How old is your body really?

An algorithm for estimating biological age, published by Morgan Levine and colleagues in the journal Aging in 2018. It was built on NHANES III data from 9,926 people and tested on 6,209 more in NHANES IV, and it combines nine routine clinical blood markers with chronological age into a single score that predicts death, illness and physical decline better than age alone. This calculator borrows the framework and maps lifestyle factors onto the markers it uses, because it can't take your blood.

It's a population estimate, not a clinical result. The weights come from very large studies, UK Biobank with 502,489 people and NHANES with more than 30,000, and at group level they line up well with biological age measured in a lab. For any one person the calculator is blind to genetics, existing illness and medication, so read your number as a direction rather than a diagnosis, and as a reason to get the blood test if the direction worries you.

Slightly, on the surface. In UK Biobank men run 1 to 2 years older biologically than women of the same chronological age, but almost all of that is behaviour rather than biology: men smoke more, drink more heavily and carry more weight. Adjust for those and the difference shrinks to under half a year. This calculator compares you with people of your own sex, so the percentile isn't distorted by the gap.

Earlier than you'd think. The Dunedin Study, which has followed 954 New Zealanders since birth, found measurable differences in the pace of ageing by 26, and by 38 its participants had biological ages spanning 28 to 61, a 33-year spread inside one birth cohort. UK Biobank shows the spread widening with age, from a standard deviation of about 3 years among people in their thirties to about 6 years among people in their sixties. The choices compound.

Depends what you'd do with the answer. Clinical tests, whether epigenetic or blood-based, cost $199 to $499 a go. If you already exercise, sleep and don't smoke, a test can show whether genes or environment are holding a gap open that habits can't close. For someone sedentary, short of sleep or still smoking, this free estimate tells you almost everything the test would, because you already know what to change. Where paid testing earns its money is repeated over time, to see whether what you're doing is working at the cellular level.

Bryan Johnson is the tech founder who has spent something like $2 million a year since 2023 trying to drive his biological age as low as it will go, and his results put the phrase into ordinary conversation. His regime of dozens of supplements, a fixed diet and constant medical monitoring isn't practical for anyone else, but the science underneath it, PhenoAge, epigenetic clocks and UK Biobank, is the same research this calculator draws on. The consistent finding from that research is that 80 to 90% of the benefit comes from the boring, free things: exercise, sleep, not smoking and a healthy weight. What's expensive makes headlines and moves the needle very little.

Yes, within limits, and there's a clinical trial as well as the cohort data. The TRIIM trial (Fahy et al. 2019, Aging Cell) recorded an average epigenetic age reversal of 2.5 years over twelve months in a small group of volunteers. In the general population, Li et al. (2023) found UK Biobank participants who improved their exercise, sleep and diet over a ten-year follow-up slowed their ageing measurably. The two biggest single moves remain giving up smoking, worth up to 5 to 7 years back within a decade, and getting active after being sedentary, worth 2 to 4 years over five. Bodies respond faster than the people living in them expect. (Sources: Fahy GM et al. Aging Cell. 2019;18(6):e13028. Li X et al. Nature Communications. 2023;14:7793.)

Epigenetic age is one way of measuring biological age, not another name for it. Biological age is the broad idea of how old your body is functioning compared with the population. Epigenetic clocks such as Horvath's and Hannum's (both 2013) read patterns of DNA methylation, chemical marks on DNA that shift predictably as we age. PhenoAge (Levine 2018) uses clinical blood markers instead of methylation but tracks the epigenetic clocks closely, and in several validation studies it predicts health outcomes better than they do. Telomere length, composite biomarker scores and functional tests like grip strength and walking speed are other routes to the same idea. This calculator is calibrated against PhenoAge norms.

Exercise, by a distance. Li et al. (2023) found UK Biobank participants meeting the WHO's 150 minutes of moderate exercise a week tested 2 to 4 years younger than sedentary people of the same age, even after adjusting for everything else. Smoking is the strongest accelerator, adding 4 to 10 years for current heavy smokers in Levine's 2018 data, and sleep comes third, with regular nights under six hours worth 2 to 4 years of extra age. Together they compound into a gap of 8 to 12 years between someone who exercises, sleeps and has never smoked and someone with the opposite habits. If you want a cheap way to check the changes are working, blood pressure responds to them within weeks, and the population figures for life expectancy and the other health markers are gathered on the health statistics page. (Sources: Li X et al. Nature Communications. 2023;14:7793. Levine ME et al. Aging (Albany NY). 2018;10(4):573-591.)

Yes, and usually by more than people think. Chronological age is years since birth and advances one year at a time. Biological age is how old your body is at the level of cells and organs, and it speeds up or slows down with lifestyle, genes and environment. Levine's PhenoAge work showed that clinical markers predict mortality far better than birthdays. Across UK Biobank's half a million participants, lifestyle explains most of the difference between the two ages, the median gap between them is about 4.7 years, and people who exercise and sleep well typically sit 3 to 8 years on the younger side. (Sources: Levine ME et al. Aging (Albany NY). 2018;10(4):573-591. Li X et al. Nature Communications. 2023;14:7793.)