The short answer: Not reliably for one person, at least not yet. Epigenetic clocks such as GrimAge and DunedinPACE predict disease and death in large research groups. But a single consumer result can move by years from technical noise alone, and different tests often disagree. Treat the number as a rough research signal, not a diagnosis.
What does a biological age test actually measure?
Most consumer tests read DNA methylation, small chemical tags on DNA that shift as you get older. An algorithm, called an epigenetic clock, turns the pattern of tags at selected sites into an age or a speed.
The clocks fall into three rough generations:
- Horvath clock (2013). Trained to predict calendar age. It uses 353 CpG sites and was built from about 8,000 samples across 51 tissues and cell types.
- PhenoAge (2018) and GrimAge (2019). Trained on health signals, not birthdays. PhenoAge was built from clinical measures linked to lifespan and healthspan. GrimAge combines methylation stand-ins for seven blood proteins and smoking pack-years.
- DunedinPACE (2022). Gives a speed, not an age. It was trained on decline in 19 indicators of organ-system health tracked over two decades in people born in 1972 and 1973.
Other products measure telomere length (the protective caps on chromosomes) or combine routine blood markers into a single score. Only DunedinPACE was designed to answer the question "how fast am I aging right now?" The others estimate where you are, not your rate of change.
Do the results predict disease and lifespan?
In large groups, yes. The GrimAge paper reported that it predicted time to death, coronary heart disease, and cancer better than earlier clocks, using validation data from thousands of people. The PhenoAge team reported that it beat earlier clocks for mortality, cancers, physical function, and Alzheimer's disease.
DunedinPACE was linked to morbidity, disability, and mortality in five outside datasets. Its effect sizes were similar to GrimAge, and it added some prediction beyond GrimAge.
Telomeres are weaker. A 2018 meta-analysis of 121,749 people found that each standard deviation shorter telomere length meant only a 9% higher risk of death (hazard ratio 1.09). The authors noted that measurement method and age at testing drove much of the variation between studies.
The catch is that group prediction is not individual prediction. A clock can shift average risk across 10,000 people and still tell you little about your own future.
How much can your result change from one test to the next?
More than most buyers expect. A 2022 study in Nature Aging found that technical noise produced deviations of up to 9 years between replicates of the same sample for six prominent clocks. That is the lab measuring the same DNA twice, not your body changing.
The same team built principal-component versions of those clocks. These brought most replicates within 1.5 years of each other. Note that the authors disclosed licensing ties to a company that sells an epigenetic test.
DunedinPACE was built to limit this problem. Its developers excluded methylation probes with low test-retest reliability, and they reported high test-retest reliability for the final measure.
Different methods also disagree with each other. In 964 middle-aged adults in the Dunedin Study, researchers compared 11 measures of biological aging, including telomeres, three clocks, and blood-marker composites. They found low agreement between the measures, and effect sizes against aging outcomes were modest. In other words, two "biological age" tests may not measure the same thing.
Are biological age tests reviewed by the FDA?
Generally not as medical devices. Many are offered as laboratory developed tests, which are designed and run inside one lab. In May 2024 the FDA issued a rule to treat these tests as devices. A federal court vacated that rule on March 31, 2025, and the FDA formally reverted its regulation on September 19, 2025.
So the FDA does not routinely review whether a lab's aging score is accurate or useful before it is sold. If a company claims FDA clearance, ask for the clearance number and check what it covers. A clearance for one lab method does not validate the aging algorithm built on top of it.
Does knowing your biological age change your health?
There is little direct evidence that it does. We found no large trial showing that people who learn their epigenetic age live longer or get sick less.
The closest evidence comes from genetic risk scores. A 2016 BMJ meta-analysis of 18 trials found that giving people DNA-based risk estimates did not change smoking, diet, physical activity, or motivation to change. The authors concluded the expectation that such results change behavior "is not supported by existing evidence." Aging scores are a different test, but the same human problem applies: a number alone rarely changes habits.
Clocks are more useful in research, where they serve as early readouts in trials. In the CALERIE trial, 220 adults were randomized to 25% calorie restriction or their usual diet for 2 years. Restriction slowed DunedinPACE but did not significantly change PhenoAge or GrimAge, and the effects were small.
That split shows a key limit. Whether you "got younger" can depend on which clock you happen to use.
What should you track instead?
The measures with the strongest links to healthy aging are cheaper and already known to your doctor. Blood pressure, LDL cholesterol, blood sugar, and waist size are covered in our guide to metabolic health markers to track. Fitness markers such as VO2 max also predict outcomes, and you can improve them with training you can feel.
The habits that move these markers are the same ones reviewed in the science of longevity habits: regular exercise, not smoking, good sleep, and a mostly whole-food diet.
Be wary when a test report leads straight to a product list. If an aging score is used to sell pills, check the evidence for each ingredient on its own, for example at supplements.md.
If you still want a test, use it carefully:
- Pick a pace measure such as DunedinPACE if you want a rate, not an age.
- Use the same company, sample type, and time of day for any repeat test.
- Ignore changes smaller than the test's known noise. Ask the company for its replicate error.
- Bring any worrying result to a clinician, along with your standard labs.
The bottom line
Biological age tests measure something real. GrimAge and DunedinPACE track disease and death risk in large research cohorts. But one result can drift by years from lab noise, different clocks disagree, and the FDA does not routinely review these tests. There is no evidence yet that knowing your number improves your health. Your blood pressure, fitness, and daily habits remain better guides to how you are aging.
Last updated: October 2026. This article is for informational purposes only and does not constitute medical advice. Talk with your doctor before using a biological age result to make health decisions.