Mushroom
Five human trials, the largest 41 people, and a mechanism whose named compounds failed the very test they are sold on.
Hericium erinaceus; Hericium erinaceum (older spelling, used in the 2003 case report); yamabushitake (Japanese); lion's mane mushroom.
Cognition and mood, dried fruiting-body powder: roughly 1 to 3 g a day. 1.8 g/day for 28 days in healthy adults aged 18-45 (Docherty 2023); about 2.88 g/day of powder (3 g/day of tablet mass at 96% powder) for 16 weeks in adults 50-80 with mild cognitive impairment (Mori 2009); a single 1 g dose in the only acute trial (La Monica 2023). Erinacine A-enriched mycelium, a different product and not interchangeable with the above: 1.05 g/day (three 350 mg capsules standardised to 5 mg erinacine A per gram), delivering about 5.25 mg of erinacine A a day, for 49 weeks in mild Alzheimer's disease (Li 2020). Upper end tried: 10 g/day for four weeks produced no effect on any cognitive or metabolic measure (Grozier 2022, as described in Docherty 2023). No trial has compared one dose against another, and the preparations differ by tissue type and marker compound, so this is the range that has been tried, not a range that has been optimised. There is no dose-response curve to draw.
None established. Neither the European Food Safety Authority nor the US National Academies has set a tolerable upper intake level for lion's mane. The only quantitative safety ceiling in the sourced literature is an animal one: a commercial lion's mane powder produced no subchronic toxicity in rats at doses up to 2000 mg per kg of body weight per day and no genotoxicity (Mahadevan 2025). That was the highest dose tested rather than a demonstrated limit, and a rat no-observed-adverse-effect level is not a human upper intake limit. No human dose-finding or toxicity study exists.
The only human measurements are acute cognitive tests. A single 1.8 g dose sped up one timed task an hour after swallowing it, and a single 1 g dose shifted reaction time at one and two hours. How fast the compounds themselves reach the bloodstream or the brain in a person is unknown, because no human pharmacokinetic study of lion's mane has been published.
One trial measured what happened after stopping. Sixteen weeks of daily fruiting-body powder raised a cognitive screening score in older adults with mild cognitive impairment, and four weeks after they stopped taking it the score had fallen significantly. No trial has measured how long a single dose lasts.
Whole fresh or dried mushroom — sold as food and cooked like any other edible mushroom; this is the form the fruiting-body research uses. Capsules of dried fruiting-body powder — commonly 500 to 600 mg a capsule; trials used 1 to 1.8 g a day. Erinacine A-enriched mycelium capsules — 350 mg a capsule standardised to 5 mg of erinacine A per gram, giving about 5.25 mg of erinacine A across three capsules a day. Blended mycelium and fruiting-body extract — one European trial product was 80 percent bulk mycelium and 20 percent fruiting-body extract, at 550 mg a day. Mycelium and fruiting body grown on oats — the form of a commercial US powder its own maker submitted for toxicology testing, and the reason a mushroom powder can be substantially grain. Memory and focus supplement blends — lion's mane was among the eighteen most common ingredients in a survey of 103 memory supplements sold in US pharmacies and grocery stores.
Lion's mane regrows your brain. It is the mushroom that switches on nerve growth factor, so it repairs nerves, rebuilds memory, clears brain fog and protects you against dementia. The hericenones and erinacines in it cross the blood-brain barrier and do the work.
A systematic review searching to June 2024 found five randomised controlled trials in humans, none analysing more than forty-one people. The most cited of them, in older adults with mild cognitive impairment, saw scores rise over sixteen weeks and fall again four weeks after intake stopped, and an independent expert panel reviewing it in 2024 reported that nobody had repeated it. Trials in healthy adults are split, with one finding a single timed task got faster and another finding nothing at all at five times the dose. The mechanism is where the claim frays first: hericenones C, D and E were tested individually for nerve growth factor and did not raise it, and the erinacines that do have brain data are a mycelium compound not detectable in fruiting-body tissue, measured crossing into the brain of a rat at several hundred times the per-kilogram dose any human trial has given.
The mushroom is sold on a mechanism it was tested for and failed. The compounds named on the label are the ones that did not raise nerve growth factor when tested alone; the ones that did reach a brain reached a rat's, and are made by the part of the fungus a fruiting-body capsule leaves out.
In 41 people with mild Alzheimer's, 49 weeks on 1.05 g a day beat placebo on a score for shopping and handling money, and on neither cognitive test.
Sixty-eight people were recruited and 41 finished, against the researchers' own target of 60. The company that makes the capsule funded the trial and pays five of the authors. Three of the four people who quit over side effects were taking lion's mane.
Mycelium grown in liquid culture made all four erinacines; six fruiting-body samples had none. They are the only lion's mane compounds measured entering a brain.
The mycelium here grew in stirred liquid, not on grain, so this settles which tissue makes erinacines and not how much of a grain-grown powder is mushroom. Only four erinacines were measured, and the assay's detection floor is not stated.
One US maker describes its own powder, in a peer-reviewed paper, as mycelium and fruiting body cultured on oats. How much leftover grain is in it goes unreported.
This paper set out to test safety, not to measure composition, so the oat description is an incidental disclosure rather than a finding. No independent laboratory has ever assayed a commercial lion's mane product and published the result.
A panel with nothing to sell reviewed lion's mane in 2024. It found plausible cell and animal work, and one very small human trial whose gain faded and nobody repeated.
This panel carries no new data of its own and was answering a narrow question about one disease, so its refusal to recommend lion's mane there is not a verdict on memory claims in general. What does travel is its flat report that the memory finding stands unrepeated.
Across 24 observational studies, diets containing mushrooms of any species tracked with better thinking and mood. The 10 trials in the same review came out mixed.
People who eat mushrooms differ from people who do not in diet quality, income and how much they move, and any of those could be doing the work. The observational half covers all mushroom species, so none of that signal belongs to lion's mane in particular.
A search of three databases from 2010 to 2024 returned 11 papers on mushrooms for nerve repair, five of them lion's mane. Every one was laboratory or animal work.
The review's verdict that lion's mane is the most promising of the four is a ranking inside eleven papers, partly earned by having the most of them and by being safer than fly agaric. Several of those five come from one research group, so they are less independent than the count suggests.
Trials have used 1 g once, 1.8 g daily for four weeks, about 2.9 g for 16 weeks, and 1.05 g of mycelium for 49 weeks. None was ever put against another.
These are not four points on one scale. Fruiting-body powder and mycelium are different tissue carrying different compounds, so a gram of one is not a gram of the other, and the range above is what has been tried rather than what has been optimised.
Lion's mane is among the 18 most common ingredients in 103 US memory supplements. Under US law the maker answers for safety and for the truth of the claims.
The same review sorts those eighteen ingredients by evidence and puts lion's mane in its best-supported group, but that sorting is one team's judgement with no stated inclusion rules and no bias assessment. Its own closing paragraph says the category lacks a solid scientific foundation.
Hot-water extracts of eight mushrooms were dripped onto human platelets. Jelly ear and king oyster cut clumping; lion's mane did not, and none changed clotting times.
Nobody swallowed anything: these were extracts applied to platelets in a tube, at concentrations matching no serving or capsule. The same paper's introduction cites an earlier report that a lion's mane extract did interfere with one platelet receptor, so even the tube evidence is split.
A commercial lion's mane powder passed guideline animal toxicology at up to 2000 mg per kilogram of body weight a day, with no DNA damage. No human measure was taken.
2000mg per kilogram was the highest dose tried, not a limit anyone established, and a rat no-harm level is not a human intake limit. No human safety measure was taken. The company selling the powder ran and wrote it. An expert panel review records one anaphylactic reaction to fresh lion's mane.
In 41 healthy adults aged 18 to 45, one 1.8 g dose sped a colour-word task and raised errors on it. Over 28 days, only the placebo group's word recall improved.
The 28-day stress result that gets quoted came in at p=0.051, just outside the usual cutoff for calling a result real. The authors call this a pilot too small to conclude from. What was in the capsule was described only as a proprietary mushroom complex.
In 40 healthy adults, one 1 g dose sped reaction time by 9.4 milliseconds over their own score two hours earlier. No comparison put it ahead of placebo.
Placebo improved on its own baseline too, which is what within-arm comparisons look like when practice and expectation are doing the work. Nothing here runs longer than two hours. The sponsor sells the tea extract used as the comparator.
Rats given erinacine A had it in the brain within an hour. Their 50 mg per kilogram of body weight was 670 times what the one human trial worked out to.
Nobody has measured what lion's mane does in a human bloodstream, in any form, so none of this is confirmed in a person. Erinacines are made by the mycelium, so a fruiting-body capsule is not carrying the compound this study followed.
A review of everything published between 2000 and June 2024 found 26 studies on lion's mane. Five were randomised trials in people; more than half was cell or animal work.
The search covered PubMed alone, so anything indexed only elsewhere or published in another language is missing, and nothing after June 2024 is included. The review's own design labels are wrong in places: it files one placebo-controlled trial as a cohort study.
Lion's mane's cognitive benefits come from small, short, often maker-funded trials, while independent trials in healthy adults found nothing against placebo. It is less a dispute than too little evidence to have one.
Chaga grown on grain separated cleanly from wild chaga on every chemical fingerprint, and came out starch-rich. Grain is the same route used across mushroom supplements.
No lion's mane product was tested here; this is a different fungus and none of these numbers transfers. The analysts work for a supplier selling fruiting-body extracts in competition with grain-grown mycelium ones, which is a direct commercial interest in this result.
Two lion's mane compounds made nerve cells sprout long, branching fibres, and mice fed them did better at memory tasks. The route was not the one named on labels.
No person received either compound. The doses that worked on cultured cells and in mice have no translation into a capsule, and blocking that receptor only partly stopped the effect, so even the pathway is unsettled.
Whole lion's mane extract raised nerve growth factor in cultured human brain cells. Hericenones C, D and E, each tested on its own, raised none of it.
The cells were a human tumour line, not ordinary brain tissue, and the mouse arm fed animals 5% lion's mane powder for a week, a far larger share of the diet than anyone eats. Nerve growth factor messenger levels are a molecular reading, not a memory test.
Three research groups each point to a different active compound in lion's mane, and the one printed on most labels failed when tested alone. None has been tested in a person, so what does the work is still open.
Nineteen batches of reishi bought in the United States were assayed for the compounds their labels claimed. Nothing equivalent has ever been published for lion's mane.
This is reishi, a different fungus, and no lion's mane product was in the sample. Nineteen batches is small, the sampling method is undescribed, and 'matched the label' means matched by these analysts' own chemical criteria, which another method would score differently. The data predate 2017.
In one trial, 15 people took about 2.9 g of lion's mane daily for 16 weeks and scored higher on a dementia screen. Four weeks after stopping, the gain had faded.
The lead author works in a mushroom producer's own laboratory, and an independent expert panel reviewing the trial in 2024 reported that nobody had repeated it. The outcome was a brief dementia screening questionnaire, not a full memory assessment.
No findings tagged to that goal.
What the evidence says about combining Lion's Mane (Hericium erinaceus) with each of these. The marker is how well established the interaction is; a caution is about the hazard, which is a different question.
8 open questions
Search runs entirely in your browser. Nothing you type is sent anywhere.