How we read research
Most of what is written online about longevity treatments is written by people selling them. This page explains how we handle the research, so you can check whether we are doing it properly.
We read the paper, not the press release
If we cite a study, someone here has opened it and read the methods. We do not summarise a study from a news article, a clinic's website, another blog, or the abstract alone when the full text is available.
Every citation carries the journal, the year and the PubMed ID so you can find it in seconds. If a claim on this site has no citation, it is a mistake, and we want to know about it.
What we look at in a study
Was there a control group. This is the single most important question and the one most often skipped. A study where everyone got the treatment and everyone improved tells you very little, because people improve for many reasons over three months.
Was it blinded, and was there a sham. For anything where expectation could plausibly do the work, the only test that settles it is comparing the real thing against a convincing fake. A number of popular treatments look impressive until a sham arm turns up.
How many people. Twenty five people is a pilot study. It generates a hypothesis. It does not establish that something works.
How big is the variation. A result written as up 37 percent, plus or minus 52 percent, is telling you the spread between individuals is larger than the average effect, which means some participants went the other way. We report that spread when it matters, because leaving it out turns an unstable finding into a headline.
Who paid, and who wrote it. A study funded or authored by people who sell the treatment is not automatically wrong. It does mean the result needs a control group to protect it from the people who benefit from it. We say plainly when the authors have a commercial interest.
Was it in humans. A finding in mice or in a dish is a reason to run a human trial, not a reason to book a session. If a result comes from animals, we say so.
Does it match what is actually sold. This one catches more claims than any other. A study run at a medical dose, pressure or duration does not tell you about a gentler version sold in a wellness package. Where the two differ, we say so.
What we do with the answer
We report what the research found, including when it is unhelpful. If the best evidence for a treatment is null, that goes on the page. If a positive result comes from a small uncontrolled study run by the manufacturer, we print the result and the caveat together, in the same breath, rather than burying one at the bottom.
We do not extrapolate to you. Research describes what happened on average to a group of people who are not you, under conditions that may not match yours.
We publish negative findings. A page that only reports good news is marketing with citations attached.
Where our sources come from
Primarily PubMed and the peer-reviewed literature, plus Cochrane reviews where one exists, since a systematic review of every trial beats any single trial. We use Europe PMC and preprint servers where relevant, and we label a preprint as a preprint, because it has not been peer reviewed.
For regulatory questions we go to the source: the HPRA in Ireland, the EMA, and the relevant professional bodies.
The limits of this
This is done carefully but it is done by a researcher, not a clinician, and it is not peer review. We can tell you what a study found and how strong it looks. We cannot tell you what you should do, and we do not try to.
If you spot something we have read wrong, email hello@gerovia.com with the page and the paper. We would rather be corrected than be confidently wrong in public.