluding 90-day deaths, HR = 2.724 (95% CI: 1.412–5.255), P = 0.003.

Stratified analysis and interaction tests

Forest plots (Fig 3) demonstrated the impact of TMAO on all-cause mortality across various subgroups between Group 1 and Group 2 including age (<65 years and ≥65 years), gender (female and male), hypertension (Yes and No), smoke (Yes and No), diabetes (Yes and No), eGFR (<90 mL/min/1.73 m2 and ≥90 mL/min/1.73 m2), LDL-C (<2.6 mmol/L and ≥2.6 mmol/L), hs-TnT (≤0.014 ng/mL and >0.014 ng/mL), NT-proBNP (<300 pg/mL and ≥300 pg/mL) (Fig 3). No significant interactions were observed between plasma TMAO levels and any of the predefined factors (P for interaction >0.05), including Sex (P for interaction= 0.394), Age (P for interaction = 0.789) and Diabetes (P for interaction=0.207), indicating that the impact of TMAO on all-cause mortality was consistent across these subgroups.

Fig. 3.

Subgroup analysis of all-cause mortality risk by TMAO levels stratified by clinical and laboratory parameters. Forest plot showing hazard ratios (squares) for all-cause mortality over a median follow-up period of 39 months, comparing Group 1 (TMAO ≤317.62 ng/mL) and Group 2 (TMAO >317.62 ng/mL). Bars represent 95% CIs. Wide CIs in some subgroups reflect limited sample sizes and event rates. The multivariable-adjusted model accounts for confounding factors, including age, gender, hypertension, smoking status, diabetes, eGFR, LDL-C, hs-TnT, NT-proBNP. Cut-off points for the covariates are as follows: age (<65 vs. ≥65 years), eGFR (<90 vs. ≥90 mL/min/1.73 m2), LDL-C (<2.6 vs. ≥2.6 mmol/L), hs-TnT (≤0.014 vs. >0.014 ng/mL), and NT-proBNP (<300 vs. ≥300 pg/mL).

Discussion

This study demonstrates that elevated baseline plasma TMAO levels are significantly associated with an increased risk of all-cause mortality in Chinese patients with CAD, regardless of the presence of diabetes, after adjustment for some conventional risk factors, including comorbidities (e.g., hypertension), demographic characteristics , and some cardiorenal indices (NT-proBNP and eGFR).

Although both baseline plasma TMAO levels and diabetes prevalence were significantly higher in patients who died, the present study demonstrated that the association of TMAO for all-cause mortality remained unaffected by concomitant diabetes. The interrelationship between TMAO and diabetes have been gradually studied and reported over the past decade. Animal models suggest that TMAO may affect glucose metabolism and promote diabetes [24–26]. In human observational studies, TMAO levels are often elevated in individuals with diabetes [27–29]. One prospective Chinese cohort study found that higher TMAO levels were associated with increased diabetes risk over 8.9 years follow-up [29], while a U.S. study found no such association over 12 years [30]. These inconsistencies underscore the complex relationship between TMAO and diabetes.

Our study corroborated the association of TMAO for all-cause mortality, as reported in earlier clinical research. Specifically, elevated plasma TMAO levels were associated with either MACEs (encompassing all-cause mortality) or all-cause mortality alone in two stable CAD cohorts, regardless of diabetic status, over 3- and 5-year follow-ups [17, 31]. Furthermore, an independent association between elevated TMAO levels and poor prognosis (all-cause death and myocardial infarction) was observed over 2-year follow-up in a separate cohort with acute myocardial infarction[18]. Our findings diverge from those reported in a northern Chinese cohort, in which the association between TMAO and MACEs was observed only in CAD patients with diabetes. The absence of this association in their non-diabetic subgroup may be attributable to methodological differences in TMAO stratification [19]. Whereas the prior study categorized TMAO levels by tertiles, our analysis employed an optimal cutoff value [19].

Our study also showed that the deceased cohort had higher plasma TMAO levels and a lower eGFR, which was consistent with some clinical reports linking higher plasma TMAO levels to worse survival in patients with chronic kidney disease [22, 32]. TMAO is bidirectionally associated with renal function since it is both cleared by the kidneys and associated with impaired renal function and renal fibrosis [16, 33]. Consistent with prior research, our results support the existence of a robust association between elevated TMAO levels and increased mortality risk in both stable CAD and acute myocardial infarction cohorts, even after adjusting for the cardiorenal indicator eGFR [18, 31].

Our findings add to a growing body of evidence supporting the potential clinical application value of novel biomarker TMAO beyond traditional cardiac troponins in patients with CAD. In resent years, several studies have evaluated alternative biomarker-based strategies for rapid rule-out and risk stratification in patients with acute coronary syndrome (ACS)，which is a very severe type of CAD. For instance, a dual marker strategy (hs-cTnT and copeptin) ensures a better reliable acute myocardial infarction rule-out with low 12-month outcome rates for the special emergency department [34]. Among patients diagnosed with non-ST-elevation myocardial infarction, cardiac myosin-binding protein C (cMyBP-C) , comparable to hs-cTnT, have demonstrated excellent diagnostic accuracy [35]. The GRACE 1.0 score serves as a key instrument for evaluating mortality risk in acute coronary syndrome. Mustafa Y et al. found that a low GRACE 1.0 score combined with ≤1 elevated biomarker including C-reactive protein, NT-proBNP, D-dimers, eGRF, copeptin and hemoglobin significantly improves mortality prediction for patients with suspected non-ST-elevation acute coronary syndrome [36], and dual-marker strategies (DMS) combining hs-cTnT with either cMyBP-C or copeptin provide highly reliable and safe protocols for non–ST-segment myocardial infarction rule-out [37]. While these biomarkers are primarily used for immediate diagnosis and short-term risk stratification for CAD patients in the emergency setting. An increasing body of research has demonstrated a strong association between elevated plasma TMAO levels and adverse outcomes in CAD patients. Meanwhile, TMAO may capture different pathophysiological pathways—such as gut microbiota-derived metabolite accumulation, inflammation, and potential attenuation of antiplatelet drug efficacy—that are not directly reflected by cardiac troponins. Therefore, further research is warranted to explore the value of TMAO, alone or in combination with other biomarkers, in identifying risk stratification and long-term prognosis among patients with critical coronary heart disease, including acute coronary syndrome.

Previous research have found that TMAO is closely associated with pathophysiological processes of CAD. Emerging evidence further suggests that TMAO may influence the efficacy of antiplatelet therapy, which is central to CAD management. Aspirin and clopidogrel are the most commonly prescribed antiplatelet agents. Zhu et al. demonstrated elevated TMAO levels appeared to partially attenuate the antiplatelet effects of low-dose aspirin (81 mg per day) in healthy individuals [12]. Additionally, TMAO has been found to significantly increase Carboxylesterase 1 protein activity and clopidogrel hydrolysis in the liver, leading to reduced formation of clopidogrel's active metabolite and impaired platelet response in mice [38]. In a rat model of ischemia and reperfusion, TMAO also mitigated the antiplatelet aggregation of clopidogrel [13]. These findings suggest that elevated plasma TMAO may weaken the response to antiplatelet drugs, potentially increasing the risk of major adverse cardiovascular events, including all-cause mortality, in CAD patients prescribed aspirin or clopidogrel.

This study has several limitations. First, this was a retrospective, single‑center observational analysis, which may limit the generalizability of our findings to other populations or settings, and residual confounding cannot be fully excluded. In addition, the study protocol and analysis plan were not pre‑registered in a public repository，and the relatively small number of all‑cause deaths (n=40) and the modest sample size increase the risk of model overfitting and limit statistical power. Therefore, the findings should be considered exploratory and hypothesis-generating, and require validation in larger, pre‑registered, multi‑center prospective cohorts. Furthermore, we acknowledge that the optimal cut‑off value for TMAO derived using the Youden index may be subject to optimistic bias, and external validation in an independent cohort is needed before clinical application. Second, data on dietary intake, which could influence plasma TMAO levels, were not available. Third, we were unable to reliably ascertain cause‑specific mortality, because follow‑up was predominantly conducted by telephone (family members were often unwilling to disclose details of the death), and many out‑of‑hospital deaths occurred without medical documentation. Consequently, we could not analyze cardiovascular mortality separately. Fourth, the absence of inflammatory biomarkers or gut microbiota profiling limits mechanistic insight, which we plan to address in subsequent studies. Finally, the median follow‑up duration of 39 months, while informative for mid‑term outcomes, is relatively short for defining long‑term (e.g., ≥5 years) mortality in CAD patients; extended follow‑up is warranted to more comprehensively assess long‑term survival.

This study demonstrates that elevated plasma TMAO levels are independently associated with an increased risk of all-cause mortality in patients with CAD in South China, irrespective of diabetes status. These findings highlight the prognostic value of TMAO as a potential biomarker for risk stratification in CAD. Our future research should focus on elucidating the underlying mechanisms linking TMAO to adverse outcomes.

Abbreviations

CKD

Chronic kindney disease

CAD

Coronary Artery Disease

UA

Unstable angina

SA

Stable angina

UA

Unstable angina

MI

Myocardial infarction

ICM

Ischemic cardiomyopathy

TG

Triglyceride

TC

Total cholesterol

HDL-C

High density lipoprotein cholesterol

LDL-C

Low-density lipoprotein cholesterol

eGFR

Estimated glomerular filtration rate

NT-proBNP

N-terminal pro-brain natriuretic peptide

hs-TnT

High-sensitivity troponin T

LVEF

Left ventricular ejection fraction

TMAO

Trimethylamine N-oxide

ACEI

Angiotensin-Converting Enzyme Inhibitor

ARB

Angiotensin Receptor Blocker

Authors’ contributions

Hao Zhang: Conceptualization, Formal analysis, Writing–original draft, Investigation, Methodology. Yuan-yuan Peng: Conceptualization, Methodology, Investigation, Project administration. Xi-ya Lu: Methodology, Investigation, Data curation. Yu-xing Li: Methodology, Investigation, Data curation. Wen-zhi Li: Methodology, Investigation, Data curation. Yi Hu: Methodology, Investigation, Data curation. Li Wan: Methodology, Investigation, Data curation. Bi-lian Chen: Conceptualization, Methodology, Investigation, Writing - review & editing, Supervision, Project administration.

Funding

This work was supported by grants from the Changsha Science and Technology Bureau, Hunan province, China (KQ2202367); the Science and Technology Department of Hunan Province (2022JJ70156), Hunan Key Laboratory for Bioanalysis of Complex Matrix Samples (FZJZ-202307); and the National Natural Science Foundation of China (82474013).

Data availability

The datasets used and analysed during the current study are available from the corresponding author on reasonable request.

Declerations

Ethics approval and consent to participate

The study was conducted in accordance with the Declaration of Helsinki (as revised in 2024) and received approval from the Ethics Committee of Xiangya Hospital, Central South University (Approval number: 2024040396).

Consent for publication

Not applicable

Competing interests

The authors declare no competing interests.