Mushroom Beta-Glucans and Insulin Signaling: How It Works

Mushrooms & Blood Sugar

4 entry points

Mushroom compounds do not share one blood sugar mechanism. Across the mechanism literature they act at four separable points in the insulin signaling cascade — the receptor and its first relay, the AMPK energy switch, the PTP1B off-switch, and ceramide accumulation. This is our mapping of the studies below, not a framework any single paper proposes.

Rodents & cells

every mechanism study cited here measured insulin signaling markers in rats, mice or cultured cell lines — none measured them in people

Myriocin

the compound behind the Cordyceps ceramide finding; commercial extracts contain it in variable amounts and it is not approved for human use (Li et al., 2022)

FYGL

the single Ganoderma lucidum proteoglycan shown to inhibit PTP1B — not a class of compounds (Yang et al., 2018)

View as image ↗

Mushroom beta-glucans and related fungal compounds improve blood sugar control in animal models, but not through a single shared mechanism. Read the mechanism studies side by side and they separate into four groups. Some compounds re-activate the insulin receptor and its first relay protein. Others bypass the receptor entirely, switching on an energy sensor called AMPK. One inhibits the phosphatase that shuts the insulin receptor off. And one works by reducing a class of lipids that accumulate in metabolic disease and interfere with insulin signaling.

That four-way split is our reading of the literature rather than a framework any single review proposes, and it matters for a practical reason: the compounds are not interchangeable. The maitake fraction that acts on the insulin receptor is a different molecule, from a different species, with a different research base than the Cordyceps constituent that lowers ceramides.

Two recent reviews cover this ground. Aramabašić Jovanović and colleagues (2021, Journal of Fungi) surveyed mushroom polysaccharides and terpenoids, concluding that polysaccharides lower blood glucose by reducing glucose absorption, increasing pancreatic beta-cell mass, and enhancing insulin signaling pathways, while terpenoids act as alpha-glucosidase inhibitors and as insulin sensitizers through PPARγ activation.^[1] Németh and colleagues (2025, International Journal of Molecular Sciences) reviewed edible medicinal mushrooms more broadly as candidate agents against insulin resistance, cataloguing their habitats, growing conditions and bioactive components alongside anti-diabetic, anti-hyperlipidemic, antioxidant and anti-inflammatory effects.^[2] Neither review organizes the evidence the way this article does. For the clinical evidence on specific mushroom species, mushrooms for type 2 diabetes: research guide and species overview covers the human trial data.

The Insulin Signaling Cascade: Where Mushroom Compounds Enter

The cascade normally runs like this. Insulin binds its receptor on the cell surface. The activated receptor phosphorylates insulin receptor substrate-1 (IRS-1), which recruits PI3K, which activates Akt. Akt signaling moves GLUT4 transporters — the channels that let glucose into the cell — from internal storage vesicles to the cell membrane. Glucose enters. Blood glucose falls.

Mushroom compounds enter this relay at four separable points:

1. At the receptor and IRS-1. Certain maitake fractions and the maitake glycoprotein SX-fraction restore activity at the top of the cascade, re-starting a relay that high glucose had stalled.

2. Via AMPK, on a parallel track. AMPK (adenosine monophosphate-activated protein kinase) is the cell's energy sensor. Activating it raises glucose uptake through a route that does not require a working insulin receptor — which is the appeal in insulin-resistant tissue. AMPK is not part of the insulin cascade; it is a separate pathway that converges on the same endpoint.

3. By inhibiting PTP1B. PTP1B is the phosphatase that terminates insulin receptor signaling by stripping off the activating phosphate groups. Inhibit it and the same amount of insulin produces a longer signal.

4. By reducing ceramides. Ceramides are lipids that accumulate in obesity and in insulin-responsive tissues, where they impair insulin sensitivity. Lowering them is an indirect route: it does not push the cascade, it removes something that was suppressing it.

Four Entry Points, One Relay

Where Each Mushroom Compound Meets the Insulin Cascade

Two mechanisms act on the relay itself. One runs on a parallel track that never touches the receptor. One acts on the cascade’s surroundings rather than the cascade. And one happens in the gut before insulin is involved at all. Grouping them as “mushrooms improve insulin signaling” hides every distinction that matters.

Before the cascade — in the gut

[13] [14]

Alpha-glucosidase inhibition

Ganoderma compounds SKG-3 (IC50 4.6 µg/mL) and ganoderol B (IC50 48.5 µg/mL) slow the enzyme that cleaves dietary carbohydrate into glucose. Less glucose arrives, so the cascade has less to do. Both are in vitro enzyme assays — neither tested whether eating the mushroom reproduces the effect.

1. 1

   Insulin receptor

   Insulin binds; receptor autophosphorylates

   Entry point 1^{[3] [5] [6]}

   Maitake F2 & F3; maitake SX-fraction (a glycoprotein)

   Both maitake fractions raised IR activity and phospho-IR at Tyr1361 in diabetic rats. SX-fraction reversed high-glucose inactivation of IR in L6 muscle cells.

2. 2

   IRS-1

   First relay protein; passes the signal down

   Entry point 1^{[3] [5] [10]}

   Maitake F2 & F3; SX-fraction; FYGL

   F2 and F3 lowered the inhibitory Ser307 phosphorylation on IRS-1; FYGL phosphorylated IRS-1 in L6 cells.

3. 3

   PI3K

   Recruited by IRS-1

   Downstream effect^{[3] [10]}

   Maitake F3 only; FYGL

   F3 raised PI3K and Akt mRNA. Its sibling fraction F2 did not — F2 went to PTP1B instead.

4. 4

   Akt

   Drives GLUT4 to the membrane

   No mushroom compound in this article was shown to act at Akt directly.

5. 5

   GLUT4 at the membrane

   Glucose enters the cell; blood glucose falls

   Shared terminus^{[8] [9] [10]}

   Oyster mushroom powder; Ganoderma extract; FYGL

   Measured directly in the AMPK studies. Notably NOT measured in the maitake F2/F3 study, despite being widely attributed to it.

Entry point 2 — parallel track

[7] [8] [9]

AMPK

Cordycepin (as cordycepin monophosphate), whole Ganoderma extract, oyster mushroom powder. AMPK is not a step in the insulin cascade — it is a separate energy-sensing pathway that reaches GLUT4 without passing through the receptor, IRS-1, PI3K or Akt. That is precisely why it is attractive in insulin-resistant tissue.

Converges on node 5 (GLUT4), bypassing nodes 1–4.

Entry point 3 — the off-switch

[10] [3]

PTP1B

FYGL, a Ganoderma proteoglycan, and maitake’s F2 fraction. PTP1B strips the activating phosphates back off the receptor and IRS-1. Inhibiting it does not add signal; it stops the signal being cancelled. Acts on nodes 1 and 2 by subtraction.

Caveat: FYGL also activated AMPK and raised GLUT4 — the categories leak.

Entry point 4 — the surroundings

[11]

Ceramide accumulation

Myriocin-containing Cordyceps extract inhibits serine palmitoyltransferase, the first committed step of ceramide synthesis. Ceramides accumulate in obesity and impair insulin sensitivity, so this route works by removing interference rather than by pushing the cascade.

Drawn beside the cascade, not on it: Akt phosphorylation was never measured in this study.

Reading the map

The numbered rail is the insulin relay itself; green boxes mark where a compound acted on it. The three coloured panels on the right are deliberately drawn off the rail, because that is what the studies show: AMPK runs parallel to it, PTP1B inhibition works by subtraction from it, and the ceramide route acts on the tissue environment around it. Every box on this map comes from a rodent or cell-culture study. None of these nodes has been measured in a human trial.

Source: Xiao 2015; Konno 2013; Preuss 2007; Hawley 2020; Lee 2020; Asrafuzzaman 2018; Yang 2018; Li 2022; Kim 2004; Fatmawati 2011 (see references)

Direct Pathway Activation: Maitake at the Receptor

Maitake (Grifola frondosa) has the most detailed insulin-signaling data of any single mushroom species, and it is also the species most often described inaccurately.

Xiao and colleagues (2015, Food & Function) tested two polysaccharide-enriched fractions, F2 and F3, in diabetic rats — not mice, as this study is frequently reported.^[3] Both fractions significantly lowered fasting serum glucose, fasting serum insulin and HOMA-IR against diabetic controls (P < 0.05). Both increased insulin receptor activity and mRNA, raised phosphorylated IR at Tyr1361, and lowered the inhibitory IRS-1 phosphorylation at Ser307.

Then the two fractions diverge, and this is the part usually lost in summary. In hepatic tissue, F3 raised PI3K and Akt mRNA — the classical downstream route. F2 did not; instead it suppressed PTP1B expression, which is the off-switch mechanism described later in this article. The authors' own conclusion was cautious: F2 and F3 are presumed to improve insulin resistance, triggered by reactivation of IR and IRS-1.

One Citation, Two Mechanisms

Where Maitake F2 and F3 Stop Agreeing

Xiao et al. (2015) tested two polysaccharide fractions from the same mushroom in the same diabetic rats and measured the same markers. The fractions behave identically for six rows and then split cleanly. Summaries that cite “maitake F2 and F3” as one finding are averaging away the most informative part of the study.

Marker-by-marker comparison of maitake polysaccharide fractions F2 and F3 in diabetic rats, showing six shared results followed by three divergent results and one marker never measured.

Marker	Fraction F2	Fraction F3
Fasting serum glucoseOutcome measure	Decreased (P < 0.05)	Decreased (P < 0.05)
Fasting serum insulinOutcome measure	Decreased (P < 0.05)	Decreased (P < 0.05)
HOMA-IRInsulin resistance index	Decreased (P < 0.05)	Decreased (P < 0.05)
Insulin receptorActivity and mRNA	Increased	Increased
Phospho-IR (Tyr1361)Activating phosphorylation	Increased	Increased
IRS-1 (Ser307)Inhibitory phosphorylation	Decreased	Decreased
IRS-1 mRNAFirst divergence↓ the fractions diverge	Not increased	Increased
PI3K and Akt mRNAThe classical downstream route	Not increased	Increased
PTP1B expressionThe off-switch	Inhibited	Not reported
GLUT4The endpoint everyone attributes to this study	Never measured	Never measured

F2’s route

Re-activates the receptor and IRS-1, then works by subtraction — suppressing PTP1B, the phosphatase that cancels the insulin signal. F2 belongs in the off-switch section of this article, not the PI3K section.

F3’s route

Re-activates the receptor and IRS-1, then works by amplification — raising IRS-1, PI3K and Akt expression. This is the classical downstream cascade, and it is F3’s result alone.

Two things to carry away

First, the model was diabetic rats, and the pathway measurements were made in hepatic tissue — not the muscle tissue most people picture when they think about glucose uptake. Second, the authors reported the glucose, insulin and HOMA-IR reductions as statistically significant rather than as percentages; the “42–45%” figure widely attributed to this study does not appear in the published abstract, which is why it does not appear in this article either.

Source: Xiao et al., 2015, Food & Function (PMID 26311233)

Two clarifications about what this study did not show. It did not measure GLUT4 — the endpoint most often attributed to it. And it did not report a percentage reduction in blood glucose; the widely circulated "42–45%" figure for these fractions does not appear in the published abstract and we could not verify it against the source, so we do not repeat it.

The maitake SX-fraction is a separate story that is often merged with the F2/F3 work. SX-fraction is a glycoprotein, not a beta-glucan. Konno and colleagues showed in cultured skeletal muscle L6 myotubes that a high-glucose environment (35 mM) inactivated the insulin receptor, IRS-1 and Akt, and that SX-fraction reversed that inactivation and restored glucose uptake, with insulin as a positive control producing near-identical results.^[5] The same group's 2020 review in World Journal of Diabetes summarized the animal and limited clinical evidence for SX-fraction alongside this in vitro mechanism.^[4] That review is a narrative synthesis by the investigators who produced the underlying data, not an independent systematic review, and it should be read with that in mind.

Preuss and colleagues (2007) tested the same SX-fraction in spontaneously hypertensive rats — a hypertension model rather than a diabetes model.^[6] SX-fraction lowered circulating glucose after a glucose challenge without raising circulating insulin, and improved sensitivity to injected insulin, comparably to the antidiabetic drug pioglitazone. No signaling markers were measured in that study; the insulin-sensitizing effect was inferred from the glucose and insulin curves.

For the maitake-specific clinical and mechanistic evidence, maitake SX-fraction and insulin sensitivity covers this compound in depth.

AMPK: A Parallel Track, Not the Insulin Cascade

In type 2 diabetes the insulin receptor is desensitized. AMPK activation offers a way around it: it raises glucose uptake through a route that does not depend on a responsive insulin receptor. It is worth being precise that AMPK is not a step in the insulin pathway — it is a separate energy-sensing pathway that converges on the same downstream effector, GLUT4.

The most mechanistically precise study in this entire literature is Hawley and colleagues (2020) in Cell Chemical Biology — and it corrects a claim repeated almost everywhere else.^[7]

Mechanism Correction

Does cordycepin activate AMPK directly?

The claim

“Cordycepin is an AMP mimetic that binds the AMPK gamma subunit directly, switching the enzyme on the way low cellular energy does.”

A compression of the Hawley 2020 finding that has propagated through supplement marketing and secondary summaries.

What research shows

Misleading

Cordycepin is a pro-drug. It enters cells through adenosine transporters and is metabolized into cordycepin mono-, di- and triphosphates. AMPK activation correlates with the level of cordycepin monophosphate specifically, and not with cordycepin itself or the other nucleotides. It is the monophosphate metabolite that mimics AMP.

The gamma subunit is still the site of action — activation in cells was blocked by a gamma-subunit mutation that prevents AMP activation, and in cell-free assays cordycepin monophosphate reproduced all three effects of AMP on AMPK. The correction is about which molecule does the binding, and it matters because it makes the effect dependent on cellular uptake and phosphorylation, not on how much cordycepin is in the capsule.

Hawley et al., 2020, Cell Chemical Biology (PMID 31991096)

The Missing Two Steps

Cordycepin Is a Pro-Drug, Not an AMP Mimic

The usual summary jumps straight from the capsule to the target: “cordycepin binds the AMPK gamma subunit.” Hawley et al. (2020) showed there are two events in between, and that the molecule which actually mimics AMP is one the label never mentions.

1. 01

   Cordycepin in the extract

   3'-deoxyadenosine

   The major bioactive agent of Cordyceps militaris. At this stage it is inert with respect to AMPK.

2. 02

   Transport into the cell

   via adenosine transporters

   Cordycepin does not diffuse in. It uses the same transporters as adenosine, so uptake depends on transporter availability rather than on dose alone.

   Failure point: no transport, no effect.

3. 03

   Phosphorylation by cellular metabolism

   mono- / di- / triphosphate

   The cell converts cordycepin into all three phosphorylated forms. At high concentrations these can almost replace the cell’s own adenine nucleotides.

   Failure point: no phosphorylation, no effect.

4. 04

   Cordycepin monophosphate

   the actual AMP mimic

   AMPK activation correlates with this metabolite specifically — not with cordycepin itself, and not with the di- or triphosphate. This is the molecule that does the work.

5. →

   AMPK gamma subunit

   In cell-free assays, cordycepin monophosphate reproduced all three effects of AMP on AMPK. In intact cells, activation was abolished by a gamma-subunit mutation that specifically prevents activation by AMP — two independent lines of evidence converging on the same binding site.

Why the correction matters

A direct binder’s effect scales with how much you take. A pro-drug’s effect depends on transport capacity and on the cell’s phosphorylating machinery — both saturable, both tissue-dependent, neither controlled by the dose on the label. The correction moves the bottleneck from the bottle to the cell.

The study’s own caution

At high concentrations cordycepin nucleotides can almost replace the cell’s native adenine nucleotides, and knocking out AMPK made cells more sensitive to cordycepin’s cytotoxic effects. AMPK activation here appears partly protective against the compound itself — not a detail that survives into marketing copy.

Source: Hawley et al., 2020, Cell Chemical Biology (PMID 31991096) — cultured cells and cell-free assays only; no animal or human glucose endpoint

Mechanism

Hawley et al. (2020) · Cell Chemical Biology

Population

Cultured cells and cell-free AMPK assays; no animals, no humans

Intervention

Cordycepin (3'-deoxyadenosine), the major bioactive agent of Cordyceps militaris

Comparator

AMP; AMPK-knockout cells; a gamma-subunit mutation blocking AMP activation

Finding

Cordycepin is converted intracellularly to cordycepin monophosphate, and AMPK activation tracks that metabolite rather than cordycepin itself. Cordycepin monophosphate reproduced all three effects of AMP on AMPK in cell-free assays.

positive Definitive on mechanism, silent on outcome — this study establishes how AMPK is activated, not that taking Cordyceps improves anyone's blood sugar.

Hawley et al., 2020, Cell Chemical Biology (PMID 31991096)

Ganoderma lucidum reaches AMPK by a route that has not been resolved to a specific compound. Lee and colleagues (2020, Nutrients) fed C57BL/6 mice a high-fat diet supplemented with 1%, 3% or 5% Ganoderma extract powder for 12 weeks.^[8] The 3% and 5% groups gained less liver and fat tissue weight, showed improved glucose and insulin tolerance, and had higher adiponectin. AMPK activation was increased, and with it the expression and activation of insulin receptor, IRS-1, Akt and GLUT1/4, alongside suppression of the lipogenic genes FAS, SCD1 and SREBP1c and reduced hepatic lipid accumulation.

Two caveats on that study. The intervention is a whole extract powder at 1–5% of total diet, so no single compound can be credited. And what fell in the liver was lipid accumulation and lipogenic gene expression — hepatic glucose output was not measured, despite being commonly attributed to this work.

The AMPK-to-GLUT4 link has been shown more directly elsewhere. Asrafuzzaman and colleagues (2018) fed oyster mushroom powder at 5% of diet to streptozotocin-induced diabetic rats for eight weeks and measured the two markers directly: phosphorylated AMPK rose in muscle and adipose tissue, GLUT4 mRNA rose significantly in both, and serum glucose fell.^[9]

For more on how AMPK activation from mushrooms compares across species, Chaga and AMPK activation: the insulin sensitivity connection covers Chaga's contribution to this pathway.

PTP1B Inhibition: Removing the Off-Switch

The insulin receptor is self-limiting. Once active it sets in motion its own shutdown, largely through PTP1B, a phosphatase that removes the activating phosphate groups from the receptor and from IRS-1. Inhibiting PTP1B has been a target of pharmaceutical antidiabetic research for decades.

Yang and colleagues (2018, Food & Function) characterized one compound here, not two: FYGL (Fudan-Yueyang Ganoderma lucidum), a proteoglycan their group had previously shown to competitively inhibit PTP1B in vitro.^[10] In ob/ob mice, FYGL lowered blood glucose, reduced body weight and ameliorated insulin resistance, with reduced PTP1B expression and increased phosphorylation of PTP1B's targets in skeletal muscle. In rat L6 myoblasts, FYGL entered cells readily, raised insulin-stimulated glucose uptake, suppressed PTP1B at the mRNA level, phosphorylated IRS-1, and activated PI3K and Akt.

Mechanism

Yang et al. (2018) · Food & Function

Population

ob/ob mice and PTP1B-transfected rat L6 myoblasts

Intervention

FYGL, a proteoglycan extracted from Ganoderma lucidum

Comparator

Untreated ob/ob controls; non-transfected L6 cells

Finding

FYGL lowered blood glucose, body weight and insulin resistance in ob/ob mice, reduced PTP1B expression, and in L6 cells raised insulin-stimulated glucose uptake while phosphorylating IRS-1 and activating PI3K and Akt.

preliminary Single research group, single proteoglycan, animal and cell models only. The compound is not a commercially available product.

Yang et al., 2018, Food & Function (PMID 29215104)

The tidy story would be that PTP1B inhibition and AMPK activation are two clean, separate mechanisms. This study undermines that story usefully: FYGL also raised AMPK phosphorylation and consequently upregulated GLUT4, promoting its transport to the plasma membrane. A compound entering the cascade at the off-switch also lit up the parallel pathway. Mechanistic categories are analytical conveniences here, not walls.

Maitake's F2 fraction belongs in this section too. As noted above, F2 suppressed PTP1B expression in the Xiao study while its sibling fraction F3 worked through PI3K and Akt.^[3]

The Ceramide Route: Actually a Myriocin Finding

Li and colleagues (2022, Scientific Reports) reported the most surprising result in this literature, and the one whose framing is most often lost.^[11]

Ceramides accumulate in insulin-responsive tissues in obesity and contribute to impaired insulin sensitivity. Serine palmitoyltransferase (SPT) catalyses the first committed step of ceramide synthesis, and inhibiting SPT with a compound called myriocin reduces body weight and improves insulin sensitivity in rodents. Myriocin was originally isolated from the fungus Isaria sinclairii and is abundant in closely related species including Cordyceps.

That is the actual chain of reasoning, and it changes the practical reading. The researchers screened commercially available Cordyceps extracts that people are already buying and found they contain myriocin in variable amounts. Treating obese mice with a human-equivalent dose of a myriocin-containing Cordyceps extract reduced ceramide accrual, increased energy expenditure, prevented diet-induced obesity, improved glucose homeostasis and resolved hepatic steatosis.

Two things this study did not show, both frequently attributed to it. It did not measure Akt phosphorylation, so the intuitive story that lowering ceramides "unblocks Akt" is background biology rather than a finding of this paper. And the mechanisms the authors did attribute the benefit to were increased adipose tissue browning, improved brown adipose function, improved hepatic insulin sensitivity, and shifts in gut microbes including Clostridium and Bilophila — a broader and messier picture than a single unblocked signaling step. The authors also note plainly that myriocin itself is not approved for human use.

Gut-Level Mechanisms: Alpha-Glucosidase Inhibition

Not all of mushrooms' blood sugar effects run through insulin signaling. Alpha-glucosidase is the intestinal enzyme that cleaves complex carbohydrates into absorbable glucose; inhibiting it slows glucose entry into circulation and blunts the post-meal spike. This is the mechanism of the drug acarbose, and it operates entirely before insulin signaling begins.

Two Ganoderma studies establish this directly. Kim and colleagues (2004) isolated SKG-3 from Ganoderma lucidum fruiting bodies and characterized it as a highly specific, reversible, competitive alpha-glucosidase inhibitor with an IC50 of 4.6 µg/mL and no activity against other glycosidases tested.^[13] Fatmawati and colleagues (2011) isolated ganoderol B from the same species, with an alpha-glucosidase IC50 of 48.5 µg/mL; the neutral fraction it came from inhibited the enzyme more strongly (IC50 88.7 µg/mL) than the acarbose positive control (IC50 336.7 µg/mL) in that assay.^[14] Both are in vitro enzyme assays; neither tested whether eating the mushroom reproduces the effect.

Beta-glucan structure matters here, and the evidence for that is narrower than usually claimed. Kim and colleagues (2005) prepared beta-glucans of 30–50 kDa from Agaricus blazei and hydrolysed them with a bacterial endo-beta-(1→6)-glucanase into mainly di- and tri-saccharides.^[12] In diabetic rats, both the intact beta-glucans and the oligosaccharides showed anti-hyperglycemic, anti-hypertriglyceridemic, anti-hypercholesterolemic and anti-arteriosclerotic activity — but the smaller oligosaccharides had roughly twice the anti-diabetic activity of the intact beta-glucans. Processing and molecular weight change the effect size substantially. Note that this study measured whole-animal outcomes only: it did not run an alpha-glucosidase assay and it did not measure insulin secretion, both of which are sometimes attributed to it.

Aramabašić Jovanović et al. (2021) place these gut-level mechanisms alongside the intracellular ones in their review, with terpenoids as the compound class most associated with alpha-glucosidase inhibition.^[1] For the postprandial side of this picture, mushroom dietary fiber and postprandial blood sugar spikes covers the human trial data on adding mushrooms to meals.

Where the Evidence Stops

Every mechanism described above was measured in rats, mice or cultured cells. That is not a hedge; it is the literal scope of the reference list for this article.

Evidence Reach

1. Cell and cell-free studiescompleted

   Strongest tier here. Hawley 2020 resolved cordycepin's AMPK mechanism to the level of a specific metabolite; Konno 2013 and Yang 2018 showed receptor, IRS-1 and Akt effects in L6 muscle cells; Kim 2004 and Fatmawati 2011 characterized alpha-glucosidase inhibition with IC50 values.

2. Animal studiescompleted

   Well populated. Xiao 2015 (diabetic rats), Lee 2020 and Li 2022 (high-fat-fed mice), Yang 2018 (ob/ob mice), Asrafuzzaman 2018 and Kim 2005 (diabetic rats), Preuss 2007 (hypertensive rats). Pathway markers were measured alongside glucose outcomes.

3. Small human trialscurrent

   Few and heterogeneous. Hsu 2007 randomized 72 patients on metformin and gliclazide to Agaricus blazei extract or placebo for 12 weeks and found lower HOMA-IR and higher adiponectin. Trials exist for Agaricus, Coprinus, Ganoderma, Grifola and Pleurotus, but none measured PI3K, Akt, AMPK, PTP1B or ceramide.

4. Large randomized trialsmissing

   None for isolated mushroom fractions at doses corresponding to the animal data.

5. Meta-analysis of pathway endpointsmissing

   Not possible — the constituent trials do not exist.

The mechanistic tier is unusually detailed for a natural-products literature, and the human tier is unusually thin. The gap between them is the honest headline: we know how these compounds act on cells better than we know whether they help people.

Hawley 2020; Xiao 2015; Yang 2018; Li 2022; Lee 2020; Konno 2013; Kim 2004; Fatmawati 2011; Kim 2005; Asrafuzzaman 2018; Hsu 2007; Lindequist 2020 (see references)

Lindequist and colleagues (2020) reviewed the clinical trial evidence directly and stated the imbalance plainly: several medicinal mushrooms show hypoglycemic activity in vitro and in animal studies, but in contrast to the high number of experimental results, only a few clinical trials or case reports have been published — for mushrooms of the genera Agaricus, Coprinus, Ganoderma, Grifola and Pleurotus.^[15]

The best-designed of those is Hsu and colleagues (2007), a randomized, double-blind, placebo-controlled trial in 72 Chinese patients with type 2 diabetes already taking metformin and gliclazide.^[16] After 12 weeks on 1,500 mg daily of Agaricus blazei extract, the treatment group had a significantly lower HOMA-IR than placebo (3.6 vs 6.6, p = 0.04) and a 20% rise in plasma adiponectin while the placebo group fell 12%. That is a real human insulin-sensitivity signal — in an add-on setting, in one population, at one dose, with no pathway markers measured.

So the honest summary is a two-part statement. The mechanistic evidence is genuine, specific and in places excellent — Hawley 2020 is a better piece of molecular pharmacology than most natural-product mechanism papers ever produce. And the inference that these same pathways drive the modest effects seen in the few human trials remains an inference, because no human trial has measured them.

Practical Safety

Moderate

Blood-sugar-active mushrooms alongside diabetes medication

Concern

Several mechanisms described here overlap with prescription drugs. Metformin also acts through AMPK; acarbose also inhibits alpha-glucosidase; the one good human trial here tested Agaricus blazei on top of metformin and gliclazide, a sulfonylurea that can cause hypoglycemia on its own. Adding a blood-sugar-active supplement to that regimen is a pharmacological change, not a neutral addition.

What the research says

Anyone taking glucose-lowering medication should treat mushroom supplements as something to discuss with their prescriber and to monitor for, rather than something to start quietly. Increased glucose monitoring during the first weeks is the practical safeguard, since any additive effect will show up there first.

Particularly relevant for: People with type 1 or type 2 diabetes, prediabetes, or reactive hypoglycemia; anyone on insulin, sulfonylureas, metformin or alpha-glucosidase inhibitors

What to do

Do not stop, reduce or substitute prescribed diabetes medication on the basis of mechanism research in rodents. This article is research education, not medical advice.

Hsu et al., 2007 (PMID 17309383); Lindequist et al., 2020 (PMID 33389851)

For more on how specific mushroom species perform in human insulin resistance research, Agaricus blazei and insulin resistance: human trial evidence provides one of the more direct human data sets in this area.

The Bottom Line

Mushroom compounds act on blood sugar at four separable points, and the compounds are not interchangeable. Maitake fractions and the maitake glycoprotein SX-fraction restore activity at the insulin receptor and IRS-1 — though maitake's F2 and F3 fractions diverge downstream, F3 through PI3K/Akt and F2 through PTP1B suppression. Cordycepin activates AMPK on a parallel track, but only after cells convert it to cordycepin monophosphate; cordycepin itself is a pro-drug, not a direct AMP mimic. A single Ganoderma proteoglycan, FYGL, inhibits PTP1B — while also activating AMPK, which shows how porous these categories are. And the Cordyceps ceramide result is really a myriocin result, dependent on a compound present at variable and unlabelled levels in commercial extracts. Separately, alpha-glucosidase inhibition by Ganoderma compounds acts in the gut, before insulin signaling begins. All of this was measured in rodents and cell lines. Human trials in this area are few, none has measured these pathway markers, and the best of them found a modest insulin-sensitivity benefit from Agaricus blazei added on top of existing diabetes medication.

Frequently Asked Questions

What are beta-glucans and how do they affect blood sugar?

Beta-glucans are polysaccharides — long-chain sugar polymers — found in the cell walls of mushrooms and certain grains. In animal studies, mushroom beta-glucan fractions have lowered fasting glucose, fasting insulin and HOMA-IR, with effects traced to reactivation of the insulin receptor and IRS-1 and, depending on the fraction, either increased PI3K/Akt expression or suppressed PTP1B (Xiao et al., 2015). Molecular weight matters: hydrolysing Agaricus blazei beta-glucans into di- and tri-saccharides roughly doubled their anti-diabetic activity in diabetic rats (Kim et al., 2005). Note that two of the most cited mushroom compounds in this area — SX-fraction and FYGL — are not beta-glucans at all; they are a glycoprotein and a proteoglycan respectively.

Do beta-glucans from mushrooms lower blood glucose?

In animal models, yes. Maitake fractions F2 and F3 significantly lowered fasting serum glucose, fasting insulin and HOMA-IR versus diabetic controls in rats (Xiao et al., 2015); Agaricus blazei beta-glucans and their hydrolysed oligosaccharides showed anti-hyperglycemic activity in diabetic rats (Kim et al., 2005). We deliberately do not quote the "42–45% glucose reduction" figure that circulates for the maitake fractions — it does not appear in the published abstract and we could not verify it. In humans the evidence is much thinner: few trials exist, and almost none test isolated beta-glucan fractions at doses corresponding to the animal data (Lindequist et al., 2020).

How do mushroom compounds activate AMPK?

The best-characterized case is cordycepin from Cordyceps militaris, and the common description of it is wrong. Cordycepin does not bind AMPK directly. Hawley et al. (2020) showed it enters cells via adenosine transporters and is metabolized to cordycepin mono-, di- and triphosphates; AMPK activation correlates specifically with cordycepin monophosphate, which mimics AMP at the gamma subunit. Cordycepin is a pro-drug. Separately, whole Ganoderma lucidum extract raised AMPK activation in high-fat-fed mice without the responsible compound being identified (Lee et al., 2020), and oyster mushroom powder raised phosphorylated AMPK and GLUT4 expression in muscle and adipose tissue of diabetic rats (Asrafuzzaman et al., 2018).

Which mushroom has the most beta-glucan?

Beta-glucan content varies by species, growing conditions, substrate and extraction method, and it is not standardized across commercial products — most labels do not specify which fractions are present or at what concentration. Shiitake, oyster, maitake and reishi all contain substantial beta-glucan fractions. For insulin signaling specifically, the deepest research base belongs to maitake and Ganoderma lucidum, but be aware that the most-studied maitake and Ganoderma compounds in this area (SX-fraction, FYGL) are not beta-glucans. Highest beta-glucan content and best blood sugar evidence are not the same question.

Do beta-glucans improve insulin sensitivity in humans?

The human evidence is limited but not absent. The strongest single result is Hsu et al. (2007): a randomized, double-blind, placebo-controlled trial in 72 patients with type 2 diabetes already on metformin and gliclazide, where 1,500 mg daily of Agaricus blazei extract for 12 weeks produced a significantly lower HOMA-IR than placebo (3.6 vs 6.6, p = 0.04) and a 20% rise in adiponectin. Beyond that, Lindequist et al. (2020) found only a small number of clinical trials and case reports across the Agaricus, Coprinus, Ganoderma, Grifola and Pleurotus genera. None of these trials measured the signaling markers described in this article, so the link between the rodent mechanisms and the human results is an inference.

How do beta-glucans compare to medication for blood sugar?

The research does not support a direct comparison. Pharmaceutical antidiabetic drugs have large human trial evidence bases with precisely measured effect sizes; mushroom research in this area is mostly preclinical, with limited human data and no head-to-head trials. Some mechanisms genuinely overlap — metformin also acts through AMPK, acarbose also inhibits alpha-glucosidase — but shared mechanism does not imply comparable effect size, and an in vitro IC50 comparison against acarbose (as in Fatmawati et al., 2011) is an enzyme assay result, not a clinical one. Notably, the one good human trial in this space tested a mushroom extract added to metformin and gliclazide, not instead of them. Nothing here supports replacing prescribed diabetes medication, and this article is research education rather than medical advice.

Sources & References

1. 1RTHM-01322·Aramabašić Jovanović, Jelena et al. (2021). “The Effects of Major Mushroom Bioactive Compounds on Mechanisms That Control Blood Glucose Level..” Journal of fungi (Basel.Study breakdown →PubMed →↩
2. 2RTHM-02157·Németh, Zsuzsanna et al. (2025). “The Importance of Edible Medicinal Mushrooms and Their Potential Use as Therapeutic Agents Against Insulin Resistance..” International journal of molecular sciences.Study breakdown →PubMed →↩
3. 3RTHM-00410·Xiao, Chun et al. (2015). “Hypoglycemic effects of Grifola frondosa (Maitake) polysaccharides F2 and F3 through improvement of insulin resistance in diabetic rats..” Food & function.Study breakdown →PubMed →↩
4. 4RTHM-01222·Konno, Sensuke (2020). “SX-fraction: Promise for novel treatment of type 2 diabetes..” World journal of diabetes.Study breakdown →PubMed →↩
5. 5RTHM-00533·Konno, Sensuke et al. (2013). “Possible hypoglycemic action of SX-fraction targeting insulin signal transduction pathway..” International journal of general medicine.Study breakdown →PubMed →↩
6. 6RTHM-00229·Preuss, Harry G et al. (2007). “Enhanced insulin-hypoglycemic activity in rats consuming a specific glycoprotein extracted from maitake mushroom..” Molecular and cellular biochemistry.Study breakdown →PubMed →↩
7. 7RTHM-00618·Hawley, Simon A et al. (2020). “Mechanism of Activation of AMPK by Cordycepin..” Cell chemical biology.Study breakdown →PubMed →↩
8. 8RTHM-01228·Lee, Hyeon A et al. (2020). “Ganoderma lucidum Extract Reduces Insulin Resistance by Enhancing AMPK Activation in High-Fat Diet-Induced Obese Mice..” Nutrients.Study breakdown →PubMed →↩
9. 9RTHM-00939·Asrafuzzaman, Mohammad et al. (2018). “Oyster mushroom functions as an anti-hyperglycaemic through phosphorylation of AMPK and increased expression of GLUT4 in type 2 diabetic model rats..” Journal of Taibah University Medical Sciences.Study breakdown →PubMed →↩
10. 10RTHM-01022·Yang, Zhou et al. (2018). “A novel PTP1B inhibitor extracted from Ganoderma lucidum ameliorates insulin resistance by regulating IRS1-GLUT4 cascades in the insulin signaling pathway..” Food & function.Study breakdown →PubMed →↩
11. 11RTHM-00779·Li, Ying et al. (2022). “Cordyceps inhibits ceramide biosynthesis and improves insulin resistance and hepatic steatosis..” Scientific reports.Study breakdown →PubMed →↩
12. 12RTHM-00082·Kim, Yea-Woon et al. (2005). “Anti-diabetic activity of beta-glucans and their enzymatically hydrolyzed oligosaccharides from Agaricus blazei..” Biotechnology letters.Study breakdown →PubMed →↩
13. 13RTHM-00138·Kim, Shin-Duk et al. (2004). “Isolation and characterization of alpha-glucosidase inhibitor from the fungus Ganoderma lucidum..” Journal of microbiology (Seoul.Study breakdown →PubMed →↩
14. 14RTHM-00392·Fatmawati, Sri et al. (2011). “Ganoderol B: a potent α-glucosidase inhibitor isolated from the fruiting body of Ganoderma lucidum..” Phytomedicine : international journal of phytotherapy and phytopharmacology.Study breakdown →PubMed →↩
15. 15RTHM-00632·Lindequist, Ulrike et al. (2020). “Medicinal Mushrooms for Treatment of Type 2 Diabetes: An Update on Clinical Trials..” International journal of medicinal mushrooms.Study breakdown →PubMed →↩
16. 16RTHM-00114·Hsu, Chung-Hua et al. (2007). “The mushroom Agaricus Blazei Murill in combination with metformin and gliclazide improves insulin resistance in type 2 diabetes: a randomized, double-blinded, and placebo-controlled clinical trial..” Journal of alternative and complementary medicine (New York.Study breakdown →PubMed →↩