The Longevity Desk

The record

Corrections

What this site published, got wrong, and fixed. A publication that has never corrected anything is not error-free; it is not looking.

6 corrections on record

What counts

An entry appears here when something was published, a reader could have believed it, and it was wrong. Errors caught in drafting never reached anyone and are not listed; neither are typos or rewording that leaves the facts unchanged, for the same reason the revision dates on entries only move when a fact does. Each correction names the page, what it said, what is true, and — where it is instructive — how the error happened.

  1. How we read research
    What it said
    That no sourcing information appears on this site, and, in a hand-written conflicts paragraph, that this publication takes no money from anyone selling anything.
    What is true
    Both had quietly stopped being true. The vendor sourcing table publishes exactly that kind of information, and it carries affiliate links. The page now states what is actually withheld — dosing guidance and protocols — and its conflicts section is derived from the same data the vendor table is built from, so it cannot assert something the site has stopped doing.
    How it happened
    A disclosure written as prose has to be remembered on the day it stops being true, and the day it stops being true is the day it turns from stale into false.
  2. What it said
    That 26 compounds have no published human dose.
    What is true
    That overstates it. The empty cell covers three different situations: nobody dosed at all, only animals dosed, or people dosed but not comparably. LL-37 has four human trials behind it, on wounds and into tumours rather than under the skin, and GW-0742 has a real human figure from a doping detection study that was not measuring whether it worked. The honest claim, and the one the page makes now, is that there is no published dose the circulating convention can be measured against.
    How it happened
    A shortening of the longer, correct phrase, made to fit a summary line.
  3. What it said
    That batch #HXAC8 was "listed at 10.32% analytical purity and marked VERIFIED", that no laboratory was named, and that the certificate library does not survive being read.
    What is true
    All three were wrong. The certificate reports 99.51% purity against a ≥95% specification, a PASS. The 10.32 is the mass in milligrams against a 10 mg label, printed by the vendor under a column headed Analytical Purity — their mislabelling, but one this page repeated. ILS Laboratories is named, with a street address, ISO/IEC 17025 accreditation and a lab director's signature.
    How it happened
    The vendor's index was read and the certificates underneath it were not opened. Two other rows carried the same error and were corrected the same day.
  4. What it said
    That no laboratory is named anywhere.
    What is true
    True of the library index, false of the certificates beneath it, which name ILS Laboratories with a street address, a lab director's signature and ISO/IEC 17025 accreditation. Batch BPC10-05202026 reports 99.84% purity against a 95% specification, identity confirmed, sterility clear, endotoxins at 0.071 EU/mL, heavy metals below USP <233> limits, and prints the HPLC chromatogram itself, which no other vendor on the table does.
    How it happened
    The same failure as the Optimum Formula row, found the same way.
  5. What it said
    Nothing factually, but the row's outbound link was presented under a disclosure stating that such links earn a commission.
    What is true
    That link pointed at a domain which redirected without carrying the referral parameter, so it earned nothing while returning a healthy 200. It now points at the final destination directly. Logged here because the disclosure described what the link did, and for that period the description was wrong.
    How it happened
    Found by auditing every vendor link for referral survival rather than by anything visibly breaking, which is the only way this class of fault surfaces.
  6. What it said
    That a discount code could be compared against a competing public offer on the site, first as matching it and then as worth less on a first order and more afterwards.
    What is true
    There was no competing offer to compare against. The 20% figure is an affiliate commission rate, dropping to 10% on repeat orders, not a customer discount. The customer-facing rate is 15% either way. The comparison was removed rather than restated.
    How it happened
    A banner advertising an affiliate's commission was read as a public promotion. The row shipped two versions of the same misreading.

Found something wrong that is not on this list? The about page says how to reach us, and a verified error goes here regardless of how small it is or who found it.