How to Read a Peptide Study: A Plain-Language Guide

Almost every peptide claim online is backed by "studies." That one word covers everything from a dish of cells to a trial of seventeen thousand people. This guide explains how to tell them apart, using real examples from the peptide literature, and how to check a claim yourself in a few minutes.

Start with the model: cells, animals, or people

The first question is always: what was actually studied?

In vitro means "in glass," usually cells growing in a dish. In vivo means inside a living organism (glossary). Cell studies are good at showing a mechanism, but they skip everything a body does to a drug: absorption, dilution in the blood, breakdown by enzymes, and clearance by the kidneys.

Example: PNC-27 killed a human cervical cancer cell line at low concentrations (an IC50 of 12.4 μM) but did not kill a normal cervical cell line (Krzesaj 2025). That is a real finding about cells in culture. It says nothing yet about whether the peptide reaches a tumor in a person, at what dose, or with what side effects.

Animal studies add a whole organism, but not a human one. BPC-157 is the clearest case. In rats with a surgically cut knee ligament, BPC-157 given by injection, in drinking water, or as a cream improved functional, biomechanical and tissue-level healing (Cerovecki 2010). Results like this explain the peptide's popularity. But a 2025 systematic review of its musculoskeletal research included 36 studies, of which 35 were preclinical and one was clinical (Vasireddi 2025). "Preclinical" is the label for everything done before human testing.

Human studies are what ultimately answer questions about people, and even they vary enormously in quality.

Clinical trial phases

Drugs in development move through numbered phases:

• Phase 1 is small and focused on safety and pharmacokinetics (how the body absorbs and clears the drug). The phase 1b trial of the triple agonist GLP-3 R enrolled 72 people with type 2 diabetes across five dose groups, and its primary outcome was safety and tolerability (Urva 2022).
• Phase 2 is larger and explores dose and early signs of benefit.
• Phase 3 is large and confirmatory. The STEP 1 trial of semaglutide enrolled 1,961 adults (Wilding 2021).

A phase number describes a trial's purpose, not its quality. Efficacy numbers from a phase 1 trial are early hints, not proof.

Randomization, placebo, and blinding

Three design features do most of the work of making a result believable:

• Randomized: chance, not the researcher, decides who gets the drug, so the groups start out comparable (RCT).
• Placebo-controlled: a comparison group receives an inactive look-alike. People improve for many reasons besides the drug, including natural recovery and expectation.
• Blinded: in a double-blind trial, neither participants nor investigators know who is receiving what, which keeps expectations from coloring the results.

STEP 1 had all three. Adults were randomly assigned 2:1 to once-weekly semaglutide 2.4 mg or placebo for 68 weeks, both groups alongside a lifestyle intervention. Body weight changed by −14.9% with semaglutide and −2.4% with placebo (Wilding 2021). Notice that the placebo group lost weight too. Without that group, all of the loss would have been credited to the drug. The difference between the groups, 12.4 percentage points, is the estimate of the drug's effect.

Now compare a BPC-157 knee pain study: a retrospective chart review in which 16 patients were contacted by phone and asked to rate their pain before injection and how much the injection had helped. No standardized measurement tools were used and there was no control group; 14 of 16 reported relief (Lee 2021). That design cannot separate the peptide from placebo effects, the natural course of knee pain, or imperfect memory. It does not show that BPC-157 fails. It shows that this study cannot tell.

A comparison against another active drug is a middle case. A Selank study in 62 patients with generalized anxiety disorder or neurasthenia compared it with the anti-anxiety drug medazepam and found similar anti-anxiety effects (Zozulia 2008). Without a placebo arm, "similar to another drug" does not tell you how much either one beats no treatment.

Sample size

Small studies are noisy, and rare side effects stay invisible in them. An intravenous BPC-157 safety pilot enrolled two people, both of whom had received the infusion before, and reported no adverse effects (Lee 2025). A 2025 review counted only three pilot studies of BPC-157 in humans in total (McGuire 2025).

For scale, the SELECT cardiovascular trial of semaglutide randomized 17,604 patients (Lincoff 2023). A side effect that affects one person in a thousand could show up in a trial that size. It could not show up in a trial of two.

Surrogate endpoints versus outcomes that matter

A surrogate endpoint is a lab value or measurement that stands in for the thing you actually care about. In healthy adults, a single injection of CJC-1295 raised average growth hormone levels 2- to 10-fold for six days or more and IGF-1 levels 1.5- to 3-fold (Teichman 2006). The trial's main outcome measures were those hormone levels and the drug's pharmacokinetics. Raising a hormone is not the same as showing a benefit to body composition, recovery, or health, and this trial did not set out to measure those.

A hard endpoint is an event that matters in its own right. SELECT counted cardiovascular deaths, heart attacks and strokes: 6.5% of the semaglutide group versus 8.0% of the placebo group had one over a mean follow-up of 39.8 months (Lincoff 2023).

When one lab produces most of the evidence

Independent replication is how science catches honest errors, unconscious bias, and quirks of one lab's methods. For some peptides, a single group produces most of the published work. BPC-157 is an example: when we ran a PubMed search for "BPC 157" while writing this guide, roughly three-quarters of the records listed the same author, including the rat ligament study above. That is not evidence of wrongdoing. It means the findings have not yet been widely reproduced by unrelated teams, and it is a fact you can check yourself (step 5 below).

What a PMID is

PubMed is the free index of biomedical literature run by the US National Library of Medicine. Every record gets a PubMed ID, or PMID, a unique number, and the address pubmed.ncbi.nlm.nih.gov followed by that number always opens the same record. A PMID proves a paper exists and is indexed. It does not prove the paper is good, or that it says what someone claims. Our methodology page explains how we verify every citation on this site against its PubMed record.

How to check a claim yourself

1. Find the citation. A claim with no source should be treated as unsupported.
2. Open the PubMed record and confirm that the paper is about the compound and the question being claimed.
3. Read the abstract for the model (cells, animals, people), the number of subjects, and whether there was a control group, randomization, and blinding. The "publication type" field shows labels such as Randomized Controlled Trial or Review.
4. Look for the number. If a claim cites a specific percentage, it should appear in the paper.
5. Look around it. Check for a retraction notice, search the compound's name, and look at whose names recur on the author lists. A meta-analysis that pools several independent trials usually outweighs any single study.

Every peptide page on this site shows its research status and links each study to its PubMed record, so you can run these checks as you read.

Sources

• HDM-2-Targeting Peptide PNC-27 Kills Cervical Cancer Cells but not Normal Cervical Cells. Ann Clin Lab Sci, 2025. PMID 40750238
• Pentadecapeptide BPC 157 (PL 14736) improves ligament healing in the rat. J Orthop Res, 2010. PMID 20225319
• Emerging Use of BPC-157 in Orthopaedic Sports Medicine: A Systematic Review. HSS J, 2025. PMID 40756949
• LY3437943, a novel triple GIP, GLP-1, and glucagon receptor agonist in people with type 2 diabetes: a phase 1b, multicentre, double-blind, placebo-controlled, randomised, multiple-ascending dose trial. Lancet, 2022. PMID 36354040
• Once-Weekly Semaglutide in Adults with Overweight or Obesity. N Engl J Med, 2021. PMID 33567185
• Intra-Articular Injection of BPC 157 for Multiple Types of Knee Pain. Altern Ther Health Med, 2021. PMID 34324435
• [Efficacy and possible mechanisms of action of a new peptide anxiolytic selank in the therapy of generalized anxiety disorders and neurasthenia]. Zh Nevrol Psikhiatr Im S S Korsakova, 2008. PMID 18454096
• Safety of Intravenous Infusion of BPC157 in Humans: A Pilot Study. Altern Ther Health Med, 2025. PMID 40131143
• Regeneration or Risk? A Narrative Review of BPC-157 for Musculoskeletal Healing. Curr Rev Musculoskelet Med, 2025. PMID 40789979
• Semaglutide and Cardiovascular Outcomes in Obesity without Diabetes. N Engl J Med, 2023. PMID 37952131
• Prolonged stimulation of growth hormone (GH) and insulin-like growth factor I secretion by CJC-1295, a long-acting analog of GH-releasing hormone, in healthy adults. J Clin Endocrinol Metab, 2006. PMID 16352683

This article is for educational and research purposes only and is not medical advice.