Changelog
Every grade change from the last 30 days, with the studies behind it. When a grade moved because we changed how we process studies, we say so.
A change is listed once the new grade has held for at least a day and a half. Grades that moved back and forth near a boundary are left out until they settle.
2 new studies, and recalculated after a method update: More complete extraction (Oct 18, 2026)
No new studies. Recalculated after a method update: More complete extraction (Oct 18, 2026)
No new studies. Recalculated after a method update: More complete extraction (Oct 18, 2026)
3 new studies, and recalculated after a method update: More complete extraction (Oct 18, 2026)
No new studies. The grade was recalculated from the studies already linked.
No new studies. The grade was recalculated from the studies already linked.
No new studies. The grade was recalculated from the studies already linked.
No new studies. The grade was recalculated from the studies already linked.
No new studies. The grade was recalculated from the studies already linked.
Changes to how we extract, link or score studies. Grades can move after these without any new research.
Oct 20, 2026
Complete study counts
Grades for the most-studied supplements were calculated from an incomplete set of findings that varied from night to night, which made some grades flip back and forth. Every grade is now calculated from the complete set, so some grades moved once.
Oct 18, 2026
More complete extraction
We re-read 4,708 papers with an improved, benchmark-tested extraction step that records each finding separately. This added about 20% more findings per paper, so some grades moved without any new studies.
Oct 12, 2026
Caught up on missed papers
A pipeline fault had left 4,527 papers added between July and October without extracted findings. We extracted them and recalculated grades. Grades that moved on this date reflect studies we should have counted earlier, not new research.
Mistakes on our pages that readers reported or we found, and what we changed.
Oct 19, 2026
Macular health: vitamin E form corrected
The protocol listed 400 IU of mixed tocopherols. AREDS, the trial the protocol is based on, used 400 IU a day of synthetic vitamin E (dl-alpha-tocopheryl acetate). The protocol now names the form that was tested. Reported by a reader.
Oct 19, 2026
Macular health: who it is for, copper added, omega-3 removed
AREDS and AREDS2 found that the formula slowed progression in people who already have intermediate age-related macular degeneration, not in people without it. The protocol now says so, adds the 2 mg of copper both trials paired with zinc, and drops omega-3, which AREDS2 found did not slow progression.
Oct 19, 2026
Dry eye: omega-3 evidence narrowed
A 2026 meta-analysis of 27 randomized trials found that long-chain omega-3 (EPA and DHA) helped screen-related and contact-lens dry eye, but not dry eye from meibomian gland dysfunction or unknown causes, and plant omega-3 (ALA) did not help. The protocol now says which kinds of dry eye the evidence supports. Reported by a reader.
Omega-3 forms and dry eye etiologies, meta-analysis of 27 RCTs (BMC Ophthalmology 2026)
A two-minute quiz builds a daily plan of habits, diet and supplements around your goals and what you already take. Free, no account needed.
Start the quiz