CROSS-COMPARISON / THE REGULATORY LADDER
Four compounds, ranked twice: once by evidence, once by law
The two rankings mostly agree, and where they part company is where the interesting reading is.
The short version
This page puts the four compounds side by side twice. The first ranking is by how much human evidence exists. The second is by where each one sits in law. They line up more closely than a sceptic might expect — and the places where they come apart are the ones worth understanding.
Semaglutide and tirzepatide are approved prescription medicines with large randomised trials behind them. NAD+ has real human trials but sits in a supplement category that is itself being argued over. BPC-157 has an enormous animal literature, almost no human one, and no approval anywhere.
So far, so tidy. The complication is that neither ranking is a fixed property of a molecule. Approval is a decision about a filed application in one country on one date, and evidence accumulates on a schedule set by whoever chose to fund a trial. Read together, the two columns describe a process, not a verdict.
The four at a glance
| Compound | What it is | Regulatory category | Human evidence | Prohibited in sport |
|---|---|---|---|---|
| BPC-157 | Synthetic 15-amino-acid peptide from a gastric-juice protein | Not approved as a medicine anywhere; distributed for laboratory research use only | Three small pilot studies; almost everything else is rodent work [2] | Yes — at all times, under the non-approved-substances category |
| NAD+ | Redox coenzyme, not a peptide; studied via the precursors NMN and NR | Dietary supplement, with NMN's supplement status contested; injectable forms typically compounded | Randomised trials showing the blood marker rises [10] [7]; clinical benefit unproven [6] | No — NAD+, NMN, NR and nicotinamide are not prohibited |
| Semaglutide | 31-amino-acid engineered GLP-1 analogue | Approved prescription medicine across several indications and two formulations | Outcome trials in 17,604 [13] and 3,533 [12] participants, plus weight trials [14] | Not specifically prohibited |
| Tirzepatide | Synthetic dual GIP and GLP-1 receptor agonist | Approved prescription medicine, indication by indication from May 2022 [18] | Phase 3 trials in 2,539 [20] and 1,879 [21] participants, plus a head-to-head [11] | Not specifically prohibited |
The sport column is the only one in that table with a single global answer, because the World Anti-Doping Agency's list applies across signatory countries at once. Every other column is a national answer wearing a general voice.
Ranked by evidence maturity
Reading the four by the weight of their human record produces a clear ranking, and the gaps between the rungs are larger than the rungs themselves.
Semaglutide sits at the top on volume and endpoint hardness. Its trials measure things that cannot be argued about — death, myocardial infarction, stroke, kidney failure — in populations of tens of thousands [13] [12], and its weight trial reported a mean change of -14.9% against -2.4% with placebo at 68 weeks [14].
Tirzepatide is close behind and gaining, with a 2,539-participant placebo-controlled phase 3 trial [20], a 1,879-participant active-comparator trial in diabetes [21], and a dedicated meta-analysis of nine trials and 9,871 participants examining two specific safety signals [19]. Its outcome-trial record is younger, which is a statement about calendars rather than quality.
NAD+ occupies a genuinely awkward middle. Its trials are real and randomised, and they consistently show the thing they set out to move: whole-blood NAD+ rises with precursor supplementation [10] [7]. What they have not shown is that the rise buys anything durable, and a 2025 review said so directly [6].
BPC-157 is last by a wide margin, and not because its animal data are weak. They are broad and consistent [5] [4]. There is simply almost no human record: three small pilot studies as of 2025 reviews, one of them a two-person intravenous safety report [1] [2], and a foundational literature concentrated in a single research group [2].
Where the two rankings come apart
Compare the evidence ladder with the legal ladder and three mismatches stand out.
NAD+ has more human evidence than its category implies. As a dietary supplement it is regulated more lightly than either approved medicine here, yet it has randomised, placebo-controlled, dose-ranging human trials [10] [7] [8]. Category is not a proxy for evidence, in either direction.
BPC-157's uniform non-approval hides a non-uniform reason. "Not approved anywhere" reads like a global verdict but is mostly the absence of a filing anywhere. What differs by country is the handling of a substance in that position — whether it may be compounded, how a research-use label is treated in law — and the United States example, its 2023 placement in a category of bulk substances not eligible for compounding under section 503A, is a decision about a pathway rather than about the pharmacology.
The two approved medicines are approved differently in different places on different dates. Tirzepatide's own record shows a first approval for type 2 diabetes in May 2022 [18], a later one for chronic weight management, and a later one again for obstructive sleep apnoea. Semaglutide accumulated indications on its own schedule. Neither sequence is universal, and a reader who takes one country's current list as the state of the world will be wrong somewhere.
One further asymmetry belongs here. Semaglutide's compounding episode — an availability expansion during a declared shortage, then a contraction when the shortage was declared resolved in 2025 — moved what people could obtain twice without moving the trial record once. On the evidence ladder, nothing happened at all.
How to read a study conducted somewhere else
The practical payoff of all of the above is a short list of things worth checking before a country of origin is allowed to influence a judgement about a trial.
Design before flag. The NMN trial cited here is multicentre, double-blind and randomised [7]; SURMOUNT-1 is double-blind and placebo-controlled [20]; SURMOUNT-5 and SURPASS-2 are open-label [11] [21]. Open-label is a real limitation on subjective endpoints, and it is a limitation wherever the trial was run.
The comparator outranks the country. SURPASS-2 compared tirzepatide with semaglutide 1 mg [21] — a diabetes amount, not the higher weight-management one. A comparator choice changes the meaning of a result far more than a national border does.
Endpoints are different currencies. A hazard ratio for major cardiovascular events [13] and a percentage change in body weight [14] do not convert into one another. Neither does a percentage rise in a blood coenzyme [10] against a change in how far somebody can walk [7].
Approval status is not a quality grade. A molecule missing from one country's approved list may have failed there, may never have been filed there, or may still be pending. Presence somewhere is not evidence of safety for any particular person; absence somewhere is not evidence of failure. These are the two mirror-image errors this whole site exists to name.