Two people are sixty. One takes the stairs without thinking about it. The other plans the day around their back. Same birthday, a decade apart in everything that matters. You already know which one you are drifting toward, because you have watched it happen to people you know. The open question is whether anyone can hand you the number.
What is biological age?
Your real age counts birthdays. Your biological age is an estimate of how worn your cells are, and the two can disagree by years in either direction.
Think of it this way. Your birthday is the year stamped on the chassis. Your biological age is the mileage. A well-maintained 2010 model outruns a neglected 2020 every time.
So how does a lab read the mileage? It looks at tiny chemical tags on your DNA, called methylation marks, that switch on and off in a fairly predictable pattern as the years pass. Add them up and you get an estimate. Scientists call it an epigenetic clock. Steve Horvath built the first widely used one at UCLA in 2013, and versions of it sit behind every study below.
At the scale of a whole population, it works. Across thousands of people, these clocks predict disease and death better than a birth date does. That much is well established, including by the researchers who are hardest on the test. The argument is about you.
Can you actually lower your biological age?
Possibly, a little. But before you trust any headline about it, notice the pattern underneath all three of the famous studies: the study that proves the product was usually paid for by the people selling it. In this field that is not a scandal. It is the default. Here they are.
The strongest one is a real trial. CALERIE randomly assigned 220 healthy adults to eat less for two years. The target was 25% fewer calories; people actually managed 11.9%. One clock showed them aging about 2 to 3% slower. Two other clocks showed nothing at all, in the same people, from the same blood. And the line you have read everywhere, that this means a 10 to 15% lower chance of dying early, is not a result of this trial at all. It is the authors' own careful guess, borrowed from separate research. CALERIE measured no deaths.
The "three years younger in eight weeks" study was real and randomized: 18 men treated, 20 as controls. It was paid for by Metagenics, whose supplements were the treatment, and corrected in March 2024, rewriting its abstract, its results and two of its figures.
The trial that started the whole "reversal" story, TRIIM, was nine men with no control group, funded by a company its lead author co-founded and holds shares in. Six months on, the effect was drifting back.
The same shape sits inside the NMN trial everyone quotes.
Why can't a test just tell you your number?
Because the clock disagrees with itself.
In 2022, researchers ran the obvious experiment almost nobody had bothered with. They took one blood sample, split it in two, and put both halves through the same clock. The two halves came back a median of 1.8 to 2.4 years apart. In the worst cases, 8.6 years apart. Same blood. Same day. Same machine.
So if a test tells you that you are three years younger than your birthday, and the test cannot tell three years from nothing, the three years are not information. They are static.
One person put that to the test in public. A writer sent his own sample to seven companies over two days and paid $3,143.99 for the answers. They came back ranging from 24.9 to 38. That is one man's shopping trip, not a study. But it is a thirteen-year spread from one body in one week, and it is exactly what the split-sample data predicts.
The person saying this loudest is not a critic. It is Steve Horvath, who built the clock. His own score moved four years in six months. "You cannot directly relate it to how long you'll live," he told MIT Technology Review in 2022. Morgan Levine, who built one of the other clocks, told the same magazine that an individual reading can be off by decades. In 2026, Horvath told a precision-medicine conference that you still cannot send one sample to two properly validated labs and get the same answer back.
That 2022 paper also published a statistical fix, and it works. The corrected clocks are not the ones being sold to you.
Should you buy an at-home biological age test?
Not to answer the question you actually want answered.
If you are buying it to learn how old you really are, it cannot tell you. That is not an accusation against any one company; it is just what the numbers above mean.
There is one honest use: read the slope, never the number. One lab, at least three tests, six months apart, same conditions each time, and ignore any single reading. That is a real discipline. Almost nobody keeps it, which is why the test usually sells a feeling rather than a fact.
The price moved the wrong way, too. The full panel this article once pointed readers to has gone from $229 to $499. The $229 tier still exists, but it now buys a smaller, different test. Nothing about the instrument improved. What improved was the story around it.
If you want numbers that move and mean something, the boring ones are better. How out of breath you get on a hill. How hard you can squeeze. Your blood pressure. Your HbA1c, which is your average blood sugar over about three months. Your hs-CRP, a cheap blood test for hidden inflammation. They cost less, they repeat cheaply, and they have decades of outcomes standing behind them.
What changed in June 2026?
Something real, and it is not in a bottle.
In 2020, David Sinclair's lab at Harvard published a result in Nature that read like science fiction. They used a harmless virus to carry three genes into the eye cells of mice, asking those cells to reload an older, cleaner copy of their own instructions. Mice whose optic nerves had been crushed could see again. Naturally aged mice got sharper sight back.
For six years the honest caption on that was: mice only, humans someday.
Someday arrived on 9 June 2026. Life Biosciences, which Sinclair co-founded, dosed the first human being ever with this kind of cell reprogramming. It is a Phase 1 safety trial, in people with glaucoma, asking whether it is safe inside an eye. It is not an anti-aging treatment, and nobody can buy it.
Notice what that trial measures, though. Not a clock reading. Whether a person can see. That is the bar this whole field has to clear, and it is the reason the number in your inbox does not matter yet.
What actually works while we wait?
The things that move your biology are the ones you already suspect. Each comes with an honest asterisk.
Sleep is when your body does most of its repair. Seven hours or more, and morning light to set the clock that runs it.
Get out of breath. Cardiorespiratory fitness, which is how well your heart and lungs handle hard work, is one of the strongest predictors of how long people live, and unlike a methylation score, you can feel it change.
Eat less; a window is one way to do it. Salk's early trial found benefits from a 10-hour eating window, and it is an easy rule to hold. Be fair about it though: two larger trials since, TREAT and one in the New England Journal, found the window added nothing beyond simply eating less. So it is a tool, not a trick. We go through that whole argument here.
Heat. University of Eastern Finland researchers followed 2,315 men for about twenty years. Those who used a sauna 4 to 7 times a week had lower rates of dementia and heart disease. It watched people rather than testing them, so it cannot prove cause.
Cold. Winter swimmers studied at the University of Copenhagen had more active brown fat and better blood sugar control. Also a snapshot, not a trial. Here is what cold actually does.
You have two ages. The first is on your licence and it is honest. The second is real, and for now nobody can hand you its number without a margin of error wide enough to drive through. That is not bad news. It just means the answer was never going to arrive in the post. It arrives in how you sleep tonight. Here is how I live with that.
