Methodology
The Evidence Framework
Four instruments, kept deliberately separate: how mature the evidence is, which way it points, how often it has survived retesting, and how far the public claim has travelled from the result.
Instrument one
Evidence Maturity — E0 to E5
The scale does not ask how true a claim is. It asks how far the proposition has travelled into direct human testing.
No meaningful evidence base
No credible evidence meaningfully supports the specific proposition.
Preclinical or indirect
Laboratory, animal, mechanistic or otherwise indirect evidence exists, but qualifying direct human evidence is absent.
Preliminary human evidence
Human exposure or early human signals exist, but causal confidence remains limited because the evidence is uncontrolled, weakly controlled or otherwise preliminary.
Controlled human evidence
At least one meaningful controlled human experiment directly tests the proposition.
Substantial clinical evidence
Large, rigorous or otherwise mature clinical evidence exists for the defined intervention and proposition.
Mature replicated clinical evidence
Multiple strong, substantially independent human programmes support a well-defined proposition.
E0
No meaningful evidence base
E1
Preclinical or indirect
E2
Preliminary human evidence
E3
Controlled human evidence
E4
Substantial clinical evidence
E5
Mature replicated clinical evidence
The historical qualifier
H — Historical qualifier
H — Historical evidence remains informative, but transport into contemporary treatment is a distinct uncertainty. H is not a downgrade. It does not mean old, obsolete, inferior or invalid, and evidence is never graded lower merely because of its age. H records that the surrounding medicine — staging, standard of care, comparators, supportive treatment and reporting conventions — has changed since the evidence was produced, so the question of whether the result carries into present-day practice must be assessed separately from the quality of the original work.
E4-H+ therefore reads as: substantial clinical evidence (E4), historical in origin (H), supporting the defined proposition (+), with transport into contemporary treatment held open as its own uncertainty.
Age is not the limitation. Transportability is.
Instrument two
Direction — + / Ø / ±
If direction is hidden inside evidence maturity, the scale quietly becomes a positivity ladder. Science needs somewhere to put a good no.
+
Supporting direction
The best direct evidence supports the defined proposition.
Ø
Not supporting
The controlled evidence does not support the proposed benefit under the tested conditions.
±
Mixed
Credible evidence points in materially different directions.
This lets us write E3+ and E3Ø without contradiction. Both may represent serious controlled human experiments. Their answers differ. Mixed evidence, meanwhile, should never become a fog into which every inconvenient null result disappears.
Reading a null result
A controlled null result is evidence
Direction Ø means the controlled evidence did not support the defined proposition under the tested conditions. It does not mean that no evidence exists, that the study failed, or that the evidence is weak because it is null. Direction and maturity are recorded separately precisely so that a well-conducted negative result keeps its standing.
A record graded E3Ø with R2 is therefore relatively mature, partly replicated controlled human evidence whose answer is null — not thin or unreliable evidence.
Instrument three
Replication — R0 to R5
Imagine two interventions, both rated E3+. One rests on a single positive randomized trial; the other has survived four substantially independent programmes. Those should not produce the same confidence.
No replication
No qualifying independent repetition.
Single qualifying experiment
The proposition rests primarily on one qualifying programme.
Limited repetition
Partial repetition, related studies or limited independent support exist, but the programme remains immature.
Meaningful independent replication
Substantially similar propositions have been reproduced independently in a clinically meaningful way.
Strong replication
Multiple substantial programmes support the same well-defined proposition.
Mature robust replication
Broad, repeated, substantially independent evidence supports the proposition across a mature programme.
Evidence maturity asks how far and how well the question has been tested. Replication asks how often the answer survived being tested again. One large study is not many studies, and several heterogeneous studies do not automatically constitute replication.
Instrument four
Claim Gap — CG0 to CG4
Claim Gap measures the distance between the sentence in public and the experiment underneath it.
CG0
Direct
The sentence states what the experiment actually found, in its own terms.
CG1
Reasonable synthesis
A fair summary, provided relevant strain, population and clinical-context limitations are not hidden when they matter.
CG2
Broader than the measured endpoint
Plausible shorthand, but the intervention identity or the outcome has started to loosen.
CG3
A different proposition
Outcome, mechanism or duration has expanded beyond what was measured.
CG4
The evidence has largely disappeared
The direct experiment no longer supports the claimed endpoint.
Worked example — Maitake, Chapter 18
- CG0In one small randomized trial, Y10M Maitake improved MoCA-J scores relative to placebo in healthy older Japanese adults.
- CG1Preliminary controlled evidence suggests Y10M Maitake may improve selected cognitive performance in older adults.
- CG2Maitake may support cognition. — the strain identity has disappeared.
- CG3Maitake improves memory and prevents cognitive decline. — outcome and duration have expanded.
- CG4Maitake helps prevent dementia. — the direct experiment no longer supports the claimed endpoint.
Where the movement happened
The Evidence Transfer Vector
Claim Gap tells us how far a claim has travelled. It does not tell us where the movement occurred. ETV records the dimensions across which evidence is being carried — a map of the bridges, not a score.
T
Taxon
Did evidence move between species, strains or taxonomic identities?
I
Intervention
Did evidence move from one preparation, formulation or material to another?
P
Population
Did evidence move to people unlike those actually studied?
O
Outcome
Did one measured endpoint become a broader or different claim?
C
Clinical context
Did evidence move into a different treatment environment?
R
Replication
Did limited evidence become established certainty?
D
Exposure
Did evidence move between different doses, routes, durations or otherwise materially different exposure conditions?
X
Experimental level
Did evidence move from cells or animals into human claims?
ETV is deliberately not a score. Mouse tumour suppression becoming a human cancer treatment claim may cross fewer labelled dimensions than a modest transfer between two closely related human populations, yet the inferential leap is vastly greater. ETV tells us where to inspect, not what verdict to issue.
Two bridges, assessed separately
TTC and ITC
Taxonomic Transfer Confidence can be high while Intervention Transfer Confidence is low — and that combination is common.
TTC
Taxonomic Transfer Confidence
How confidently a result recorded under one biological identity can be carried to another: between species, strains, historical names and modern molecular taxonomy.
Example from the Atlas: historical Coriolus versicolor → Trametes versicolor: HIGH.
ITC
Intervention Transfer Confidence
How confidently a result belonging to one preparation, formulation, dose or route can be carried to another product sold under the same name.
Example from the Atlas: PSK → an arbitrary Turkey Tail product: LOW.
Assessed separately
Safety — S1 to S3
Safety is assessed independently of efficacy. A strong efficacy programme does not establish harmlessness, and a null or immature efficacy programme says nothing about risk. Every safety grade is specific to the exposure, preparation, dose, route, duration and population actually studied.
No material safety signal in the evidence examined
Within the exposures, preparations and populations actually studied, no material safety signal was identified. This is a statement about what was observed, not a declaration that the organism or any product made from it is universally safe.
Qualified or context-dependent considerations
Tolerability is broadly reported, but meaningful conditions apply — route, dose, duration, co-treatment, preparation identity or population. The qualification is part of the grade and travels with it.
Specific hazard signal requiring attribution
A specific, documented hazard signal is attached to a defined exposure or preparation. S3 does not mean universally dangerous; it means a real signal exists that must be attributed to the exposure that produced it rather than to a common name.
Intermediate notations such as S1-2 record that the evidence sits between two states rather than resolving cleanly into one. No S grade describes a mushroom, a common name or a product category as a whole: it describes what was observed for a defined exposure. S1 is not a declaration of universal safety, and S3 is not a declaration of universal danger.
Assessed separately
Conservation — C0 to C3, and C+
Conservation is assessed independently of efficacy. Rarity does not imply potency, wild origin does not imply clinical superiority, and a conservation concern is never evidence for or against a therapeutic claim. Cultivation frequently improves both sustainability and reproducibility, because an authenticated cultivated strain can be characterized, deposited and repeated.
C0
No material conservation concern
Supply does not depend on pressure against wild populations, typically because the organism is routinely cultivated.
C1
Sourcing caution
An Atlas sourcing caution rather than a formal threat classification: wild collection under uncertain identity or uncertain provenance creates avoidable ecological and analytical risk.
C2
Material collection pressure
Demand exerts meaningful pressure on wild material, habitat or collection practice.
C3
Serious wild-population concern
Wild harvesting raises serious conservation concern. C3 is an ecological statement; it says nothing about whether the material is clinically effective.
C+
Conservation-positive potential
Cultivation, fermentation or manufacture can reduce dependence on wild harvest while improving authentication, standardization and batch archiving — a scientific gain as well as an ecological one.
A conservation grade carries no therapeutic meaning in either direction. C3 does not mean clinically ineffective, and a wild-collected material is not more potent for being wild. Combined notations such as C0/C+ record both the current sourcing position and the improvement that authenticated cultivation would offer.
What would resolve it
Resolution Zone (RZ)
Resolution Zone records the type and stage of work that would actually reduce the remaining uncertainty for a record — for example authenticating the biological material, characterizing the preparation, reanalysing like with like, or running a clinically meaningful confirmatory trial.
- It is not an evidence grade.
- It is not a treatment recommendation.
- It is not a numerical quality score, and a higher number is not worse evidence.
Resolution Zone entries are written as prose because the honest answer is usually a sequence of steps rather than a single label. Two records can share a maturity grade and still need entirely different work next.
The reader's audit
Six questions to carry into any claim
When somebody says studies show this mushroom works, these questions turn one sentence back into several scientific ones.
- 01
Was it tested in humans? If not, the evidence is preclinical or indirect.
- 02
Was there an appropriate control? If not, causation remains uncertain.
- 03
Did the intervention actually outperform the control? Not merely: did the treatment group improve?
- 04
Has substantially the same proposition been independently reproduced? Not merely: are there many papers about the mushroom?
- 05
What exactly did participants receive?
- 06
What exactly improved?
Scepticism is not the art of never being convinced. If the intervention is well characterized, the comparator appropriate, the endpoint clinically meaningful, the analysis sound, the effect substantial and the result independently replicated, then confidence should rise.