This page explains how articles in the Limitless Labs research library are sourced, checked, and maintained. We link the evidence behind research claims, distinguish study types, and correct errors when identified.
Primary literature only
Articles in this library are built from primary sources: peer-reviewed papers indexed in PubMed, and trial registrations and results posted to ClinicalTrials.gov. Where a finding has not yet reached a journal — for example, topline results announced by a trial sponsor — we may summarize the announcement, but we label it as company-reported and unreviewed, and we link the original release rather than paraphrasing coverage of it.
What we do not treat as sources: blog posts, forum threads, vendor marketing pages, supplement-industry press, or other summaries of summaries. If a claim appears in our articles, you should be able to click through to the study behind it. Secondary sources such as review articles are cited where they are themselves peer-reviewed and add synthesis the primary papers do not.
Machine-verified citations
References carrying a PubMed identifier are checked against NCBI records. Automated checks cover identifier existence, publication title, journal, publication year, and agreement with indexed DOI metadata where available, as well as coverage in the study database. A matching record does not validate the interpretation of a study or its author list. Registry entries, sponsor announcements, and sources without PubMed identifiers require separate review; they are not covered by the PubMed check.
This exists because fabricated and mangled citations are a real failure mode of AI-assisted writing, and this library is drafted with AI assistance under human editorial review. Citation checks help identify errors, but do not guarantee that an article is error-free. Assessing methods, populations, outcomes, and limitations requires reading the source; automated metadata checks cannot replace that work.
Original writing
We do not republish third-party text. Articles are written from the studies themselves — methods, results, and limitations sections — not assembled from abstracts or lifted from other summaries. Direct quotations, where used, are short, marked as quotations, and attributed. Study data (effect sizes, sample sizes, endpoints) are reported as published, with the citation number pointing at the paper they came from.
How articles are updated
The literature moves, and pages here are living documents rather than archived posts. Each article carries an "updated" date in its header. We revise an article when a relevant trial reads out or publishes, when topline results we previously summarized appear in peer-reviewed form (the peer-reviewed version then replaces the press release as the cited source), when a paper we cite is corrected or retracted, or when a regulatory status changes. Updates go through the same citation verification as new articles. We do not silently rewrite conclusions: where new data changes the picture an article painted, the article says what changed.
Research framing
Everything in this library summarizes published scientific literature for an audience of researchers. Compounds sold by Limitless Labs are supplied for laboratory research use only. Library articles therefore do not provide medical advice, do not describe or endorse human use of research materials, and do not make claims of medical effectiveness for anything sold on this site. Trial results are reported as findings about study populations under clinical supervision — statements about what researchers observed, not promises about what a product does. Where an article covers an approved pharmaceutical (as with semaglutide), the research-grade material we sell is not that pharmaceutical, and the article says so plainly.
Corrections
If you find an error — a misread result, a citation pointing at the wrong paper, a claim that overreaches its source — tell us via the contact page. Substantive corrections are made in the article, the "updated" date is refreshed, and material errors (anything that changes a stated finding, not typos) are noted at the bottom of the corrected article. We would rather be corrected by a reader than wrong in public.