Can biological age be slowed? What the trials on epigenetic clocks actually found

Tests that report a "biological age" often come with a second promise: that the number can be brought down. A handful of trials have now measured DNA methylation clocks before and after an intervention. This article goes through what those trials did, what they found, and why a lower clock reading is not yet the same thing as a longer or healthier life.

If you want the background first, see how biological age is calculated and what epigenetic clock tests actually measure.

What "slowing" can mean for a clock

Two kinds of DNA methylation measures appear in these trials, and they answer different questions.

• Clocks such as PhenoAge and GrimAge give an age in years. PhenoAge's methylation model was trained to predict a clinical score fitted to mortality, and that score includes chronological age; GrimAge was fitted to time-to-death and also includes chronological age. Predictions of age-related disease come from later tests of the finished clocks.

• Pace measures such as DunedinPACE give a rate: how many "years" of biological change happen per calendar year. It was built from repeated measurements of organ-system decline in a single birth cohort followed into midlife.

A trial can move one kind of measure and not the other. That already happened in the calorie-restriction trial described next.

Calorie restriction: the CALERIE trial

CALERIE randomised 220 healthy adults without obesity, aged roughly 21 to 50, to a prescribed 25% calorie restriction or their usual diet for two years. In practice the restriction group cut intake by about 12% on average, not 25%.

• In a later analysis of stored blood samples from 197 participants with a baseline sample and at least one follow-up, the restriction group showed a slower pace of ageing on DunedinPACE than the control group.

• The same analysis found no significant change in PhenoAge, GrimAge or the other clocks tested.

• The authors describe the effect size as small.

What this does not say: the trial was not designed or powered to show fewer diseases or deaths, and the methylation analysis was done after the fact (post hoc). It shows that a measure can move under a controlled intervention in healthy, non-obese adults — not that the movement translates into outcomes.

Omega-3, vitamin D and exercise: the DO-HEALTH trial

DO-HEALTH tested vitamin D (2,000 IU a day), omega-3 (1 g a day) and a simple home exercise programme, alone and in combination, in generally healthy, active adults aged 70 and older in five European countries, for three years. A post hoc analysis measured four methylation measures in 777 participants.

• Omega-3 alone slowed PhenoAge and DunedinPACE compared with no omega-3, and slowed GrimAge2 when each treatment arm was compared with placebo.

• For PhenoAge, the three treatments together showed an additive effect.

• The size of the effect was about three to four months of clock time over three years — small, by the authors' own description.

What this does not say: these are changes in blood markers in a subset of a trial whose main endpoints were different. It is not evidence that omega-3 extends life. Subgroup analyses suggested a stronger omega-3 benefit, and somewhat larger additive effects on PhenoAge, in participants who started with lower blood levels of DHA and EPA. For how the blood level itself is read, see the omega-3 index article.

The small studies behind the headlines

Much of the popular claim that biological age can be "reversed" comes from two small studies.

• TRIIM (2019). Nine men aged 51–65, of ten recruited, received a one-year combination of growth hormone, DHEA and metformin intended to regenerate the thymus. Their average epigenetic age was about 1.5 years below baseline at the end. There was no control group; the "compared to no treatment" figure is a comparison with the expected rate of ageing, not with randomised controls.

• An 8-week diet and lifestyle programme (2021). Forty-three healthy men aged 50–72 were randomised to a programme of diet, sleep, exercise, relaxation and supplements, or to no intervention. In the saliva samples used for the clock comparison (18 in the programme group and 20 controls), the programme group scored 3.23 years lower on the original Horvath 2013 clock than controls. The change within the programme group itself did not reach statistical significance.

Both are early signals in small groups over short periods. Neither reported disease incidence or deaths.

Why a single before-and-after reading can mislead

Clock readings carry measurement noise. In one analysis, running the same DNA sample twice produced differences of up to nine years on six widely used clocks. A modified method based on principal components brought most repeat readings within 1.5 years of each other.

For a person comparing two test results, this matters more than any trial: a drop of a year or two between two commercial tests can be noise rather than change. See also what the gap between biological and chronological age means.

What is still missing

A marker is useful as a trial target only if changing it has been shown to change the outcome it stands for. For epigenetic clocks this has not yet been shown. A 2023 framework from researchers in the field sets out the validation steps a biomarker of ageing would need before use in clinical trials and practice, and describes the lack of agreed standards as a current obstacle.

What the evidence supports today:

• Some interventions can shift some methylation measures by a small amount in controlled trials.

• Different clocks disagree, and a change on one is not a change on all.

• No trial has yet shown that lowering a clock reading leads to fewer diseases or a longer life.

What to do with your own number

A clock result is best treated as one research-grade reading among many, not as a target to chase. If you track it, repeating the same test from the same lab, under similar conditions, is the least noisy comparison — and any change is worth discussing with a doctor alongside measures that already have established meaning, such as blood pressure, lipids and fitness.

This material is for information only and does not replace a consultation with a doctor.

Sources

• Waziry R et al. Effect of long-term caloric restriction on DNA methylation measures of biological aging in healthy adults from the CALERIE trial. Nat Aging. 2023. PMID 37118425

• Kraus WE et al. 2 years of calorie restriction and cardiometabolic risk (CALERIE): exploratory outcomes of a multicentre, phase 2, randomised controlled trial. Lancet Diabetes Endocrinol. 2019. PMID 31303390

• Bischoff-Ferrari HA et al. Individual and additive effects of vitamin D, omega-3 and exercise on DNA methylation clocks of biological aging in older adults from the DO-HEALTH trial. Nat Aging. 2025. PMID 39900648

• Fahy GM et al. Reversal of epigenetic aging and immunosenescent trends in humans. Aging Cell. 2019. PMID 31496122

• Fitzgerald KN et al. Potential reversal of epigenetic age using a diet and lifestyle intervention: a pilot randomized clinical trial. Aging (Albany NY). 2021. PMID 33844651

• Higgins-Chen AT et al. A computational solution for bolstering reliability of epigenetic clocks. Nat Aging. 2022. PMID 36277076

• Levine ME et al. An epigenetic biomarker of aging for lifespan and healthspan. Aging (Albany NY). 2018. PMID 29676998

• Lu AT et al. DNA methylation GrimAge strongly predicts lifespan and healthspan. Aging (Albany NY). 2019. PMID 30669119

• Belsky DW et al. DunedinPACE, a DNA methylation biomarker of the pace of aging. eLife. 2022. PMID 35029144

• Moqri M et al. Biomarkers of aging for the identification and evaluation of longevity interventions. Cell. 2023. PMID 37657418