How we check claims

Every factual claim in a published article is checked against the source it cites. This page describes the steps, how we handle studies that disagree, and the claims we deliberately avoid.

The steps

1. Find the best available evidence. We look for systematic reviews, meta-analyses, randomised trials and guidelines first, in medical databases and publisher sites.
2. Read the source, not a summary of it. At a minimum the abstract and results; numbers are taken from the study itself.
3. Record what kind of study it is. Design, size, who was studied and how long, so the article can describe it fairly.
4. Match each claim to its source. If the source does not say it, the claim is removed or reworded. Practical tips that are not from a study, such as a starting number of repetitions, are presented as suggestions, not findings.
5. Label the evidence. Each article gets an evidence label.
6. Check links and dates before publishing, and when updating.

When studies disagree

We say so. For example, in adults forward head posture is associated with neck pain in one meta-analysis, while a later study found no difference. The article reports both and explains what the disagreement means. We give more weight to larger, better designed studies and to reviews, and we say when a question is open.

What we do not claim

• That a posture causes damage when the evidence shows only an association.
• That a product, supplement or gadget works when trials are missing.
• That an app score is a diagnosis.
• Precise numbers that a study did not report.

Mistakes

If you find a claim that does not match its source, please tell us at support@bodysignals.app. Corrections are listed on corrections and updates.