Biological age tests are the most emotionally compelling product in longevity and among the hardest to interpret responsibly. Both of those facts deserve more attention than they usually get.
What an epigenetic clock measures
DNA methylation patterns — small chemical marks on the genome that change predictably with age — can be fed into an algorithm trained to predict chronological age. The best clocks do this remarkably well, within a few years.
The interesting part is the residual. When the clock predicts you're older than your birth certificate says, that gap correlates with mortality and disease risk in large cohorts. That correlation is real and has been replicated.
Three limitations worth knowing
Test-retest variability
Some commercial clocks show variability of several years between samples drawn from the same person on the same day. If the measurement noise approaches the size of the effect you're tracking, the number is not telling you much.
Correlation is not a target
A marker that predicts an outcome does not automatically improve that outcome when you change it. Moving the number is only valuable if the underlying biology moved with it, and that link is not established for most interventions.
Clocks disagree
Different clocks trained on different populations and endpoints can give the same sample meaningfully different answers. There is no single authoritative measure of biological age.
A more useful framing
Functional measures — VO2 max, grip strength, gait speed, fasting insulin, ApoB — are cheaper, more reproducible and more directly actionable. They also respond to interventions in ways you can feel. Start there, and treat the clock as a curiosity you track on the side.