Epigenetic age tests

An epigenetic age test estimates your biological age by reading DNA methylation, small chemical tags called methyl groups that sit on specific spots of your DNA. The tags never change your genetic code, but they change which genes get used, a bit like sticky notes that switch genes on or off. As you age, these tags shift in a predictable way, and that predictability is what makes an age estimate possible.^[1] Depending on the clock, an algorithm uses anywhere from 13 of these spots to more than a thousand^{[1, 4, 22]} and turns the pattern into one number: your biological age at the cellular level. That number is a research estimate, not a clinical diagnosis.

Getting from a tube of saliva to that number takes five steps:

1. You collect a sample at home: saliva or a few drops of blood, depending on the test.
2. The lab extracts the DNA.
3. The lab measures methylation at specific CpG sites, the spots where the DNA letter C sits right before a G and where methylation usually lands.
4. An algorithm compares your pattern with reference data.
5. You get a report with your biological age.

That report usually leads with a biological age. Many tests add a pace of aging score (how many biological years you age per calendar year) or separate ages for systems such as immunity or metabolism. Some also attach lifestyle suggestions, and from kit-in-mailbox to result usually takes two to six weeks. Which of these numbers you get, and what they mean, depends on the clock the lab runs.

The clock a lab runs decides what your number means, because each of the main clocks was trained to answer a different question. Horvath's and Hannum's clocks were built to match calendar age.^{[1, 2]} PhenoAge and GrimAge track the risk of disease and death, and DunedinPACE measures how fast you are aging right now.^{[3, 4, 6]} That is why a kit that runs several clocks on one sample can hand you several biological ages at once, which is more birthdays than anyone asked for.

The table shows what each clock estimates and how often a retest makes sense:

Sources: the original paper of each clock [1–6]; the retest intervals are our rule of thumb, explained at the end of this section.

Steve Horvath's 2013 clock was the first that works across many tissue types^[1] and Hannum's blood clock from the same year is often run next to it as a sanity check.^[2] Morgan Levine's PhenoAge went a step further and was anchored to clinical markers linked to mortality rather than to calendar age alone.^[3] In studies PhenoAge predicted mortality, cancer, physical functioning and Alzheimer's disease better than earlier clocks.^[3]; that is a group-level association, not a forecast of your own healthy years.

GrimAge, built by Ake Lu with Horvath as senior author, was designed to predict mortality^[4] It estimates smoking history and seven blood proteins from your methylation pattern, so no extra blood test is needed.^[4] Its newer version GrimAge2 also estimates CRP (an inflammation marker) and HbA1c (long-term blood sugar), and it predicts mortality better than the original.^[5] Many providers do not say which version they run, so check that your report says "GrimAge2" rather than just "GrimAge".

DunedinPACE works differently, because it reports biological years per calendar year from a single sample. A pace of 1.0 is average and lower means slower aging, and in its own paper it predicted health outcomes about as strongly as GrimAge.^[6]

Which one to pick?

For tracking change, DunedinPACE is the clock that has moved in randomized trials. In a two-year calorie-restriction trial it slowed by about 2 to 3%, while the more reliable versions of PhenoAge and GrimAge did not change.^[11] The effect was small, though, and the two groups' trajectories overlapped by close to 90%, so a single person's change says little.^[11] For a one-off reading of health risk, GrimAge2 or PhenoAge fit the question better, and many commercial kits blend several clocks into one report.

Which clock sits inside which product

Most kits sold in Germany, Austria and Switzerland run one of these clocks or a vendor's own model, and two well-known US kits do not ship here at all. Prices follow in the cost section.

• TruDiagnostic TruAge: three clocks in one finger-prick kit, OMICmAge, SYMPHONYAge (ages of 11 organ systems) and DunedinPACE.^[24]
• neotes bioAge: GrimAge method on a dried blood spot, analysed in a German specialist lab, with results in about 4 weeks; neotes does not say which GrimAge version it uses.^[23]
• epiAge: saliva test from a German seller in Starnberg that reads just 13 CpG sites by NGS sequencing, its own small panel rather than the Horvath clock.^{[21, 22]}
• MyAgingTests: home blood GrimAge test from the Clock Foundation, the nonprofit behind Steve Horvath's clocks, with an international kit for the UK, the EU and Canada.^[28]
• MyDNAge: Horvath-style clock that has offered blood or urine sampling, which is rare (not re-checked in October 2026).
• NOVOS Age: biological age, a pace of aging built on DunedinPACE, 11 organ ages and telomere length, but it ships only within the US and Canada.^[25]
• Elysium Index: PhenoAge-lineage saliva clock (developed with Morgan Levine),^[26] but Elysium ships and bills Index only to US addresses, so it is not an option from Germany, Austria or Switzerland.^[27]
• Muhdo: UK saliva test that reports biological age plus scores such as inflammation and immunity.^[30]
• GlycanAge: not epigenetic at all. It measures IgG N-glycans (sugar chains on your antibodies) as a marker of inflammatory age, so it complements a methylation clock rather than replacing one.

How often should you retest?

No study has worked out how often a consumer should retest, and the lab noise on the older clocks is the main reason to go slowly. When the same sample is measured twice, the older clocks typically differ by about 1 to 2.5 years and occasionally by up to about 9 years.^[9] PhenoAge swung the most, up to 8.6 years, and GrimAge the least.^[9] Even the typical swing is larger than one year of aging, so a yearly retest on these clocks mostly shows noise. PC-clock versions condense the information from tens of thousands of sites into principal components, which cuts the typical swing to well under a year, and over 90% of their repeat measurements agreed within 1 to 1.5 years.^[9]

Our rule of thumb follows from that noise: at most once a year for DunedinPACE, and every 18 to 36 months for the older clocks. Use the same provider and sample type each time, and retest only after a change you have kept up for months. Otherwise there is nothing real for the clock to pick up.

Most epigenetic age kits cost about 200 to 500 euros per test as of October 2026,^{[21, 23, 24]} and in Germany, Austria and Switzerland you pay all of it yourself, retests included. Prices move often, so treat the table as orientation and confirm the current price on the provider's own site before you order:

Prices from the providers' own shops, checked in October 2026.

Two things decide what you really pay: where the kit ships from and how often you retest. A yearly DunedinPACE kit is a very different annual bill from a one-off GrimAge reading. None of it is reimbursed by statutory insurance in the DACH region, so you pay the full cost yourself, including any shipping, import VAT and handling fees.

What lands on top of a US price

Shipping origin matters because a kit sent from outside the EU counts as an import. On a 500-euro kit from the US, the extra charges typically come to about 100 to 130 euros in Germany or Austria, so the kit lands closer to 600 to 630 euros. In Switzerland they come to about 50 to 60 euros. The money goes to three places:

• Import VAT from the first euro. Kits shipped from the US owe import VAT on their full value: 19% in Germany and 20% in Austria. Switzerland is outside the EU and charges 8.1% import VAT^[31] which Swiss buyers also owe on parcels from the EU.
• Customs duty is a separate charge. Since 1 July 2026 the EU charges a temporary flat €3 duty per item type in parcels worth up to €150, until 1 July 2028.^{[14, 16]} The seller or importer declares and pays it.^[14] A 300 to 500 euro kit is above that line anyway, so the flat duty does not apply to it.
• Carrier handling. DHL or UPS add around €6 to €20.

Ordering from Germany or Austria avoids most of this, because epiAge and neotes are sold by German companies and no import VAT applies. neotes also analyses samples in a German specialist lab.^[23] Muhdo ships from the UK, so check at checkout whether import VAT is added.^[30]

Before you order

Check first that the provider ships to your country at all, because NOVOS and Elysium do not ship to Germany, Austria or Switzerland.^{[25, 27]} Then look at what happens to what you send. Your sample and its methylation data count as genetic and health data, which EU data protection law treats as specially protected.^[32] So check where the lab sits (EU, UK or US) and whether it stores your sample or uses it for research. Also check whether data goes to partners and how to have it deleted.

For 200 to 500 euros, an epigenetic test buys you a number that is good at sorting large groups of people by age but can be off by several years for one person.^[1] That group-level accuracy comes from the first clock. In its original paper the Horvath clock correlated with calendar age at r = 0.96 (1.0 would be a perfect match), and half of its estimates fell within 3.6 years of true age.^[1]

That r = 0.96 was measured across all ages from newborns to old age, and such a wide age range flatters any correlation. For one adult the typical miss of a few years matters more than the correlation. Mortality- and pace-trained clocks such as GrimAge and DunedinPACE are not meant to match calendar age at all, so the figure does not describe them.

The miss is not random, either. In older adults the Horvath clock tends to come in younger than true age, an error seen across tissues and datasets.^[15] Accuracy also depends on the tissue and on whether a clock was validated in people like you.

What the experts say

The researchers who build and study these clocks are just as careful about what one result can do. Luigi Ferrucci, scientific director at the US National Institute on Aging, told NPR in 2024: "At this point, if you want to do it, it must be based on curiosity." He also cautions that you should be ready for a result you do not like.^[12]

Steve Horvath, who built the first multi-tissue clock, said in 2022: "Epigenetic clocks will never replace clinical markers, [but] the clocks add value to them".^[13] So read a clock next to your blood pressure, glucose and lipids, not instead of them. A concrete medical question belongs with your doctor, because a methylation report is not a diagnosis.

Both cautions rest on limits built into the clocks themselves. A result is a snapshot of one sample, and on the older clocks that same sample can shift by 1 to 2.5 years between lab runs.^[9] Different clocks also often give different ages for the same person. There is also no agreed standard for what a "normal" biological age should be, and how well one result predicts your long-term health is still being worked out.

How to read your result

Treat your epigenetic age as one research number that sits next to your ordinary health markers. A clock age lower than expected for your calendar age, or a pace below 1.0, is linked to lower risk in population studies.^{[4, 6]} These are group trends, not clinical cut-offs, and no study has shown that one person's result, or a change in it, means a longer or healthier life.

That gap between group data and your own life is why the result should not drive decisions. Do not start, stop or change medicines, hormones or supplements because of an epigenetic age, and do not skip screening such as blood pressure, lipid or cancer checks. No clock is validated to guide treatment.^[13] Be wary of reports that end in a supplement or coaching offer. Do not test while you are ill or under unusual stress, since epiAge itself warns this skews results, and follow the kit's sampling rules; epiAge, for example, is sold for adults from 25.^[21]

An epigenetic age test is worth its 200 to 500 euros mainly if you buy it out of curiosity and can live with a noisy number. That is also how the scientific director of the US National Institute on Aging frames it.^[12] Insurance does not help, since none of this is covered by GKV in Germany, the Austrian social insurers or Swiss basic insurance (OKP), and the bill grows with every retest.

A test makes sense when:

• You are making real lifestyle changes and want a rough outside check on them.
• You like data and stay motivated by seeing numbers move.
• You can afford the kit, and a repeat, without straining your budget.
• You understand the limits and will not over-read a single number.
• You are working with a doctor who reads the result next to your clinical markers.

Skip it when:

• You have not yet put the basic lifestyle changes in place.
• The cost would hurt your budget.
• A bad number would worry you for weeks.
• You want a single test to diagnose your overall health.
• You expect it to tell you how long you will live.
• You hope it will tell you whether to start or stop a medicine or supplement.

You do not need a test to improve your health. The lifestyle changes linked to lower biological age are the same whether you test or not: exercise, decent food, sleep and managing stress, none of which waits for a lab report. A test is most useful for people already doing that work who want a rough check on the direction. For everyone else, free tools and a routine blood test are the better first step.

You can estimate where you stand without a 200-euro kit. Free online tools, a few movement tests at home and routine blood work from your doctor cover the basics.

Free tools.

Our Photo Age Test uses AI to estimate how old you look from visible aging cues in a selfie. It is not as precise as an epigenetic clock, but it is free, needs no sign-up and takes seconds. The photo is processed on EU servers and discarded right away.

Our Pace of Aging Test is a free questionnaire that estimates your biological age from habits tied to it, such as sleep, exercise, diet, stress and social connection.

Movement tests you can run at home:

• Your resting heart rate (lower is generally better)
• How many push-ups you can do
• Whether you can sit down on the floor and stand up again without using your hands
• How long you can stand on one leg next to a wall or a sturdy chair, first with eyes open; try eyes closed only if that feels safe^[20]
• How fast you walk a set distance
• How strong your grip is (compare your grip strength by age and sex)

Skip the balance and floor-rising tests if you have balance problems or have fallen recently, or do them with someone nearby.

Blood work through your doctor. Which values make sense for you is a question for your GP; these are the usual candidates:

• Fasting glucose and HbA1c (metabolic health)
• Lipid panel (heart and blood vessel health)
• CRP (inflammation)
• Complete blood count
• Vitamin D only if your doctor sees a reason, since the Endocrine Society advises against routine testing in healthy people^[19]

In Germany, statutory insurance (GKV) usually pays for these tests when your GP (Hausarzt) orders them for a medical reason. Pure wellness panels are self-pay, and the Austrian and Swiss systems work similarly. Unlike a clock age, these values give you and your doctor something to act on, so a talk with your GP about which checks suit you is the better place to start; an epigenetic test can come later if you are curious and the budget allows.