Rolling thread: post a claim you have seen, somebody will check it

Good one. Advice that is right in general and wrong in application is harder to argue with than advice that is just wrong.

From a clinic's page:

"NAD+ therapy restores cellular energy at the source."

"At the source" is the tell, and it is doing something quite specific.

NAD+ is genuinely central to metabolism — that part is textbook. The unevidenced steps are, in order: that an infusion meaningfully raises it inside the cells you care about, that raising it there produces the effects being sold, and that "cellular energy" corresponds to anything a person experiences.

Three inferential steps, all unstated, in a sentence that sounds like biochemistry. evidence: early is where it actually sits.

Verdict: real biology, three unevidenced leaps, phrased to sound like one fact.

The construction is worth naming since it recurs: a true mechanistic statement, followed by an outcome claim, joined by nothing at all. The reader supplies the join and then feels they worked it out themselves, which makes it far stickier than being told.

That should be in the first post of this thread. Half of what gets submitted here is that structure.

"It's the same peptide the clinics use, just without the markup."

The most consequential one posted here so far, because it is the one that actually moves money and needles.

"The same peptide" may well be true at the level of the molecule. What it omits is everything the markup was paying for: manufacture under inspection, batch records, a validated process, an identity and purity specification somebody is accountable for, and a chain of custody.

An approval is not a certificate attached to a molecule. It is a judgement about a molecule and the system that produced it, and only one of those two things is being offered at the lower price.

Verdict: possibly true as stated, and the omission is the entire product.

And the part a buyer cannot check is exactly the part that was removed. You can assess a molecule's identity with a lab. You cannot assess a quality system with a PDF.

Two hundred posts away from needing a second thread, but a running index of the patterns so far, since they repeat more than the compounds do:

1. True number, false sentence. The count is real, the species or the endpoint is not.
2. Borrowed evidence. Data for a related molecule, cited for the one being sold.
3. Mechanism then outcome, joined by nothing. The reader supplies the inference.
4. "Up to", "basically", "clinically". One word carrying the whole claim.
5. Same molecule, different system. True and beside the point.

Keep them coming.