Your chronological age is fixed - a count of orbits around the sun. Your biological age is a measurable property of your cells that can reflect how you've lived in those years, what you've survived, and how your body is currently functioning. The two numbers are related. They are not the same number.
Epigenetic clocks - algorithms trained on DNA methylation patterns across the genome - produce a biological age estimate that may predict mortality, disease onset, and healthspan with a precision no blood panel alone may achieve. The best-validated clocks can now respond to interventions: they can move with lifestyle change, and they can move with treatments like hormone therapy. That can make them useful not just as a snapshot, but as a feedback mechanism.
What DNA methylation measures
DNA methylation is a chemical modification - the addition of a methyl group to cytosine bases - that changes gene expression without altering the underlying DNA sequence. Methylation patterns can shift predictably with age across hundreds of genomic sites (Horvath, 2013). Steve Horvath's 2013 discovery that these patterns could be used to construct an accurate biological age estimate launched the field of epigenetic clocks.
Methylation is not random. It can be influenced by diet, exercise, sleep, chronic stress, smoking, alcohol, and hormonal status. That is what can make it modifiable - and what may make it a more informative reflection of biological age than any single biomarker.
The three clocks worth knowing
Horvath (2013) - the original multi-tissue clock
Trained on 8,000 samples across 51 tissue types, Horvath's clock was the proof of concept (Horvath, 2013). It remains the reference standard. Its limitation is that it was optimised for accuracy across tissues, not for sensitivity to interventions - it can move more slowly in response to lifestyle changes than later clocks.
PhenoAge (Levine, 2018) - the clinical state clock
PhenoAge was trained on clinical biomarkers correlated with mortality - albumin, creatinine, glucose, CRP, lymphocyte percentage, red cell volume, and others - then mapped to a methylation model (Levine et al., 2018). It can reflect current physiological state more than Horvath's clock does, which may make it more responsive to acute interventions and more sensitive to inflammation.
GrimAge (Lu, 2019) - the mortality prediction clock
GrimAge may be the strongest predictor of time-to-death and time-to-first-disease of any clock tested to date (Lu et al., 2019). It incorporates smoking-associated methylation patterns and seven plasma protein surrogates (Lu et al., 2019). A GrimAge five years older than chronological age can predict mortality risk equivalent to someone five years older - a clinically meaningful difference.
Which clock to use
For a general longevity baseline, GrimAge may be the most predictive (McCrory et al., 2021). For tracking intervention response, PhenoAge may be the most responsive. Most commercial testing now reports both alongside Horvath as the reference.
What can move the clocks
The Fitzgerald et al. 2021 pilot trial, in 43 men aged 50-72, found that an 8-week diet and lifestyle intervention (methylation-supportive diet, sleep, exercise, stress management) may have lowered Horvath clock age by 3.23 years compared with controls (Fitzgerald et al., 2021). A 2023 Nature Aging analysis of the CALERIE trial found that two years of caloric restriction may have slowed the pace of aging on the DunedinPACE measure in healthy adults, with no significant change on clocks including PhenoAge and GrimAge (Waziry et al., 2023). The effects were small.
- Smoking can add 5+ years to biological age on the GrimAge clock - it may be the largest single modifiable factor.
- Sustained aerobic exercise may be associated with 4–9 year reduction across multiple clock systems in longitudinal studies.
- Obesity - particularly visceral adiposity - can accelerate biological aging independently of BMI.
- Chronic psychological stress can accelerate methylation age; trauma history may be detectable epigenetically (Harvanek et al., 2026).
The menopause effect - and what HRT may do
A 2016 PNAS paper by Morgan Levine found evidence that menopause may accelerate epigenetic aging of blood (Levine et al., 2016). Early evidence from smaller studies may suggest that estrogen therapy may attenuate this acceleration, but the data is preliminary and the optimal timing and formulation are not yet established.
What is established: biological age is not fate. It is a measurement that can respond to how you live. Testing it once gives you a number. Testing it annually gives you a trajectory. Testing it before and after a significant intervention gives you feedback. That feedback loop is what can make epigenetic testing clinically useful rather than merely interesting.
References
- Horvath, S. (2013). DNA methylation age of human tissues and cell types. Genome Biology, 14(10), R115. PMID 24138928
- Levine, M. E., et al. (2018). An epigenetic biomarker of aging for lifespan and healthspan. Aging, 10(4), 573-591. PMID 29676998
- Lu, A. T., et al. (2019). DNA methylation GrimAge strongly predicts lifespan and healthspan. Aging, 11(2), 303-327. PMID 30669119
- McCrory, C., et al. (2021). GrimAge Outperforms Other Epigenetic Clocks in the Prediction of Age-Related Clinical Phenotypes and All-Cause Mortality. Journals of Gerontology: Series A, 76(5), 741-749. PMID 33211845
- Fitzgerald, K. N., et al. (2021). Potential reversal of epigenetic age using a diet and lifestyle intervention: a pilot randomized clinical trial. Aging, 13(7), 9419-9432. PMID 33844651
- Waziry, R., et al. (2023). Effect of long-term caloric restriction on DNA methylation measures of biological aging in healthy adults from the CALERIE trial. Nature Aging, 3(3), 248-257. PMID 37118425
- Harvanek, Z. M., et al. (2026). A Conceptual Framework for Understanding the Effects of Stress on Epigenetic Aging. Annual Review of Psychology, doi:10.1146/annurev-psych-020226-034950. PMID 42809826
- Levine, M. E., et al. (2016). Menopause accelerates biological aging. Proceedings of the National Academy of Sciences of the United States of America, 113(33), 9327-9332. PMID 27457926
