Specialty Testing Guide › TruAge Biological Age Testing
An epigenetic clock is a formula that reads chemical tags on your DNA, called methylation marks, and turns them into a single number. Depending on the clock, that number is an estimated age, a predicted risk, or a speed of aging.
Clocks have changed a lot since the first ones appeared in 2013. Early versions mostly guessed your birthday. Newer versions try to predict health, disability and death, and the newest try to measure how fast you are aging right now. Knowing which kind you are looking at is the first step to reading any result sensibly.
Key takeaways
- Each clock is a statistical model built from methylation patterns at many spots on the genome, not a direct look at cell damage.
- The earliest clocks were trained to match calendar age, while later ones were trained on health outcomes such as death and chronic disease.
- Pace-of-aging tools report a rate, where 1.0 means aging about one biological year per calendar year.
- Clocks predict risk well across large groups, but one person's result carries measurement noise and is not a diagnosis.
What an epigenetic clock actually reads
Your DNA sequence stays the same for life. On top of it sit small chemical tags that help control which genes are switched on in each cell. The most studied tag is a methyl group attached at sites called CpG sites.
At some of these sites, the share of DNA copies carrying a tag drifts in a predictable way as people get older. Researchers measured hundreds of thousands of sites in large groups of people, then used statistics to find the combination of sites that best predicted a target. That combination, with its weights, is the clock.
So a clock is only as meaningful as the target it was trained on. That is the main difference between the three generations.
First generation: trained on calendar age
The first widely used multi-tissue clock, published by Steve Horvath in 2013, was built from 8,000 samples covering 51 healthy tissues and cell types.1 It uses 353 CpG sites and estimates age in years for most tissues.
Around the same time, a blood-based model by Hannum and colleagues showed the same thing in blood: methylation tracks age closely.
The catch is that a clock trained to match your birthday works best when it matches your birthday. The leftover gap between clock age and real age was hoped to reflect biology. In an Irish study of older adults, the Horvath and Hannum gaps did not predict walking speed, frailty, cognitive test scores or death, while GrimAge did.2
Second generation: trained on health and mortality
Second-generation clocks were trained on outcomes that matter. PhenoAge was built in two steps, first from routine blood chemistry linked to mortality, then from methylation, and it outperformed earlier clocks for death, cancer, physical function and Alzheimer's disease.3
GrimAge goes further: it combines methylation-based estimates of seven blood proteins and of smoking history into a single lifespan predictor. In the original validation data, it predicted time to death, coronary heart disease and cancer better than the clocks before it.4
Third generation: measuring the pace of aging
DunedinPACE, often called a third-generation tool, asks a different question. Instead of how old you seem, it asks how fast you are aging. It was trained on people from Dunedin, New Zealand, born in 1972 and 1973, whose decline across 19 measures of organ-system health was tracked over two decades.5
The result is a rate. A value of 1.0 means about one year of biological aging per calendar year, compared with the reference group.
| Generation | Example clocks | Trained to predict |
|---|---|---|
| First | Horvath, Hannum | Calendar age |
| Second | PhenoAge, GrimAge, GrimAge2 | Health decline and time to death |
| Third | DunedinPACE | Speed of decline in organ-system measures |
What clocks can and cannot tell you
In large studies, people whose second- and third-generation clocks run fast are more likely to develop age-related disease and to die sooner. That is a real and useful finding for research.
For one person, the picture is less precise. A study of repeated measurements on the same samples found technical noise produced deviations up to 9 years between replicates for six prominent clocks.6 Newer methods reduce that noise, but a small change between two tests may not be real.
A clock result is also not a diagnosis. It cannot tell you that you have heart disease or diabetes, and there is little evidence yet that moving your number changes what actually happens to your health. Your blood pressure, cholesterol, blood sugar, weight and smoking status remain the measures that standard care acts on.
Where to go from here
If you want to understand the two clock types you are most likely to see on a report, read what DunedinPACE measures and how GrimAge works. For the full picture of biological age testing at our Conshohocken clinic, start at the TruAge guide hub.
Frequently asked questions
Is epigenetic age the same as biological age?
It is one way to estimate biological age. Other approaches use routine blood chemistry, physical tests or telomere length, and these methods often disagree with each other because they capture different parts of aging.
Can a saliva or cheek swab clock be compared with a blood clock?
Not directly. Methylation patterns differ between tissues, and many clocks were built for blood. Compare results only from the same tissue, the same lab and the same clock.
Why do different clocks give me different ages?
Each one was trained on a different target and uses a different set of DNA sites. A clock built to predict mortality and one built to match birthdays are answering different questions, so their numbers will not line up.
Does a younger clock age mean I am healthy?
No. It is a statistical estimate. You can have a favorable clock result and still have high blood pressure or an early cancer that only standard screening would find.
Related reading
This article is for general education and is not medical advice. Decisions about testing and treatment should be made with your clinician. If you have severe symptoms, call 911. For a suspected poisoning, call Poison Control at 1-800-222-1222.
