Almost nothing in this field measures lifespan directly, for the obvious reason that it would take a lifetime. So researchers measure a marker instead: something faster to observe that is believed to track the outcome.
That is a necessary and sensible substitution. It also introduces the single largest gap between what was studied and what gets claimed.
What a surrogate marker is
Any measurable stand-in for an outcome nobody can wait for: a blood value, an imaging result, a calculated biological age, an inflammatory marker, a measure of cellular activity.
The claim being made whenever one is used, usually silently, is that moving the marker moves the outcome. That claim is separate from the study and frequently untested.
The gap in one sentence
Something can reliably move a marker and have no effect on how long or how well anyone lives, and this has happened repeatedly in medicine with drugs far better studied than any supplement.
Why the link breaks
A marker is usually a signal of an underlying process rather than the process itself. Changing the signal without changing what produces it accomplishes nothing, in the same way that adjusting a thermometer does not alter the temperature.
Markers can also be downstream of several processes at once, so a change may reflect something entirely unrelated to the mechanism being claimed.
Biological age tests specifically
These deserve their own caution, because they are sold directly to consumers and are unusually persuasive. Most are trained to predict chronological age or mortality from a set of measurements. That the estimate responds to an intervention shows the measurements changed. It does not show that anything about ageing changed, and different tests routinely disagree with each other on the same sample.
Interesting, worth following, and not a health outcome.
What raises confidence in a marker
Established markers exist and some are excellent. What separates them is that the link to real outcomes has itself been tested, that interventions moving the marker have gone on to change the outcome in trials, and that the relationship holds across different populations and different ways of moving it.
How Longevity Claim Verification handles it
A claim resting on a marker is reported as a marker result, with the outcome it is standing in for named and the strength of that link stated separately. Where the link has never been established, the claim is unsupported for the outcome even when the marker result is solid.