Epigenetic clocks are biological tools that calculate a person’s biological age by measuring DNA methylation (chemical tags on the genome that shift predictably as we age). The most accurate versions estimate age to within 3 to 5 years from a blood or saliva sample, making them the most precise aging biomarkers currently available to researchers.
Your DNA Is Tagged With a Ticking Record of Time
Every cell in your body carries the same genetic sequence, yet a liver cell and a neuron behave completely differently. That difference comes from epigenetics (the system of chemical switches that controls which genes are active without changing the underlying DNA letters). The most studied epigenetic switch is DNA methylation, the process by which a methyl group (a carbon atom bonded to three hydrogen atoms) attaches to specific cytosine bases in the genome and typically silences nearby genes.
What scientists discovered over the past two decades is that methylation patterns across hundreds of specific genomic sites change in a remarkably consistent, age-related way across virtually all human tissues. The strong correlation between aging and DNA methylation levels has been recognized since the late 1960s, but the first robust demonstration that methylation data from saliva could generate age predictors came from a UCLA team that included Steve Horvath, who later developed the most widely cited epigenetic clock. These predictable patterns became the foundation for mathematical models now called epigenetic clocks. Feed the methylation data from those sites into a trained algorithm, and the algorithm outputs a biological age estimate.
The gap between that biological age and your chronological age (years since birth) is one of the most actively studied numbers in longevity science today.
How Epigenetic Clocks Actually Work
Epigenetic clocks work by training a mathematical model on DNA methylation data from thousands of people of known age, then using that model to estimate biological age from a new sample. Researchers identify which methylation sites change most consistently with age, assign each site a mathematical weight, and produce a formula that converts raw methylation readings into an age estimate.
The measurement itself requires sequencing or array-based profiling of the genome. The most common platform used in research is the Illumina EPIC array, a chip that reads methylation levels at over 850,000 CpG sites (locations in the genome where cytosine is followed by guanine, the preferred attachment points for methyl groups). DNA methylation is measured at each site using bisulfite conversion, a chemical process that transforms unmethylated cytosines so they can be distinguished from methylated ones during sequencing.
| Step | What Happens |
|---|---|
| 1. Sample collection | Blood, saliva, or tissue is collected from the subject |
| 2. Bisulfite conversion | A chemical process converts unmethylated cytosines so they can be distinguished from methylated ones |
| 3. Array or sequencing | Methylation levels at CpG sites are measured, expressed as beta values from 0 (unmethylated) to 1 (fully methylated) |
| 4. Clock algorithm | A trained regression model multiplies each site’s beta value by its learned weight and sums the results |
| 5. Age output | The algorithm returns a biological age estimate in years |
The entire process can be completed in a clinical laboratory in two to three days. As sequencing costs fall, some newer approaches use targeted sequencing rather than full arrays, which may bring costs below $100 per sample in the near future.
The Major Clocks Researchers Use Today
Five epigenetic clocks dominate current research, each trained on different populations and optimized for different outcomes. Understanding the key differences matters for interpreting research findings correctly.
| Clock | Developer | Year | CpG Sites Used | Primary Prediction Target |
|---|---|---|---|---|
| Horvath Clock | Steve Horvath | 2013 | 353 | Chronological age across tissues |
| Hannum Clock | Gregory Hannum | 2013 | 71 | Chronological age in blood |
| PhenoAge | Morgan Levine | 2018 | 513 | Biological age using clinical biomarkers |
| GrimAge | Lu et al. | 2019 | 1,030 | Mortality and disease risk |
| DunedinPACE | Belsky et al. | 2022 | 173 | Pace of aging over time |
The Horvath Clock, published in Genome Biology in 2013, was a landmark because it was trained on data from more than 8,000 samples spanning 51 healthy tissues and cell types, becoming the first clock that worked across nearly every cell type in the body. Before this breakthrough, age predictors were tissue-specific: a clock built on blood cells was useless for analyzing brain tissue. The Hannum clock, published the same year, used whole blood from 656 individuals aged 19 to 101 and achieved correlation with chronological age above 91%. The Horvath clock remains the most cited epigenetic clock in scientific literature.
GrimAge, published in the journal Aging in 2019, shifted the field by training its model on time-to-death and smoking-pack-years rather than chronological age alone, making it one of the strongest predictors of mortality risk. A NHANES-based study of 2,105 adults followed for a median of 17.5 years found that GrimAge acceleration was the strongest predictor of overall mortality among all tested clocks, and was the only clock that independently predicted cardiovascular mortality, with a hazard ratio of 1.61 per five-year acceleration (Mendy and Mersha, GeroScience, 2025).
DunedinPACE, published in eLife in 2022, introduced a different philosophy: rather than measuring biological age as a snapshot, it measures the speed at which a person is aging, expressed as years of biological aging per calendar year. A DunedinPACE score of 1.0 means aging at the population average rate, while a score of 1.2 means aging 20% faster than average. It was built from the Dunedin Longitudinal Study, which followed roughly 1,000 individuals born in 1972 to 1973 in New Zealand, tracking 19 biomarkers across 20 years.
Why Biological Age Diverges From Chronological Age
Biological age diverges from chronological age because genetics, lifestyle, environment, and disease all alter the rate at which DNA methylation patterns drift over time. Research published in Nature Aging and JAMA has confirmed that individuals whose biological age runs ahead of their chronological age face elevated risk for heart disease, type 2 diabetes, dementia, and various cancers.
Several well-replicated factors consistently accelerate or slow epigenetic aging.
Factors associated with accelerated epigenetic aging:
- Smoking (one of the strongest known accelerators, with effects detectable across multiple clock models)
- Obesity, particularly excess visceral fat (fat stored around internal organs)
- Chronic psychological stress and elevated cortisol levels
- Sedentary behavior and low physical activity
- Poor diet quality, especially diets high in processed foods and added sugars
- Sleep deprivation and untreated sleep apnea
- Alcohol overconsumption
- Socioeconomic disadvantage and cumulative adversity in childhood
Factors associated with slower epigenetic aging:
- Regular aerobic and resistance exercise
- Caloric restriction and dietary patterns resembling the Mediterranean diet
- High dietary fiber intake
- Non-smoking status
- Normal body weight and waist circumference
- Strong social connections
- Higher education levels (likely a proxy for compounding lifestyle and stress factors)
These associations do not yet prove causation in every case. Longitudinal studies and intervention trials are actively working to establish which lifestyle changes reliably reverse epigenetic age acceleration rather than simply correlating with lower scores.
What Epigenetic Clocks Reveal About Disease
Epigenetic clocks can predict disease onset and mortality years before symptoms appear, making them among the most promising candidates for clinical risk stratification (the process of sorting patients by their probability of future illness).
Research using post-mortem brain samples has produced some of the most striking disease associations. A study of 721 dorsolateral prefrontal cortex specimens from participants in the Religious Orders Study and Rush Memory and Aging Project found that epigenetic age acceleration correlated with neuritic plaques, amyloid load, and pathologic Alzheimer’s disease diagnosis (Grodstein et al., Neurobiology of Aging, 2021). An earlier analysis of 700 prefrontal cortex samples from the same cohort found that epigenetic age acceleration was associated with lower global cognitive functioning, episodic memory, and working memory in Alzheimer’s patients, and that the acceleration itself is heritable with a heritability estimate (h2) of 0.41 (Levine et al., Aging, 2015). A cortical-specific clock trained on post-mortem tissue showed even stronger results, linking each standard deviation increase in cortical clock age to a 90% greater likelihood of a pathologic Alzheimer’s disease diagnosis.
In cardiovascular research, accelerated GrimAge has been independently associated with worse scores on the American Heart Association’s Life’s Simple 7 cardiovascular health scale, as shown in analyses of the CARDIA and Framingham Heart Study cohorts (Zheng et al., Circulation Research, 2022). GrimAge acceleration consistently outperforms other clocks in predicting cardiovascular disease incidence independent of traditional risk factors like cholesterol and blood pressure.
Studies of cancer biology have found that tumor tissue consistently shows significant epigenetic age acceleration compared with adjacent normal tissue from the same patient. An analysis of 6,000 cancer samples from 32 datasets found that all 20 cancer types studied exhibited age acceleration, with an average of 36 years of excess biological aging beyond expected age (Horvath S, Genome Biology, 2013). Certain cancers, including colorectal cancer and glioblastoma (an aggressive brain tumor), show particularly pronounced clock dysregulation.
Centenarians (people who live to 100 or beyond) consistently show lower epigenetic ages relative to their chronological ages when measured in midlife. Whether that slower epigenetic pace is a cause or consequence of exceptional longevity remains an open and compelling question that multiple ongoing studies are working to resolve.
Research into epigenetic age and immune function has revealed another important layer. The immune system relies on rapid cell turnover and continuous renewal of white blood cells, which means that blood-based epigenetic clocks are particularly sensitive to immune aging, a process researchers call immunosenescence (the gradual deterioration of immune function with age). Conditions that accelerate immune aging, including chronic viral infections like untreated HIV and long-term autoimmune diseases, appear as strong biological age accelerators in blood-derived clock measurements. Conversely, populations with very low chronic inflammation tend to show biological ages five to ten years younger than their chronological counterparts, consistent with the hypothesis that reducing systemic inflammation is one of the most impactful levers for slowing epigenetic aging.
Measuring Your Own Epigenetic Age: Consumer Tests
Several companies now offer consumer epigenetic age testing through mail-in kits, with prices ranging from $299 to $549 depending on the panel. Results are typically returned within two to three weeks of sample receipt.
| Company | Approximate Price (USD) | Sample Type | Clock Used |
|---|---|---|---|
| TruDiagnostic | $299 to $549 | Blood (fingerprick) | DunedinPACE, GrimAge, multiple |
| Elysium Health Index | $299 per test | Saliva | Proprietary (Horvath-based) |
| MyDNAge | $299 | Blood or urine | Proprietary multi-tissue clock |
| Epigenomics | $399 | Blood | Multiple validated clocks |
The reliability of consumer tests varies considerably. Some companies use validated, peer-reviewed clocks directly. Others use proprietary algorithms whose exact CpG sites and training datasets are not publicly disclosed, making independent verification difficult. The American College of Medical Genetics does not yet endorse epigenetic age testing for routine clinical use, citing the need for standardized reference ranges and clinical outcome validation across diverse populations.
A single test result is most useful when interpreted as a starting point rather than a definitive judgment. Repeat testing over six to twelve months following a lifestyle intervention tends to be more informative than a one-time snapshot, particularly when using DunedinPACE, which is more sensitive to short-term changes than static biological age clocks.
Interventions That Appear to Reverse Epigenetic Age
Epigenetic age can be reversed, not merely stabilized, according to evidence from multiple clinical trials, challenging the longstanding assumption that biological aging moves in only one direction.
The TRIIM trial (Thymus Regeneration, Immunorestoration, and Insulin Mitigation), led by Dr. Gregory Fahy and published in Aging Cell in 2019, enrolled nine healthy men aged 51 to 65 and treated them for one year with recombinant human growth hormone, metformin, and DHEA (Fahy GM et al., Aging Cell, 2019). The trial was designed to study thymus regeneration, but epigenetic analysis by Steve Horvath revealed a striking secondary finding: participants showed a mean epigenetic age approximately 1.5 years younger than baseline after one year of treatment, translating to a 2.5-year reduction compared with what would have been expected without treatment. The GrimAge predictor showed a 2-year decrease in epigenetic age that persisted six months after discontinuing treatment. The authors described it as the first reported increase in predicted human lifespan achieved through a clinically accessible aging intervention.
The CALERIE trial (Comprehensive Assessment of Long-term Effects of Reducing Intake of Energy), published in Nature Aging in 2023, randomized 220 healthy non-obese adults to 25% caloric restriction or a freely-eaten control diet for two years (Waziry R et al., Nature Aging, 2023). The caloric restriction group showed significant reductions in DunedinPACE scores compared with controls. The intervention effect represented a 2 to 3% slowing in the pace of aging, which researchers calculated translates to a 10 to 15% reduction in mortality risk, comparable to the benefit of helping someone stop smoking. Notably, neither PhenoAge nor GrimAge showed significant changes, confirming that DunedinPACE is more sensitive to short-term intervention effects than snapshot biological age clocks.
Emerging research is also examining the following:
- Rapamycin (an immunosuppressant drug that inhibits the mTOR pathway, a cellular growth and metabolism regulator) in off-label longevity contexts
- NAD+ precursors such as NMN (nicotinamide mononucleotide) and NR (nicotinamide riboside), which support cellular energy production
- Senolytics (compounds that selectively clear senescent cells, aged cells that stop dividing but resist death and secrete inflammatory signals)
- Partial cellular reprogramming using Yamanaka factors (a set of four proteins that can reset a cell’s epigenetic state toward a younger profile)
None of these interventions have yet been validated in large, long-term human trials powered to show mortality reduction. The field is moving rapidly, with dozens of longevity-focused clinical trials currently recruiting across the United States.
Limitations and Scientific Debates
Epigenetic clocks carry important limitations that researchers and consumers alike need to understand before acting on results.
The clocks were largely trained on European-ancestry populations. Performance in populations of African, Latino, East Asian, and South Asian ancestry varies, sometimes significantly, because baseline methylation patterns differ across ancestral groups. A study of 1,100 primarily hypertensive African Americans in the GENOA cohort confirmed GrimAge associations with cardiometabolic risk but specifically cautioned that training on predominantly European samples limits confident cross-population generalization (Biram et al., BMC Medical Genomics, 2021). Efforts to train more population-diverse clocks are underway but incomplete.
Tissue specificity is a persistent challenge. A blood-based clock reflects the aging of blood cells, which turn over rapidly, rather than the aging of slow-turnover tissues like neurons or cardiac muscle. Research published in Genome Biology demonstrated that the Horvath clock systematically underestimates biological age in tissues from older individuals, with the effect most pronounced in the cerebellum, likely because CpG sites approach saturation at extreme ages. Comparing clock scores across tissue types without appropriate calibration can produce misleading results.
There is active scientific debate about what the clocks actually measure. The methylation patterns they track correlate with aging but may reflect cellular turnover rates, cumulative DNA damage, or immune system changes rather than a fundamental aging mechanism. As one analysis in Genome Biology noted, if a CpG site played a direct causal role in aging, the mortality it caused would make it less likely to be observed in older individuals, suggesting the clock sites capture an emergent property of the aging epigenome rather than a direct driver of aging.
Finally, regression to the mean affects repeat testing. A person who tests at an unusually high or low biological age on a first measurement will tend to score closer to average on a repeat test regardless of any intervention, an artifact of statistical noise rather than genuine biological change.
The Future of Epigenetic Clocks in Medicine
Epigenetic clocks are on a trajectory toward routine clinical use alongside cholesterol and blood pressure as standard biomarkers of cardiovascular and metabolic health. The National Institute on Aging has funded multiple large consortium studies to validate clock performance in clinical populations, including the CALERIE Research Network (grant R33AG070455) and the Dunedin Study follow-up (grant R01AG032282). The FDA is actively evaluating whether biological age could qualify as a surrogate endpoint (an accepted proxy for clinical benefit) in drug approval trials targeting aging-related conditions.
Second-generation clocks that incorporate not just DNA methylation but also histone modifications (chemical tags on the proteins around which DNA is wrapped), transcriptomic data (patterns of gene expression), and proteomic signatures (protein levels in the blood) are already showing improved predictive accuracy in research settings. Multi-omics clocks (those integrating multiple biological data types simultaneously) may eventually deliver biological age estimates accurate to within one to two years.
The prospect of validating anti-aging drugs using epigenetic clocks as an endpoint rather than waiting decades for mortality data could accelerate the entire field of longevity medicine. A trial that shows a drug reliably reduces DunedinPACE or GrimAge acceleration over two to three years could, if validated, provide regulatory justification for approval far faster than conventional outcomes-based trials. The CALERIE trial’s finding that a 2 to 3% DunedinPACE reduction translates to a mortality benefit comparable to smoking cessation illustrates the potential clinical significance of even modest clock improvements.
One underappreciated dimension of the field is the intersection of epigenetic aging and mental health. A study published in Translational Psychiatry measured GrimAge in 49 unmedicated individuals with major depressive disorder and 60 healthy controls, finding that people with depression showed significantly greater GrimAge acceleration, with a median of 2 years of excess cellular aging that persisted even after controlling for smoking status and BMI (Han et al., Translational Psychiatry, 2021). Multiple population studies, including analyses from the UK Biobank (a biomedical database containing information from 500,000 UK participants), have found similar acceleration patterns in individuals with PTSD and severe anxiety.
Workplace and environmental exposures are another active frontier. Studies of industrial workers exposed to air pollution, heavy metals, and pesticides show clock acceleration proportional to exposure duration and intensity, providing a molecular mechanism for long-observed disparities in life expectancy between high-pollution and low-pollution communities. Environmental justice researchers are increasingly collaborating with epigenetics labs to document how neighborhood-level exposures drive biological aging disparities across racial and economic groups in the United States.
The lifestyle choices most consistently linked to slower epigenetic aging are the same ones associated with better cardiovascular, metabolic, and cognitive health. Regular physical activity, a diet rich in whole foods and fiber, adequate sleep, stress management, and not smoking are not merely generic health advice. They are, according to the epigenetic clock literature, reliably associated with younger biological ages at the molecular level.
Frequently Asked Questions
What exactly is an epigenetic clock?
An epigenetic clock is a mathematical model that calculates biological age by measuring DNA methylation patterns at specific sites in the genome. These chemical tags on DNA shift in predictable ways as a person ages, allowing researchers to estimate how old someone’s cells are regardless of their birth year. The Horvath Clock, trained on more than 8,000 samples from 51 tissue types, can estimate chronological age to within 3 to 5 years from a blood or saliva sample.
How is biological age different from chronological age?
Chronological age is simply the number of years since you were born, while biological age reflects how well or poorly your cells and tissues are aging relative to the average person your age. Two people born in the same year can have biological ages that differ by 10 or more years based on genetics, lifestyle, and disease history. A biological age younger than your chronological age generally signals lower disease risk and better long-term health outcomes.
Can you actually reverse your epigenetic age?
Evidence from clinical trials suggests that certain interventions can reduce epigenetic age by measurable amounts. The 2019 TRIIM trial published in Aging Cell found that a combination of growth hormone, DHEA, and metformin reduced epigenetic age by approximately 2.5 years compared to no treatment over one year in nine men. The 2023 CALERIE trial published in Nature Aging showed that caloric restriction slowed the pace of aging by 2 to 3% as measured by DunedinPACE, an effect researchers equate to a 10 to 15% reduction in mortality risk. No intervention has yet been proven in large-scale studies to extend human lifespan through epigenetic age reduction.
What does GrimAge measure and why does it matter?
GrimAge is an epigenetic clock trained to predict time-to-death and smoking exposure rather than chronological age, making it one of the strongest predictors of mortality risk currently available. In a NHANES cohort study of 2,105 adults followed for a median of 17.5 years, GrimAge acceleration was the most significant predictor of overall mortality among all tested clocks, and the only one to independently predict cardiovascular mortality. Each five-year increase in GrimAge acceleration carried a hazard ratio of 1.61 for cardiovascular death in that cohort.
How much does an epigenetic age test cost?
Consumer epigenetic age tests typically range from $299 to $549 depending on the company and the number of clocks included in the analysis. Some panels report a single clock estimate while others simultaneously report multiple clocks such as GrimAge, PhenoAge, and DunedinPACE. Insurance does not currently cover these tests, and most are marketed as wellness tools rather than clinical diagnostic devices.
What lifestyle factors speed up epigenetic aging the most?
Smoking is consistently identified as one of the strongest accelerators of epigenetic aging across multiple clock models, with effects detectable even at moderate levels of tobacco use. Obesity, chronic stress, physical inactivity, and poor diet quality also reliably associate with faster biological aging in population studies. Among all modifiable factors, smoking cessation and maintaining a healthy body weight have the largest and most consistent evidence base for slowing epigenetic age acceleration.
Are epigenetic clocks accurate for all ethnic groups?
Most major epigenetic clocks were trained predominantly on individuals of European ancestry, which limits their accuracy in other populations. A GENOA cohort study of 1,100 African American adults confirmed GrimAge associations with cardiometabolic risk but found that training on European cohorts limits confident cross-population application. Next-generation clocks trained on more diverse datasets are in development and are expected to improve cross-population accuracy significantly.
What is DunedinPACE and how is it different from other clocks?
DunedinPACE measures the speed of biological aging rather than estimating a fixed biological age. Derived from the Dunedin Longitudinal Study following roughly 1,000 individuals born in 1972 to 1973 in New Zealand, it expresses the aging rate as a number where 1.0 equals the population average pace. The 2023 CALERIE trial showed it was significantly more sensitive to caloric restriction than PhenoAge or GrimAge, which showed no significant change, making DunedinPACE particularly useful for evaluating lifestyle and drug interventions over months rather than years.
Can epigenetic clocks detect cancer?
Epigenetic clocks cannot currently diagnose cancer, but research shows tumor tissue consistently exhibits significant biological age acceleration compared with normal tissue from the same patient. An analysis of 6,000 cancer samples from 32 datasets found that all 20 studied cancer types showed age acceleration averaging 36 years beyond expected biological age. Researchers are exploring methylation-based liquid biopsies as early cancer detection tools, and the Galleri multi-cancer early detection test uses related methylation biology, though it employs different signatures from aging clocks specifically.