MSACL 2026 Abstract

Lipidomes Over Time: Individuality, Sex Differences, and Network Dynamics in a Longitudinal Adult Cohort of Lausanne

Nicolas Goss (1, 2, 3), Théo Cavinato (2, 3), Ko Ko Maung (5), Jessica Medina (1), Rebecca Borreggine (1), Tony Teav (1), Pedro Marquez Vidal (5), Hector Gallart-Ayala (1), Zoltán Kutalik (2, 3, 4), Julijana Ivanisevic (1)
(1) Metabolomics and Lipidomics Platform, Faculty of Biology and Medicine, University of Lausanne, Lausanne, Switzerland, (2) Department of Computational Biology, Faculty of Biology and Medicine, University of Lausanne, Lausanne, Switzerland, (3) Swiss Institute of Bioinformatics (SIB), University of Lausanne, Lausanne, Switzerland, (4) University Center for Primary Care and Public Health, Lausanne, Switzerland, (5) Department of Medicine, Internal Medicine, Lausanne University Hospital (CHUV) and University of Lausanne, Lausanne, Switzerland

Nicolas Goss (Presenter) University of Lausanne

Presenter Bio: Nicolas is a PhD candidate at the University of Lausanne, at the interface of the Metabolomics & Lipidomics Facility and the Statistical Genetics Group. Trained as a biologist at the Johannes Gutenberg-University in Mainz and the University of Cologne, he specializes in mass spectrometry–based analyses. His interdisciplinary research focuses on lipid biology in health and disease, integrating mass spectrometry technologies with statistical analyses across genetics, metabolism, and clinical outcomes.

Relevant Financial Disclosures (within past 24 months, reported on Jun 02, 2026) No relevant financial relationship(s) to disclose.

INTRODUCTION:
The global obesity epidemic and the rising burden of cardiometabolic diseases represent major challenges for modern medicine. Beyond established risk factors such as age, sex, body mass index, and smoking status, dysregulation of lipid metabolism is closely linked to cardiometabolic health outcomes. In clinical practice, lipid metabolism is typically assessed using bulk measures such as LDL-/HDL-cholesterol and total triacylglycerols. While these markers have improved risk prediction and advanced our understanding of disease, they provide only a coarse view of lipid biology and fail to capture the extensive structural and functional diversity of the circulating lipidome. As a result, traditional cardiometabolic risk equations often do not adequately capture biological heterogeneity in patients of intermediate risk. Advances in mass spectrometry now enable comprehensive characterization of circulating lipid species, offering the potential to improve both biological interpretability and predictive power. State-of-the-art methodologies allow quantitative measurement of lipid species with high sensitivity and specificity. While initial studies have begun to explore these previously inaccessible layers of lipid metabolism from multiple perspectives (1,2), current datasets remain limited by heterogeneous disease statuses, inclusion of non-fasted participants, small cohort sizes, scarce longitudinal data, and suboptimal lipid coverage. These limitations bias data interpretation, hinder the integration of findings across studies, and constrain a unified view linking genetics, molecular phenotypes, and health outcomes.

In this study, we generated a large-scale, quantitative lipidomics dataset targeting 20 lipid classes and 944 species in a Swiss cohort of 2577 apparently healthy adults (40-80y) at baseline, with a 10-year follow-up. Co-regulated lipids were grouped into clusters, reflecting that most lipids act in coordinated networks rather than as individual molecules. This approach captures lipid biology more accurately, improves interpretability and reduces data dimensionality. The resulting framework enables the integration of genetic, metabolic, and clinical perspectives.

OBJECTIVES:
The objectives of this study are to (1) define co-regulated lipid modules using large-scale longitudinal lipidomics data, (2) characterize their trajectories across the adult lifespan, and (3) investigate their genetic determinants and predictive capacity for cardiometabolic health outcomes.

METHODS:
Participants were drawn from the CoLaus cohort, with fasted plasma samples and clinical data collected at 5-year intervals. Lipidomics analyses included 2577 apparently healthy individuals at baseline, with follow-up measurements in 2235 and 1925 participants after five and ten years, respectively. Plasma lipid species were quantified using high-throughput LC–MS/MS–based lipidomics with stable isotope–labeled standard spike-in (75 species), quality control and batch correction procedures (3). Lipid species were grouped into 32 co-regulated clusters using k-means clustering. Lipid trajectories were assessed using mixed-effects models to account for inter-individual variability. Genetic associations were evaluated using genome-wide association analyses. Predictive capacity of lipid clusters is currently being assessed in the context of cardiovascular (SCORE2) and diabetes risk (several diabetes risk scores).

RESULTS:
We quantified 683 lipid species across >6,700 samples and 83 batches with high precision (CV across internal QCs < 20%, and d-ratio < 40%). These comprehensive lipid signatures show a high level of individuality, with samples from the same individual clustering together, at timescale of 10 years. More than half of the measured lipids displayed significant sex differences, with SMs and PCs found in significantly higher concentrations in females. In contrast, males showed higher concentrations of TGs and specific species of LPLs.

Unsupervised clustering identified 32 co-regulated lipid modules reflecting distinct metabolic pathways. While many clusters corresponded to conventional lipid classes, several classes were distributed across multiple clusters, indicating the presence of distinct metabolic networks within the same class. This was supported by longitudinal analyses, which revealed divergent age-related trajectories within lipid classes. For example, triacylglycerol-rich clusters show two distinct trajectory patterns. Genome-wide association analyses further showed that 22 of 32 clusters were associated with at least one significant genetic locus.

CONCLUSION:
We present a large-scale, high-coverage, longitudinal lipidomics dataset which demonstrates the strong sex-specificity and individuality of circulatory lipid signatures. Cluster-based analysis reveals a nuanced organization of lipid metabolism that extends beyond conventional lipid classes. The observed genetic associations further support the biological relevance of these lipid networks and provide a framework to link molecular lipid profiles to underlying pathways and cardiometabolic health.

REFERENCES:
1. Beyene HB, Olshansky G, Smith AAT, Giles C, Huynh K, Cinel M, et al. High-coverage plasma lipidomics reveals novel sex-specific lipidomic fingerprints of age and BMI: Evidence from two large population cohort studies. PLoS Biol. 2020 Sep 1;18(9):e3000870. doi:10.1371/JOURNAL.PBIO.3000870 PubMed PMID: 32986697.
2. Tabassum R, Rämö JT, Ripatti P, Koskela JT, Kurki M, Karjalainen J, et al. Genetic architecture of human plasma lipidome and its link to cardiovascular disease. Nature Communications 2019 10:1. 2019 Sep 24;10(1):4329-. doi:10.1038/s41467-019-11954-8 PubMed PMID: 31551469.
3. Medina J, Borreggine R, Teav T, Gao L, Ji S, Carrard J, et al. Omic-Scale High-Throughput Quantitative LC−MS/MS Approach for Circulatory Lipid Phenotyping in Clinical Research. Anal Chem. 2023;95:3168–79. doi:10.1021/acs.analchem.2c02598.