The Mainstream-Medicine Skeptical Case on Peptides — And Where I Push Back

If you search "are peptides safe," you'll find two camps shouting past each other. I run a peptide-education site with an affiliate deal, so you'd be right to wonder whether I'll give you the honest version of the other side's argument. Here it is — as strong as I can make it, from actual doctors who have nothing to gain by saying it.

Who's making this argument

The clearest version I've found is a crossover conversation between four MDs — Spencer Nadolsky, Karl Nadolsky, Austin Baracki, and Jordan Feigenbaum (Docs Who Lift × Barbell Medicine). All four have openly said they could brand their own peptide line for real money and choose not to. That disclosure matters — it heads off the easy dismissal that peptide-skeptical doctors are just protecting pharma turf. These aren't industry mouthpieces. They're evidence-standard advocates, and I want to engage with the strongest version of what they're saying, not a strawman.

Their core argument: pick a standard and say it out loud

Their claim isn't "peptides bad." It's structural: anyone recommending a compound should be able to name, out loud, what evidence standard they're using. Their preferred bar is some form of a placebo-controlled human trial — not necessarily full FDA approval, just a real structured threshold.

Their challenge to the peptide world is direct: what's your standard? If the honest answer is "I've seen people feel better on it," that's a standard — but it's one with a bad track record. The CAST trial found antiarrhythmic drugs that successfully suppressed arrhythmias actually increased mortality. The CETP inhibitor torcetrapib raised HDL and killed more patients anyway, through an off-target effect nobody predicted. Fen-phen is the textbook cautionary tale. Biological plausibility is a notoriously unreliable predictor of what a trial will actually show, and without the trial, you can't tell which of your intuitions are the wrong ones.

They're not FDA absolutists, either — they explicitly don't defend the agency as infallible. Their point is narrower: have some structured evidence floor, even a modest one, and say so plainly.

The supply-chain argument — and I mostly agree with it

This is the part of their case I find hardest to argue with, because it's not opinion, it's a published lab analysis. Van Wagoner et al., 2017, in JAMA (PMID 29183075), bought 44 products sold online as SARMs and ran them through mass spectrometry. The results: 9% contained no active compound at all, 25% contained substances not listed on the label, only 41% actually had the labeled amount of active ingredient, and 39% contained other unapproved drugs entirely.

Their argument: if research-chemical SARMs sold through the same "not for human consumption" gray-market channel look like that, assuming peptides from the same lanes are reliably what the label says is unsupported.

I agree with the data completely. Where I part ways is the conclusion. They conclude: don't use peptides. I conclude: use only vendors with documented third-party Certificates of Analysis, verified purity, and US manufacturing — and skip the unverified gray market entirely. Same evidence, different prescription.

Where they draw the line on specific compounds

They're not blanket-skeptical. They endorse the GLP-1 class (semaglutide, tirzepatide) as having "mountains of evidence," and they endorse tesamorelin and PT-141 for their actual FDA-approved indications. Their skepticism concentrates on the research-chemical peptides: BPC-157 (no published human RCT exists, and ~90% of the research comes from one lab group), TB-500 (topical wound-healing data exists, injected systemic use doesn't), CJC-1295 (they cite a trial shutdown after a cardiac event), MK-677 (a documented congestive-heart-failure signal), and melanotan (rhabdomyolysis and acute kidney injury cases — on this one, I agree without reservation; it's the highest-risk compound in the category). See the BPC-157 skeptical-coverage companion piece for the deepest version of that specific argument.

"If it worked, pharma would sell it" — this is where I disagree most

Their version: if BPC-157, CJC-1295, MK-677, and similar compounds actually worked, pharma economics would have brought them to market by now. The fact that several were abandoned mid-development is evidence they don't work as advertised.

I think this proves less than it sounds like it proves. Insulin's own history is the counter-example — Banting and Best gave away the original patent for a dollar. Many research peptides share the same structural problem: they're old, unpatentable in their natural form, and the trial economics don't pencil out when the compound would compete against an already-patented blockbuster in the same indication, or targets a market too small to justify a billion-dollar trial program, or treats an indication — healthy-aging and longevity — that the FDA doesn't currently recognize as an approvable disease category at all. Pharma walking away from a compound is one data point. It can mean the molecule failed, or it can mean the math failed. Those are different claims, and the "if it worked they'd sell it" argument usually doesn't distinguish them. That said, it's not zero-content either — the BPC-157 development history and the AOD-9604 trial failure are real yellow and red flags worth knowing about, not evidence to wave away.

The FDA's own funding structure is worth knowing about

Separate from the specific-compound debate, there's a structural point worth stating plainly: the FDA's drug-review budget is roughly 77% funded by user fees paid by the companies whose drugs it reviews (Congressional Research Service R44750), a rate schedule set under the Prescription Drug User Fee Act. A peer-reviewed BMJ study (Bien & Prasad, 2016) found that of 26 FDA cancer and hematology reviewers who left the agency between 2001 and 2010, 15 — 57% — later worked for or consulted with the pharmaceutical industry. Scott Gottlieb went from FDA Commissioner to Pfizer's board within two months of leaving the agency. The pharmaceutical industry spent $391.5 million on federal lobbying in 2024 alone (OpenSecrets). None of this is hidden — it's public-record data, and it's the reason "trust the FDA" isn't a neutral position to hold on either side of this debate. It doesn't mean FDA reviewers are corrupt; most are careful career scientists. It means the incentive gradient is real, and it cuts against the "no FDA approval means unsafe" framing just as much as it cuts against blind trust in industry.

Where we actually agree

Strip out the social-media noise and the real overlap is bigger than either side's loudest voices admit: the gray-market supply chain is a documented mess and verified sourcing solves the same problem both camps are worried about. FDA-approved peptides used for their approved indications are genuinely well-supported. Anecdote isn't a randomized trial, and nobody serious argues otherwise. And a handful of specific compounds — melanotan chief among them — carry real, well-documented risk that no amount of "signal vs. flood" framing changes.

Where we actually disagree

Four things, honestly: how much weight animal evidence deserves when it's the entire evidence base for a compound (I weigh it more than they do); how much weight decades of consistent practitioner observation deserves (I treat it as meaningful signal, they treat it as anecdote); how to read pharma abandonment (I weigh the economics explanation alongside the "it didn't work" explanation, they lean harder on the latter); and whether an informed adult using a verified-vendor, well-sourced compound is a defensible path (I think it is; they think unsupervised use sits outside responsible practice).

None of those are stupid disagreements on either side. They're honest disagreements about how to weigh evidence and risk under real uncertainty — which is exactly why I think you deserve to see both arguments instead of just mine. Have your own evidence standard, and be able to say it out loud. That's the actual takeaway, whichever side you land closer to.

See also: the compounding-pharmacy pathway (503A/503B) for how the FDA-approved / compounded / research-chemical pathways actually differ.

Educational information only, not medical advice.

Sources: Docs Who Lift × Barbell Medicine crossover discussion, Spencer Nadolsky MD, Karl Nadolsky MD, Austin Baracki MD, Jordan Feigenbaum MD (2025); Van Wagoner RM, Eichner A, Bhasin S, Deuster PA, Eichner D, "Chemical Composition and Labeling of Substances Marketed as Selective Androgen Receptor Modulators and Sold via the Internet," JAMA 2017;318(20):2004–2010, PMID 29183075; Echt et al., CAST trial, NEJM 1991;324:781–788; Barter et al., ILLUMINATE (torcetrapib) trial, NEJM 2007;357:2109–2122; Bien J, Prasad V, "Public health and prescription drugs: a research letter on FDA medical reviewers' subsequent employment in industry," BMJ 2016;354:i5055; Congressional Research Service R44750, "FDA Human Medical Product User Fee Programs"; OpenSecrets, Pharmaceuticals/Health Products Lobbying Profile 2024; Association for Molecular Pathology v. Myriad Genetics, Inc., 569 U.S. 576 (2013); DiMasi JA, Grabowski HG, Hansen RW, "Innovation in the pharmaceutical industry: New estimates of R&D costs," J Health Econ. 2016;47:20-33.