DNA Methylation and Biological Age, Explained
Two people born in the same year can age at very different speeds. DNA methylation is one of the most promising ways scientists have found to measure that difference.
KEY POINTS
- Chronological age counts years. Biological age estimates how far your body has aged.
- Epigenetic clocks estimate biological age from DNA methylation at selected CpG sites.
- In large studies, a biological age above chronological age is linked with higher risk of disease and death.
- For an individual, the estimate carries uncertainty and is not a diagnosis. Repeated measurements with the same method are most informative.
Chronological age and biological age
Your chronological age is simply the time since you were born. Your biological age is an estimate of how far your cells, tissues and organs have aged. The two often differ, and the gap can be meaningful: across large populations, people whose biological age runs ahead of their calendar age tend to have a higher risk of age-related disease.
There is no single “true” biological age. Researchers estimate it from different signals — blood biomarkers, physical function, and, increasingly, DNA methylation.
How methylation changes as we age
DNA methylation is a chemical mark placed on the DNA, mostly at CpG sites, where the letter C is followed by G. At each site, a test measures the share of DNA copies in the sample that carry the mark, from 0% to 100%.
As we age, methylation at some sites rises steadily, at others it falls, and overall patterns become less tightly controlled — a process sometimes called epigenetic drift. Because some of these changes are so consistent, they can be used to estimate age.
How epigenetic clocks work
An epigenetic clock is a statistical model. Researchers measure methylation at hundreds of thousands of sites in thousands of people, then use machine learning to find the combination of sites that best predicts an age-related outcome. Applying that formula to a new sample gives an estimate.
First generation — trained on calendar age. The Horvath clock (2013) uses 353 CpG sites and works across many tissues. The Hannum clock (2013) uses 71 sites measured in blood.
Second generation — trained on health. PhenoAge (2018) was trained on a “phenotypic age” built from clinical blood markers, and GrimAge (2019) on time to death. Both predict disease and mortality better than first-generation clocks.
Third generation — pace of aging. DunedinPACE (2022) was trained on how fast 19 biomarkers changed over two decades in the same group of people. It estimates how fast someone is aging right now, rather than how old they are.
What the number can and cannot tell you
Epigenetic age estimates are strongly associated with health outcomes in population studies. For an individual, they are estimates with a margin of error, and technical factors in the laboratory can shift results; newer methods aim to make clocks more reliable on repeat testing.
That is why a single result should be read as a signal, not a verdict, and why repeated measurements with the same method are most useful. Early trials suggest that some interventions can change these measures — for example, two years of moderate calorie restriction slightly slowed DunedinPACE in the CALERIE trial — but whether such changes translate into longer, healthier lives is still being studied.
An estimated biological age is not a clinical diagnosis. Decisions about your health should be made with a qualified professional.
Evidence at a glance
ESTABLISHED SCIENCE
- DNA methylation at specific sites changes predictably with age.
- Epigenetic age estimates are associated with disease and mortality risk across large populations.
EMERGING RESEARCH
- How precisely clocks measure an individual, and how reliably on repeat testing.
- Whether interventions that change clock results also improve long-term health.
HOW EPIX AI USES IT
- We measure methylation at around 930,000 CpG sites and use our own age clock, with marks linked to genes active across 40 tissues.
- After 6 months, you retest only your epigenetics, and we recalibrate your model.
References
- Horvath S. DNA methylation age of human tissues and cell types. Genome Biology. 2013;14:R115.
- Hannum G, et al. Genome-wide methylation profiles reveal quantitative views of human aging rates. Molecular Cell. 2013;49:359–367.
- Levine ME, et al. An epigenetic biomarker of aging for lifespan and healthspan. Aging. 2018;10:573–591.
- Lu AT, et al. DNA methylation GrimAge strongly predicts lifespan and healthspan. Aging. 2019;11:303–327.
- Belsky DW, et al. DunedinPACE, a DNA methylation biomarker of the pace of aging. eLife. 2022;11:e73420.
- Horvath S, Raj K. DNA methylation-based biomarkers and the epigenetic clock theory of ageing. Nature Reviews Genetics. 2018;19:371–384.
- Higgins-Chen AT, et al. A computational solution for bolstering reliability of epigenetic clocks. Nature Aging. 2022;2:644–661.
- 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;3:248–257.
See your own genetics and epigenetics
One blood sample: whole genome sequencing, DNA methylation analysis and an estimate of your biological age.
This article is for education and is not medical advice. An estimated biological age is not a clinical diagnosis; please discuss any health decisions with a qualified professional. Prof. Milena Georgieva is Epix AI’s co-founder and Chief Science Officer; videos are credited to their original publishers.