DunedinPACE and aging speed: how to read a 0.57x result

Your report arrives with a 0.57, and for a moment it feels like a shortcut: am I aging at almost half speed? The short answer is no. The number may be interesting, but it is not a literal conversion between biological years and calendar years.

DunedinPACE analyses DNA methylation patterns in one sample and turns them into a relative estimate of aging pace. A low result does not tell you how long you will live, measure every organ or prove that a treatment has rejuvenated you.

That does not make the test useless. It puts it in the right place: a promising research tool that may complement other information when we know what question it is trying to answer and how uncertain the answer is.

What 0.57x means, in one sentence

It means that this sample, analysed with this version of the model, showed a pattern compatible with a pace below the reference used to build it. The “x” makes the value look like a precise physical speed, but the scale is statistical.

The speedometer analogy only goes so far. A car can show 57 km/h in a direct physical unit. DunedinPACE combines signals from 173 CpG sites to approximate a complex biological process. Its value is centred near 1 in the development population, but 0.57 does not mean every tissue accumulates exactly 0.57 years of damage per year.

The result may suggest	The result cannot prove
Where the sample sits on that model's scale	Exactly how many years you have left
A signal associated with physiological aging in cohorts	That heart, brain, muscle and immunity age at one shared rate
A hypothesis to compare with the rest of your health	That an intervention has prevented disease
A possible trend when measurements are comparable	That a small difference is an exact biological change

To place it alongside other models, our guide to epigenetic clocks and biological age explains why Horvath, PhenoAge, GrimAge and DunedinPACE do not answer exactly the same question.

How DunedinPACE was built

Its starting point is unusual and valuable. The Dunedin cohort followed people born in the same period and recorded changes across several physiological systems over many years. Researchers then trained a blood DNA-methylation signature to approximate that longitudinal pace from one collection.

The original study evaluated the model in other datasets and reported high test-retest reliability plus associations with morbidity, disability and mortality. More recently, a BASE-II analysis compared 14 biomarkers in 1,083 adults aged 60 to 80. DunedinPACE had the strongest and most consistent association with mortality among the candidates studied.

That matters, but it remains a population finding. An association that helps separate risk between groups does not automatically become an individual prognosis. Even the complete BASE-II model had modest discrimination. Age, blood pressure, physical function, symptoms and medical history do not disappear because a molecular marker is elegant.

Biological age, risk and pace are not synonyms

PhenoAge and GrimAge were trained to reflect signals related to phenotype, disease or mortality. DunedinPACE tries to capture rate. This is a conceptual difference, not a ranking from best to worst.

It also explains why two reports may seem contradictory. You can receive an epigenetic age close to your chronological age and a relatively high or low pace at the same time. Each algorithm selects different variables, populations and targets. Comparing their decimals as if they were identical thermometers creates more confusion than information.

Before celebrating or worrying, ask which exact assay was used, which tissue was sampled, which algorithm version ran and which quality checks the sample passed. A serious report should also explain reproducibility and uncertainty instead of presenting a decimal as a margin-free truth.

What intervention studies have found

The available trials invite curiosity and caution. In CALERIE, 220 adults without obesity were assigned to calorie restriction or their usual diet for two years. The methylation analysis was post hoc: DunedinPACE slowed by a small amount, while PhenoAge and GrimAge did not change significantly.

In DO-HEALTH, another post hoc analysis studied 777 older adults for three years. Omega-3 produced small effects across several clocks, while vitamin D and exercise added a signal for PhenoAge. The estimates amounted to a few months on the modelled scales, not months of life regained. We cover the study in more detail in our review of omega-3 and biological aging.

Both studies used intermediate markers. They did not show that changing DunedinPACE prevented heart attacks, dementia, frailty or death. They also do not justify severe calorie restriction or supplements to “improve the clock”. An intervention is worthwhile when its full clinical balance is favourable, not when it optimises one number.

Why two results may not be comparable

Methylation depends on tissue and cell composition. Blood, saliva, buccal cells, PBMCs and dried blood spots are not equivalent. A cross-tissue study found differences in how several clocks behaved, so changing sample type can break the series even when the commercial label looks familiar.

Batch, laboratory, platform, bioinformatic processing, algorithm version and collection quality also matter. The person's state on the collection day adds another layer. A recent infection, weight change or medication can coincide with a result, but we should not claim causation without data.

This is why a small difference should not be narrated as “I rejuvenated”. First check that the instruments, tissues and conditions were genuinely comparable.

When to repeat the test

No clinical guideline sets a universal six-month or annual schedule. Repeating may be reasonable when you define the question in advance, keep the same laboratory and method, allow enough time for the expected change and know which sensible decision could follow each result.

If no decision would change, another measurement may add cost and anxiety without improving health. When you do repeat it, record the date, sample type, recent events, medication and assay version. A trend is interpretable only when the method remains stable.

How to use the number without letting it take over

Start with biomarkers that can change clinical decisions: blood pressure, lipids, glucose, kidney function, body composition and other measurements chosen for your context. Add strength, aerobic fitness, sleep, symptoms and habits. They usually lead to clearer actions.

1. Check the method: assay, tissue, version, laboratory and quality control.
2. Ask for context: which population defines the reference and how uncertain the result is.
3. Look for coherence: compare it with cardiometabolic risk, function and clinical trajectory.
4. Define the decision: do not repeat or change treatment if you do not know which question it will answer.
5. Skip the competition: the lowest number does not win when real health fails to follow.

The most sensible reading of a 0.57x is less dramatic and more useful: it looks like a favourable signal within one model. The next step is to ask whether it is reliable, comparable and coherent with the person behind the sample.

Frequently asked questions about DunedinPACE

What does a DunedinPACE result of 0.57 mean?

It means the methylation pattern in that sample produced an estimated pace below the model reference. It does not mean every organ is aging at 57% speed or that the person will live almost twice as long.

Is DunedinPACE a biological age?

Not exactly. Age clocks try to estimate an age or acceleration relative to chronological age. DunedinPACE estimates a rate of biological change on a relative scale built from longitudinal data.

Can it predict how long I will live?

No. It has been associated with morbidity, disability and mortality in population groups, but it cannot provide a date or a reliable life expectancy for one person.

Does a lower value prove rejuvenation?

No. A change may be interesting, but it does not prove less disease, better function or a longer life. It must be checked against the rest of the person's health and the measurement context.

Why can it change between two tests?

Sample type and quality, laboratory, algorithm version, technical processing and biological state at collection can all matter. Small differences should not be interpreted as exact biological changes.

Can blood and saliva results be compared?

Not directly. Tissues have different methylation patterns, and clocks do not always behave the same in blood, saliva, buccal cells or dried blood spots.

How often should I repeat it?

There is no universal interval supported by clinical guidelines. Repeating it makes sense only when the question is defined, the same method is used and the result could change a reasonable decision.

Should it change a treatment on its own?

No. Treatment decisions should rest on clinical indications, benefits, risks and validated markers. DunedinPACE can add context, but it does not replace that assessment.

Sources

1. Belsky DW, Caspi A, Corcoran DL, et al. “DunedinPACE, a DNA methylation biomarker of the pace of aging”. eLife. 2022;11:e73420. Original paper.
2. Levine ME, Lu AT, Quach A, et al. “An epigenetic biomarker of aging for lifespan and healthspan”. Aging. 2018;10:573-591. Europe PMC.
3. Lu AT, Quach A, Wilson JG, et al. “DNA methylation GrimAge strongly predicts lifespan and healthspan”. Aging. 2019;11:303-327. Europe PMC.
4. Waziry R, Ryan CP, Corcoran DL, 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;3:248-257. Paper.
5. Bischoff-Ferrari HA, Vellas B, Rizzoli R, 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. Paper.
6. Vetter VM, Junge MP, et al. “Comparing fourteen consensus biomarkers of aging: epigenetic pace of aging as the strongest predictor of mortality in BASE-II”. Biomarker Research. 2026. Paper.
7. Apsley AT, Caspi A, Moffitt TE, et al. “Cross-tissue comparison of epigenetic aging clocks in humans”. Aging Cell. 2025;24:e14451. Europe PMC.
8. Nelson et al. “Do we actually need aging clocks?”. npj Aging. 2025. Critical review.

This article is informational and does not replace individual medical assessment. An epigenetic clock should not be used on its own to diagnose, treat or change medication.