Method
How a page earns its place.
Short version: no number without a named study, no cause from an association, and the limit printed next to the finding.
The editorial rule
Every numerical headline maps to a named study or systematic review and carries its design, population and comparator. Observational association stays association. Biomarkers are not promoted into disease outcomes. Animal and mechanistic findings are labelled as such.
How sources are checked
- Each DOI is looked up in Crossref and the doi.org handle service. Title, authors, journal and year must match what the page says.
- Each paper is matched to its PubMed record. We read the abstract, and the full text where it is open access, and confirm every figure quoted on the site against it.
- Where a figure or label did not match, we corrected the page and logged it below.
All eleven sources on the site were checked on 17 November 2026. The sources page lists what was confirmed for each.
Corrections log
- 17 November 2026. Circadian light: the cited study was described as a “controlled study of 104 participants”. It is a retrospective within-participant analysis of two inpatient studies (104 volunteers in the room-light comparison, 116 in total). Labels corrected.
- 17 November 2026. Four source titles now match the published titles (Feinstein 2018, Alkurd 2024, Søberg 2021, Li 2008). Thermal exposure now says the sauna result comes from one Finnish cohort of men. The forest study now states its 12 participants were healthy men aged 35–56.
- Earlier edition (v1.3 to v1.4). Removed an unsupported “+22% interoceptive accuracy” figure, a causal “40% cardiovascular mortality reduction” wording, an unsupported “−23% cognitive performance” headline, and a claim that morning light is “the highest-leverage intervention”.
Review status
Summaries are written and checked against the cited papers by the Purecohere team. They are not medical advice and have not been reviewed or endorsed by any health professional or expert panel. Read them as a guide to the papers, and read the papers.
What this is not
- Not a diagnosis, treatment plan, personalised recommendation or monitoring service.
- Not a score, streak or optimisation target, and not a claim that more is better.
- Not a substitute for a doctor or other qualified health professional where your own health matters.
Independence
There are no adverts, affiliate links, sponsorships or supplement sales on Purecohere. The free pages will stay free to read.
Report an error
If a figure or claim looks wrong, write to hello@3dcoded.com with the page and the paper. Corrections are logged here.