Multi-Omics in Precision & Preventive Medicine: From Molecular Insights to Better Health Outcomes

Healthcare is gradually moving beyond diagnosing and treating disease once symptoms appear. Advances in molecular diagnostics now make it possible to characterise individual biology in greater depth, identify some risk factors earlier and inform more personalised preventive strategies.

Yet no single biomarker captures this complexity. Genetic predisposition, molecular regulation, immune activity, metabolism and physiological function are interconnected, and changes in one system may influence others long before a clinically recognisable disease develops. Multi-omics offers a framework for connecting these layers — from inherited susceptibility to measurable physiological function.

This article explains what multi-omics measures, how it may contribute to three clinical applications — disease prevention and risk stratification, personalised health optimisation, and healthy ageing — and where the evidence is established, emerging or still uncertain.

Key takeaways

• Multi-omics connects layers of biology, from inherited DNA through molecular regulation to clinical and functional phenotype. Each layer changes at its own speed and answers a different question.
• Integrating the layers is feasible, but the deep-profiling studies that show it have been small and largely descriptive.
• In disease prevention and risk stratification, molecular data are studied as a complement to established clinical assessment. Most models show associations at population level, not proven benefit for individuals.
• In personalised health optimisation, pharmacogenetics is the most mature example. Elsewhere, the working principle is to measure a baseline, make a change and measure again.
• In healthy ageing, twelve interconnected hallmarks show that ageing is not one process. Ageing clocks measure different biology, and no single clock is a full measure of ageing.
• Responsible use rests on analytical validity, clinical validity and clinical utility, with results interpreted in the full clinical picture.

Understanding multi-omics: from upstream biology to downstream function

Multi-omics refers to the integrated analysis of multiple classes of biological molecules and their interactions. Rather than examining isolated measurements, the approach seeks to connect different levels of biological organisation.

Multi-omics layers, exposures and phenotype

Inherited biology flows downward through the molecular layers to phenotype; exposures act on every layer except the germline genome.

Upstream — inherited biology

Genomics

Molecular regulation

Epigenomics

Transcriptomics

Proteomicsincl. glycomics · protein glycosylation

Metabolomics

Downstream — physiological expression

Clinical & functional phenotype

Exposome

Diet, medication, lifestyle

Act directly on the epigenome, transcriptome, proteome and metabolome, on the microbiome, and on phenotype.

Host–microbiome interface

Microbiome

Interacting microbial ecosystem

Exchanges proteins and metabolites with the host; microbial metabolites such as short-chain fatty acids also modify host epigenetic marks.

shapes composition

Each layer reflects a different part of biology and changes at its own speed:

Layer	What it reflects	How quickly it changes
Genomics	Genetic predisposition, biological pathways and medication response	Essentially stable over a lifetime
Epigenomics	DNA methylation, regulatory states and ageing-associated molecular patterns	Slowly, with age, environment and lifestyle
Transcriptomics	Gene expression: which genes are being read into RNA in a given cell or tissue	Within hours to days, in response to signals, stress and daily rhythms
Proteomics	Protein abundance, signalling and biological responses	Acute-phase proteins such as C-reactive protein (CRP) within hours to days; many other proteins over weeks to months
Glycomics	Sugar structures (glycans) attached to proteins such as antibodies, which take part in immune regulation	Relatively stable over months; shifts with age, hormonal change and inflammation
Metabolomics	Metabolic pathways, substrate utilisation and metabolic phenotype	Over hours to weeks, with diet, medication and activity
Microbiome	The microbial communities of the gut and their metabolic activity	Within days with diet, antibiotics or illness, around a core composition that is more stable
Clinical & functional phenotype	Cardiometabolic health, organ function, physical performance and clinical outcomes	Varies; many measures are followed routinely in clinical care

We describe the glycome in more depth in our note on inflammaging and immune ageing.

Why integration matters

Biological systems do not work in isolation. A genetic variant may have little effect until gene regulation, diet or medication brings it into play. The same inflammatory process can appear as altered gene regulation, shifted protein levels and changed metabolite profiles, each telling part of the story. Integrating the layers aims to show patterns that no single layer reveals on its own [1].

Several deep-profiling studies have explored this in practice. In an early example, one individual was profiled repeatedly across genomic, transcriptomic, proteomic and metabolomic layers over 14 months, and the profiles shifted during two viral infections and the emergence of elevated blood glucose [2]. Later studies followed larger groups: 108 individuals with dense longitudinal data and lifestyle coaching [3], and around a hundred participants profiled over several years, many of whom received clinically relevant findings [4].

These studies show that integrated profiling is feasible, but they were small, intensive and largely descriptive. They were not designed to show that multi-omic profiling improves long-term health outcomes.

Three clinical applications

In practice, the layers matter in three ways. They may add to established clinical assessment when deciding who needs closer follow-up, inform more individual choices about medication and lifestyle, and describe how biological ageing unfolds in a particular person. The evidence differs between the three, so each section ends with a short note on where it stands.

Disease prevention and risk stratification

Risk stratification groups people by their estimated likelihood of developing a condition, so that monitoring and preventive care can be matched to need. In current practice it rests on established clinical factors: age, sex, family history, blood pressure, lipids, glucose, smoking and validated clinical scores.

Molecular layers are being studied as additions to this foundation:

• Genomics. Rare monogenic variants with established clinical management are already used in defined clinical contexts. Polygenic scores, which add up the small effects of many common variants, can identify people at the extremes of inherited susceptibility for some common conditions [5]. Added to established clinical scores, however, they have generally improved prediction only modestly [6], and they perform less well in people of non-European ancestry, because most of the underlying studies were conducted in European-ancestry populations [7].
• Proteomics. Large plasma protein panels have been used to build models associated with several health outcomes in cohort studies [8].
• Metabolomics. In a large population cohort, metabolomic profiles from a single blood sample were associated with the later onset of many common conditions, and in some cases added information to established predictors [9].
• Epigenomics. Some DNA methylation-based measures, such as GrimAge, are associated with mortality and age-related morbidity in population studies [10].

The common pattern is association at population level. A model that separates groups well in a cohort does not automatically give a reliable estimate for an individual, and it does not show that acting on the result changes outcomes. Established clinical biomarkers remain the basis of preventive assessment, with molecular data as a possible complement.

Evidence status: established for clinical factors and validated clinical scores, and for monogenic variants in defined clinical contexts. Emerging for polygenic scores and molecular prediction models. Whether adding them improves outcomes for individuals is still uncertain.

Personalised health optimisation

Pharmacogenetics is the most mature example of personalisation. For selected gene–drug pairs, international consortia publish guidelines on how genotype should inform drug choice or dosing [11]. In a large European implementation trial, genotype-guided prescribing with a 12-gene panel was associated with fewer clinically relevant adverse drug reactions in patients with an actionable result [12].

Beyond pharmacogenetics, the evidence is earlier. Studies combining molecular profiles with coaching or individual feedback have reported changes in clinical biomarkers over months [3], but they were not randomised against usual care. Metabolic and protein markers can describe a person’s current state, such as lipid patterns, fatty acid balance or inflammatory activity.

A practical principle follows: measure a baseline, make a change, and measure again. A shift in a molecular marker shows that biology has responded. On its own, it does not show that health outcomes will improve.

Evidence status: established for pharmacogenetic guidance on selected gene–drug pairs. Emerging for repeat molecular measurements to follow response to lifestyle or treatment changes. Whether multi-omic-guided personalisation improves long-term outcomes is still uncertain.

Healthy ageing

Ageing is not just one process. In 2023, López-Otín and colleagues updated the framework of ageing biology to twelve interconnected hallmarks [13]: genomic instability; telomere attrition; epigenetic alterations; loss of proteostasis; disabled macroautophagy; deregulated nutrient-sensing; mitochondrial dysfunction; cellular senescence; stem cell exhaustion; altered intercellular communication; chronic inflammation; and dysbiosis.

Primary hallmarks

Causes of molecular damage

• Genomic instability
• Telomere attrition
• Epigenetic alterations
• Loss of proteostasis
• Disabled macroautophagy

Antagonistic hallmarks

Responses to that damage

• Deregulated nutrient-sensing
• Mitochondrial dysfunction
• Cellular senescence

Integrative hallmarks

Consequences for tissue function

• Stem cell exhaustion
• Altered intercellular communication
• Chronic inflammation
• Dysbiosis

The authors group the hallmarks into three categories. Primary hallmarks are causes of molecular damage. Antagonistic hallmarks are responses to that damage, which may be protective at first but become harmful when they persist. Integrative hallmarks arise when the accumulated damage can no longer be compensated, and they affect how tissues function. The hallmarks influence one another, and none of them explains ageing on its own.

They also play out in different biological layers. Genomic instability and epigenetic alterations concern DNA and its regulation. Loss of proteostasis and senescence-associated signalling concern proteins. Deregulated nutrient-sensing and mitochondrial dysfunction concern metabolism, and dysbiosis concerns the gut microbiome. No single layer reflects all twelve, which is why a multi-omics approach is relevant to ageing. How well any individual measurement reflects a given hallmark in a living person is still an active research question.

Ageing research increasingly treats biological age as something that can be estimated, not only counted in years. Several layers now have their own ageing clocks:

• Epigenetic clocks estimate age or the pace of ageing from DNA methylation patterns [14, 15].
• Proteomic models estimate the ageing of individual organs from plasma proteins. In cohort studies, organs that appeared older than expected were associated with later disease and mortality [16].
• Glycan-based [17], metabolomic [18] and clinical-chemistry models [19] capture further aspects of ageing.

Deep longitudinal profiling suggests that people may age along different biological pathways, described as “ageotypes”, such as metabolic, immune, liver and kidney ageing [20].

Different clocks measure different biology, and their estimates often disagree for the same person [20, 21]. No single clock is a full measure of ageing. Researchers in the field have described the validation steps ageing biomarkers need before they can be used to evaluate longevity interventions [22]. Until then, a lower biological age estimate after an intervention is a signal worth following, not proof that ageing has slowed.

Evidence status: the hallmarks are a widely used research framework. Ageing clocks and ageotypes are emerging research and monitoring tools. Whether changing a biological age estimate changes health is still uncertain.

Where the evidence stands

Evidence status for the three clinical applications of multi-omics

Application	Established	Emerging	Uncertain
Disease prevention and risk stratification	Clinical factors and validated clinical scores. Monogenic variants in defined clinical contexts.	Polygenic scores alongside clinical assessment. Proteomic, metabolomic and methylation-based prediction models.	Whether adding molecular models improves outcomes for individuals. Performance across ancestries.
Personalised health optimisation	Pharmacogenetic guidance for selected gene–drug pairs.	Repeat molecular measurements to follow response to lifestyle or treatment changes.	Whether multi-omic-guided personalisation improves long-term outcomes or is cost-effective.
Healthy ageing	The hallmarks of ageing as a research framework.	Epigenetic, proteomic and glycan-based ageing clocks. Ageotypes.	Whether changing a biological age estimate changes health. Which clocks are suitable for evaluating interventions.

Using multi-omics responsibly

For clinicians and individuals alike, three questions apply to every measurement:

• Analytical validity: is the marker measured accurately and reproducibly?
• Clinical validity: is it meaningfully associated with relevant outcomes, in a population like the person being tested?
• Clinical utility: does using the result lead to better decisions or outcomes?

Integration brings its own challenges: different measurement platforms, different timescales, missing data, and far more variables than participants. Statistical models can find patterns that do not replicate. Validation in independent populations therefore matters as much as the measurements themselves.

A few practical points follow:

• More data does not automatically mean better decisions. Each additional measurement adds the possibility of uncertain or incidental findings.
• Results should be interpreted in the context of the full clinical picture, ideally with a healthcare professional.
• Repeat measurements say more than a single snapshot, provided the analytical and biological variability of the marker is known.
• Genetic and other molecular data are sensitive personal data and need appropriate consent and protection.

How BioAge Diagnostics’ tests relate to these layers

BioAge Diagnostics offers measurements in several of the layers described here. Each test reports on one layer:

Layer	Test	Sample
Genomics	Genetics & Pharmacogenetics	Saliva
Epigenomics	Advanced Biological Age & Epigenetics	Blood
Proteomics	Advanced Proteomics	Buccal swab
Glycomics	Immune Ageing & Glycan Biology	Finger-prick blood
Metabolomics	Metabolic Resilience	Finger-prick blood

The Advanced Proteomics test analyses proteins from buccal cells, which is a different sample and a different set of proteins from the plasma proteomics studies cited above. Those studies therefore do not establish the clinical validity of buccal-cell protein profiles, and the evidence for this sample type is at an earlier stage than for plasma proteomics. A UK Biobank research project is investigating an explainable machine-learning model for proteomics-based biological age from buccal swabs [23]; this is an ongoing research project, not a published validation study. Results from separate tests are reported separately; combined, they are not a validated composite score. Non-diagnostic. We recommend discussing your results with a healthcare professional.

Conclusion

Multi-omics offers a way to connect inherited susceptibility, molecular regulation and physiological function, and to follow how they change over time. Its most established uses today are specific: pharmacogenetic guidance for selected drugs, alongside the established clinical biomarkers that remain the foundation of preventive care. Broader applications in risk stratification, health optimisation and healthy ageing are promising, but much of the evidence is still associative.

The value lies in careful interpretation: knowing what each layer measures, how reliable it is, and where the evidence stops. More measurements are useful when they lead to better questions and better follow-up, and they work best alongside clinical judgement, not in place of it.

Scientific references

1. Hasin Y, Seldin M, Lusis A. Multi-omics approaches to disease. Genome Biol. 2017. doi:10.1186/s13059-017-1215-1
2. Chen R, et al. Personal omics profiling reveals dynamic molecular and medical phenotypes. Cell. 2012. Cell full text
3. Price ND, et al. A wellness study of 108 individuals using personal, dense, dynamic data clouds. Nat Biotechnol. 2017. PubMed 28714965
4. Schüssler-Fiorenza Rose SM, et al. A longitudinal big data approach for precision health. Nat Med. 2019. PubMed 31068711
5. Khera AV, et al. Genome-wide polygenic scores for common diseases identify individuals with risk equivalent to monogenic mutations. Nat Genet. 2018. PubMed 30104762
6. Elliott J, et al. Predictive accuracy of a polygenic risk score-enhanced prediction model vs a clinical risk score for coronary artery disease. JAMA. 2020. PubMed 32068818
7. Martin AR, et al. Clinical use of current polygenic risk scores may exacerbate health disparities. Nat Genet. 2019. PubMed 30926966
8. Williams SA, et al. Plasma protein patterns as comprehensive indicators of health. Nat Med. 2019. PubMed 31792462
9. Buergel T, et al. Metabolomic profiles predict individual multidisease outcomes. Nat Med. 2022. PubMed 36138150
10. Lu AT, et al. DNA methylation GrimAge strongly predicts lifespan and healthspan. Aging (Albany NY). 2019. PMC6366976
11. Relling MV, Klein TE. CPIC: Clinical Pharmacogenetics Implementation Consortium of the Pharmacogenomics Research Network. Clin Pharmacol Ther. 2011. PubMed 21270786
12. Swen JJ, et al. A 12-gene pharmacogenetic panel to prevent adverse drug reactions: an open-label, multicentre, controlled, cluster-randomised crossover implementation study. Lancet. 2023. doi:10.1016/S0140-6736(22)01841-4
13. López-Otín C, et al. Hallmarks of aging: an expanding universe. Cell. 2023. doi:10.1016/j.cell.2022.11.001
14. Horvath S. DNA methylation age of human tissues and cell types. Genome Biol. 2013. doi:10.1186/gb-2013-14-10-r115
15. Belsky DW, et al. DunedinPACE, a DNA methylation biomarker of the pace of aging. eLife. 2022.
16. Oh HS, et al. Organ aging signatures in the plasma proteome track health and disease. Nature. 2023. doi:10.1038/s41586-023-06802-1
17. Krištić J, et al. Glycans are a novel biomarker of chronological and biological ages. J Gerontol A Biol Sci Med Sci. 2014. doi:10.1093/gerona/glt190
18. van den Akker EB, et al. Metabolic age based on the BBMRI-NL ¹H-NMR metabolomics repository as biomarker of age-related disease. Circ Genom Precis Med. 2020.
19. Levine ME, et al. An epigenetic biomarker of aging for lifespan and healthspan. Aging (Albany NY). 2018.
20. Ahadi S, et al. Personal aging markers and ageotypes revealed by deep longitudinal profiling. Nat Med. 2020.
21. Belsky DW, et al. Eleven telomere, epigenetic clock, and biomarker-composite quantifications of biological aging: do they measure the same thing? Am J Epidemiol. 2018.
22. Moqri M, et al. Biomarkers of aging for the identification and evaluation of longevity interventions. Cell. 2023. doi:10.1016/j.cell.2023.08.003
23. UK Biobank. Using explainable machine learning to provide a proteomics-based biological age from buccal swabs. Approved research project. ukbiobank.ac.uk

Medical disclaimer

This article is intended for professional education and general health information. It describes published research and does not describe the performance of any individual test. Multi-omic prediction models and ageing clocks are emerging tools. They are not established diagnostic tools and are not validated for guiding individual treatment. The article does not replace advice from a healthcare professional.

Contact. Questions about this article or our tests: info@bioagediagnostics.com or +45 22 70 47 24.