Biological Age vs Chronological Age: What's the Difference?

Two people turn 50 on the same day. One runs half marathons, sleeps seven hours and has the blood pressure of a thirty-year-old. The other is pre-diabetic, sedentary and short of breath climbing stairs.

Chronologically identical. Biologically, not remotely. That gap is the entire idea behind biological age — and it is a genuinely useful concept that has been packaged into consumer tests considerably less reliable than most buyers realise.

For where measurement fits alongside intervention, see our complete longevity supplements guide. To see current UK options, browse the full range of anti aging supplements at Welzo.

The two ages defined

The premise is that two 45-year-olds can have substantially different rates of organ function decline and disease risk, and that a measurement capturing that difference would be more informative than a birth certificate.

That premise is reasonable. Everything difficult in this field concerns the second row of that table: how accurately can it actually be estimated?

What creates the gap in the first place

Before measuring anything, it is worth being clear about what the measurement is trying to capture. The divergence between the two ages is driven by a fairly well-agreed set of factors:

• Cardiometabolic status — blood pressure, glucose regulation, lipids and body composition. These are measurable, treatable, and account for a large share of the difference between people of the same age.
• Cardiorespiratory fitness — among the strongest single predictors of healthspan, and highly modifiable.
• Muscle mass and strength, which decline predictably without resistance training.
• Smoking, which is directly encoded into GrimAge via methylation surrogates for smoking pack-years [1,5].
• Chronic inflammation, captured clinically by markers such as CRP, which feeds into PhenoAge [5].
• Sleep, stress and environmental exposures — which, as the reliability section below shows, also affect the measurement itself.

Notice how many of those appear on a standard blood panel. That observation runs through the rest of this article.

How biological age is measured

Note that these are not different measurements of one underlying number. They capture different aspects of ageing, and there is no reason to expect them to agree.

The three generations of epigenetic clocks

Understanding which generation a test uses matters more than the brand name on the box.

First generation — trained to guess your birthday

Horvath and Hannum, published around 2013, were trained to predict chronological age from methylation patterns. They do that well — which is a strange thing to pay for, since you already know your age. Their clinical utility is limited because predicting the thing you already know is not prognostic.

Second generation — trained on health outcomes

This was the important shift. PhenoAge, developed by Morgan Levine at Yale, incorporated clinical biomarkers including CRP, glucose and albumin into its training, making it more reflective of physiological health [5]. GrimAge, developed by Steve Horvath and Ake Lu, uses DNA methylation surrogates for plasma proteins associated with mortality — including PAI-1, adrenomedullin and smoking pack-years — and predicted mortality more accurately than any prior clock [1,5].

GrimAge acceleration — biological age exceeding chronological age — predicts cardiovascular disease, cancer and all-cause mortality independently of traditional risk factors [1].

Third generation — pace rather than position

DunedinPACE represents a conceptual shift. Instead of estimating a static age, it estimates how fast you are currently ageing. It was developed from the Dunedin cohort — 1,037 people born in New Zealand in 1972–73 and followed for five decades — by tracking the rate of decline across 19 biomarkers of organ function measured repeatedly over time, then identifying the methylation signature associated with faster or slower decline [1,6].

A DunedinPACE score of 1.0 means you are ageing at roughly one year per calendar year. A score of 0.85 means about 15% slower [1].

Its predecessor, DunedinPoAm, was built from a 12-year span with data at only three time points — a constraint that limited its accuracy for assessing rate of ageing and its usefulness in intervention trials. DunedinPACE was introduced to address those limitations [7].

The reliability problem

This is the section that should change how you read your result, and it is almost never covered in consumer marketing.

Technical reliability is excellent. Biological reliability is not.

A 2025 analysis evaluated epigenetic clocks across four independent datasets, using replicate samples processed under different laboratory conditions, and separated two very different questions [3]:

• Technical reliability — if you split one sample and run it twice, do you get the same number? Here most clocks performed excellently, with intraclass correlation coefficients above 0.9. The laboratory process is reproducible.
• Biological reliability — if you sample the same person twice under ordinary conditions, do you get the same number? Here pooled estimates showed most clocks achieving only "moderate" to "good" reliability (ICC roughly 0.4–0.7), with no clock reaching the "excellent" range [3].

The counterintuitive part

Technical reproducibility did not predict biological stability — the correlation between the two was essentially zero. Some of the most technically robust clocks, including GrimAgeV2 and DunedinPACE, were among the most biologically fragile [3].

That is the opposite of what most people would assume. A test can be beautifully consistent in the laboratory and still give you a different answer next month for reasons that have nothing to do with how you have aged.

An important caveat about this evidence

The reliability analysis described above was published as a preprint and had not completed peer review at the time of writing [3]. We are giving it prominence because it is directly relevant, methodologically detailed and consistent with longstanding concerns in the field — but it should be read as strong preliminary evidence rather than settled fact, and we would apply the same caution we apply to preprints anywhere else.

When clocks disagree

The clearest illustration comes from the CALERIE trial — a randomised controlled trial in which 220 non-obese participants were assigned to two years of 25% caloric restriction or an ad libitum control diet.

The analysis, published in Nature Aging in 2023, found that caloric restriction slowed DunedinPACE significantly compared with controls — while PhenoAge and GrimAge showed no statistically significant change [4].

Sit with what that means for a consumer. Two years of a demanding, well-controlled intervention. One clock says it worked. Two clocks say it did not. Your verdict on your own two years of effort depends on which test you bought.

The reasonable interpretation is that different clocks capture different aspects of the ageing process [4] — which is scientifically interesting and commercially awkward, because it means no single result is the answer.

What actually predicts outcomes

Here is the fact that reframes the whole category.

PhenoAge — one of the better-performing clocks — was built on clinical blood markers. Using data from the NHANES III study, covering 9,926 adults with more than 23 years of mortality follow-up, researchers identified nine blood markers that collectively predict 10-year survival with around 90% accuracy [5].

Nine ordinary blood markers. Not methylation arrays. The epigenetic clock was subsequently trained to approximate that clinical composite from DNA.

Functional measures you can do at home

These cost nothing, are strongly linked to outcomes in older adults, and are considerably more stable than a methylation reading taken after a stressful week.

• Grip strength. One of the most consistent single predictors of function and mortality risk in ageing research. A cheap dynamometer measures it.
• Five-times sit-to-stand. Time yourself rising from a chair five times without using your arms. This is the endpoint used in NMN trials [8].
• Gait speed. Time a comfortable-pace walk over a fixed distance. Walking speed maintenance was an endpoint in NAD⁺ precursor research [9].
• Cardiorespiratory fitness. Among the strongest predictors of healthspan — see zone 2 training for longevity.
• Balance. Single-leg stand time, eyes open.

Record these once, then again in six months. The comparison is more informative than a single epigenetic reading, and the measurement is free.

Why researchers use these rather than clocks

It is worth noticing that the trials of longevity compounds do not typically use epigenetic age as their endpoint. They use sit-to-stand time [8], walking speed [9], muscle strength and endurance. That is not an oversight — functional measures are stable, meaningful to the person being measured, and directly relevant to independence in later life.

If the researchers designing the studies choose grip strength and gait speed over a methylation panel, that is a reasonable signal about which measurement carries more practical weight.

Supporting all of this: adequate protein, vitamin D for muscle function, and creatine, which has among the largest evidence bases of any supplement for preserving lean mass [12].

Should you buy a biological age test?

If you do buy one

• Check which clock it uses. A first-generation clock predicting chronological age tells you little.
• Sample under consistent conditions — same time of day, similar fasting state, not during illness or an unusually stressful period [3].
• Do not over-interpret a single result. A three-year difference may be within the noise.
• Expect disagreement if you run more than one test.

What to measure instead

These are the markers that underpin the clinical composites, and each one is actionable.

• HbA1c — average blood glucose over two to three months; central to metabolic ageing.
• Cholesterol and lipids — cardiovascular risk, and treatable.
• Full blood count — includes immune and red cell markers used in clinical age composites [5].
• Liver function including albumin, a component of PhenoAge [5].
• Kidney function — declines predictably with age and is worth tracking.
• Vitamin D, ferritin and thyroid function — common, correctable, and frequently the actual explanation for feeling older than you are.

For a single comprehensive baseline, the Full Body MOT Health Check, health and lifestyle blood test or Welzo Well-Human advanced blood test cover most of the above in one draw. If cardiovascular risk is the concern specifically, see the heart disease risk blood test.

Can you lower it?

Possibly — with important caveats. The CALERIE trial found two years of 25% caloric restriction slowed DunedinPACE significantly, though not the other clocks tested [4]. That is the strongest randomised evidence available, and it required a demanding two-year intervention to move one measure out of three.

Given the reliability findings [3], apparent improvements after short interventions should be treated sceptically — particularly if the follow-up sample was taken under different conditions from the first.

The interventions with the best evidence for healthspan remain unglamorous and are covered in how to lower your biological age: sleep, resistance training, cardiorespiratory fitness, not smoking, dietary quality and treating raised blood pressure or glucose. See also the hallmarks of ageing and Blue Zones longevity.

On the supplement side, our assessments are in the minimal supplement stack, the NAD supplement guide, urolithin A and — for the sceptical counterweight — longevity supplements that probably don't work.

When to speak to a doctor

A biological age result is not a diagnosis and should not be treated as one. If a consumer test returns a concerning number, that is a reason to review conventional risk factors with your GP — blood pressure, cholesterol, glucose, weight, smoking and family history — not a reason to panic or to self-treat.

Speak to your GP promptly rather than relying on any ageing test if you have persistent unexplained fatigue, unintended weight loss, breathlessness, chest pain, or a new symptom lasting more than a few weeks. These require clinical assessment, and no ageing clock substitutes for it.

Equally, a reassuring biological age result does not cancel out a raised blood pressure reading or a family history of heart disease. Treat it as one input among many.

Frequently asked questions

What is the difference between biological age and chronological age?

Chronological age is time since birth — fixed and certain. Biological age is an estimate of how much physiological deterioration has occurred relative to the average person of your years. Two 45-year-olds can differ substantially in organ function decline and disease risk, and biological age attempts to capture that difference.

Are biological age tests accurate?

Technically reproducible, biologically less so. A 2025 analysis found most epigenetic clocks achieved excellent technical reliability but only moderate to good biological reliability, with no clock reaching the excellent range — and short-term factors including meals, stress and pollution exposure caused substantial fluctuations in results [3]. That analysis was a preprint at the time of writing.

Which biological age test is best?

Among epigenetic clocks, second and third-generation options — GrimAge and DunedinPACE — have the strongest outcome validation, with GrimAge acceleration predicting cardiovascular disease, cancer and all-cause mortality independently of traditional risk factors [1]. First-generation clocks such as Horvath and Hannum were trained to predict chronological age, which you already know. For actionability and cost, a clinical blood panel is often the better purchase.

Why do different biological age tests give different results?

Because they measure different things. In the CALERIE trial of two years of 25% caloric restriction, DunedinPACE slowed significantly while PhenoAge and GrimAge showed no significant change [4]. Different clocks capture different aspects of ageing, so disagreement is expected rather than a sign that one is broken.

Can I actually lower my biological age?

The strongest randomised evidence is the CALERIE trial, where two years of 25% caloric restriction slowed DunedinPACE — though not PhenoAge or GrimAge [4]. Given that everyday factors move these readings [3], be cautious about apparent improvements after short interventions. Sleep, training, cardiorespiratory fitness and managing blood pressure and glucose remain the best-evidenced levers.

Is a blood test better than an epigenetic test?

For most people, yes — on cost, reliability and actionability. PhenoAge, one of the better clocks, was built from clinical blood markers: nine markers identified from NHANES III data on 9,926 adults with over 23 years of follow-up predict 10-year survival with around 90% accuracy [5]. Measuring those markers directly gives you much of the same signal in a form your GP can act on.

What free measures tell me about biological age?

Grip strength, five-times sit-to-stand, gait speed, balance and cardiorespiratory fitness are all linked to outcomes in ageing research, cost little or nothing, and are more stable than a single methylation reading. Record them now and repeat in six months — the trend matters more than any single value.

Does a high biological age mean I will die sooner?

No. GrimAge acceleration is associated with higher risk at a population level [1], but an association across thousands of people does not predict any individual's outcome. It is one input alongside blood pressure, glucose, smoking status, fitness and family history — and unlike a birth certificate, most of those are modifiable.

How often should I test?

Not often, and not reactively. Given the biological instability documented in repeat measures [3], frequent testing mostly measures noise. Annual testing under consistent conditions — same time of day, similar fasting state, not during illness or a stressful period — is more informative than quarterly testing under varying ones.

Are telomere tests worth it?

Telomere length was an early candidate for measuring biological ageing but has high measurement variability between labs and methods, and it has largely been superseded by epigenetic clocks and clinical composites in research. It is not the strongest use of a testing budget.

References

1. Biological Age Test: 5 Methods Compared by Accuracy & Actionability — summarising GrimAge composition and mortality prediction, and DunedinPACE development from the Dunedin cohort. https://mitohealth.com/guide/biological-age-testing
2. Belsky DW, et al. DunedinPACE, a DNA methylation biomarker of the pace of aging. eLife. 2022. https://elifesciences.org/articles/73420
3. Biological versus Technical Reliability of Epigenetic Clocks and Implications for Disease Prognosis and Intervention Response. bioRxiv, 2025. Preprint — not certified by peer review at the time of writing. https://www.biorxiv.org/content/10.1101/2025.10.13.682176v1.full
4. The Complete Guide to Epigenetic Clocks — reporting the CALERIE randomised trial (220 participants, two years of 25% caloric restriction), published in Nature Aging 2023, in which DunedinPACE slowed significantly while PhenoAge and GrimAge did not. Read the summary
5. Biological Age Test: Epigenetics vs Blood Biomarkers — describing PhenoAge development from NHANES III (9,926 adults, 23+ years mortality follow-up; nine blood markers predicting 10-year survival with ~90% accuracy) and the generations of epigenetic clocks. https://lolahealth.com/blogs/longevity/biological-age-testing-explained
6. EpigeneticAgePipeline: an R package for comprehensive assessment of epigenetic age metrics from methylation microarrays — noting DunedinPACE's 173 CpG sites and associations with morbidity, disability and mortality. https://www.biorxiv.org/content/10.1101/2024.10.24.620090.full.pdf
7. Epigenetic Clocks: Beyond Biological Age, Using the Past to Predict the Present and Future — on DunedinPoAm's limitations and the development of DunedinPACE. https://pmc.ncbi.nlm.nih.gov/articles/PMC12539533/
8. Kim M, Seol J, Sato T, et al. Effect of 12-Week Intake of Nicotinamide Mononucleotide on Sleep Quality, Fatigue, and Physical Performance in Older Japanese Adults. Nutrients. 2022;14(4):755. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8877443/
9. Morifuji M, Higashi S, Ebihara S, Nagata M. Ingestion of β-nicotinamide mononucleotide increased blood NAD levels, maintained walking speed, and improved sleep quality in older adults. GeroScience. 2024;46(5):4671–4688. https://pmc.ncbi.nlm.nih.gov/articles/PMC11336149/
10. NAD⁺ precursor supplementation in human ageing: clinical evidence and challenges. Nature Metabolism. 2025. https://www.nature.com/articles/s42255-025-01387-7
11. NHS. Vitamin D — Vitamins and minerals. https://www.nhs.uk/conditions/vitamins-and-minerals/vitamin-d/
12. Kreider RB, Kalman DS, Antonio J, et al. International Society of Sports Nutrition position stand: safety and efficacy of creatine supplementation in exercise, sport, and medicine. Journal of the International Society of Sports Nutrition. 2017;14:18. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5469049/
13. US Preventive Services Task Force; Mangione CM, Barry MJ, et al. Vitamin, Mineral, and Multivitamin Supplementation to Prevent Cardiovascular Disease and Cancer: US Preventive Services Task Force Recommendation Statement. JAMA. 2022;327(23):2326–2333. https://pubmed.ncbi.nlm.nih.gov/35727271/

All product photography in this article is owned by Welzo and served from the Welzo media library. Research positions correct at the time of publication.