Corrections log
Everything we have got wrong, since launch
We correct on the page, keep the old wording visible in the note, and add a permanent row here. We do not silently edit, we do not delete an article to make an error disappear, and we do not remove old rows.
Correction
A published fact, figure, quotation, reference, or interpretation was incorrect. We replace it with accurate information and explain what was wrong and why it changed. Corrections affecting safety receive a prominent notice on the article.
Clarification
A statement needed more context or could reasonably be misunderstood. We revise the wording and explain the distinction so readers can understand the evidence and its limits.
Update
New research, official guidance, or regulatory developments have changed the information available. We revise the relevant content, record the development, and update the review date where appropriate.
We described a 2018 rodent study as measuring blood pressure continuously; it sampled at fixed intervals. The onset-timing paragraph was rewritten and the claim regraded from Supported to Contested.
S. AydınAgency guidance reissued June 2026. Novel-food classification wording aligned to the new text; check date advanced.
R. MehtaA sentence comparing two case series could be read as implying a threshold amount. Rewritten. No quantity was stated in either version.
M. LindqvistNesting altitude range was given from a secondary source and was too narrow at the lower bound. Corrected against the original survey.
P. GurungThe comparison table’s onset row implied direct comparability between two differently-measured literatures. A caveat row was added above it.
I. ThapaBook and chapter reference for the Anabasis passage was wrong. Corrected, and the translation used is now named inline.
P. GurungRewritten after clinical review to remove drug-class generalisations that were not directly evidenced, and to route the whole question to a pharmacist or prescriber.
T. WeberA worked example of a report layout was mistaken by several readers for an endorsement of a testing method. Reframed and relabelled.
I. ThapaShow earlier corrections, 2025 Eight entries shown. The 2025 rows load as their own page so the archive stays linkable without JavaScript.
Tell us we got something wrong
Found an inaccurate statement, outdated source, or misleading explanation? Send us the article link, quote the relevant passage, and explain what you believe needs correcting. A reliable supporting source helps us assess your report.