Direct answer
Does L-ergothioneine support memory or cognitive health?
The human evidence is suggestive but inconclusive.
Prospective observational studies have associated higher circulating ergothioneine levels with more favorable cognitive or dementia outcomes, while lower levels have been associated with poorer cognition and faster cognitive decline. These studies do not show that supplementation caused those differences.
Two randomized supplementation trials have also reported selected signals. A 19-person pilot involving adults with mild cognitive impairment reported findings in learning performance and a neurodegeneration-related biomarker. A larger 147-person trial did not improve its primary composite-memory outcome versus placebo, although it reported selected secondary or exploratory findings.
Bottom line: The evidence supports further research, but it does not establish that L-ergothioneine supplementation improves memory, preserves cognitive health, treats mild cognitive impairment, or prevents dementia.
This draft separates a null primary outcome from secondary or exploratory findings and separates blood-level associations from supplementation effects.
Status before human approval
Evidence verdict
Publication approval not yet granted. Direction, evidence stage, and certainty remain unassigned until the human reviewer accepts or changes the analysis.
- Overall result
- Review in progress
- Direction
- Not yet assigned
- Evidence stage
- Not yet assigned
- Certainty
- Not yet assigned
- Human supplementation evidence
- Two randomized trials identified
- Independent replication
- Not established
- Last evidence search
- September 20, 2026
- Review version
- Draft 0.10
Claims being reviewed
These statements are not equivalent. A biomarker finding cannot establish treatment, and an association cannot establish prevention.
| Claim variant | Claim type | Current boundary |
|---|---|---|
| Supports cognitive health | Broad wellness claim | Requires qualification; direct supplementation evidence is preliminary. |
| Improves memory or learning | Functional claim | Not established consistently versus placebo on prespecified outcomes. |
| Slows cognitive decline | Disease-progression claim | Not established by the available trials. |
| Prevents dementia or Alzheimer’s disease | Prevention claim | Unsupported; prospective associations do not prove prevention. |
| Treats mild cognitive impairment | Treatment claim | Unsupported; one 19-person pilot cannot establish treatment efficacy. |
| Protects brain cells or reduces neurodegeneration | Mechanism or biomarker claim | Biologically plausible, but biomarkers and models do not establish a patient-relevant benefit. |
Why cognitive benefit has been proposed
The case begins with biological plausibility: ergothioneine is actively transported and has antioxidant and cellular-protection chemistry. Preclinical work has investigated brain distribution, oxidative stress, neuronal models, and learning or memory outcomes in animals.
Human interest increased because lower circulating ergothioneine has been associated with cognitive impairment and later decline in prospective cohorts. Two randomized trials then tested direct supplementation. That progression makes cognition a reasonable research question, but it does not make every step evidence of the same claim.
Strongest supporting case
A reasonable supporting advocate would argue that several evidence streams point in the same direction. Prospective cohorts in different populations associate higher circulating ergothioneine with more favorable cognitive or dementia outcomes. A 19-person randomized pilot in mild cognitive impairment reported signals in learning and neurofilament light. A later 147-person randomized trial confirmed dose-dependent exposure and reported selected reaction-time, subjective prospective-memory, and sleep findings.
The advocate would also note that the larger trial enrolled relatively high-functioning adults, had a 16-week duration, and included participants with comparatively high baseline ergothioneine. Those features could reduce the room or time available to detect cognitive change. Together, the data justify better targeted and longer trials.
This is the strongest reasonable case—not a conclusion that the claim has been proved.
Evidence by category
Two trials were identified. The larger trial did not show a sustained or between-group benefit on its primary composite-memory outcome. Both trials reported selected exploratory, secondary, subjective, or biomarker signals.
No additional isolated-ergothioneine intervention was identified that independently replicates a clinically meaningful cognitive outcome.
Two prospective cohorts support an association between circulating ergothioneine and later cognitive or dementia outcomes. They do not show that supplementation causes the difference.
Animal models support biological plausibility and include learning or memory observations. Species, experimental exposure, and model conditions limit translation.
Transport, antioxidant chemistry, and cellular-stress work explain why the question is being studied. They cannot determine a patient-relevant cognitive benefit.
Evidence docket
The docket emphasizes the registered or stated primary question, direct comparisons, limitations, and independence—not the number of positive phrases in an abstract.
| Study | Design and population | Intervention | Primary outcome or main question | Main result | Limitations and conflicts | Weight |
|---|---|---|---|---|---|---|
| Zajac et al. (2025) ↗Study record → | Randomized, double-blind, placebo-controlled Adults 55–79 with subjective memory complaints n=147; 16 weeks | 10 or 25 mg/day | Change in CNS Vital Signs composite memory at week 16 | No significant treatment effect on the primary composite-memory outcome. A 25 mg within-group improvement at week 4 was not sustained; selected secondary or subjective outcomes showed signals. | High-functioning cohort, short duration, multiple secondary analyses, several within-group findings, and no sustained primary-outcome benefit. Funded by Phyto Tech Corp (Blue California); one author was employed by Blue California and the tested ingredient was branded ErgoActive®. | Preliminary clinical; important but not confirmatory |
| Yau et al. (2024) ↗Study record → | Randomized, double-blind, placebo-controlled pilot Adults 60+ with mild cognitive impairment n=19; 1 year | 25 mg three times weekly | Cognitive, safety, and neurodegeneration-related measures | The paper reported a learning-assessment signal and a different neurofilament-light trajectory versus placebo. | Extremely small pilot with multiple measures; inadequate power for prevention or treatment conclusions and no independent replication. Authors declared no potential conflicts; PubMed lists non-U.S. government research support. | Hypothesis-generating clinical evidence |
| Meng et al. (2025) ↗Study record → | Prospective observational cohort Community-dwelling Japanese adults 65+ without dementia n=1,344; Median 11.2 years | No intervention; baseline serum ergothioneine | Incident all-cause dementia and subtypes | Higher serum ergothioneine was associated with lower incidence of all-cause, Alzheimer, and non-Alzheimer dementia after multivariable adjustment. | Residual confounding and reverse causation remain possible; measured blood level is not a supplementation trial. Several authors were affiliated with Suntory Global Innovation Center; funding and complete disclosures require consideration. | Strong observational signal; no causal proof |
| Wu et al. (2022) ↗Study record → | Prospective observational memory-clinic cohort Older adults attending memory clinics in Singapore n=470; Up to 5 years | No intervention; baseline plasma ergothioneine | Cognitive and functional trajectories | Lower baseline plasma ergothioneine was associated with poorer baseline performance and faster decline in several domains. | Clinical cohort and observational design cannot establish causality or show that supplementation changes outcomes. The paper acknowledged provision of ergothioneine by Tetrahedron; full funding disclosures remain part of the independence assessment. | Supportive observational evidence |
Scientific cross-examination
Primary versus secondary outcomes
The 147-person trial’s primary composite-memory outcome was not significantly improved versus placebo at week 16. A temporary within-group change and selected secondary findings cannot replace that primary result.
Power and multiplicity
The 19-person trial is too small for a stable treatment estimate across numerous cognitive and biomarker measures. Testing many outcomes and time points raises the risk that some findings occur by chance.
Direct comparisons
Improvement from baseline within a treatment arm is not proof of benefit. The relevant efficacy question is whether change differs from placebo under the prespecified analysis.
Clinical significance
A reaction-time interaction, a subjective questionnaire change, or a neurofilament biomarker trajectory does not automatically translate into preserved independence, delayed dementia, or a noticeable cognitive benefit.
Causation and generalizability
Serum or plasma levels may reflect diet, health status, metabolism, or other factors. Memory-clinic cohorts and older Japanese community cohorts may not generalize to every supplement user.
Independence and replication
The larger trial was industry-funded and included an industry-employed author. The positive signals have not been independently replicated in a confirmatory trial designed around the same endpoint.
Supporting-side rebuttal
The primary outcome was null.
Supporting responseParticipants began near the high end of normal cognition and may have had limited room for improvement; longer follow-up or lower-baseline-status enrollment could be more sensitive.
The pilot was extremely small.
Supporting responseIts one-year duration, controlled design, learning signal, and biomarker result make it useful for planning—not dismissing—larger trials.
Observational studies cannot prove causation.
Supporting responseProspective timing, multivariable adjustment, and findings in more than one cohort reduce some alternative explanations, even though they cannot eliminate them.
Industry involvement lowers confidence.
Supporting responseIndustry funding does not invalidate a study. Registration, blinding, transparent methods, full reporting, and independent replication determine how much confidence it deserves.
Working adjudication
This is a question-by-question working assessment, not a final approved rating.
- Is cognitive benefit biologically plausible?
- Yes. Transport and preclinical evidence make the hypothesis reasonable.
- Is there relevant human evidence?
- Yes. Randomized trials and prospective cohorts are relevant.
- Has a supplementation benefit been demonstrated reliably?
- Not established. The larger trial’s primary memory outcome was null, and the pilot is too small for a definitive conclusion.
- Has the result been independently replicated?
- No confirmatory replication of the same clinically meaningful cognitive endpoint was identified.
- Is prevention or treatment of cognitive decline established?
- No. The available evidence does not establish prevention of dementia or treatment of mild cognitive impairment.
- Is additional research justified?
- Yes. Longer, adequately powered, preregistered trials in well-characterized populations are justified.
Provisional evidence boundary
Ergothioneine reaches the circulation; cognition is a legitimate research target; prospective associations and trial signals justify further study.
Whether supplementation produces a reproducible, clinically meaningful cognitive benefit, for whom, at what dose, and over what duration.
Prevention of dementia or Alzheimer’s disease, treatment of mild cognitive impairment, or reliable memory improvement in healthy adults.
Independent, adequately powered trials with prespecified cognitive outcomes, appropriate multiplicity control, direct between-group effects, and clinically meaningful follow-up.
Publication status: Human review is in progress. This evidence boundary is not a final approved ruling.
Website-language ruling
- “Ergothioneine is being studied for cognitive health.”
- “Human studies include prospective cohorts and two randomized supplementation trials.”
- “The evidence is preliminary and does not establish prevention or treatment.”
- “May support memory” only when immediately paired with the null primary outcome and lack of replication.
- “Associated with lower dementia risk” only when clearly labeled observational and noncausal.
- “Neuroprotective” only as a mechanistic or research description—not a demonstrated clinical effect.
- “Prevents dementia” or “prevents Alzheimer’s disease.”
- “Treats mild cognitive impairment.”
- “Clinically proven to improve memory.”
- “Reverses cognitive decline” or “repairs the brain.”
Safety context
The two cognitive trials reported tolerability and monitored selected clinical safety measures within their tested populations and durations. That information does not establish universal, indefinite, pregnancy, pediatric, interaction, or disease-specific safety.
The cognitive evidence also does not establish an optimal dose. Study doses are research exposures, not personalized recommendations. See the separate safety review and dosage guide.
Primary and authoritative sources
- Zajac et al. 2025 randomized trial ↗
- Yau et al. 2024 randomized pilot ↗
- Meng et al. 2025 Hisayama cohort ↗
- Wu et al. 2022 memory-clinic cohort ↗
- NCT03641404 trial registration ↗
- ACTRN12621001607864 trial registration ↗
Commercial pages were reviewed only to inventory claims, not to establish efficacy.
Transparency record
- First published
- August 22, 2026
- Last evidence search
- September 20, 2026
- Last substantive update
- September 27, 2026
- Review version
- Draft 0.10
- Evidence cutoff
- September 20, 2026
- Methodology
- Adversarial Evidence Review Pilot Pilot 1.0
- AI assistance
- Codex assisted with evidence discovery, extraction, comparison, adversarial questioning, and drafting. AI output was not treated as a source. No multi-model consensus is claimed.
- Human oversight
- Human review in progress; publication approval not yet granted.
- Ownership and conflicts
- Published by L-Ergothioneine.com. No active affiliate product links are included in this review. Study-level funding and conflicts are reported in the docket.
- Corrections and challenges
- Material corrections will be dated and explained. Evidence challenges should identify the disputed statement and a primary or authoritative source.
- Revision history
- Draft 0.10: revised the direct answer and key-evidence navigation; corrected the 2022 memory-clinic cohort attribution and DOI. Draft 0.9: expanded docket, primary-outcome analysis, cross-examination, rebuttal, and provisional language ruling.
Educational information only. This page does not diagnose, prevent, or treat cognitive impairment and does not provide personalized supplement advice.
Read the full disclaimer →