Almost every longevity claim you encounter arrives detached from its evidence. Someone credible says a thing, it gets repeated, and by the third repetition the hedges are gone and the study it came from is a hyperlink nobody clicked. Longevity Claim Verification is the work of following that link and reporting honestly what is at the other end.
What a claim is, and what verifying it means
A claim is a specific, checkable statement about what something does: this supplement extends lifespan, this protocol reverses a marker, this practice reduces a risk. Vague encouragement is not a claim and is not verifiable. "Sleep matters" cannot be checked. "Eight hours of sleep reduces all-cause mortality by thirty percent" can.
Verifying it means establishing four things: that the cited evidence exists, that it studied what the claim says it studied, that it found what the claim says it found, and that the finding is strong enough and repeated enough to carry the weight being put on it. A claim can pass the first three and fail the fourth, and that case is the most common by a wide margin.
The core problem
The citation is usually real. It usually does not say what the claim says it says. Most of the distance between health claims and reality is created in that gap, not by outright fabrication.
The six checks
1. Does the evidence exist? Sometimes there is no study, only a confident assertion that has been repeated until it acquired the texture of evidence. This is the rarest failure and the easiest to report.
2. Was it in humans? An enormous share of longevity claims rest on work in mice, worms, yeast or cells. That work is legitimate and it is not a claim about people. Lifespan extension in a nematode is a real finding about nematodes.
3. Who was studied, and are you them? A result in 400 post-menopausal women with existing deficiency does not transfer to a healthy 34-year-old man. Population is the check most often skipped, because the result is real and the transfer is silent.
4. How big was the effect, actually? Relative numbers dominate headlines because they are larger. A risk falling from two in a thousand to one in a thousand is a fifty percent reduction and a one-in-a-thousand improvement. Both descriptions are true. Only one of them tells you whether to change anything.
5. Has it been repeated? A single study is a reason to be interested, not a reason to act. Fields with a replication problem produce striking single results routinely, and nutrition and longevity research are two of them.
6. Who paid for it, and who benefits? Funding does not make a study wrong and it is a reason to look harder. A supplement trial funded by the company selling the supplement is not disqualified, it is downgraded in confidence, and the reader is told.
Three outcomes, not two
Most claim checking in public produces a binary: true or debunked. Real evidence rarely cooperates, so verification here has three states.
Supported. Human evidence, in a relevant population, with an effect size that matters, repeated. Rarer than anyone expects.
Unsupported. The evidence does not carry the claim. This usually means the study was in animals, in a different population, much smaller in effect than reported, or unreplicated. It is not the same as saying the claim is false.
Unknown. Nobody has done the work properly yet. This is an honest and frequent outcome for longevity questions, where good long-horizon human trials are expensive and slow. Rendering unknown as unsupported would be tidier and it would be a lie, so it is kept as its own state.
Why we do not say debunked
Debunked implies the claim was shown to be false. Almost always, what has actually happened is weaker and more interesting: the claim outran its evidence. Saying debunked would be the same category of error the original claim made, just pointing the other way, and it would put us in the position of overstating a negative while criticising someone for overstating a positive.
It also encourages the reader to sort people into trustworthy and untrustworthy, which is the habit that created the problem. A careful researcher can make one overreaching claim. Someone who is usually wrong can be right about a specific thing. The claim gets checked, not the person.
What this is not
It is not medical advice. Nothing here tells anyone what to take, stop, or do about their own health. Verification reports what the evidence supports. Decisions about a specific body belong to the person in it and their doctor.
It is not a verdict on a person. A claim is checked in isolation from who made it, including when who made it makes the answer obvious.
It is not contrarianism. The point is not that popular claims are wrong. Some are well supported, and those get reported as supported with the same detail. A checker whose answer is always negative has stopped checking and started performing.
The objection worth taking seriously
The fair criticism: who checks the checker? A site that grades other people's evidence is asking for trust it has not itself demonstrated, and "we read the study" is exactly what everyone in this space says.
The only real answer is that the working is shown. Every verification names the specific evidence, states which of the six checks it passes and fails, and links the source, so a reader who disagrees can go and look. That is a weaker claim than authority and it is the honest one: the value is in doing the legwork visibly, not in being believed.
The second fair criticism is that demanding replicated human evidence for everything is a bar that would have rejected many things that later turned out to be true, and that a strict standard has its own cost. That is correct. It is why unknown exists as a separate state rather than being collapsed into unsupported, and why an unsupported reading says what evidence would change it.