Lion's Mane

Lion's Mane

Hericium erinaceus is a real fungus with a real chemistry: hericenones in the fruiting body, erinacines in the mycelium, and a rodent nerve-growth-factor literature that the market has renamed neurogenesis. The human file is a handful of small trials, non-interchangeable products, and outcomes that do not survive that renaming.

Compiled by South Beach Longevity · 20 August 2026 Copyright 2026 Series SBL-41 / SP-LIONS-MANE · Register A scientific article Sources peer-reviewed human trials, chemistry papers, labelled animal or in-vitro work, and product-quality analytics · verified NCBI records Constraint This document describes published research. It is not medical advice. No human use, dose, route or schedule is recommended anywhere in this document.

How to read this document Every finding is labelled, in the sentence that reports it, by the kind of study that produced it. A fruiting-body powder is not an erinacine-standardised mycelium. A mouse hippocampal-neurogenesis assay is not a human memory trial. A 15-versus-15 Japanese MCI study is not a dementia programme. Where two results conflict, both are given. Amounts and durations appear only as reported experimental parameters, always with the population attached. Nothing here is a recommendation. Findings are graded in place as established, strongly supported, emerging, plausible, or speculative. Two further labels mark careful absences rather than verdicts: not established, where the evidence is too thin to place a claim on the ladder at all — untested or insufficient, an absence of proof rather than disproof; and not supported, where the weight of evidence leans against a claim but stops short of a formal refutation.

01 What this document is, and four things it is not

This article is a research review of lion's mane — Hericium erinaceus (Bull.) Pers., also filed in older papers as Hericium erinaceum, and sold under Yamabushitake, houtou, hedgehog fungus, bearded tooth, and pom-pom. It covers fungal taxonomy, the fruiting-body versus mycelium split, hericenones, erinacines, nerve-growth-factor (NGF) hypotheses, extraction, and central-nervous-system exposure. It then reads the actual human file for mild cognitive impairment, cognition, mood, anxiety and depression, sleep, and neurodegeneration. It ends on product quality: grain-grown mycelium, beta-glucan measurement, starch, marker standardisation, and species authentication.

Four things follow immediately.

First, this is not a neurogenesis product. Ryu, Hong, Kang, and colleagues reported that a H. erinaceus extract reduced anxiety- and depression-like behaviour and increased hippocampal neurogenesis markers in adult mice (Ryu et al., 2018). That is a rodent behavioural and histology paper. No human trial in this file measured adult hippocampal neurogenesis. Treating the mouse finding as a label claim is the most expensive error in the aisle.

Second, it is not proof that fruiting body and mycelium are the same exposure. Kawagishi's group isolated hericenones from the fruiting body and erinacines from cultured mycelia as two NGF-stimulant classes (reviewed in Friedman, 2015; Li et al., 2018). Li and colleagues stated the operational split directly: erinacine A "exists only in fermented mycelium but not in its fruiting body" (Li et al., 2018). A grain-grown "100% mycelium" powder is not Mori's Yamabushitake tablets, and Mori's tablets are not an erinacine A capsule.

Third, it is not a large human evidence base. The load-bearing human studies can be counted on two hands. Mori, Inatomi, Ouchi, Azumi, and Tuchida randomised 30 Japanese adults with mild cognitive impairment (Mori et al., 2009). Nagano and colleagues randomised 30 women to four weeks of cookies (Nagano et al., 2010). Saitsu and colleagues tested fruiting-body capsules for 12 weeks (Saitsu et al., 2019). Li and colleagues' 49-week mild Alzheimer's pilot analysed 41 completers of an erinacine A–enriched mycelium (Li et al., 2020). Docherty, Doughty, and Smith tested 1.8 g in 41 healthy adults aged 18–45 years (Docherty, Doughty, and Smith, 2023). Grozier and colleagues found no cognitive or metabolic-flexibility signal after 10 g/day for four weeks in a college-age sample (Grozier et al., 2022). Surendran and colleagues found no global acute cognitive or mood effect of a fruiting-body extract in 18 healthy younger adults (Surendran et al., 2025). That is the file. It is not a pipeline.

Fourth, it is not medical advice and it is not a dosing guide. Amounts appear only as study facts. This document recommends no product, extract, capsule, cookie, muffin, tea, or schedule for any person.

The working thesis is narrower than the aisle. H. erinaceus has a split chemistry that is established. The NGF hypothesis is strongly supported in cell and rodent systems and not established as a human mechanism. The human cognitive signal is emerging in one small MCI fruiting-body trial and one industry-adjacent mild-AD mycelial pilot, and not supported as a general nootropic. Mood and sleep claims are emerging at best and confounded where diet or mixed products sit underneath them. Neurogenesis marketing is speculative.

02 Fungal taxonomy

Hericium erinaceus is a wood-decay basidiomycete in the family Hericiaceae. It fruits as a white, hanging mass of spines on dead or dying hardwood — the object that earned the lion's-mane, bearded-tooth, and pom-pom names. Older literature uses Hericium erinaceum; the names travel together in PubMed and should not be treated as two species. East Asian culinary and Kampo/TCM use of Yamabushitake and houtou is historical context, not a clinical indication (Friedman, 2015; Phan, David, Naidu, Wong, and Sabaratnam, 2015). Grade for identity of the wild fungus: established.

Related Hericium species — including H. americanum and H. coralloides — share the genus and not automatically the hericenone or erinacine profile. A DNA-barcode that says "Hericium" is not a certificate that the bottle matches Mori's powder or Li's mycelium. Section 23 returns to authentication. Grade for species-level interchangeability: not supported.

The life cycle is the product-quality problem in miniature. The fruiting body is the mushroom. The mycelium is the vegetative network. In industrial practice the mycelium is often grown on sterilised grain. The harvested material is then mycelium plus residual grain. Calling that powder "100% mushroom" is a labelling move, not a chemical one. Section 03 and Part Four keep that object in view.

03 Fruiting body versus mycelium

The two tissues are not two brand stories. They are two chemistries.

Hericenones are aromatic compounds isolated from fruiting bodies. Erinacines are cyathane diterpenoids isolated from mycelia. Friedman treated that partition as the load-bearing fact in a Journal of Agricultural and Food Chemistry review of fruiting bodies, mycelia, and their bioactive compounds (Friedman, 2015). Li, Lee, Chen, Chou, and Wang restated it for the erinacine literature: the NGF-relevant diterpenoid class is a mycelial product (Li et al., 2018). Li, Chang, Lin and colleagues used the same split to justify testing an erinacine A–enriched mycelium rather than a fruiting-body powder in mild Alzheimer's disease (Li et al., 2020). Grade: established as a chemical geography; strongly supported as a reason the human trials are not pooled.

Both tissues also make polysaccharides, sterols, and ordinary fungal biomass. Those shared fractions do not collapse the hericenone/erinacine split. A "full-spectrum" label that mixes an unspecified fruiting-body extract with grain mycelium is a third object. Vigna and colleagues used exactly that third object — 80% bulk mycelia and 20% fruiting-body extract in a commercial capsule — in their overweight mood-and-sleep study (Vigna et al., 2019). That trial cannot be spent as a fruiting-body result or as an erinacine A result.

Nagano and colleagues wrote, in the opening of a human cookie trial, that "hericenones and erinacines isolated from its fruiting body stimulate nerve growth factor (NGF) synthesis" (Nagano et al., 2010). The sentence is a category error. Erinacines were isolated from mycelia (Li et al., 2018; Friedman, 2015). The trial remains a trial. The mechanism sentence is not used as a map.

04 Hericenones

Hericenones are meroterpenoid aromatics. Kawagishi and colleagues isolated hericenones C, D, and E from the fruiting body as stimulators of NGF synthesis in astroglial cell systems; later isolations added hericenones F–J and related structures (discovery sequence as reviewed by Friedman, 2015, and by Ma, Shen, Yu, Li, Zhang, and Dai, 2010, as cited in later NGF reviews). Phan, David, Naidu, Wong, and Sabaratnam placed those compounds inside a broader culinary-medicinal mushroom survey (Phan et al., 2015). Lai and colleagues isolated hericenones from fruiting bodies cultivated under tropical conditions and reported NGF-mediated neurite outgrowth in PC12 cells through MEK/ERK and PI3K-Akt pathways (Lai et al., 2014). Grade for existence and fruiting-body origin: established. Grade for NGF activity in cell systems: strongly supported. Grade for a measured human hericenone exposure after an oral powder: not established.

Li and colleagues recorded a useful negative inside the mycelial literature: when the constituents of Mori's fruiting-body effect were examined, "hericenones failed to stimulate NGF gene expression in primary cultured rat astrocytes" in at least one follow-up, which is why that group turned to erinacine-enriched mycelium (Li et al., 2018). The sentence is not a demolition of Kawagishi. It is a reason not to treat "hericenones = NGF = Mori" as a closed chain. Mori measured a cognitive scale, not hericenones in plasma (Mori et al., 2009).

05 Erinacines

Erinacines are cyathane diterpenoids from cultured mycelia. Kawagishi's 1994 isolation of erinacines A, B, and C as strong NGF-synthesis stimulators is the foundation paper of that class (Kawagishi et al., 1994, as cited with DOI in Li et al., 2018, and Docherty, Doughty, and Smith, 2023). Later papers added erinacines E–I and others (Lee et al., 2000, as cited in Li et al., 2020). Li and colleagues' 2018 review is the most useful single map of the mycelial file — and it is a Grape King Bio author paper. That affiliation travels with the citation (Li et al., 2018).

Erinacine A is the marker that made a human trial possible. The Li 2020 mild-AD capsules were specified as 350 mg mycelia containing 5 mg/g erinacine A, three capsules per day, for 49 weeks (Li et al., 2020). That is a rare act of standardisation. It is also why that trial cannot be read as evidence for an uncharacterised fruiting-body powder. Grade for mycelial origin of erinacines: established. Grade for erinacine A as the only human-standardised marker in this file: strongly supported. Grade for commercial bottles matching 5 mg/g: not established unless the label and a certificate say so, and most do not.

06 Polysaccharides and the rest of the mushroom

H. erinaceus fruiting bodies and mycelia also yield beta-glucans, other polysaccharides, sterols, fatty acids, and lectins (Friedman, 2015; He, Wang, Zhang, Zhang, and Sun, 2017). He and colleagues reviewed extraction, structure, and claimed biological activities of polysaccharides from fruiting body, mycelium, and culture broth (He et al., 2017). Those fractions are real. They are also the easiest place for a grain-grown mycelial powder to hide starch. Phenol-sulfuric "total polysaccharides" counts glucose from leftover grain. McCleary and Draga published an enzymatic method specifically to measure beta-glucan in mushrooms and mycelial products and to separate it from alpha-glucan (McCleary and Draga, 2016). Section 21 is the quality home of that method. Grade for the existence of fungal polysaccharides: established. Grade for "high polysaccharides" on a grain-mycelium label as a beta-glucan claim: not supported without an enzymatic split.

07 Extraction

Extraction is not a single verb. Hot-water extracts concentrate polysaccharides. Ethanol or hydroalcoholic extracts pull hericenones and other lower-polarity fruiting-body metabolites. Mycelial erinacines are typically recovered from fermented mycelium with organic solvents in the chemistry papers, then formulated as biomass capsules in the human EAHE trial (Li et al., 2018; Li et al., 2020). Surendran and colleagues used a commercial fruiting-body extract described as 95% H. erinaceus extract and 5% maltodextrin, given as 3 g of a 10:1 extract, produced by water and ethanol processing (Surendran et al., 2025). Vigna's fruiting-body portion was an extract on top of bulk mycelia (Vigna et al., 2019). Mori used 96% Yamabushitake dry powder in tablets, not an extract (Mori et al., 2009).

Those four objects are not interchangeable. A 10:1 extract is a concentration claim, not a hericenone assay. A dry powder is a biomass claim. An erinacine A–enriched mycelium is a marker claim. Collapsing them into "lion's mane extract" is how the aisle writes a meta-analysis that the laboratory cannot run. Grade: established that extraction route changes the object; speculative that any one route is "the" nootropic form.

08 NGF-related hypotheses

The controlling mechanistic claim is that hericenones and erinacines increase NGF synthesis, and that NGF then supports neurons. Mori, Obara, Hirota and colleagues reported NGF-inducing activity of H. erinaceus in 1321N1 human astrocytoma cells (Mori et al., 2008). Lai and colleagues reported hericenone-linked neurite outgrowth in PC12 cells (Lai et al., 2014). The erinacine cell and rodent file is larger: NGF stimulation, ischaemia models, MPTP parkinsonism models, and APP/PS1 Alzheimer's mice (Li et al., 2018; Tsai-Teng et al., 2016; Kuo et al., 2016, as cited in Li et al., 2020). Grade for in-vitro NGF induction by named compounds: strongly supported. Grade for in-vivo rodent neurotrophic and lesion-model effects of erinacine A or EAHE mycelium: strongly supported as animal pharmacology. Grade for demonstrated NGF induction in a living human brain after an oral commercial product: not established.

NGF itself does not make the leap easy. Endogenous NGF is a large protein with poor oral bioavailability and limited blood–brain-barrier passage as a drug. The Hericium hypothesis is not that people absorb NGF from a mushroom. It is that small molecules induce NGF in situ. That is a coherent hypothesis. It is also unmeasured in the human trials. Mori, Nagano, Saitsu, Docherty, Grozier, and Surendran did not report CSF NGF. Li 2020 reported circulating BDNF among other biomarkers, not central NGF (Li et al., 2020). Vigna reported circulating pro-BDNF and BDNF, not NGF (Vigna et al., 2019). The hypothesis is available. The human chain is not.

09 BDNF and other neurotrophins

BDNF entered the file because it is easier to draw from blood than NGF is to draw from brain. Vigna and colleagues found an increase in circulating pro-BDNF after eight weeks of a mixed mycelium/fruiting-body capsule, without a significant change in mature BDNF (Vigna et al., 2019). Li and colleagues reported that BDNF fell in the placebo arm of the mild-AD mycelium trial and did not show the same fall in the EAHE arm, alongside other biomarker shifts (Li et al., 2020). Chiu, Lee, Chen and colleagues reported antidepressant-like effects of erinacine A–enriched mycelium in mice with BDNF/PI3K/Akt/GSK-3β signalling as the proposed path (Chiu et al., 2018). Grade for circulating BDNF as a human Hericium endpoint: emerging and not a brain biopsy. Grade for BDNF as the mechanism of any cognitive scale: plausible in animals, speculative in the human trials.

Chong, Phan, David and colleagues reviewed H. erinaceus and depressive disorder largely from preclinical work plus the small Nagano and Vigna human studies (Chong, Phan, Sabaratnam, David, and Naidu, 2020). A review can organise a hypothesis. It cannot enlarge n=30.

10 Neurogenesis marketing

"Neurogenesis" on a lion's-mane label is almost always a borrowing of Ryu et al., 2018, or of neurite-outgrowth photographs from PC12 or primary-neuron dishes (Lai et al., 2014; Phan et al., 2013). Ryu's paper measured behaviour and hippocampal markers in mice after an extract (Ryu et al., 2018). Neurite outgrowth in a dish is not the birth of new dentate-gyrus neurons in an adult human. ALSUntangled, reviewing the species for amyotrophic lateral sclerosis, stated the human constraint without ornament: a very small double-blind MCI trial showed a temporary cognitive improvement that has yet to be replicated, and there are no ALS studies (Muhanna et al., 2024). Grade for mouse hippocampal neurogenesis after some extracts: emerging. Grade for human neurogenesis as a product claim: not supported.

The marketing move has a second fault. It treats NGF induction, neurite outgrowth, neuroprotection, and neurogenesis as one word. They are not. A cell that sprouts a neurite is not a new neuron. A rodent that explores an open field longer is not a person who grew a hippocampus. Brandalise, Roda, Ratto and colleagues, in a 2023 Journal of Fungi review, collected the bench file and still called for wider clinical trials rather than declaring the shoreline reached (Brandalise et al., 2023). That is the correct distance.

11 CNS exposure

Does an oral powder put hericenones or erinacines into the human central nervous system? The honest answer is that the human measurement is missing.

Hu, Li, Lin and colleagues reported absolute bioavailability, tissue distribution, and excretion of erinacine S after administration in a mycelial context — a pharmacokinetic paper, not a cognition trial (Hu et al., 2019). Chen, Lin, Li and colleagues gave 5 mg/kg erinacine A intravenously to a single landrace pig, detected the compound in cerebrospinal fluid at 15 minutes, saw the highest CSF concentration at 30 minutes (5.26 ± 0.58 μg/L), and measured 77.45 ± 0.58 μg/L in brain tissue (Chen et al., 2024). Six metabolites, including erinacine B, appeared in plasma and brain. That is emerging evidence that erinacine A can reach mammalian CSF and brain after intravenous dosing in one pig. It is not oral human pharmacokinetics. It is not hericenone pharmacokinetics. Li and colleagues' 2021 sleep-disturbance paper asserted that "a large body of evidence indicates that erinacine can pass the blood-brain barrier" in the course of a mouse study (Li et al., 2021). Assertions of that kind are not a substitute for a human oral CSF series.

Fruiting-body hericenones have a thinner exposure file. No verified human CSF hericenone series sits in the reviewed record. Surendran and colleagues closed their acute fruiting-body study by asking for the time to peak concentration of H. erinaceus bioactives in the human brain — which is an admission that they did not have it (Surendran et al., 2025). Grade for IV erinacine A in one pig: emerging. Grade for oral human CNS exposure of either class: not established.

12 Why rodent NGF findings do not become human claims

The generalisation has four breaks.

The compound is often different. Mouse EAHE or isolated erinacine A is not Mori's fruiting-body powder and is not a grain-mycelium starch blend (Li et al., 2018; Mori et al., 2009).

The endpoint is different. NGF mRNA in astrocytes, neurite length in PC12 cells, Morris-water-maze latency, and a Revised Hasegawa Dementia Scale derivative are four objects (Mori et al., 2008; Lai et al., 2014; Tsai-Teng et al., 2016; Mori et al., 2009).

The exposure is different. Oral grams of biomass in a person are not milligrams per kilogram of a purified diterpenoid in a mouse, and they are not 5 mg/kg intravenous erinacine A in a pig (Chen et al., 2024).

The sample is different. APP/PS1 mice, MPTP mice, sleep-disturbed mice, and 15 Japanese adults with MCI do not share a likelihood ratio (Tsai-Teng et al., 2016; Li et al., 2021; Mori et al., 2009).

Phan and colleagues already wrote the restraint into a 2015 review: the mushroom file is mostly in vitro, and "clinical trials with humans are needed" (Phan et al., 2015). The trials that arrived later did not close the restraint. They populated it with small n. Figure 2 is the map of those four breaks.

13 How small the human file is

ALSUntangled's summary is still the right size: one very small double-blind MCI trial with a temporary improvement, unreplicated as a programme, and no ALS data (Muhanna et al., 2024). Menon, Jalal, Arshad, Nawaz, and Kashyap's 2025 systematic review gathered five randomised trials, three pilot clinical trials, and a thin remainder, then reported a combined weighted mean Mini-Mental State Examination increase of 1.17 points from one RCT and one pilot — a number that should be read as "the reviewers had almost nothing to pool," not as a treatment effect (Menon et al., 2025). Brandalise and colleagues likewise counted a limited clinical set against a large preclinical bulk (Brandalise et al., 2023).

Tiny trials are not a footnote. They are the design. A 15-person arm cannot carry a neurodegenerative indication. A 18-person acute crossover cannot carry a nootropic category. A college muffin study that goes null is more informative than a narrative review that goes loud (Grozier et al., 2022). The rest of Part Three is the file, not a highlight reel.

14 Mild cognitive impairment

Mori, Inatomi, Ouchi, Azumi, and Tuchida performed a double-blind, parallel-group, placebo-controlled trial in 50- to 80-year-old Japanese men and women with mild cognitive impairment (Mori et al., 2009). After two weeks of preliminary examination, 30 subjects were randomised to two 15-person groups. The Yamabushitake group took four 250 mg tablets containing 96% Yamabushitake dry powder three times a day for 16 weeks — a study fact, not a recommendation — and was then observed for four weeks after stopping. At weeks 8, 12, and 16 the Yamabushitake group showed significantly increased scores on a cognitive function scale based on the Revised Hasegawa Dementia Scale compared with placebo. Scores rose with duration of intake and fell significantly at week 4 after termination. Laboratory tests showed no adverse effect in that sample. Grade: emerging as a small, apparently reversible scale signal in Japanese MCI on a fruiting-body dry powder. Not established as a treatment for MCI. Not a licence to treat the same powder as an erinacine drug.

What the paper does not show matters as much as what it does. There is no hericenone or erinacine blood level. There is no MRI. There is no replication with the same product in an independent sample of comparable size. Saitsu later reported an MMSE movement on fruiting-body capsules in a different, not-MCI-restricted design (Saitsu et al., 2019). That is not a replication of Mori. ALSUntangled was right to call the 2009 result unreplicated (Muhanna et al., 2024).

15 Cognition outside MCI

Saitsu, Nishide, Kikushima, Shimizu, and Ohnuki ran a randomised, double-blind, placebo-controlled, parallel-group, 12-week fruiting-body study with Mini-Mental State Examination, Benton visual retention, and a standard verbal paired-associate learning test (Saitsu et al., 2019). MMSE alone showed a significant improvement and a claim of prevented deterioration. Benton and S-PA are not presented as the same success. The authors speculate that hericenones and other compounds act on neural networks, and they close with "dementia prevention" language the design cannot carry. Grade: emerging for an MMSE shift on that fruiting-body product; not supported as dementia prevention; insufficient as a multi-test cognitive claim.

Docherty, Doughty, and Smith tested acute (60 minutes) and 28-day effects of 1.8 g H. erinaceus — study fact — in 41 healthy adults aged 18–45 years, double-blind, placebo-controlled, parallel groups (Docherty, Doughty, and Smith, 2023). After a single dose, Stroop performance was faster (p = 0.005). After 28 days, subjective stress showed a trend (p = 0.051). The authors report null and limited negative findings and warn on sample size. Sempera Organics provided the investigational product and had no role in design or analysis; the authors declare no other conflict. The product is not an erinacine A–assayed mycelium in the published methods summary. Grade: emerging as a single-task acute speed signal in healthy young adults; not supported as a general cognitive enhancer; insufficient as chronic cognition.

Grozier, Alves, Killen, Simpson, O'Neal, and Waldman gave 10 g/day H. erinaceus or placebo for four weeks as two identical muffins in a single-blind, parallel study of 24 college-age adults (Grozier et al., 2022). Graded exercise and dual-task Stroop plus mental arithmetic, before and after fatigue, showed no significant interaction or main effect on metabolic flexibility or cognition (all p > 0.05). Grade: strongly supported as a null in that young, high-intake, muffin design. A null of this kind is not "failed replication of Mori." It is a different population, product, and endpoint. It still blocks the "works for everyone" sentence.

La Monica and colleagues compared acute 1 g Nordic-grown lion's mane, 650 mg guayusa extract, and placebo in a randomised, double-blind, crossover design with Go/No-go, N-back, and Serial 7s (La Monica et al., 2023). Lion's mane improved some reaction-time and working-memory indices. Guayusa, a caffeinated holly, improved more indices and raised blood pressure. The design is useful as a reminder that an acute "focus" bottle is often a caffeine trial in costume. Grade for isolated lion's mane as an acute nootropic in this crossover: emerging / task-specific. Grade for reading the paper as a caffeine-free victory: not supported.

Surendran and colleagues gave 3 g of a 10:1 fruiting-body extract or placebo to 18 healthy adults aged 18–35 years in an acute double-blind crossover, with testing at 90 minutes (Surendran et al., 2025). Composite global cognition and mood did not differ. The pegboard test improved. Two authors were employed by a company that uses H. erinaceus in drinks; that affiliation is in the conflict block. Grade: not supported as an acute global cognitive or mood effect; emerging as a possible task-specific motor/psychomotor blip that needs a design built for it.

Taken together, cognition outside MCI is a scatter of small, product-heterogeneous studies with one important null and several task-level positives that do not survive compositing. The aisle's "clinically studied for focus" sentence is spending Mori.

16 Mood, anxiety, and depression

Nagano, Shimizu, Kondo and colleagues randomised 30 women to four weeks of H. erinaceus cookies or placebo cookies and scored the Kupperman Menopausal Index, CES-D, Pittsburgh Sleep Quality Index, and an Indefinite Complaints Index (Nagano et al., 2010). CES-D and ICI fell after HE relative to that group's own baseline. Versus placebo, two ICI items ("insentive" and "palpitatio" in the published wording) were significantly lower; concentration, irritation, and anxiety only tended to be lower. The authors themselves suggest a mechanism different from NGF-enhancing action. Grade: emerging for a short, small, menopausal-adjacent mood signal on cookies; not established as an antidepressant; not an NGF demonstration.

Vigna, Morelli, Agnelli and colleagues recruited 77 adults with overweight or obesity and a mood, sleep, or binge-eating complaint, all under a low-calorie diet, and compared diet alone with diet plus eight weeks of "Micotherapy Hericium" — 80% bulk mycelia and 20% fruiting-body extract, three capsules/day as labelled in the paper (Vigna et al., 2019). Depression scores did not differ significantly between groups. Anxiety improved about 12.6% after the supplement. Binge-eating scores fell in both arms — the diet is doing work. Circulating pro-BDNF rose; mature BDNF did not. Drop-outs clustered in the mushroom arm (three at T1, five at T2) and none in control. Grade: emerging for anxiety on a mixed product plus diet; not supported as an isolated antidepressant effect; confounded for binge eating; not a fruiting-body trial and not an erinacine A trial.

Docherty's 28-day stress trend (p = 0.051) is a trend, recorded as such (Docherty, Doughty, and Smith, 2023). Surendran's acute mood composites did not move (Surendran et al., 2025). Chong's 2020 review is a preclinical-heavy narrative (Chong et al., 2020). Grade for lion's mane as a treatment of major depression or anxiety disorders: insufficient.

17 Sleep

Nagano included PSQI and did not make sleep the headline success versus placebo (Nagano et al., 2010). Vigna reported improved nocturnal rest on the mixed product plus diet (Vigna et al., 2019). Li, Lee, Lo and colleagues — a Grape King–affiliated mouse study — found that 150 mg/kg mycelium ameliorated tail-suspension-test-induced anxiety-like behaviour and reversed a dark-period NREM disturbance in mice (Li et al., 2021). That paper is not a human sleep trial. Its abstract also repeats the fruiting-body/mycelium hericenone-and-erinacine muddle. Grade for human sleep: emerging at best, confounded by diet and mixed products, not established. Grade for mouse sleep-architecture rescue by mycelium: emerging as animal data.

18 Neurodegeneration

Li, Chang, Lin and colleagues conducted the longest human Hericium trial in this file: a 49-week, double-blind, placebo-controlled pilot of erinacine A–enriched mycelia in mild Alzheimer's disease, registered as NCT04065061 (Li et al., 2020). Three 350 mg capsules/day at 5 mg/g erinacine A is a study fact. Forty-nine participants were randomised; 41 completers (17 men, 24 women) were analysed. The placebo arm lost Cognitive Abilities Screening Instrument score; the EAHE arm improved Mini-Mental State Examination; Instrumental Activities of Daily Living differed between groups; contrast sensitivity favoured EAHE; several blood biomarkers and some diffusion-tensor metrics moved. Four subjects dropped for abdominal discomfort, nausea, or skin rash. Grape King Bio Inc. paid salaries for five authors and provided research materials. The conflict statement says the company had no additional role in design or analysis. That disclosure stays in the sentence, not a footnote.

Grade: emerging as a single-site, industry-adjacent, modest-n pilot of a characterised mycelial product in mild AD. Not established as disease modification. Not evidence for fruiting-body powders. Not a replication of Mori, because the product and the disease stage are different. A larger, independently sponsored trial would be the falsifier. It is not in this file.

Mori, Obara, Hirota and colleagues had already shown fruiting-body effects on amyloid-β(25–35)–induced learning and memory deficits in mice (Mori et al., 2011). Tsai-Teng, Chin-Chu, Li-Ya and colleagues showed that erinacine A–enriched mycelia attenuated cerebral Aβ plaque burden in APPswe/PS1dE9 mice after 30 days (Tsai-Teng et al., 2016). Those papers explain why Li 2020 was run. They do not enlarge Li 2020.

ALSUntangled found no ALS cell, animal, or human studies and did not support use in ALS (Muhanna et al., 2024). Grade for ALS: insufficient. Grade for Parkinson's disease in humans: insufficient. The MPTP mouse file is preclinical (Kuo et al., 2016, as cited in Li et al., 2020).

19 Safety

LiverTox (NIDDK), updated 5 January 2024, classifies lion's mane as an unlikely cause of clinically apparent liver injury (likelihood score E). The chapter reports no convincing hepatotoxicity signal in the small human trials and notes typical gastrointestinal complaints — abdominal discomfort, nausea, or diarrhoea — usually in fewer than 10% of treated subjects in those trials (NIDDK LiverTox, 2024). One acute hypersensitivity reaction after oral lion's mane is recorded there. ALSUntangled notes an anaphylactic case after consumption of fresh mushroom (Muhanna et al., 2024). Li 2020 recorded drop-outs for abdominal discomfort, nausea, and skin rash (Li et al., 2020). Vigna's drop-outs sat in the supplement arm (Vigna et al., 2019). Mori reported no laboratory adverse effect in 16 weeks (Mori et al., 2009).

Rodent toxicology of erinacine A–enriched mycelium — 28-day and 13-week oral feeding, genotoxicity, and high-dose acute work — is a Grape King–adjacent file and is not a human safety database (Li et al., 2014a; Li et al., 2014b; Lee et al., 2019, as cited in Li et al., 2018, and Li et al., 2020). Grade for short-term tolerability of studied oral biomasses in small adult samples: strongly supported as "generally well tolerated, with GI and occasional rash/hypersensitivity." Grade for long-term, high-intake, multi-product, or diseased-liver safety: insufficient. Grade for hepatotoxicity as a class effect: not supported by LiverTox. This is not a licence for any person.

Hericenone B has been reported to inhibit collagen-induced platelet aggregation in a chemistry/pharmacology paper (Mori et al. isolation literature as extended by later hericenone B work; see Friedman, 2015, for the compound class). That is not a clinical bleeding trial. It is a reason not to invent a "blood-thinning" claim or to ignore a theoretical interaction in people already on antiplatelet drugs — and also a reason not to scare healthy readers with a dish finding.

20 Fruiting body versus grain-grown mycelium

A fruiting body grown on wood or supplemented sawdust is mostly fungus. A mycelium grown on grain is fungus plus grain unless the grain is removed and the remainder is assayed. McCleary and Draga built an enzymatic beta-glucan method for mushrooms and mycelial products because "mycelial products" were already a commercial category that needed a starch-aware assay (McCleary and Draga, 2016). Friedman distinguished fruiting-body and mycelial chemistries rather than treating biomass as one powder (Friedman, 2015). Li and colleagues stated that erinacine A is a fermented-mycelium compound, not a fruiting-body compound (Li et al., 2018).

The quality failure is therefore specific. A bottle labelled "100% lion's mane mycelium" grown on rice or oats can be mostly cereal starch by weight, with a fungal film. Alpha-glucan will be high. Beta-glucan will be low. Erinacine A may be present or absent depending on the strain and fermentation, and is almost never reported. A fruiting-body powder can be low in erinacines by construction and still be an honest mushroom. The two failures are different. The aisle uses one word for both.

Vigna's 80/20 mix is at least labelled as a mix (Vigna et al., 2019). Many retail powders are not. Grade for the chemical non-identity of fruiting body and mycelium: established. Grade for grain-grown mycelium as an equivalent of Mori's tablets: not supported.

21 Beta-glucan versus starch

McCleary and Draga (2016) is the load-bearing analytical paper. It measures beta-glucan enzymatically in mushrooms and mycelial products and distinguishes alpha-glucan. Phenol-sulfuric acid "total glucans" or "polysaccharides ≥50%" on a grain-mycelium label can be a starch test. He and colleagues' polysaccharide review is useful for structure and extraction and is not a substitute for that split (He et al., 2017). Grade for enzymatic beta-glucan minus alpha-glucan as the minimum polysaccharide honesty test: strongly supported. Grade for a phenol-sulfuric number on a mycelial-on-grain powder: not load-bearing.

A high beta-glucan fruiting-body powder still has not been shown to be Mori's cognitive object. Beta-glucan is an identity and quality marker, not the NGF hypothesis. Confusing the two is how a glucan assay becomes a nootropic claim.

22 Erinacine and hericenone standardisation

The only human trial in this file that specified a diterpenoid marker was Li 2020: 5 mg/g erinacine A in EAHE mycelia (Li et al., 2020). Hu et al., 2019, and Chen et al., 2024, are pharmacokinetic papers on erinacines, not retail certificates. Hericenone assays on commercial fruiting-body powders are scarce in the indexed trial methods. Surendran used a 10:1 extract with 5% maltodextrin and did not publish a hericenone milligram amount (Surendran et al., 2025). Mori used 96% dry powder without a hericenone number (Mori et al., 2009). Docherty published 1.8 g of a Sempera Organics product without a public marker table in the abstract-level record (Docherty, Doughty, and Smith, 2023).

Without a marker, "standardised lion's mane" is a marketing phrase. With erinacine A at 5 mg/g, the product is still only as good as the trial that used it — one mild-AD pilot with industry salaries on the author line. Grade for the existence of a workable erinacine A specification: established as chemistry. Grade for that specification as a property of the retail category: not supported.

23 Species authentication

ITS or similar DNA barcoding can confirm Hericium and, with a proper reference, H. erinaceus. It cannot certify hericenone or erinacine content, and it cannot see starch. A barcode on a grain-mycelium powder that matches H. erinaceus is consistent with a thin fungal film on rice. Microscopy, substrate declaration, enzymatic glucans, and a diterpenoid or hericenone HPLC panel do different jobs. Friedman and He provide chemical and polysaccharide context, not a forensic protocol (Friedman, 2015; He et al., 2017). Grade for DNA as a species screen: strongly supported. Grade for DNA as a quality certificate: not supported.

24 What remains, and what would change the thesis

The justified remainder is small.

H. erinaceus is a characterised basidiomycete with a real hericenone/erinacine split (Friedman, 2015; Li et al., 2018). Fruiting-body dry powder moved a Japanese MCI scale in 15 people and the effect receded after stopping (Mori et al., 2009). An erinacine A–enriched mycelium produced mixed cognitive and biomarker signals in 41 mild-AD completers in a Grape King–salaried pilot (Li et al., 2020). Mood and sleep findings are smaller, shorter, or diet-entangled (Nagano et al., 2010; Vigna et al., 2019). Healthy-young cognition is null, task-specific, or caffeine-adjacent (Grozier et al., 2022; Docherty, Doughty, and Smith, 2023; La Monica et al., 2023; Surendran et al., 2025). Neurogenesis is a mouse and dish word (Ryu et al., 2018; Lai et al., 2014). Oral human CNS exposure of the named compounds is unshown (Chen et al., 2024, is one IV pig). Grain-grown mycelium fails a starch-aware glucan test unless it passes McCleary and a marker assay (McCleary and Draga, 2016).

What would change the thesis: a pre-registered, independently sponsored, product-characterised trial in MCI or mild AD with a locked primary cognitive endpoint and a measured hericenone or erinacine exposure; or a human oral CSF series. Until then, the aisle is running ahead of the file.

Menon and colleagues' 2025 review is useful as a reminder that even a systematic search still returns a handful of human papers and a long preclinical tail (Menon et al., 2025). That imbalance is the result, not a temporary embarrassment.

25 Standing constraint

References

Evidence handling

Human randomised and controlled trials are labelled as such in the reporting sentence. Animal and cell findings are labelled as animal or cell findings and are never phrased as human outcomes. Industry salaries and product provision are named beside the trial they funded or supplied. Conflicting results are kept in the same section: Grozier and Surendran sit next to Docherty; Vigna's null depression sits next to its anxiety signal; Mori's unreplicated MCI scale sits next to Li's different product in mild AD.

Grades mean this:

• Established — independently replicated observation with stable methods (taxonomy; hericenone/erinacine tissue split; enzymatic vs phenol-sulfuric glucans as a quality problem).
• Strongly supported — consistent evidence with residual limits (in-vitro NGF induction; short-term GI-predominant tolerability in small trials; McCleary as the honest glucan assay).
• Emerging — real signal, small or single-context (Mori 2009; Li 2020; Nagano mood items; pig IV CSF).
• Plausible — mechanistically coherent, unmeasured in the relevant human compartment (oral hericenone CNS exposure; BDNF as the mood path).
• Speculative — marketing or over-read mechanism (human neurogenesis; interchangeable products; dementia prevention).
• Not supported / insufficient — stated where the file cannot carry the claim.

Matrices for human trials, preclinical work, extract and product quality, and safety are filed with the project apparatus. They use author–year and study type. Amounts are study facts. No row is a use instruction.

Original analysis, compilation and graphics in this document are original works generated from the cited numerical results; no third-party figure has been reproduced.