Biological Age Tests: What They Measure and What They Miss

Two people turn 55 on the same day. One climbs three flights without thinking about it. The other stops on the landing. The idea behind biological age is that this difference should be measurable, and that the measurement would be more useful than the birthday.

The idea is sound. The products built on it are at very different stages of readiness.

The three kinds of test on sale#

They get talked about as one thing and they are not.

Epigenetic clocks. These read DNA methylation, chemical tags that sit on DNA and change in patterns as people age. Algorithms trained on thousands of samples convert those patterns into an age estimate. This is the serious end, and the research behind it is real.

Telomere tests. Telomeres are protective caps on the ends of chromosomes that shorten as cells divide. A kit measures average length in white blood cells and converts it to an age. This is the weakest end.

Questionnaire calculators. You answer questions about smoking, activity, sleep, blood pressure and weight, and arithmetic produces a number. These measure nothing biological at all, and are the most transparent about it.

What epigenetic clocks do well and badly#

Across large populations, epigenetic clocks work. Groups whose epigenetic age runs ahead of their birthday go on to have more disease and shorter survival. That association has been reproduced many times and is why the field is taken seriously.

For one person, three problems stack up.

Different clocks disagree. There is no single clock. First-generation clocks were trained to predict chronological age. Later ones were trained to predict mortality or the speed of ageing, which is a different target. Run several on one blood sample and the answers can sit years apart, all of them internally valid.

The measurement is noisy. Test the same person twice a few weeks apart and the result can move by a year or more with nothing having changed. Some of that is biological and some is laboratory variation. For a number people are meant to act on, that is a lot of wobble.

The sample matters. Most clocks are built on blood. The methylation pattern in blood is not the pattern in your liver, brain or heart, and a recent infection changes the mix of white cells being read.

None of this makes the science wrong. It makes a single consumer result a weak basis for a decision.

Telomere tests#

Telomeres do shorten with age on average. The word "average" carries the whole problem.

The spread between healthy people of the same age is very wide, so a single result tells you little about the person holding it. Measurement methods vary in reliability, and the most reliable ones are not usually what consumer kits use. And the relationship runs both ways: several conditions shorten telomeres rather than being caused by short ones.

There are real medical telomere tests, used in specialist clinics to investigate rare inherited syndromes. That is not the same product as a mail-in kit.

The honest comparison#

The right-hand column is the one that decides whether a test is worth doing. A number with no action behind it is entertainment, however sophisticated the laboratory that produced it.

Can you lower it#

Yes, in the sense that the number moves. Stopping smoking, losing excess weight, treating high blood pressure, exercising and sleeping properly all shift measured epigenetic age in studies.

Whether that means the underlying ageing changed, or only that a marker moved, is exactly the question the field has not closed. Markers that move without the outcome moving have a long history in medicine.

The practical position is straightforward. Everything that lowers a biological age reading is worth doing for reasons already established, decades before anyone could measure methylation. You do not need the test to justify the habit, and you should not need the habit justified.

Why our own health age tool exists, and what it is not#

This site has a health age test. It asks eight questions and returns a number.

It is worth being precise about what that is. It is not a measurement of your cells. It is arithmetic on factors with established effects on life expectancy: smoking, movement, sleep, blood pressure, weight, alcohol, diet and how often you see people. The output is your age adjusted by those factors.

It earns its place for one reason. Because each input is something you can change, the tool can show what each change is worth, and let you toggle one and watch the number move. That is a decision aid, not a diagnosis, and it takes two minutes rather than a laboratory.

The eight questions are also, not by coincidence, close to the eight habits that carry most of the evidence on living longer.

If you have already bought one#

Plenty of people arrive holding a result. It is not wasted, as long as it is read for what it is.

Find out which clock produced it. A first-generation clock trained to predict chronological age is answering a different question from a clock trained to predict mortality or pace of ageing. A report that does not name its method is not worth reasoning from.

Look for the confidence interval, if there is one. Serious reports give a range. A result reported as a single year, with no range, is presenting a noisy measurement as a precise one.

Do not retest to check. A second test a few weeks later will differ, and the difference will tell you about the assay rather than about you. If you retest at all, leave a year and use the same laboratory and the same clock, because switching either makes the comparison meaningless.

Read the inputs, not the output. Most reports come with a lifestyle questionnaire and a list of contributing factors. That list, not the headline number, is the part with actions attached.

The organ-specific clocks#

The next generation of this research is more interesting than the consumer version. Rather than one number for the whole body, these methods estimate ageing separately for the heart, brain, liver, kidney and immune system, usually from proteins in blood rather than from DNA methylation.

The appeal is obvious. A person whose kidneys are ageing quickly and whose heart is not has a specific problem with specific management, which is a far more useful output than a single figure.

This is early work. It is being published rather than sold, which is a good sign about where it sits, and it is worth knowing about mainly so that a current consumer test is not mistaken for it.

At work#

Biological age testing has started to appear inside workplace wellness programmes and executive health screens, usually bundled with a body composition scan and a lipid panel.

Two cautions belong with that.

The first is medical. A screening test with no agreed threshold and no defined action produces anxiety and follow-up investigations without a matching benefit. That is the general problem with unfocused screening, and it applies here with more force than usual because the measurement is noisier than a blood test.

The second is about privacy. A result suggesting accelerated ageing is health information, and where it ends up matters. Anyone offered this through work is entitled to ask who sees the result, how long it is kept, and whether it is linked to their name.

The parts of an executive health screen that earn their place are the boring ones already in every national screening programme: blood pressure, lipids, blood sugar, and age-appropriate cancer screening.

What I get asked#

People arrive with a printed result saying they are 9 years older than their birthday, and they want to know what to do about it.

The conversation always ends in the same place. Whatever the report said, the useful questions are what their blood pressure is, whether they smoke, how many days a week they move, how they sleep, and when they last had the screening appropriate for their age. Those five answers explain most of any biological age result, and each one has something behind it that a doctor can act on.

I have never seen a biological age result change what should be done. I have seen a good one used as a reason to defer a blood pressure check, which is the specific harm worth naming.

The bottom line#

Epigenetic clocks are real science at an early stage, sold as a finished consumer product. Telomere kits are weaker still. Questionnaire calculators measure nothing biological and are honest about it, which makes them more useful than they sound. If you want to know how you are ageing, the answer is already available from a blood pressure cuff, an HbA1c, a waist measurement and how you feel on the third flight of stairs. Those cost almost nothing, agree with each other, and each comes with something you can do.