Do Peptides Work? What the Evidence Shows

A patient came in recently with a printed menu from a wellness clinic. Fourteen peptides, each with a short description, each priced by the month. She wanted to know what was known about them.

That turned out to be a longer conversation than either of us expected, because the list combined compounds that have almost nothing in common. Some have been studied in tens of thousands of people. For others, the human research has not really begun.

The word "peptide" was the only thing connecting them, and it is not a word that carries information about evidence.

What the Word Actually Means

A peptide is a short chain of amino acids. That is the whole definition. It describes what a molecule is built from, in roughly the way that "liquid" describes water and gasoline equally well.

Insulin is a peptide. So is semaglutide, the active ingredient in Ozempic and Wegovy. So is tirzepatide. These have been studied in tens of thousands of patients across years of randomized trials, with published cardiovascular outcomes.

BPC-157 is also a peptide. So are epitalon, TB-500, and CJC-1295.

Same chemical family. Very different amounts of information behind them.

So when you encounter a peptide, whether on a clinic menu or in a podcast, the category tells you nothing on its own. These are three questions that tend to organize the picture.

Question One: Is It FDA-Approved, and for What?

The second half of that question carries most of the meaning, because approvals are always specific to a condition and a population.

Tesamorelin is the only FDA-approved growth-hormone-releasing peptide. It does reduce visceral fat. It is approved for excess abdominal fat in patients with HIV-associated lipodystrophy, and its label states that it is not indicated for weight loss management. The label also reports a threefold increase in new diabetes compared with placebo, along with elevated IGF-1 levels in a substantial share of treated patients.

Elamipretide, sometimes referred to as SS-31, received FDA approval in September 2025. That approval is for Barth syndrome, a rare inherited mitochondrial condition.

Sermorelin is often described as FDA-approved. The approved product, Geref, was discontinued and is no longer marketed in the United States. What is available today is compounded.

In each case the word "approved" is accurate. What it refers to varies a great deal. The useful follow-up is: approved for which condition, in which patients, and how close is that to my situation.

Question Two: How Much Human Evidence Exists?

This question tends to produce the widest range of answers.

BPC-157. A systematic review published in the HSS Journal in July 2025 examined studies from 1993 through 2024 and identified 36. Thirty-five were preclinical, meaning animal or laboratory work. One involved humans: a retrospective review of twelve patients who received a knee injection. The authors reported that no clinical safety data were found.

TB-500. One Phase I safety trial in humans, and no efficacy trials. A 2026 review in Frontiers in Aging also notes that TB-500 is a synthetic fragment rather than the full-length natural molecule it is named for, so research on the natural version does not transfer directly.

CJC-1295. Phase 2 development was discontinued. The same review lists long-term cardiovascular and cancer risks as unknown.

Ipamorelin. Preclinical work plus limited human data, with no long-term safety data. It currently appears on the FDA's list of bulk substances that may present significant safety risks. The agency's entry cites published reports of serious adverse events, including a death, following intravenous administration.

Epitalon. Rodent studies have shown lifespan extension in the range of 12 to 24 percent. There are no Western randomized controlled trials and no long-term safety data.

MOTS-c. The first registered human trial testing its effect on insulin sensitivity began in February 2026 and is still enrolling.

GHK-Cu. No clinical trials. The evidence comes from cell cultures and skin models. Systemic safety has not been characterized.

AOD-9604. Roughly 893 people across six randomized trials, at doses ranging widely, for up to 24 weeks. Safety was indistinguishable from placebo. It did not go on to become an approved obesity treatment.

Two patterns are worth noticing. Most of these compounds have far less human data than their visibility suggests. And in the one case where extensive human testing was done, the result was informative in both directions: reassuring on safety, and not sufficient on benefit.

Question Three: What Would We Measure to Know?

This one is practical rather than evaluative.

Whatever a compound is meant to do, there is usually a way to check. If the goal is body composition, that is measurable. If it is insulin sensitivity, that is measurable. If it is inflammation, recovery, or sleep, there are ways to track each.

The value of asking is that it converts an open-ended protocol into something with a defined endpoint. Baseline measurements before starting, a repeat interval, and a stated result that would mean continuing or stopping. Without those, a course of treatment can run indefinitely without anyone being able to say whether it did anything.

This applies equally to well-studied drugs and to compounds with little research behind them. It is simply how you find out.

What We Do Not Know: Do Peptides Work?

;

This is the part that is easy to state badly in either direction, so I want to be precise about it.

For most of the compounds above, we do not know whether they work. That is not the same as knowing that they do not work. Absence of evidence is a real gap, not a finding.

We also do not know whether they are harmful. That is not the same as knowing that they are safe. When the FDA's scientists reviewed several of these compounds ahead of the July 2026 advisory meeting, the concern they documented was not that harm had been demonstrated. It was that peptides given by injection or nasally carry a recognized potential for immune reactions, and that the data needed to characterize that potential, including impurity profiles and aggregation behavior, was largely missing.

Tesamorelin is a useful illustration of why that gap matters. It is the most thoroughly studied compound in this group, and that study is precisely what revealed a threefold increase in new diabetes. That risk was not visible from the mechanism. It became visible because the trials were large enough and long enough to show it. For most compounds on a typical peptide menu, no comparable trials exist, which means a similar signal would not yet have surfaced whether or not it is there.

There is a separate and more concrete uncertainty around what is in a given vial. Independent laboratory testing reported by NBC in July 2026 found that roughly 30 percent of tested peptide vials failed, whether through mislabeling, incorrect dosing, or contamination. That testing came from a commercial laboratory rather than a peer-reviewed study, so the specific figure should be held loosely. But it points at a real question that applies to any injected product obtained outside the standard pharmaceutical supply chain.

A Summary of the Current Evidence

Putting It Together

Peptides as a class include some of the most consequential drugs in metabolic medicine. They also include compounds whose human research is just beginning. Both statements are true at the same time, and the shared name is what makes that hard to see.

If you are looking at a list, you do not need to evaluate the biology of each item. Three answers will tell you most of what the current science supports: what it is approved for, how many people have taken it under study, and what you would measure to find out whether it is working for you.

Where those answers are thin, the accurate conclusion is not that a compound is dangerous, and not that it is fine. It is that the question is still open. That is a reasonable thing to know before deciding, and a reasonable thing to revisit as more research arrives.

Common Questions

Are peptides FDA approved?

Some are. Insulin, semaglutide, tirzepatide, tesamorelin and elamipretide are all FDA-approved peptides, each for a specific condition. BPC-157, TB-500, CJC-1295, ipamorelin, epitalon, MOTS-c and GHK-Cu are not approved for any indication. The word covers both groups, which is the source of most of the confusion.

Do peptides work for weight loss?

The peptides with substantial weight loss evidence are the approved GLP-1 medications. In the one large head-to-head trial, tirzepatide produced 20.2 percent weight loss compared with 13.7 percent for semaglutide over 72 weeks. Compounds marketed specifically as fat loss peptides have not demonstrated comparable results. AOD-9604 was tested in about 893 people across six randomized trials and did not establish benefit sufficient to become an approved treatment.

What is the difference between peptides and GLP-1 medications like Ozempic?

There is no difference in category. Semaglutide, the active ingredient in Ozempic and Wegovy, is a peptide. The distinction people are usually reaching for is between approved medications and unapproved compounds, not between peptides and GLP-1s.

Are peptides safe?

It depends on which one, and for most the answer is not established. For approved peptides the risks are documented, which is why tesamorelin's label reports a threefold increase in new diabetes. For unapproved compounds, safety has not been demonstrated in either direction. A thin evidence base is not a finding of danger, and it is not a finding of safety.

What does "research use only" mean on a peptide label?

It is a labeling choice by the seller rather than a regulatory category. In warning letters issued in 2026, the FDA stated that when a website markets a product for effects such as appetite suppression or fat oxidation, that marketing establishes the product is intended as a drug for human use regardless of the disclaimer.

How much does peptide therapy cost?

Published pricing at Los Angeles clinics generally runs from about $200 to $800 per month depending on the protocol. Cost varies with the compound, the dosing schedule, and whether baseline labs and ongoing monitoring are included.

References

The findings in this post come from the following. Each is publicly available, and I have linked them so you can read the primary material yourself.

1. HSS Journal. Systematic review of BPC-157, July 2025. Study counts and the finding that no clinical safety data were identified.

2. Frontiers in Aging. Review of peptide therapeutics in aging, April 2026. Evidence status for TB-500, CJC-1295, ipamorelin, epitalon, and GHK-Cu.

3. U.S. Food and Drug Administration. Bulk drug substances for use in compounding that may present significant safety risks. Current listings and the stated concern for each substance.

4. U.S. Food and Drug Administration. Briefing materials, Pharmacy Compounding Advisory Committee, July 2026. The agency review staff's assessment, including the immunogenicity discussion.

5. U.S. Food and Drug Administration. EGRIFTA (tesamorelin) prescribing information. Approved indication, the statement on weight loss, and the diabetes and IGF-1 findings.

6. Clinical trial registry. MOTS-c metabolic trial, NCT07505745. Registration date and enrollment status.

7. Stealth Biotherapeutics. FDA accelerated approval of elamipretide for Barth syndrome, September 2025.

8. Sermorelin drug information. Discontinuation of the approved product.

9. Journal of Endocrinology and Metabolism. Safety review of AOD-9604, 2013. Trial count, participant numbers, and safety findings.

10. NBC Washington. Reporting on independent laboratory testing of peptide vials, July 23, 2026. Note that this testing was conducted by a commercial laboratory and has not been peer reviewed.

Reviewed and current as of August 2026. This is an active area of research, and I will update this post as new evidence is published.

Want the plain version of stories like this one?

I write a short letter every other week that translates what is happening in metabolic and longevity medicine, with the evidence laid out plainly. Sign up at whiteolivedpc.com.