Can Wearables Estimate Biological Age?

You went for a run, and your watch told you that you’re 47. You are 39. Or the ring on your finger announced that your heart is younger than you are, which was nice of it. Humanic Blue is asking what a wrist or finger sensor can really know about biological age.

It’s a slightly strange moment in history: the same gadget that counts your steps now has thoughts about your age. And since the question underneath is really about living longer and better, we wanted to know what’s behind those numbers.

The short answer is mixed. Wearables can estimate a fitness-related age, and measures behind it—treadmill-assessed fitness and resting heart rate—are associated with mortality in large groups, including a 2018 study of about 122,000 patients in JAMA Network Open and a 2016 meta-analysis of 1.2 million people in CMAJ.

What they can’t do is measure epigenetic biological age. The wearable features reviewed here do not read the chemical marks on DNA called methylation, and none of the consumer “age” features we reviewed has published validation against death, heart attack or disability. A 2026 study showed that Oura Ring pulse signals can support vascular-age estimation, but it tested a research model rather than Oura’s proprietary consumer score and did not study clinical outcomes. WHOOP’s white paper lists longitudinal validation as future work.

Each device gets to its number a different way. So we went through the manuals, the white papers and the validation studies, one route at a time.

How do Garmin Fitness Age and Apple Cardio Fitness differ?

The first route estimates VO2 max—the maximum oxygen your body can use per minute and kilogram—then looks up the age when that value would be typical. A laboratory measures VO2 max with graded exercise and a face mask. Your watch has no mask, which is an important plot point.

Garmin uses a proprietary Firstbeat algorithm to interpret heart rate and pace, then compares the estimate with age-and-sex tables. Apple also estimates VO2 max from heart-rate response during qualifying outdoor walks, runs or hikes with adequate GPS and signal quality; treadmill workouts do not count. Garmin expresses a comparison as Fitness Age. Apple’s current support page describes a VO2 max estimate and categories compared with people of the same age and sex, rather than a native “fitness age” in years. Neither is an age measured inside you.

The scientific ancestor is Norway’s HUNT Fitness Study and its free NTNU Fitness Calculator, built from several thousand adults who completed treadmill testing. Fitness age repackages a serious clinical measure. The open question is how well the wrist guesses it.

How wrong is the watch’s VO2 max?

Evidence label: Supported, with wide individual error.

Across five validation studies, the average error ranged from small to an underestimation of roughly 4 to 6 ml/kg/min. The spread around those averages was wider, which is the part that matters when one person looks at one watch. The newest Apple Watch studies, for example, found enough individual variation to limit the estimate’s clinical use even when it remained useful for studying groups.

Because a fitness age is a lookup of that estimated VO2 max, the error travels with it. Two people with identical lab results can get different fitness ages from the same brand of watch. The study-by-study numbers are below.

Con: the algorithm is licensed and proprietary, the population tables aren’t yours, and a software update can move your number without your body changing.

One Garmin study added a wrinkle for anyone who’s highly trained: its estimates were less accurate in that group than among moderately trained athletes. In this study, the better-trained participants were more likely to have their fitness underestimated.

Is Oura’s Cardiovascular Age or WHOOP Age the same thing?

No. Oura and WHOOP take two different routes.

Oura’s Cardiovascular Age uses the ring’s light-based pulse signal to estimate pulse-wave velocity, the speed at which a pulse travels through arteries, which tends to rise as arteries stiffen. It needs at least 14 nights of recent data and reports whether cardiovascular age is below, aligned with or above calendar age. Oura says the ring does not directly measure pulse-wave velocity, is not a medical device and may not give valid results for some conditions or pacemakers.

A 2026 study of about 160 adults showed that a research model could estimate chronological age from Oura pulse signals with about seven years of average absolute error, and older-looking signals were associated with higher systolic blood pressure. It did not test Oura’s proprietary feature or future clinical outcomes. One senior author advised Oura, and two authors later joined an Oura–university laboratory.

WHOOP Age is closer to a scorecard. It combines nine sleep, activity, fitness and body-composition metrics, compares them with recommendations and converts estimated risk into years. Meet every recommendation and WHOOP Age equals calendar age. Company questionnaire data linked the score with self-rated health and chronic conditions, but WHOOP’s own white paper says longitudinal validation is future work.

So the ring’s vascular estimate and the strap’s guideline score cannot be compared with each other. Neither formula is fully auditable by the wearer—which is quite a lot of trust to place in a number that may start commenting on your birthday. The product methods, samples and disclosures

What would make two wearable age labels meaningfully comparable?

Has any wearable “age” been tested against actual outcomes?

Yes, in research rather than a current consumer feature. CosinorAge used one week of wrist movement from about 76,000 UK Biobank participants and about 3,400 US adults. Each additional model year was associated with 8–12% higher mortality risk. The evidence is promising, but the UK sample was overwhelmingly White and healthier than average, and one week may not represent a person’s usual rhythm.

Separate Stanford research repeatedly sampled about 100 adults and found that molecular changes clustered in the mid-40s and early 60s rather than moving smoothly. That suggests a single polished “age” may iron out the biologically interesting wrinkles. Neither study validates a current consumer age feature. Research showing a pattern can be detected is different from evidence that a device’s commercial score predicts health outcomes.

What separates an age model used in research from an age feature sold on a device?

Why do VO2 max and resting heart rate matter so much anyway?

Because both are well-documented risk markers. In about 122,000 treadmill patients followed for a median 8.4 years, higher fitness tracked with lower mortality. A meta-analysis of 46 cohorts covering 1.2 million people likewise associated each additional ten resting beats per minute with 9% higher mortality risk.

Those are strong population associations, not proof that a consumer “age” adds information beyond its ingredients. The raw measurements have evidence. The birthday costume still needs to earn its keep.

When the number is most likely to be wrong

If you take a beta blocker or calcium channel blocker, Apple’s paper reports estimated error fell “from 11.8 ± 4.0 ml/kg/min to 1.6 ± 3.1 ml/kg/min” only once the medication was entered in the Health app. That shows why medication information can materially affect the estimate. It does not provide a fixed conversion from the error to years of fitness age.

If you’re highly trained, Engel’s Forerunner data shows underestimation of 9 to 10%, which could turn a VO2 max of 55 into an estimate near 50 millilitres per kilogram per minute. Those are fitness units, not ages.

And if the number is sitting on your mind, Doherty’s advice applies: “Use your fitness tracker as a guide, but not as your coach, your doctor, or your judge.”

One observation about the situation. Several of these features refresh weekly, and one projects your aging pace from the last 30 days. It’s possible to spend an hour a week managing a number whose error band is wider than the change you’re chasing. Optimizing every second of your existence is probably not the purpose of having more seconds.