Biological age is an estimate of how old your body seems from biological markers, as opposed to chronological age, the years since you were born. Many consumer tests estimate it from chemical marks on the DNA in your blood cells, using a model called an epigenetic clock. The science is real, but one person's result is noisy.
That last part matters if you're thinking about paying for a test. According to Stanford Medicine, no direct-to-consumer biological age clock is approved by the FDA as a medical diagnostic tool. This guide covers what these tests measure, how the main clocks were built, how accurate they are, where AI fits, and what the research says about moving the number.
Last reviewed: October 20, 2026.
Key Takeaways
- Chronological age is the years you've been alive. Biological age is a model's estimate of how your body is aging.
- There's no single agreed way to measure biological age. Researchers use DNA methylation, telomeres, RNA, proteins, metabolites and composite scores.
- The first epigenetic clocks, from Hannum and Horvath, were published in 2013. Newer clocks such as GrimAge (2019) and DunedinPACE (2022) were built to predict outcomes or the pace of aging instead.
- A 2022 study in Nature Aging found technical noise alone could shift results by up to 9 years between repeat samples for six well-known clocks.
- In the CALERIE trial, two years of calorie restriction slowed DunedinPACE slightly but didn't significantly change PhenoAge or GrimAge.
- Consumer tests cost several hundred dollars, aren't covered by typical insurance, and aren't used clinically by mainstream clinicians, according to Stanford Medicine.
What is biological age?
Biological age is a measure of how a person's organs and cells are changing over time, as Stanford Medicine puts it. The idea is that cells change in fairly predictable ways as we get older. Some people's cells look older or younger than average for their age.
It isn't meant to be a snapshot of your current health; routine checkups already do that. It's meant to capture the trajectory: how fast the underlying processes of aging seem to be moving. Stanford researchers point out that scientists are still working out how biological age links to lifespan and to healthspan, the years lived in good health.
What's the difference between biological age and chronological age?
Chronological age is a fact. Biological age is an estimate, and the estimate depends on which test made it.
| Chronological age | Biological age | |
|---|---|---|
| What it is | Years since birth | A model's estimate of how your body is aging |
| How it's measured | Your birth date | DNA methylation, blood proteins, routine blood markers or other data |
| Can two tests disagree? | No | Yes. Different tests measure different parts of your biology |
| Can it change? | Only forward, one year at a time | It can move faster or slower than calendar age, and a single result has error built in |
How is biological age measured?
Biological age is measured with several kinds of "predictors," and no single method is agreed on. A 2017 review in EBioMedicine by Jylhävä, Pedersen and Hägg grouped them into six types. It called the epigenetic clock the most promising, but said its true value as a biomarker of aging still needed longitudinal confirmation.
| Predictor | What it looks at | What researchers say |
|---|---|---|
| Epigenetic clocks | Chemical marks (methylation) on DNA, often from blood | Called the most promising type in the 2017 review; many commercial versions read only the DNA in blood cells |
| Telomere length | The protective caps on the ends of chromosomes | One of the oldest proposed markers, but studies suggest it may not be the most accurate |
| Transcriptomic | RNA, the readout of which genes are switched on | Can be noisy, because gene activity changes quickly |
| Proteomic | Levels of proteins in the blood | Some proteins may track how individual organs are aging |
| Metabolomic | Small molecules and markers from metabolism | Consumer versions often use routine blood markers already checked at a physical; their precision is unclear |
| Composite scores | Combinations of several types of markers | The 2017 review saw combining predictors as a promising direction |
The six types come from the 2017 review. The notes in the last column come from that review and from Stanford Medicine researchers interviewed in September 2026.
What is an epigenetic clock?
An epigenetic clock is a statistical model that estimates age from DNA methylation, the chemical tags that help switch genes on and off. As we age, methylation at certain spots on the genome (called CpG sites) changes in a fairly regular way, so a model can learn to read age from it.
Two clocks published in 2013 started the field. Hannum and colleagues, writing in Molecular Cell, built a model from more than 450,000 CpG markers measured in the blood of 656 people aged 19 to 101. Horvath's clock, in Genome Biology, was built from 8,000 samples covering 51 healthy tissues and cell types, and it uses 353 CpG sites.
Horvath reported a median error of 3.6 years in his test data. In plain terms, the clock's estimate was within 3.6 years of people's real age for half of them.
How have epigenetic clocks changed since 2013?
The first clocks were trained to match chronological age. Later ones were trained to predict outcomes instead, such as lifespan or the pace of physical decline.
| Clock | Year | What it was built to capture |
|---|---|---|
| Hannum | 2013 | Chronological age, from blood samples of 656 adults |
| Horvath | 2013 | Chronological age, across most tissues and cell types |
| PhenoAge (Levine) | 2018 | "Phenotypic age," a score built from clinical measures, so the clock reflects differences in lifespan and healthspan |
| GrimAge (Lu) | 2019 | Lifespan, using DNA-methylation stand-ins for seven blood proteins and for smoking pack-years |
| DunedinPACE (Belsky) | 2022 | The pace of aging: how fast 19 measures of organ-system integrity declined over two decades in a New Zealand birth cohort |
DunedinPACE works differently from the others. Instead of giving you an age, it gives a rate, and its developers built it only from probes with good test-retest reliability.
How accurate are biological age tests?
Biological age tests are useful in research groups, but a single person's result can be off by several years. Three findings explain why:
- Technical noise. Higgins-Chen and colleagues reported in Nature Aging (2022) that technical noise produced deviations of up to 9 years between replicate samples for six prominent epigenetic clocks. Retrained versions of the same clocks brought most replicates within 1.5 years of each other.
- Built-in error. Stanford epidemiologist David Rehkopf estimates roughly three years of error in many epigenetic clocks. A result saying you're three years "older" could just as easily mean you're three years younger.
- Different tests, different answers. Two tests can disagree because they measure different things. As Stanford's Tony Wyss-Coray put it, many commercial epigenetic tests are "really just giving you the age of blood cells."
Stanford researchers also note that most of these tests are slow to respond. A real change in your habits can take years to show up in an epigenetic blood test.
Where does AI fit into biological age?
Machine learning is built into these tools. Epigenetic clocks are statistical models trained on large datasets to find age-related patterns. Horvath's clock and DunedinPACE were both built with elastic-net regression, a standard machine-learning method; Horvath's model picked its 353 CpG sites automatically.
Deep learning has been tried too. In 2016, a team at Insilico Medicine, a company developing AI-built biomarkers and drugs, trained 21 deep neural networks on more than 60,000 routine blood test results. The combined model predicted chronological age with a mean absolute error of 5.55 years.
The five markers it leaned on most were albumin, glucose, alkaline phosphatase, urea and red blood cells. The authors disclosed that their company planned to use such markers commercially.
Another approach is organ-level clocks. Stanford's Wyss-Coray lab tracks thousands of blood proteins to estimate how individual organs are aging. For more on the science and the people behind it, read our pillar guide to AI, longevity and David Sinclair.
Can you lower your biological age?
Research hasn't settled it. One randomized trial shows a small slowing in one measure, not a reversal. The CALERIE trial randomly assigned 220 healthy adults to cut calories by 25% or eat as usual for two years.
A 2023 post hoc analysis in Nature Aging found the calorie-restriction group's pace of aging, measured by DunedinPACE, slowed. Their PhenoAge and GrimAge results didn't change significantly, and the authors called the effect sizes small.
That fits what Stanford researchers say more broadly. They point to decades of research linking a longer life to everyday habits: a diet rich in vegetables, whole grains and lean protein, regular aerobic exercise and strength training, enough quality sleep, handling stress well and strong social ties. How reliably any of them shows up on a biological age test is less established.
Should you take a biological age test?
A biological age test is a curiosity purchase, not a medical one. Stanford notes these tests are sold as wellness or informational products. Researchers who study aging suggest it probably won't tell you much you didn't already suspect. Rehkopf points to a "pretty decent correlation" between how healthy people say they are and whether they test older or younger.
If you do take one, a few questions keep the result in proportion:
- Which clock does it use? A first-generation clock, GrimAge and DunedinPACE answer different questions.
- What sample does it read? A blood-based epigenetic test reflects blood cells, not your other organs.
- Does it report an age or a pace? A pace score isn't an age, so don't compare it with your birthday.
- Will you test the same way again? If you repeat it, use the same company and the same time of day, since the mix of blood cells can shift within a day, and read small changes as noise.
- Is it selling you something? Wyss-Coray warned about companies selling "some kind of powder" alongside tests that's "never been tested in humans in a clinical trial."
Stanford geneticist Anne Brunet sums it up: "People need to understand that it's more complex than this one number, and that they should not freak out if a clock says they're older than their calendar age."
We sell supplements, and we don't claim that any of them changes a biological age result. If you're comparing products for other reasons, our no-hype guide to longevity supplements explains what to check on the label.
Frequently asked questions
How do you determine your biological age?
Biological age is estimated with a model, usually an epigenetic clock that reads DNA methylation in a blood sample. Other approaches use blood proteins, routine blood markers or telomere length. Different methods can give different numbers for the same person.
How accurate is a biological age test?
Not very precise for one person. A 2022 study found technical noise could shift results by up to 9 years between repeat samples for six well-known clocks, and a Stanford researcher estimates about three years of error in many clocks.
What's the difference between biological age and chronological age?
Chronological age is the number of years since you were born. Biological age is an estimate of how your body is aging, based on biological markers. Chronological age is fixed; biological age depends on the test used.
Is epigenetic aging reversible?
Research hasn't settled it. In the CALERIE trial, a randomized study, two years of calorie restriction slowed the pace of aging measured by DunedinPACE, but it didn't significantly change PhenoAge or GrimAge.
How much does a biological age test cost?
Consumer biological age tests typically cost several hundred dollars and aren't covered by typical insurance plans, according to Stanford Medicine (2026). None is FDA-approved as a diagnostic tool.
Are free biological age calculators accurate?
Their precision is unclear. Stanford researchers say tools based on vital signs and routine metabolic markers often don't have strong outcome data behind them, so treat a free calculator's number as rough.
Sources
- Williams S. Biological age versus chronological age: What the science says. Stanford Medicine Insights. September 23, 2026.
- Jylhävä J, Pedersen NL, Hägg S. Biological Age Predictors. EBioMedicine. 2017.
- Hannum G, Guinney J, et al. Genome-wide methylation profiles reveal quantitative views of human aging rates. Mol Cell. 2013.
- Horvath S. DNA methylation age of human tissues and cell types. Genome Biol. 2013.
- Levine ME, et al. An epigenetic biomarker of aging for lifespan and healthspan. Aging (Albany NY). 2018.
- Lu AT, et al. DNA methylation GrimAge strongly predicts lifespan and healthspan. Aging (Albany NY). 2019.
- Belsky DW, et al. DunedinPACE, a DNA methylation biomarker of the pace of aging. eLife. 2022.
- Higgins-Chen AT, et al. A computational solution for bolstering reliability of epigenetic clocks. Nat Aging. 2022.
- Waziry R, Ryan CP, et al. Effect of long-term caloric restriction on DNA methylation measures of biological aging in healthy adults from the CALERIE trial. Nat Aging. 2023.
- Lane E, Mamoshina P, et al. Deep biomarkers of human aging: Application of deep neural networks to biomarker development. Aging (Albany NY). 2016.
Disclosure: OPA Nutrition sells supplements, including products mentioned in this article. This content is for education only and is not medical advice. Statements about dietary supplements have not been evaluated by the Food and Drug Administration. These products are not intended to diagnose, treat, cure, or prevent any disease. Talk to your healthcare provider before using any supplement, especially if you take medication or have a medical condition.
