Biological age
Five ways to measure biological age, five different answers
Methylation clocks, telomere length, blood biomarker algorithms, glycan analysis, and composite scores measure different things. Disagreement between them is expected, not anomalous.
A 38-year-old takes three tests. Methylation says 44. A blood biomarker algorithm says 33. The telomere result says average. None of these is wrong, because none of them is measuring the same thing.
The short version
- At least five fundamentally different approaches are sold under the heading of biological age.
- They measure different biology, so agreement between them was never a reasonable expectation.
- Blood biomarker panels are the least fashionable and the most actionable, because their inputs respond to intervention in months.
- Telomere length has the weakest link to actionable individual decisions despite being the most intuitive to explain.
What each method looks at
DNA methylation clocks read chemical marks at specific genomic sites and feed them into a model trained to predict age or mortality. Horvath, Hannum, PhenoAge, GrimAge and DunedinPACE are the recurring names, and they were trained on different outcomes, which is one reason they disagree.Telomere length measures the protective caps on chromosomes. Intuitive, widely marketed, and weakly predictive at the individual level.Blood biomarker algorithms combine routine clinical values, inflammation markers, metabolic values, kidney and liver function, into a composite. Least exotic, most responsive to change.Glycan analysis examines sugar structures attached to circulating proteins, which shift with inflammatory state. Newer and less validated.Composite scores blend several of the above, which improves the marketing and complicates the interpretation.Why disagreement is the expected result
A methylation clock trained to predict mortality and a blood panel algorithm trained to predict physiological function are answering different questions. When they diverge, that divergence is information about which aspects of your biology look better or worse, not evidence that one test malfunctioned.
The trouble is that all five report their answer in years, which invites direct comparison and creates the impression that four of them must be broken.
Which one to use, if any
If the goal is to track whether something you are doing is working, blood biomarker panels are the pragmatic choice. Their inputs are standard clinical assays with known error characteristics, they move within three to six months in response to genuine changes, and every component is individually meaningful to a clinician.
If the goal is research participation or curiosity about long-horizon risk, a third-generation methylation clock is the more scientifically interesting instrument, with the reliability caveats that apply to all of them.
If the goal is a number to post, any of them will do, and that is the use case most of this market is actually serving.
Common questions
Which is most validated?
Third-generation epigenetic clocks have the strongest research literature behind them for predicting outcomes at population scale. That is a different question from which is most useful to you individually.
Can I compare results across providers?
Generally no, even within the same method. Different platforms, models, and reference populations make cross-provider comparison unreliable.
Sources
- Biological age test: five methods compared — https://mitohealth.com/guide/biological-age-testing
- Biological age tests: what epigenetic clocks actually measure — https://ubiehealth.com/doctors-note/biological-age-testing-epigenetic-clock-guidance4262q2