How we work
The sources we use, how every claim is checked, and what the labels on our reviews mean.
What we do, and what we do not do
We report what published sources say, and we link to them. We do not give medical advice, recommend treatments or doses, or tell you what to do. We do not judge or grade the strength of evidence. We describe the type of evidence and quote official bodies.
The sources we use
We use these sources, in this order of preference:
- Official bodies. Medicines regulators such as the FDA, the EMA and the MHRA, and in the United Arab Emirates the Ministry of Health and Prevention, the Dubai Health Authority and the Department of Health Abu Dhabi where relevant. Also the World Health Organization, national guideline bodies such as NICE, and specialist medical societies.
- Systematic reviews and meta-analyses published in indexed journals, including Cochrane reviews.
- Randomised controlled trials in people, published in indexed journals.
- Other studies in people, such as cohort, case-control and case series studies. We say which kind each one is.
- Animal and laboratory studies. These are always labelled as such, and are never presented as evidence of an effect in people.
The sources we do not use
We do not use these as the source for a factual claim: company or clinic websites, supplement sellers, press releases, news articles, blogs, social media, podcasts, answers from AI tools, and journals that are not indexed in PubMed or an equivalent database. A preprint, which is a study that has not yet been checked by other scientists, may be mentioned only with the words "not yet peer reviewed".
No source, no claim
Every factual statement on Longevello has a numbered source with a working link: a PubMed record, a DOI, or the official page. If we cannot find a source for a statement, the statement is removed.
- The source must say it. Each claim is checked against the text of the source itself, not against memory or someone else's summary.
- Numbers are copied exactly, with their units, the people who were studied and how long the study lasted.
- No stretching. Animal results are not described as applying to people. One small study is not described as "studies show". A link between two things is not described as one causing the other.
- Both directions. If trials disagree, or a harm has been reported, we say so.
- Regulatory status has a country and a date, for example "not approved as a medicine by the FDA as of November 2026".
Every source is opened and checked
Each source is retrieved from PubMed or from the official website that published it. The PubMed record or DOI must lead to the paper we name. A reference that does not exist, or that points to a different paper, stops the review until it is fixed.
The passage that supports each claim is read in the abstract or full text of the source, and we keep a record of the claim, the source, the supporting passage, the type of study, who was studied and the date it was checked. Full text is not always available. Where only the abstract could be read, our record says so. We also check each paper for a retraction or correction notice.
A second, independent check
Before a review is published, it goes through a second check that is made separately from the first. This check starts from the finished review and its list of sources only, and goes back to each source to test every claim. Anything that is not supported, is overstated or is missing context is reported. Disagreements are settled by going back to the source. If they cannot be settled, the claim is removed.
What the type of evidence tags mean
Each review carries one tag describing the type of evidence behind it. The tag is set by a fixed rule, not by opinion. It tells you what kind of studies exist. It is not a score and does not say how good or reliable the studies are.
What "Evidence at a glance" means
Each review has a short list of four questions, each answered yes or no. Every answer has its own source.
- Have human trials been published?
- Have animal or laboratory studies been published?
- Is it approved by a medicines regulator? If yes, we name which regulator and for what use.
- Is it recommended in an official guideline? If yes, we name which guideline.
A "yes" is not an endorsement, and a "no" is not a warning. The answers describe what exists.
We do not rate evidence or give advice
We do not give evidence a score, a grade or a star rating, and we do not tell readers what to do. Our reviews contain no advice, no dosing instructions, no promises and no superlatives. Each review includes a section on the limits of the evidence and on safety, and the line: "It does not give medical advice and does not replace your own doctor."
How often pages are reviewed
Every review shows the date it was first published, the date it was last checked and the date it was last updated. The updated date changes only when the content changes.
- We currently check every source link each month.
- We currently review every article against its sources about every six months.
- A review is looked at sooner if a source changes or is withdrawn.
- A page reflects what its sources said on the date it was last checked. Information published after that date may not be included.
How corrections work
Every review has a "Report a correction" link. When an error is reported or found, we check it against the source. If the page is wrong, we correct it promptly and add a note at the foot of the page with the date of the change. Read more on our corrections page.
- Curated from published research
- Every claim linked to its source
- Nothing sold, no adverts