Wolverine Peptide Study: Use Is Up, Proof Isn't

The new Wolverine peptide study doesn't tell you whether BPC-157 works. It answers a different question: how many people are using it, and who they are. Researchers used AI to read de-identified doctors' notes and found 1,039 patients whose charts recorded BPC-157 use, with new users each quarter rising 33-fold between 2020 and 2026. New users became more male and younger over time. Of those whose source was written down, about a third got it from gray-market sellers. The authors say themselves that their data can't show whether it works, and the paper hasn't been peer reviewed yet.

A naming note. The paper's title calls BPC-157 itself the "Wolverine peptide." Online, "Wolverine" more often means BPC-157 paired with TB-500, which is also what Promise's Wolverine blend contains. This study is about BPC-157.

As of September 23, 2026, the day this article was written, the paper was still a preprint, meaning a draft shared publicly before other scientists have formally reviewed it. It was posted on September 7 (Venkatakrishnan et al., 2026). Reuters covered it on September 15, framing the story around Health and Human Services Secretary Robert F. Kennedy Jr.'s public support for the peptide.

What the Wolverine peptide study found

BPC-157 is hard to count. It has no National Drug Code, the number pharmacies and insurers use to track a medicine, so it doesn't show up in prescription records. The team, from the data company nference, went where use does get written down: clinicians' notes.

They used a large language model (AI software that reads and sorts text) to search notes with names and identifying details removed, drawn from a network of U.S. health systems that pool their records. Reuters reported the search covered more than 15 million medical records. Physicians checked the AI's work: it was right 96% of the time on which other drugs people used, but only 74% on non-drug care like physical therapy.

Treat 1,039 as a floor, not a count. The study's leader, Venky Soundararajan, told Reuters: "Most likely, very few doctors are even asking their patients, 'Are you taking BPC-157?'"

Who takes BPC-157, and where they get it

By 2026, seven in ten new users in these notes were men, and their average age was 49, six years younger than in 2020. Among the 972 users whose race was recorded, 95.9% were White, compared with 78% of the wider patient population in the network.

The reasons written down were mostly pain (33% of all users), gut problems such as reflux, inflammatory bowel disease or gastritis (14%), an old injury (11%) and healing after surgery (10%).

Where people got it was recorded for only 205 users, about 1 in 5. Among them, 36% named a compounding pharmacy and 35% a gray-market peptide vendor, a seller working outside the prescription system. A chart note records what the patient said; it can't confirm that a seller was a licensed pharmacy filling a prescription.

What it means if you're weighing BPC-157

One number in the paper matters more than the headline. Among 80 users whose start date could be pinned down, they had been taking BPC-157 for an average of 60 days before a clinician first wrote it down. By then they had already picked a source, an amount and a schedule on their own. Reuters also reported that when people stopped, it was most often because a doctor advised them to.

A prescribed route changes that order of events. A licensed provider looks at your history, your other medicines and why you're asking before anything is dispensed, and can say no. A licensed U.S. compounding pharmacy prepares the prescription and labels what's in the vial. If something goes wrong, there's a prescriber and a pharmacist to call and a record to trace. At Promise, a licensed provider reviews every request, and not everyone qualifies.

None of that makes the evidence stronger. It makes clear who decided and what was dispensed. Our look at gray market peptides covers the other side of that contrast.

The medication list matters more than usual here. Half the users, 525 people, were also taking something else: most often testosterone (17.0%), then NSAIDs, the ibuprofen family of painkillers (14.7%), TB-500 (12.9%) and corticosteroids (11.7%). How BPC-157 interacts with any of these hasn't been studied in people, which is the kind of question a prescriber weighs before a first dose rather than two months after. The Wolverine peptide stack page covers what's known about pairing it with TB-500, which comes down to one rat study.

What the study can't tell you about benefits and side effects

The notes recorded improvement in pain, healing or function for 79% of the 354 users whose response was written down, and worsening for 5%. That sounds encouraging until you see the gaps. For 685 of the 1,039 users (65.9%), no response was recorded at all. Nobody was compared with a similar group not taking it. And many were getting physical therapy, surgery or other drugs at the same time.

The authors say it directly: these reports "cannot establish efficacy" or separate BPC-157 from other treatments, the placebo effect (feeling better because you expect to) and biases in who gets written up. Flynn McGuire of the University of Utah, who wasn't involved, told Reuters there was "no placebo or untreated comparator." A 2026 review in Sports Medicine also describes the placebo effect as part of how peptides appear to work, amplified by social media (Mendias and Awan, 2026).

Side effects are the other blind spot. The notes documented neuropsychiatric effects (changes involving mood, thinking or the nervous system) in 1.4% of users, injection-site hypersensitivity (an allergy-type reaction where the shot went in) in 1.4% and gut problems in 1.3%. The authors call these likely under-reported, and the method explains why: a note holds only what a patient mentions and a clinician thinks worth recording. Mild reactions, ones blamed on something else, and anything felt by people who never told a doctor all go missing. A low number here means rarely written down, not necessarily rare.

It isn't an FDA decision, either. July's advisory vote is easy to read as a green light. The committee recommended adding two forms of BPC-157 to the list of ingredients pharmacies may compound with, but that advice is nonbinding. As of September 23, 2026, the day this article was written, FDA's meeting page showed no final decision, and the Federal Register carried no proposed rule naming BPC-157. BPC-157 has never had an approved product and is not FDA-approved; the compounded formulation offered here is not FDA-approved either. A licensed provider may still prescribe a compounded formulation, and that decision is between you and your doctor.

What to watch next

Peer review. The preprint may change before a journal publishes it, and reviewers are likely to push on the AI methods and the numbers above.

A decision from FDA. The next real step would be a proposed rule naming one form of BPC-157, both or neither, followed by public comment. The September BPC-157 update tracks that process.

Controlled trials. As of September 23, 2026, the day this article was written, ClinicalTrials.gov listed a placebo-controlled trial of about 120 adults with hamstring strains that was still recruiting, and a small placebo-controlled pilot after rotator-cuff surgery, first posted September 3 and not yet recruiting. Trials with a comparison group can answer what chart notes can't. Until they report, the human evidence is the handful of small studies in BPC-157 human studies.