Can Exercise Slow Your Biological Age? What Epigenetic Clocks Show

What an epigenetic clock is

Your chronological age is how many birthdays you have had. Your biological age is a guess at how worn the machinery is.

The best current way to guess it works like this. Small chemical tags sit on your DNA and shift in fairly predictable ways as you get older. A model reads the pattern of those tags in a blood sample and returns an age. If that number lands above your real age, the gap has a name: epigenetic age acceleration.

There is more than one clock, and they were built for different jobs. Horvath and Hannum predict your chronological age. GrimAge and PhenoAge were trained on health outcomes instead, so they track how people actually fare. The clocks do not agree with each other. That turns out to matter here.

What the big review found

A 2026 systematic review and meta-analysis in The Lancet Healthy Longevity searched six databases and pulled in 44 studies covering 145,465 people, with average ages from 24 to 78. Across those studies, more physical activity generally went with a lower methylation age, although plenty of the individual results did not reach significance on their own. (Shan 2026.)

Then they pooled what could be pooled, which came to seven cross-sectional studies:

Two things worth sitting with. The clock with the strongest link is GrimAge, the one built to predict how people actually do rather than what their birth certificate says. And the effects are small. A tenth of a standard deviation is a real signal in a sample of that size, not a transformation.

The catch, stated plainly

The pooled result comes from cross-sectional studies. Everyone was measured once. So the finding is that active people have younger clocks, which is not the same as finding that becoming active makes your clock younger. The review's authors say so themselves: the evidence is predominantly cross-sectional, that limits causal inference, and trials are needed to determine whether physical activity can causally modify ageing. (Shan 2026.)

Plenty could explain the gap without exercise doing the work. Healthier people find it easier to be active in the first place. Money, sleep, smoking and stress all move with activity, and all of them touch methylation too. Until a long trial measures the same people before and after, the honest answer is that the two travel together.

What the trials do show

Exercise clearly changes methylation. A 2026 systematic review in Clinical Epigenetics went through 29 human exercise trials. One session changed very little. Sustained training was different: it remodelled methylation on genes tied to metabolism, inflammation and immune function. How much depended on the dose and on who was exercising. (Yang 2026.)

That is a different claim from the clock one, and a firmer one. Training rewrites the marks on genes that run your metabolism and your immune system. Whether the composite score we call biological age follows is the open question.

What to do with this

Nothing here changes the advice, which is the reassuring part. Every plausible reading of this research points the same way: keep moving, keep doing it for years, and make it something that loads your muscles rather than only your lungs. Muscle turns out to be the tissue with the most direct line to how well people age, and we wrote about that separately.

Two things I would skip. A consumer biological age test will hand you a number from one clock, and you have just seen that the clocks disagree with each other on this very question. And anything sold on the promise of reversing your epigenetic age is running well ahead of the evidence. The trial that would justify that language has not been done.

What is left is the boring version that works: sessions you will still be doing in a year. Ours happen to be an hour and three quarters, adults only, in groups of four, and hard enough that nobody checks their watch. If you are working out how often, we built a weekly plan from the trial doses.

FAQ

Should I get my biological age tested?

There is no need, and the review above is the reason. The clocks disagree about the effect of the thing you are asking them to measure, so a single number from a single clock is not a good basis for a decision. If you want a health check, a GP with a blood panel is better value and better evidence.

Does the type of exercise matter?

The clock review measured overall activity in MET-minutes per week, so it cannot separate lifting from walking from bouncing. The trials review found the response varied with dose and population, which suggests it might, but nobody has ranked exercise types against these clocks yet.

Is it too late if I have not been active?

The average ages in these studies ran up to 78, and the association held across that range. That is not proof of a benefit from starting late, for all the reasons above, but there is no age at which the evidence tells you to stop bothering. Ask your GP first if you have a condition or you are starting from nothing.

Does this mean exercise makes me younger?

It means active people's cells read younger on two of the four clocks tested. The study design cannot tell us which way the arrow points. I would rather leave that open than sell you something the research has not earned.

Want the version that actually keeps happening? Guided sessions at Run The Wall in Brookvale are adults-only, groups of four, and coached on every jump. Book a session, or read why muscle is the tissue worth protecting in Muscle Is the Organ of Longevity.

References

• Shan, J., et al. (2026). Physical activity and biological age measured by DNA methylation clocks: a systematic review and meta-analysis. The Lancet Healthy Longevity, 7(4), 100835. doi:10.1016/j.lanhl.2026.100835 (PMID 42068988)
• Yang, Y., Zhu, F., & Chen, A. (2026). Effects of exercise on peripheral blood DNA methylation and related epigenetic markers: a systematic review of human trials. Clinical Epigenetics, 18(1), 123. doi:10.1186/s13148-026-02123-y (PMID 42069681)