PubMed
Find the study, then read its design, population, endpoints and corrections. Being indexed does not certify a paper’s conclusions.
Explore PubMedFollow the evidence.
Every record is a starting point. The sources below let you trace a statement back to the molecule, experiment or regulatory document behind it.
ClinicalTrials.gov
Check the intervention, planned outcomes, recruitment history and posted results. Registration alone is not evidence of benefit.
Explore ClinicalTrials.govPubChem
Match a specific structure, salt and stereochemistry. Similar names and shared formulas do not make compounds equivalent.
Explore PubChemFDA
Use the product label and approval documents for the stated jurisdiction. Approval belongs to a defined medicine and indication.
Explore FDAHow a record earns its place.
This is an expanding editorial reference, not a systematic review. The current pass checked the identity and evidence statements published here against selected original studies, indexed abstracts, trial records and regulatory documents. Full texts were not available for every study; relevant limits appear in the individual record.
- Resolve the molecule. Separate analogs, fragments, conjugates, mixtures and nonpeptide compounds before considering outcomes.
- Read the experiment. Keep cell and animal findings separate from human interventions. Identify the actual formulation, population and measured endpoint.
- Preserve uncertainty. Retractions, negative results, small samples and unmatched molecular forms remain visible. An evidence band describes study context, not effect size or safety.
- Date the review. Each record states its review date. New evidence and changed labels can alter a summary; the linked source is the place to check.
Unverified dosing schedules, cycling instructions and undocumented confidence scores are not used as evidence. Identity-limited entries provide context while making the missing match explicit. Send a source or correction.
The source register.
Direct references for each compound. A source may establish an identity distinction or a limitation rather than support a positive effect.