Coronary artery disease is associated with particular change of serum metabolome: a case-control study

Abstract

Introduction: Cardiovascular disease (CVD) is a significant cause of mortality worldwide. Preventive programs are trying to reduce the burden of the disease. Recent advances in metabolomics profiling open a new avenue for developing complementary CVD evaluation strategies.

Objectives: The aim of the study was to investigate whether a metabolomic profile can provide an additional characterisation of individuals with coronary artery disease (CAD).

Methods: The study included 167 participants with CAD aged 41-79 years. A control group was formed of 166 individuals without CAD, gender- and age-matched to the study group. A total of 188 metabolites were profiled in serum by liquid chromatography-tandem mass spectrometry. After clearing the data, associations between 132 metabolites and CAD presence were analysed using multiple linear regression models.

Results: We observed significant differences in serum metabolic profiles between analysed groups on various levels. However, a deeper analysis revealed sphingomyelin 41:1 (SM 41:1) as the main metabolite independently associated with CAD after correction for classical CV risk factors. Its concentration was lower in the CAD group (median 9.79 µmol/L, interquartile range (IQR) 7.92-12.23) compared to control one (median 13.60 µmol/L, IQR 11.30-16.15) (p < 0.001). Further analysis showed that SM 41:1 concentration was inversely correlated with CAD, current smoking, and hypertension; and positively associated with female gender and non-HDL level.

Conclusions: CAD patients present lower plasma concentrations of SM 41:1 than healthy subjects. A better understanding of the biological function of sphingomyelin in CAD patients may help develop therapeutic approaches and risk stratification in this group.

Keywords: Sphingomyelin 41:1; Cardiovascular disease; Coronary artery disease; Metabolomics; Principal component analysis.

Fig. 1

Study flow diagram

Fig. 2

Metabolites significantly associated with CAD in multiple linear regression models adjusted for BMI, GFR, and diabetes (Model 1) and its association with CAD in Model 2 (adjusted for BMI, GFR, diabetes, smoking status, non-HDL cholesterol, alanine aminotransferase, hsCRP, statin therapy)

Fig. 3

Association of SM 41:1 concentration with clinical, echocardiographic, and laboratory parameters; ^aCategorical variable; ^bHistory or fasting glucose ≥ 126 mg/dL or OGTT120 min glucose ≥ 200 mg/dL or HbA1c ≥ 6.5%; ^cSystolic > 140 mmHg and/or diastolic > 90 mmHg; ^dThe analysis was not performed since almost the entire group declared a disease; BMI body mass index, BPd blood pressure dystolic, BPs blood pressure systolic, eGFR estimated glomerular filtration rate, HbA1c glycated haemoglobin A1c, HDL-C high density lipoprotein cholesterol, HR heart rate, hsCRP high sensitivity C-reactive protein, hs-TnT high sensitivity troponin T., LAVI left atrial volume index, LDL-C low density lipoprotein cholesterol, LVEF left ventricular ejection fraction, LVMI left ventricular mass index, nonHDL-C non high density lipoprotein cholesterol, NT-proBNP n-terminal pro-brain natriuretic peptide, TG triglicerydes, WHR waist-hip ratio

Fig. 4

Association of SM 41:1 concentration with study group, age, sex, body mass index (BMI), blood pressure systolic (BPs), hypertension, diabetes, smoking status, non high density lipoprotein cholesterol (nonHDL-C), glycated haemoglobin A1c (HbA1c), and statin dose equivalent to simvastatin. Panel A presents multiple regression model for sphingomyeline 41:1 (SM 41:1) (Model 2) separately performed for coronary artery disease (CAD) and control group. Panel B presents multiple regression model sphingomyeline 41:1 (SM 41:1) (Model 3) for entire study population. ^aCategorical variable; ^bEquivalent to simvastatin. ^cHistory or fasting glucose ≥ 126 mg/dL or OGGT 120 min glucose ≥ 200 mg/dL or HbA1c ≥ 6.5%

Fig. 5

PCA analysis: A Factor loadings of two principal components (PC) most differentiating coronary artery diseases (CAD) and healthy subjects, including metabolites with factor loading > 0.6, B PCA biplot of 2 factors differentiating CAD and healthy subjects the most