"Natural" does not mean "inert". Mushroom supplements have a
small but real body of documented harm — and a large body of
tolerability data. Both are collected here without marketing
retouch.
Two categories of data
Human mushroom safety is known from two different types of
sources, and they should not be conflated:
Phase I and interventional work — where
tolerability was measured deliberately: trametes extract up
to 9 g/day in women after breast-cancer treatment,
Cordyceps cicadae mycelium for 3 months in healthy
adults, Phase I Agaricus blazei in 78 patients in
remission, C. militaris in people with already
elevated liver transaminases. Here there are numbers: doses,
duration, laboratory panels.
Case reports — where harm happened in real
life and was published. These are not "weak data" but the
canonical form of pharmacovigilance: this is exactly how
rare and unexpected reactions that RCTs cannot catch are
recorded.
Documented harm
The most notable signal is chaga and oxalate
nephropathy: three published cases from Japan and
Korea, all with biopsy-confirmed oxalate crystals in renal
tubules. In one case the oxalate content of the chaga powder
was measured directly — 14.2 g per 100 g, which at daily
intake produced a load 2–5 times the usual dietary one. The
common denominator of the cases: concentrated powder,
grams a day, months and years.
The second class is drug interactions:
maitake and INR elevation on warfarin (the Hanselin 2010
case), cautions on reishi and anticoagulants/immunosuppressants
from the MSKCC database. The full map is in our article
"Mushroom supplements and medicines".
What tolerability studies show
Where safety was measured deliberately, the picture is
reassuring: Phase I trials found no toxicity at working doses,
and laboratory panels (liver, kidneys, blood) stayed within
normal ranges. But these studies share a common limit: they
are short (weeks to months) and run in specific groups — they
say "at this dose, these people had no problems", not "safe
always and for everyone".
The practical boundary
Documented harm converges into one formula:
concentrate + large dose + long duration + vulnerable
system (kidneys, anticoagulation, immunosuppression).
Culinary doses of mushrooms do not fall into this formula.
Grams of chaga powder daily — falls into it squarely.
Related species
Chaga (Inonotus obliquus) — the only mushroom with a series of biopsy-confirmed kidney complications.
A 57-year-old man ate self-collected wild mushrooms consistent with Hypholoma fasciculare and, about 10 days later — without a violent onset — developed pancytopenia (suppression of all blood cell lines). The diagnosis was reached by exclusion; with supportive care the blood counts fully recovered.
A review of trace-element speciation in edible mushrooms (arsenic, chromium, selenium, mercury, antimony): analytical methods, literature data, risk assessment. Key findings: chromium barely bioaccumulates, antimony is mostly Sb(V), mercury builds up on contaminated sites, ~10% of samples exceed inorganic-arsenic limits.
A 19-year review of 118 poison-center reports involving gyromitrin-containing false morels: 91.5% were Gyromitra esculenta, and the picture was mostly gastrointestinal plus hepatotoxicity rather than the textbook seizures. The danger of false morels is documented reality, not legend.
In the mycelium of Clitocybe rivulosa, 4′-phosphomuscarine — a phosphorylated precursor with almost no affinity for the M3 muscarinic receptor — was identified as the major natural product. Upon cellular injury (chewing, digestion) it is dephosphorylated into lethal muscarine: the mushroom stores a 'safe' form and pulls the trigger at the moment of being eaten.
First broad mapping of gyromitrin across Discinaceae: the toxin was found throughout the G. esculenta group and in G. leucoxantha — but is not spread across the whole genus. The finding recalibrates our cards: 'gyromitra = gyromitrin' is a simplification this study was the first to test quantitatively.
Systematic review and meta-analysis of 33 studies on mushroom-poisoning mortality: reported case fatality ranged 0–40% across series, pooled estimate 2.87%. Of 16 liver-transplant patients, 14 survived. The worst prognosis is in patients with hepatic and renal involvement; early admission reduces transplant need.
Amatoxins cause ~95% of fatal mushroom poisonings — and 'all Lepiota' are traditionally suspected. Phytochemistry 2022 was the first broad test of 16 species: toxins were found only in a monophyletic clade (subincarnata, brunneoincarnata, elaiophylla) — 12 species are clean. So 'lepiota = amatoxin' is itself a simplification.
Diaz's 2021 review systematises nephrotoxic mushroom poisonings: while amatoxins hit the liver, orellanine webcaps and newer species (Amanita smithiana, A. proxima) attack the kidneys — acute renal failure can develop within days without signs of hepatitis. A toxidromic approach by onset of renal failure allows earlier diagnosis; hemodialysis is often indicated, kidney transplantation is the fallback.
Clinical safety assessment of fermented Cordyceps cicadae mycelium: 49 participants took 1.05 g of lyophilized granules daily for 3 months. Blood analysis before and after showed no significant changes in kidney, liver, electrolyte or lipid markers; participants reported no adverse effects or complaints.
The Lagrange–Vernoux review on hydrazinic toxins: Gyromitra esculenta's disputed 'edibility' is explained by individual susceptibility; the authors suggest Verpa bohemica is likewise inedible. A transient cerebellar syndrome from heavy true-morel meals and a GI 'haemolytic' syndrome from raw or undercooked morels are described.
The first systematic review of medicinal-mushroom interactions with chemotherapy: 213 papers (77 clinical) on co-using reishi and turkey tail with cytotoxics. The headline — clinically significant pharmacokinetic interactions are essentially absent, but study quality is low and specific combination signals call for caution.
A counter-intuitive result: urban R. cyanoxantha held LESS cadmium (0.39 mg/kg) than forest ones (3.15–14.1 mg/kg). Soil properties drove accumulation more than urbanisation itself — an 'edible forest mushroom' is not automatically cleaner than a city one.
In Inocybe virosa, liquid chromatography confirmed muscarine, and gas chromatography found volatile toxic metabolites (methyl palmitate, phenolic compounds, phytol). Further chemical confirmation that 'innocent' white fibercaps carry a muscarinic charge.
A Korean RCT in people with mild liver dysfunction (ALT 1.5–3× above normal): Cordyceps militaris extract 1.5 g/day vs. placebo with a full lab panel, liver CT and symptom scales. The preparation was tolerated safely — neither hematology nor biochemistry showed alarming shifts; the authors state protection against fatty-disease progression as a possibility, not a proven fact.
The best-documented harm from a "medicinal" mushroom: three published cases of oxalate nephropathy during regular chaga powder intake. Measured oxalate content in chaga — 14.2 g/100 g, exceeding ordinary dietary load 2–5×; outcomes range from reversible acute kidney injury to end-stage renal disease. A case-report format card — the canonical data type for safety.
A study of mercury in 20 populations of C. caperatus: median cap concentration 0.81–2.4 mg/kg dry weight, bioconcentration factor 18–120 — even in clean forests the mushroom accumulates Hg above background. The authors' conclusion: occasional consumption is safe, but frequent season-long picking is not.
The classic phase I 'how much can you take': women after chemo- and radiotherapy for breast cancer took turkey tail at 3, 6 or 9 g/day for six weeks. Of nine adverse events, seven were mild, one moderate, one severe; the preparation was judged tolerable up to 9 g/day. Incidental immune shifts were trends: lymphocyte rises at 6–9 g and NK-activity gains at 6 g/day.
Phase I in 78 cancer patients in remission: A. blazei granules at 1.8, 3.6 or 5.4 g/day for 6 months. Adverse events in 9 patients (12%) — mostly gastrointestinal (nausea, diarrhea); one case of liver dysfunction judged a food allergy. No dose dependence was found. The authors conclude general safety, apart from a possible allergic reaction.
A case report from Annals of Pharmacotherapy: a patient on a stable warfarin dose developed a rise in INR (a coagulation measure) after adding maitake extract. After the extract was stopped, INR returned to target. A single documented signal — but a clinically important one: anyone on anticoagulants should monitor INR when taking maitake.
A descriptive review of world literature 1951–2004 collecting 28,018 confirmed mushroom poisonings. Diaz proposed an expanded syndromic classification — 14 syndromes grouped by onset time (early <6 h, late 6–24 h, delayed ≥1 day) and target organ — and formally described four new ones: accelerated nephrotoxicity (A. proxima, A. smithiana), rhabdomyolysis (T. equestre, R. subnigricans), erythromelalgia (Clitocybe spp.) and delayed neurotoxicity (Hapalopilus rutilans).
Part of a larger mapping of amatoxins/phallotoxins: in Conocybe lactea no amatoxins were found — but phallotoxins were, the first detection outside the genus Amanita (in small amounts). Meanwhile C. filaris (=Pholiotina rugosa) is confirmed as a true amatoxin bearer. Both toxins and species refuse to fit the 'genus = toxin' frame.
The largest contemporary review of amatoxin-poisoning treatment: 2,108 hospitalised patients from North American and European literature over ~20 years. Penicillin G was used most often but showed little efficacy; thioctic acid and steroids showed none. Silybin (alone or in combination) and N-acetylcysteine as monotherapy were the only regimens statistically associated with survival — the analysis laid the foundation of the modern silibinin + NAC + supportive-care protocol.