Biological Age Clocks
An umbrella term covering very different methods for estimating the 'true' pace at which an organism is aging — from well-validated second-generation epigenetic clocks to far less proven commercial tests based on glycans or the microbiome. Not all of them carry a comparable level of evidence.
Number of studies
3
Safety
High
Time to effects
Not applicable — this is a category of measurement tools, not an intervention; meaningfully assessing a change in score over time requires repeated measurements months or years apart.
Who it's for
Table of contents
TL;DR
An umbrella term covering very different methods for estimating the 'true' pace at which an organism is aging — from well-validated second-generation epigenetic clocks to far less proven commercial tests based on glycans or the microbiome. Not all of them carry a comparable level of evidence.
- →Helps organize the highly varied and rapidly growing market of 'biological age' tests by actual level of evidence rather than marketing alone
- →Reveals that a single 'biological age' result from one method need not match the result from another — different clocks measure different layers of biology (DNA, proteins, sugars, bacteria) and aren't simple interchangeable substitutes
- →Second/third-generation clocks (epigenetic, proteomic) are used as endpoints (surrogates) in clinical trials of anti-aging interventions, shortening the time needed to assess their potential efficacy
| Definition | An umbrella term covering various mathematical models estimating the pace of aging from biological data |
|---|---|
| Main families | Epigenetic (DNA methylation), proteomic/organ, glycan-based (IgG), microbiome-based |
| Best validated | Second/third-generation epigenetic clocks (PhenoAge, GrimAge) — repeatable links to mortality across many independent cohorts |
| Least mature | Microbiome clocks (OMAA Score, MicroAge) — single publications, no independent replication yet |
| Key distinction | Generation I (trained on chronological age) vs. generation II/III (trained on mortality/morbidity) |
| Large comparison | Mavrommatis et al. 2025 — 14 epigenetic clocks, 18,859 people, 174 endpoints |
| Status | Research and partly commercial tools, none replaces standard medical diagnostics |
Understand
Overview
"Biological age clocks" is an umbrella term for a wide variety of mathematical and laboratory tools whose shared goal is to estimate how fast a given organism is aging — as opposed to chronological age, which is simply the number of years since birth. The problem is that this single label covers methods with wildly different levels of scientific maturity: from epigenetic clocks based on DNA methylation, which have a decade of research behind them across hundreds of thousands of samples and repeatable links to mortality, to brand-new, single-publication models based on the oral microbiome or commercial glycan tests whose independent validation is only just beginning.
The main families of biological age clocks, each described in more detail in its own entry on this site, are: epigenetic clocks (DNA methylation — see epigenetic age acceleration), proteomic clocks based on plasma proteins, including organ clocks that estimate the age of individual organs separately (see organ biological age), tests based on IgG antibody glycosylation (see GlycanAge), and the newest, still very early category of microbiome clocks (see OMAA Score and MicroAge). A separate, complementary concept is intrinsic capacity, promoted by the WHO — this is not a molecular clock but a functional clinical assessment that often correlates with biomolecular clock results, yet measures a different layer of reality (function rather than cellular biology).
Who can realistically benefit from this? People who encounter marketing for commercial 'biological age' tests and want to know which category of method they're looking at and its actual level of validation before paying for a result and treating it as a reliable diagnosis. The key organizing principle: the longer the research history, the larger and more diverse the validation cohorts, and the stronger the link to hard endpoints (death, a specific disease) — independently confirmed by groups other than the method's creators — the more solid the evidence. By this standard, second- and third-generation epigenetic clocks are the best established today, proteomic/organ clocks have solid but younger evidence, and glycan- or microbiome-based clocks remain a promising but still early stage of research.
Mechanism of action
Despite differences in input data, most biological age clocks work on the same general statistical scheme. Researchers collect a large sample of people with known chronological ages and measure some biological signal that changes with age — the methylation level of thousands of CpG sites in DNA, the concentrations of thousands of plasma proteins, the glycosylation profile of IgG antibodies, or the bacterial composition of the microbiome. A machine learning algorithm (most often regularized regression, e.g., elastic net) learns to combine this data into a single value that best matches chronological age or — in more advanced, so-called second- and third-generation clocks — directly matches the risk of death and morbidity, independent of chronological age.
This generational distinction, best described for epigenetic clocks, is key to assessing the credibility of any new clock, regardless of the type of input data: first-generation clocks (e.g., Horvath, Hannum) were trained purely to predict chronological age, making them good age calculators but weaker predictors of health. Second-generation clocks (e.g., PhenoAge, GrimAge) were trained on composite clinical indicators and real mortality data, making them noticeably stronger predictors of disease and death — a large, independent comparison of 14 epigenetic clocks in a sample of nearly 19,000 people (2025) confirmed that second- and third-generation clocks consistently outperformed first-generation clocks in predicting 174 different diseases and all-cause mortality. This same generational logic — the newer the model trained directly on hard endpoints, the stronger the predictor — applies to a lesser extent (given their shorter research history) to proteomic, organ, glycan, and microbiome clocks as well.
Choosing a biological signal
DNA methylation, plasma proteins, IgG glycans, or microbiome composition — each clock family relies on a different type of molecular data.
Training the model on a large cohort
A machine learning algorithm combines thousands of input variables into a single value, trained on a sample of people with known chronological age or known health outcomes.
Choosing the training target (generation)
A model trained on chronological age (generation I) gives an accurate 'age calculator'; a model trained on mortality and morbidity (generation II/III) gives a stronger health predictor.
Validation in independent cohorts
A clock's credibility grows with the number and diversity of independent populations in which its links to hard endpoints have been confirmed by groups other than the method's creators.
Evidence: moderate — based on 3 studies in this database.
Benefits
Common myths
MythAll 'biological age clocks' have a similarly solid level of scientific evidence since they share a similar marketing name.
FactThe level of evidence differs drastically between categories — second/third-generation epigenetic clocks have a decade of research behind them across hundreds of thousands of samples, while some microbiome or glycan clocks today rest on single publications and require independent replication.
MythIf two different 'biological age' tests give a different result for the same person, at least one of them must be wrong.
FactDifferent clocks measure different layers of biology (DNA methylation, plasma proteins, glycans, bacteria) and aren't direct substitutes for each other — a discrepancy in results is expected and partly reflects the fact that aging isn't one single, uniform process.
MythA commercial 'biological age' test result is a precise, individualized medical diagnosis.
FactEven the best-validated clocks base their predictive power on large population studies — a single result has limited precision for a specific person and doesn't replace standard clinical diagnostics.
Check your profile
Not sure which supplements actually make sense for you?
Answer a few short questions about your lifestyle, diet, sleep, and goals. VitMode will build your profile and show supplements worth considering — with reasoning and evidence strength.
Recommendations take your answers and the strength of the scientific evidence into account. A supplement's popularity has no bearing on whether it gets recommended.
Practice
Frequently asked questions
There isn't one universally best clock — it depends on the research question. For predicting mortality and morbidity, second/third-generation epigenetic clocks (e.g., GrimAge, DunedinPACE) and proteomic/organ clocks have the strongest evidence today; glycan and microbiome clocks are promising but have a much shorter history of independent validation.
They don't have to, and often don't — they measure different layers of biology (DNA, proteins, sugars, bacteria), which can age at slightly different rates within the same person. A discrepancy in results doesn't necessarily mean a measurement error.
Biological age clocks rely on molecular biomarkers (DNA, proteins, sugars, bacteria) and require a lab sample. Intrinsic capacity is a WHO concept assessing a person's functional physical and mental capacities (e.g., gait, memory, vision) through a clinical assessment, without needing to analyze biological samples — it's a different layer of measuring aging, partly complementary.
It depends on your expectations — as a tool for tracking your own trends over time (e.g., the impact of a lifestyle change), they can be interesting, especially when based on well-validated second-generation epigenetic clocks. They shouldn't, however, be treated as a precise medical diagnosis or a substitute for clinical tests ordered by a doctor.
What to combine with
Good combinations
Epigenetic Age Acceleration — Epigenetic clocks are today's best-validated family of biological age clocks
Organ Biological Age — Organ clocks break a single score down into components specific to individual organs
Safety
Side effects & contraindications
Possible side effects
Contraindications
No significant contraindications at typical doses.
Is it worth taking?
Who it's for
- People considering buying a commercial 'biological age' test who want to know which category of method they're buying and its actual level of validation
- People interested in the scientific context behind the various types of aging biomarkers described separately on this site
Not for
- No significant contraindications at typical doses.
Evidence
Worth knowing
A large, independent comparison of 14 epigenetic clocks in a sample of nearly 19,000 people (2025) found that second- and third-generation clocks consistently outperformed first-generation clocks in predicting 174 diseases and mortality.
Organ clocks can reveal that one specific organ (e.g., the heart) may be aging noticeably faster than the rest of the body in the same person — information unavailable from a single, averaged score.
The WHO's intrinsic capacity is a separate, functional concept for assessing aging — it isn't based on molecular biomarkers but on a clinical assessment of five domains of function (cognition, vitality, locomotion, psychological well-being, senses).
Studies
DNA methylation-based biomarkers and the epigenetic clock theory of ageing
Moderate evidenceHorvath S, Raj K. · Nature Reviews Genetics · 2018
A foundational review of epigenetic clock theory, describing the different generations of methylation clocks and the logic behind their increasing predictive accuracy for health and mortality.
View studyAn unbiased comparison of 14 epigenetic clocks in relation to 174 incident disease outcomes
Moderate evidenceMavrommatis C, Belsky DW, Ying K, Moqri M, Campbell A, Richmond A, Gladyshev VN, Chandra T, McCartney DL, Marioni RE. · Nature Communications · 2025
A large, independent comparison of 14 popular epigenetic clocks in a sample of 18,859 people as predictors of 174 diseases and all-cause mortality over 10 years of follow-up; second- and third-generation clocks consistently outperformed first-generation clocks.
View studyOrgan aging signatures in the plasma proteome track health and disease
Moderate evidenceOh HSH, Rutledge J, Nachun D, et al. · Nature · 2023
The study introducing 11 organ-aging clocks based on plasma proteomics as a separate, younger family of biological age clocks that complements epigenetic clocks.
View studySources & bibliography
Citations are illustrative for this demo version and require full bibliographic verification by the editorial team before production publication.
Compare with similar entries
About the authors of this entry
Author
dr Anna KowalczykEditor-in-Chief, Molecular Biology
Anna studied molecular biology at the University of Warsaw, then spent eight years after her PhD in a lab researching the mechanisms of cellular aging and autophagy. She stumbled into science journalism almost by accident — frustrated by how easily her field's findings get oversimplified in the media, she started a blog explaining the biology of aging in plain language. That blog became the seed of VitMode. Today Anna oversees the entire editorial process, holding every piece to the same rigor her old lab demanded: primary sources, methodology checks, and honesty about the limits of the evidence. Outside work, she's a dedicated boulderer.
121 publications on this site
Medical review
dr Piotr ZielińskiEndocrinologist
Piotr has practiced endocrinology for more than fifteen years, mostly in male hormonal disorders and metabolic health. He joined VitMode as a scientific consultant because, as he jokes, he got tired of explaining the same testosterone questions at every appointment and decided to write the answers down properly, once. He reviews content on hormone therapy, supplement pharmacology and drug interactions, making sure articles never turn into encouragement to self-supplement in situations that genuinely need diagnostics and medical supervision. His professional motto — "evidence first, enthusiasm second" — has come up more than once with a patient who arrived with a supplement plan they found online.
174 publications on this site
Related entries
4.1Epigenetic Age Acceleration
The difference between an age estimated from a DNA methylation pattern and chronological age — one of the most statistically well-documented aging biomarkers, strongly linked to mortality in population studies, though still of limited diagnostic value for a single individual.
4.2Organ Biological Age
The concept that individual organs — the heart, brain, liver, or kidneys — can age at markedly different rates within the same person, measurable through analysis of tissue-specific plasma proteins rather than a single, averaged 'biological age' score.
4.1Intrinsic Capacity
A World Health Organization (WHO) concept defining 'true' functional aging as the sum of five domains of physical and mental ability — an alternative to molecular biomarkers, assessed in a clinical exam rather than a lab.
4.0GlycanAge
A commercial biological-age test based on the glycosylation pattern of IgG antibodies — it grew out of real research from the Croatian Genos/Gordan Lauc group, but independent validation of the product itself remains limited.
4.0OMAA Score
The OMAA Score (Oral Microbiome Aging Acceleration) is a numerical index calculated as the difference between the age predicted from an oral microbiome sample and a person's actual age — in the study that introduced it, it was linked to elevated mortality and frailty risk.
3.9MicroAge
A machine-learning model that estimates biological age from the composition of the saliva microbiome — a very early, emerging approach to aging biomarkers, distinct from the mortality-focused OMAA Score.
4.4Telomeres and Telomerase
Protective 'caps' on the ends of chromosomes that shorten with every cell division — one of the most recognizable, though still imperfect, biomarkers of cellular aging.
4.1Oral Microbiome Aging
The composition of bacteria colonizing the mouth changes in a predictable way with age — predictable enough that researchers have started building models to estimate biological age from a simple saliva sample.
Comments (2)
- KW
Kasia W. 2 weeks ago
Very clearly explained, especially the interactions section — I hadn't seen it laid out this well anywhere else.
- MT
Marek T. a month ago
Are you planning to update this with the newest study from this year? I saw an interesting meta-analysis.