We checked 103 of the most-repeated claims in longevity, supplements, devices, hormones and tests against the published evidence. Just 8 hold up cleanly. Here is what the evidence actually says, in numbers.
The largest single group is Mixed: claims with a kernel of truth that gets stretched well past what the data shows. Add Unproven (plausible but untested in humans) and most of the longevity conversation turns out to be neither settled science nor outright nonsense.
That is the honest middle the marketing skips. We keep Unproven apart from Unsupported on purpose: not shown is not the same as shown to be false.
The pattern nobody sells
50% of the evidence sits at low or very-low certainty (grade C or D). The science behind most longevity claims is thin, short-term, or built on surrogate markers rather than outcomes that matter.
And where the evidence is strongest, grade A, 6 of 8 of those checks contradicts the popular claim. When the science is most certain, it is most likely to say the hype is wrong.
Share of each category we rated Misleading or Unsupported (categories with at least three checks).
3 of 3 checks
6 of 8 checks
6 of 13 checks
2 of 5 checks
3 of 8 checks
Grade-A checks, where the science is most certain
Read the study, not the thread.
These figures are computed live from Caveat’s 103 published and in-review checks, current as of June 2026. Every verdict traces to primary sources and a graded read of the evidence. These are population averages, though: what holds for most people may not hold for any one person. You are not the average.