We review brain supplements, and the honest case against buying one is strong. The largest, longest and best-funded trials in this field have mostly failed: ginkgo in roughly 5,900 people over five to six years, B vitamins in about 22,000 people, cocoa flavanols in more than 5,700, vitamin D in two separate trials, high-dose CoQ10 stopped for futility. Where positive results exist they are small and tied to a specific population. Add a documented adulteration problem and several real safety signals, and the reasonable default is to spend the money on sleep, a blood test, or nothing.

The claim

The claim we are examining is the category’s founding assumption: that a healthy adult who takes a brain supplement will think, remember or concentrate measurably better than the same adult who does not. Not that a specific nutrient corrects a specific deficiency — that is a different and defensible claim — but that supplementation is a general upgrade available to anyone with a credit card.

This page argues against that assumption. We are a site that writes about these compounds for a living, which makes the argument worth stating plainly rather than burying it: if you read only one of our pages before buying something, we would rather it were this one or the tier list.

Where it came from

Two decades of observational epidemiology built the expectation. People who eat more fish have lower dementia rates in cohort studies. People with low vitamin D have higher dementia risk. People with high homocysteine do worse cognitively. Each of those associations is real, replicated and, taken alone, persuasive.

What followed was the standard sequence: a plausible mechanism, a small positive trial, a supplement industry, and then — years later and much more quietly — the large randomized trials designed to test whether the association was causal. The associations largely did not survive that test. Frail, housebound and already-declining people go outside less and have lower vitamin D; that is reverse causation, not a treatable deficiency of brain function.

What the evidence actually shows

This is the part the category does not put on its labels. Below are the landmark trials in cognitive supplementation — chosen because they were large, long, or otherwise decisive, not because they were negative.

TrialIntervention and populationDurationResult
GEM (Ginkgo Evaluation of Memory)EGb 761 240 mg/day, ~3,069 older US adultsMedian 6.1 yearsNo reduction in all-cause dementia or Alzheimer’s incidence
GuidAgeGinkgo, ~2,854 French older adults with memory complaints5 yearsNo reduction in progression to Alzheimer’s
Cochrane ginkgo review (2026)82 studies, 10,613 participantsSearched to Nov 2024Probably little or no difference at six months in mild cognitive impairment1
Clarke et al.Homocysteine-lowering B vitamins, 11 trials, ~22,000 individualsPooledNo significant effect on cognitive aging2
QE3CoQ10 1,200 and 2,400 mg/day, early Parkinson’s diseaseTerminated earlyStopped for futility — no evidence of benefit3
VitaMINDVitamin D in adults with mild-to-moderate deficiencyRandomizedNo significant cognitive benefit4
Finnish Vitamin D TrialD3 in healthy older adultsMulti-yearNo reduction in incidence of diagnosed dementia5
COSMOS cocoa arm500 mg cocoa flavanols/day, >5,700 combined3 yearsNo cognitive effect at 1 year or across 3 years6
Omega-3 trials (per NIH ODS)500–2,000 mg/day EPA+DHA, healthy older adults and established Alzheimer’s6–24 monthsNo significant cognitive improvement7
EFSA assessment (2024)Creatine and cognitive function health claim—Cause-and-effect not established; claim rejected8
Jorissen et al.Soy-derived phosphatidylserine in age-associated memory impairmentRandomizedNo significant benefit on cognition9
Cochrane huperzine A6 RCTs, 454 Alzheimer’s patientsUp to 36 weeksEvidence inadequate to make any recommendation10
Cochrane Panax ginsengHealthy participantsPooledLack of convincing evidence of a cognitive-enhancing effect11

The one large trial that worked was not a brain supplement

COSMOS randomized thousands of older adults to a daily multivitamin-mineral or placebo for three years and tested cognition by telephone, on the web and in person. COSMOS-Mind, with 2,262 participants, reported global cognition improving relative to placebo — a slowing of cognitive aging estimated at roughly 60%, about 1.8 years over three years. COSMOS-Web, with more than 3,500 participants, found a memory benefit at one year sustained through three.612

The product was Centrum Silver. Benefits were larger in participants with a history of cardiovascular disease, which points at correcting marginal micronutrient insufficiency rather than at any pharmacological effect. The most successful intervention in this literature costs a few cents a day and contains no botanicals at all. That is the shape of the case against brain supplements: not that nothing helps, but that what helps is cheap, boring and already in a pharmacy.

Even the positive results are smaller than they sound

Creatine’s memory effect across eight trials in 225 healthy participants was a standardized mean difference of 0.29 — and it was concentrated almost entirely in adults aged 66 to 76, where it reached 0.88, against roughly zero (0.03) in participants aged 11 to 31.13 Bacopa’s meta-analysed benefit, after twelve weeks or more of a standardized extract, was a Trail Making Test B time shortened by 17.9 milliseconds and a choice reaction time reduced by 10.6 milliseconds.14 Those are genuine measurements. They are also milliseconds.

You often do not know what is in the bottle

A 2020 analysis in Neurology: Clinical Practice tested US-sold cognitive-enhancement supplements and detected five unapproved drugs — omberacetam (noopept), aniracetam, phenibut, vinpocetine and picamilon — frequently at doses above pharmacological norms and often undeclared or mis-declared on the label.15 A proprietary blend compounds the problem legally rather than chemically: it conceals per-ingredient amounts, so a label can list twelve impressive botanicals in 500 mg total without any of them reaching a dose that has ever been studied.

The risks are not zero

⚠️ The safety signals worth knowing before you buy

Ashwagandha: NIH LiverTox assigns it a likelihood score of B — a likely cause of clinically apparent liver injury, with roughly 23 published cases as of the 2024 update, latency typically two to twelve weeks, mostly cholestatic or mixed. Rare cases required transplant; Denmark and Iceland have restricted or advised against these products.16 Alpha-GPC: a Korean national cohort of more than 12 million adults found a 43% higher ten-year hazard of stroke in users, with a dose-response gradient — observational, confounded by indication, and still the largest safety datum on the ingredient.17 Vitamin B6: chronic high doses cause sensory peripheral neuropathy, reported at intakes as low as 50 mg/day with prolonged use.18 Ginkgo: ginkgolides inhibit platelet-activating factor, so anyone on warfarin, a DOAC, clopidogrel or aspirin should clear it with a prescriber and stop it at least two weeks before surgery.1

Why the claim persists

The failures are quiet and the mechanisms are loud. A press release accompanies a positive pilot study in 30 people. A five-year trial in 3,000 people that finds nothing is a paragraph in a journal. Nobody markets a null result.

Subjective improvement is easy to produce. A supplement bought deliberately, taken daily and expected to work will usually be experienced as working, particularly when the outcome being judged is “how sharp did I feel today”. This is why blinding exists, and why uncontrolled self-experiments cannot settle the question.

Regulation moves slowly. The FTC sued the marketer of Prevagen in January 2017 over memory claims for apoaequorin, and it was December 2024 before a federal court ordered the company to stop making the deceptive memory and cognitive-improvement claims.19 Eight years is a long time to sell a product to a worried customer.

Community reports — not trial data

The most common pattern in supplement forums is a strong first two weeks that fades, followed by a dose increase or a new addition to the stack. That is exactly what an expectation effect looks like when it decays. It is also what a real effect looks like when tolerance develops. Uncontrolled reports cannot distinguish the two, which is the whole reason placebo control was invented.

What to do instead

The case against brain supplements is not a case for doing nothing. It is a case for spending attention where the effect sizes are larger.

  • Sleep first. Caffeine’s half-life is about five to six hours, so afternoon coffee measurably degrades sleep architecture — and sleep is the largest cognitive variable available to you. Our methods pages cover this without selling anything.
  • Test, do not guess. B12 deficiency causes real and sometimes reversible cognitive symptoms, and the people at risk are identifiable: adults over 50, vegans, people on metformin or long-term acid suppression, people with pernicious anaemia or after bariatric surgery. A blood test answers this; a capsule does not.
  • Rule out the causes that need a doctor. Persistent cognitive change is worth a clinical conversation before it is worth a purchase — see the brain fog hub.
  • If you are over 65, the multivitamin is the one with the trial. Check its B6 content, and separate iron- or calcium-containing products from levothyroxine by four hours.
  • If you still want to experiment, buy single ingredients at studied doses rather than blends, give them the duration the trials used, and tell your prescriber. How we rate evidence explains what counts as a studied dose.

What would change our mind

A large, preregistered, independently funded randomized trial — a thousand healthy adults or more, twelve months or longer, one named compound at a disclosed dose, with a prespecified primary cognitive endpoint — reporting a benefit that survives sensitivity analysis and is then replicated by a second group with no commercial relationship to the first. COSMOS demonstrates that this scale of trial is achievable in this field. No botanical brain supplement has one.

Frequently asked questions

Are brain supplements a waste of money?

For a healthy, well-fed adult with no deficiency, the trial evidence does not support expecting a cognitive benefit from a botanical brain supplement. The exceptions are narrow: correcting an actual deficiency, and a daily multivitamin-mineral in adults over 65, where COSMOS found a modest effect.

Why do you review these products if you think this?

Because people buy them anyway, and the difference between a bacopa product at a studied dose and a twelve-botanical proprietary blend is real and worth explaining. Our editorial policy sets out where we do and do not carry commercial links — debunk pages like this one never do. If you want to see these criteria applied to real products, our NeuroPrime review covers a blend that does not disclose per-ingredient doses, and our Neuro-Thrive review covers one that prints every dose.

Does the failure of one ingredient mean the whole category fails?

No, and we have not argued that. The argument is cumulative: the largest trials of the most-promoted ingredients failed, the surviving effects are small and population-specific, the products are inconsistently what they say they are, and some carry documented harms. Any one of those alone would be weak. Together they set the default.

What about all the positive studies I have seen cited?

Check three things: how many participants, how long, and who paid. In this field, positive results cluster in trials of 30 to 100 people run for 8 to 12 weeks, frequently with manufacturer involvement, and they are rarely replicated by an independent group.

Related reading

These statements have not been evaluated by the Food and Drug Administration. Nothing here is intended to diagnose, treat, cure or prevent any disease.

This article is information, not medical advice. Talk to a doctor or pharmacist about your own situation, especially if you take prescription medication.