Plasma Lipidomics for Cardiovascular Risk Prediction: Key Biomarkers and Clinical Insights

1. Lipidomics for Mapping ASCVD Lipid Metabolism

Lipidomics is a major branch of metabolomics focused on the systematic analysis of lipid composition, abundance, and fine structure in biological samples. It covers major lipid categories, including glycerolipids, glycerophospholipids, sphingolipids, sterol lipids, fatty acids, and glycolipids, and can quantify thousands of individual lipid species in plasma. Its key advantage is that it moves beyond the limitation of traditional biochemical assays that measure total lipid levels without distinguishing molecular structures. Lipidomics can reveal fine structural differences such as fatty acyl chain length, double-bond number and position, and sn-position configuration; these features directly shape the biological functions of lipid molecules [3][4].

Current lipidomics platforms generally fall into two major categories and can support the full workflow of biomarker discovery, validation, and clinical translation:

• Untargeted lipidomics: Built on ultra-high-performance liquid chromatography coupled with high-resolution mass spectrometry, this approach enables unbiased screening of detectable lipid molecules in samples and is well suited for discovering novel biomarkers;
• Targeted lipidomics: Built on ultra-high-performance liquid chromatography coupled with triple-quadrupole mass spectrometry, this approach uses multiple reaction monitoring (MRM) for highly sensitive absolute quantification of defined lipid molecules and is suitable for biomarker discovery, validation, and clinical translation [4].

Emerging technologies, including ion mobility mass spectrometry and imaging lipidomics, can further resolve lipid spatial distribution and fine structural features, providing stronger technical support for mechanistic studies and the translation of lipid biomarkers [6].

2. Plasma Lipidomics Biomarkers for ASCVD Risk Prediction

Over the past decade, lipidomics studies based on large prospective cohorts worldwide have identified multiple ASCVD risk prediction biomarkers that are independent of, and complementary to, traditional lipid indicators. By lipid class, these biomarkers can be grouped into the following major categories, broadly ordered by the current strength of evidence.

2.1 Sphingolipid Biomarkers in ASCVD Risk Prediction

Sphingolipids are core structural components of cell membranes and participate in key ASCVD-related pathological processes, including apoptosis, inflammation, and oxidative stress. Among them, ceramides (Cer) are currently the most extensively studied lipidomics biomarkers and have achieved the most mature level of clinical translation [10].

Ceramides are central intermediates in sphingolipid metabolism and consist of a sphingosine backbone linked to acyl chains of varying lengths. Their biological functions show marked acyl-chain specificity. In 2016, a study of three independent cohorts (n>5000) published in European Heart Journal first showed that plasma Cer (d18:1/16:0), Cer (d18:1/18:0), Cer (d18:1/24:0), and the Cer (d18:1/24:1)/Cer (d18:1/24:0) ratio were significantly and positively associated with major adverse cardiovascular events (MACE). These associations were independent of all traditional ASCVD risk factors, including LDL-C [5]. In the same year, a study from the Mayo Clinic team published in Circulation further validated this finding and clarified the prognostic value of ceramides in patients with stable coronary artery disease and acute coronary syndromes [34]. Since then, more than 20 independent cohorts worldwide, covering primary prevention, secondary prevention, statin-treated populations, patients with diabetes, patients with chronic kidney disease, and other groups, have validated the predictive value of ceramides [6]:

• Incremental value for risk stratification: A study based on the ARIC cohort (n=9200) showed that adding a four-ceramide risk score to the traditional ASCVD pooled cohort equations (PCE) risk score significantly increased the C-statistic by 0.023, achieved a net reclassification improvement (NRI) of 15.8%, and achieved an integrated discrimination improvement (IDI) of 0.018. Notably, the ceramide score reclassified 30% of intermediate-risk individuals into high-risk or low-risk groups, helping refine clinical intervention decisions [7];
• Value for residual risk assessment: In statin-treated participants in the JUPITER study, individuals with the highest baseline ceramide levels had a 2.3-fold higher risk of MACE than those with the lowest levels. This association was independent of post-treatment LDL-C levels, helping address an important gap in residual risk assessment that traditional lipid indicators do not fully capture [8].

Beyond ceramides, other sphingolipid molecules have also shown potential predictive value. The effects of sphingomyelin (SM) subtypes are highly heterogeneous: SM (16:0) and SM (18:0), which contain saturated fatty acids, are positively associated with MACE risk, whereas SM (24:1), which contains long-chain polyunsaturated fatty acids, is negatively associated with risk [9]. Glycosphingolipids such as glucosylceramide (GluCer) and lactosylceramide (LacCer), as well as sphingosine-1-phosphate (S1P), have also been associated with ASCVD incidence and plaque stability, although their clinical predictive value still requires validation in large cohorts [6,10].

2.2 Glycerolipid Biomarkers Beyond Total Triglycerides

Traditional assays measure total plasma TG levels, but TG represents a highly heterogeneous class of molecules. Hundreds of TG subtypes with different fatty acyl chain compositions can be detected in plasma, and their associations with ASCVD risk depend strongly on fatty acyl chain length and degree of unsaturation. This molecular heterogeneity is a key blind spot of conventional total TG testing [11]. Multiple large cohort studies have consistently shown that:

• TG subtypes containing short-chain (carbon number <=16), saturated, or monounsaturated fatty acids, such as TG (48:0), TG (50:1), and TG (52:2), are significantly and positively associated with ASCVD incidence risk, independent of traditional total TG levels [12];
• TG subtypes containing long-chain (carbon number >=20) polyunsaturated fatty acids, especially omega-3 fatty acids, such as TG (54:6), TG (56:7), and TG (58:8), are significantly and inversely associated with ASCVD risk [22].

Lipidomics analysis from the Framingham Heart Study showed that 14 TG subtypes were significantly associated with incident coronary heart disease. Among them, the risk associations of eight short-chain saturated or monounsaturated TG species were independent of total TG and other traditional risk factors [12]. This finding helps explain the heterogeneous relationship between total TG and ASCVD risk and provides lipidomics-level mechanistic evidence for the cardioprotective effects of omega-3 fatty acids [22]. Diacylglycerols (DG) are core intermediates in TG synthesis and degradation and also serve as important lipid second messengers. Plasma DG subtypes such as DG (16:0/18:1) and DG (18:1/18:1) are significantly and positively associated with MACE risk and provide incremental predictive value beyond traditional risk factors [13].

2.3 Glycerophospholipids, Lipoprotein Function, and Vascular Homeostasis

Glycerophospholipids are major structural lipids in cell membranes and lipoproteins. They include multiple subclasses, such as phosphatidylcholine (PC), phosphatidylethanolamine (PE), lysophospholipids, and plasmalogens. Their molecular composition directly influences lipoprotein structure and function, as well as vascular endothelial homeostasis [19].

2.3.1 Lysophospholipid Biomarkers in Atherosclerosis Inflammation

Lysophospholipids are hydrolysis products of glycerophospholipids. Among them, lysophosphatidylcholine (LPC) is the most abundant lysophospholipid in plasma and a major active component of oxidized low-density lipoprotein (ox-LDL). LPC participates in multiple ASCVD-related pathological processes, including endothelial injury, foam cell formation, and plaque instability [14]. Lipidomics studies have found that total plasma LPC levels are negatively associated with ASCVD risk, whereas specific LPC subtypes can show opposite directions of association. LPC (16:0) and LPC (18:0), which contain saturated fatty acids, are positively associated with MACE risk, while LPC (22:6) and LPC (20:5), which contain polyunsaturated fatty acids, are negatively associated with risk [14]. A nested case-control study based on the UK Biobank (n>10000) showed that LPC (16:0) significantly improved the predictive performance of traditional risk models, with an NRI of 12.3% [15]. In addition, lysophosphatidylethanolamine (LPE), lysophosphatidic acid (LPA), and other subtypes have also been associated with ASCVD risk, and LPA has become a potential target for anti-atherosclerotic drug development [16].

2.3.2 Plasmalogen Biomarkers in Oxidative Stress and Plaque Stability

Plasmalogens are a specialized class of glycerophospholipids containing a vinyl ether bond. They have potent antioxidant and anti-apoptotic properties and function as endogenous protective factors in the vascular endothelium [17]. Multiple studies have consistently shown that plasma plasmalogen levels are significantly and inversely associated with ASCVD incidence risk. For example, each 1-standard-deviation increase in subtypes such as plasmenyl-PC (16:0/20:4) and plasmenyl-PE (18:0/22:6) is associated with a 20%-30% reduction in MACE risk, independent of traditional risk factors [17]. Plasmalogen levels are also directly associated with coronary plaque stability. Plasma plasmalogen levels are significantly lower in patients with vulnerable plaques than in patients with stable plaques, suggesting that plasmalogens may serve not only as risk prediction biomarkers but also as noninvasive markers of plaque instability [18].

2.4 Sterol Lipids and Fatty Acids Beyond LDL-C

Traditional assays measure the total levels of sterol-related lipid indicators such as TC, LDL-C, and HDL-C. They cannot distinguish the fatty acid composition of cholesteryl esters (CE), nor do they fully reflect the metabolic status of free cholesterol (FC) [21]. Lipidomics studies have found that CE fatty acid composition is closely associated with ASCVD risk: saturated or monounsaturated CE species such as CE (16:0) and CE (18:1) are positively associated with MACE risk, whereas polyunsaturated CE species such as CE (20:5) and CE (22:6) are negatively associated with risk [20]. In addition, the FC/CE ratio is a core marker of cholesterol esterification efficiency, and elevated levels are associated with increased ASCVD risk independently of total cholesterol levels [21].

Free fatty acids (FFA) are central intermediates in lipid metabolism, and different FFA subtypes have markedly different cardiovascular effects. Saturated fatty acids, including palmitic acid (16:0) and stearic acid (18:0), are positively associated with ASCVD risk, whereas omega-3 polyunsaturated fatty acids, including EPA (20:5) and DHA (22:6), are negatively associated with risk [22]. Fatty acid derivatives such as eicosanoids, including prostaglandins and leukotrienes, are key mediators of inflammatory responses and have also been associated with ASCVD risk and plaque stability. However, their detection remains technically challenging, and clinical translation still depends on further analytical advances [23].

Beyond individual sterol and fatty acid species, lipoprotein-level heterogeneity also helps explain why conventional LDL-C and HDL-C measurements do not fully capture ASCVD risk. Small, dense LDL particles have been linked to greater atherogenicity [24], while HDL particles are functionally heterogeneous and vulnerable to compositional and functional remodeling [25]. Prospective lipoprotein subclass profiling further shows that lipoprotein subclasses can provide additional information for coronary heart disease risk assessment beyond conventional lipid measures [26].

3. Clinical Translation of Lipidomics Biomarkers for ASCVD

3.1 Clinical Use of Lipidomics Biomarkers in ASCVD Risk Stratification

i. Optimizing risk stratification for primary ASCVD prevention: Traditional PCE scores have limited predictive accuracy in intermediate-risk populations, whereas lipidomics biomarkers can improve risk discrimination, particularly by reclassifying intermediate-risk individuals and reducing the likelihood of over-intervention or under-intervention. Clinical testing products based on ceramide scores (CERT1 and CERT2) have been approved in the United States for ASCVD risk assessment and are referenced in clinical consensus guidance from the American College of Cardiology (ACC) [27].

ii. Assessment of residual cardiovascular risk: After lipid-lowering therapies such as statins and PCSK9 inhibitors, a substantial proportion of patients still experience MACE even when LDL-C targets are achieved. Lipidomics biomarkers, represented by ceramides and specific TG subtypes, can help predict residual risk after lipid-lowering therapy and guide more intensive intervention strategies [2][8].

iii. Prognostic assessment in secondary prevention populations: In secondary prevention populations, including patients with acute coronary syndrome or myocardial infarction, lipidomics biomarkers can help predict recurrent cardiovascular events and all-cause mortality, supporting more individualized secondary prevention strategies [5][6].

iv. Efficacy monitoring and target development for lipid-lowering drugs: Lipidomics can comprehensively evaluate how lipid-lowering drugs affect lipid metabolic networks and can help identify new therapeutic targets. Inhibitors targeting ceramide synthase have entered preclinical research and may represent a potential class of anti-ASCVD therapeutics [28].

3.2 Translational Challenges for Lipidomics Biomarkers in Cardiovascular Risk Prediction

i. Standardization and harmonization of testing technologies: Sample preparation methods, instrument platforms, quantification strategies, and lipid annotation standards for lipidomics testing still vary substantially across laboratories, leading to limited comparability of results. This remains one of the major barriers to clinical translation. Internationally harmonized lipidomics testing standards, quality control systems, and reference materials are urgently needed [29].

ii. Distinguishing association from causality: Most lipidomics biomarkers have been discovered through observational cohort studies, which can demonstrate associations but cannot establish causality. Mendelian randomization studies, gene-edited animal models, and cellular functional experiments are needed to validate causal relationships between lipid molecules and ASCVD and to distinguish true pathogenic factors from biomarkers that merely accompany disease processes [30].

iii. Large-sample validation across populations and centers: Most current studies of lipidomics biomarkers have been conducted in European and American White populations, with insufficient validation in Asian, African, and other ethnic populations. In particular, large, long-term follow-up cohort studies in Chinese populations remain relatively limited. Multicenter and multiethnic validation studies are urgently needed to ensure biomarker generalizability [31][33].

iv. Testing cost and clinical accessibility: The current cost of lipidomics testing remains much higher than that of conventional lipid testing, which limits its use in routine clinical practice. Wider adoption of mass spectrometry technologies, together with automation and scaling of analytical workflows, is expected to gradually reduce testing costs and support clinical translation [6].

v. Multi-omics integration and big-data analysis: The predictive value of any single lipid biomarker is limited. Integrating lipidomics with genomics, transcriptomics, proteomics, and gut microbiome data, combined with big-data methods such as machine learning and artificial intelligence, may enable more precise ASCVD risk prediction models. This represents an important direction for future development [32].

4. Future Directions for ASCVD Lipidomics Biomarkers

Dysregulated lipid metabolism is a central pathological basis of ASCVD. Traditional lipid indicators capture only a limited portion of lipid metabolic complexity, whereas lipidomics technologies provide unprecedented tools for comprehensively mapping lipid metabolic networks and discovering novel ASCVD risk prediction biomarkers [3][4].

Over the past decade and more, numerous prospective lipidomics studies worldwide have identified novel biomarkers that complement traditional lipid indicators. Sphingolipids represented by ceramides, structure-specific glycerolipid subtypes, lysophospholipids, plasmalogens, and other lipid molecules have been validated in multiple independent cohorts for ASCVD risk prediction. These biomarkers offer particular value for risk stratification in individuals with normal traditional lipid profiles, populations with residual risk after statin treatment, and intermediate-risk populations [5][6][7][8].

Although the clinical translation of lipidomics biomarkers still faces challenges related to testing standardization, causal validation, and cost control, continued advances in mass spectrometry, the establishment of international standardization systems, and deeper multi-omics integration will likely expand the role of lipidomics in ASCVD precision prevention, risk stratification, efficacy monitoring, and target development [29][32]. In the future, lipidomics biomarkers may enter routine clinical testing and be applied together with traditional lipid indicators to support individualized precision management of ASCVD and ultimately reduce the global burden of CVD.