Guide — Grading the Evidence
Naval architects test hulls in a tank because it is the only way to ask the question. Nobody then sells the tank result as sea trials. The supplement aisle does exactly that, and this guide shows you where.
Tank Trials
A compound is added to prostate cancer cells in a dish. The cells die. This is a real observation, it gets published, and within a year it is on a bottle. What has actually been shown is that the compound kills those cells at that concentration in that dish — which is also true of bleach, of alcohol, and of a great many things nobody proposes swallowing.
The first gap is getting into the body at all. A concentration achievable in a dish may be impossible to reach in living tissue at any tolerable dose, and much of what is swallowed is broken down before it arrives anywhere useful. The second gap is the animal model: helpful, necessary, and a poor predictor. A long list of compounds that worked in mice have gone on to do nothing in people.
Neither gap is a flaw in the research. Cell work and animal work are how the questions get asked; nobody could run human trials on every promising molecule. The failure is entirely in the translation — a finding produced at one level being sold at another. This guide covers the three levels, what each can genuinely establish, and how to tell from a page which one it is actually describing.
Each answers a different question. Reading one as another is where the whole category goes wrong.
The guide covers what these studies are for, why the concentrations used are often far above anything achievable in a person, and why the phrase “shown to inhibit cancer cells” is technically accurate and almost always misleading in a marketing context.
The question to ask: at what concentration, and could a human ever reach it? A great many substances are cytotoxic in a dish. That property alone has never made anything a treatment.
This step answers the questions a dish cannot: does it get absorbed, does it reach the tissue, is it tolerated. The guide explains why translation from rodent to human fails so often, and why a mouse result is a reason to run a trial rather than a reason to buy anything.
The question to ask: was this ever tested in people? If a product's evidence page cites only animal work, that is not weak human evidence — it is an absence of human evidence, which is a different thing entirely.
Even here the levels continue: a handful of volunteers over six weeks and a randomised trial of thousands over years are not the same evidence. The guide covers how to read the size, the duration, and the design without needing statistical training.
The question to ask: how many people, for how long, and compared against what? These three facts are usually in the abstract, and they settle more than any conclusion sentence will.
Contents
Digital Guide · PDF Format
Three levels of evidence and the questions that separate them — a method that outlasts whichever compound is in fashion.
Tax included · Instant download · One-time payment
Read it at your own pace. If what arrives does not match what this page sets out, the full amount is returned within 14 days of purchase, by email, with no reason requested.
Notice: This is not a medical portal and provides no diagnostic or treatment service. What is sold here is a method for grading evidence. It names no brand and endorses no product, and it is not a verdict on any substance in either direction — thin evidence means the question is open, not that the answer is no. Nothing here is a reason to start or stop anything on your own. If you have symptoms, pain, or any concern, book an appointment.
FDA Statement: These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease. Intended for adults 18 years of age or older.
This site is not operated by or affiliated with Google. Google is a registered trademark of Alphabet Inc.