Biological age. The honest version.

Two kinds of age

The underlying idea is uncontroversial and well supported. Ageing, biologically, is the accumulation of molecular and cellular damage over time — and the World Health Organization's own standing reference on ageing describes those changes as neither linear nor consistent, and only loosely tied to how many years a person has lived.

The illustration it gives is the useful one: some people in their eighties have physical and mental capacities similar to many people in their thirties, while others decline much earlier — and much of that variation is attributed to environment and personal circumstances rather than to years lived.

So *chronological age is a poor individual predictor of health and function* is a well-founded statement. What does not follow from it is that anybody can currently put a reliable number on the difference for one particular person. That is the whole subject of the rest of this page.

What researchers actually measure

When ageing research talks about biological age, it is usually talking about molecular measures taken from blood or tissue — not about anything a wearable can see.

The best-known family is the epigenetic clocks. A 2018 review in Nature Reviews Genetics set out the concept: patterns of DNA methylation can produce accurate age estimates across tissues and across the whole life course, and the review proposes a theory linking those patterns to biological ageing. Two things are worth holding onto about it. Its strongest established claim is accuracy against chronological age — that these measures can guess how old you are. The claims about measuring the *rate* of biological ageing are framed there as hypothesis and research direction. And the review is co-authored by the researcher who developed the original method, which makes it the reference explanation of the field rather than an independent assessment of it.

More recent work has tried to measure pace rather than age. A 2022 paper in eLife describes a measure built from blood DNA methylation and trained on how 19 indicators of organ-system integrity declined within individuals across four timepoints spanning two decades in a birth-cohort study. It was engineered specifically for reliability, and it was associated with illness, disability and mortality in the groups studied.

Note what that is and is not. It is a research biomarker derived from blood, with associations observed at group level. It is not derived from any wearable signal, it does not establish what any individual's number means for them, and it was trained in a single national birth cohort, which leaves generalisation across populations an open question its own authors acknowledge.

What the field itself says is not settled.

The most important source on this page is a 2019 critical appraisal in Genome Biology written by 21 of the field's own principal investigators — including the developer of the original clock. That authorship matters: this is not scepticism from outside.

It confirms that clocks are a highly accurate molecular correlate of chronological age, and then lists what is unresolved. There is no single measure or gold standard of biological ageing to validate any clock against. Most of the changes clocks pick up are small and may reflect shifts in which cells are present in a sample rather than change inside the cells themselves. Statistical significance does not establish that methylation changes drive ageing rather than merely accompanying it. And the underlying genetic and epigenetic data are strongly biased toward populations of European ancestry, which leaves accuracy across other populations unknown.

Its sharpest statement is the one most often ignored: clock associations predict all-cause mortality at a population level, but not at an individual level. The paper explicitly warns against clocks becoming a de facto measure of an individual's health, and raises ethical concerns about predicting one person's ageing trajectory.

The field's own consensus on standards, published in Nature Medicine in 2024 by around thirty leading researchers, points the same way: it sets out how biomarkers of ageing *should* be validated before use, and treats rigorous validation as the precondition for real-world use rather than as something already achieved.

A change of several years can be measurement noise.

This is the finding that should make anyone cautious about before-and-after messaging, and it comes from inside the field again.

A 2022 paper in Nature Aging tested how reliable the underlying data are by running the same biological sample twice and comparing the answers. Across six prominent clocks, technical noise alone produced differences of up to nine years between two measurements of the same sample. The authors put it plainly: it would be misleading for a clock to indicate that a person had aged nine years when the difference comes entirely from technical variation — and they state that a change in someone's measured biological age after starting a treatment or a lifestyle change would, on that basis, be suspect.

That paper's own contribution is a fix — retrained versions of those clocks that bring most repeat measurements within about a year and a half — so it should not be read as *epigenetic clocks do not work*. It should be read as: a few years of movement in an age number is not, by itself, evidence that anything happened to you.

The part nobody advertises

Everything above is about blood, cells and laboratories. Now the consumer question, which has a short and uncomfortable answer.

No consumer wearable-derived age number is validated against anything. The source pack behind this page went looking for peer-reviewed validation of a wearable age estimate against an epigenetic or biomarker measure of ageing, and found none that met its quality bar. Not a weak one. None.

What these numbers descend from is a different and more modest idea: *fitness age*. The best-documented version, published in 2011 in Medicine & Science in Sports & Exercise, estimates cardiorespiratory fitness from a handful of easily collected facts — age, a body measurement, self-reported leisure-time activity and resting heart rate — and was developed in 4,637 healthy adults, explaining 61% of the variation in men and 56% in women. Its own authors describe it as suitable for a rough assessment of cardiorespiratory fitness.

That is a useful thing. It is also, precisely, a fitness estimate expressed on an age scale — not a measure of biological ageing, and never validated as one. The credibility risk in this whole topic is content that blurs the two, letting evidence about molecular clocks lend authority to a number derived from a wrist. This page will not do that, and it is worth noticing which pages do.

For the avoidance of doubt: the estimate described above is not how Omnisio's Bio-Age is arrived at. How it is arrived at is not something we publish.

What an age number must never be read as.

Some of these are claims made routinely in this category. None of them is supported by the research described on this page.

• Not a measure of how fast you are ageing. That claim belongs to research biomarkers taken from blood, and even there it is framed as an area of active work rather than a settled capability.
• Not a life-expectancy figure. Clock associations with mortality hold at population level and explicitly not at individual level. A number is not a forecast about you.
• Not evidence that anything you did worked. A movement of a few years can be measurement noise even in laboratory measures engineered for reliability.
• Not equivalent to an epigenetic clock. A wearable-derived number and a DNA-methylation measure are different objects with different evidence behind them, and the evidence for one does not transfer to the other.
• Not a diagnosis, a screening result or a risk assessment. No source on this page supports any of those readings, and phrasing an age number that way moves it into territory it has no evidence to occupy.
• Not equally accurate for everybody. The underlying research data are strongly biased toward populations of European ancestry, and the field itself names that as unresolved.

Where Omnisio fits

Omnisio shows a Bio-Age reading: the long arc, the parts it is built from, and how fast it is moving. Having spent this page establishing what an age number cannot be, here is what ours is.

It is a wellness reading, and — like every other reading in the app — an estimate derived from consumer-grade sensors rather than a medical-grade measurement. It is not an epigenetic clock, it is not equivalent to one, and nothing in the research on this page validates it. While it is still settling, Omnisio marks it as still settling, and the mark comes off the moment it is not — which is the same instinct as giving no score at all on a night that could not be read: an unconfident number should look unconfident.

Omnisio is a wellness product for general lifestyle use. Neither the app nor the devices used with it hold FDA, CE-MDR or TITCK medical device clearance, and nothing in it is intended to diagnose, treat, cure or prevent any disease. Your Bio-Age is not a statement about your health, your risk, or how long you will live, and Omnisio does not present it as one.

Where this comes from

Seven published sources, named plainly. Each was confirmed live on 8 September 2026.

• Source: World Health Organization, 2025. Standing fact sheet on ageing and health: biological ageing as accumulated molecular and cellular damage, with those changes described as neither linear nor consistent and only loosely associated with age in years.
• Source: Nature Reviews Genetics, 2018. The reference review of epigenetic clocks and the clock theory of ageing. Its established claim is accuracy against chronological age; claims about the rate of biological ageing are framed as theory and research direction. Co-authored by the developer of the original method.
• Source: eLife, 2022. A blood DNA-methylation measure of the pace of ageing, trained on within-individual decline in 19 indicators of organ-system integrity across four timepoints over two decades in one birth cohort. Associations with illness, disability and mortality are at group level.
• Source: Genome Biology, 2019. Critical appraisal by 21 of the field's principal investigators: no gold standard to validate clocks against, most changes small and possibly reflecting cell composition, correlation not causation, ancestry bias in the underlying data, and prediction of mortality at population but not individual level.
• Source: Nature Aging, 2022. Reliability study across six prominent clocks: technical variation alone produced differences of up to nine years between repeat measurements of the same sample, and the authors state that a change after a treatment or lifestyle change would on that basis be suspect. Retrained versions bring most repeats within about a year and a half.
• Source: Nature Medicine, 2024. Consensus-style review by around thirty researchers on how biomarkers of ageing should be validated, treating validation standards as still being established rather than met.
• Source: Medicine & Science in Sports & Exercise, 2011. The fitness-age estimate: cardiorespiratory fitness estimated without an exercise test in 4,637 healthy adults, explaining 61% of variation in men and 56% in women, described by its authors as suitable for a rough assessment of fitness. It is a fitness estimate on an age scale, not a measure of biological ageing.

Questions