Promise chose 36 compounds for this peptide evidence map from those marketed or widely discussed by U.S. telehealth and wellness sellers. Semaglutide, tirzepatide and oxytocin serve as reference points with approved medicines. This is not a complete list of peptides. 5-Amino-1MQ, MK-677, SLU-PP-332, NAD+, NR and NMN are not peptides.

ClinicalTrials.gov, the U.S. public trial registry, lists studies in which researchers assign people a treatment or test. Of the 28 compounds on this list with no matching FDA product, 14 had no completed study of that kind. For 11 of those 14, PubMed also yielded no qualifying paper on a trial that assigned people to groups by chance. Older studies can escape either search.

AI models screened records against written rules and checked their decisions against the registry or PubMed record.

Promise has a commercial interest because several listed compounds are prescribed through it.

At a glance

  • Across the full list, the PubMed search found no qualifying randomized-trial paper for 12 compounds. PubMed is a public index of medical papers.
  • Automated sums across all study groups in unscreened completed registrations gave 120,835 enrollments for semaglutide and 40,069 for tirzepatide, while one hand-screened BPC-157 gummy study reported 40 people. The sums include controls and can count a person more than once.
  • The LL-37 search returned 42 papers. At least 32 measured the body's own peptide or how much it made. Three papers qualified as trials that gave LL-37 to people.
  • A separate review traced 101 more human studies, including 15 for Semax. It followed selected leads, including papers already in PubMed that the trial filter missed.

Compounds with no completed study in the registry search

  • Also no qualifying PubMed paper — 5-Amino-1MQ, Dihexa, Epitalon/Epithalon, IGF-1 LR3, KPV, Melanotan II, MOTS-c, PE-22-28, Pinealon, Semax and SLU-PP-332.
  • The other three — CJC-1295, DSIP and Selank had qualifying PubMed papers. The Melanotan II crossover and one Selank paper describe giving the compounds to people, but their reports leave chance assignment unclear. Melanotan II crossover · Selank combination trial

BPC-157 also had no qualifying PubMed paper, but it had a completed registration.

Peptide evidence map by compound

Compound Registered Completed Reported enrollment PubMed Other studies FDA product
Oxytocin 662† 410† 74,674†¶ 2,023† / Not verified Not reviewed Pitocin, Syntocinon (first year unverified)
Semaglutide 574† 264† 120,835† 318† / Not verified Not reviewed Ozempic, Rybelsus, Wegovy (2017)
Tirzepatide 225†‡ 72† 40,069† 137† / Not verified Not reviewed Mounjaro, Zepbound (2022)
SS-31 (elamipretide) 30 22 1,413 13 / 10 3 Forzinity (2025)
Melanotan I (afamelanotide) 22 17 581 6 / 8 Not reviewed Scenesse (2019)
Tesamorelin§ 21‡ 13 1,260 21 / 10 11 Egrifta, Egrifta SV, Egrifta WR (2010)
PT-141 (bremelanotide)§ 9 8 2,308¶ 12 / 10 16 Vyleesi (2019)
Sermorelin§ 1 1 Not reported¶ 5 / 5 4 Geref (1990, discontinued)
Nicotinamide riboside (NR; separate precursor)§ 124 57 3,150 29 / 26 Not reviewed None
Glutathione (administered drug)§ 67 44 3,369¶ 32 / 28 Not reviewed None
Thymosin alpha-1 (thymalfasin)§ 65 22 5,336¶ 55 / 51 1 None
Nicotinamide mononucleotide (NMN; separate precursor)§ 35 16 905 16 / 14 Not reviewed None
Retatrutide 33‡ 22 7,887 11 / 6 Not reviewed None
TB-500 / thymosin beta-4 (combined union) 20‡ 12 2,149 6 / 6 3 None
Thymosin beta-4 (full protein) 19 12 2,149 6 / 6 3 None
MK-677 (ibutamoren) 14 7 756¶ 17 / 14 2 None
Kisspeptin§ 13 8 376 18 / 16 2 None
NAD+ (administered NAD+ itself)§ 11 4 175 1 / 1 9 None
VIP (vasoactive intestinal peptide / aviptadil)§ 8 4 754¶ 21 / 18 Not reviewed None
BPC-157 4‡ 1 40 0† / 0† 5 None
GHK-Cu§ 3‡ 1 43 1 / 1 4 None
Ipamorelin 2 2 437 2 / 2 0 None
LL-37 2 1 4 3 / 3 0 None
CJC-1295 1 0 0 1 / 2 1 None
Epitalon/Epithalon 1 0 0 0† / 0† 3 None
Melanotan II§ 1‡ 0 0 0 / 0 2 None
MOTS-c 1‡ 0 0 0 / 0 0 None
TB-500 (fragment name) 1‡ 0 0 0† / 0† 0 None
5-Amino-1MQ 0† 0† 0† 0† / 0† 0 None
Dihexa 0 0 0 0 / 0 0 None
DSIP (delta sleep-inducing peptide)§ 0 0 0 2 / 2 13 None
IGF-1 LR3 0† 0† 0† 0† / 0† 0 None
KPV 0† 0† 0† 0† / 0† 0 None
PE-22-28 0† 0† 0† 0† / 0† 0 None
Pinealon§ 0† 0† 0† 0† / 0† 5 None
Selank§ 0† 0† 0† 1 / 1 3 None
Semax§ 0 0 0 0 / 0 15 None
SLU-PP-332 0† 0† 0† 0† / 0† 0 None
  • Registered includes interventional studies at any stage. Researchers assign the treatment or test.
  • Completed means the registry marks the study Completed.
  • Reported enrollment adds the actual counts for all study groups in completed registrations, including controls.
  • PubMed gives papers, then trials. Several papers can share a trial. One paper can cover several trials.
  • Other studies counts verified additions from other registries or publications outside the PubMed filter.
  • FDA product names a matching product in FDA records. An approval covers a specific product for a specific use and does not extend to other products containing the same molecule. Compounded medications are not FDA-approved. The year is the earliest confirmed. “None” means no match. The downloads give marketing status. Drugs@FDA data files

† An automated search or screen supplied the number. A marked PubMed total counts candidates. A marked zero means none came back. The download gives frozen and repeat counts for the nine empty searches checked again on November 20, 2026.

‡ The marked counts include eight registrations from one sponsor that list the same single site. ClinicalTrials.gov does not verify recruitment for any sponsor. The CSV also gives counts with these records set aside.

§ Some records remain unresolved. Three DSIP papers leave chance assignment unclear. Three registrations name growth hormone-releasing hormone (GHRH) without identifying sermorelin. The downloads list the records.

¶ The CSV counts completed records missing actual enrollment. “Not reported” means all completed records lack it.

The Other studies cells add to 105 because one study appears under both Selank and Semax and the thymosin beta-4 rows overlap (101 distinct studies), while nine rows carry ‡ for eight registrations because the TB-500 fragment registration also appears in the combined row.

“Not reviewed” means the separate search did not cover the compound. A zero means it found no eligible addition. Use the combined thymosin beta-4 entry for the full protein and TB-500 fragment together. All additions involved the full protein. NR and NMN count separately.

What stands out

Four BPC-157 registrations, one marked completed

The completed record describes 40 people in one group taking a gummy containing BPC-157. It had no posted results. ClinicalTrials.gov study record

As of November 20, 2026, the day this article was written, a recent case series describes people who received BPC-157 alongside other treatments. This case series, a report on a group of patients, fell outside the five additional BPC-157 studies in the review. Case series in Frontiers in Oral Health

How enrollment was counted for semaglutide and other compounds

For retatrutide, completed records report 7,887 enrollments. For GHK-Cu, 43 people took part in a skin-care study. Every assigned night cream contained GHK-Cu alongside other ingredients. The study could not isolate GHK-Cu from the rest of the regimen. GHK-Cu registry record

Sermorelin’s completed record has no enrollment field labeled Actual. The linked paper reports participants, but the registry sum uses the registry field. Sermorelin study record

Compounded versions are not FDA-approved, although an FDA-approved sermorelin product was once marketed and has since been discontinued. FDA notice on Geref

Why LL-37 and VIP lost so many search hits

One LL-37 trial assigned people to peptide cream or placebo. The count excludes studies of bacteria that release LL-37. Miranda and colleagues’ LL-37 trial

For vasoactive intestinal peptide, or VIP, 21 of 99 PubMed candidates qualified. Many measured the body’s own VIP. Some older papers describe people receiving VIP but leave chance assignment unclear.

What the MOTS-c papers measured

All four PubMed candidates measured MOTS-c made by the body. None tested giving it to people. The registry lists one MOTS-c study plan.

Eight registrations share a sponsor

The counts include eight registrations from one sponsor that list the same single site. All list their status as Recruiting. ClinicalTrials.gov does not verify recruitment for any sponsor. Two descriptions use the phrases “This example trial” and “this public example”.

Why thymosin alpha-1 has more papers than trials

The 55 qualifying papers describe 51 identifiable trials. TESTS has a report and a correction. Older papers without a shared trial number may describe the same people. TESTS trial report in The BMJ

What these numbers cannot tell you

To assess benefits or risks, you need the results, who took part and how the study was run. A registered study may never post results. A small BPC-157 pilot counts once, just as a much larger trial does. Lee and Burgess’s pilot study

The 101 additional studies included 44 randomized studies, 20 non-randomized trials, 4 case series and 33 studies of other or unclear design. A study can appear under multiple compounds or in both databases. Adding columns would count some trials twice.

Some counts still come from automated searches. The 318 semaglutide papers are candidates. Their eligibility and trial groupings remain unverified. Registry studies also vary in stage. How clinical trial stages differ

How we counted

On November 20, 2026, we searched ClinicalTrials.gov and PubMed using compound names and their aliases. We used the registry’s study type and status. On PubMed, we used its filter for randomized human trials.

AI models worked independently from written rules and checked decisions against the registry or PubMed record. A further review settled disagreements against the primary record. For a study to qualify, people had to receive the compound as part of what researchers were testing. We set aside diagnostic tests, comparison arms used solely to study another treatment, and papers about a different molecule. We grouped papers by trial where the records allowed it.

The separate review followed leads in other registries and publications. Older work, including papers in Russian, can be hard to find. Some papers lack an abstract or accessible methods. The downloads give the full method, the underlying study records and reasons for excluding earlier candidates.

Promise's interest

Promise is a telehealth company. Prescriptions written through its prescriber network are filled by licensed U.S. compounding pharmacies. Compounded medications are not FDA-approved. Several compounds in this table are prescribed through Promise, so Promise has a commercial interest in how they are perceived. This map reports what the registries and the literature contain and takes no position on any compound.

A licensed provider reviews every request at Promise, and not everyone qualifies. The patient and provider make prescribing decisions.

How to cite and download the map

Cite Promise. Peptide Evidence Map. Version 2, November 20, 2026. Link here and name the compound and column. The separate BPC-157 guide describes the publications. BPC-157 human studies

Corrections are welcome at max@mypromise.com. Confirmed corrections are logged in the download’s change log with the date.