What is biological age and how is it measured?

Two numbers, two different stories

Chronological age tells you how long you have been alive. Biological age tells you how well your body is holding up relative to that time. Two 50-year-olds can have biological ages a decade apart — one body showing the wear patterns of a 40-year-old, the other closer to 60. The gap between these numbers matters, because biological age predicts health outcomes far better than the date on a birth certificate.

For most of the history of medicine, biological age was an intuitive concept with no reliable way to measure it. That changed over the past decade. Advances in molecular biology have produced a new class of tools — epigenetic clocks — that can estimate biological age from a blood sample with striking accuracy, and predict mortality and disease risk in ways that go well beyond what age alone can tell us.

How epigenetic clocks work

Every cell in your body contains the same DNA sequence. What changes with age — and differs between people of the same age — is the pattern of chemical modifications layered on top of that sequence. The most well-studied of these modifications is DNA methylation: the attachment of methyl groups to specific locations on the DNA called CpG sites. Methylation patterns at thousands of CpG sites change in consistent, predictable ways as people age, and these patterns can be read from a standard blood sample.

Epigenetic clocks are algorithms trained on methylation data from large populations to estimate biological age. They work like a molecular fingerprint of ageing: your DNA sequence is the hardware, and the methylation pattern is the software — accumulating errors over time in ways that the clock can detect. First-generation clocks like the Horvath clock (2013, 353 CpG sites, trained on 8,000 samples across 51 tissue types) and the Hannum clock were trained to predict chronological age as accurately as possible, achieving correlations of around 0.96. Their limitation: predicting your age is not the same as predicting your health.

Second-generation clocks — and why they matter more

The more clinically relevant clocks are trained on health outcomes rather than chronological age. PhenoAge, developed by Morgan Levine in 2018, uses 513 CpG sites and was trained incorporating nine clinical biomarkers from the NHANES III study. Each year of PhenoAge acceleration is associated with a 4.5% increase in mortality risk. It significantly outperforms first-generation clocks for predicting 10- and 20-year mortality, cancer, Alzheimer's disease, and physical functioning.

GrimAge (2019) and its updated successor GrimAge2 (2024) go further still, using DNA methylation surrogates for plasma proteins — including markers of inflammation, kidney function, and metabolism — as their training targets. In a study of 18,859 individuals, GrimAge2 acceleration was associated with an all-cause mortality hazard ratio of 1.54 per standard deviation (p = 7.1 × 10⁻⁶²). Current smokers show 7.91 years of GrimAge acceleration on average — a useful benchmark for understanding what the clock is sensitive to.

DunedinPACE (2022) operates differently. Rather than estimating static biological age (an odometer), it estimates the current pace of ageing (a speedometer). It was developed from a 20-year follow-up of the Dunedin birth cohort, tracking 19 organ-system integrity indicators over time. Its test-retest reliability is ICC = 0.96 — important for an intervention tracking tool. Faster DunedinPACE is associated with a hazard ratio of 1.49 for developing mild cognitive impairment or dementia (ADNI cohort), and with steeper cognitive decline trajectories in the Framingham Heart Study (n=2,296).

Beyond epigenetics: newer measurement approaches

Epigenetic clocks are not the only game in town. Proteomic clocks — trained on patterns of plasma proteins rather than methylation — have emerged as powerful alternatives. A 2024 Nature Medicine study developed a clock using 204 plasma proteins from the UK Biobank (r = 0.94 with chronological age) and validated it for predicting 18 chronic diseases and all-cause mortality across UK Biobank, China Kadoorie, and FinnGen. A simplified 20-protein panel captured 91% of the accuracy of the full model. Organ-specific proteomic clocks, developed in 2025, can now estimate the biological age of individual organs — the brain, heart, liver — and predict 20-year disease risk with organ specificity that blood methylation cannot provide.

Metabolomic clocks offer a third angle. The MileAge clock (2024) was developed using 54 NMR biomarkers from 250,341 UK Biobank participants and showed an all-cause mortality hazard ratio of 1.51 — comparable to leading epigenetic clocks. Because metabolomics captures real-time biochemical state, these clocks are more responsive to short-term changes (diet, exercise, medication) than DNA methylation. GlycanAge, which measures IgG glycosylation as a proxy for immune system ageing, is a fourth approach — available commercially via finger-prick test.

Can you actually reduce your biological age?

Several interventions have now been tested against epigenetic clocks in clinical trials. The most compelling data comes from the CALERIE trial — a randomised controlled study of 220 participants maintained on 25% caloric restriction for 2 years. The DunedinPACE result: caloric restriction slowed the pace of biological ageing by 2–3%, an effect size comparable to quitting smoking. Notably, PhenoAge and GrimAge did not change significantly — a reminder that different clocks measure different aspects of ageing, and interventions affect them differently.

Exercise shows consistent observational associations with lower epigenetic age acceleration — cardiorespiratory fitness inversely correlates with epigenetic ageing across multiple clocks. Structured exercise training has been shown to induce epigenomic rejuvenation in blood and skeletal muscle, though controlled intervention data is less robust than for caloric restriction. Vitamin D supplementation has been associated with 2.6 years of lower DNA methylation age acceleration after 12 months in some cohorts. A 2026 Nature Medicine study also found that daily multivitamin supplementation slowed epigenetic clock progression — a more nuanced finding than prior research suggested.

Getting tested in Australia

Several Australian providers offer epigenetic age testing. HealthScreen provides exclusive Australian access to GrimAge testing via UCLA Horvath Lab technology. Vively offers DNAm PhenoAge testing through over 4,000 collection centres nationally. Melbourne Functional Medicine provides TruDiagnostic TruAge testing, and Elixa Longevity Centre offers comprehensive epigenetic assessments. Costs typically range from AUD $550–795. International options include TruDiagnostic, Elysium Index, and myDNAge, each using different underlying clock methodologies.

Before testing, two caveats deserve emphasis. First, commercial test accuracy is imperfect: TruDiagnostic's own estimates suggest approximately 58% accuracy in detecting meaningful changes — meaning a single result should not be over-interpreted. Second, no commercial epigenetic test has prospective clinical validation demonstrating that its specific output predicts disease or drives better clinical decisions in individuals. The clocks underlying the tests have strong population-level predictive validity; translating that to individual clinical utility is a separate and still-incomplete project.

The bottom line

Biological age testing represents one of the most scientifically grounded frontiers in longevity medicine. Second-generation epigenetic clocks, particularly GrimAge2 and DunedinPACE, have substantial evidence linking acceleration to mortality and disease across large cohort studies. Proteomic and metabolomic clocks offer complementary perspectives. The strongest intervention evidence points to caloric restriction and exercise as meaningful brakes on biological ageing — both of which are achievable without a test result. Testing is useful for establishing a baseline and tracking trajectory over time, but the fundamentals that move the needle do not require measuring it first.

References

1. Horvath, S. (2013). DNA methylation age of human tissues and cell types. Genome Biology, 14(10).
2. Levine, M. E., et al. (2018). An epigenetic biomarker of aging for lifespan and healthspan. Aging (Albany NY), 10(4).
3. Lu, A. T., et al. (2019). DNA methylation GrimAge strongly predicts lifespan and healthspan. Aging (Albany NY), 11(2).
4. Lu, A. T., et al. (2024). DNA methylation GrimAge version 2. Aging (Albany NY), 14(12).
5. Belsky, D. W., et al. (2022). DunedinPACE: a DNA methylation biomarker of the pace of aging. eLife, 11.
6. Oh, H. S-H., et al. (2024). Organ aging signatures in the plasma proteome track health and disease. Nature Medicine.
7. Nature NPJ Aging (2025). MileAge: metabolomic-based aging clock using 54 NMR biomarkers (n=250,341).
8. Waziry, R., et al. (2023). Effect of long-term caloric restriction on DNA methylation measures of biological aging in healthy adults (CALERIE trial). Nature Aging.
9. Nature Communications (2025). Unbiased comparison of 14 epigenetic clocks across 18,859 individuals.
10. JAHA (2025). MESA study: GrimAge and cardiovascular disease outcomes (n=1,264).
11. Nature Medicine (2026). Daily multivitamin supplementation slows epigenetic clock progression.