How we write

We start from the original study

We read the paper itself, not a press release or another article about it. Study details are checked on PubMed or the journal page, including whether a paper has been corrected or retracted.

We report four things about every study

1. Who took part: age, health status and how many people.
2. How long the study ran.
3. What the diet or supplement was compared with.
4. What was measured, such as weight, blood sugar, cholesterol or events like heart attacks.

A short trial in a small group answers a different question from a long trial in thousands of people, so we say which kind of evidence a claim rests on.

We keep numbers exact

Doses, study sizes and results are copied from the source with their units. When two sources disagree, we say so and name both. When a number cannot be confirmed, we leave it out.

We update when the evidence changes

Guides are reviewed when a major new trial or official recommendation is published. A change to the facts is recorded as a change to the page date.

What we do not do

We do not give individual medical advice, and we do not describe a supplement as a treatment for a disease.