The most useful question about a peptide study is specific: what material was tested, in what model, against what comparison, and with what result? A long reference list cannot replace those details. A study can provide useful evidence for one question while leaving another unanswered.
This guide explains how to read research findings and their limits. It does not provide a treatment or self-administration protocol.
Primary-source examples checked October 7, 2026.
1. Identify where the research happened
In vitro research examines material outside an intact organism, for example in cultured cells. Animal research studies a living model with its own biology and experimental conditions. These studies can investigate mechanisms and help frame further questions; a finding in either setting is not itself a demonstrated benefit in people. The FDA describes preclinical research as work that precedes testing in humans. FDA overview of preclinical research.
The source of the cells matters, but “human cells” does not mean “human clinical trial.” A cultured human cell line does not reproduce a whole person’s exposure, distribution, metabolism, other tissues or long-term outcomes. Keep the model beside the finding whenever summarizing it.
2. Ask what the human study actually tested
| Evidence type | Useful information | Main reading check |
|---|---|---|
| Observational study | Patterns or associations in measured characteristics. | Was a peptide administered at all, or were its naturally occurring levels measured? |
| Uncontrolled intervention or case series | What happened to the reported participants after an intervention. | Without a comparable control group, what other explanations remain? |
| Randomized controlled trial | A comparison between groups assigned by chance. | Check the actual material, comparator, participant numbers, primary outcome, follow-up and missing data. |
| Review or evidence synthesis | An organized account of earlier studies. | Which original studies support the claim, and are the same participants counted more than once? |
NIH distinguishes observational studies from clinical trials, in which an intervention is assigned. Randomization can strengthen a comparison, but it does not make a tiny, short or poorly reported trial conclusive. A trial’s phase also describes its development role; it is not a standalone quality grade. NIH explanation of clinical studies; FDA overview of clinical research.
3. Read the findings together with their limits
Selank: a human cell line remains a cell experiment
A 2017 study examined expression of 84 genes in IMR-32 neuroblastoma cells. Selank alone did not change the measured mRNA levels under the tested conditions; combinations with other substances produced different patterns. This is a specific laboratory finding. It neither demonstrates a therapeutic benefit nor proves that the compound has no effect in every setting. Primary study: Filatova and colleagues, 2017.
MOTS-c: measuring a human response is different from administering the peptide
A 2021 paper combined mouse experiments with measurements from 10 young men around an exercise session. The human component measured endogenous MOTS-c in muscle and circulation; the participants were not being tested for a benefit from administered MOTS-c. Reading the mouse intervention and human measurements as if they were one human treatment trial would change the meaning of the paper. Primary study: Reynolds and colleagues, 2021.
GHK-Cu: report the null findings alongside the favorable result
In a randomized study of topical skin-care regimens after CO2 laser resurfacing, 13 participants completed follow-up. Objective and blinded assessments did not find the stated between-group improvements in redness, wrinkles or overall skin quality; a patient questionnaire favored the GHK-Cu regimen. Both findings belong in a summary. This small topical study does not establish effects of a catalog powder in a different formulation or setting. Primary study: Miller and colleagues, 2006.
4. Separate a study plan from its results
A trial registration can describe an intended design, estimated enrollment and outcomes. Those are not findings. Look for actual enrollment, completion status, posted results and a publication that reports the same study. A listing also does not mean the US government has evaluated the study’s safety and science. NIH guidance on finding a clinical trial.
When reading a record on ClinicalTrials.gov, check whether enrollment is estimated or actual, when the record was last updated, and whether results are posted. Recruiting status and planned enrollment do not show that participants completed the study or that an intervention worked. Registry status can lag actual activity, so retain the update date and read the complete record.
An abstract, sponsor announcement and full paper also provide different amounts of information. Label what you actually read. If only an abstract is available, do not fill missing methods or outcomes with assumptions. If several sources describe the same trial, count it as one trial when discussing the evidence base.
If the registry lists a study but you cannot find its findings, use Why Can’t I Find Results for a Peptide Clinical Trial? to compare results tabs, completion dates, record versions and linked publications.
5. Use a claim-to-source checklist
- Match the material. Check the exact peptide, complex, fragment, formulation or analog. Similar names do not make materials interchangeable.
- Name the model. Cells, animals and people belong in separate rows. For people, distinguish measurements of natural levels from an administered intervention.
- Identify the comparison. Note the control, randomization and blinding where reported. State when a study is uncontrolled.
- Count the right units. Record participants and completed follow-up separately. Repeated samples from one person are not additional participants.
- Name the outcome. A gene-expression change, laboratory marker and clinical outcome are different endpoints. A proposed biomarker is not automatically a validated substitute for how someone feels or functions. FDA explanation of surrogate endpoints.
- Keep uncertainty visible. Read effect sizes, confidence intervals, missing data, null findings and follow-up. “Not statistically significant” is not the same as proving identical effects.
- Separate efficacy from safety. A favorable outcome does not settle safety, and a small study with few reported events cannot rule out uncommon or long-term harms.
- Check the record. Save the primary source and check for corrections, retractions or later results. Keep the source-check date separate from a substantive content-update date.
A useful summary follows this pattern: “In this model and study design, the investigators measured this outcome; this limitation prevents a broader conclusion.” Avoid replacing it with an unqualified benefit claim or a letter grade unsupported by an explicit evidence assessment.
Explore the research library and the BPC-157 evidence profile for compound-specific context. For the separate question of what a particular material’s laboratory documentation establishes, read How to read a peptide COA. Product documentation does not establish clinical efficacy, and a published study does not verify a supplier’s batch.
WORKED EVIDENCE EXTRACTION
6. Separate the evidence rows within one paper
A paper can combine different models and experimental questions. Make a separate reading record for each relevant component, rather than attaching one broad claim to the whole citation.
- 01 · Model
Identify the cells, animals or participants actually studied.
- 02 · Exposure
Separate an administered material from a naturally occurring substance being measured.
- 03 · Outcome
Keep the measured endpoint beside the finding and its limits.
MOTS-c: two components, two different conclusions
| Component | What was studied | Boundary of the conclusion |
|---|---|---|
| Human exercise measurements | Ten young men exercised on a stationary bicycle. Investigators measured endogenous MOTS-c in muscle and blood. | The reported exercise-associated changes are not a trial demonstrating benefit from administered MOTS-c. |
| Mouse intervention experiments | Other parts of the paper examined administered MOTS-c and physical-performance outcomes in mice. | Those intervention findings belong to the mouse experiments; they cannot be relabeled as human treatment outcomes. |
The human component and the mouse interventions share a paper, but not a single study population or exposure. Read the published abstract and figure captions and the original article before quoting a result. This extraction example was checked October 21, 2026; it is not a reanalysis of the raw data.
Count people and measurements separately
In a hypothetical study of 10 people sampled four times each, there are 40 sample collections and 10 participants. Repeated observations do not become 40 different people. Record both numbers and use the paper’s stated analysis population; do not invent participant counts from the number of specimens or plotted points.
For study-design context, see NIH’s explanation of clinical studies. For the difference between a biological marker and a clinical outcome, see the FDA endpoint guide.
Continue with the MOTS-c evidence profile. The separate laboratory catalog entries are MOTS-C – 10 mg and MOTS-C – 20 mg; this study does not verify those products.
7. Reuse this study-reading worksheet
Create a separate record for each model, intervention and outcome you intend to summarize. Preserve the units, follow-up and qualifications that determine what the result means.
Open the copyable evidence worksheet
Citation / DOI / trial identifier: Source URL and date checked: Reading scope: abstract / full paper / registry / other: Exact material and formulation: Model or participant population: Assigned intervention or observed exposure: Comparator / randomization / blinding: Participants or animals enrolled / analyzed: Number of observations or specimens: Outcome, measurement unit and follow-up: Reported result and uncertainty: Missing data, null findings and reported harms: Correction, retraction or later-result check: What this finding supports: What this finding does not establish:
For a worked comparison across different patient settings, read What Is the Difference Between Semax and Selank? Use a separate worksheet for each study before comparing its findings.
For the sample-to-stock question, use the COA review worksheet. Keep the research finding and the material’s documentation separate when recording a conclusion.