How to Lower Biological Age: What Randomized Trials Actually Show

TL;DR: Only a handful of randomized trials have tested whether a lifestyle or supplement intervention changes DNA methylation measures of biological age. Caloric restriction (CALERIE), omega-3 (DO-HEALTH) and an 8-week diet-and-lifestyle program (Fitzgerald pilot) each moved at least one clock, but the effects were small, differed from clock to clock, and none of these trials shows that people actually live longer. Treat any promise to “reverse your biological age” with caution.

Table of Contents

• The Short Answer: What Does the Research Show?
• What Is Biological Age and Why Does It Matter for Aging?
• Key Studies Explained
• Uncontrolled Pilots: Why TRIIM Is Not in the Table
• What the Evidence Actually Supports (and What It Doesn’t)
• The Limits: Small Effects, Clock Disagreement, Reliability
• Practical Takeaways
• FAQ
• Conclusion
• References

The Short Answer: What Does the Research Show?

Search for “how to lower biological age” and you will find confident lists of habits and supplements. Very few of those claims come from randomized controlled trials that actually measured biological age before and after an intervention. When you restrict yourself to that kind of evidence, the list becomes short: three randomized trials with DNA methylation data are strong enough to discuss in detail, and all three come with important caveats.

The pattern across them is consistent. An intervention can shift one or two epigenetic clocks by a modest amount, other clocks in the same people often do not move, and no trial has yet linked those shifts to fewer diseases or a longer life. The table below includes only randomized trials, ranked by strength of evidence.

Study	n	Finding	Evidence Quality
DO-HEALTH (Bischoff-Ferrari, Nature Aging 2025): omega-3, vitamin D, exercise, 3 years	777 (age 70+)	Omega-3 1 g/day slowed PhenoAge, GrimAge2 and DunedinPACE, not GrimAge. Vitamin D and exercise alone: no effect on any clock. All three combined: additive benefit on PhenoAge only. Effects about 3 months over 3 years.	Moderate: large randomized trial, but post hoc clock analysis without formal multiple-testing correction
CALERIE (Waziry, Nature Aging 2023): caloric restriction, 2 years	220	Slowed DunedinPACE. No significant change in PhenoAge or GrimAge. Small effect sizes.	Moderate: randomized trial, post hoc analysis, one of several clocks moved
Fitzgerald pilot (Aging 2021, corrected 2024): 8-week diet and lifestyle program	43 men (18 analyzed in treatment group)	Horvath clock (saliva) 3.23 years lower vs control (p=0.018). Multi-component program, so no single cause can be identified.	Low: small pilot, short duration, first-generation clock

What Is Biological Age and Why Does It Matter for Aging?

Chronological age is simply the number of years you have been alive. Biological age is an attempt to estimate how far your body has actually moved along the aging process. The most widely used tools are epigenetic clocks: algorithms that read DNA methylation, the chemical tags on DNA that change in predictable ways as we get older.

Not all clocks measure the same thing. First-generation clocks such as the Horvath clock were trained to predict chronological age. Later clocks such as PhenoAge and GrimAge were trained on blood markers and mortality risk, and GrimAge2 is an updated version of GrimAge. DunedinPACE, developed by Daniel Belsky and colleagues at Columbia and Duke, does something different: it estimates the current pace of aging, based on how quickly 19 indicators of organ-system integrity declined over two decades in the Dunedin Study cohort. In the original paper, DunedinPACE showed high test-retest reliability and was associated with morbidity, disability and mortality (Belsky et al., eLife 2022).

This matters because a trial can report that an intervention “lowered biological age” when only one clock moved. When you read a claim, always ask which clock was used, whether it was measured in blood or saliva, and whether other clocks agreed. If you want a practical comparison of the tests you can actually buy, see our guide to biological age tests compared.

Key Studies Explained

DO-HEALTH: Omega-3, Vitamin D and Exercise in Adults Over 70

DO-HEALTH was a 3-year randomized trial in generally healthy, active adults aged 70 and older. It used a factorial design: participants received vitamin D (2,000 IU per day), omega-3 (1 g per day, provided as 330 mg EPA plus 660 mg DHA from marine algae), a simple home strength exercise program (30 minutes, three times a week), all three, or placebo and control combinations. In 2025, Heike Bischoff-Ferrari and colleagues published a post hoc analysis in Nature Aging of DNA methylation clocks in 777 of the 1,006 Swiss participants (59% women, mean age 75), measured at the start and after 3 years.

The results were specific to the clock and to the intervention:

• Omega-3 alone slowed PhenoAge, GrimAge2 and DunedinPACE (standardized effects of about 0.16 to 0.32 units). It did not significantly change the original GrimAge.
• Vitamin D alone and the exercise program alone were not associated with changes in any of the clocks.
• All three combined showed an additive benefit on PhenoAge only. There was no additive effect on GrimAge, GrimAge2 or DunedinPACE.
• In absolute terms, the effects corresponded to roughly 2.9 to 3.8 months of slower biological aging over 3 years.

That is a real signal from a well-run randomized trial, but it needs careful framing. The clock analysis was post hoc, the authors did not apply a formal correction for multiple testing (they looked for consistent patterns across confidence intervals instead), there were only two time points, and the participants were a healthier, more active Swiss subgroup than the general population of older adults. The authors also note that there is no gold-standard measure of biological aging and that a 3-year trial cannot tell us whether these changes translate into longer survival. We covered the omega-3 side of this research in more depth in omega-3 and biological aging.

The null result for exercise deserves a note. It means this specific home program did not shift these clocks in already active 70-year-olds over 3 years. It does not mean exercise is irrelevant to healthy aging; the much broader evidence on physical activity is reviewed in exercise and biological age.

CALERIE: Two Years of Caloric Restriction

CALERIE is the only long-term randomized trial of caloric restriction in healthy, non-obese adults. Participants were randomized to a target of 25% caloric restriction or to their usual diet for 2 years. In a post hoc analysis published in Nature Aging in 2023, Ryan Waziry, Daniel Belsky and colleagues analyzed DNA methylation in blood samples from the 220 randomized adults.

Caloric restriction slowed the pace of aging as measured by DunedinPACE. It did not significantly change biological age estimates from the other clocks, including PhenoAge and GrimAge. The authors described the treatment effects as small and stated that conclusive evidence would require long-term follow-up on chronic disease and mortality. Two further caveats: 25% was the target, and the restriction participants actually achieved was lower; and a single clock moving while others did not is a reason for caution, not a headline.

The Fitzgerald Pilot: An 8-Week Diet and Lifestyle Program

In 2021, Kara Fitzgerald and colleagues published a small randomized pilot trial in Aging (Albany NY). Forty-three healthy men aged 50 to 72 were randomized to an 8-week program or to a control group. The program combined a diet rich in foods thought to support methylation, exercise, sleep and relaxation guidance, and supplemental probiotics and phytonutrients. Biological age was measured in saliva with the first-generation Horvath clock from 2013.

At 8 weeks, DNA methylation age in the treatment group was 3.23 years lower than in the control group (p=0.018). A 2024 correction revised the within-group figures but left that between-group result unchanged. According to the corrected data, the treatment group (18 men analyzed) was 2.04 years younger than at baseline (p=0.043), and the control group was 1.10 years older (p=0.191, not significant). Of the 18 treated men, 8 showed a reduction, 9 were unchanged and 1 increased. An earlier 2022 correction concerned data sharing only and did not change the results.

This is an interesting early signal, but it is weak evidence. The sample was small, the follow-up lasted just 8 weeks, the clock is a first-generation measure read from saliva, and because the program had many components, there is no way to know which element (if any) was responsible.

Uncontrolled Pilots: Why TRIIM Is Not in the Table

You will often see the TRIIM study (Thymus Regeneration, Immunorestoration and Insulin Mitigation) cited as proof that biological age can be reversed. Published by Gregory Fahy and colleagues in Aging Cell in 2019, it tested a 1-year protocol built around growth hormone combined with DHEA and metformin. The authors reported a mean epigenetic age about 1.5 years below baseline after 1 year, and a GrimAge decrease of about 2 years that persisted six months after treatment stopped.

TRIIM was a single-arm pilot: it was not randomized and had no control group. Without a comparison group, you cannot separate the effect of the treatment from measurement noise, regression to the mean or other changes over the year. That is why it sits outside the ranked table here. It is a hypothesis worth testing in a proper trial, not an established way to lower biological age, and it involves hormonal drugs that require medical supervision.

What the Evidence Actually Supports (and What It Doesn’t)

Reasonably supported: in randomized trials, some interventions can shift at least one DNA methylation clock in the expected direction. Omega-3 at 1 g/day in adults over 70 (DO-HEALTH) and caloric restriction in healthy adults (CALERIE) are the best-documented examples, and in both cases the change was measurable but small.

Weakly supported: that a short, multi-component lifestyle program can lower a saliva-based Horvath clock within weeks (Fitzgerald pilot). The direction is encouraging, but the trial is too small and too short to be conclusive.

Not supported: that any of these interventions makes people live longer or reduces disease because it changed a clock. That link has not been tested in these trials. Also not supported: that vitamin D or the DO-HEALTH exercise program on their own changed any clock, or that results in one population (for example, Swiss adults over 70) apply to everyone.

The Limits: Small Effects, Clock Disagreement, Reliability

The effects are small. The DO-HEALTH effects amount to roughly three months of slower biological aging over three years. CALERIE’s authors also called their effects small. These are not the “10 years younger” results that appear in marketing.

Clocks disagree. In CALERIE, DunedinPACE moved while PhenoAge and GrimAge did not. In DO-HEALTH, omega-3 changed GrimAge2 but not the original GrimAge, and the combined intervention added a benefit only on PhenoAge. Different clocks capture different aspects of aging, so a result on one clock does not generalize to “biological age” as a whole.

Measurement noise is real. Albert Higgins-Chen and colleagues showed in Nature Aging (2022) that technical noise alone can produce deviations of up to 9 years between replicate samples for six prominent epigenetic clocks. Their principal-component versions of those clocks brought most replicates within 1.5 years. When a single test reports that you are two years younger, that difference can be within the noise of some clocks.

A clock change is not a longer life. Epigenetic clocks are associated with disease and mortality in population studies, but no trial so far has shown that lowering a clock score through an intervention extends lifespan or healthspan. Until long-term outcome data exist, clock changes are best read as early, surrogate signals.

Practical Takeaways

• Judge claims by study design. Look for randomized trials with a control group. Single-arm pilots and before-and-after testimonials cannot show cause and effect.
• Ask which clock. A claim that something “lowers biological age” should name the clock, the tissue (blood or saliva) and whether other clocks agreed.
• Expect small changes. The best trial data show effects of months, not years, over several years of intervention.
• If you test yourself, test consistently. Use the same test, the same lab and similar conditions, and do not over-interpret a single result given the measurement noise described above.
• Keep the fundamentals in perspective. Diet quality, regular physical activity and sleep have broad health evidence that does not depend on clock results.

On omega-3 specifically: DO-HEALTH tested 1 g per day of omega-3 (330 mg EPA plus 660 mg DHA from marine algae) in adults aged 70 and older. That is the dose the trial studied, not a recommendation for anyone to start supplementing. Commercial products differ in source, formulation and EPA to DHA ratio, and the trial results cannot be assumed to apply to any particular product or to younger people. If you are considering omega-3, discuss it with your doctor, especially if you take blood thinners.

Disclosure: affiliate link. I earn a small commission at no additional cost to you if you purchase through this link.

FAQ

Can you actually lower your biological age?

Randomized trials show that some interventions can shift certain epigenetic clocks by a small amount. Whether that means your body is truly aging more slowly, and whether it adds years of healthy life, has not yet been demonstrated.

Which intervention has the strongest evidence?

Among randomized trials with DNA methylation data, DO-HEALTH (omega-3, 3 years, 777 adults over 70) and CALERIE (caloric restriction, 2 years, 220 adults) are the strongest. Both were post hoc analyses, and both found small, clock-specific effects.

Did vitamin D lower biological age in DO-HEALTH?

No. On its own, vitamin D at 2,000 IU per day was not associated with changes in any of the clocks. The only additive effect of combining vitamin D, omega-3 and exercise was seen on PhenoAge.

Should I trust a home epigenetic test that says I am younger?

Interpret it cautiously. Technical noise can produce large differences between replicate measurements for some clocks, and different clocks can disagree. A single result is a rough estimate, not a verdict.

Conclusion

The honest answer to “how to lower biological age” is that the best randomized evidence is still thin. Omega-3 in older adults and caloric restriction in healthy adults moved some clocks by a small amount, a short lifestyle program moved a first-generation saliva clock in a small pilot, and the most dramatic “reversal” result came from a trial with no control group. Clocks disagree, measurements are noisy, and no trial has yet shown that changing a clock adds healthy years. The fundamentals of diet, activity and sleep remain the most defensible foundation.

If you want the full system behind these decisions, I wrote it all down in The MVHK Protocol.

References

• 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. Nature Aging. 2025. PMID: 39900648.
• Waziry R, et al. Effect of long-term caloric restriction on DNA methylation measures of biological aging in healthy adults from the CALERIE trial. Nature Aging. 2023. PMID: 37118425.
• 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. Corrections: PMID: 35907207 (2022) and PMID: 38488762 (2024).
• Fahy GM, et al. Reversal of epigenetic aging and immunosenescent trends in humans. Aging Cell. 2019. PMID: 31496122.
• Belsky DW, et al. DunedinPACE, a DNA methylation biomarker of the pace of aging. eLife. 2022. PMID: 35029144.
• Higgins-Chen AT, et al. A computational solution for bolstering reliability of epigenetic clocks: implications for clinical trials and longitudinal tracking. Nature Aging. 2022. PMID: 36277076.