Atlas chapter 20 · Cognition

Tremella

Tremella fuciformis

E3++ — Supports the tested propositionR1. One evidence-owning Tremella-alone controlled cognition programme exists (the 2018 randomized, double-blind, placebo-controlled supplementation trial). Independent Tremella-alone replication remains absent or insufficient. A 2023 randomized, double-blind, placebo-controlled cognition study of a multi-component formula containing fish roe, snow fungus and yeast is not an independent replication of the Tremella-alone proposition: the formula owns its own evidence, and ingredient presence is not replication. Multiple dose arms, multiple outcomes, neuroimaging findings, systematic reviews, mechanistic evidence and further publications from the same programme likewise do not constitute independent Tremella-alone replication. R1 therefore sits beside E3+.

Proposition / claim: Placebo-relative improvement in subjective memory complaints and selected short-term-memory/executive-function measures in adults with subjective cognitive impairment
Evidence owner: 2018 randomized double-blind placebo-controlled supplementation trial: 600 or 1200 mg/day for eight weeks
Dominant uncertainty: Independent replication and effect robustness

Organism: Tremella
Tested intervention: Tremella fuciformis supplementation at 600 mg/day or 1200 mg/day for eight weeks
Evidence target: Standardized preparation
Intervention description: The evidence belongs first to the tested 2018 supplementation programme: Tremella fuciformis supplementation administered at 600 mg/day or 1200 mg/day for eight weeks in a randomized, double-blind, placebo-controlled trial. FA-047 does not establish strain identity, extract ratio, polysaccharide or beta-glucan concentration, fruiting-body versus mycelial composition, manufacturing equivalence, batch composition or cultivation conditions, and none of these are inferred here. Intervention-transfer boundary: studied preparation to a demonstrably equivalent preparation is MODERATE-HIGH only where manufacturing and composition are reproducible; studied preparation to a generic Tremella supplement is LOW-MODERATE without equivalence data; a Tremella-alone trial does not automatically transfer to a formula containing Tremella. Species identity is not proof of preparation equivalence.
Population: 75 community-dwelling adults aged 40-65 with subjective cognitive complaints / subjective cognitive impairment but without clinical cognitive impairment; 69 completed the programme. Subjective cognitive complaints are not clinically diagnosed cognitive impairment or dementia, and this population is neither an Alzheimer's disease population nor a general elderly population.
Comparator: Placebo, n = 15 at allocation. The placebo arm was half the size of either active arm (600 mg/day n = 30; 1200 mg/day n = 30), a materially small control arm that remains a limitation of the programme. Placebo composition is not inferred beyond FA-047.
Outcome: REGISTERED OUTCOME HIERARCHY. Trial registration NCT02377024 records as PRIMARY outcomes: (1) change from baseline in neurocognitive test batteries (CANTAB); (2) change in brain structure/morphometry; (3) change in brain function (BOLD fMRI); (4) change in brain metabolites (MRS). SMCQ change from baseline was registered as a SECONDARY outcome, as were adverse events at weeks 4 and 8. The registration was first posted 3 March 2015, approximately 13 months after enrolment began in February 2014; it is therefore RETROSPECTIVE relative to trial commencement and must not be described as a prospectively registered outcome hierarchy. PUBLICATION REPORTING / EMPHASIS. The 2018 publication reports placebo-relative improvement in the Subjective Memory Complaints Questionnaire (SMCQ) and in selected standardized short-term-memory and executive-function measures. FA-047 describes SMCQ as the primary endpoint and the standardized short-term-memory/executive-function measures as secondary. This differs from the registered hierarchy. SMCQ is NOT the trial's registered primary endpoint. Neuroimaging (including voxel-based morphometry) was registered within the primary outcome set but was treated and interpreted as exploratory in the publication and in the FA-047 presentation; it is therefore not simply an exploratory outcome, and the registration/publication divergence is itself part of the record. ATLAS EVIDENCE OWNERSHIP. Imaging biomarkers are not treated as the grade-owning clinical outcome merely because they were registered primary outcomes. The Atlas proposition continues to be owned by the participant-relevant cognitive findings (subjective memory complaints and selected short-term-memory/executive-function measures), reported as placebo-relative improvements rather than within-group change, on the explicit condition that the registered-versus-publication outcome-hierarchy discrepancy is disclosed as above. Exploratory-treated morphometric findings remain mechanistic observations and do not own the clinical efficacy proposition.
Evidence maturity (E): E3+
Direction: + — Supports the tested proposition
Historical flag (H): Not specified at dossier level
Replication (R): R1. One evidence-owning Tremella-alone controlled cognition programme exists (the 2018 randomized, double-blind, placebo-controlled supplementation trial). Independent Tremella-alone replication remains absent or insufficient. A 2023 randomized, double-blind, placebo-controlled cognition study of a multi-component formula containing fish roe, snow fungus and yeast is not an independent replication of the Tremella-alone proposition: the formula owns its own evidence, and ingredient presence is not replication. Multiple dose arms, multiple outcomes, neuroimaging findings, systematic reviews, mechanistic evidence and further publications from the same programme likewise do not constitute independent Tremella-alone replication. R1 therefore sits beside E3+.
Claim Gap (CG): Ladder. CG0 - DIRECT: an 8-week randomized placebo-controlled trial found greater improvement in subjective memory complaints and selected cognitive measures with the tested T. fuciformis supplementation programme than with placebo. CG1 - FAIR SYNTHESIS: preliminary controlled evidence suggests a defined Tremella intervention may influence selected cognitive outcomes in adults with subjective cognitive impairment. CG2-3: "Tremella improves memory and concentration" - preparation, population and outcome breadth have expanded. CG3: "Tremella protects the ageing brain" - duration and neuroprotection have appeared. CG4: "Tremella prevents dementia" - the existing trial does not establish this proposition. Claim ceiling: CG1 for preparation-specific cognition wording.
Evidence Transfer Vector: For the broad claim "Tremella prevents age-related cognitive decline". I: tested preparation to generic Tremella. P: adults aged 40-65 with subjective complaints to the ageing population generally. O: subjective memory complaints and selected test performance to prevention of cognitive decline. D: eight weeks to long-term prevention. R: one controlled programme to an established preventive effect.
Taxonomic Transfer Confidence: HIGH for the present evidence programme. Taxonomy is not the dominant uncertainty here. Stable species naming does not, however, establish preparation equivalence, and HIGH taxonomic transfer confidence must not be allowed to inflate intervention transfer confidence.
Intervention Transfer Confidence: LOW-MODERATE across unspecified preparations. Studied preparation to a demonstrably equivalent preparation: MODERATE-HIGH if manufacturing and composition are reproducible. Studied preparation to a generic Tremella supplement: LOW-MODERATE without equivalence data. Tremella-alone trial to a multi-component formula: not automatic.
Safety: S1-2
Conservation: C0/C+
Resolution zone: RZ-2-3
Next best experiment: Run a larger, independent, randomized, double-blind, placebo-controlled trial using the same or a demonstrably equivalent T. fuciformis preparation, with balanced allocation or a sufficiently large placebo group, one prespecified primary cognitive endpoint, and a clearly defined comparable population. Characterize the intervention and archive the batch. Avoid making exploratory neuroimaging or post-hoc subgroup findings carry the efficacy conclusion. Include post-treatment follow-up if feasible, and publish regardless of result. Investigator independence matters: a new group, new recruitment and new site testing the same essential proposition.
Primary references: GRADE-OWNING / PRIMARY / CONTROLLED POSITIVE — Ban S, Lee SL, Jeong HS, Lim SM, Park S, Hong YS, Kim JE. Efficacy and Safety of Tremella fuciformis in Individuals with Subjective Cognitive Impairment: A Randomized Controlled Trial. Journal of Medicinal Food. 2018;21(4):400-407. DOI 10.1089/jmf.2017.4063. PMID 29319408. Verified in full text: 88 screened, 75 randomized 2:2:1 to low-dose TF 600 mg/day (n=30), high-dose TF 1200 mg/day (n=30) or indistinguishable placebo (n=15), six capsules daily in three divided doses, 8 weeks; 69 completed (93.3% / 90.0% / 93.3%); compliance >90% in all arms. Community-dwelling adults aged 40-65 with subjective cognitive complaints and no clinical impairment; exclusions included MMSE <=24 or CDR >=0.5, axis I disorder, CNS medication within 3 months, IQ <=80, MRI contraindication. Analysis: intent-to-treat, mixed-effects models for repeated measures with treatment group, visit and group-by-visit interaction as fixed effects and the within-subject factor as random; SMCQ adjusted for age, sex, IQ and baseline; Stata v13.1; alpha 0.05 two-tailed. Placebo-relative contrasts: SMCQ high dose z=-3.16, P=.002, low dose z=-2.68, P=.007 (essentially unchanged with HDRS added: z=-3.25, P=.001 and z=-2.70, P=.007); short-term memory (CANTAB pattern-recognition-memory immediate recall + spatial span composite) high dose z=3.35, P=.001, low dose z=3.31, P=.001; executive function (WCST) high dose z=2.33, P=.02, low dose z=1.75, P=.08. No high- versus low-dose difference on any outcome. ANALYTICAL PROVENANCE: Cook's distance was estimated for each observation and observations with Cook's distance >0.058 were removed — four from the SMCQ analysis, three from short-term memory, four from executive function. The paper reports no prespecification of this procedure, no arm-of-origin for the excluded observations and no sensitivity analysis retaining them. The published placebo-relative findings are verified for the reported analysis, but robustness to the Cook's-distance exclusions is not established. OUTCOME-HIERARCHY DISCREPANCY: the publication presents SMCQ as primary and short-term memory, executive function and gray-matter volume as secondary; the registration (NCT02377024) registers CANTAB and multimodal neuroimaging as primary and SMCQ as secondary — see the registration entry. FUNDING: Ministry of Public Safety and Security (MPSS-Fire Fighting Safety-2016-86); National Research Foundation of Korea / Ministry of Science, ICT and Future Planning (2015M3C7A1028376); and a grant from Pulmuone Holdings. Author Disclosure Statement: "No competing financial interests exist." Whether Pulmuone supplied the capsules or took part in design, conduct, analysis, interpretation, manuscript preparation or the publication decision is NOT REPORTED. PREPARATION LIMITATION: the article names the intervention only as "TF" / TF extract and states that the detailed preparation method is described elsewhere (its reference 2). Species authentication, strain, fruiting body versus mycelium, extraction method, polysaccharide or beta-glucan content, standardization criteria, dose per capsule composition, batch identity and cultivation are NOT REPORTED in the trial publication.
TRIAL REGISTRATION / OUTCOME-HIERARCHY PROVENANCE — ClinicalTrials.gov NCT02377024. "Effects of Tremella Fuciformis on Improvement in Cognitive-Bio-Markers of Cognitive Functions in Healthy Adults With Subjective Cognitive Impairment: An 8-Week Prospective Randomized Double-Blind Placebo-Controlled Trial With Multimodal Neuroimaging and Neurocognitive Assessments." Lead sponsor Ewha Womans University; responsible party In Kyoon Lyoo. Actual enrolment 75. Verified against the registry record: REGISTERED PRIMARY OUTCOMES — (1) change from baseline in neurocognitive function (test batteries such as CANTAB); (2) change from baseline in brain structure; (3) change from baseline in brain function (BOLD); (4) change from baseline in brain metabolites (MRS). REGISTERED SECONDARY OUTCOMES — change from baseline in SMCQ score at 8 weeks; number of participants with adverse events (weeks 4 and 8). TIMING: study start February 2014; first submitted 26 February 2015; first posted 3 March 2015 — approximately thirteen months after enrolment began. Registration is therefore RETROSPECTIVE relative to trial commencement and must not be described as a prospectively registered outcome hierarchy. This entry exists to preserve the registration-versus-publication hierarchy discrepancy, not to resolve it.
INTERVENTION CHARACTERIZATION (CITED PREPARATION SOURCE) — Park HJ, Shim HS, Ahn YH, et al. Tremella fuciformis enhances the neurite outgrowth of PC12 cells and restores trimethyltin-induced impairment of memory in rats via activation of CREB transcription and cholinergic systems. Behavioural Brain Research. 2012;229(1):82-90. This is reference 2 of Ban 2018 and is the source to which the trial defers for "the detailed preparation method of TF supplementation." It is a preclinical rat and PC12-cell study (TF 25/50/100 mg/kg orally for 21 days); several of its authors are affiliated with the Institute of Food & Culture, Pulmuone Holdings Co., Ltd. Its role here is intervention-characterization provenance only. Whether it materially characterizes the human trial capsule — species authentication, strain, extract fraction, polysaccharide content, standardization, batch — is NOT YET RECOVERABLE from indexed material. This entry is not evidence for the cognition proposition and is not replication.
SYSTEMATIC REVIEW / SECONDARY SYNTHESIS / SAFETY AND RISK-OF-BIAS CONTEXT — Cave AE, Chang DH, Munch GW, Steiner GZ. Nutraceuticals and cognition in older adults with and without subjective cognitive impairment: a systematic review. Systematic Reviews. 2023;12:143. DOI 10.1186/s13643-023-02301-6. PMID 37592293. PMCID PMC10433666. Twenty-one trials from 7666 screened full texts; nine in SCI populations. Independently corroborates Ban 2018: community-dwelling adults, 65F:10M, mean age 53.83 (SD 5.5), range 40-65; MMSE >=25 and CDR <=0.5 as stated by the review (the trial itself states exclusion at MMSE <=24 or CDR >=0.5 — a minor secondary-source imprecision, recorded, not adopted); 1200 mg/day and 600 mg/day arms, two capsules three times daily, matched placebo with placebo ingredients not reported; 8 weeks; CANTAB pattern-recognition-memory immediate recall and spatial span, WCST; improvement in short-term memory and executive function. RoB 2: Ban 2018 rated OVERALL "SOME CONCERNS" — one of six such trials among twenty-one; the driver is the randomisation domain, for which Ban 2018 is not among the fourteen trials judged low risk, while domains 2 to 5 (effect of assignment, effect of adherence, missing outcome data, measurement of outcomes, selection of reported results) are all judged low risk. The review concludes low overall quality of evidence for nutraceuticals in this population. Secondary synthesis only — NOT replication.
FORMULA EVIDENCE / REPLICATION EXCLUSION — Lin YK, Lin YH, Chiang CF, Jingling L. Effectiveness of Fish Roe, Snow Fungus, and Yeast Supplementation for Cognitive Function: A Randomized, Double-Blind, Placebo-Controlled Clinical Trial. Nutrients. 2023;15(19):4221. DOI 10.3390/nu15194221. PMID 37836504. PMCID PMC10574613. Sixty-four adults with subjective cognitive complaints, placebo (n=32) versus a multi-component beverage formula containing fish roe, snow fungus (Tremella fuciformis) and yeast (n=32), 8 weeks, cognitive testing at weeks 0, 4 and 8. The tested entity is the three-component formula; no arm isolates Tremella, so any effect is not experimentally attributable to Tremella. The formula owns its own evidence. NOT Tremella-alone replication; the replication exclusion is retained.
Source chapter: Chapter 20, The Fungal Evidence Atlas