Tags (every claim carries one, no third tag):
- MEASURED = a numeric value, rate, effect size or count that I read directly off the source named in the row.
- READ = a qualitative statement read directly off the source named in the row.
An untagged claim is not written. "Because X" is inferred unless X was measured, and any inference is marked INFERRED and shows its hops.
Grades (as set by the commission):
- A = randomized human trials or meta-analysis
- B+ = human data, or independently replicated animal work
- B = mechanistic, cell or animal
- C = a single isolated result or case report
- NG (not graded) = position statement, clinical guidance document, religious ruling, news report, or any source that is not a study. Carried here because it is the governing text in practice, not because it is evidence of effect.
Identifiers. Real DOI, PMID or PMC only. Where an identifier was not visible on the source I opened, the row says so in plain words. No identifier in this file was derived, guessed or reconstructed.
dryfasting.com has been published since 2014. This evidence record was compiled on 22 November 2026. Every row names the fasting modality the study actually ran, the system it was measured in and the grade. Where no source carries a DOI or PMID the row says so rather than supplying one. Nothing is sold on this site.