The Evidence Grade Register — changes Every grade movement in the register, dated, with the reason behind it. 2026-09-22T06:53:10.852Z https://grade.nexagenlabs.com/ Suryaprakash Tripathy Talking therapy: held at B, reason revised https://grade.nexagenlabs.com/talking-therapy/#2026-09-22 2026-09-22T00:00:00Z The source was identified and the register had been describing it wrongly. This entry said the pooled estimate compared people with themselves before and after treatment, with no control group held throughout, and treated that as one of the reasons for the grade. It is not what the analysis does: it compares the change in the therapy groups against the change in usual care. The description has been removed from every field that carried it. The grade holds at B, because the two other reasons are untouched — the trials disagree with one another, with heterogeneity for depression above fifty percent, and at twelve months no effect could be detected in either depression or anxiety. Exercise: A to B https://grade.nexagenlabs.com/exercise/#2026-09-16 2026-09-16T00:00:00Z The only source on this claim is the roundtable consensus statement, which summarises randomised trials rather than reporting one. No trial reporting physical function or fitness as an endpoint could be cited here. The ceiling on a guideline is B, and the grade moves to meet it rather than the ceiling being set aside. Garlic: D to C https://grade.nexagenlabs.com/garlic/#2026-09-16 2026-09-16T00:00:00Z Randomised human outcome data exists. The Shandong Intervention Trial reported reduced gastric cancer mortality with garlic supplementation, with incidence not reduced. The earlier reason, that there was no trial evidence of benefit, was wrong. Seed oils: held at D, reason revised https://grade.nexagenlabs.com/seed-oils/#2026-09-16 2026-09-16T00:00:00Z Grade held, reason inverted. The D is not because the question is open but because the evidence supporting the accusation is weak. Prospective cohort meta-analysis found no significantly increased risk, and the entry was rewritten around that rather than around an absence. Astragalus: D to C https://grade.nexagenlabs.com/astragalus/#2026-09-15 2026-09-15T00:00:00Z Randomised human outcome data exists. The earlier description of this entry as having no outcome evidence was wrong. Beta-blockers: held at C https://grade.nexagenlabs.com/beta-blockers/#2026-09-15 2026-09-15T00:00:00Z Grade held, entry rewritten. The observational evidence is null rather than favourable, and in studies corrected for immortal time bias the signal reverses. Dichloroacetate: held at D https://grade.nexagenlabs.com/dichloroacetate/#2026-09-15 2026-09-15T00:00:00Z Grade held, two errors corrected. Bioavailability is near-complete so the concentration argument does not apply, and the neuropathy is dose-dependent and reversible rather than a hard ceiling. Eradicating H. pylori: newly graded A https://grade.nexagenlabs.com/h-pylori-eradication/#2026-09-15 2026-09-15T00:00:00Z Newly graded. Pooled randomised trials: incidence down 36 percent, mortality down 22 percent. Essiac: held at D https://grade.nexagenlabs.com/essiac/#2026-09-15 2026-09-15T00:00:00Z Grade held deliberately. Every harm signal is animal or cell culture, and an E in this book requires human evidence pointing the wrong way. Folate: D to E https://grade.nexagenlabs.com/folate/#2026-09-15 2026-09-15T00:00:00Z Cole et al. JAMA 2007: more advanced adenomas in the folic acid arm, relative risk 1.67, and more than twice the rate of three-or-more adenomas. Graviola, soursop: D to X https://grade.nexagenlabs.com/graviola-soursop/#2026-09-15 2026-09-15T00:00:00Z Annonacin linked to an irreversible parkinsonian syndrome, and no intake level that could be considered without concern could be established. Mebendazole: held at C https://grade.nexagenlabs.com/mebendazole/#2026-09-15 2026-09-15T00:00:00Z Grade held, reasoning corrected. No trial with a mebendazole-free comparator exists, so the evidence base is thinner than the book claimed rather than more balanced. Milk thistle: held at C https://grade.nexagenlabs.com/milk-thistle/#2026-09-15 2026-09-15T00:00:00Z Grade held. A randomised trial was located, and the entry is better supported than the earlier draft assumed. Statins: C to D https://grade.nexagenlabs.com/statins/#2026-09-15 2026-09-15T00:00:00Z Pooled randomised trials gave an overall survival hazard ratio of 0.94 with an interval crossing one, against an observational literature of over a million patients claiming 30 to 40 percent mortality reductions. Talking therapy: A to B https://grade.nexagenlabs.com/talking-therapy/#2026-09-15 2026-09-15T00:00:00Z Moved from A to B during the citation audit, on the heterogeneity between the trials and the absence of any detectable effect at twelve months. A third reason was recorded at the time and was wrong; it was withdrawn on 2026-09-22, and the entry above, dated 2026-09-22, says what it was. The grade did not move again. Treating major depression: newly graded A https://grade.nexagenlabs.com/treating-major-depression/#2026-09-15 2026-09-15T00:00:00Z Newly graded. SMaRT Oncology-2: 62 percent response against 17 percent in usual care. Vitamin C, intravenous: held at D https://grade.nexagenlabs.com/vitamin-c-intravenous/#2026-09-15 2026-09-15T00:00:00Z Grade held. The reason was corrected: this is not a compound that cannot reach its target concentration, it is one that can and has never been tested for outcomes. Vitamin C, oral: held at D https://grade.nexagenlabs.com/vitamin-c-oral/#2026-09-15 2026-09-15T00:00:00Z Grade held. The oral and intravenous routes were separated, and the concentration difference between them was found to be the reason the historical trials disagreed, rather than an absence of evidence.