Skip to main content

Epigenetic vs glycomic clocks

DNA marks or antibody sugars? What each kind of clock reads, what studies that measured both found, and how to choose.

4 min read
Contents

What does each method read?

Both kinds of test read a blood sample, but they read very different things in it.

Epigenetic clocks
Reads
Methylation marks on the DNA of blood cells
Built to estimate
Calendar age (the first clocks); health risk scores, time until death in a study, or how fast the body changes over years (later clocks)
Result
An age in years, or a pace of ageing
Also moves with
The mix of immune cells in the sample
Glycomic clocks
Reads
Sugar chains (glycans) on the most common antibody in blood (IgG)
Built to estimate
Calendar age (the published formula)
Result
An age in years
Also moves with
Sex hormones, for example around menopause
Schematic drawings, not to scale.

For methylation, a formula worked out from data on large groups of people turns the pattern of tags into an age (how epigenetic clocks work). Later methylation clocks were built to estimate different things, so two of them can answer different questions (why aging clocks disagree). The best-known published glycan age formula combines three sugar patterns. Separately, laboratories independent of the glycan test's developers have also found that IgG glycans shift with age.

What happened when both were measured in the same people?

The key idea here is the age gap: how far a clock puts a person above or below their calendar age.

Illustration of an age gap: the distance between a person's calendar age and the age a clock gives them. No real values.

The largest comparison we found

Age gaps barely lined up

ORCADES, a research study of people in Orkney, Scotland

Researchers compared 15 clocks built from different kinds of measurements, including glycan and methylation clocks: 11 they built themselves and four published ones. Methylation data were available for about 1,000 participants.

The methylation clocks tended to agree with each other. But people with a bigger glycan age gap did not tend to have a bigger gap on the two best-known published methylation clocks, Horvath's and Hannum's. The statistical link between them (the correlation) was close to zero, as we read it from an archived copy of the paper's figure. With methylation clocks the team rebuilt from its own data, the link was still weak.

Before you read into it

  • The study only used early clocks built to guess calendar age, so it cannot tell us how newer clocks, or any test you can buy today, would compare.
  • The authors caution that a large part of these gaps may simply be noise.
  • The weak link on its own does not prove that the two tests track different kinds of ageing.

New Zealand · 964 people

Disagreement is not unique to glycans

A study compared eleven ways of measuring biological ageing, none of them glycan-based, and also found low agreement between them.

Plasma-exchange trial · 42 people who completed it

Set up for a different question

A small randomised trial of plasma exchange (a procedure that removes and replaces the liquid part of the blood) measured methylation clocks from a commercial lab and antibody sugar patterns from a commercial glycan kit in the 42 people who completed it. It was set up to study plasma exchange, not to compare the tests. Its main report gives no glycan age, so it cannot show how the two age numbers compare, and it does not validate either home test.

The missing study

We found no study that ran a consumer methylation test and a consumer glycan test on the same people and compared their results.

Does a study result apply to the test you buy?

Almost all of this evidence is about research versions of each method, run in study laboratories. A test you buy adds its own steps.

In a study

  1. Sample collected for the study
  2. Study laboratory
  3. Formula as tested
  4. Study finding

A test you buy

  1. Its own sample collection
  2. Its own lab steps
  3. Its own software
  4. Your result
When a study shows that a formula works, that applies to the formula as tested in the study; it does not automatically apply to a product's result.

For example, the developers of GlycanAge say its commercial result is calculated with its own unpublished formula. Our GlycanAge evidence summary and the others on Compare explain what is known about each product's own result, and that is what our ratings look at (methodology).

Can you combine the two numbers?

Not in any established way.

Both tests give you an age in years, so it is tempting to compare them directly or to average them. But the research we reviewed gives no established way to turn one result into the other or to average them into a single age. A gap between your results does not, by itself, mean either test is wrong.

If you test again

  • Methylation scores, for example, can shift between blood samples taken only hours or days apart.
  • None of the sources we reviewed says how big a change in one person’s result would be meaningful, for either kind of test.
  • A lower number is a change in the test's reading, not proof that you are healthier. See how often to test.

How to choose

Start with the question you want answered.

  1. Do you want a result built around health outcomes?
    Some methylation clocks were built using health risk or how fast the body changes over years. In large research studies they have been linked to the risk of dying, including by researchers who did not create them. These links describe groups of people, not a forecast for you, and they apply to the published formulas, so check how a specific test reports them.
  2. Are you curious about a different part of your biology?
    A glycan test reads sugars on your antibodies. Research links these to age, sex hormones and markers of inflammation. So far, the published formula has mainly been checked against calendar age, and its links to future health are less well studied.
  3. Are your hormones changing, for example around menopause?
    Glycan results are known to shift with hormone levels. A change during that time may simply reflect the hormones. It is not a diagnosis.
  4. Do you want one test to confirm the other?
    The research we reviewed does not show that either kind of test can check the other's result.
  5. Do you want to track change over time?
    Using the same test each time may make results easier to compare, but companies can change their methods and formulas. Read small changes with caution.

Which tests use which method?

Compare shows what is known about each one, test by test.

Compare testsFind your test

Sources

Show all 17 sources, with our notes on authors' ties to test makers where relevant

This article was prepared with AI assistance from a research ledger that a separate AI reviewer checked against the sources, and its text was reviewed against the cited sources by another separate AI session. No external expert validation. Sources checked 20 and 21 October 2026. See our methodology or report a correction.