Biological age: are the tests worth it?
In short: your "biological age" is an estimate of how worn your body is, most often calculated from chemical marks on DNA (epigenetic clocks). In large studies, these clocks predict well the risk of disease and death in a group of people. But for an individual, the tests sold online (about 200 to 450 euros) remain unreliable: the same sample can give results several years apart, and no standard governs them. Serious studies show that these clocks can be slowed a little (calorie restriction, multivitamins in people over 60), but the effect amounts to a few months and we do not yet know whether it lengthens life. Our advice: keep your money for what already measures your health, such as blood pressure, a standard blood test, your strength and your walking speed.
Updated 11 November 2026.
"I'm 55, but my body is 48." Lines like this are everywhere on social media, test result attached. Companies sell saliva or blood kits that promise to reveal your "real" age, and sometimes the supplements meant to bring it down. Behind the marketing lies real science, born some fifteen years ago and booming in 2026. Here is what it knows, what it does not know yet, and what it means for you.
The verdict at a glance
| The promise | What studies show | Level of evidence |
|---|---|---|
| "The test reveals your real age" | Good prediction for groups; a margin of error of several years for one person | High (cohort studies), low for individual use |
| "Results are comparable from one lab to another" | False: no standard, unpublished methods, large gaps | Moderate (technical studies, journalists' experiments) |
| "You can get 3 years younger in 8 weeks" | Small pilot study, not confirmed | Low |
| "A multivitamin slows ageing" | About 4 months gained over 2 years on some clocks, in people in their seventies | Moderate (one randomised trial, small effect) |
| "Eating less slows ageing" | 2 to 3% slower ageing on one clock, over 2 years | Moderate (one randomised trial) |
| "Slowing the clock lengthens life" | Plausible, not yet demonstrated | Not proven |
Chronological age, biological age: what are we talking about?
Your chronological age is the one on your identity card. Your biological age is an estimate of the state of your body: two people aged 70 can be in very different health, one still running, the other struggling to climb the stairs.
The problem is that ageing cannot be measured with a single instrument. Researchers have therefore built indicators by combining measurements that change with age: blood tests, muscle strength, lung capacity and, above all since 2013, DNA methylation.
How epigenetic clocks work
DNA is the same in all our cells and hardly changes during our lifetime. But it carries small chemical tags, methyl groups, which attach at certain places and help switch genes on or off. This is called epigenetics: what happens "on top of" the genes.
With age, these tags shift in a fairly regular way. Researchers therefore took thousands of blood samples, measured methylation at hundreds of thousands of places on the DNA, and asked a statistical program to find the combination that best predicts age. The result is an epigenetic clock: a formula that turns a blood (or saliva) sample into an estimated age.
There are now dozens of them, usually grouped into three generations.
| Generation | Examples | What they predict |
|---|---|---|
| 1st (2013) | Horvath and Hannum clocks | Chronological age. A gap with real age is interpreted as "acceleration". |
| 2nd (2018-2019) | PhenoAge, GrimAge | The risk of disease and death, based on health markers and smoking |
| 3rd (2022) | DunedinPACE | The speed of ageing, like a speedometer rather than an odometer |
What the clocks can do
In large studies, the finding is solid: people whose clock "runs ahead" have, on average, more age-related diseases and higher mortality, even at the same chronological age. Second- and third-generation clocks are the most informative.
Two studies published in 2026 refined this picture.
- The pace matters as much as the starting point. In Tuscany, the InCHIANTI study followed 699 adults for up to 24 years, with several blood samples. Those whose clock accelerated most over time had a higher risk of death, regardless of their epigenetic age at the start. Level of evidence: observational, good quality.
- Not all clocks measure the same thing. A team from the University of Southern California analysed blood from 3,227 participants in the large US Health and Retirement Study, measuring both methylation and gene activity. The five most widely used clocks reflect partly different processes: immunity and inflammation for some, metabolism and cell growth for others. Two tests that do not rely on the same clock are therefore not quite talking about the same "age".
That is the great strength of these tools: in research, they make it possible to test a treatment in two years instead of waiting thirty years to see who falls ill.
The problem with tests sold online
Moving from a research tool to an individual test is another matter. Several obstacles remain.
Technical noise. In 2022, a team at Yale University measured the same blood sample several times. With six common clocks, results varied by up to 9 years from one measurement to the next. As one of the researchers put it, a test could say 50, then 59 on the same blood. The team developed a corrected version (based on "principal components") that brings the gap down to under 1.5 years for most samples. The laboratory still has to use it.
Gaps between companies. Several public experiments have shown this. The American biologist Matt Kaeberlein, an ageing specialist, sent 8 kits to 4 companies on the same day: his biological ages ranged from 43 to 61. The editorial team of the specialist site Lifespan.io found a gap of nearly 15 years by sending the same sample twice to the same company.
No standards. Unlike medical tests, these tests are not validated for clinical use. Most companies do not say which clock they use or how they process the sample. At a precision medicine conference in 2026, specialists in the field acknowledged that there are still no reference standards for obtaining the same result in two laboratories.
The advice at the end. Whether your age comes out "young" or "old", the report's recommendations are the same: move, eat well, sleep, don't smoke. Some companies add a range of supplements, an obvious conflict of interest.
Price and data. Expect to pay about 200 to 450 euros per test, not reimbursed by compulsory health insurance. And you are handing your DNA, sensitive health data, to a company often based outside Europe: read its privacy policy before sending anything.
In an interview clip with the biologist Rhonda Patrick, Morgan Levine explains why the first clocks gave such variable results, and how her team reduced this noise.
Can you slow your biological age?
That is the question that sells. A few randomised trials (trials in which chance decides who receives the intervention, which makes it possible to conclude there is an effect) give cautious answers.
Eating less. In the US CALERIE trial, 220 healthy adults of normal weight or slightly overweight were randomly assigned to cut their calories by 25% for two years, or to eat normally. The restriction group saw its ageing slow by 2 to 3% according to the DunedinPACE clock. The other clocks did not move. According to the authors, such a gap would correspond in other studies to 10 to 15% lower mortality, but this link was not tested in the trial itself. Level of evidence: randomised trial, small effect.
A multivitamin in people in their seventies. In March 2026, the journal Nature Medicine published an analysis of the US COSMOS trial: 958 people aged about 70 took a common multivitamin, a cocoa extract or a placebo for two years. The multivitamin slowed two clocks (PhenoAge and GrimAge, in their corrected versions), equal to about four months less biological ageing over two years. The effect was clearer in those whose clock was already running fast at the start. Cocoa changed nothing. Several experts stressed that the effect, while statistically significant, was very small, in mostly white participants in good health. Level of evidence: randomised trial, pre-specified analysis, small effect.
The stress of fearing old age. A New York University study published in early 2026 asked 726 American women about their anxiety about ageing. Those most worried about their future health had a faster DunedinPACE clock; fear of losing attractiveness or fertility had no measurable effect. But the study only takes a snapshot at one point in time, and the link weakened markedly once smoking, alcohol and physical activity were taken into account. Level of evidence: observational, causation not established.
What these studies have in common: the effects are real but modest, and they point towards advice we already know.
Does the body age in bursts?
In 2024, a Stanford University team followed 108 adults aged 25 to 75 in California, with repeated samples over sometimes several years: blood, stool, skin, nose, mouth. Of more than 135,000 molecules and microbes analysed, about 81% changed in a non-linear way, with two peaks of change around age 44 and around age 60.
Around 44, the changes mainly concerned the metabolism of fats, alcohol and caffeine, the heart, the skin and the muscles. Around 60, immunity and the kidneys were added. The peaks also appeared in men, suggesting that menopause is not the only explanation.
The authors remain cautious: the sample is small, and these peaks could reflect lifestyle changes at these ages rather than an internal clock. Level of evidence: observational, small cohort. The main lesson: your forties and sixties are good times to take stock with your doctor.
Does this concern me? Should I take a test?
| Your situation | Our view |
|---|---|
| Simple curiosity | Why not, if budget is not an issue. Take the figure as an indication, not a diagnosis. |
| You want to track the effect of your efforts | Only with the same laboratory, the same clock (ideally a corrected version or DunedinPACE), and measurements at least a year apart. A difference of one or two years means nothing. |
| The test then offers you supplements | Be wary: the seller is interpreting the result. |
| You feel anxious about ageing | Better to skip it: a bad figure can worry you for no reason, and anxiety itself does not help. |
| You have symptoms or an illness | The test is not a medical tool. See your doctor. |
What already measures your ageing, at little cost
Several simple measurements predict future health as well as, or better than, a DNA test, and they tell you what to do.
- Blood pressure. Blood pressure that is too high wears down the arteries for years without any symptoms. See our article Taking your blood pressure: why and how?.
- A standard blood test. Blood sugar, cholesterol, kidney function, inflammation: these reimbursed tests say a lot. The PhenoAge clock was in fact built from nine routine blood tests.
- Walking speed. In older people, slow walking is one of the best signals of a higher risk of losing independence and of death, according to an analysis of 34,000 people published in JAMA in 2011.
- Hand strength. In the international PURE study, every 5 kg loss of grip strength was associated with a 16% higher risk of death. A dynamometer at the physiotherapist's or doctor's is enough to measure it.
- Breath and stamina. Climbing two flights of stairs without stopping, walking briskly for half an hour: these are concrete benchmarks.
- Waist size, which reflects fat around the organs.
And to slow ageing, the levers are well known, and they are the same ones the tests invoke: move every day, keep your muscles working, eat a varied, little-processed diet, don't smoke, limit alcohol, sleep well, stay socially connected, and have your blood pressure, weight and blood sugar checked. To understand what happens in ageing cells, see also our article on cellular senescence.
True or false?
"My biological age is a precise figure." False. It is a statistical estimate, with a margin of error that can reach several years depending on the clock and the laboratory.
"If my biological age is lower, I will live longer." Partly true. It holds on average in large studies, not necessarily for you.
"You can get several years younger in a few weeks." False. Solid trials show effects of a few months over two years.
"All biological age tests are equally good." False. They do not use the same clocks and do not measure exactly the same thing.
"Multivitamins are the key to youth." False. The effect seen in COSMOS is small, and nothing yet shows it prolongs life. They replace neither a good diet nor physical activity.
Key takeaways
Biological age is a serious concept and epigenetic clocks are real scientific tools. They are transforming research on ageing, because they make it possible to test treatments much faster.
But in 2026, the test sold online remains an expensive gadget for individual use: unreliable, unstandardised, and with no advice different from what you already know. If you want to know how you are ageing, start with your blood pressure, a blood test, your strength and your stamina. And if you want to slow the clock, the best methods are free.
Going further
Towards more complete clocks. Researchers are now combining methylation with other measurements: gene activity (the 2026 Health and Retirement Study paper created "transcriptomic scores" that sometimes predict frailty or diabetes better than the clocks alone), blood proteins, or organ-specific clocks (heart, brain, kidneys).
Clocks as judges of trials. The CALERIE and COSMOS trials are examples: instead of waiting decades, researchers check whether the clock slows down. But it still has to be shown that slowing the clock really means living longer in good health. Participants in CALERIE and COSMOS are still being followed to answer this question.
Ageing is not one single thing. The Stanford team had already described "ageotypes" in 2020: in some people it is mainly the kidneys or the liver that age fast, in others metabolism or immunity. A single figure cannot sum it all up.
Test yourself
Sources
Ahadi, S., Zhou, W., et al. (2020). Personal aging markers and ageotypes revealed by deep longitudinal profiling. Nature Medicine, 26(1), 83-90. https://doi.org/10.1038/s41591-019-0719-5
Arpawong, T. E., Cole, S., Badhesha, H., Kim, J. K., Beam, C., Siegmund, K., Klopack, E. T., Thyagarajan, B., & Crimmins, E. M. (2026). How epigenetic clocks tick: unpacking the black box by deciphering biological pathways and transcriptomic signatures of accelerated aging. npj Aging. https://doi.org/10.1038/s41514-026-00446-x
Belsky, D. W., Caspi, A., et al. (2022). DunedinPACE, a DNA methylation biomarker of the pace of aging. eLife, 11, e73420. https://doi.org/10.7554/eLife.73420
Fitzgerald, K. N., Hodges, R., Hanes, D., Stack, E., Cheishvili, D., Szyf, M., Henkel, J., Twedt, M. W., Giannopoulou, D., Herdell, J., Logan, S., & Bradley, R. (2021). Potential reversal of epigenetic age using a diet and lifestyle intervention: a pilot randomized clinical trial. Aging, 13(7), 9419-9432. https://doi.org/10.18632/aging.202913
Higgins-Chen, A. T., Thrush, K. L., Wang, Y., Minteer, C. J., Kuo, P.-L., Wang, M., Niimi, P., Sturm, G., Lin, J., Moore, A. Z., Bandinelli, S., Vinkers, C. H., Vermetten, E., Rutten, B. P. F., Geuze, E., Okhuijsen-Pfeifer, C., van der Horst, M. Z., Schreiter, S., Gutwinski, S., … Levine, M. E. (2022). A computational solution for bolstering reliability of epigenetic clocks: implications for clinical trials and longitudinal tracking. Nature Aging, 2(7), 644-661. https://doi.org/10.1038/s43587-022-00248-2
Horvath, S. (2013). DNA methylation age of human tissues and cell types. Genome Biology, 14(10), R115. https://doi.org/10.1186/gb-2013-14-10-r115
Kaeberlein, M. [Optispan]. (2025). Epigenetic age tests are wrong... [Video]. YouTube. https://www.youtube.com/watch?v=blALlDEAxkI
Kuo, P.-L., Moore, A. Z., Tanaka, T., Belsky, D. W., Lu, A. T., Horvath, S., Bandinelli, S., & Ferrucci, L. (2026). Longitudinal changes in epigenetic clocks predict survival in the InCHIANTI cohort. Nature Aging, 6(3), 534-540. https://doi.org/10.1038/s43587-026-01066-6
Leong, D. P., Teo, K. K., et al. (2015). Prognostic value of grip strength: findings from the Prospective Urban Rural Epidemiology (PURE) study. The Lancet, 386(9990), 266-273. https://doi.org/10.1016/S0140-6736(14)62000-6
Levine, M. E., Lu, A. T., et al. (2018). An epigenetic biomarker of aging for lifespan and healthspan. Aging, 10(4), 573-591. https://doi.org/10.18632/aging.101414
Li, S., Hamaya, R., Pereira, A. C., et al. (2026). Effects of daily multivitamin-multimineral and cocoa extract supplementation on epigenetic aging clocks in the COSMOS randomized clinical trial. Nature Medicine. https://doi.org/10.1038/s41591-026-04239-3
Lifespan.io. (n.d.). When doctors prescribe horoscopes: the trouble with biological age tests. https://lifespan.io/news/when-doctors-prescribe-horoscopes-the-trouble-with-biological-age-tests/
Patrick, R. [FoundMyFitness]. (2022). Are consumer epigenetic age tests reliable? [Video]. YouTube. https://www.youtube.com/watch?v=RGUARXDtnZw
Rodrigues, M., Bather, J. R., & Cuevas, A. G. (2026). Aging anxiety and epigenetic aging in a national sample of adult women in the United States. Psychoneuroendocrinology, 184, 107704. https://doi.org/10.1016/j.psyneuen.2025.107704
Shen, X., Wang, C., Zhou, X., Zhou, W., Hornburg, D., Wu, S., & Snyder, M. P. (2024). Nonlinear dynamics of multi-omics profiles during human aging. Nature Aging, 4. https://doi.org/10.1038/s43587-024-00692-2
Studenski, S., Perera, S., et al. (2011). Gait speed and survival in older adults. JAMA, 305(1), 50-58. https://doi.org/10.1001/jama.2010.1923
The Scientist. (2026). Epigenetic clocks power biological age data, but why do results differ across tests? https://www.the-scientist.com/epigenetic-clocks-power-biological-age-data-but-why-do-results-differ-across-tests-74347
Waziry, R., Ryan, C. P., et al. (2023). Effect of long-term caloric restriction on DNA methylation measures of biological aging in healthy adults from the CALERIE trial. Nature Aging, 3, 248-257. https://doi.org/10.1038/s43587-022-00357-y
Yale School of Medicine. (2022). A computational solution for bolstering reliability of epigenetic clocks. https://medicine.yale.edu/news-article/a-computational-solution-for-bolstering-reliability-of-epigenetic-clocks-implications-for-clinical-trials-and-longitudinal-tracking/