Metabonomic Biomarkers of Plaque Burden and Instability in Patients With Coronary Atherosclerotic Disease After Moderate Lipid‐Lowering Therapy

Zhendong Mei; Lili Xu; Qingxia Huang; Chenhao Lin; Mengyao Yu; Shalaimaiti Shali; Hongyi Wu; Yijing Lu; Runda Wu; Zhen Wang; Lingfeng Luo; Zhonghan Sun; Liang Sun; Juying Qian; Guochong Chen; Huiru Tang; Kang Yao; Yan Zheng; Yuxiang Dai; Junbo Ge

doi:10.1161/JAHA.124.036906

. 2024 Dec 10;13(24):e036906. doi: 10.1161/JAHA.124.036906

Metabonomic Biomarkers of Plaque Burden and Instability in Patients With Coronary Atherosclerotic Disease After Moderate Lipid‐Lowering Therapy

Zhendong Mei ^1,^2,^*, Lili Xu ^1,^3,^*, Qingxia Huang ^4,^*, Chenhao Lin ^2,^*, Mengyao Yu ⁵, Shalaimaiti Shali ¹, Hongyi Wu ¹, Yijing Lu ¹, Runda Wu ¹, Zhen Wang ¹, Lingfeng Luo ², Zhonghan Sun ², Liang Sun ⁶, Juying Qian ¹, Guochong Chen ⁷, Huiru Tang ⁴, Kang Yao ^1,^✉, Yan Zheng ^1,^2,^6,^✉, Yuxiang Dai ^1,^✉, Junbo Ge ^1,³

PMCID: PMC11935549 PMID: 39655754

Abstract

Background

Contemporary risk assessment in patients with coronary atherosclerotic disease (CAD) often relies on invasive angiography. However, we aimed to explore the potential of metabolomic biomarkers in reflecting residual risk in patients with CAD after moderate lipid‐lowering therapy.

Methods and Results

We analyzed serum metabolomic profile among 2560 patients with newly diagnosed CAD undergoing moderate lipid‐lowering therapy, through nuclear magnetic resonance spectroscopy and quantified 175 metabolites, predominantly lipoproteins and their components. CAD severity was evaluated using Gensini score for plaque burden and circulating cardiac troponin T levels for plaque instability. The association of metabolites with CAD severity was examined using multivariate linear regression, and the underlying potential causality was explored using a 2‐sample Mendelian randomization approach. Two composite metabolomic indices were constructed to reflect CAD severity using least absolute shrinkage and selection operator linear regression, and their associations with risk of major adverse cardiac events during a median follow‐up of 3.8 years were evaluated using Cox models. Our investigation revealed that triglycerides and apolipoprotein B in low‐density lipoprotein particles displayed stronger associations with CAD severity compared with the clinically used low‐density lipoprotein cholesterol marker. In large high‐density lipoprotein, components like cholesterol, cholesterol esters, triglyceride, apolipoprotein A1/A2 showed inverse associations with CAD severity. Certain metabolites, including apolipoprotein B and dihydrothymine, showed a putative causal link with Gensini score. Notably, per standard deviation increase in Gensini score–based metabolomic index was associated with 14.8% higher major adverse cardiac event risk (hazard ratio, 1.148 [95% CI, 1.018–1.295]) independent of demographic factors, medication use, and disease status.

Conclusions

Our findings highlight the potential of nuclear magnetic resonance–based metabolomics in identifying novel biomarkers of plaque burden and instability. Metabolites related to plaque burden may facilitate noninvasive assessment of CAD prognosis.

Keywords: coronary atherosclerosis severity, metabolomics, prognosis, risk assessment

Subject Categories: Cardiovascular Disease, Functional Genomics

Nonstandard Abbreviations and Acronyms

CE: cholesterol ester
cTnT: cardiac troponin T
FDR: false discovery rate
GRAND: Genetics and Clinical Characteristics of Coronary Artery Disease in the Chinese Young Adults
IDL: intermediate‐density lipoprotein
IV: instrumental variable
MACE: major adverse cardiac event
MR: Mendelian randomization
NAG: N‐acetyl‐glycoprotein
NMR: nuclear magnetic resonance
RuLAS: Rugao Longevity and Aging Study

Clinical Perspective.

What Is New?

Lipid subfractions and components profiled by nuclear magnetic resonance spectroscopy were associated with coronary plaque burden and instability among patients with coronary atherosclerotic disease after moderate lipid‐lowering therapy.
The metabolomic index of plaque burden, instead of the plaque stability, was associated with a higher risk of major adverse cardiac events.
A 2‐sample Mendelian randomization approach revealed potential causal associations of certain metabolites, including apolipoprotein B and dihydrothymine, with plaque burden and stability.

What Are the Clinical Implications?

Noninvasive metabolomic profiling could introduce new prognostic methods for CAD, providing a potential alternative to traditional, invasive coronary angiography.

The increasing prevalence of coronary atherosclerotic disease (CAD) constitutes a heavy public health burden worldwide. ¹ Recognized as a condition influenced by traditional risk factors such as hypertension, diabetes, and dyslipidemia, CAD remains a leading cause of morbidity and death. ² , ³ , ⁴ While lipid‐lowering therapies are pivotal in CAD management, they offer incomplete protection against adverse cardiovascular events, ⁵ highlighting the need for identifying new risk factors and targets for therapy.

Metabolic disturbances, especially dyslipidemia, not only contribute to the onset of atherosclerosis but are also integral to its advancement. ² Standard clinical lipid profiles, which measure parameters like low‐density lipoprotein cholesterol (LDL‐C), and high‐density lipoprotein cholesterol (HDL‐C), lack the granularity to distinguish lipoproteins by size, density, or composition—factors potentially critical to CAD risk stratification. ⁶ Advanced nuclear magnetic resonance (NMR) metabolomics has emerged as a powerful tool, providing nuanced insights into lipoprotein particles and their lipids, categorized by size. ⁷ Although metabolic signatures have been investigated in cardiovascular disease risk within Western cohorts, ⁸ , ⁹ , ¹⁰ there is a paucity of large‐scale research focusing on the metabonomic characteristics of Asian populations with CAD and the direct links between metabolite profiles and the severity of coronary atherosclerosis.

In the current study, we measured 175 circulating metabolic parameters using NMR metabolomic profile in 2560 patients with CAD receiving moderate lipid‐lowering therapy based on a large multicenter study (GRAND [Genetics and Clinical Characteristics of Coronary Artery Disease in the Chinese Young Adults] ¹¹ ). We aimed to evaluate the association of metabolomic biomarkers with plaque burden and instability in these CAD patients and further explored the potential causality using the 2‐sample Mendelian randomization (MR) approach. In addition, we assessed the capability of metabolomic biomarkers to predict the risk of major adverse cardiovascular events (MACEs).

Methods

The data supporting the findings of this study are available from the corresponding authors upon reasonable request. The study protocol was approved by the Institutional Review Board of Zhongshan Hospital, Fudan University, Shanghai, China (Approval No. B2017‐051). All participants provided written informed consent.

Study Population

This study was conducted on the basis of the GRAND study ¹¹ in an age‐range extended study population. CAD was defined by coronary angiography as luminal narrowing of >50% in at least 1 main coronary artery, including the left main artery, left anterior descending artery, left circumflex artery, and right coronary artery. This condition may manifest as occult CAD, stable angina pectoris, unstable angina pectoris, non–ST‐segment–elevation myocardial infarction (MI), or ST‐segment–elevation MI. Patients who (1) were severely sick with a limited life expectancy of <1 year, (2) had malignancies, or (3) were pregnant or planning to become pregnant were excluded from the study. Finally, a total of 3047 participants of all‐age range newly diagnosed with CAD by coronary angiography were included from 38 cardiac centers in China from May 2017 to April 2019. Among them, 2560 participants receiving moderate lipid‐lowering therapy provided qualified blood samples for metabolomics profiling and were included in the current metabolomic analysis. The blood samples were collected on the same day as the coronary angiography. All included participants completed the evaluation of the Gensini score, and 2383 also underwent measurement of circulating cardiac troponin T (cTnT) levels (Figure S1). This study defined moderate lipid‐lowering therapy as therapy with a theoretical reduction of >40% in LDL‐C levels (eg, atorvastatin 20 mg, rosuvastatin 10 mg, and pravastatin 40 mg), according to the available literature on the efficacy of statins in Asian patients with CAD. Combination therapy of a statin with ezetimibe was taken into account. Proprotein convertase subtilisin/kexin type 9 inhibitors had not been widely used when the study was conducted and therefore were not taken into consideration. Self‐reported sex, body weight, body height, smoking status, and biochemistry indices, including HDL‐C, LDL‐C, total cholesterol, and triglycerides, along with estimated glomerular filtration rate (eGFR) were obtained from electronic medical records. Body mass index (BMI) was computed by dividing the weight (in kilograms) by the square of height (in meters).

Evaluation of Coronary Atherosclerosis Severity

Coronary atherosclerosis severity was evaluated using Gensini score and circulating cTnT level, which stand for plaque burden and plaque instability respectively. Gensini score is an angiography‐based scoring system to quantify the coronary atherosclerosis burden, in which a higher score accounts for a more severe proximal lesion by combining the degree of luminal narrowing and the location of narrowing. ¹² Calculation of the Gensini score was initiated by giving a severity score to each coronary stenosis and by summation of individual coronary segment scores. ¹² , ¹³ , ¹⁴ Serum cTnT levels reflect the severity of myocardial injury and are routinely analyzed as a biomarker in the clinical diagnosis of MI, which is often caused by the rupture of unstable atherosclerotic plaques and subsequent thrombosis and occlusion in the coronary artery. A higher level of cTnT accounts for the presence of more unstable atherosclerotic plaques and a larger range of myocardial injury. ¹⁵ cTnT was measured using venous blood samples that were obtained at admission and by an automated analyzer using a high‐sensitivity assay (Roche Diagnostics, Indianapolis, IN). The lower limit of high‐sensitivity cTnT that can be reproducibly measured with a coefficient of variation <10% was 0.003 ng/mL, and the 99th percentile upper range limit was 0.014 ng/mL. Clinical, biochemical, and procedural data were recorded for all patients.

Follow‐Up Visits

A total of 2204 patients were followed up via the clinic visit or telephone interview to record the occurrence of major adverse cardiac events (MACEs), which were defined as the composite of all‐cause death, nonfatal MI, stroke, ischemia‐driven revascularization, and progression of coronary atherosclerosis. The progression of coronary atherosclerosis was characterized by the incidence of new stenosis, with a minimum of 50% constriction in a previously normal vessel, or an escalation in the grade of preexisting stenosis by >20% within any coronary artery.

Measurement of NMR‐Based Metabolomics

Serum Sample Preparation

A fasting blood sample was collected from each participant at admission. Serum was separated by centrifugation at 2200 to 2500 rpm for at least 15 minutes and stored at −80 °C. Then, 320 μL of each serum sample and equivalent phosphate buffer saline (0.085M, pH 7.4, containing 10% D₂O and 4.644 mM sodium 3‐trimethylsilyl‐propionate‐2,2,3,3‐d4) were mixed, and 600 μL of the mixture was transferred into a 5 mm NMR tube for NMR analysis.

NMR Analysis

All NMR measurements were performed at 310K on a Bruker Avance III HD 600 MHz NMR spectrometer equipped with a 5 mm broad‐band inverse probe (Bruker Biospin GmbH, Germany) according to protocols previously published. ¹⁶ Two ¹H NMR spectra were acquired using NOESYGPPR1D and LEDBPGPPR2S1D pulse sequences, and 32 transients were recorded for all spectra with the spectral width of 20 ppm and 96K data points. Quantitative data for 41 low‐molecular‐weight metabolites and 112 lipoproteins together with their components were obtained, respectively, using the Bruker IVDr Quantification in Serum (B.I.Quant‐PS™) and Lipoprotein Subfraction Analysis (B.I.LISA™) methods.

These lipoproteins consisted of very‐low‐density lipoprotein (VLDL; 0.950–1.066 kg/L), intermediate‐density lipoprotein (IDL; 1.006–1.019 kg/L), low‐density lipoprotein (LDL; 1.019–1.063 kg/L), high‐density lipoprotein (HDL; 1.063–1.210 kg/L), 15 subfractions (5 VLDLs, 6 LDLs, and 4 HDLs) and their components (apolipoproteins, cholesterol esters [CEs], cholesterol, free cholesterol, phospholipids, and triglycerides). The VLDL was fractioned into 5 density subfractions; the LDL was fractioned into 6 density subfractions (LDL1, 1.019–1.031 kg/L; LDL2, 1.031–1.034 kg/L; LDL3, 1.034–1.037 kg/L; LDL4, 1.037–1.040 kg/L; LDL5, 1.040–1.044 kg/L; and LDL6, 1.044–1.063 kg/L); and the HDL subfractions in 4 density classes (HDL1, 1.063–1.100 kg/L; HDL2, 1.100–1.112 kg/L; HDL3, 1.112–1.125 kg/L; and HDL4, 1.125–1.210 kg/L). ¹⁶ , ¹⁷ In addition, 25 lipoproteins (such as CEs in lipoprotein and their subfractions, cholesterol in non‐HDL, and total phospholipids) were calculated on the basis of the above quantitative data ¹⁸ , ¹⁹ and 8 extra parameters for N‐acetyl‐glycoproteins (NAG1, NAG2) and fatty acids were obtained from the LEDBPGPPR2S1D experiment. ²⁰ , ²¹ , ²² , ²³ A total of 186 metabolite parameters were directly measured. Metabolites with missing rates >50% were removed (n=11). Missing values (under detectable levels) for other metabolites were imputed with half of the minimum value for a given metabolite. Ultimately, 175 parameters were obtained to characterize the metabolomic phenotypes of each human serum sample. The rank‐based inverse normal transformation was applied to the metabolite data before subsequent analyses. Figure S2 shows the Pearson correlation between the included metabolites among all study participants. Table S1 shows the Pearson correlation coefficients and corresponding P values between lipids (total cholesterol, LDL‐C, HDL‐C, and triglycerides) measured by clinical chemistry and that by NMR spectroscopy.

Statistical Analysis

Characteristics of the participants were presented as mean±SD or percentages where appropriate. Student t test for continuous variables and χ² test for categorical variables were used to compare the participant characteristics. Before analyses, circulating levels of cTnT were log transformed, as they were right skewed. The Euclidean dissimilarity metric was calculated on the basis of the rank‐based inverse normal transformation –transformed levels of metabolites to assess variability in metabolite profiles across samples. Subsequently, the associations of metabolite profile with Gensini score and cTnT were assessed using the permutational multivariate ANOVA with 999 permutations. A dimension reduction was further performed through principal coordinate analysis based on the Euclidean dissimilarity metric. The differences in the top 2 principal coordinate loading scores across the levels of Gensini score and cTnT were evaluated by employing ANOVA or Student t test, where appropriate. Multivariable linear regression models with adjustment for age, sex, smoking status, BMI, eGFR, medication use for dyslipidemia, and disease status of hypertension and diabetes were used to assess the associations of metabolites with Gensini score and cTnT levels. Missing values of BMI (n=16), smoking (n=6), and eGFR (n=71) were imputed by the multiple imputation using the R mice package (R Foundation for Statistical Computing, Vienna, Austria). ²⁴ The Benjamini–Hochberg false discovery rate (FDR) method was used to address the multiple testing issue, with a predefined target rate of 0.05.

Stratified analyses were conducted to assess potential effect modification by age (<55 or ≥55 years for men, <65 or ≥65 years for women, according to the definition of premature CAD by the American College of Cardiology/American Heart Association ²⁵ ) and BMI (<24 or ≥24 kg/m²). The interactions of stratified factors and metabolites on Gensini score and cTnT were estimated by including a multiplicative factor in the multivariable linear regression models. Sensitivity analyses were performed by excluding participants with eGFR <60 mL/min per 1.732 m² or those with diabetes.

We constructed 2 metabolomic indices based on coronary plaque burden (ie, Gensini score–based metabolomic index) and plaque instability (ie, cTnT‐based metabolomic index). In each metabolomic index, the metabolites were selected by using the linear regression with least absolute shrinkage and selection operator penalty implemented with R package “glmnet.” ²⁶ All metabolites with nonzero coefficients in the least absolute shrinkage and selection operator regression models were included to construct the metabolomic index weighted by the corresponding coefficients. Hazard ratios (HRs) and 95% CIs of the metabolomic indices with the risk of MACE were estimated by using Cox proportional hazards models with adjustment for the above‐mentioned covariates. The proportional hazards assumption of Cox models was tested on the basis of the Schoenfeld residuals, and no evidence of violation was detected.

All analyses were performed using R version 4.0.3 (https://www.r‐project.org/).

Two‐Sample MR

Data Source of Exposure

The genome‐wide association study (GWAS) summary statistics of metabolites used in the MR analyses were obtained from a metabolomics GWAS study involving a total of 10 792 Chinese individuals. Specifically, the study included 3013 individuals from the National Survey of Physical Traits cohort (61.8% women; mean age, 48.9±12.6 years), 5666 individuals from the Shanghai Changfeng Study ²⁷ (57.3% women; mean age, 63.7±9.6 years), and 2113 individuals from the RuLAS (Rugao Longevity and Aging Study) ²⁸ (54.2% women, mean age, 78.8±4.9 years). Notably, the GWAS study used the identical metabolomics platform as the one used in this study. The level of each metabolite was transformed to a normal distribution using the rank‐based inverse normal transformation. Single‐nucleotide polymorphisms (SNPs) with low frequency (minor allele frequency <5%), deviation from Hardy–Weinberg equilibrium (P<1×10⁻⁵) and low imputation quality(<0.4) were excluded. GWAS was conducted using a linear mixed model under an additive genetic model in BOLT‐LMM version 2.4.1 software (Broad Institute, Cambridge, MA), with adjustment for age, sex, use of lipid‐lowering medication, and the top 5 genetic ancestry principal components. We included 153 metabolites that overlapped with the present study and had >2 genome‐wide significantly associated SNPs, which were included in subsequent MR analyses.

To ensure the validity of instrumental variables (IVs), 3 key assumptions should be satisfied: (1) the IV must associate with the exposure; (2) the IV should be independent of confounding factors in the exposure–outcome association; and (3) the IV should affect the outcome only through the exposure. For each metabolite, we first chose metabolite‐associated SNPs at the threshold of genome‐wide significance (P<5×10⁻⁸) as candidate IVs. Subsequently, to select those independent associations, 1000 Genomes Project Phase 3 version 5 East Asian samples data were used as the reference panel to calculate the linkage disequilibrium) between the SNPs. Among those SNPs in linkage disequilibrium (window size, 10000kb; r ²<0.1), only the SNPs with the lowest P values were chosen. Finally, we calculated the F statistic of each SNP to evaluate the strength of IV, and those SNPs with F<10 were excluded. ²⁹

Data Source of Outcome

To obtain the effect estimates from relevant SNPs on the outcome, that is, Gensini score and cTnT, we performed GWASs within the GRAND cohort individuals who had genotyping data. Specifically, we included 4830 samples for the GWAS of Gensini score and 5265 samples for the GWAS of cTnT. We further excluded SNPs with minor allele frequency <0.05, low imputation quality <0.4, and P value of Hardy–Weinberg equilibrium test <1×10⁻⁵. The GWASs were conducted using linear regression model via PLINK version 1.9 software, ³⁰ with adjustment of age, sex, and the top 5 genetic ancestry principal components.

MR Analysis

GWAS summary statistics of exposures and outcomes were harmonized to make alignment on effect alleles before every MR analysis. Palindromic SNPs (ie, SNPs with A/T or G/C) with intermediate allele frequencies (minor allele frequency >0.42) were excluded. We used the inverse variance weighted method to estimate the causal effect of exposure on outcome in our main results. The inverse variance weighted method estimates the Wald ratio for each SNP and then combines them by a meta‐analysis approach to estimate the overall causal effect. ³¹ Besides, several sensitivity analyses robust to horizontal pleiotropy were used to assess the reliability of our findings, including MR‐Egger regression, the median‐based method, and the mode‐based method. Briefly, the MR‐Egger regression is based on the weaker Instrument Strength Independent of Direct Effect assumption, which allows it to evaluate the existence of directional pleiotropy with an extra test for the intercept term. ³² The weighted mode requires that the largest subset of IVs, which estimate the same causal effect be valid IVs, ³³ and the weighted median method assumes that valid IVs provide more than half of the weight. ³⁴ Furthermore, leave‐one‐out analysis was performed to determine whether the causal signal was driven by a single SNP. The potential directional pleiotropy was detected by the hypothesis test for the intercept of MR‐Egger regression. ³⁵ Cochran's Q statistic was used to quantify the heterogeneity of IVs. To address the multiple testing issues, the Benjamini–Hochberg FDR method was used. The putative causal relationship was determined by FDR <0.05 in the inverse variance weighted measurement and consistent directions in other MR methods. All MR analyses were performed using the R package TwosampleMR version 0.5.6. ³⁶ This study followed the Strengthening the Reporting of Observational Studies in Epidemiology Using Mendelian Randomization guideline. ³⁷

Results

GRAND Cohort Characteristics

In this cohort of 2560 patients with CAD who underwent metabolomics measurements, the average age was 59.7±13.4 years, and the average BMI was 25.2±3.3 kg/m², with 78% being men. The distributions of clinical characteristics by BMI groups (normal weighted group versus overweight/obese group) are presented in the Table. The patients who were overweight/obese tended to be younger, men, and former smokers; have a higher systolic and diastolic blood pressure, and a higher fasting plasma glucose; and be patients with diabetes (all P<0.05). As for the lipid profile, patients who were overweight/obese were more likely to have a lower level of HDL‐C and a higher level of triglycerides (both P<0.001). The proportions of clinical administration of antihypertensive and glucose‐lowering drugs were similar between these 2 groups (Table). During a median follow‐up period of 3.8 (interquartile range, 3.3–3.9) years, incident MACEs were documented in 312 patients. These included 28 all‐cause deaths, 32 cases of nonfatal MI, 35 strokes, 278 cases of ischemia‐driven revascularization, and 201 patients with progression of coronary atherosclerosis.

Table .

Baseline Characteristics of the Study Participants

	Normal weight (N=917)	Overweight/obese (N=1627)	P value
Age, y	62.7±12.6	58.0±13.6	<0.001
Men	655 (71.4)	1327 (81.6)	<0.001
BMI, kg/m²	21.9±1.7	27.0±2.5	<0.001
Smoker			<0.001
Never	553 (60.3)	817 (50.2)
Former	186 (20.3)	533 (32.8)
Current	178 (19.4)	277 (17.0)
Drinker	141 (15.4)	254 (15.9)	0.48
Systolic blood pressure, mm Hg	130.0±19.4	132.0±19.0	0.03
Diastolic blood pressure, mm Hg	76.5±12.0	78.9±11.9	<0.001
Glycated hemoglobin, %	6.3±2.3	6.4±1.3	0.40
Fasting plasma glucose, mmol/L	6.6±2.8	7.0±2.9	0.02
Total cholesterol, mmol/L	4.0±1.3	4.1±2.5	0.17
Triglycerides, mmol/L	1.7±1.3	2.2±4.3	<0.001
HDL‐C, mmol/L	1.2±0.5	1.1±0.9	<0.001
LDL‐C, mmol/L	2.1±1.2	2.2±1.3	0.06
eGFR, mL/(min × 1.73 m²)	84.2±18.2	86.6±19.5	0.003
Hypertension	698 (76.1)	1290 (79.3)	0.10
Diabetes	341 (37.2)	672 (41.3)	0.04
Use of antihypertensive drugs	617 (67.3)	1138 (69.9)	0.18
Use of glucose‐lowering drugs	136 (14.8)	268 (16.5)	0.30
MACEs during the follow‐up
Total	101 (11.0)	209 (12.8)	0.08
All‐cause death	13 (1.4)	15 (0.9)	0.60
Nonfatal MI	10 (1.1)	22 (1.4)	0.31
Stoke	16 (1.7)	18 (1.1)	0.42
Ischemia‐driven revascularization	85 (9.8)	192 (13.1)	0.02
Progression of coronary atherosclerosis	55 (6.3)	145 (9.9)	0.004

Open in a new tab

Values are mean±SD or number (percentage). P values were estimated by the t test for continuous variables and χ² test for categorical variables. A total of 16 participants with missing BMI values are not shown in this table. Additionally, 2 cases of MACEs with missing BMI values are not shown in this table. BMI indicates body mass index; eGFR, estimated glomerular filtration rate; HDL‐C, high‐density lipoprotein cholesterol; LDL‐C, low‐density lipoprotein cholesterol; MACEs, major adverse cardiovascular events; and MI, myocardial infarction.

Different Roles of Lipid Components in Coronary Plaque Burden and Instability

The overall metabolite profile was significantly associated with the Gensini score and cTnT (both permutational multivariate ANOVA P < 0.001; Figure S3). The first principal coordinate loading score was significantly different across the levels of Gensini score and cTnT (both P < 0.001; Figure S3). For the serum level of overall lipid components (Figure 1), the overall circulating concentrations of apolipoprotein A1 and apolipoprotein A2 were inversely associated with, while those of apolipoprotein B, CE, cholesterol, and free cholesterol were positively associated with both the coronary plaque burden (indicated by Gensini score) and instability (indicated by cTnT level). Besides, the serum levels of total phospholipid and triglyceride displayed positive associations with cTnT level but no association with Gensini score.

The dots and error bars are β coefficients and corresponding 95% CIs for the associations of lipid components (rank‐based inverse normal transformed) with Gensini score and circulating cTnT level (log transformed) estimated from multivariable linear regression models with adjustment for age, sex, smoking status, body mass index, estimated glomerular filtration rate, medication use for dyslipidemia, and disease status of hypertension and diabetes. AB indicates apolipoprotein B; CAD, coronary atherosclerotic disease; CE, cholesterol ester; CH, cholesterol; cTnT, cardiac troponin T; FC, free cholesterol; PL, phospholipids; PN, particle number per liter; TG, triglycerides; and nH0CH, cholesterol in non‐HDL.

The particle number per liter serum of IDL and LDL was positively associated with Gensini score and cTnT level. For VLDL, the concentrations of its components (ie, apolipoprotein B, CEs, cholesterol, free cholesterol, phospholipid, and triglyceride) were positively associated with cTnT levels (β ranging from 0.09 for triglyceride to 0.19 for CEs; FDR<0.05), but not associated with Gensini score. For IDL, the circulating concentrations of its internal components were positively associated with both plaque severity indicators, except for its triglyceride with Gensini score. Likewise, the patterns in associations with Gensini score and cTnT level of all the components within LDL were similar to, though stronger than, those within IDL (Figure 1). Among them, the circulating concentrations of triglyceride and apolipoprotein B within LDL displayed stronger associations with Gensini score and cTnT compared with the LDL‐C concentration, which is the most commonly measured parameter in clinical practice (associations with Gensini score: triglyceride within LDL, β=6.70; apolipoprotein B within LDL, β=5.48; and LDL‐C concentration, β=4.23; associations with cTnT: triglyceride within LDL, β=0.43; apolipoprotein B within LDL, β=0.37; and LDL‐cholesterol concentration, β=0.30).

For HDL particles, its cholesterol, CEs, triglyceride, apolipoprotein A1, and apolipoprotein A2 circulating concentrations were inversely associated with both plaque severity indicators, while free cholesterol concentrations within HDL were not significantly associated with Gensini score or cTnT levels. Phospholipids concentration of HDL was inversely associated with Gensini score but not associated with cTnT (Figure 1). Of note, there were stronger inverse associations of triglyceride, apolipoprotein A1 and apolipoprotein A2 concentrations within HDL, compared with cholesterol concentrations within HDL, that is, HDL‐C in clinical practice, with both Gensini score and cTnT level.

Different Lipid Subfraction Signatures of Coronary Plaque Burden and Instability

We further investigated the associations of different lipoprotein subfractions (Figure 2), which were classified by densities, with plaque burden and instability. It is worth mentioning that none of the VLDL subfractions showed significant associations with Gensini score, while nearly all the components of VLDL were consistently and positively associated with cTnT levels, with CEs within VLDL particles showing the strongest associations.

The dots and error bars are β coefficients and corresponding 95% CIs for the associations of lipid subfractions (rank‐based inverse normal transformed) with Gensini score and circulating cardiac troponin T level (log transformed) estimated from multivariable linear regression models with adjustment for age, sex, smoking status, body mass index, estimated glomerular filtration rate, medication use for dyslipidemia, and disease status of hypertension and diabetes. AB indicates apolipoprotein B; CAD, coronary atherosclerotic disease; CE, cholesterol ester; CH, cholesterol; cTnT, cardiac troponin T; FC, free cholesterol; HDL, high‐density lipoprotein; LDL, low‐density lipoprotein; PL, phospholipids; PN, particle number per liter; TG, triglycerides; and VLDL, very‐low‐density lipoprotein.

Concentrations of all the components within measured LDL subfractions were positively associated with both the Gensini score and cTnT level. Among them, the components within LDL1 and LDL3 particles showed the strongest associations with both the Gensini score and cTnT level (Figure 2). As for HDL subfractions, almost all components (especially apolipoprotein A1) within HDL2, HDL3, HDL4, but not HDL1, were inversely associated with Gensini score. Contrarily, concentrations of CEs, cholesterol, and phospholipids within HDL1 showed significant and positive associations with Gensini score (FDR <0.05). Notably, triglyceride concentrations within each HDL subfraction were inversely associated with Gensini score. The pattern in association with cTnT level of components within HDL subfractions was similar to the association with Gensini score (Figure 2).

Associations of Amino Acids and Other Metabolites With Plaque Burden and Instability

In general, while only a few measured amino acids were significantly associated with Gensini score, more amino acids showed associations with cTnT levels (FDR <0.05) (Figure 3). For example, the serum levels of glycine, histidine, isoleucine, lysine, ornithine, threonine, and valine were inversely associated, while that of phenylalanine was positively associated with cTnT levels. Among the ketone bodies, 3‐hydroxybutyric acid levels were positively associated with both the Gensini score (β=3.03) and cTnT level (β=0.17; both FDR <0.05). Both acetone and acetoacetic acid exhibited a strong association with cTnT level but no associations with Gensini score. The percentage of polyunsaturated fatty acid in total fatty acid was inversely associated with both Gensini score and cTnT level, and that of monounsaturated fatty acid in total fatty acid was positively associated with cTnT level only. It is noteworthy that the serum levels of NAG1 (N‐acetylglucosaminyl/N‐acetylgalactosaminyl‐glycoproteins), NAG2 (N‐acetylneuraminyl‐glycoproteins) and their ratio (the NAG1 to NAG2 ratio), which are inflammatory biomarkers, were positively associated with cTnT level but not associated with Gensini score (Figure 3).

The dots and error bars are β coefficients and corresponding 95% CIs for the associations of amino acids and other metabolites (rank‐based inverse normal transformed) with Gensini score and circulating cTnT level (log transformed) estimated from multivariable linear regression models with adjustment for age, sex, smoking status, body mass index, estimated glomerular filtration rate, medication use for dyslipidemia, and disease status of hypertension and diabetes. AAA indicates aromatic amino acids (phenylalanine+tyrosine+tryptophan); BCAA, branched‐chain amino acids (leucine+valine+isoleucine); CAD, coronary atherosclerotic disease; cTnT, cardiac troponin T; MUFAp, the proportion of monounsaturated fatty acid in total fatty acid; NAG1, N‐acetylglucosaminyl/N‐acetylgalactosaminyl‐glycoproteins; NAG2, N‐acetylneuraminyl‐glycoproteins; PUFAp, the proportion of polyunsaturated fatty acid in total fatty acid; SFAp, the proportion of saturated fatty acid in total fatty acid; and UFAp, the proportion of unsaturated fatty acid in total fatty acid.

Secondary Analyses of the Associations Between Metabolites and Severity of Atherosclerosis

Compared with that in older patients, there were stronger positive associations of the serum lipid component levels with cTnT level in younger patients (all P‐interaction <0.05; Figure S4), and the associations of lipid components with Gensini score were in general consistent across age groups (Figure S4). Other metabolites, such as amino acids, fatty acids, and inflammation markers, also exhibited stronger associations with cTnT levels in younger patients compared with that in older patients (P‐interaction <0.05; Figure S5). Remarkably, NAG1 and NAG2 displayed stronger associations with cTnT level but weaker associations with Gensini score in younger patients. Associations of lipid components and metabolites with coronary atherosclerosis were generally consistent among patients with different BMI statuses (Figures S6 and S7). Sensitivity analyses, excluding participants with eGFR <60 mL/min per 1.732 m² or those with diabetes, yielded consistent association results for both the Gensini score (Figure S8A) and cTnT levels (Figure S8B).

The association patterns of metabolites with plaque burden, as indicated by the Gensini score, were generally consistent with those observed for plaque instability, as indicated by cTnT levels (Figure S9). Especially, several metabolites (eg, triglyceride within LDL1, cholesterol, and phospholipid within LDL3, NAG1) displayed stronger positive associations with cTnT compared with that with Gensini score, while the percentage of polyunsaturated fatty acid in total fatty acid showed stronger inverse associations with cTnT level compared with those with Gensini score. Cholesterol esters, cholesterol, and triglyceride within VLDL5 displayed significantly positive associations with cTnT level as well as inverse associations with Gensini score (Figure S9).

Causal Effects of Circulating Metabolites on Plaque Burden and Instability

The summary information of IVs for each metabolite is present in Table S2. The F statistics for all IVs used in the present study were >10. The genome‐wide association analysis shows that the rs72665765 variants near RP11‐333A23.4 were associated with cTnT at genome‐wide significance (P=1.74×10⁻⁸; Figure S10 and Table S3). The rs72665765 was in high linkage disequilibrium with rs28581409 (R ²=0.753 in East Asian samples according to the 1000 Genome Project Phase 3), which was reported in previous GWASs for cTnT. ³⁸ The loci of rs58721068 (EDNRA; P=1.04×10⁻⁹), rs9349379 (PHACTR1; P=1.49×10⁻⁸), rs10811654 (CDKN2B‐AS1; P=5.22×10⁻¹¹) and rs1808757 (CTSH; P=1.39×10⁻¹¹) were associated with Gensini score at genome‐wide significance (Figure S10 and Table S3).

We found several potential causal effects of circulating metabolites on Gensini score (FDR <0.05; Table S4 and Figure 4). Genetically predicted higher level of circulating apolipoprotein B (β=0.23; SE, 0.04; FDR, 6.17×10⁻⁶), apolipoprotein B particle number per liter (0.23±0.04; FDR, 6.17×10⁻⁶), CE (0.23±0.06; FDR, 5.88×10⁻⁴), cholesterol (0.30±0.06; FDR, 2.88×10⁻⁵), free cholesterol (0.40±0.07; FDR, 6.17×10⁻⁶) and cholesterol in non‐HDL (0.31±0.06; FDR, 2.02×10⁻⁵) were associated with higher Gensini score. These MR results were consistent with observational findings, while the genetically predicted higher apolipoprotein A1 and apolipoprotein A2 were not associated with the Gensini score. Furthermore, genetically predicted higher levels of LDL components were suggestively associated with higher Gensini score (β ranging from 0.13 to 0.16; P<0.05), except for its triglyceride. For specific lipid subfraction signatures, we found that genetically predicted higher levels of 6 components of LDL‐6, 4 components of LDL‐2, L3CE, and L4CE were associated with higher Gensini score (β ranging from 0.17 to 0.42; FDR <0.05). Besides, IDPL (0.17±0.06; FDR, 0.017) and V5CE (0.31±0.12; FDR, 0.047) were also causally associated with higher Gensini score. Meanwhile, genetically predicted higher levels of dihydrothymine (−0.39±0.14; FDR, 0.038) and alanine (−0.42±0.15; FDR, 0.047) were associated with lower Gensini score, while NAG1/NAG2 was associated with higher Gensini score (0.23±0.08; FDR, 0.047). The observed direction of other sensitivity analyses aligned with the inverse variance weighted results. However, since only 2 SNPs were included for 4 components of LDL‐2, dihydrothymine, and alanine, other MR methods' sensitivity analyses were not applicable. The MR‐Egger intercept test showed limited evidence of unbalanced pleiotropy (P>0.05 for intercepts). Additionally, we did not find evidence of circulating metabolites having a causal effect on the level of cTnT (all FDR >0.05; Figure S11). However, genetically predicted higher V5CE (β=0.27), H1A1 (β=0.08), and glucose (β=0.30) were suggestively associated with higher cTnT (all P<0.05).

The dots and error bars represent the beta coefficients and corresponding 95% CIs of the causal associations of metabolomics features with the Gensini score estimated by the inverse variance weighted Mendelian randomization method. AB indicates apolipoprotein B; CE, cholesterol ester; CH, cholesterol; FC, free cholesterol; HDL, high‐density lipoprotein; IDL, intermediate‐density lipoprotein; LDL, low‐density lipoprotein; PL, phospholipids; PN, particle number per liter; TG, triglycerides; and VLDL, very‐low‐density lipoprotein.

Metabolomic Indices and Risk of MACEs

A Gensini score–based metabolomic index was constructed on the basis of 33 selected metabolites. These metabolites predominantly encompass the components of triglycerides, apolipoprotein A1, CEs, and total lipids in lipoproteins. Additionally, it includes certain amino acids with established associations with coronary artery health, such as trimethylamine N‐oxide and glycine, along with glucose, and a few inflammation biomarkers including NAG (Table S5 and Figure S12A). Per‐SD increase in baseline Gensini score–based metabolomic index was associated with a 14.8% increase in the risk of MACEs during a median follow‐up of 3.8 years independent of age, sex, smoking, BMI, eGFR, medication use, and disease status (HR, 1.148 [95% CI, 1.018–1.295]; P=0.024; Figure 5A). The higher risk of MACEs associated with Gensini score–based metabolomic index was more pronounced among participants with elevated cTnT levels (P‐interaction=0.004; Table S6). However, the associations were consistent across patients of different ages, sex, and BMI (all P‐interactions ≥0.46, Table S6). No significant association was observed between cTnT‐based metabolomic index (Table S7 and Figure S12B) and MACE risk (P>0.05; Figure 5B and Table S6).

Patients were right censored if they lost follow‐up or had no major adverse cardiac events at the time of data freeze (December 31, 2021). P value was estimated by log‐rank test. HRs were estimated from multivariable Cox proportional hazards regression models with adjustment for age, sex, smoking status, body mass index, estimated glomerular filtration rate, medication use for dyslipidemia, and disease status of hypertension and diabetes. CMI indicates cTnT‐based metabolomic index; and GMI, Gensini score‐based metabolomic index; and HR, hazard ratio.

Discussion

Among 2560 patients with newly diagnosed CAD, this study comprehensively examined the associations of residual risks of lipid components and subfractions, amino acids, and other metabolites with coronary plaque burden and instability. The results showed that the components within LDL and HDL and their subfractions had generally similar associations with plaque burden and instability, while the components within VLDL and its subfractions were positively associated with plaque instability but not associated with plaque burden. In addition, other metabolites, including amino acids, ketone bodies, fatty acids, and inflammation markers, demonstrated varying degrees of association with plaque burden and instability. Furthermore, MR analysis revealed that metabolites such as apolipoprotein B and dihydrothymine, showed a causal relationship with Gensini score.

Studies using experimental animal models have conclusively shown that LDL‐C is a key factor in the pathogenesis of atherosclerosis. ³⁹ This is further validated by human data, such as the significantly elevated risk of premature atherosclerotic cardiovascular disease observed in individuals with familial hypercholesterolemia. ⁴⁰ In the present study, the concentrations of all the components within LDL subfractions showed consistently positive associations with both plaque burden (Gensini score) and instability (cTnT level). Our observation is consistent with the known effect of LDL‐C on the risk of coronary atherosclerosis and MI demonstrated by previous genetic and metabonomic studies. ⁸ , ⁴¹ , ⁴² , ⁴³ We identified inverse associations of HDL2/3/4 (referred to as medium and large HDL particles) and cholesterol concentrations within these lipoproteins with plaque burden and instability, consistent with previous metabonomic studies on HDL subfractions and the risk of coronary atherosclerosis and MI. ⁸ , ⁴³ However, almost all the components within HDL1 (referred to as small HDL particles) were positively associated with plaque burden and instability. This is in contrast with the consistent associations of the components within other HDL subfractions with plaque burden and instability, suggesting a specific role of HDL1 different from other HDL subfractions as well as opposing effects of triglyceride and other components within HDL1. Most investigations of HDL in CAD have used a composite measure of the cholesterol concentrations across HDL lipoproteins. The collective findings of the present and previous studies ⁶ , ⁴⁴ indicated that the inverse associations of HDL‐C may be limited to medium and large HDL particles and could partially explain the controversial data on the effect of HDL‐C. ⁴⁵ , ⁴⁶

We also identified several associations of fatty acids and amino acids with coronary plaque severity, which are in general in line with findings from previous studies in cardiovascular risk. For example, the inverse association of polyunsaturated fatty acid in total fatty acid with plaque burden and instability was in line with existing evidence that omega‐3 polyunsaturated fatty acids have antihypertriglyceridemia effects and mitigate inflammation, thereby reducing the risk of atheroslerosis. ⁴⁷ , ⁴⁸ In addition, the positive association of the inflammatory biomarker NAG2 with plaque instability, echoed the previous reports on NAG2 and atherosclerosis in patients with diabetes and hypercholesterolemia. ⁴⁹ , ⁵⁰

Previous cross‐sectional or prospective studies reported that both total and LDL cholesterol levels tended to increase with increasing age in young or middle‐aged adults ⁵¹ , ⁵² , ⁵³ , ⁵⁴ while decrease with advancing age in older participants. ⁵⁵ , ⁵⁶ , ⁵⁷ , ⁵⁸ HDL cholesterol levels did not vary by age in most cross‐sectional studies, ⁵¹ , ⁵⁷ , ⁵⁹ , ⁶⁰ while they decreased with increasing age in the majority of the prospective studies. ⁵³ , ⁵⁵ , ⁶¹ , ⁶² In our study, younger patients (men aged <55 years and women aged <65 years) displayed stronger positive associations with plaque instability with regard to the components within IDL/VLDL/LDL particles and weaker associations with plaque burden for the components within IDL/VLDL. The inverse associations of the components within HDL with plaque burden and instability were similar between the older and younger groups, except for free cholesterol and phospholipid concentration. These results suggested that the role of different lipid components in the associations with CAD may be dependent on age. Moreover, the Gensini‐based metabolomic index showed positive associations with MACE, providing a novel strategy for predicting prognosis using noninvasive metabolite measurements instead of invasive coronary angiography.

There are noteworthy limitations in the current study. First, our study could not infer the causal relationship between the plasma metabolites and plaque severity due to its observational nature. However, we have employed MR approaches to explore the potential causal effects. In addition, we lacked replication studies to reproduce our novel findings. Furthermore, other residual confounding factors, such as diet, may exist. Additionally, the use of lipid‐lowering medications before enrollment might have influenced the assessment of different lipid components and the interpretation of the results of this study.

Metabonomic biomarkers continue to present residual risks in patients with CAD, even after statin therapy. Our study highlights a pronounced association of lipid subfractions and components, as profiled by NMR spectroscopy, with coronary plaque burden and instability, noting varied characteristics across different age cohorts. These findings suggest that noninvasive metabolite profiling could herald new prognostic approaches for CAD, offering a potential alternative to traditional, invasive coronary angiography.

Sources of Funding

This work was supported by grants from the National Key R&D Program of China (2021YFC2500500, 2022YFC3400700), Shanghai Clinical Research Center for Interventional Medicine (19MC1910300), Clinical Research Plan of Shanghai Hospital Development Center (SHDC2020CR1007A), the National Natural Science Foundation of China (81973032), and Shanghai Municipal Science and Technology Major Project (Grant No. 2017SHZDZX01). Dr Dai was supported by the Program of Shanghai Academic Research Leader (22XD1423300). Dr Mei was supported by the fellowship of the China Postdoctoral Science Foundation (2022M710786). Dr Xu was supported by National Natural Science Foundation of China (82100466).

Disclosures

None.

Supporting information

Data S1

JAH3-13-e036906-s002.pdf^{(3.8MB, pdf)}

Tables S1–S7

Figures S1–S12

JAH3-13-e036906-s001.xlsx^{(418.1KB, xlsx)}

Acknowledgments

The authors thank all the participants and the research staff of the 38 recruiting centers for their valuable contributions. Drs Dai and Zheng designed the study. Drs Mei, Huang, and Lin analyzed and interpreted the data. Drs Dai, Mei, Xu, Huang, Lin, and Zheng drafted the manuscript. Drs R. Wu, H. Wu, Lu, Shali, Wang, Luo, and Z. Sun contributed to the acquisition of data. Drs Yu, L. Sun, and Chen provided suggestions on the study and data interpretation. Drs Ge, Dai, Zheng, Yao, Qian, and Tang supervised the entire study. All authors read and approved the final manuscript.

This manuscript was sent to Jacquelyn Y. Taylor, PhD, PNP‐BC, RN, FAHA, FAAN, Associate Editor, for review by expert referees, editorial decision, and final disposition.

Supplemental Material is available at https://www.ahajournals.org/doi/suppl/10.1161/JAHA.124.036906

For Sources of Funding and Disclosures, see page 14.

Dr. Chenhao Lin is now also with the College of Food Science, Fujian Agriculture and Forestry University, Fuzhou, Fujian, China.

Contributor Information

Kang Yao, Email: yao.kang@zs-hospital.sh.cn.

Yan Zheng, Email: yan_zheng@fudan.edu.cn.

Yuxiang Dai, Email: dai.yuxiang@zs-hospital.sh.cn.

References

1. Ezzati M, Obermeyer Z, Tzoulaki I, Mayosi BM, Elliott P, Leon DA. Contributions of risk factors and medical care to cardiovascular mortality trends. Nat Rev Cardiol. 2015;12:508–530. doi: 10.1038/nrcardio.2015.82 [DOI] [PMC free article] [PubMed] [Google Scholar]
2. Tzoulaki I, Elliott P, Kontis V, Ezzati M. Worldwide exposures to cardiovascular risk factors and associated health effects: current knowledge and data gaps. Circulation. 2016;133:2314–2333. doi: 10.1161/CIRCULATIONAHA.115.008718 [DOI] [PubMed] [Google Scholar]
3. Timmis A, Townsend N, Gale C, Grobbee R, Maniadakis N, Flather M, Wilkins E, Wright L, Vos R, Bax J, et al. European Society of Cardiology: cardiovascular disease statistics 2017. Eur Heart J. 2018;39:508–579. doi: 10.1093/eurheartj/ehx628 [DOI] [PubMed] [Google Scholar]
4. Knuuti J, Wijns W, Saraste A, Capodanno D, Barbato E, Funck‐Brentano C, Prescott E, Storey RF, Deaton C, Cuisset T, et al. 2019 ESC Guidelines for the diagnosis and management of chronic coronary syndromes. Eur Heart J. 2020;41:407–477. doi: 10.1093/eurheartj/ehz425 [DOI] [PubMed] [Google Scholar]
5. Ridker PM. Residual inflammatory risk: addressing the obverse side of the atherosclerosis prevention coin. Eur Heart J. 2016;37:1720–1722. doi: 10.1093/eurheartj/ehw024 [DOI] [PubMed] [Google Scholar]
6. Wurtz P, Havulinna AS, Soininen P, Tynkkynen T, Prieto‐Merino D, Tillin T, Ghorbani A, Artati A, Wang Q, Tiainen M, et al. Metabolite profiling and cardiovascular event risk: a prospective study of 3 population‐based cohorts. Circulation. 2015;131:774–785. doi: 10.1161/CIRCULATIONAHA.114.013116 [DOI] [PMC free article] [PubMed] [Google Scholar]
7. Soininen P, Kangas AJ, Wurtz P, Suna T, Ala‐Korpela M. Quantitative serum nuclear magnetic resonance metabolomics in cardiovascular epidemiology and genetics. Circ Cardiovasc Genet. 2015;8:192–206. doi: 10.1161/CIRCGENETICS.114.000216 [DOI] [PubMed] [Google Scholar]
8. Tzoulaki I, Castagne R, Boulange CL, Karaman I, Chekmeneva E, Evangelou E, Ebbels TMD, Kaluarachchi MR, Chadeau‐Hyam M, Mosen D, et al. Serum metabolic signatures of coronary and carotid atherosclerosis and subsequent cardiovascular disease. Eur Heart J. 2019;40:2883–2896. doi: 10.1093/eurheartj/ehz235 [DOI] [PMC free article] [PubMed] [Google Scholar]
9. Talmor‐Barkan Y, Bar N, Shaul AA, Shahaf N, Godneva A, Bussi Y, Lotan‐Pompan M, Weinberger A, Shechter A, Chezar‐Azerrad C, et al. Metabolomic and microbiome profiling reveals personalized risk factors for coronary artery disease. Nat Med. 2022;28:295–302. doi: 10.1038/s41591-022-01686-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
10. Fromentin S, Forslund SK, Chechi K, Aron‐Wisnewsky J, Chakaroun R, Nielsen T, Tremaroli V, Ji B, Prifti E, Myridakis A, et al. Microbiome and metabolome features of the cardiometabolic disease spectrum. Nat Med. 2022;28:303–314. doi: 10.1038/s41591-022-01688-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
11. Shali S, Dai Y, Hongyi W, Qian J, Zheng Y, Yao K, Ge J. Clinical and genetic characteristics of coronary artery disease in Chinese young adults: rationale and design of the prospective Genetic characteristics of coRonary Artery disease in ChiNese young aDults (GRAND) study. Cardiol Plus. 2021;6:65–72. doi: 10.4103/2470-7511.312594 [DOI] [Google Scholar]
12. Gensini GG. A more meaningful scoring system for determining the severity of coronary heart disease. Am J Cardiol. 1983;51:606. doi: 10.1016/s0002-9149(83)80105-2 [DOI] [PubMed] [Google Scholar]
13. Cui G, Li Z, Li R, Huang J, Wang H, Zhang L, Ding H, Wang DW. A functional variant in APOA5/A4/C3/A1 gene cluster contributes to elevated triglycerides and severity of CAD by interfering with microRNA 3201 binding efficiency. J Am Coll Cardiol. 2014;64:267–277. doi: 10.1016/j.jacc.2014.03.050 [DOI] [PubMed] [Google Scholar]
14. Dandona S, Stewart AF, Chen L, Williams K, So D, O'Brien E, Glover C, Lemay M, Assogba O, Vo L, et al. Gene dosage of the common variant 9p21 predicts severity of coronary artery disease. J Am Coll Cardiol. 2010;56:479–486. doi: 10.1016/j.jacc.2009.10.092 [DOI] [PubMed] [Google Scholar]
15. Thygesen K, Alpert JS, Jaffe AS, Chaitman BR, Bax JJ, Morrow DA, White HD; Executive Group on behalf of the Joint European Society of Cardiology/American College of Cardiology/American Heart Association/World Heart Federation Task Force for the Universal Definition of Myocardial I . Fourth universal definition of myocardial infarction (2018). Circulation. 2018;138:e618–e651. doi: 10.1161/CIR.0000000000000617 [DOI] [PubMed] [Google Scholar]
16. Jimenez B, Holmes E, Heude C, Tolson RF, Harvey N, Lodge SL, Chetwynd AJ, Cannet C, Fang F, Pearce JTM, et al. Quantitative lipoprotein subclass and low molecular weight metabolite analysis in human serum and plasma by (1)H NMR spectroscopy in a multilaboratory trial. Anal Chem. 2018;90:11962–11971. doi: 10.1021/acs.analchem.8b02412 [DOI] [PubMed] [Google Scholar]
17. Magnani HN, Howard AN. A quantitative method for blood lipoproteins using cellulose acetate electrophoresis. J Clin Pathol. 1971;24:837–845. doi: 10.1136/jcp.24.9.837 [DOI] [PMC free article] [PubMed] [Google Scholar]
18. Xia M, Ma S, Huang Q, Zeng H, Ge J, Xu W, Wu Q, Wu L, Li X, Ma H, et al. NAFLD‐related gene polymorphisms and all‐cause and cause‐specific mortality in an Asian population: the Shanghai Changfeng Study. Aliment Pharmacol Ther. 2022;55:705–721. doi: 10.1111/apt.16772 [DOI] [PubMed] [Google Scholar]
19. Chen YM, Zheng Y, Yu Y, Wang Y, Huang Q, Qian F, Sun L, Song ZG, Chen Z, Feng J, et al. Blood molecular markers associated with COVID‐19 immunopathology and multi‐organ damage. EMBO J. 2020;39:e105896. doi: 10.15252/embj.2020105896 [DOI] [PMC free article] [PubMed] [Google Scholar]
20. Xu W, Wu J, An Y, Xiao C, Hao F, Liu H, Wang Y, Tang H. Streptozotocin‐induced dynamic metabonomic changes in rat biofluids. J Proteome Res. 2012;11:3423–3435. doi: 10.1021/pr300280t [DOI] [PubMed] [Google Scholar]
21. Liu M, Tang H, Nicholson JK, Lindon JC. Use of 1H NMR‐determined diffusion coefficients to characterize lipoprotein fractions in human blood plasma. Magn Reson Chem. 2002;40:S83–S88. doi: 10.1002/mrc.1121 [DOI] [Google Scholar]
22. Wu Q, Huang QX, Zeng HL, Ma S, Lin HD, Xia MF, Tang HR, Gao X. Prediction of metabolic disorders using NMR‐based metabolomics: the Shanghai Changfeng study. Phenomics. 2021;1:186–198. doi: 10.1007/s43657-021-00021-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
23. Zhu X, Huang Q, Ma S, Chen L, Wu Q, Wu L, Ma H, Li X, Li Q, Aleteng Q, et al. Presence of sarcopenia identifies a special group of lean NAFLD in middle‐aged and older people. Hepatol Int. 2023;17:313–325. doi: 10.1007/s12072-022-10439-z [DOI] [PubMed] [Google Scholar]
24. van Buuren S, Groothuis‐Oudshoorn K. mice: multivariate imputation by chained equations in R. J Stat Softw. 2011;45:1–67. doi: 10.18637/jss.v045.i03 [DOI] [Google Scholar]
25. Arnett DK, Blumenthal RS, Albert MA, Buroker AB, Goldberger ZD, Hahn EJ, Himmelfarb CD, Khera A, Lloyd‐Jones D, McEvoy JW, et al. 2019 ACC/AHA guideline on the primary prevention of cardiovascular disease: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Circulation. 2019;140:e596–e646. doi: 10.1161/CIR.0000000000000678 [DOI] [PMC free article] [PubMed] [Google Scholar]
26. Friedman J, Hastie T, Tibshirani R. Regularization paths for generalized linear models via coordinate descent. J Stat Softw. 2010;33:1–22. doi: 10.18637/jss.v033.i01 [DOI] [PMC free article] [PubMed] [Google Scholar]
27. Gao X, Hofman A, Hu Y, Lin H, Zhu C, Jeekel J, Jin X, Wang J, Gao J, Yin Y, et al. The Shanghai Changfeng study: a community‐based prospective cohort study of chronic diseases among middle‐aged and elderly: objectives and design. Eur J Epidemiol. 2010;25:885–893. doi: 10.1007/s10654-010-9525-6 [DOI] [PubMed] [Google Scholar]
28. Liu Z, Wang Y, Zhang Y, Chu X, Wang Z, Qian D, Chen F, Xu J, Li S, Jin L, et al. Cohort profile: the Rugao longevity and ageing study (RuLAS). Int J Epidemiol. 2016;45:1064–1073. doi: 10.1093/ije/dyv101 [DOI] [PubMed] [Google Scholar]
29. Burgess S, Thompson SG; Collaboration CCG . Avoiding bias from weak instruments in Mendelian randomization studies. Int J Epidemiol. 2011;40:755–764. doi: 10.1093/ije/dyr036 [DOI] [PubMed] [Google Scholar]
30. Chang CC, Chow CC, Tellier LC, Vattikuti S, Purcell SM, Lee JJ. Second‐generation PLINK: rising to the challenge of larger and richer datasets. Gigascience. 2015;4:7. doi: 10.1186/s13742-015-0047-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
31. Burgess S, Butterworth A, Thompson SG. Mendelian randomization analysis with multiple genetic variants using summarized data. Genet Epidemiol. 2013;37:658–665. doi: 10.1002/gepi.21758 [DOI] [PMC free article] [PubMed] [Google Scholar]
32. Bowden J, Davey Smith G, Burgess S. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol. 2015;44:512–525. doi: 10.1093/ije/dyv080 [DOI] [PMC free article] [PubMed] [Google Scholar]
33. Hartwig FP, Davey Smith G, Bowden J. Robust inference in summary data Mendelian randomization via the zero modal pleiotropy assumption. Int J Epidemiol. 2017;46:1985–1998. doi: 10.1093/ije/dyx102 [DOI] [PMC free article] [PubMed] [Google Scholar]
34. Bowden J, Davey Smith G, Haycock PC, Burgess S. Consistent estimation in Mendelian randomization with some invalid instruments using a weighted median estimator. Genet Epidemiol. 2016;40:304–314. doi: 10.1002/gepi.21965 [DOI] [PMC free article] [PubMed] [Google Scholar]
35. Burgess S, Thompson SG. Interpreting findings from Mendelian randomization using the MR‐Egger method. Eur J Epidemiol. 2017;32:377–389. doi: 10.1007/s10654-017-0255-x [DOI] [PMC free article] [PubMed] [Google Scholar]
36. Hemani G, Zheng J, Elsworth B, Wade KH, Haberland V, Baird D, Laurin C, Burgess S, Bowden J, Langdon R, et al. The MR‐base platform supports systematic causal inference across the human phenome. elife. 2018;7:e34408. doi: 10.7554/eLife.34408 [DOI] [PMC free article] [PubMed] [Google Scholar]
37. Skrivankova VW, Richmond RC, Woolf BAR, Davies NM, Swanson SA, VanderWeele TJ, Timpson NJ, Higgins JPT, Dimou N, Langenberg C, et al. Strengthening the reporting of observational studies in epidemiology using mendelian randomisation (STROBE‐MR): explanation and elaboration. BMJ. 2021;375:n2233. doi: 10.1136/bmj.n2233 [DOI] [PMC free article] [PubMed] [Google Scholar]
38. Yang Y, Bartz TM, Brown MR, Guo X, Zilhao NR, Trompet S, Weiss S, Yao J, Brody JA, Defilippi CR, et al. Identification of functional genetic determinants of cardiac troponin T and I in a multiethnic population and causal associations with atrial fibrillation. Circ Genom Precis Med. 2021;14:e003460. doi: 10.1161/CIRCGEN.121.003460 [DOI] [PMC free article] [PubMed] [Google Scholar]
39. Bentzon JF, Otsuka F, Virmani R, Falk E. Mechanisms of plaque formation and rupture. Circ Res. 2014;114:1852–1866. doi: 10.1161/circresaha.114.302721 [DOI] [PubMed] [Google Scholar]
40. Tsao CW, Preis SR, Peloso GM, Hwang SJ, Kathiresan S, Fox CS, Cupples LA, Hoffmann U, O'Donnell CJ. Relations of long‐term and contemporary lipid levels and lipid genetic risk scores with coronary artery calcium in the framingham heart study. J Am Coll Cardiol. 2012;60:2364–2371. doi: 10.1016/j.jacc.2012.09.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
41. White J, Swerdlow DI, Preiss D, Fairhurst‐Hunter Z, Keating BJ, Asselbergs FW, Sattar N, Humphries SE, Hingorani AD, Holmes MV. Association of lipid fractions with risks for coronary artery disease and diabetes. JAMA Cardiol. 2016;1:692–699. doi: 10.1001/jamacardio.2016.1884 [DOI] [PMC free article] [PubMed] [Google Scholar]
42. Holmes MV, Asselbergs FW, Palmer TM, Drenos F, Lanktree MB, Nelson CP, Dale CE, Padmanabhan S, Finan C, Swerdlow DI, et al. Mendelian randomization of blood lipids for coronary heart disease. Eur Heart J. 2015;36:539–550. doi: 10.1093/eurheartj/eht571 [DOI] [PMC free article] [PubMed] [Google Scholar]
43. Holmes MV, Millwood IY, Kartsonaki C, Hill MR, Bennett DA, Boxall R, Guo Y, Xu X, Bian Z, Hu R, et al. Lipids, lipoproteins, and metabolites and risk of myocardial infarction and stroke. J Am Coll Cardiol. 2018;71:620–632. doi: 10.1016/j.jacc.2017.12.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
44. Inouye M, Kettunen J, Soininen P, Silander K, Ripatti S, Kumpula LS, Hamalainen E, Jousilahti P, Kangas AJ, Mannisto S, et al. Metabonomic, transcriptomic, and genomic variation of a population cohort. Mol Syst Biol. 2010;6:441. doi: 10.1038/msb.2010.93 [DOI] [PMC free article] [PubMed] [Google Scholar]
45. Kumar A, Sathian B. Correlation between lipid profile and troponin I test results in patients with chest pain in Nepal. Asian Pac J Trop Biomed. 2013;3:487–491. doi: 10.1016/S2221-1691(13)60101-X [DOI] [PMC free article] [PubMed] [Google Scholar]
46. Li XL, Guo YL, Zhu CG, Xu RX, Qing P, Wu NQ, Jiang LX, Xu B, Gao RL, Li JJ. Relationship of high‐density lipoprotein cholesterol with periprocedural myocardial injury following elective percutaneous coronary intervention in patients with low‐density lipoprotein cholesterol below 70 mg/dL. J Am Heart Assoc. 2015;4:e001412. doi: 10.1161/JAHA.114.001412 [DOI] [PMC free article] [PubMed] [Google Scholar]
47. Liu QK. Triglyceride‐lowering and anti‐inflammatory mechanisms of omega‐3 polyunsaturated fatty acids for atherosclerotic cardiovascular risk reduction. J Clin Lipidol. 2021;15:556–568. doi: 10.1016/j.jacl.2021.05.007 [DOI] [PubMed] [Google Scholar]
48. Simonetto M, Infante M, Sacco RL, Rundek T, Della‐Morte D. A novel anti‐inflammatory role of omega‐3 PUFAs in prevention and treatment of atherosclerosis and vascular cognitive impairment and dementia. Nutrients. 2019;11:2279. doi: 10.3390/nu11102279 [DOI] [PMC free article] [PubMed] [Google Scholar]
49. Seres‐Noriega T, Gimenez M, Perea V, Blanco J, Vinagre I, Pane A, Ruiz S, Cofan M, Mesa A, Esmatjes E, et al. Quantification of glycoproteins by nuclear magnetic resonance associated with preclinical carotid atherosclerosis in patients with type 1 diabetes. Nutr Metab Cardiovasc Dis. 2021;31:2099–2108. doi: 10.1016/j.numecd.2021.03.021 [DOI] [PubMed] [Google Scholar]
50. Malo AI, Girona J, Ibarretxe D, Rodriguez‐Borjabad C, Amigo N, Plana N, Masana L. Serum glycoproteins A and B assessed by (1)H‐NMR in familial hypercholesterolemia. Atherosclerosis. 2021;330:1–7. doi: 10.1016/j.atherosclerosis.2021.06.905 [DOI] [PubMed] [Google Scholar]
51. Heitmann BL. The effects of gender and age on associations between blood lipid levels and obesity in Danish men and women aged 35–65 years. J Clin Epidemiol. 1992;45:693–702. doi: 10.1016/0895-4356(92)90046-p [DOI] [PubMed] [Google Scholar]
52. Schaefer EJ, Lamon‐Fava S, Cohn SD, Schaefer MM, Ordovas JM, Castelli WP, Wilson PW. Effects of age, gender, and menopausal status on plasma low density lipoprotein cholesterol and apolipoprotein B levels in the Framingham Offspring Study. J Lipid Res. 1994;35:779–792. doi: 10.1016/S0022-2275(20)39173-2 [DOI] [PubMed] [Google Scholar]
53. Anderson KM, Wilson PW, Garrison RJ, Castelli WP. Longitudinal and secular trends in lipoprotein cholesterol measurements in a general population sample. The Framingham Offspring Study. Atherosclerosis. 1987;68:59–66. doi: 10.1016/0021-9150(87)90094-3 [DOI] [PubMed] [Google Scholar]
54. Wilson PW, Anderson KM, Harris T, Kannel WB, Castelli WP. Determinants of change in total cholesterol and HDL‐C with age: the Framingham study. J Gerontol. 1994;49:M252–M257. doi: 10.1093/geronj/49.6.m252 [DOI] [PubMed] [Google Scholar]
55. Ferrara A, Barrett‐Connor E, Shan J. Total, LDL, and HDL cholesterol decrease with age in older men and women. The Rancho Bernardo Study 1984–1994. Circulation. 1997;96:37–43. doi: 10.1161/01.cir.96.1.37 [DOI] [PubMed] [Google Scholar]
56. Laurenzi M, Mancini M. Plasma lipids in elderly men and women. Eur Heart J. 1988;9:69–74. doi: 10.1093/eurheartj/9.suppl_d.69 [DOI] [PubMed] [Google Scholar]
57. Ettinger WH, Wahl PW, Kuller LH, Bush TL, Tracy RP, Manolio TA, Borhani NO, Wong ND, O'Leary DH. Lipoprotein lipids in older people. Results from the Cardiovascular Health Study. The CHS Collaborative Research Group. Circulation. 1992;86:858–869. doi: 10.1161/01.cir.86.3.858 [DOI] [PubMed] [Google Scholar]
58. Frishman WH, Ooi WL, Derman MP, Eder HA, Gidez LI, Ben‐Zeev D, Zimetbaum P, Heiman M, Aronson M. Serum lipids and lipoproteins in advanced age. Intraindividual changes. Ann Epidemiol. 1992;2:43–50. doi: 10.1016/1047-2797(92)90036-p [DOI] [PubMed] [Google Scholar]
59. Heiss G, Tamir I, Davis CE, Tyroler HA, Rifkand BM, Schonfeld G, Jacobs D, Frantz ID Jr. Lipoprotein‐cholesterol distributions in selected North American populations: the lipid research clinics program prevalence study. Circulation. 1980;61:302–315. doi: 10.1161/01.cir.61.2.302 [DOI] [PubMed] [Google Scholar]
60. Kromhout D, Nissinen A, Menotti A, Bloemberg B, Pekkanen J, Giampaoli S. Total and HDL cholesterol and their correlates in elderly men in Finland, Italy, and the Netherlands. Am J Epidemiol. 1990;131:855–863. doi: 10.1093/oxfordjournals.aje.a115576 [DOI] [PubMed] [Google Scholar]
61. Hubert HB, Eaker ED, Garrison RJ, Castelli WP. Life‐style correlates of risk factor change in young adults: an eight‐year study of coronary heart disease risk factors in the Framingham offspring. Am J Epidemiol. 1987;125:812–831. doi: 10.1093/oxfordjournals.aje.a114598 [DOI] [PubMed] [Google Scholar]
62. Garry PJ, Hunt WC, Koehler KM, VanderJagt DJ, Vellas BJ. Longitudinal study of dietary intakes and plasma lipids in healthy elderly men and women. Am J Clin Nutr. 1992;55:682–688. doi: 10.1093/ajcn/55.3.682 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Data S1

JAH3-13-e036906-s002.pdf^{(3.8MB, pdf)}

Tables S1–S7

Figures S1–S12

JAH3-13-e036906-s001.xlsx^{(418.1KB, xlsx)}

[jah310328-bib-0001] 1. Ezzati M, Obermeyer Z, Tzoulaki I, Mayosi BM, Elliott P, Leon DA. Contributions of risk factors and medical care to cardiovascular mortality trends. Nat Rev Cardiol. 2015;12:508–530. doi: 10.1038/nrcardio.2015.82 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0002] 2. Tzoulaki I, Elliott P, Kontis V, Ezzati M. Worldwide exposures to cardiovascular risk factors and associated health effects: current knowledge and data gaps. Circulation. 2016;133:2314–2333. doi: 10.1161/CIRCULATIONAHA.115.008718 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0003] 3. Timmis A, Townsend N, Gale C, Grobbee R, Maniadakis N, Flather M, Wilkins E, Wright L, Vos R, Bax J, et al. European Society of Cardiology: cardiovascular disease statistics 2017. Eur Heart J. 2018;39:508–579. doi: 10.1093/eurheartj/ehx628 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0004] 4. Knuuti J, Wijns W, Saraste A, Capodanno D, Barbato E, Funck‐Brentano C, Prescott E, Storey RF, Deaton C, Cuisset T, et al. 2019 ESC Guidelines for the diagnosis and management of chronic coronary syndromes. Eur Heart J. 2020;41:407–477. doi: 10.1093/eurheartj/ehz425 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0005] 5. Ridker PM. Residual inflammatory risk: addressing the obverse side of the atherosclerosis prevention coin. Eur Heart J. 2016;37:1720–1722. doi: 10.1093/eurheartj/ehw024 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0006] 6. Wurtz P, Havulinna AS, Soininen P, Tynkkynen T, Prieto‐Merino D, Tillin T, Ghorbani A, Artati A, Wang Q, Tiainen M, et al. Metabolite profiling and cardiovascular event risk: a prospective study of 3 population‐based cohorts. Circulation. 2015;131:774–785. doi: 10.1161/CIRCULATIONAHA.114.013116 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0007] 7. Soininen P, Kangas AJ, Wurtz P, Suna T, Ala‐Korpela M. Quantitative serum nuclear magnetic resonance metabolomics in cardiovascular epidemiology and genetics. Circ Cardiovasc Genet. 2015;8:192–206. doi: 10.1161/CIRCGENETICS.114.000216 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0008] 8. Tzoulaki I, Castagne R, Boulange CL, Karaman I, Chekmeneva E, Evangelou E, Ebbels TMD, Kaluarachchi MR, Chadeau‐Hyam M, Mosen D, et al. Serum metabolic signatures of coronary and carotid atherosclerosis and subsequent cardiovascular disease. Eur Heart J. 2019;40:2883–2896. doi: 10.1093/eurheartj/ehz235 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0009] 9. Talmor‐Barkan Y, Bar N, Shaul AA, Shahaf N, Godneva A, Bussi Y, Lotan‐Pompan M, Weinberger A, Shechter A, Chezar‐Azerrad C, et al. Metabolomic and microbiome profiling reveals personalized risk factors for coronary artery disease. Nat Med. 2022;28:295–302. doi: 10.1038/s41591-022-01686-6 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0010] 10. Fromentin S, Forslund SK, Chechi K, Aron‐Wisnewsky J, Chakaroun R, Nielsen T, Tremaroli V, Ji B, Prifti E, Myridakis A, et al. Microbiome and metabolome features of the cardiometabolic disease spectrum. Nat Med. 2022;28:303–314. doi: 10.1038/s41591-022-01688-4 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0011] 11. Shali S, Dai Y, Hongyi W, Qian J, Zheng Y, Yao K, Ge J. Clinical and genetic characteristics of coronary artery disease in Chinese young adults: rationale and design of the prospective Genetic characteristics of coRonary Artery disease in ChiNese young aDults (GRAND) study. Cardiol Plus. 2021;6:65–72. doi: 10.4103/2470-7511.312594 [DOI] [Google Scholar]

[jah310328-bib-0012] 12. Gensini GG. A more meaningful scoring system for determining the severity of coronary heart disease. Am J Cardiol. 1983;51:606. doi: 10.1016/s0002-9149(83)80105-2 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0013] 13. Cui G, Li Z, Li R, Huang J, Wang H, Zhang L, Ding H, Wang DW. A functional variant in APOA5/A4/C3/A1 gene cluster contributes to elevated triglycerides and severity of CAD by interfering with microRNA 3201 binding efficiency. J Am Coll Cardiol. 2014;64:267–277. doi: 10.1016/j.jacc.2014.03.050 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0014] 14. Dandona S, Stewart AF, Chen L, Williams K, So D, O'Brien E, Glover C, Lemay M, Assogba O, Vo L, et al. Gene dosage of the common variant 9p21 predicts severity of coronary artery disease. J Am Coll Cardiol. 2010;56:479–486. doi: 10.1016/j.jacc.2009.10.092 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0015] 15. Thygesen K, Alpert JS, Jaffe AS, Chaitman BR, Bax JJ, Morrow DA, White HD; Executive Group on behalf of the Joint European Society of Cardiology/American College of Cardiology/American Heart Association/World Heart Federation Task Force for the Universal Definition of Myocardial I . Fourth universal definition of myocardial infarction (2018). Circulation. 2018;138:e618–e651. doi: 10.1161/CIR.0000000000000617 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0016] 16. Jimenez B, Holmes E, Heude C, Tolson RF, Harvey N, Lodge SL, Chetwynd AJ, Cannet C, Fang F, Pearce JTM, et al. Quantitative lipoprotein subclass and low molecular weight metabolite analysis in human serum and plasma by (1)H NMR spectroscopy in a multilaboratory trial. Anal Chem. 2018;90:11962–11971. doi: 10.1021/acs.analchem.8b02412 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0017] 17. Magnani HN, Howard AN. A quantitative method for blood lipoproteins using cellulose acetate electrophoresis. J Clin Pathol. 1971;24:837–845. doi: 10.1136/jcp.24.9.837 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0018] 18. Xia M, Ma S, Huang Q, Zeng H, Ge J, Xu W, Wu Q, Wu L, Li X, Ma H, et al. NAFLD‐related gene polymorphisms and all‐cause and cause‐specific mortality in an Asian population: the Shanghai Changfeng Study. Aliment Pharmacol Ther. 2022;55:705–721. doi: 10.1111/apt.16772 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0019] 19. Chen YM, Zheng Y, Yu Y, Wang Y, Huang Q, Qian F, Sun L, Song ZG, Chen Z, Feng J, et al. Blood molecular markers associated with COVID‐19 immunopathology and multi‐organ damage. EMBO J. 2020;39:e105896. doi: 10.15252/embj.2020105896 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0020] 20. Xu W, Wu J, An Y, Xiao C, Hao F, Liu H, Wang Y, Tang H. Streptozotocin‐induced dynamic metabonomic changes in rat biofluids. J Proteome Res. 2012;11:3423–3435. doi: 10.1021/pr300280t [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0021] 21. Liu M, Tang H, Nicholson JK, Lindon JC. Use of 1H NMR‐determined diffusion coefficients to characterize lipoprotein fractions in human blood plasma. Magn Reson Chem. 2002;40:S83–S88. doi: 10.1002/mrc.1121 [DOI] [Google Scholar]

[jah310328-bib-0022] 22. Wu Q, Huang QX, Zeng HL, Ma S, Lin HD, Xia MF, Tang HR, Gao X. Prediction of metabolic disorders using NMR‐based metabolomics: the Shanghai Changfeng study. Phenomics. 2021;1:186–198. doi: 10.1007/s43657-021-00021-2 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0023] 23. Zhu X, Huang Q, Ma S, Chen L, Wu Q, Wu L, Ma H, Li X, Li Q, Aleteng Q, et al. Presence of sarcopenia identifies a special group of lean NAFLD in middle‐aged and older people. Hepatol Int. 2023;17:313–325. doi: 10.1007/s12072-022-10439-z [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0024] 24. van Buuren S, Groothuis‐Oudshoorn K. mice: multivariate imputation by chained equations in R. J Stat Softw. 2011;45:1–67. doi: 10.18637/jss.v045.i03 [DOI] [Google Scholar]

[jah310328-bib-0025] 25. Arnett DK, Blumenthal RS, Albert MA, Buroker AB, Goldberger ZD, Hahn EJ, Himmelfarb CD, Khera A, Lloyd‐Jones D, McEvoy JW, et al. 2019 ACC/AHA guideline on the primary prevention of cardiovascular disease: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Circulation. 2019;140:e596–e646. doi: 10.1161/CIR.0000000000000678 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0026] 26. Friedman J, Hastie T, Tibshirani R. Regularization paths for generalized linear models via coordinate descent. J Stat Softw. 2010;33:1–22. doi: 10.18637/jss.v033.i01 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0027] 27. Gao X, Hofman A, Hu Y, Lin H, Zhu C, Jeekel J, Jin X, Wang J, Gao J, Yin Y, et al. The Shanghai Changfeng study: a community‐based prospective cohort study of chronic diseases among middle‐aged and elderly: objectives and design. Eur J Epidemiol. 2010;25:885–893. doi: 10.1007/s10654-010-9525-6 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0028] 28. Liu Z, Wang Y, Zhang Y, Chu X, Wang Z, Qian D, Chen F, Xu J, Li S, Jin L, et al. Cohort profile: the Rugao longevity and ageing study (RuLAS). Int J Epidemiol. 2016;45:1064–1073. doi: 10.1093/ije/dyv101 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0029] 29. Burgess S, Thompson SG; Collaboration CCG . Avoiding bias from weak instruments in Mendelian randomization studies. Int J Epidemiol. 2011;40:755–764. doi: 10.1093/ije/dyr036 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0030] 30. Chang CC, Chow CC, Tellier LC, Vattikuti S, Purcell SM, Lee JJ. Second‐generation PLINK: rising to the challenge of larger and richer datasets. Gigascience. 2015;4:7. doi: 10.1186/s13742-015-0047-8 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0031] 31. Burgess S, Butterworth A, Thompson SG. Mendelian randomization analysis with multiple genetic variants using summarized data. Genet Epidemiol. 2013;37:658–665. doi: 10.1002/gepi.21758 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0032] 32. Bowden J, Davey Smith G, Burgess S. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol. 2015;44:512–525. doi: 10.1093/ije/dyv080 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0033] 33. Hartwig FP, Davey Smith G, Bowden J. Robust inference in summary data Mendelian randomization via the zero modal pleiotropy assumption. Int J Epidemiol. 2017;46:1985–1998. doi: 10.1093/ije/dyx102 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0034] 34. Bowden J, Davey Smith G, Haycock PC, Burgess S. Consistent estimation in Mendelian randomization with some invalid instruments using a weighted median estimator. Genet Epidemiol. 2016;40:304–314. doi: 10.1002/gepi.21965 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0035] 35. Burgess S, Thompson SG. Interpreting findings from Mendelian randomization using the MR‐Egger method. Eur J Epidemiol. 2017;32:377–389. doi: 10.1007/s10654-017-0255-x [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0036] 36. Hemani G, Zheng J, Elsworth B, Wade KH, Haberland V, Baird D, Laurin C, Burgess S, Bowden J, Langdon R, et al. The MR‐base platform supports systematic causal inference across the human phenome. elife. 2018;7:e34408. doi: 10.7554/eLife.34408 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0037] 37. Skrivankova VW, Richmond RC, Woolf BAR, Davies NM, Swanson SA, VanderWeele TJ, Timpson NJ, Higgins JPT, Dimou N, Langenberg C, et al. Strengthening the reporting of observational studies in epidemiology using mendelian randomisation (STROBE‐MR): explanation and elaboration. BMJ. 2021;375:n2233. doi: 10.1136/bmj.n2233 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0038] 38. Yang Y, Bartz TM, Brown MR, Guo X, Zilhao NR, Trompet S, Weiss S, Yao J, Brody JA, Defilippi CR, et al. Identification of functional genetic determinants of cardiac troponin T and I in a multiethnic population and causal associations with atrial fibrillation. Circ Genom Precis Med. 2021;14:e003460. doi: 10.1161/CIRCGEN.121.003460 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0039] 39. Bentzon JF, Otsuka F, Virmani R, Falk E. Mechanisms of plaque formation and rupture. Circ Res. 2014;114:1852–1866. doi: 10.1161/circresaha.114.302721 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0040] 40. Tsao CW, Preis SR, Peloso GM, Hwang SJ, Kathiresan S, Fox CS, Cupples LA, Hoffmann U, O'Donnell CJ. Relations of long‐term and contemporary lipid levels and lipid genetic risk scores with coronary artery calcium in the framingham heart study. J Am Coll Cardiol. 2012;60:2364–2371. doi: 10.1016/j.jacc.2012.09.007 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0041] 41. White J, Swerdlow DI, Preiss D, Fairhurst‐Hunter Z, Keating BJ, Asselbergs FW, Sattar N, Humphries SE, Hingorani AD, Holmes MV. Association of lipid fractions with risks for coronary artery disease and diabetes. JAMA Cardiol. 2016;1:692–699. doi: 10.1001/jamacardio.2016.1884 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0042] 42. Holmes MV, Asselbergs FW, Palmer TM, Drenos F, Lanktree MB, Nelson CP, Dale CE, Padmanabhan S, Finan C, Swerdlow DI, et al. Mendelian randomization of blood lipids for coronary heart disease. Eur Heart J. 2015;36:539–550. doi: 10.1093/eurheartj/eht571 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0043] 43. Holmes MV, Millwood IY, Kartsonaki C, Hill MR, Bennett DA, Boxall R, Guo Y, Xu X, Bian Z, Hu R, et al. Lipids, lipoproteins, and metabolites and risk of myocardial infarction and stroke. J Am Coll Cardiol. 2018;71:620–632. doi: 10.1016/j.jacc.2017.12.006 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0044] 44. Inouye M, Kettunen J, Soininen P, Silander K, Ripatti S, Kumpula LS, Hamalainen E, Jousilahti P, Kangas AJ, Mannisto S, et al. Metabonomic, transcriptomic, and genomic variation of a population cohort. Mol Syst Biol. 2010;6:441. doi: 10.1038/msb.2010.93 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0045] 45. Kumar A, Sathian B. Correlation between lipid profile and troponin I test results in patients with chest pain in Nepal. Asian Pac J Trop Biomed. 2013;3:487–491. doi: 10.1016/S2221-1691(13)60101-X [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0046] 46. Li XL, Guo YL, Zhu CG, Xu RX, Qing P, Wu NQ, Jiang LX, Xu B, Gao RL, Li JJ. Relationship of high‐density lipoprotein cholesterol with periprocedural myocardial injury following elective percutaneous coronary intervention in patients with low‐density lipoprotein cholesterol below 70 mg/dL. J Am Heart Assoc. 2015;4:e001412. doi: 10.1161/JAHA.114.001412 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0047] 47. Liu QK. Triglyceride‐lowering and anti‐inflammatory mechanisms of omega‐3 polyunsaturated fatty acids for atherosclerotic cardiovascular risk reduction. J Clin Lipidol. 2021;15:556–568. doi: 10.1016/j.jacl.2021.05.007 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0048] 48. Simonetto M, Infante M, Sacco RL, Rundek T, Della‐Morte D. A novel anti‐inflammatory role of omega‐3 PUFAs in prevention and treatment of atherosclerosis and vascular cognitive impairment and dementia. Nutrients. 2019;11:2279. doi: 10.3390/nu11102279 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah310328-bib-0049] 49. Seres‐Noriega T, Gimenez M, Perea V, Blanco J, Vinagre I, Pane A, Ruiz S, Cofan M, Mesa A, Esmatjes E, et al. Quantification of glycoproteins by nuclear magnetic resonance associated with preclinical carotid atherosclerosis in patients with type 1 diabetes. Nutr Metab Cardiovasc Dis. 2021;31:2099–2108. doi: 10.1016/j.numecd.2021.03.021 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0050] 50. Malo AI, Girona J, Ibarretxe D, Rodriguez‐Borjabad C, Amigo N, Plana N, Masana L. Serum glycoproteins A and B assessed by (1)H‐NMR in familial hypercholesterolemia. Atherosclerosis. 2021;330:1–7. doi: 10.1016/j.atherosclerosis.2021.06.905 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0051] 51. Heitmann BL. The effects of gender and age on associations between blood lipid levels and obesity in Danish men and women aged 35–65 years. J Clin Epidemiol. 1992;45:693–702. doi: 10.1016/0895-4356(92)90046-p [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0052] 52. Schaefer EJ, Lamon‐Fava S, Cohn SD, Schaefer MM, Ordovas JM, Castelli WP, Wilson PW. Effects of age, gender, and menopausal status on plasma low density lipoprotein cholesterol and apolipoprotein B levels in the Framingham Offspring Study. J Lipid Res. 1994;35:779–792. doi: 10.1016/S0022-2275(20)39173-2 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0053] 53. Anderson KM, Wilson PW, Garrison RJ, Castelli WP. Longitudinal and secular trends in lipoprotein cholesterol measurements in a general population sample. The Framingham Offspring Study. Atherosclerosis. 1987;68:59–66. doi: 10.1016/0021-9150(87)90094-3 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0054] 54. Wilson PW, Anderson KM, Harris T, Kannel WB, Castelli WP. Determinants of change in total cholesterol and HDL‐C with age: the Framingham study. J Gerontol. 1994;49:M252–M257. doi: 10.1093/geronj/49.6.m252 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0055] 55. Ferrara A, Barrett‐Connor E, Shan J. Total, LDL, and HDL cholesterol decrease with age in older men and women. The Rancho Bernardo Study 1984–1994. Circulation. 1997;96:37–43. doi: 10.1161/01.cir.96.1.37 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0056] 56. Laurenzi M, Mancini M. Plasma lipids in elderly men and women. Eur Heart J. 1988;9:69–74. doi: 10.1093/eurheartj/9.suppl_d.69 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0057] 57. Ettinger WH, Wahl PW, Kuller LH, Bush TL, Tracy RP, Manolio TA, Borhani NO, Wong ND, O'Leary DH. Lipoprotein lipids in older people. Results from the Cardiovascular Health Study. The CHS Collaborative Research Group. Circulation. 1992;86:858–869. doi: 10.1161/01.cir.86.3.858 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0058] 58. Frishman WH, Ooi WL, Derman MP, Eder HA, Gidez LI, Ben‐Zeev D, Zimetbaum P, Heiman M, Aronson M. Serum lipids and lipoproteins in advanced age. Intraindividual changes. Ann Epidemiol. 1992;2:43–50. doi: 10.1016/1047-2797(92)90036-p [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0059] 59. Heiss G, Tamir I, Davis CE, Tyroler HA, Rifkand BM, Schonfeld G, Jacobs D, Frantz ID Jr. Lipoprotein‐cholesterol distributions in selected North American populations: the lipid research clinics program prevalence study. Circulation. 1980;61:302–315. doi: 10.1161/01.cir.61.2.302 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0060] 60. Kromhout D, Nissinen A, Menotti A, Bloemberg B, Pekkanen J, Giampaoli S. Total and HDL cholesterol and their correlates in elderly men in Finland, Italy, and the Netherlands. Am J Epidemiol. 1990;131:855–863. doi: 10.1093/oxfordjournals.aje.a115576 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0061] 61. Hubert HB, Eaker ED, Garrison RJ, Castelli WP. Life‐style correlates of risk factor change in young adults: an eight‐year study of coronary heart disease risk factors in the Framingham offspring. Am J Epidemiol. 1987;125:812–831. doi: 10.1093/oxfordjournals.aje.a114598 [DOI] [PubMed] [Google Scholar]

[jah310328-bib-0062] 62. Garry PJ, Hunt WC, Koehler KM, VanderJagt DJ, Vellas BJ. Longitudinal study of dietary intakes and plasma lipids in healthy elderly men and women. Am J Clin Nutr. 1992;55:682–688. doi: 10.1093/ajcn/55.3.682 [DOI] [PubMed] [Google Scholar]

PERMALINK

Metabonomic Biomarkers of Plaque Burden and Instability in Patients With Coronary Atherosclerotic Disease After Moderate Lipid‐Lowering Therapy

Zhendong Mei, PhD

Lili Xu, MD

Qingxia Huang, PhD

Chenhao Lin, PhD

Mengyao Yu, PhD

Shalaimaiti Shali, MD

Hongyi Wu, MD

Yijing Lu, PhD

Runda Wu, MD

Zhen Wang, MD

Lingfeng Luo, MPH

Zhonghan Sun, PhD

Liang Sun, PhD

Juying Qian, MD

Guochong Chen, PhD

Huiru Tang, PhD

Kang Yao, MD

Yan Zheng, MD, PhD

Yuxiang Dai, MD

Junbo Ge, MD

Abstract

Background

Methods and Results

Conclusions

Nonstandard Abbreviations and Acronyms

Clinical Perspective.

What Is New?

What Are the Clinical Implications?

Methods

Study Population

Evaluation of Coronary Atherosclerosis Severity

Follow‐Up Visits

Measurement of NMR‐Based Metabolomics

Serum Sample Preparation

NMR Analysis

Statistical Analysis

Two‐Sample MR

Data Source of Exposure

Data Source of Outcome

MR Analysis

Results

GRAND Cohort Characteristics

Table .

Different Roles of Lipid Components in Coronary Plaque Burden and Instability

Figure 1. Associations of lipid components with CAD severity.

Different Lipid Subfraction Signatures of Coronary Plaque Burden and Instability

Figure 2. Associations of lipid subfractions with CAD severity.

Associations of Amino Acids and Other Metabolites With Plaque Burden and Instability

Figure 3. Associations of amino acids and other metabolites with CAD severity.

Secondary Analyses of the Associations Between Metabolites and Severity of Atherosclerosis

Causal Effects of Circulating Metabolites on Plaque Burden and Instability

Figure 4. Causal effects of metabolomic features on Gensini score.

Metabolomic Indices and Risk of MACEs

Figure 5. Associations of metabolomic indices of Gensini score (A) and cTnT levels (B) with risk of major adverse cardiac events.

Discussion

Sources of Funding

Disclosures

Supporting information

Acknowledgments

Contributor Information

References

Associated Data

Supplementary Materials

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases