Corrections log
When something on this site is wrong, we fix it and log it here with the date. That covers wrong numbers, citations, testing listings and picks. Routine price updates are not listed.
By Erin Rose Updated October 11, 2026 Report an error
Found a problem? Email [email protected]. We read every message and aim to reply within five business days.
What we have fixed
27 fixes since March 21, 2026. Most were citations, numbers and testing listings, which is why we now check those against their source.
- Citation10
- Number or dose7
- Testing listing5
- Claim4
- Product pick1
27 fixes since March 21, 2026. Source: our corrections log.
The log, newest first
-
The media kit said 42% of US adults are low in vitamin D and 20% of women are low in iron. Our own analysis of CDC data says 22.1% and about 10% of women ages 18 to 50. A card also said PPI users have a 65% higher risk of low magnesium; the cited study found 43%. The media kit now matches its sources.
-
Two studies were listed under the wrong author name. They now show the first author PubMed lists: Virgilio 2024 and Lindbergh 2018.
-
We said Sports Research Collagen Peptides was not on the Informed Choice list, or called it Informed Sport. It is on Informed Choice, and the pages now say so.
-
Our omega-3 pages named different best-value picks. They now all name Viva Naturals.
-
A note said Nature Made Calcium 600 + D3 was not confirmed on USP's list. It is on the list.
-
We said Thorne Creatine uses Creapure, but Thorne does not say that. The pages now call it creatine monohydrate, NSF Certified for Sport.
-
We said MegaFood Blood Builder was NSF certified. It is not on the NSF, USP or Informed Sport lists.
-
We cited the NSF listing for the Blue Raspberry flavor, but we link the Watermelon pack. We now cite the Watermelon listing.
-
We said both tubs were only on NSF's basic list. Both are also NSF Certified for Sport.
-
Three magnesium glycinate products had the wrong pack sizes, so our best-value pick was wrong. The pick is now Nature Made.
-
Amazon pack sizes had changed, so cost per serving was off. Each was re-read from the label.
-
Protein amounts did not match the cited studies. They now say 1.3 to 1.8 g per kg of body weight, and leucine figures with no source were removed.
-
Ten study IDs that did not match real papers were replaced with real, checked papers.
-
The Jarrow KSM-66 cost per dose was wrong: one capsule is 300 mg. The inositol value pick was also wrong.
-
A tongkat ali claim cited a heart-surgery paper. It now cites the right review (Leisegang 2022).
-
We said 50,000 IU a week is about 6,000 IU a day. It is about 7,100.
-
We described a 100-person 2026 trial. The study is Liu 2016, with 44 people.
-
We said tongkat ali's best-studied dose was 400 mg; the trials used 200 mg. We said 42% of people lack vitamin D; CDC data says 22.1%.
-
We overstated that one probiotic cuts repeat infections. We also said LGG helps eczema, but the trial found it did not.
-
We said no one had pooled the research on these. Reviews exist, so the claim was removed.
-
The page said Nature Made was not found on USP's list, then labeled it USP Verified. It also showed $0.47 instead of $0.36. Both are fixed.
-
We cited a study of IV magnesium in heart patients. It now cites a review of magnesium taken by mouth.
-
We advised 200 to 400 mg while warning that 350 mg is the upper limit for supplements. Everyday doses now say 200 to 350 mg.
-
We said taking D3 with fat raises absorption. The cited trial found the low-fat group absorbed the most.
-
A key study had the wrong authors. Two papers from one team were also called two research groups.
-
We said the Endocrine Society sets a 30 ng/mL target. Its 2024 guideline dropped that target.
-
Started this public corrections log and added a link to report errors on every page.
-
Removed a citation (PMID 41601871) that was invalid. It was found in a citation check.
How we try to prevent errors
- We check every PubMed ID on the site against PubMed. The last full check, on October 11, 2026, found all 1,319 IDs were real. It also found 2 wrong author names, which we fixed.
- Label data comes from the NIH label database and the product’s own label, not from ads.
- Testing claims are checked against the tester’s own list, with the date we checked.
- Cost per dose is worked out by formula, shown on our methodology page.
What counts as a correction
We log a change when something we published was wrong: a claim, number, citation, testing listing, pick or price. We do not log new pages, design changes or normal price updates.