> ## Content Index > Fetch the complete content index at: https://compoundcodex.org/llms.txt > Use this file to discover other available public pages before exploring further. # Ginkgo biloba for preventing dementia - URL: https://compoundcodex.org/ginkgo-biloba/ - Published: 2026-08-25T15:43:22.000Z - Updated: 2026-08-25T15:53:12.000Z - Description: Two multi-year trials in nearly six thousand older adults, one publicly funded and one paid for by the manufacturer. Both found nothing. - Author: The Archivist - Tags: nootropics, grade-strong, verdict-not-supported, ginkgo Strong evidence Not supported Marks reviewed Aug 2026 6 sources cited Next review Aug 2027 --- ## The rare case of a settled question Most of what this publication grades is under-tested. Ginkgo is the opposite. Two very large randomised trials, running for five and six years, funded from different countries and run by different investigators, asked whether ginkgo prevents dementia in older adults. Between them they followed almost six thousand people. Both found nothing. Not a small effect, not a promising signal in a subgroup — nothing. That is an unusual and valuable result. It means the four dots at the top of this page are not a statement of confidence in the supplement. They are a statement of confidence in the answer. ## The two trials that settled it #### Large, long, placebo-controlled prevention trials GEM — United States, 120 mg twice daily, median 6.1 years[1](#c1)3,069 people GEM — hazard ratio, all-cause dementia1.12 (0.94–1.33) GEM — hazard ratio, Alzheimer's disease1.16 (0.97–1.39) GuidAge — France, EGb761, 5 years[3](#c3)2,820 people GuidAge — Alzheimer's incidence, ginkgo vs placebo4% vs 5% Combined participants5,889 The **Ginkgo Evaluation of Memory study** enrolled 3,069 adults aged 75 and over across American academic medical centres, randomised them to 120 mg of standardised extract twice daily or placebo, and followed them for a median of 6.1 years. Dementia was assessed every six months.[1](#c1) Read the hazard ratios in the table again. **Both are above 1.0.** The ginkgo group developed slightly *more* dementia than the placebo group, not less. Neither difference reached statistical significance, so the honest reading is "no effect" rather than "harm" — but there is not the faintest hint of benefit hiding in these numbers to be rescued by a larger trial. A companion analysis of the same cohort asked a softer question: forget dementia diagnoses, did ginkgo slow the ordinary rate of cognitive decline? It measured memory, attention, visuospatial ability, language and executive function separately. Annual decline did not differ between groups on any of them.[2](#c2) **GuidAge** then ran the experiment again in France, in a different population — 2,820 adults who had spontaneously complained of memory problems to their doctor — using EGb761 for five years. Probable Alzheimer's developed in 4% of the ginkgo group and 5% of the placebo group.[3](#c3) ## Who paid, and why it matters here GEM was funded by the US National Institutes of Health. GuidAge was funded by **Ipsen, which manufactures EGb761** — the very extract under test. Elsewhere on this site, manufacturer funding is a reason for caution, and our rubric caps a claim's strength when every trial traces to one sponsor. That logic does not apply symmetrically here, and it would be sloppy to pretend otherwise. Sponsor bias pushes results *toward* the sponsor's product. A manufacturer-funded trial that spends five years and enrols nearly three thousand people and then reports that its own extract did not work is evidence made *more* credible by its funding, not less. The incentive ran against the finding and the finding survived. So the independence test is met twice over: an independent public funder found nothing, and the manufacturer found nothing either. ## What ginkgo is still argued about Prevention in healthy older adults is settled. One adjacent question is not, and we grade it separately. The cognitive-decline finding is not listed as its own claim: it comes from a secondary analysis of the same GEM cohort, so grading it separately would count one trial twice. Prevents dementia or Alzheimer's disease Strong Not supported Improves symptoms in people who already have dementia Moderate Mixed The second is a genuinely different question — treating an existing disease rather than preventing a future one — and the trials there disagree with each other in ways that dose and population do not resolve. Systematic reviews have repeatedly described the evidence as inconsistent and unreliable.[4](#c4) Mixed caps strength at Moderate under our rubric, and that is where it sits. None of that rescues the prevention claim, which is the one printed on the bottles. ## Safety and unknowns - Bleeding risk Ginkgo inhibits platelet-activating factor, and case reports describe bleeding events in people also taking anticoagulants or antiplatelet drugs. The large trials did not show a clear excess of major bleeding, but the interaction is the reason surgeons routinely ask patients to stop it before an operation. - Seizure risk from seeds Ginkgo seeds contain ginkgotoxin, which can provoke seizures. Standardised leaf extracts are the form used in trials and the form sold as supplements; raw seeds are a different exposure entirely. - Tolerability Well tolerated across both large trials. Whatever else is true of ginkgo, six years of daily use in three thousand elderly people did not produce an alarming safety signal. - What you are actually buying Both trials used standardised extracts — EGb761 or its equivalent — at defined doses. Retail ginkgo products vary in extract standardisation, and a bottle that does not state its standardisation is not obviously the same substance that was tested and found not to work. ## How the grade was reached #### Evidence strength — Strong, criteria checked Qualifying randomised placebo-controlled trials2 very large, plus others ✓ Trials with 100+ participants each2, both >2,800 ✓ Total participants5,889 ✓ Independent groups2 — US public, French ✓ Published systematic reviewyes ✓ Replication across distinct populationshealthy 75+, memory-complaint 70+ ✓ Verdict is Supported or Not supported, not MixedNot supported ✓ Trial duration adequate for the claim5 and 6.1 years ✓ Every criterion met. This is one of the few claims on this site where the research base is not the limiting factor — the question was asked properly and answered clearly. ## Bottom line Ginkgo biloba does not prevent dementia and does not slow cognitive decline. Two multi-year randomised trials in nearly six thousand older adults, one publicly funded and one funded by the extract's own manufacturer, both reported no benefit. In the larger of the two, the dementia rate was marginally higher on ginkgo than on placebo. It remains widely sold for exactly the purpose it was tested for and failed at. That gap between what the evidence says and what the shelf says is the reason this publication exists. Four dots here mean the question is closed, not that the supplement works. nootropics grade-strong verdict-not-supported ginkgo ## Sources 1. DeKosky ST, Williamson JD, Fitzpatrick AL, et al. Ginkgo biloba for prevention of dementia: a randomized controlled trial. *JAMA*. 2008;300(19):2253–2262\. PMID 19017911\. [pubmed.ncbi.nlm.nih.gov/19017911](https://pubmed.ncbi.nlm.nih.gov/19017911/?ref=compoundcodex.org) 2. Snitz BE, O'Meara ES, Carlson MC, et al. Ginkgo biloba for preventing cognitive decline in older adults: a randomized trial. *JAMA*. 2009;302(24):2663–2670\. PMID 20040554\. [pubmed.ncbi.nlm.nih.gov/20040554](https://pubmed.ncbi.nlm.nih.gov/20040554/?ref=compoundcodex.org) 3. Vellas B, Coley N, Ousset PJ, et al. Long-term use of standardised ginkgo biloba extract for the prevention of Alzheimer's disease (GuidAge): a randomised placebo-controlled trial. *The Lancet Neurology*. 2012;11(10):851–859\. Funded by Ipsen. [thelancet.com](https://www.thelancet.com/journals/laneur/article/PIIS1474-4422%2812%2970206-5/abstract?ref=compoundcodex.org) 4. Birks J, Grimley Evans J. Ginkgo biloba for cognitive impairment and dementia. *Cochrane Database of Systematic Reviews*. Concluded the evidence is inconsistent and unreliable. [cochranelibrary.com](https://www.cochranelibrary.com/cdsr/doi/10.1002/14651858.CD003120.pub3/full?ref=compoundcodex.org) 5. National Center for Complementary and Integrative Health. The Ginkgo Evaluation of Memory (GEM) study. Trial funder's summary. [nccih.nih.gov](https://www.nccih.nih.gov/health/the-ginkgo-evaluation-of-memory-gem-study?ref=compoundcodex.org) 6. ClinicalTrials.gov. Ginkgo Biloba Prevention Trial in Older Individuals. NCT00010803\. [clinicaltrials.gov/study/NCT00010803](https://clinicaltrials.gov/study/NCT00010803?ref=compoundcodex.org) **The Compound Codex** — marks are assigned under the [public editorial standards](https://compoundcodex.org/editorial-standards/) in force on the day of publication, and every mark is [open to challenge](https://compoundcodex.org/challenge/). Informational only. Nothing here is medical advice, and nothing here endorses the use of any compound discussed. © 2026 The Compound Codex · Every claim cited, none of them prescriptive.