Biological Age: What Those “You’re Really 34” Scores Actually Measure

Somewhere in your life there is a sixty-year-old who out-hikes everyone on the trail and a forty-five-year-old who gets winded on the stairs. Same calendar, very different bodies. That everyday observation is the seed of a serious scientific idea: two people born in the same year can age at different rates, and “biological age” is an attempt to quantify where your body actually is on that journey — not how many birthdays you have collected.

The science behind the idea is real, and parts of it are among the most striking findings in modern aging research. But the number your watch or app hands you under the label “fitness age” or “body age” is not that science, and the commercial tests that do use the real science carry error bars their marketing rarely mentions. This article walks through what biological age genuinely means, which versions of it predict health outcomes, what consumer scores measure instead — and what, practically, is worth doing about any of it.

One Birthday, Two Ages

Chronological age is the single strongest risk factor for most chronic diseases — heart disease, cancer, and dementia all climb steeply with the calendar. But as a measure of an individual body it is blunt, because it treats every 50-year-old as identical. Aging researchers have therefore spent decades hunting for biomarkers of aging: measurements that capture how far the underlying biology has actually progressed in one particular person.

A useful biological-age measure has to clear two bars. It should track calendar age reasonably well across a population — otherwise it is not measuring aging at all. And, more importantly, when it disagrees with the calendar, the disagreement should carry information: someone whose measured age runs five years ahead of their birthdays should, on average, face worse health prospects than someone whose measure runs five years behind. Without that second property, the number is decoration.

There is also a practical reason researchers want such a measure. If you run a trial of an intervention meant to slow aging, you cannot wait thirty years to count deaths — you need an endpoint that responds sooner. A validated biological-age biomarker would be that endpoint: a way to see, within a few years, whether a drug, a diet, or a training programme is bending the curve. That is why so much effort has gone into building and validating these clocks, and why the field holds them to a harder standard than consumer marketing does.

Epigenetic Clocks: The Research Gold Standard

The measure that transformed the field reads chemical marks on your DNA. Methylation — small chemical tags attached to specific positions on the genome — helps switch genes on and off, and the pattern of those tags drifts with age in a surprisingly lawful way. In 2013, Steve Horvath showed that the methylation state of just 353 positions could predict a person’s calendar age across dozens of tissue and cell types with remarkable accuracy. The result was so clean that it earned the nickname “the epigenetic clock.”

Horvath, S. (2013). “DNA methylation age of human tissues and cell types.” Genome Biology, 14(10), R115.

First-generation clocks had a quirk, though: they were trained to predict calendar age itself — a strange target, given that the whole point is to do better than the calendar. Second-generation clocks fixed this by training on health instead. PhenoAge was built against a composite of clinical biomarkers that predicts mortality; GrimAge was assembled from methylation surrogates of blood-plasma proteins and smoking exposure. Both predict lifespan and healthspan markedly better than clocks trained on the calendar alone.

Levine, M.E. et al. (2018). “An epigenetic biomarker of aging for lifespan and healthspan.” Aging, 10(4), 573–591.

Lu, A.T. et al. (2019). “DNA methylation GrimAge strongly predicts lifespan and healthspan.” Aging, 11(2), 303–327.

The Clocks Predict Real Outcomes

This is not a laboratory curiosity. In cohorts of older adults, people whose blood methylation age ran ahead of their calendar age died earlier, on average, than people whose clocks ran behind — and the association held after accounting for the usual risk factors. An accelerated clock, in other words, carries prognostic information that the calendar does not.

Marioni, R.E. et al. (2015). “DNA methylation age of blood predicts all-cause mortality in later life.” Genome Biology, 16, 25.

Reviews of the whole field reach a consistent verdict: among the many candidate measures of biological age — telomere length, blood-chemistry composites, physiological panels, methylation clocks — epigenetic clocks and composite clinical-biomarker scores currently show the strongest links to aging outcomes. Telomere length, the candidate with the best public relations, performs considerably worse than its reputation suggests. It is worth pausing on the second name in that pair: a well-chosen panel of ordinary clinical markers — the kind a standard blood draw already measures — competes respectably with the fancy molecular clocks. The mundane numbers your doctor already orders carry much of the signal.

Jylhävä, J., Pedersen, N.L. & Hägg, S. (2017). “Biological Age Predictors.” EBioMedicine, 21, 29–36.

Your Watch Is Not Reading Your DNA

None of that machinery lives on your wrist. When a watch or an app announces that you are “really 34,” it has not read a single methylation mark. Consumer “fitness age” and “body age” scores are, almost always, your estimated VO2max — the maximum rate at which your body can use oxygen — compared against population norms by age, sometimes blended with resting heart rate, heart-rate variability, activity level, or body composition. If your estimated VO2max matches the average 34-year-old’s, congratulations: you are “really 34.” It is a fitness percentile wearing an age costume.

That does not make it useless — far from it. Cardiorespiratory fitness is one of the strongest modifiable predictors of long-term health in the entire exercise literature, and a “young” fitness age is genuine good news. But be clear about what it is: a fitness measure, not a measure of cellular aging. It can swing within weeks when you train or detrain, while methylation clocks move slowly over years. It knows nothing about your smoking history or your metabolic health beyond what leaks into heart-rate numbers. Treat the two as different instruments that happen to share a unit.

Ironically, the property that disqualifies fitness age as a measure of cellular aging — its responsiveness — is exactly what makes it useful day to day. A number that improves after eight weeks of consistent training is a feedback loop; a number that budges only over years is a report card. Use the watch score as the former and never mistake it for the latter: when it drops by five “years” after a good training block, your cardiovascular system got fitter — which is worth celebrating on its own terms — but your cells did not travel back in time.

What Actually Moves Every Credible Proxy

Here is the convergence that makes the whole debate matter less than it seems: whichever credible measure of aging you favour, the behaviours associated with a younger reading are the same unglamorous list. Observational studies link them to slower epigenetic clocks; exercise physiology links them to a better fitness age; clinical medicine links them to better biomarker panels. Nobody’s data points to a shortcut.

• Build cardio fitness deliberately. Mostly-easy aerobic volume plus a small dose of genuinely hard work each week is the standard recipe for raising VO2max — the engine behind every fitness-age score.
• Lift something twice a week. Muscle and strength decline with age and predict independence late in life; resistance training is the countermeasure, and your grip and leg strength are trackable numbers.
• Guard your sleep like a training session. Short, irregular sleep degrades exactly the metabolic and inflammatory markers that second-generation clocks were trained on.
• Eat mostly unprocessed food, and don’t smoke. Smoking is so tightly linked to accelerated aging that GrimAge literally incorporates a methylation signature of smoke exposure.
• Pick proxies you can measure, and judge trends. A VO2max estimate, resting heart rate, strength numbers, and a yearly blood panel are movable, trackable, and cheap. Read months, not days.

Where Lamplit Fits In

Lamplit deliberately does not hand you a biological-age score — given the noise in the underlying measurements, a single made-up number would be marketing, not information. What it gives you instead is one place to track the proxies that actually move. Logging your workouts and strength sessions, sleep, nutrition, and lab-test results is free, so a blood panel becomes a trend across years rather than a PDF lost in your inbox. With Lamplit Pro, biomarker sync brings your VO2max estimate, resting heart rate, HRV, and body composition in automatically from Apple Health or Health Connect — your watch’s measurements land in the same timeline as your training and your labs.

The payoff is exactly the practice this article recommends: instead of buying an age, you watch your own movable numbers bend. A resting heart rate drifting down across a training block, a deadlift creeping up, a lab marker improving between two annual panels — that is biological-age thinking without the pseudo-precision. On the Pro plan, Genie’s weekly recap summarizes those trends in plain language, so the feedback loop that keeps you training actually closes.

Honest Limits

The commercial methylation kits deserve their own caveat. Repeat testing has shown that the same person’s sample can return epigenetic ages a few years apart, and different labs and algorithms produce different numbers — which means a single result cannot tell you whether last year’s routine “made you younger.” More fundamentally, no intervention has been proven to durably reverse epigenetic age in humans; the trials behind “age reversal” headlines are small, short, and awaiting replication. Treat any product promising to subtract years from your cells with the skepticism it has earned. And remember what these scores are for: they are research tools and rough compasses, not diagnoses. If a health concern is driving your interest in your biological age, that conversation belongs with a clinician, not a kit.

One more distinction keeps expectations honest: everything encouraging in this literature is true at the level of groups. A cohort with faster clocks dies earlier on average; a population that exercises reads younger on average. That does not mean your individual result this month carries the same certainty — individual predictions from any of these measures remain wide. The clocks are excellent epidemiology and, for now, mediocre personal fortune-tellers.

The Bottom Line

Biological age is a genuine scientific concept, and epigenetic clocks — especially the second generation trained on health outcomes — really do predict lifespan and healthspan at the population level. Your watch’s “fitness age” is a different, simpler thing: a fitness metric mapped onto age norms, well worth improving and wrongly named. Consumer methylation kits sit in between — real science, noisy at the individual level. The good news is that every credible version of the number responds to the same inputs. A younger reading, on any instrument, is a byproduct of training, sleeping, and eating well for years. It is not a product you can buy.

Start free on Lamplit — track your workouts, sleep, labs, and strength in one place, and watch the numbers that actually measure your trajectory move in the right direction.