Gathering the record
Relay is organizing the evidence and source boundaries.
CNTF-derived experimental peptide
An experimental peptide, usually called P021, discussed for memory and brain aging. The exact sequence matters because the name can be ambiguous.
Evidence is explained by question. Dosing references show their own check dates.
60-second overview
Why people look into it, what studies found, and what is still unknown.
Educational information only. Evidence reviewed through July 31, 2026. Published studies, sponsor announcements, and trials without results are labeled separately. Evidence gaps apply to the specific finding beside them.
Research questions
Each label shows how much evidence supports that question.
What studies show
No controlled administered-human study of P021 was located.
This is the strongest human-evidence boundary in the current record.
Mechanistic findings explain biological plausibility; they do not establish a human outcome.
Relay keeps this uncertainty visible rather than converting it into a prediction.
The safety snapshot reflects the studied record and its limitations.
Regulatory and identity status determine how the evidence should be interpreted.
Risks & unknowns
Current evidence cannot yet resolve: Human identity, pharmacokinetics, target engagement, clinical outcomes, and longer-term neurologic safety.
No direct evidenceAn absence of adequate evidence is not evidence of benefit, harm, or equivalence.
Open sourceThe available information does not reliably show how often particular side effects occur.
Evidence boundaryUnder-detected or under-reported events must not be described as rare.
Relay has not verified which side effects led people to stop treatment in studies.
Evidence boundaryThis is an evidence gap, not proof that discontinuations did not occur.
Relay has not verified a compound-specific warning summary for this profile.
Evidence boundaryPreclinical; no FDA-approved P21/P021 product. Human identity, pharmacokinetics, target engagement, clinical outcomes, and longer-term neurologic safety.
This profile does not contain verified, product-specific emergency instructions.
Evidence boundaryRelay does not infer emergency guidance from study discontinuations, mechanism, or incomplete adverse-event reporting.
Dosing references
Each table labels its source and applies only to the named route, product, or study. An online example or community report is not an approved dose.
The reviewed cognition-related evidence is preclinical. P21 is not interchangeable with a parent protein or another neurotrophic peptide.
What would fill this gap: An exact-identity human administration protocol.
Dosing review checked Oct 20, 2026. Additional sources are available below.No route-specific preparation instructions were verified. A different route or product does not supply them.
Starting dose: the amount used at the beginning of the named label schedule or trial.
Maintenance: the ongoing amount described by that product label.
Trial target: the amount assigned to a study group. Different groups may stop increasing at different targets.
Published research investigated P021 in transgenic and aged mouse models, measuring learning behavior alongside neurogenesis, synaptic, and disease-like pathology markers.
Population: Mouse models
Design: Controlled animal studies
Mechanisms
P021 was designed to retain neurotrophic signaling while limiting parent-protein liabilities. Published outcome evidence is preclinical and should not be confused with the unrelated p21 cell-cycle protein.
A plausible mechanism can explain why a study was attempted. It does not prove that the compound improves a human outcome.
Research reference
A reference for the studies, populations, and routes behind this profile.
Animal and mechanistic evidence only; marketed naming can be ambiguous.
Research depth describes how large and mature the overall record is. It does not tell you whether the results were positive.
Human identity, pharmacokinetics, target engagement, clinical outcomes, and longer-term neurologic safety.
Confidence describes how reliable and consistent the conclusions are. It can be high for one outcome and low for another.
The strongest evidence type shows what the best-supported conclusions are actually based on.
Human findings are more directly relevant than animal or laboratory results, but they still apply only to the populations and outcomes studied.
Route-specific record
These are exposures used in cited research for a specific route and population—not an instruction for an individual.
Half-life describes how long it takes the measured amount in the body to fall by half. It is not a dosing recommendation.
Evidence can change by administration method. Findings from one route should not automatically be applied to another.
Evidence boundary: No verified controlled study administered P021 to human participants. Animal behavior and surrogate brain markers cannot establish human cognition benefit, disease prevention, exposure, or safety. Amounts shown describe cited research exposure—not a dosage recommendation.
Studies
Study arms are reported for transparency. They describe what researchers did in a defined record and do not transfer across identities, routes, formulations, or populations.
A later trial phase can ask a more mature question, but it does not guarantee a positive result, regulatory approval, or relevance outside the studied population.
Controlled animal studies
Mouse models
P021 improved selected learning and pathology measures in transgenic and aged mouse models.
Study administration: P021 in feed or experimental formulations
Limitations: Animal-model and formulation findings do not establish human outcomes.
Animal and laboratory study
Rodents and neural systems
Endpoints were not fully described in the current record.
Preclinical studies reported neurogenesis and synaptic markers alongside behavioral outcomes.
Study administration: Not stated in the current record.
Limitations: Surrogate markers and animal behavior do not establish clinical benefit.
Community
Built by the community
Every structured report adds useful context about research setup, outcomes, and unwanted effects as the community dataset grows.
Publicly anonymous by default. Your account keeps reports editable and helps reduce duplicate submissions.Sources
PMID 20600002
PMID 26114951