Medicinal Mushrooms and Autoimmune Disease: What the Research Actually Tested — and What Must Not Be Ignored

In brief5 points · 1-minute read

• Across 13 autoimmune diseases, only one randomized controlled trial has tested any of the four mushrooms in humans — and it missed its primary endpoint.
• That trial gave reishi together with San Miao San to 65 rheumatoid-arthritis patients for 24 weeks; ACR20 was reached by 15% versus 9.1%, not significant.
• In ten of the diseases, including multiple sclerosis, lupus and Hashimoto's, there is no human study at all — only mouse or rat models, or nothing.
• Injected β-glucans triggered chronic arthritis in genetically predisposed mice in four studies, but none of them used reishi, cordyceps, lion's mane or turkey tail.
• Cordyceps alongside cyclosporine was measured in 202 kidney-transplant recipients; methotrexate, biologics and JAK inhibitors have zero interaction studies with any of the four.

Questions about this article? Ask the Triterra bot ›

Reishi (Ganoderma lucidum, “lingzhi” in Chinese medicine), cordyceps (“dong chong xia cao”), lion’s mane (Hericium erinaceus, “yamabushitake”) and turkey tail (Trametes versicolor, “yun zhi”) are medicinal mushrooms marketed all over the world as “immune modulators” — and therefore, supposedly, suitable for people living with an autoimmune disease as well.

The honest answer: across the 13 autoimmune diseases we screened, there is exactly one randomized controlled trial in humans with any of these four mushrooms — and it missed its primary endpoint. In ten of the diseases there is not a single human study. And against the modern drugs for autoimmune disease — methotrexate, biologics, JAK inhibitors — the number of interaction studies is zero.

What does exist: one randomized trial in rheumatoid arthritis; a randomized trial in inflammatory bowel disease with a different mushroom (Agaricus); a Cochrane review of cordyceps in kidney-transplant recipients — the only population in which a mushroom was deliberately given alongside immunosuppressants; trials in healthy volunteers that measured immune cells; and, above all of those, a mountain of mouse and cell-culture studies that the internet presents as evidence in humans. This page separates them, line by line.

Can you take medicinal mushrooms with an autoimmune disease?

That is a question for your treating physician, not for us — and not out of legal caution, but because not one clinical trial has asked it. What was measured: mushroom extracts changed immune cells measurably in healthy volunteers, and cordyceps was deliberately given alongside immunosuppressants in kidney-transplant recipients. What that means in active autoimmune disease — nobody has tested.

We grow and sell mushroom extracts, which is exactly why we will say this plainly here: autoimmune disease is treated with drugs, not with extracts. The drug you are taking now prevents damage that cannot be undone — to a joint, a nerve, a kidney, a gland. A mushroom extract will not do that. Do not stop or change a prescribed treatment on your own — not because of what you read here, and not because of what you read anywhere else.

And why this question is different from every other question about supplements: in autoimmune disease the immune system attacks the body’s own tissue — a joint, myelin, the thyroid, the gut — and the drug suppresses or blocks part of it. A supplement that changes immune activity touches precisely the variable the drug is trying to control, and these mushrooms change it measurably. In a randomized double-blind trial in healthy adults aged 18–55, 84 days of reishi β-glucan significantly raised CD3, CD4 and CD8 lymphocytes, NK cells and IgA. In 79 healthy adults, 8 weeks of a cordyceps mycelium extract (1.68 grams a day) raised the cytotoxic activity of NK cells by 38.8% versus placebo. And in a Cochrane meta-analysis of five trials in cancer patients, reishi raised CD3 by 3.91%, CD4 by 3.05% and CD8 by 2.02% — small changes, from studies whose quality was rated “not satisfactory”, but changes. The assumption that “it’s natural, it does nothing to the immune system” contradicts human data. What it does in the body of a person with active autoimmune disease — no study has answered. Every chapter below states the system in which the finding was measured: in humans, in mice, or in a test tube.

What happens in the body when β-glucan meets the immune system?

β-glucan binds a receptor called Dectin-1 — which sits on monocytes, macrophages, neutrophils and dendritic cells — and additional receptors (CR3 and others). The binding sets off a signalling cascade that reaches the cell nucleus. What comes out of it — inflammation or tolerance — depends on the ligand, the context and the host, not on the receptor alone. Most of what is known here was measured in mice and in cultured cells.

The basic biology is well established: work from 2004 showed that Dectin-1 binds β-glucans with (1→3) and (1→6) linkages and is expressed mainly on monocytes, macrophages, neutrophils and dendritic cells — in isolated mouse and human cells. And a 2022 review summarized that the outcome of binding is not fixed: the same receptor can produce inflammatory cytokines, or a “tolerant” response — depending on the ligand, and the pathway can be tuned in either direction.

Two more findings. “Trained immunity”: β-glucan from the Candida cell wall programmed human monocytes in culture to respond more strongly to the next challenge, through the Dectin-1–Akt–mTOR–HIF-1α axis — an advantage against infection, but a 2024 review linked the maladaptive version of the same mechanism to chronic inflammation and to comorbidities, including arthritis. And the size of the gap: a systematic review of dietary polysaccharides found 62 publications on rodents against only 15 controlled human studies. We wrote at length about β-glucans and the immune system in medicinal mushrooms and the immune system and in how mushrooms modulate the immune system. Here it is enough to say: the receptor is a doorbell, not a switch. Who opens the door depends on who is home.

“Mushrooms modulate rather than stimulate” — is there any basis for that?

Partly — and the phrasing itself is wrong. In the pharmacological literature, “immunomodulator” is the umbrella term and “immunostimulant” is a subcategory of it. Saying “modulates rather than stimulates” is like saying “a vehicle, not a car”. There is genuine mechanistic evidence for a regulatory arm, in mice; and there is evidence, in those same mice, for the opposite direction.

Let us start with what genuinely supports the slogan, because there is such a thing. In work from 2006, zymosan — a yeast cell wall, a ligand for Dectin-1 and TLR2 — made mouse dendritic cells secrete abundant IL-10 and almost no IL-12p70, and made macrophages secrete TGF-β; T cells exposed to an antigen together with it did not respond to a repeat challenge: tolerance, not inflammation. In work from 2014, the same zymosan promoted regulatory T cells (Treg) in NOD mice and protected them from autoimmune diabetes. That is the regulatory arm, and it is real — in a mouse, and with zymosan, which is neither reishi nor turkey tail.

Now the other side, which almost no website tells. SKG mice are genetically prone to autoimmune arthritis, but in a microbially clean environment they do not develop it. In work from 2005, a single intraperitoneal injection of zymosan, curdlan or laminarin triggered severe chronic arthritis in them — and blocking Dectin-1 prevented it. Same receptor, same family of molecules. And in work from 2007, schizophyllan — a fungal β-glucan, a registered drug in Japan — caused more severe arthritis in those same mice than laminarin did.

And the Treg claim, the heart of the slogan, contradicts itself in the literature. In NOD mice zymosan raised Treg; in tumour-bearing mice, a reishi polysaccharide (GLPS) actually lowered Treg accumulation and function and suppressed FoxP3 expression via miR-125b. Same family of molecules, opposite direction, depending on the model. Anyone who writes “reishi raises Treg” without naming the model has not read both studies.

And in humans? In the only trial in active autoimmune disease — 65 RA patients, 24 weeks — no immunomodulatory effect was measured, in any direction. In healthy volunteers, as we saw, the direction was up. The only fair phrasing: the effect is context-dependent — both activation and suppression have been measured, sometimes in the same experiment; and in a person with active autoimmune disease nothing has been measured at all. “Modulates”, in a marketer’s mouth, usually means “not measured”. We went deeper into this question in reishi and the immune system: modulation or stimulation?

What did the only randomized trial in an autoimmune disease find?

A 2007 double-blind trial: 65 patients with active rheumatoid arthritis, on top of their existing DMARDs, received 4 grams of reishi + 2.4 grams of San Miao San a day, or placebo, for 24 weeks. The primary endpoint, ACR20, was reached by 15% versus 9.1% — not significant. CD4, CD8, NK and B cells and five inflammatory cytokines in plasma did not change at all.

What did improve: the pain score and the patient’s global assessment — two self-reported, secondary measures. 13 patients reported 22 mild adverse events, 8 in the herbal group and 14 in the placebo group. And the authors’ conclusion, word for word: “no significant antioxidant, anti-inflammatory or immunomodulatory effect could be demonstrated”.

Two points most websites leave out. First: the reishi was given together with San Miao San, a separate herbal formula; even had there been a positive result, it could not have been attributed to reishi — neither the improvement in pain nor the safety profile. Second: anyone who cites this study as “controlled trials showed significant improvement in RA patients” has not read it. We wrote about arthritis more broadly in do mushrooms help with arthritis — what the science says, and on the narrower question of the inflammatory response in does reishi affect the inflammatory response?

What is known about medicinal mushrooms and multiple sclerosis?

In humans — nothing. Zero trials, zero case series. A PubMed query combining mushroom extract, multiple sclerosis and patients returns zero results. What exists is two mouse models: in the EAE model, reishi reduced demyelination when given early as prevention; in the cuprizone model, reishi and lion’s mane suppressed demyelination — but lion’s mane did not improve motor function.

We write this chapter knowing that customers with MS have contacted us, and that whoever is reading it now may be living with the disease. So we will spell out exactly what there is. In a 2026 study, EAE mice — a model that mimics one aspect of multiple sclerosis — given reishi before disease onset showed lower severity, less demyelination and a shift of microglia toward M2 polarization. The dose was not stated, the dosing was preventive rather than therapeutic, and these were mice.

In a 2022 study, mice were fed for five weeks on a diet containing 5% mushroom together with cuprizone, a substance that causes demyelination. Reishi and lion’s mane prevented weight loss and suppressed demyelination. But on the rotarod test — motor function — only Pleurotus, reishi and enokitake improved. Lion’s mane, the mushroom marketed more than any other for “nerve health”, did not improve motor function. On cordyceps and turkey tail there is not even an animal study.

What this means if you live with MS: your disease-modifying therapy was measured in patients — relapses, MRI lesions, disability progression. The mushrooms were measured in mice. Those are two different worlds. Do not stop or change a prescribed treatment on your own, and if you are considering any supplement at all — your neurologist needs to know before, not after.

What was found in inflammatory bowel disease — Crohn’s and ulcerative colitis?

For our four mushrooms — no human study. There is a randomized trial of a different mushroom, Agaricus blazei (the AndoSan extract): 50 ulcerative-colitis patients, 21 days. The symptom score fell from 5.88 to 4.50, but faecal calprotectin — the only objective marker of gut inflammation — did not change at all. And of 18 cytokines, only one moved in each disease.

The trial was single-blinded, 24 on the intervention and 26 on placebo. Symptoms and fatigue improved significantly; then came the test of the inflammation itself — faecal calprotectin and blood counts — no change. In the companion cytokine study, in Crohn’s only IL-2 fell significantly and in colitis only IL-5; TNF-α, IL-6, IL-17, IL-1β and IL-10 did not change. The authors themselves wrote that the effect was “limited” and supported an anti-inflammatory effect “only marginally”. People felt better; the gut, by the measurement, did not.

On our mushrooms there are mice only: ganoderic acid A from reishi, in a DSS-induced colitis model, prevented colitis and preserved the integrity of the epithelium. A hypothesis to test — not a basis for a decision.

And what about psoriasis, Hashimoto’s, lupus, type 1 diabetes and the other diseases?

Zero human research in every one of them. Some have a single mouse or rat model; in others — psoriatic arthritis, celiac disease, alopecia, myasthenia — dedicated PubMed queries returned zero results. The table that follows is the most important chapter on this page: 13 diseases, and what actually exists for each one.

Even in the rodent rows the findings are not clear-cut: in the Sjögren’s model the area of lymphocytic infiltration fell, but the incidence of the inflammation itself did not differ from controls; and the LZ-8 protein that prevented diabetes in mice was, in culture, an activator of human blood cells in the presence of monocytes, and a suppressor only in their absence. Results in rodents are a hypothesis to be tested in humans. That test has not been done.

Has a worsening of autoimmune disease following β-glucan been documented?

Yes — in mice, by injection. Four studies documented that β-glucans triggered or aggravated autoimmune arthritis in genetically predisposed mice. The fair caveat, which must not be left out: all four by intraperitoneal injection, in mice built to become ill, and none of them with reishi, cordyceps, lion’s mane or turkey tail.

Here are the four. In work from 2005, a single intraperitoneal injection of zymosan, curdlan or laminarin triggered severe chronic arthritis in SKG mice — and only transient arthritis in ordinary mice; blocking Dectin-1 prevented it. The same year, particulate β-glucan from Candida served as an adjuvant for inducing arthritis in DBA/1 mice — and the disease was more severe than the one induced with complete Freund’s adjuvant, the standard of the field. In work from 2007, schizophyllan — a β-glucan from the fungus Schizophyllum commune, a registered drug in Japan — caused more severe arthritis in SKG mice than laminarin, with high serum IL-6; the authors concluded that fungal β-glucans “are capable of inducing and aggravating autoimmune diseases such as RA”. And in work from 2018, Nod2-deficient SKG mice exposed to β-glucan developed a dramatically aggravated form, with expansion of Th17 cells.

And now the caveat, in full, because without it this chapter becomes fear-mongering. Route of administration: all four studies injected the β-glucan straight into the abdominal cavity — like the difference between drinking a glass of water and injecting it into a vein. Injection bypasses the gut, the liver and every control mechanism along the way. The host: an SKG mouse is not an RA patient; it is a mouse with a defined mutation, built to mimic one aspect of the disease — a flight simulator, not a flight. The substance: zymosan is yeast cell wall, laminarin comes from seaweed, curdlan from a bacterium, schizophyllan from a fungus that is not one of our four. We looked for an autoimmune model in which one of our four mushrooms made a disease worse — and found none.

What may be concluded: the molecule is not “gentle” by nature; the outcome is decided by host genetics, route of administration and dose. And what may not be concluded: neither that swallowing an extract is dangerous, nor that it is safe. The measurement — by mouth, in humans with autoimmune disease — has simply not been done.

And has a flare been documented in humans?

For our four mushrooms — we searched explicitly and found not a single case report. But there is a documented series of three patients who flared — two flares of pemphigus vulgaris and one case of dermatomyositis — close in time to taking other immune-stimulating supplements: echinacea and spirulina. Supplements are not subject to mandatory pharmacovigilance reporting. That is a data void, not a safety certificate.

The series was published in 2004. One of the patients carried a polymorphism in the TNF-α promoter that increases its production — a hint that genetic background, as in SKG mice, determines who reacts. In other words: there is a documented precedent for an immune-stimulating supplement coinciding in time with a flare. It is not about mushrooms.

And why the absence of reports on mushrooms is not proof: drugs are bound to an adverse-event reporting system; supplements are not. A rheumatologist whose patient flares will usually not suspect the supplement and will not report it. And in the other direction: there is no documented case showing that medicinal mushrooms cause a flare. Whoever writes that “mushrooms are dangerous for autoimmune patients” is making it up exactly as much as whoever writes that they “balance”.

What happens when medicinal mushrooms are combined with immunosuppressants?

Here there actually are human data — and precisely because of them, this is not a decision to make alone. In a trial of 202 kidney-transplant recipients, the group that received cordyceps (1.0 gram ×3 a day) had significantly lower cyclosporine doses in months 2–6, and lower blood trough levels in months 3–6. The combination is done deliberately in China, under monthly blood monitoring. Outside monitoring — no.

Two reservations about the cyclosporine rows, without which the table misleads. First: in the 202-patient study the physicians deliberately reduced the cyclosporine dose in the cordyceps group, so the fall in trough levels may stem from the dose reduction rather than from a direct pharmacokinetic interaction. It is not permissible to write “cordyceps lowers cyclosporine” as a mechanistic fact; what is permissible: in clinical series in which cordyceps was given, cyclosporine doses and levels were significantly lower. Second: the preparations there were fermented-mycelium capsules, not a liquid fruiting-body extract, and carrying the finding over from a capsule to a bottle is not scientifically justified. On cordyceps and the kidney — in kidney health, and every interaction outside the autoimmune field — in are there drug interactions?

And what about methotrexate, biologics and JAK inhibitors?

Never tested — and that is the whole answer. A PubMed search for an interaction or pharmacokinetic study between methotrexate, TNF blockers, rituximab, IL-6 blockers or JAK inhibitors and any of the four mushrooms returns zero results. “No risk was found” and “no one looked for a risk” are two different sentences, and you are in the second one.

All the human literature on combining mushrooms with immunosuppressants comes from organ-transplant recipients on cyclosporine and tacrolimus — drugs whose blood level is measured every month. An RA patient on methotrexate or a Crohn’s patient on adalimumab is not monitored that way. The conclusion “the combination is safe because they do it in China” ignores that in China it is done inside a monitoring framework.

The only mechanistic link is theoretical: JAK inhibitors are metabolized substantially through CYP3A4, and a reishi polysaccharide inhibited CYP3A in rat liver microsomes. But inhibition in a test tube, at concentrations orders of magnitude above what is reached by swallowing, does not predict inhibition in a person — MSKCC itself writes “clinical relevance is not known”. A flag for a study to be done, not proof of an interaction. And what is true of every row in the table is doubly true here: do not stop or change a prescribed treatment on your own, and tell your physician about every supplement you take. Even if it is “natural”. Especially if it is “natural”.

What is known about raw-material quality — and why does it matter here in particular?

Because someone who suppresses their immune system with a drug cannot afford a contaminant. In an analysis of 17 production batches of wild cordyceps, 88.24% exceeded the arsenic limit. And a more disturbing finding: a fungus of the genus Tolypocladium, isolated from a wild cordyceps fruiting body, contained cyclosporine A in its mycelial extract — a genuine immunosuppressant.

The second finding deserves an explanation. Fungi of the genus Tolypocladium produce cyclosporine routinely. In a 2006 study such a fungus was isolated from a wild cordyceps fruiting body, and cyclosporine A was identified in its mycelial extract. The meaning: raw material sold as “cordyceps” may contain a different fungus that itself produces an immunosuppressant, in a quantity nobody has measured. For a person on methotrexate that is not a technical detail.

And on arsenic: in the 17 batches, copper, lead, cadmium and mercury were within the limit in all of them — but arsenic exceeded it in 88.24%, concentrated in the larval body at 7 to 12 times the level in the stroma. NCCIH, the US institute for complementary medicine, puts it directly: “What’s listed on the label of a dietary supplement may not be what’s in the product”. The only answer is a certificate of analysis for what you are actually buying — how to read a test report (COA) explains what to look for in it.

What we actually measure in our bottle

After thirteen chapters of “not measured”, here is what is. We grow the mushrooms on our farm in the Galilee and harvest fruiting body — not mycelium grown on grain, not imported powder that nobody knows what happened to on the way. That difference is not marketing: mycelium-on-grain usually contains less than 7% β-glucan, because most of its weight is the grain itself; dried, imported fruiting body sits one step above it; and fresh fruiting body from the farm, going to extraction the same day, sits on the third step. As far as we know, we are among the very few in the world who move the mushroom from harvest to extraction with no drying step at all.

The extraction is triple: water, alcohol and heat, with the alcohol stage alone lasting seven weeks. On the page about extraction ratios we explained why 1:3 is a ratio and not “triple extraction” — and why both numbers matter. The finished product contains 32% alcohol, and we write that here explicitly, because for someone taking methotrexate who has been told to limit alcohol, that is a figure the physician needs to see.

Then the numbers. We sent the finished extracts — not the raw material — for testing at TÜV Austria, and measured β-glucan on a dry basis. We also measured α-glucan, which is starch: it is the chemical marker for grain, and if it is not detected — there is no grain in there. These are the results for the cordyceps extract, the reishi extract, the lion’s mane extract and the turkey tail + reishi extract:

All our tests are public, and you can read them on the lab results page. That is what we know how to measure and prove about what is in the bottle. What it will do for autoimmune disease was not measured — so it is not written.

What this page does not say — and what we do not claim

We do not claim that medicinal mushrooms treat, relieve, prevent or “balance” any autoimmune disease — no human study supports it, and under DSHEA a dietary supplement may not claim to diagnose, treat, cure or prevent disease. We do not claim they are safe for someone taking immunosuppressants, methotrexate, biologics or JAK inhibitors — because nobody has tested it. And we do not claim the opposite either: there is no case report of a flare following any of the four mushrooms, and the mouse studies of worsening are by injection, in genetically predisposed animals, with other molecules. We do not present a finding from mice as though it were measured in a person, nor a finding in healthy volunteers as though it were measured in patients. We do not write “safe” about anything that was not measured. And we do not offer an extract — ours or anyone else’s — as a replacement for, an addition to, or an improvement on the treatment your physician prescribed.

If you have an autoimmune disease — what do you actually do?

You continue your drug treatment exactly as prescribed. If you are considering any supplement — mushroom or otherwise — you bring it to your treating physician before you start, with the ingredient list, the alcohol content and the certificate of analysis. If the physician agrees, you start after scheduled blood tests, so that there is a baseline for comparison.

A few practical points. On cyclosporine or tacrolimus — that is the family for which human series exist in which blood levels differed; the nephrologist who monitors you needs to know. On methotrexate — our extract contains 32% alcohol, a figure the physician needs to see. Before surgery — in 40 healthy volunteers reishi did not change coagulation measures, and you still inform the surgeon and the anaesthetist about every supplement. And if what troubles you is joint pain that is not autoimmune — gout, for example — that is a different page: medicinal mushrooms and gout. General side effects, unrelated to autoimmunity, are covered in do mushrooms have side effects?

The bottom line

Across 13 autoimmune diseases there is one human trial with any of the four mushrooms, and it missed its primary endpoint and demonstrated no immune effect. “Modulates rather than stimulates” is a category error: the same family of molecules produced tolerance in one mouse and chronic arthritis in another, and raised immune cells in healthy humans. The combination with cyclosporine and tacrolimus was measured in kidney-transplant recipients and changed a variable that physicians monitor; the combination with methotrexate, biologics and JAK inhibitors has never been measured. There is no report of a flare from mushrooms — and there is also no system that would have caught one. What was measured in our bottle — β-glucan and α-glucan at TÜV Austria — we publish. What it will do for your disease was not measured — so it is not written.

Frequently asked questions

I have rheumatoid arthritis and I’m on methotrexate. Can I take reishi?

Not one study has tested methotrexate against any of the four mushrooms. The only RA trial (65 patients, 24 weeks) gave reishi + San Miao San on top of existing drugs and found 8 mild adverse events in the herbal group versus 14 on placebo — but it was not designed to measure an interaction. Note also the 32% alcohol in the extract. The decision belongs to your rheumatologist, before and not after.

Is there any research on medicinal mushrooms and multiple sclerosis?

In humans — none. A PubMed query combining mushroom extract, multiple sclerosis and patients returns zero results. There are two mouse studies: in the EAE model reishi reduced demyelination when given preventively, and in the cuprizone model reishi and lion’s mane suppressed demyelination — but lion’s mane did not improve motor function. Mice are not patients. Do not change a disease-modifying therapy because of a mouse study.

If mushrooms “balance” the immune system, isn’t that safe for autoimmune patients?

“Balance” is not a measurement. What was measured in healthy humans is an increase — CD3, CD4, CD8, NK cells and IgA — one direction. In mice, both tolerance (zymosan → IL-10, Treg) and chronic arthritis (injected β-glucan in SKG mice) were measured. The direction depends on context, and in a person with active autoimmune disease nothing has been measured. “Modulates” here means “unknown”.

I’m a kidney-transplant recipient on cyclosporine. Is cordyceps dangerous for me?

Here there actually are data, and because of them this is not a decision to make alone. In a trial of 202 kidney-transplant recipients, cyclosporine doses and blood trough levels were significantly lower in the group that received cordyceps — possibly because of deliberate dose reduction, possibly because of an interaction. In China the combination is given deliberately, under blood monitoring. Outside monitoring, and without your nephrologist — no.

I’m on a TNF blocker or rituximab. What is known about combining them with mushrooms?

Nothing — zero studies. That is not “safe”, that is “not tested”. Biologics are proteins that are broken down differently from chemical drugs, so a metabolic interaction is theoretically less likely — but the more relevant question is the effect on the immune system itself, and nobody has measured that. Update your treating physician before adding anything.

Are there reports of an autoimmune flare from medicinal mushrooms?

We searched explicitly and found no report on any of the four mushrooms. There is a series of three patients — two pemphigus flares and one dermatomyositis case — close in time to taking other immune-stimulating supplements, echinacea and spirulina. Supplements are not subject to mandatory reporting the way drugs are, so the absence of reports also reflects the absence of a reporting system. A data void, not a safety certificate.

What about Crohn’s and ulcerative colitis?

For our four mushrooms — no human study. There is a randomized trial of an Agaricus blazei extract: 50 ulcerative-colitis patients, 21 days, symptom score down from 5.88 to 4.50. But faecal calprotectin, the objective marker of gut inflammation, did not change at all, and of 18 cytokines only one moved in each disease. People felt better; the gut, by the measurement, did not.

I read that β-glucan causes arthritis. Is that true?

In genetically predisposed mice, by intraperitoneal injection — yes, in four studies, and in one of them a fungal β-glucan (schizophyllan) did it more severely than laminarin. But: injection bypasses the gut and the liver; an SKG mouse is not an RA patient; and none of the studies was done with reishi, cordyceps, lion’s mane or turkey tail. It is evidence that the molecule is not “gentle” — not evidence that swallowing it is dangerous, nor that it is safe.

So why does the whole internet say mushrooms help with autoimmunity?

Because most of it cites mouse, test-tube and mechanism studies as if they were human evidence, and sometimes reports the only human RA trial as a “significant improvement” when its primary endpoint was not met. We sell mushroom extracts and we will not do that. “No evidence” is the finding — and it is the fairest one for anyone living with the disease.

Scientific sources (peer-reviewed)

Every source is accompanied by the exact finding, the system in which it was measured, and a PubMed link. For sources whose original title carries a cordyceps species name, the title is omitted and the finding, authors and journal are kept.

Clinical evidence in humans

• Reishi 4 grams + San Miao San 2.4 grams a day, 24 weeks, 65 active-RA patients on top of DMARDs, double-blind: ACR20 in 15% versus 9.1% (not significant); no change in CD4/CD8/NK/B or in IL-18, IP-10, MCP-1, MIG, RANTES; 22 mild adverse events (8 versus 14). “No significant anti-inflammatory or immunomodulatory effect could be demonstrated” — Li EK, et al. Arthritis and Rheumatism, 2007. View on PubMed
• Agaricus blazei extract (AndoSan), 50 ulcerative-colitis patients, 21 days, randomized single-blind (24 versus 26): symptom score 5.88→4.50 (p=0.001); faecal calprotectin and blood counts unchanged — Therkelsen SP, et al. PLoS One, 2016. View on PubMed
• The same extract, 50 Crohn’s and colitis patients: of 18 cytokines, in Crohn’s only IL-2 fell and in colitis only IL-5; the systemic effect “limited”, anti-inflammatory “only marginally” — Therkelsen SP, et al. Scandinavian Journal of Immunology, 2016. View on PubMed
• Cochrane review, cordyceps in kidney-transplant recipients: 5 trials, 447 participants; 4 (265) versus azathioprine, 1 (182) cordyceps + low-dose cyclosporine versus standard dose — no difference in survival, graft loss or rejection; “limited, low-quality evidence”, unclear risk of bias, follow-up up to one year — Hong T, et al. Cochrane Database of Systematic Reviews, 2015. View on PubMed
• 202 kidney-transplant recipients (93 cordyceps 1.0 gram ×3 a day / 109 controls): cyclosporine doses significantly lower in months 2–6 and blood trough levels lower in months 3–6 (P<.05); acute rejection 11.83% versus 15.60% (not significant); unblinded — Li Y, et al. Transplantation Proceedings, 2009. View on PubMed
• 67 kidney-transplant recipients (25/42): the cyclosporine or tacrolimus dose 12 weeks after surgery significantly lower in the cordyceps mycelium-preparation group (P<.05, P<.01); no difference in graft survival or rejection — Ding CG, et al. Chinese Journal of Integrated Traditional and Western Medicine, 2009. View on PubMed
• β-1,3;1,6-glucan from reishi, healthy adults aged 18–55, 84 days, double-blind versus placebo: significant rise in CD3, CD4, CD8, the CD4/CD8 ratio, NK cells, serum IgA and NK cytotoxicity; no change in kidney and liver function (participant number and dose not stated in the abstract) — Chen SN, et al. Foods, 2023. View on PubMed
• Cordyceps mycelium extract (laboratory strain) 1.68 grams a day, 8 weeks, 79 healthy adults (39/40), double-blind: NK cytotoxic activity +38.8% ± 17.6% versus placebo (P<0.019) — Jung SJ, et al. BMC Complementary and Alternative Medicine, 2019. View on PubMed
• Cochrane review, reishi in cancer, 5 trials: CD3 +3.91% (95% CI 1.92–5.90), CD4 +3.05% (1.00–5.11), CD8 +2.02% (0.21–3.84); methodological quality “generally not satisfactory”; minor adverse effects (nausea, insomnia) — Jin X, et al. Cochrane Database of Systematic Reviews, 2016. View on PubMed
• Cochrane review, reishi for cardiovascular risk factors, 5 trials, 398 participants (1.4–3 grams a day, 12–16 weeks): HbA1c WMD ‎-0.10% (95% CI ‎-1.05 to 0.85) — zero effect; adverse event RR 1.67 (0.86–3.24, not significant), not serious — Klupp NL, et al. Cochrane Database of Systematic Reviews, 2015. View on PubMed
• Randomized double-blind trial, 40 healthy volunteers, reishi 1.5 grams a day, 4 weeks: no significant difference in coagulation function, fibrinogen, vWF, PFA-100 and thromboelastography; all values within the normal range — Kwok Y, et al. Anesthesia and Analgesia, 2005. View on PubMed
• Series of 3 cases: two pemphigus vulgaris flares and one dermatomyositis case close in time to taking echinacea and spirulina (not mushrooms); one patient carried the ‎-308A polymorphism in the TNF-α promoter — Lee AN, Werth VP. Archives of Dermatology, 2004. View on PubMed

Mechanism — receptors, regulation, and the counter-evidence (mice and cultured cells)

• Dectin-1 binds β-glucans with (1→3) and (1→6) linkages; expressed on monocytes/macrophages, neutrophils and dendritic cells, and at low levels on a T-cell subset (isolated cells, mouse and human) — Herre J, et al. Molecular Immunology, 2004. View on PubMed
• Review: the outcome of Dectin-1 signalling varies — cytokines and ROS on one hand, tolerant responses on the other — depending on the ligand; the pathway can be tuned toward inflammation or regulation — Mata-Martínez P, et al. Frontiers in Immunology, 2022. View on PubMed
• Zymosan (a TLR2/Dectin-1 ligand) made mouse DCs secrete abundant IL-10 and almost no IL-12p70, and splenic macrophages secrete TGF-β; T cells exposed to antigen + zymosan did not respond to a repeat challenge — immunological tolerance — Dillon S, et al. The Journal of Clinical Investigation, 2006. View on PubMed
• Zymosan promoted a Treg response against a β-cell antigen and protected NOD mice from autoimmune diabetes, with a rise in IL-10- and Foxp3-positive T cells — Karumuthil-Melethil S, et al. Diabetes, 2014. View on PubMed
• A single intraperitoneal injection of zymosan, curdlan or laminarin triggered severe chronic arthritis in genetically predisposed SKG mice (and only transient arthritis in ordinary mice); blocking Dectin-1 prevented it — Yoshitomi H, et al. The Journal of Experimental Medicine, 2005. View on PubMed
• In SKG mice, by intraperitoneal injection: Candida β-glucan and schizophyllan (a fungal β-glucan, a registered drug in Japan) caused more severe arthritis than laminarin, with high serum IL-6; the authors: fungal β-glucans “are capable of inducing and aggravating autoimmune diseases such as RA” — Hida S, et al. Biological & Pharmaceutical Bulletin, 2007. View on PubMed
• Particulate Candida β-glucan as an adjuvant for inducing collagen-induced arthritis in DBA/1 mice: more severe disease than that induced with complete Freund’s adjuvant, with anti-CII autoantibodies — Hida S, et al. Journal of Autoimmunity, 2005. View on PubMed
• Nod2-deficient SKG mice exposed to β-glucan developed dramatically aggravated arthritis with Th17 expansion and elevated IL-17A in synovial fluid; neutralizing IL-17A eased it — Napier RJ, et al. Journal of Immunology, 2018. View on PubMed
• The opposite direction to the Treg claim: reishi polysaccharide (GLPS) in hepatoma-bearing mice lowered Treg accumulation and function and inhibited FoxP3 expression by raising miR-125b — Li A, et al. Journal of Translational Medicine, 2015. View on PubMed
• LZ-8 from reishi: mitogenic to human PBMC in the test tube in the presence of monocytes, suppressive in their absence; prolonged skin-graft survival in mice and completely prevented autoimmune diabetes in NOD mice — van der Hem LG, et al. Transplantation, 1995. View on PubMed
• Candida β-glucan trains human monocytes in culture — a shift to aerobic glycolysis through the Dectin-1–Akt–mTOR–HIF-1α axis; inhibiting any one of them blocks the training — Cheng SC, et al. Science, 2014. View on PubMed
• Review: maladaptive myeloid training is linked to the perpetuation of chronic inflammation and to comorbidities including arthritis and periodontitis (human + mouse) — Mitroulis I, et al. Cardiovascular Research, 2024. View on PubMed
• Systematic review of dietary polysaccharides: 62 publications on rodents against 15 controlled human studies; the literature “is insufficient to support broad structure/function generalizations” — Ramberg JE, et al. Nutrition Journal, 2010. View on PubMed
• Reishi polysaccharide dose-dependently inhibited CYP2E1, CYP1A2 and CYP3A in rat liver microsomes; the authors: drug pharmacokinetics “may be altered” — Wang X, et al. Biological & Pharmaceutical Bulletin, 2007. View on PubMed

Rodent models by disease

• EAE mice (multiple sclerosis model): early preventive reishi reduced severity, inflammatory infiltration, demyelination and microglial activation, and promoted M2 polarization; dose not stated — Zhang L, et al. Chinese Medicine, 2026. View on PubMed
• C57BL/6 mice, 5 weeks on a diet with 5% mushroom + cuprizone: reishi and lion’s mane prevented weight loss and suppressed demyelination; rotarod motor function improved with Pleurotus, reishi and enokitake — not lion’s mane — Yamashina K, et al. International Journal of Medicinal Mushrooms, 2022. View on PubMed
• MRL/lpr mice (lupus model), reishi + San Miao San, 10 mice per group: anti-dsDNA fell (P<0.05); Treg and Breg rose; IL-21, IL-10 and IL-17A fell — Cai Z, et al. Chinese Medicine, 2016. View on PubMed
• Reishi spores in 32 female NOD mice (Sjögren’s model), 10 weeks: the area of lymphocytic foci fell to 50% of controls (P<0.05); the incidence of salivary-gland inflammation did not differ from controls (P>0.05) — Qi G, et al. Chinese Medical Journal, 2009. View on PubMed
• Bailing capsule (a cordyceps preparation) in rats with autoimmune thyroiditis: TgAb and TPOAb fell, TSH normalized, IFN-γ and IL-12 fell, IL-4 and IL-10 rose; dose and number of animals not stated — Wang Q, et al. Journal of Traditional Chinese Medicine, 2024. View on PubMed
• Ganoderic acid A from reishi in an IMQ-induced psoriasis mouse model: reduced skin thickness, erythema and scaling through inhibition of GSDMD-pathway pyroptosis; dose not stated — Xiao C, et al. Tissue & Cell, 2025. View on PubMed
• Ganoderic acid A in a DSS-induced colitis mouse model: prevented colitis, preserved epithelial and mucus-layer integrity, activated AhR and induced IL-22 via gut bacteria — Kou RW, et al. Journal of Agricultural and Food Chemistry, 2024. View on PubMed

Raw-material quality

• ICP-MS on 17 production batches of wild cordyceps: copper, lead, cadmium and mercury within the limit in all; arsenic exceedance rate 88.24%, concentration in the larval body 7–12 times that in the stroma — Zhou L, et al. China Journal of Chinese Materia Medica, 2017. View on PubMed
• A fungus of the genus Tolypocladium isolated from a wild cordyceps fruiting body: cyclosporine A identified in its mycelial extract — an immunosuppressant that Tolypocladium fungi produce routinely — Leung PH, et al. Journal of Applied Microbiology, 2006. View on PubMed

Institutional sources — not peer-reviewed articles

Information pages of medical bodies. They summarize the literature and sometimes rely on anecdotal reports; they are not first-grade evidence, but they are what the medical establishment actually tells patients.

• [Institutional] MSKCC — About Herbs: Reishi Mushroom. “Reishi can enhance immune response” (immunosuppressants); “Reishi can increase the risk of bleeding”; CYP2E1/1A2/3A in the test tube — “clinical relevance is not known”. Memorial Sloan Kettering Cancer Center
• [Institutional] MSKCC — About Herbs: Cordyceps. “No major side effects have been reported”; a case report of increased bleeding after a tooth extraction; an additive hypoglycaemic effect in the laboratory — “Clinical relevance has yet to be determined”. Memorial Sloan Kettering Cancer Center
• [Institutional] NIH LiverTox — Lingzhi, Reishi. “Likelihood Score: D (possible rare cause of clinically apparent liver injury)”; in small placebo-controlled trials aminotransferase levels did not change. NIH LiverTox
• [Institutional] NIH LiverTox — Cordyceps. “Likelihood Score: E (unlikely cause of clinically apparent liver injury)”; the page has no drug-interaction section at all. NIH LiverTox
• [Institutional] NIH LiverTox — Lion’s Mane. “Likelihood Score: E”; but “has not been subjected to prospective trials of its safety”; at least one acute hypersensitivity reaction has been documented. NIH LiverTox
• [Institutional] NCCIH (NIH) — Dietary and Herbal Supplements. “What’s listed on the label of a dietary supplement may not be what’s in the product”; “Supplements you buy from stores or online may differ in important ways from products tested in research studies”. NCCIH

Read next

• Medicinal mushrooms and the liver — what was measured in humans
• Medicinal mushrooms and blood sugar — what the research actually tested
• Medicinal mushrooms and hair loss — what was tested and what was not
• Reishi and the immune system: modulation or stimulation?
• Medicinal mushrooms and the immune system (β-glucans)
• How mushrooms modulate the immune system
• Are there drug interactions?
• Do mushrooms help with arthritis — what the science says
• Does reishi affect the inflammatory response?
• Medicinal mushrooms and gout
• Kidney health
• Turkey tail and inflammatory processes
• Reishi and seasonal allergies
• Do mushrooms have side effects?
• Regulation of mushroom supplements in Israel
• Lab results — β-glucan
• The research hub

This page is educational and does not constitute medical advice. It describes what published trials and case reports recorded — and what was not recorded; it is not a prediction of what any individual will experience, and it is not a recommendation to start, stop or combine any supplement with any medication — those decisions belong with your physician. Mushroom extracts are dietary supplements, not drugs, and they do not cure, prevent or treat autoimmune disease or any other disease. Drugs for autoimmune disease prevent irreversible damage: do not stop or change a prescribed treatment on your own — under any circumstances, and no matter what you read, here or elsewhere. Consult your treating physician before adding any supplement, especially if you take immunosuppressants, methotrexate, biologics, JAK inhibitors, steroids or anticoagulants, before surgery, and in pregnancy or while breastfeeding. The extract contains 32% alcohol. *These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease.*