The biggest brain-supplement trials since 2018 mostly found nothing

The memory shelf at a large pharmacy usually holds two kinds of bottles. One promises focus and recall, with a long list of botanicals, mushroom extracts and chemical-sounding compounds. A few feet away sits a plain daily multivitamin that says nothing about your brain at all.

Ask which one has better evidence behind it, and the answer from the biggest trials and reviews since 2018 runs against intuition. Studies of fish oil, cocoa extract and B vitamins in thousands of older adults mostly found nothing. The one product that moved memory scores in a big trial was the plain multivitamin, and that effect was small and came from a single study program. Meanwhile, independent labs testing specialty focus blends have found drugs never approved in the United States.

The evidence up to 2018, including a federally funded review of 38 supplement trials, a six-year ginkgo trial and a long legal fight over one memory supplement’s ads, is laid out in the library’s piece on why a legal label claim can still be an illegal advertisement. This article picks up from there.

Fish oil, B vitamins and vitamin E: big trials, flat lines

Observational research has been kindest to omega-3 fats. The National Center for Complementary and Integrative Health (NCCIH) says they have the best record of any dietary factor studied against cognitive decline in older people, usually measured by how much fish people eat. Capsules are a separate question. In 2022, a team led by Jae H. Kang at Brigham and Women’s Hospital reported thinking results from VITAL, a national trial of marine omega-3 capsules and vitamin D whose heart findings appear in the library’s omega-3 article. Among 3,424 healthy adults aged 60 and older tested by phone over about 2.8 years, and 794 more tested in person over two years, the omega-3 group did no better than the placebo group.

That matched a 2012 Cochrane review, which combined three trials in cognitively healthy older people and found no benefit on thinking scores after as long as 40 months. None of those trials measured whether anyone developed dementia.

A 2026 trial at the University of Southern California sharpened the point. Researchers gave high-dose DHA, one of the main omega-3s in fish oil, or a placebo to 365 adults aged 55 to 80 who ate little fish and had at least one dementia risk factor. Within six months, DHA’s share rose in the fluid that bathes the brain and spinal cord. After 24 months, the groups did not differ on thinking tests or brain volume. More than a third dropped out, mostly because of COVID-19. Still, it illustrates a rule worth keeping: getting a compound into the brain is not the same as improving memory.

B vitamins tell a similar story. A 2018 Cochrane review of cognitively healthy adults found 14 trials with 27,882 participants, most in their 60s and 70s and many with heart or blood vessel disease. B vitamin supplements probably had little or no effect on overall thinking at any point up to five years.

A companion Cochrane review that year looked at people with mild cognitive impairment, measurable memory or thinking trouble that stops short of dementia. Five trials with 879 people found probably little or no effect on memory, planning or processing speed over six months to two years. One trial hinted at benefit in people with high homocysteine, an amino acid that climbs when B vitamin status is low, and the reviewers said that hint needs repeating.

Those trials gave extra vitamins to people who were not deficient. A true B12 deficiency, testable and treatable, is covered alongside the library’s diet and memory trials.

Vitamin E did no better. In the mild cognitive impairment review, one trial of 516 people found high-dose vitamin E probably did not slow progression to Alzheimer’s dementia over three years. The healthy-adult review flagged a safety signal from another trial: more prostate cancer diagnoses among men taking vitamin E. A single trial of vitamin D with calcium in 4,143 people found no effect on thinking or dementia for up to 10 years.

On the library’s evidence scale, omega-3 and B vitamin supplements for keeping healthy older adults sharp are tested and not supported over the two to five years these trials ran. Whether decades of use would differ is untested.

Cocoa flavanols: three tests, three misses on the main question

Flavanols are plant compounds found in cocoa, tea and some fruits. Cocoa extract was one of two products tested in COSMOS, a trial of 21,442 US adults aged 60 and older whose main goal was preventing heart disease and cancer. Three offshoot studies measured thinking.

• COSMOS-Mind, run from Wake Forest University, tested 2,262 people by phone every year for three years. Cocoa extract was its primary question, and it had no effect on global cognition.
• COSMOS-Web tested 3,562 people online. Its prespecified cocoa result, memory after one year across everyone, was not statistically significant. A 2023 paper in the Proceedings of the National Academy of Sciences reported gains only among people whose usual diets were lower in quality or in flavanols. Two of its authors listed an affiliation with the company that supplied the extract.
• COSMOS-Clinic tested 573 people in person over two years. Cocoa extract had no effect on overall thinking, memory or executive function, the planning and self-control skills. Hints of benefit among people with poorer diets came from subgroup analyses not corrected for asking many questions at once.

When a main question misses, slicing participants into subgroups will usually turn up some group where the product seems to work. Such slices are leads, not findings. Cocoa extract for memory in well-fed older adults is tested and not supported; the idea that it helps people with poor diets is insufficient evidence, never tested as a main question.

The multivitamin signal: real, small and from one program

COSMOS also tested a standard commercial multivitamin, and here the results went the other way.

• In COSMOS-Mind, the multivitamin was the prespecified secondary question. Over three years, people taking it scored modestly higher than the placebo group on a global thinking composite, and on memory and executive function.
• In COSMOS-Web, led by Lok-Kin Yeung and Adam M. Brickman of Columbia University, the multivitamin group did better at recalling a word list right after hearing it, the primary outcome, at one year and on average across three years. No secondary outcome moved.
• In COSMOS-Clinic, with the most detailed in-person testing, the overall thinking score was not significantly different. Episodic memory, the kind you use to recall what happened yesterday, was.

In January 2024, Chirag M. Vyas, Olivia I. Okereke and colleagues pooled the three non-overlapping groups, more than 5,000 people. The multivitamin improved global cognition by 0.07 standard deviations and episodic memory by 0.06.

How big is 0.07?

A standard deviation is the typical spread of scores across a group. By our arithmetic, a shift of 0.07 would move someone ranked exactly in the middle of 100 people up to about 53rd. A trial of thousands can detect that; whether anyone would notice it at the kitchen table is another matter.

The researchers translated the effect into years: about two years of cognitive aging avoided in the pooled analysis, 3.1 years of memory change in COSMOS-Web and 1.8 years in COSMOS-Mind. The Web and Mind figures come from comparing older and younger participants at one point in time, not from watching anyone age. The COSMOS-Mind team called its estimate an imprecise yardstick and warned that practice effects, people improving at tests simply by repeating them, distort numbers like this.

Practice effects showed up plainly: in COSMOS-Mind, scores in both the cocoa and placebo groups rose for two years before leveling off, which the authors attribute to retesting. The multivitamin finding rests on the gap between groups.

Why caution still applies

• One program, one product. All three studies sit inside one parent trial and tested one multivitamin. Pooling them is not independent replication.
• Who took part. COSMOS-Mind was 89% non-Hispanic White, and its authors called for confirmation in a more diverse group.
• No dementia outcome. None of the three measured whether people developed dementia.
• An earlier null trial. In the Physicians’ Health Study II, a 12-year trial in male doctors, a cognitive substudy of 5,947 men aged 65 and older found no multivitamin effect. The COSMOS-Mind authors note that its testing began an average of 2.5 years after people started their pills, so an early benefit could have been missed.
• No mechanism yet. A September 2026 COSMOS paper tracked three blood markers of nerve-cell damage and Alzheimer’s-related proteins in about 500 people over two years. Neither supplement changed them.

Two parts of the National Institutes of Health describe this evidence differently, and the reason is the calendar, not a dispute. NCCIH’s clinician digest, dated July 2022, says multivitamins have no clear cognitive benefit in well-nourished people; it predates COSMOS-Web and the pooled analysis. The Office of Dietary Supplements’ fact sheet, updated in July 2024, summarizes all three studies and concludes the findings suggest multivitamins can help maintain or improve thinking in older adults. Neither page is a clinical guideline, and the fact sheet gives no effect sizes.

On the library’s scale, the multivitamin result is positive evidence with a contested interpretation: a consistent, small effect from one trial program, not yet confirmed outside it. Whether a multivitamin makes sense for a particular person, given diet, medicines and other supplements, is a question a clinician or pharmacist can weigh.

Single-ingredient memory products and how small their trials are

Specialty formulas build on single ingredients whose trials are a different scale from COSMOS.

Huperzine A, from a Chinese herb, is a cholinesterase inhibitor, the same drug class as several prescription Alzheimer’s medicines. A mechanism is not an outcome, though. A US phase II trial published in Neurology in 2011 randomized 210 people with mild to moderate Alzheimer’s disease, and its primary analysis found no cognitive benefit at 16 weeks; a higher dose looked better only in secondary analyses. A 2008 Cochrane review found six trials with 454 patients, most of modest quality, and summed up: “However, only one study was of adequate quality and size.” A 2012 Cochrane review in mild cognitive impairment found no eligible placebo-controlled trials at all. Verdict: insufficient evidence.

Bacopa, an herb from Ayurvedic medicine, has a 2014 meta-analysis of nine trials with 518 people. The measurable gains were small speed improvements on attention tasks, and the authors said only a large head-to-head trial against an existing medicine would settle the question. Verdict: insufficient evidence.

Lion’s mane mushroom has only a handful of small human trials. The best known, from Japan in 2009, enrolled 30 people with mild cognitive impairment; its first author worked at a mushroom company’s laboratory, and scores fell after people stopped. Later placebo-controlled studies are also small. Four in healthy adults aged 18 to 50, with 18 to 41 participants, found no overall effect, or scattered gains on single tasks after one dose. Two in middle-aged and older people reported benefits: a 2019 Japanese trial, on one of three thinking tests, and a 2020 Taiwanese pilot trial in mild Alzheimer’s disease with 41 people finishing, in which only a daily-living score differed significantly from placebo and five authors were paid by the company that supplied the study materials. Neither has been confirmed by a larger trial. The library’s mushroom-supplement article examines the 2009 trial closely.

Phosphatidylserine is the only ingredient listed under cognitive function on FDA’s list of qualified health claim letters, through letters in 2003 and an update in November 2004. By FDA’s own definition, qualified claims fall short of the significant scientific agreement standard that authorized health claims must meet, and they must carry wording that signals how limited the evidence is. FDA says it does not approve these petitions; it issues letters stating it will not object to specific language.

The rules behind structure-and-function label claims are explained in the library’s article on why supplement labels say supports and never cures.

When the focus pill contains a drug

The bigger danger in this aisle is a product that does something nobody disclosed.

In 2021, Pieter A. Cohen of Cambridge Health Alliance and Harvard Medical School, with chemists at the University of Mississippi’s National Center for Natural Products Research, bought 10 supplements online labeled as containing omberacetam, aniracetam, phenylpiracetam or oxiracetam, four drugs not approved for human use in the United States. The lab found five unapproved drugs across the products: omberacetam and aniracetam, plus phenibut, vinpocetine and picamilon.

The labels failed in both directions. Some drugs that turned up were not listed, and some listed drugs were absent. Where labels gave amounts, 9 of 12 were wrong. One product could expose a person to as many as four unapproved drugs, at up to four times a pharmaceutical dose. The paper’s conclusion: “Over-the-counter cognitive improvement supplements may contain multiple unapproved drugs.”

A 2022 follow-up by the same group tested seven supplements listing centrophenoxine, a drug prescribed in China and elsewhere that FDA has never approved for any use. All seven contained it, and only one stated an amount within 10 percent of what was inside.

FDA describes vinpocetine as synthetic, derived from plant compounds but not present in any plant, and in 2016 tentatively concluded that it does not meet the legal definition of a dietary ingredient. In June 2019, FDA warned that vinpocetine may cause miscarriage or harm fetal development, and said people who are pregnant or could become pregnant should not take it. It is sold for memory, focus, energy and weight loss, and Cohen’s team found it in a product whose label did not reveal it.

FDA has also sent warning letters and online advisory letters to 17 companies over more than 58 products claiming to prevent, treat or cure Alzheimer’s disease, many sold as supplements. The agency calls them unapproved or misbranded drugs and warns that they could keep people from a proper diagnosis and treatment.

Prescription stimulants that healthy people take for focus have their own trial record, taken up in the companion piece on study drugs.

Questions to ask before buying a cognitive supplement

This checklist is our own, built from the trials above.

1. How many people were tested, and on what? COSMOS enrolled thousands and measured memory directly. Many ingredients rest on trials of a few dozen people.
2. Did the main question succeed, or only a slice of the data? The cocoa trials missed their primary endpoints; the encouraging headlines came from subgroups.
3. Does the ingredient act like a drug? Huperzine A blocks the same enzyme as prescription Alzheimer’s drugs, which makes combining them a prescriber’s call.
4. Does the label list omberacetam, aniracetam, phenibut, picamilon, vinpocetine or centrophenoxine? Independent testing found such products often contain more, or less, than the label says.
5. Is pregnancy possible? FDA’s vinpocetine warning covers anyone pregnant or able to become pregnant, and the ingredient has turned up unlisted.
6. Is a blood thinner involved? NCCIH notes that ginkgo may raise bleeding risk and that vitamin E interacts with anticoagulant and antiplatelet drugs, a question for a pharmacist.
7. Does it promise to treat or prevent Alzheimer’s? FDA treats that claim on a supplement as an illegal drug claim.

When a memory worry needs a clinician, not a capsule

NCCIH’s consumer guidance makes the central point: complementary approaches should not become a reason to put off seeing a health care provider about memory loss. Memory trouble can come from treatable causes, including depression, reactions to medicines, and thyroid, liver or kidney problems. How that evaluation works, and which changes call for one, are spelled out in the library’s guide to memory assessment.

Situations that call for professional input:

• Memory or thinking changes that are new, getting worse or noticed by people around you. A clinician can look for the treatable causes above.
• Numbness, tingling or unsteadiness in someone taking high-dose vitamin B6. NCCIH notes that chronic high doses can cause severe, progressive nerve damage.
• Unusual bruising or bleeding in someone combining ginkgo or vitamin E with a blood thinner.
• A seizure, chest pain, fainting, confusion or severe agitation after a focus or memory product. Call 911. For questions about a product someone swallowed, without those emergency signs, the Poison Help line at 1-800-222-1222 is toll-free and answers 24 hours a day.
• Any bad reaction to a supplement can be reported to FDA at 1-888-INFO-FDA.

The takeaway

The evidence since 2018 turns the memory aisle upside down. Fish oil, cocoa extract and B vitamins went through large trials in older adults and came back flat, and vitamin E fared no better in mild cognitive impairment. A plain multivitamin produced a small, consistent memory signal, all within one trial program and one product. The specialty formulas rest on trials of a few dozen to a few hundred people, and the focus blends are where independent labs keep finding unapproved drugs. If your memory is the worry, the more useful step is an evaluation that looks for a treatable cause.

Sources

1. National Institutes of Health, Office of Dietary Supplements. “Multivitamin/mineral Supplements: Fact Sheet for Health Professionals (updated July 17, 2024).” ods.od.nih.gov
2. National Center for Complementary and Integrative Health. “Dietary Supplements and Cognitive Function, Dementia, and Alzheimer's Disease: What the Science Says (clinician digest, July 2022).” nccih.nih.gov
3. National Center for Complementary and Integrative Health. “7 Things To Know About Dietary Supplements for Cognitive Function, Dementia, and Alzheimer's Disease.” nccih.nih.gov
4. U.S. Food and Drug Administration. “Vinpocetine in Dietary Supplements (content current as of February 22, 2023).” fda.gov
5. U.S. Food and Drug Administration. “Unproven Alzheimer's Disease Products.” fda.gov
6. U.S. Food and Drug Administration. “Qualified Health Claims: Letters of Enforcement Discretion (content current as of March 28, 2024).” fda.gov
7. U.S. Food and Drug Administration. “Qualified Health Claims (content current as of March 28, 2024).” fda.gov
8. Health Resources and Services Administration. “Poison Help.” poisonhelp.hrsa.gov
9. Alzheimer's & Dementia: Translational Research & Clinical Interventions. “Kang JH, Vyas CM, Okereke OI, et al. "Marine n-3 fatty acids and cognitive change among older adults in the VITAL randomized trial." Alzheimers Dement (N Y), 2022. PubMed. DOI: 10.1002/trc2.12288.” pubmed.ncbi.nlm.nih.gov
10. Cochrane Database of Systematic Reviews. “Sydenham E, Dangour AD, Lim WS. "Omega 3 fatty acid for the prevention of cognitive decline and dementia." Cochrane Database Syst Rev, 2012. PubMed. DOI: 10.1002/14651858.CD005379.pub3.” pubmed.ncbi.nlm.nih.gov
11. EBioMedicine. “Yassine HN, et al. "CNS target engagement of high-dose DHA supplementation in older adults at risk for dementia: a randomised, double-blind, placebo-controlled trial." EBioMedicine, 2026. PubMed. DOI: 10.1016/j.ebiom.2026.106316.” pubmed.ncbi.nlm.nih.gov
12. Cochrane Database of Systematic Reviews. “Rutjes AW, et al. "Vitamin and mineral supplementation for maintaining cognitive function in cognitively healthy people in mid and late life." Cochrane Database Syst Rev, 2018. PubMed. DOI: 10.1002/14651858.CD011906.pub2.” pubmed.ncbi.nlm.nih.gov
13. Cochrane Database of Systematic Reviews. “McCleery J, et al. "Vitamin and mineral supplementation for preventing dementia or delaying cognitive decline in people with mild cognitive impairment." Cochrane Database Syst Rev, 2018. PubMed. DOI: 10.1002/14651858.CD011905.pub2.” pubmed.ncbi.nlm.nih.gov
14. Alzheimer's & Dementia. “Baker LD, Manson JE, Rapp SR, et al. "Effects of cocoa extract and a multivitamin on cognitive function: A randomized clinical trial." Alzheimers Dement, 2023 (online 2022). PubMed. DOI: 10.1002/alz.12767.” pubmed.ncbi.nlm.nih.gov
15. American Journal of Clinical Nutrition. “Yeung LK, Alschuler DM, Wall M, et al. "Multivitamin Supplementation Improves Memory in Older Adults: A Randomized Clinical Trial." Am J Clin Nutr, 2023. PubMed. DOI: 10.1016/j.ajcnut.2023.05.011.” pubmed.ncbi.nlm.nih.gov
16. Proceedings of the National Academy of Sciences. “Brickman AM, Yeung LK, Alschuler DM, et al. "Dietary flavanols restore hippocampal-dependent memory in older adults with lower diet quality and lower habitual flavanol consumption." Proc Natl Acad Sci U S A, 2023. PubMed. DOI: 10.1073/pnas.2216932120.” pubmed.ncbi.nlm.nih.gov
17. American Journal of Clinical Nutrition. “Vyas CM, Manson JE, Sesso HD, et al. "Effect of cocoa extract supplementation on cognitive function: results from the clinic subcohort of the COSMOS trial." Am J Clin Nutr, 2024 (online 2023). PubMed. DOI: 10.1016/j.ajcnut.2023.10.031.” pubmed.ncbi.nlm.nih.gov
18. American Journal of Clinical Nutrition. “Vyas CM, Manson JE, Sesso HD, et al. "Effect of multivitamin-mineral supplementation versus placebo on cognitive function: results from the clinic subcohort of the COcoa Supplement and Multivitamin Outcomes Study (COSMOS) randomized clinical trial and meta-analysis of 3 cognitive studies within COSMOS." Am J Clin Nutr, 2024. PubMed. DOI: 10.1016/j.ajcnut.2023.12.011.” pubmed.ncbi.nlm.nih.gov
19. Journal of Alzheimer's Disease. “Rist PM, Li S, Hamaya R, et al. "Effect of a multivitamin and cocoa flavanol extract on changes in neurofilament light, N-terminal tau fragment, and oligomeric amyloid-beta levels in COSMOS-clinic." J Alzheimers Dis, 2026. PubMed. DOI: 10.1177/13872877261489739.” pubmed.ncbi.nlm.nih.gov
20. Neurology. “Rafii MS, Walsh S, Little JT, et al. "A phase II trial of huperzine A in mild to moderate Alzheimer disease." Neurology, 2011. PubMed. DOI: 10.1212/WNL.0b013e318216eb7b.” pubmed.ncbi.nlm.nih.gov
21. Cochrane Database of Systematic Reviews. “Li J, Wu HM, Zhou RL, Liu GJ, Dong BR. "Huperzine A for Alzheimer's disease." Cochrane Database Syst Rev, 2008. PubMed. DOI: 10.1002/14651858.CD005592.pub2.” pubmed.ncbi.nlm.nih.gov
22. Cochrane Database of Systematic Reviews. “Yue J, Dong BR, Lin X, et al. "Huperzine A for mild cognitive impairment." Cochrane Database Syst Rev, 2012. PubMed. DOI: 10.1002/14651858.CD008827.pub2.” pubmed.ncbi.nlm.nih.gov
23. Journal of Ethnopharmacology. “Kongkeaw C, Dilokthornsakul P, Thanarangsarit P, Limpeanchob N, Norman Scholfield C. "Meta-analysis of randomized controlled trials on cognitive effects of Bacopa monnieri extract." J Ethnopharmacol, 2014 (online 2013). PubMed. DOI: 10.1016/j.jep.2013.11.008.” pubmed.ncbi.nlm.nih.gov
24. Phytotherapy Research. “Mori K, Inatomi S, Ouchi K, Azumi Y, Tuchida T. "Improving effects of the mushroom Yamabushitake (Hericium erinaceus) on mild cognitive impairment: a double-blind placebo-controlled clinical trial." Phytother Res, 2009. PubMed. DOI: 10.1002/ptr.2634.” pubmed.ncbi.nlm.nih.gov
25. Nutrients. “Docherty S, Doughty FL, Smith EF. "The Acute and Chronic Effects of Lion's Mane Mushroom Supplementation on Cognitive Function, Stress and Mood in Young Adults: A Double-Blind, Parallel Groups, Pilot Study." Nutrients, 2023. PubMed. DOI: 10.3390/nu15224842.” pubmed.ncbi.nlm.nih.gov
26. International Journal of Exercise Science. “Grozier CD, Alves VA, Killen LG, et al. "Four Weeks of Hericium erinaceus Supplementation Does Not Impact Markers of Metabolic Flexibility or Cognition." Int J Exerc Sci, 2022. PubMed. DOI: 10.70252/XZKO8571.” pubmed.ncbi.nlm.nih.gov
27. Frontiers in Nutrition. “Surendran G, Saye J, Binti Mohd Jalil S, et al. "Acute effects of a standardised extract of Hericium erinaceus (Lion's Mane mushroom) on cognition and mood in healthy younger adults: a double-blind randomised placebo-controlled study." Front Nutr, 2025. PubMed. DOI: 10.3389/fnut.2025.1405796.” pubmed.ncbi.nlm.nih.gov
28. Biomedical Research (Tokyo). “Saitsu Y, Nishide A, Kikushima K, Shimizu K, Ohnuki K. "Improvement of cognitive functions by oral intake of Hericium erinaceus." Biomed Res, 2019. PubMed. DOI: 10.2220/biomedres.40.125.” pubmed.ncbi.nlm.nih.gov
29. Frontiers in Aging Neuroscience. “Li IC, Chang HH, Lin CH, et al. "Prevention of Early Alzheimer's Disease by Erinacine A-Enriched Hericium erinaceus Mycelia Pilot Double-Blind Placebo-Controlled Study." Front Aging Neurosci, 2020. PubMed. DOI: 10.3389/fnagi.2020.00155.” pubmed.ncbi.nlm.nih.gov
30. Nutrients. “La Monica MB, Raub B, Ziegenfuss EJ, et al. "Acute Effects of Naturally Occurring Guayusa Tea and Nordic Lion's Mane Extracts on Cognitive Performance." Nutrients, 2023. PubMed. DOI: 10.3390/nu15245018.” pubmed.ncbi.nlm.nih.gov
31. Neurology: Clinical Practice. “Cohen PA, Avula B, Wang YH, Zakharevich I, Khan I. "Five Unapproved Drugs Found in Cognitive Enhancement Supplements." Neurol Clin Pract, 2021. PubMed. DOI: 10.1212/CPJ.0000000000000960.” pubmed.ncbi.nlm.nih.gov
32. Clinical Toxicology. “Cohen PA, Avula B, Khan I. "The unapproved drug centrophenoxine (meclofenoxate) in cognitive enhancement dietary supplements." Clin Toxicol (Phila), 2022. PubMed. DOI: 10.1080/15563650.2022.2109485.” pubmed.ncbi.nlm.nih.gov