Peptide evidence guides
Each guide lists the human studies, the animal and cell work, and the gaps. Approved medicines and investigational peptides are each described by the evidence behind them.
The six guides
Start with Peptides 101 if you're new to the topic.
Peptides 101
How to tell approved peptide medicines apart from pharmacy-made versions and early research, so a lab finding isn't mistaken for a proven treatment.
Foundation- Evidence in people
- Varies by product and use
GLP-1s
Approved uses, the main human trials, known risks, and how FDA-approved products differ from compounded or other unapproved versions.
Approved medicine class- Evidence in people
- Established for specific approved uses
BPC-157
What BPC-157 is, the few small human studies, the animal and cell work, the open safety questions, and the 2026 FDA compounding review.
Investigational peptide- Evidence in people
- Limited human evidence
KPV
What cell and mouse studies suggest about KPV, how far that is from proof that it helps people, and what the FDA found when it reviewed KPV in 2026.
Investigational peptide- Evidence in people
- No useful human studies found
TB-500
What TB-500 is (a lab-made piece of a natural protein), the research often wrongly credited to it, and the open questions about using it and about product quality.
Investigational peptide- Evidence in people
- No useful human studies found
MOTS-c
What MOTS-c is (a small protein coded in the DNA of mitochondria, the cell's power plants), the human exercise study, results in cells and mice, and the missing evidence for MOTS-c given to people.
Investigational peptide- Evidence in people
- No useful human studies found
Side by side
The table compares the six on human evidence, regulatory status and the main open question.
| Guide | Evidence in people | Approval status | Main open question |
|---|---|---|---|
| Peptides 101 | Varies by product and use | Mixed status; evaluate the exact product and use | Calling something a peptide says almost nothing about whether it helps people, what risks it carries, or whether results for one version apply to another. |
| GLP-1s | Established for specific approved uses | FDA-approved products exist; unapproved versions are distinct | Trial averages can't predict how one person will respond, cope with side effects, keep taking it, or do over many years. Results for an approved product also say nothing about what is in an unapproved version, how pure or sterile it is, or whether it works. |
| BPC-157 | Limited human evidence | Not FDA-approved; reviewed in the 2026 503A compounding process | The open question is whether any BPC-157 product helps people enough to outweigh its risks. Little is known about whether it works in people, how the body absorbs and clears it by the usual routes, whether the immune system reacts to it, what dose matters, product quality, and long-term safety. |
| KPV | No useful human studies found | Not FDA-approved; reviewed in the 2026 503A compounding process | No useful study has given KPV to people. So how much reaches the body, whether it helps, how the body handles it, its side effects, immune reactions, drug interactions and long-term risks are all unknown. |
| TB-500 | No useful human studies found | Not FDA-approved; reviewed in the 2026 503A compounding process | Studies of the full natural protein (thymosin beta-4), of a slightly different fragment, or of cells and animals are often cited as if they had tested TB-500 in people. None of them showed whether a TB-500 product works in people, how much reaches the body, whether the immune system reacts to it, or whether it is safe over time. |
| MOTS-c | No useful human studies found | Not FDA-approved; reviewed in the 2026 503A compounding process | The open question is whether a well-made MOTS-c product, given to people, helps enough to outweigh its risks. Studies so far measured the body's own MOTS-c or treated cells and mice, so how much reaches the body, whether it helps, its safety and its effect on lifespan in people are unknown. |
The evidence ladder
Each rung answers a different question. Evidence lower down can matter without showing that a treatment helps people.
The ladder ranks how directly evidence speaks to treatment. It doesn't grade safety, and study quality and relevance matter on every rung.
Health results in people
Studies with a comparison group that measure how people feel, get around, fall ill or are harmed. This is the most direct evidence about a treatment.
Blood tests and body measurements in people
Blood levels, lab tests and body measurements in people. They show what happens in the body without proving that people are better off.
Animal studies
Animal studies test how something might work and whether it looks safe. Animals differ from people, so results often don't carry over.
Lab studies in cells
Experiments in cells, tissue or test tubes suggest ideas to test. Only studies in people can show that a product helps people.
Untested idea or marketing claim
A proposed benefit with no direct evidence behind it. It is an untested claim, and should be described as one.
How the FDA compounding review works
The July 2026 meeting dealt with a compounding pathway set by statute, which runs separately from FDA drug approval. It has four steps.
A staff memo is a staff assessment, and an advisory committee recommendation is advice. Neither is an FDA decision.
A substance is nominated
Someone asks the FDA to consider adding an ingredient to the list that pharmacies may use to make custom medicines. That is separate from approving a drug, which needs its own application.
FDA staff assess the record
FDA staff look at what the ingredient is, how it has been used, whether it works and whether it is safe, and publish a memo on what they found.
PCAC gives nonbinding advice
The Pharmacy Compounding Advisory Committee, a panel of outside experts, discusses the questions and advises the FDA. Its advice approves nothing.
FDA considers final agency action
The FDA weighs the panel's advice and finishes its reviews before making a final decision. A meeting or a staff memo on its own is never that decision.