I get a lot of questions about peptides. Most of the people asking have already done real reading — podcasts, forums, a friend who swears by something — and what they want from me is not a lecture about FDA approval. They want to know what is actually true. So this page tries to do one thing: tell you, agent by agent, what we know, how well we know it, and what I'd want to check before using it in you.
A few ground rules for how I've written this, because they explain the shape of every entry below.
"Not FDA-approved" is a fact, not a verdict. Most of these molecules were never entered into large trials because nobody could patent them well enough to justify spending a few hundred million dollars proving they work. Absence of trials is not the same as evidence of no effect. But it also isn't evidence of effect, and I won't pretend a rat study is a human result. For each agent I'll tell you what was studied in people, in how many, at what dose, and what happened — with the numbers.
Molecules your body already makes deserve some benefit of the doubt — with limits. If a peptide occurs naturally in you, your body has enzymes that break it down and receptors that are used to seeing it. That makes it more likely to be tolerable at reasonable doses. It does not make it safe at any dose, by any route, forever. Several agents on this list are sold as "natural" when they are actually synthetic fragments, modified analogs, or not peptides at all. I flag which is which. One correction worth making up front, because I hear it constantly: the popular "growth hormone peptides" (sermorelin, CJC-1295, ipamorelin) are not pieces of growth hormone. They act on the pituitary to release your own — a real distinction that matters for both benefit and risk. Details in that section.
A contraindication is only as good as its reasoning. When I say "don't use this if you have X," I'll tell you whether that's Founded — meaning documented human harm or a regulator's label — or Theoretical — meaning it follows from the mechanism but nobody has actually observed it, and I'll note research that supports or undercuts it. Theoretical cautions are still worth respecting when the downside is serious (cancer, pregnancy). They are not the same thing as proven danger, and you deserve to know the difference.
The 2026 regulatory picture, in one paragraph. In 2023 the FDA put most of these peptides on a list of compounding ingredients that "may present significant safety risks," which effectively stopped licensed pharmacies from making them. In April 2026 the FDA removed twelve of them from that list after the original nominations were withdrawn, and in July 2026 an FDA advisory committee voted — over the objection of FDA's own scientists — to recommend allowing pharmacies to compound six (BPC-157, TB-500, KPV, MOTS-c, semax, epitalon). That vote is advisory. Until formal rulemaking finishes, none of those six is legally compoundable, and most of what's sold online is "research chemical" product with no sterility, identity, or dose verification. That last point is not a technicality: for most agents here, the biggest documented risk is not the molecule, it is the vial.
How to read the evidence label on each entry
Established — Multiple randomized trials or an approved label with consistent results.
Probable — Human trials suggest a real but modest effect, with some inconsistency.
Promising — Small or early human studies are encouraging; replication needed.
Plausible — Mechanism and animal data are credible; direct human outcome data are thin or absent.
Unknown — Not enough evidence to estimate the effect in humans at all.
Negative — Adequate human trials did not show meaningful benefit at studied doses.
"Endogenous" on a badge means the exact molecule occurs naturally in the human body. "Analog" or "fragment" means it is related to something natural but is not the same thing.
Healing and tissue repair
BPC-157, TB-500, GHK-Cu, KPV, LL-37 — the "recovery" peptides.
BPC-157
Body Protection Compound 157 · "pentadecapeptide" · sold as acetate (injectable) or arginate/"PDA" (oral)
Synthetic 15-amino-acid peptideNot endogenousEvidence: Plausible
What it is
A 15-amino-acid sequence that its developers describe as derived from a protective protein in human gastric juice. The intact peptide itself has not been shown to exist in the body, so despite the marketing it is best thought of as a synthetic molecule inspired by a natural one. It is unusually stable in stomach acid, which is why it is sold orally as well as injected.
What we know
The animal literature is large and consistently positive — tendon, ligament, muscle, gut, and blood-vessel healing in rats — but almost all of it comes from a single research group in Zagreb, and a 2025 systematic review of BPC-157 in sports medicine found 36 studies, of which 35 were in animals. In people: one randomized trial (53 patients, ulcerative colitis, 80 mg enema daily for two weeks, 2005) did not beat placebo. Beyond that, the human record is a 16-patient knee-pain chart review from one Florida clinic (87% reported relief after a single injection; no controls, no imaging), a 12-woman bladder pilot, and a two-person IV safety pilot. There is no published, peer-reviewed randomized trial showing BPC-157 works for anything in humans. The claim that the oral arginate form is "90% bioavailable" appears in no peer-reviewed study; formal pharmacokinetics in rats and dogs show a half-life under 30 minutes and no oral data at all.
Doses in the research
Animal efficacy doses are typically 10 micrograms/kg (and, oddly, 10 nanograms/kg — the studies report both working). Human exposures: 80 mg rectally, 2–4 mg into a joint, 10 mg into the bladder, 10–20 mg IV. The common online regimen of 250–500 mcg injected once or twice daily is not derived from any human study.
Contraindications and cautions
TheoreticalActive cancer. BPC-157 strongly activates VEGF/angiogenesis signaling, and tumors need blood vessels. No animal study has shown it accelerates a tumor — in one mouse colon-cancer model it reversed muscle wasting without changing tumor size — but nobody has looked in humans. I would not use it in anyone with an active malignancy, and I'd say so plainly rather than pretend the mechanism is reassuring.
TheoreticalLiver disease. Twenty-eight-day animal toxicology showed ALT and metabolic changes at higher doses. No human signal.
TheoreticalPregnancy and breastfeeding. No data. Avoid.
FoundedProduct quality. FDA's adverse-event database holds a handful of reports (injection-site reactions, one case of shortness of breath, one case of diffuse skin and gum darkening from a combined BPC/TB-500 product that recurred on re-exposure). All involved unregulated product. That's the real-world risk.
Status
Not approved anywhere. On FDA's "do not compound" list from September 2023 until April 2026; FDA advisory committee voted 8–6 in July 2026 to recommend allowing compounding, over FDA staff objection; not yet legal to compound pending rulemaking. Prohibited at all times by WADA (an athlete was banned four years for BPC-157 plus TB-500).
My takeThe mechanism is credible and the animal data are hard to dismiss, but the human evidence is a negative trial and a few uncontrolled case series. I'd keep this in the plausible-but-unproven category: low apparent toxicity, no demonstrated human benefit, and a supply chain you can't trust. If you're using it, the most useful conversation we can have is about what else is driving the injury that isn't healing — that's usually where the leverage is.
Key sources
TB-500 and Thymosin β4
Tβ4 · TB4 · "TB-500" (a 7-amino-acid fragment, residues 17–23, usually acetylated)
Thymosin β4: endogenous 43-amino-acid protein"TB-500": synthetic fragmentEvidence: Tβ4 Probable (eye) · TB-500 Unknown
What it is — and why the name matters
Thymosin β4 is a real human protein, present in nearly every cell and in wound fluid, that organizes the cell's actin skeleton and promotes cell migration. It has been through actual clinical trials. "TB-500," as the FDA and most vendors define it, is a short synthetic piece of that protein — seven amino acids with an acetyl group added, which changes its charge and behavior. Gray-market products labeled TB-500 have variously contained the fragment, the full protein, or a different fragment altogether. When someone tells me TB-500 is "natural," the honest answer is: the full protein is; the fragment you probably bought isn't, and has never been given to a human in a published study.
What we know
Full-length Tβ4 as an eye drop (RGN-259) healed corneas in a small neurotrophic keratopathy trial (6 of 10 vs 1 of 8 on placebo), but a larger European Phase 3 missed its primary endpoint in June 2025, and three dry-eye Phase 3 trials totaling more than 1,600 patients missed their co-primary endpoints. A topical Tβ4 gel for pressure ulcers (72 patients) was safe but no better than placebo. IV Tβ4 was given to 40 healthy volunteers at single doses up to 1,260 mg without serious problems, and a 2025 Chinese trial in 96 heart-attack patients found no overall reduction in infarct size. For the TB-500 fragment: FDA's 2026 review found zero human studies, zero adverse-event reports, and one lab study in which the fragment did not promote wound healing in cell culture.
Doses in the research
Full Tβ4: 0.1% eye drops; 0.01–0.1% topical gel; IV up to 1,260 mg single dose. TB-500 fragment: no human dose has ever been studied. The online pattern of 2–5 mg per week has no clinical basis.
Contraindications and cautions
TheoreticalActive cancer — and here the theory has teeth. Tβ4 drives cell migration and new blood vessel growth. In a 2003 animal study, forcing melanoma cells to overexpress Tβ4 increased tumor size, quadrupled tumor blood-vessel density, and raised lung metastases from about 11 to 47 nodules. That is animal data on the protein, not human data on the fragment, but it is a more specific signal than BPC-157's. I would not use any Tβ4 product in someone with an active or recent cancer.
TheoreticalPregnancy and breastfeeding. No data.
FoundedProduct identity. Mislabeling between fragment and full protein is documented and common.
Status
Neither form approved anywhere. TB-500 fragment: on FDA's do-not-compound list 2023–April 2026; advisory committee voted 8–6 in July 2026 to recommend allowing compounding; rulemaking pending. WADA-prohibited at all times (it is named specifically in the growth-factor section).
My takeThe full protein is a legitimate investigational drug with a mixed record in the eye and a clean safety record in healthy volunteers. The fragment sold as TB-500 is an unstudied piece of it. If the goal is a tendon or muscle that won't heal, I'd want to look at the mechanics first — that is what
osteopathic evaluation is for — and I'd be candid that injecting an unstudied fragment is a hope, not a plan.
Key sources
GHK-Cu
Copper peptide · copper tripeptide-1 · glycyl-histidyl-lysine copper
Endogenous tripeptideEvidence: Probable (topical) · Unknown (injected)
What it is
A three-amino-acid peptide that genuinely circulates in human blood, bound to copper. Levels fall with age — roughly 200 ng/mL in your twenties to about 80 by sixty — and it's released from collagen when tissue is injured, which is the basis for the "repair signal" framing. This is one of the few agents on this page where the natural-molecule argument is entirely accurate.
What we know
Applied to skin, it has a respectable record: a 1994 randomized trial of 2% GHK-Cu gel on diabetic ulcers reported about 40% better wound closure and fewer infections than vehicle, and several small cosmetic trials (12 weeks, 41–71 women) show measurable improvements in skin thickness, laxity, and wrinkle depth versus placebo, including head-to-head wins over vitamin C and retinoic acid for collagen production in a 20-person study. Injected or systemic GHK-Cu, which is how it's now being sold for "whole-body" repair and hair, has no human data at all — every systemic result is from mice, rats, or pigs. The leading review is written by the company that sells it, which doesn't make it wrong but is worth knowing.
Doses in the research
Topical: 2% gel for wounds; 0.05–1% in cosmetic products. Injectable: no human dose exists. The "50 mcg/kg daily" figure circulating online comes from vendor monographs.
Contraindications and cautions
FoundedWilson's disease. A copper-delivering agent in a copper-handling disorder is an absolute no on mechanism alone.
TheoreticalCopper load with injection. Topical use delivers trivial copper. Injecting it bypasses the skin's regulation, and nobody has measured what repeated systemic doses do to copper balance or immune response. This, not the peptide's biology, is the real unknown.
TheoreticalActive cancer. Copper supports angiogenesis; GHK modulates hundreds of genes. No human tumor data either way.
TheoreticalPregnancy. No systemic data.
Status
Topical: legal cosmetic ingredient, no prescription needed. Injectable: on FDA's do-not-compound list 2023–April 2026; removed, but not on the July 2026 advisory agenda; FDA review scheduled before February 2027. Not WADA-listed.
My takeFor skin, this is a reasonable, well-tolerated topical with real if modest data behind it — buy a product with a stated concentration and use it. For injection, you are the experiment. I would rather match the evidence: use it where it was studied.
Key sources
KPV
Lysine-proline-valine · the last three amino acids of α-MSH
Fragment of an endogenous hormoneEvidence: Plausible (animal only)
What it is
The tail end of alpha-melanocyte-stimulating hormone, a natural anti-inflammatory hormone. The tripeptide keeps the anti-inflammatory action and loses the skin-darkening action, which is an attractive idea for gut inflammation.
What we know
In mice, oral KPV is taken up by a transporter in the colon lining and reduces the severity of chemically induced colitis (Gastroenterology, 2008), with follow-up work confirming it in other inflammatory-bowel models. In humans: nothing. FDA's July 2026 review found zero clinical studies, zero pharmacokinetic data, zero adverse-event reports, and zero compounded products ever reported by outsourcing pharmacies. One lab study found it penetrates human skin poorly, which undercuts the topical products.
Doses in the research
No human dose has been studied. Online doses (200–500 mcg injected or by mouth) are vendor-derived.
Contraindications and cautions
Everything here is Theoretical, because there is no human safety data to ground anything: pregnancy (no data), active cancer (no data; it lacks the pigment-cell activity of its parent hormone), and the usual unregulated-product concern for injected forms.
Status
Not approved anywhere. Do-not-compound list 2023–April 2026; advisory committee voted 8–6 in July 2026 to recommend allowing compounding; rulemaking pending. Not WADA-listed.
My takeA tidy idea with an entirely animal evidence base. If you have inflammatory bowel disease, we have treatments with human outcome data, and I'd want you on something that's been shown to prevent damage before we talk about an adjunct that's never been dosed in a person.
Key sources
LL-37
Human cathelicidin · ropocamptide (investigational drug name)
Endogenous antimicrobial peptideEvidence: Promising (topical, large ulcers) · Unknown (injected)
What it is
The one antimicrobial peptide of its family that humans make. Your white blood cells release it at sites of infection; it punches holes in bacterial membranes and recruits immune cells. Fully natural — and that cuts both ways, as you'll see.
What we know
Topically, on hard-to-heal venous leg ulcers, a 2014 dose-finding trial found something unusual: the lowest dose (0.5 mg/mL) worked best, shrinking ulcers about 68%, while the highest dose lost its effect because LL-37 at high concentration damages your own cells too. A 2021 Phase 2b trial of 148 patients then missed its overall primary endpoint (about 25% healed in every arm), with benefit only in ulcers larger than 10 cm² (28% healed vs 8%). A 25-patient diabetic-foot trial in 2023 improved granulation tissue but not wound size. There is no human data for injected or systemic LL-37, which is how it's being sold for "immune support."
Doses in the research
Topical 0.5–1.6 mg/mL twice weekly. Higher was worse. No systemic human dose exists.
Contraindications and cautions
Theoretical — but this is the best-documented theoretical caution on the page. Psoriasis, rosacea, and lupus. In psoriasis, LL-37 is overproduced, binds to your own DNA, and drives the autoimmune inflammation — it's a validated psoriasis autoantigen. In rosacea, abnormal processing of LL-37 produces the inflammatory fragments that cause the disease. Adding more of it to someone with these conditions is adding fuel. No one has done the experiment in humans, and I would not.
TheoreticalCancer. LL-37 promotes tumor growth in breast, ovarian, and lung cancer models (and suppresses it in some gut cancers). FDA's 2023 listing specifically cited nonclinical tumor and reproductive findings.
FoundedDose ceiling. Cell damage at higher concentrations is demonstrated in the human trials, not just in a dish.
Status
Investigational (Swedish company). Do-not-compound list 2023–April 2026; removed; FDA review scheduled before February 2027. Not WADA-listed.
My takeReal molecule, real but narrow topical data, and an unusually clear reason for caution in common inflammatory skin conditions. As an injected "immune booster" it has no human basis, and the biology says the immune system already regulates this one tightly for a reason.
Key sources
↑ Back to list
The growth hormone axis
Sermorelin, CJC-1295, ipamorelin, tesamorelin, MK-677, and the older GHRPs.
First, what these actually are
Growth hormone is a 191-amino-acid protein made by your pituitary. None of the "GH peptides" below contain any piece of it. They are secretagogues — signals that tell your pituitary to release its own GH. They work through two different receptors, and that distinction is worth two minutes:
GHRH analogs (sermorelin, CJC-1295, tesamorelin) are copies or modifications of growth-hormone-releasing-hormone, a separate 44-amino-acid hypothalamic peptide. Sermorelin is literally the first 29 amino acids of your own GHRH — so that one is a natural fragment. The others are altered to last longer.
Ghrelin-receptor agonists (ipamorelin, GHRP-2, GHRP-6, hexarelin, MK-677) are fully synthetic. They aren't fragments of anything — they are small molecules designed in the 1980s–90s that happen to hit the "hunger hormone" receptor, which also releases GH. MK-677 isn't even a peptide; it's an oral small molecule that's marketed as one.
The one true GH fragment sold under this umbrella is AOD-9604, covered in the metabolic section.
Why the distinction matters: because these release your GH, the release stays under your body's normal feedback control, which caps how high it can go — a real safety advantage over injecting GH. But once GH is released, the downstream biology (IGF-1, fluid retention, blood sugar, any growth-promoting effect on tissue you'd rather not grow) is the same as with GH, just smaller. The class-wide safety section after MK-677 covers that.
Sermorelin
GHRH(1-29) · formerly Geref
Fragment of endogenous GHRH (unmodified)Formerly FDA-approved (1997–2008)Evidence: Established (children) · Promising (adults, biomarkers only)
What we know
Sermorelin was an FDA-approved drug for eleven years, for children with growth hormone deficiency (30 mcg/kg nightly), and it worked — less well than recombinant GH, which is why the manufacturer pulled it in 2008 for purely commercial reasons. That history means it has the best-characterized safety profile of anything in this section. In adults, the study everyone cites (Vittone 1997) gave a stabilized GHRH(1-29) analog to 19 people aged 55–71 at 10 mcg/kg nightly for 16 weeks: IGF-1 rose within two weeks, skin thickness increased, lean mass increased in the men only, and bone density didn't change. That is a small biomarker study — no strength, function, sleep, or quality-of-life outcomes were measured. There is no adequately powered adult trial of sermorelin for anti-aging, body composition, or sleep.
Doses in the research
Children: 30 mcg/kg under the skin at bedtime. Adults: 10 mcg/kg nightly (Vittone); 500 mcg twice daily for two weeks (Corpas 1992). The common clinic dose of 200–500 mcg nightly is in the same range but was not itself tested in a trial.
Contraindications and cautions
FoundedUntreated hypothyroidism blunts the response — fix thyroid first (from the original label).
FoundedPituitary disease or surgery. It needs a working pituitary to act on; mechanistic and on the tesamorelin label.
TheoreticalActive cancer — inherited from the GH/IGF-1 story below; no sermorelin-specific data. I treat it as a hard stop anyway.
TheoreticalBlood sugar, fluid retention. Mild and pulsatile; no glucose signal in the trials, but the studied durations are short.
FoundedAntibodies. A minority of children on long-term therapy developed anti-sermorelin antibodies, without loss of effect — this is the origin of FDA's immunogenicity concern with all compounded injectable peptides.
Status
No approved product currently exists in the U.S. Sermorelin is not on FDA's safety-risk list; pharmacies compound it under the argument that it's a component of a previously approved drug — a position FDA has not formally endorsed, so call it widely available and legally gray rather than settled. WADA-prohibited at all times.
My takeIf someone is going to use a GH secretagogue, this is the one with the longest human track record and the most physiologic profile — it's your own molecule, short-acting, and feedback-controlled. What it hasn't been shown to do is make an adult feel or function better. I'm comfortable with a monitored trial in the right person with defined goals, IGF-1 and glucose checks, and a stop date if nothing measurable changes.
Key sources
CJC-1295
With DAC (long-acting) · "without DAC" is properly Mod-GRF(1-29), a different molecule
Modified analog of endogenous GHRHNot endogenousEvidence: Plausible (raises GH/IGF-1; no outcome data)
What we know
There is essentially one human study. In 2006, two small placebo-controlled trials in healthy adults showed that a single injection of CJC-1295 with DAC raised GH 2- to 10-fold for about six days and IGF-1 1.5- to 3-fold for 9–11 days, with a half-life of about a week and no serious adverse effects over 28–49 days. That's it. No trial has ever measured whether it changes body composition, strength, recovery, or sleep in a person. The same year, the developer's 192-patient Phase 2 trial in HIV lipodystrophy was halted after one participant died. The treating physician attributed the death to a heart attack from undiagnosed coronary disease, not the drug — but the company never published the trial, never restarted it, and no one has taken the molecule into humans since. So the honest statement is: one unadjudicated death ended development, and we don't know what happened to the other 191 people. The "without DAC" version sold online has no human data at all.
Doses in the research
30–60 mcg/kg under the skin, once weekly or every other week, for up to seven weeks. Common online regimens (1–2 mg weekly) fall in that range; multi-year use has never been studied.
Contraindications and cautions
Theoretical Active cancer. Of all the secretagogues, this one produces the largest and longest IGF-1 elevation — and, because it's long-acting, a continuous rather than pulsatile GH signal, which is the opposite of the "physiologic" argument used for the class. The cancer concern is at its most credible here.
Theoretical Coronary artery disease. One death from plaque rupture in the only sizable trial, cause not established. I'd screen for it and disclose it.
Theoretical Blood sugar. Sustained GH elevation is diabetogenic; glucose wasn't reported in the human study.
Founded Pituitary disease (needs a working pituitary). Theoretical Pregnancy (no data; I'd treat it as contraindicated anyway).
Status
Not approved anywhere. Placed on FDA's safety-risk list in 2023; the nomination has since been withdrawn, so there is no pathway to legal compounding. WADA-prohibited at all times.
My takeThis is the secretagogue I'm least comfortable with, not because it's been shown to hurt anyone, but because the thing that makes it convenient — a week-long, non-pulsatile GH signal — removes the safety logic of the class, and the only real trial was stopped and buried. If someone wants GHRH stimulation, sermorelin or tesamorelin gets there with far more human data.
Key sources
Ipamorelin
Usually stacked with CJC-1295
Fully synthetic pentapeptideNot endogenous · not a GH fragmentEvidence: Negative (only trial) · Unknown (all marketed uses)
What we know
Ipamorelin's reputation as "the clean one" — GH release without cortisol or prolactin — comes from a 1998 study in rat pituitary cells, anesthetized rats, and pigs. It has never been confirmed in a human. The entire human record is a pharmacokinetic study in 48 volunteers and one randomized trial: 114 patients after bowel surgery, given 0.03 mg/kg IV twice daily for seven days to speed gut recovery. It didn't work (25 vs 33 hours to tolerate a meal, not significant). There is no human study of ipamorelin for body composition, muscle, fat, recovery, sleep, or aging, no study by the subcutaneous route, and no exposure longer than a week. A widely repeated claim that FDA restricted it because of "deaths with IV use" is false — FDA's own 2024 briefing states no deaths were reported. FDA's actual concerns were missing toxicology, possible addiction potential via the ghrelin receptor, reproductive signals in mice, and immunogenicity.
Doses in the research
0.03 mg/kg IV twice daily for seven days. The online standard of 200–300 mcg under the skin one to three times daily has no human basis — and daily use for years is roughly a hundred times the longest studied duration, by a route never studied.
Contraindications and cautions
Theoretical Active cancer (class IGF-1 concern; the elevation achievable is modest and the response fades with repeated dosing).
Theoretical Pregnancy — FDA specifically flagged reproductive/developmental findings from ghrelin-receptor agonism in mice.
Theoretical Blood sugar — ghrelin agonism acutely suppresses insulin; not documented as a clinical problem at studied doses.
Expected effect, not a caution: some appetite increase — it works through the hunger receptor.
Status
Not approved anywhere. On FDA's safety-risk list (2023, still listed for outsourcing pharmacies; nomination withdrawn for 503A). WADA-prohibited at all times.
My takeThe pharmacology is pleasant, the marketing is confident, and the human evidence is one negative trial. The selectivity claim may well be true — the animal data are consistent — but I'd say "preclinically selective, not human-confirmed." The combination with CJC-1295 has never been tested in a trial at all; it's an extrapolation from two single-agent GH-release curves.
Key sources
Tesamorelin
Egrifta · Egrifta SV · Egrifta WR
Modified analog of endogenous GHRHFDA-approved (2010)Evidence: Established (HIV belly fat) · Promising (cognition, liver fat)
What we know
This is the reference standard for the class — the only GHRH analog with a current FDA label — and it is worth reading closely because it shows what a real GH secretagogue does to a real population. In 404 HIV patients with excess abdominal fat, 2 mg daily for 26 weeks reduced visceral (deep belly) fat by 14–18% versus placebo; the effect held at 52 weeks and reversed quickly when the drug was stopped. IGF-1 rose by about 100 ng/mL: 47% of treated patients went above +2 standard deviations and 36% above +3 — meaning a third of people on an approved GHRH drug ran IGF-1 in the acromegaly-adjacent range. Five percent developed diabetes-range A1c versus 1% on placebo (hazard ratio 3.3), which is why the label requires glucose monitoring. Separately, a 20-week academic trial in 137 older adults (1 mg nightly) found a favorable effect on executive function in both healthy people and those with mild cognitive impairment, at the cost of higher fasting glucose — an interesting, unreplicated signal. Tesamorelin also reduced liver fat in HIV patients with fatty liver.
Doses in the research
2 mg under the skin daily (original trials); current formulations are 1.4 mg (SV) and 1.28 mg (WR, approved March 2025) daily. Cognition trial: 1 mg nightly.
Contraindications and cautions
These are on an FDA label, which makes them the firmest in this section — all Founded: active cancer (absolute; prior cancer must be inactive and treatment complete — the label's reasoning is simply that GH is a growth factor and IGF-1 rises); pituitary disruption (tumor, surgery, radiation, head trauma); pregnancy (animal teratogenicity); glucose intolerance (monitor A1c); fluid retention — swelling, joint pain, carpal tunnel in a meaningful fraction, resolving on stopping; hypersensitivity in 4%; and the label's own statements that it is not for weight loss and that long-term cardiovascular safety is not established after fifteen years on the market.
Status
FDA-approved for HIV-associated abdominal fat only. Any other use is off-label. WADA-prohibited at all times.
My takeReal drug, real effect, real costs. If the question is "does raising my own GH actually change anything measurable," tesamorelin is the proof that it does — and the proof of what it charges you in blood sugar and IGF-1. That is the trade you're evaluating with every agent in this section, and this is the one where the numbers are known.
Key sources
MK-677 (ibutamoren)
Ibutamoren · LUM-201 · sold as "Nutrobal"
Not a peptide — oral small moleculeFully syntheticEvidence: Established (raises GH/IGF-1; adds lean mass) · Negative (function, cognition)
What we know
The largest human dataset in this section — several thousand subjects in Merck's program — and the clearest picture of what a ghrelin-receptor agonist does over time. In a two-year randomized trial of 65 healthy adults aged 60–81 (25 mg daily), fat-free mass rose 1.1 kg versus a 0.5 kg loss on placebo; fat mass did not fall (limb fat actually rose); GH and IGF-1 returned to young-adult levels; and strength and function did not improve. Fasting glucose rose and insulin sensitivity fell. The accompanying editorial was titled "Not Yet Ready for Prime Time." A 563-patient Alzheimer's trial found no cognitive benefit despite a 73% IGF-1 rise. And a 123-patient trial in elderly hip-fracture patients was stopped early when congestive heart failure occurred in 4 of 62 on the drug versus 1 of 61 on placebo. That five-patient imbalance in frail elderly patients is the entire basis for FDA's "heart failure" designation — small numbers, biologically plausible (GH-driven fluid retention), unreplicated, and exactly the kind of signal that ends drug development. A pediatric program for growth hormone deficiency is still in Phase 3.
Doses in the research
25 mg by mouth daily — one of the few cases where the common online dose matches the studied one. Duration is where they diverge: two years under monitoring is the maximum studied.
Contraindications and cautions
Founded Heart failure or significant cardiac disease.
Founded Diabetes, prediabetes, insulin resistance. Documented in the pivotal trial — the best-documented harm of any agent in this section.
Founded Appetite and fat gain. Expect to be hungrier; the trial showed added fat, not just muscle.
Theoretical Active cancer — large, sustained IGF-1 elevation; no oncology data.
Theoretical Sleep apnea, edema-prone states, kidney disease (fluid retention); QT prolongation reported modestly in some studies; CYP3A4 interactions. Pharmacologically it opposes GLP-1 drugs on appetite, a combination people use anyway.
Status
Not approved anywhere. On FDA's safety-risk list for both pharmacy types, with the most specific rationale on the list ("potential for congestive heart failure"). Not a lawful dietary ingredient; FDA has sent warning letters to supplement sellers. WADA- and DoD-prohibited.
My takeMK-677 is honest in a way the injectables aren't: we know what it does, and it's a kilogram of lean mass you can't feel, a hungrier appetite, and worse blood sugar. For someone with metabolic risk — which is most of the people asking — the trade is backwards.
Key sources
GHRP-2, GHRP-6, and hexarelin
Pralmorelin · examorelin · the first-generation ghrelin agonists
Fully syntheticNot endogenousEvidence: Established (acute GH release) · Unknown (any clinical benefit)
What we know
All three release GH robustly in humans — acutely. GHRP-2 is approved in Japan as a one-time IV diagnostic test for GH deficiency, which is the only regulatory approval any GHRP has anywhere, and it's diagnostic, not therapeutic. The defining problem of the group is founded in human studies: unlike ipamorelin, they also raise ACTH, cortisol, and prolactin — in one study the cortisol response was comparable to giving a corticotropin-releasing-hormone test. GHRP-6 is the strongest appetite stimulant of the family. Hexarelin also binds a cardiac receptor (CD36) and improved heart function in rat models after heart attack — never tested in humans — and is the most prone to losing effect with continued dosing. No therapeutic outcome trial exists for any of the three.
Contraindications and cautions
Founded Cortisol and prolactin elevation with chronic use — which undermines the body-composition and sleep goals people take them for. Otherwise the class cautions apply (cancer theoretical, pregnancy, glucose).
Status
Not approved as therapy anywhere. GHRP-2 and GHRP-6 on FDA's safety-risk list for outsourcing pharmacies; hexarelin was never nominated. All WADA-prohibited.
My takeSuperseded. If the goal is GH release without cortisol, the field moved to ipamorelin for a reason; if the goal is a demonstrated clinical benefit, none of the three has one.
Key sources
Class-wide: the questions that apply to every GH secretagogue
Cancer. Here's the full picture rather than a slogan. Higher IGF-1 is associated with modestly higher risk of several cancers in large population studies — in the UK Biobank (394,000 people), about 8–11% higher risk of breast, prostate, and colorectal cancer per 5 nmol/L increment, within the normal range. People with acromegaly, who live for decades with pathologic GH excess, have about 45% higher overall cancer incidence. But 24,000 children given actual growth hormone for years showed no overall increase in cancer, and no relationship between dose and risk. So: the mechanism is real, the epidemiology is consistent, the magnitude is small, and no secretagogue has ever been studied with cancer as an endpoint. Active cancer is an absolute no — that's on an FDA label. A cancer history deserves an oncologist's input, not a shrug.
Blood sugar. The most reliable harm in the class, documented for MK-677 and tesamorelin. Anyone using these needs baseline and follow-up fasting glucose, insulin, and A1c.
Fluid retention. Swelling, joint aches, carpal tunnel — classic GH effects, documented on the tesamorelin label, dose-dependent and reversible.
Duration. Longest studied human exposures: tesamorelin 52 weeks, MK-677 two years, sermorelin a year in children, CJC-1295 seven weeks, ipamorelin seven days. Nobody knows what a decade looks like, and no agent has cardiovascular-outcome, cancer-surveillance, or mortality data. Stacked combinations have never been trialed at all.
Product. FDA's stated reason for restricting nearly all of these is immunogenicity from aggregated or impure peptide made outside pharmaceutical manufacturing. That's a concrete, near-term risk that has nothing to do with the endocrinology.
Athletes. Every agent in this section is prohibited by WADA at all times, prescription or not.
Where I land. The physiology is not fringe; these are real drugs with real effects. The right way to use one is as a defined therapeutic trial — specific goal, specific dose, IGF-1 and glucose at baseline and 3 months, a date to reassess, and a willingness to stop if the number you care about hasn't moved.
↑ Back to list
Semaglutide, tirzepatide, retatrutide, AOD-9604, 5-amino-1MQ.
Semaglutide
Ozempic · Wegovy · Rybelsus · oral Wegovy 25 mg (approved December 2025)
Modified analog of endogenous GLP-1FDA-approvedEvidence: Established
What we know
This is what a fully studied peptide looks like. In the pivotal obesity trial (1,961 people, 68 weeks, 2.4 mg weekly), average weight loss was 14.9% versus 2.4% on placebo; half of participants lost 15% or more, a third lost 20% or more. In 17,604 people with heart disease and obesity but no diabetes, it cut major cardiovascular events 20% and all-cause death 19% over about three years. Two-thirds of the weight comes back within a year of stopping. About 40% of the loss is lean mass, roughly what happens with any substantial weight loss. The oral 25 mg tablet approved in late 2025 produces about 13.6% loss at 64 weeks.
Doses in the research
Weekly injection titrated 0.25 → 0.5 → 1 → 1.7 → 2.4 mg over 16+ weeks. "Microdosing" is a community practice with no efficacy data.
Contraindications and cautions
FoundedPersonal or family history of medullary thyroid cancer or MEN2 — a label contraindication, but the evidence behind it is worth understanding: the tumors occurred in rats and mice, whose thyroid C-cells express far more GLP-1 receptor than ours. The human data are unresolved — a French database study suggested about 50% higher thyroid-cancer risk; a Scandinavian study of 145,000 patients found none. The label keeps the contraindication because the human question isn't settled; I'd honor it.
FoundedGallbladder disease (about 2.6% vs 1.2% in the trial), pancreatitis (rare, about 0.2 per 100 patient-years, but real), delayed gastric emptying / ileus (on the label since 2023; gastroparesis roughly 1% per year in a comparative study), diabetic retinopathy complications in people with diabetes, and aspiration under anesthesia — tell your anesthesiologist.
Founded in Europe, under review in the U.S. — NAION (a stroke of the optic nerve). European regulators concluded in 2025 it's a "very rare" side effect (about 1 extra case per 10,000 patient-years, roughly double the background risk); the U.S. label doesn't list it yet. Sudden painless vision loss in one eye means stop and get seen.
TheoreticalPregnancy — a label precaution based on animal data: stop at least two months before trying to conceive.
Not a concern per data: suicidality. FDA reviewed 91 trials and 2.2 million patients and removed the warning in January 2026. Heart disease is an indication, not a caution.
Status
FDA-approved. The compounding shortage exemption ended in early 2025; mass-market compounded semaglutide is no longer lawful, FDA has sent dozens of warning letters, and it has proposed permanently excluding it from outsourcing-pharmacy compounding. Generic semaglutide launched in Canada in 2026; U.S. patents run to about 2031. Not WADA-prohibited.
My takeEstablished, effective, and the best cardiovascular outcome data of any weight-loss drug in history. The real questions are individual: is it the highest-leverage step for you, what protects your
muscle while you lose, and what's the plan for the day you stop. Those are the conversation, not whether it works.
Key sources
Tirzepatide
Mounjaro · Zepbound
Synthetic dual GIP/GLP-1 agonistFDA-approvedEvidence: Established
What we know
In 2,539 people over 72 weeks, 15 mg weekly produced 20.9% weight loss versus 3.1% on placebo; 57% lost at least a fifth of their body weight. Head-to-head against semaglutide (751 people, 72 weeks), tirzepatide won: 20.2% vs 13.7%. In 13,299 people with diabetes and heart disease followed four years, it was non-inferior to dulaglutide on cardiovascular events (hazard ratio 0.92, superiority not quite reached) with lower all-cause mortality. Regain after stopping is substantial (about 14% of body weight back within a year).
Doses in the research
2.5 mg weekly for four weeks, then increase by 2.5 mg every four or more weeks to 5, 10, or 15 mg.
Contraindications and cautions
Same list as semaglutide: Founded MTC/MEN2 (rodent-derived, unresolved in humans), gallbladder (~2% vs 1.2%), pancreatitis (0.2%), gastric emptying, retinopathy in diabetes, anesthesia. One tirzepatide-specific item on the label: Founded oral contraceptives are absorbed less reliably — use a barrier method for four weeks after starting and after each dose increase. Hair loss in about 5%. The NAION signal is weaker and mixed for tirzepatide.
Status
FDA-approved. Compounding exemption ended February–March 2025; same enforcement picture as semaglutide. Not WADA-prohibited.
My takeCurrently the most effective approved weight-loss medication, with a cardiovascular record that's solid but not yet as striking as semaglutide's. Between the two, the choice is usually about tolerability, cost, and what your goals are — not about which one "works."
Key sources
Retatrutide
LY3437943 · "triple agonist" (GIP / GLP-1 / glucagon)
Fully syntheticNot approved anywhereEvidence: Probable (Phase 3 topline) · not yet peer-reviewed in full
What we know
In the Phase 2 trial (338 people, 48 weeks), 12 mg weekly produced 24.2% weight loss versus 2.1% on placebo, with no plateau, and an 82% reduction in liver fat. Phase 3 topline results released through 2026 report 28.3% at 80 weeks in 2,339 people without diabetes (and just over 30% at two years in the heaviest group), 20.8% in people with diabetes, and 26–29% in knee-osteoarthritis patients — the largest losses ever reported for a drug. Nausea affected roughly 30–40% at the higher doses (up to 42% at 12 mg); discontinuation for side effects ran up to 11–18% at the top doses. Two things are specific to this molecule: the glucagon component raises heart rate in a dose-dependent way (peaking around six months), and some patients report skin hypersensitivity. There is no cardiovascular outcome trial and no data beyond two years.
Doses in the research
4, 9, or 12 mg weekly (Phase 3). Every vial sold online is unverified product; FDA has warned state boards that compounded retatrutide has no legal pathway.
Contraindications and cautions
All extrapolated from the class — Theoretical until a label exists: MTC/MEN2, pancreatitis, gallbladder, pregnancy, and additionally tachyarrhythmia given the documented heart-rate rise. The Founded risk right now is the supply: a 2026 observational report on gray-market users found less weight loss than in trials and more cardiovascular effects, which is what you'd expect from unverified dosing.
Status
Not approved. Manufacturer plans to file in early 2027. Not lawfully compoundable. Not WADA-listed.
My takeAlmost certainly the next big approved drug, and the numbers are extraordinary. But "almost approved" and "sold as a research chemical" are very different things, and the heart-rate signal is exactly the kind of thing you want a label and a monitoring plan around. If someone wants triple-agonist-level results now, the honest answer is tirzepatide at full dose today, retatrutide when it's real.
Key sources
AOD-9604
hGH fragment 176-191 · "the fat-loss fragment"
Modified fragment of human growth hormoneThe one true GH fragmentEvidence: Negative (oral, weight loss) · Unknown (injected)
What we know
This is the molecule people are thinking of when they say "a piece of growth hormone": the last 16 amino acids of hGH, the region responsible for fat breakdown, with one substitution. It does not raise IGF-1 or stimulate growth. An Australian company ran six trials in nearly 900 people in the early 2000s. The 300-person 12-week study reported 2.8 kg loss on 1 mg daily by mouth versus 0.8 kg on placebo, without a dose-response. The 536-person Phase 2b that followed failed its primary endpoint — about 2 kg lost, not significantly different from placebo — and development stopped in 2007. Only the pooled safety paper was ever published; the efficacy data never were. Safety across those 900 people was clean: side effects indistinguishable from placebo, no antibodies, no effect on glucose tolerance. Note that all of that was oral or single IV dosing; there is no published human study of the injected form sold today.
Doses in the research
Oral 0.25–30 mg daily for up to 24 weeks; IV single doses up to 400 mcg/kg. The online injected regimen (250–500 mcg daily, fasting) has no trial behind it.
Contraindications and cautions
Nothing founded beyond the general absence of long-term or injected data. Theoretical cancer (argued against by the lack of IGF-1 effect), pregnancy (no data), injection immunogenicity (no antibodies seen with oral dosing). Founded: WADA prohibits it by name.
Status
Not approved as a drug. Its maker notified FDA of GRAS status as an oral food ingredient. On FDA's safety-risk list 2023–2024; nomination withdrawn, no compounding pathway. WADA-prohibited.
My takeThe safest-looking molecule in this section, and the one with the clearest negative trial. Nine hundred people, and it didn't beat placebo for weight. I'd rather spend your money on something that did.
Key sources
5-Amino-1MQ
5-amino-1-methylquinolinium · NNMT inhibitor
Not a peptide — synthetic small moleculeEvidence: Unknown (no human data)
What we know
A lab-designed molecule that blocks an enzyme (NNMT) in fat cells, which in theory raises cellular NAD+ and increases energy expenditure. In obese mice given 20 mg/kg three times daily for 11 days, weight fell 5–7% without eating less; a longer mouse study with diet produced normalization of weight. That is the entire evidence base. No human trial, no registered trial, no published toxicology, no human pharmacokinetics.
Doses in the research
Mouse only. The online 50–150 mg oral daily dose is not derived from anything.
Contraindications and cautions
All Theoretical, and some are non-trivial: NNMT is overexpressed in many cancers and is a drug target — but in some contexts it suppresses tumors, so the direction is unknown; the enzyme is the main disposal route for excess nicotinamide, so combining it with NAD+ precursors or high-dose B3 is uncharted; methylation and homocysteine effects are plausible; the liver expresses a lot of NNMT. Avoid in pregnancy and anyone under 18 on general principle.
Status
Not approved, never nominated for compounding, no legal pathway. Sold as a "research chemical." Not WADA-listed by name.
My takeInteresting biochemistry, zero human data, and no safety floor at all. This is the one where "the body knows how to handle it" doesn't apply — nothing about it is natural. Not something I'd put in a person.
Key sources
↑ Back to list
Sexual function, bonding, and pigmentation
PT-141, kisspeptin, oxytocin, melanotan I and II.
PT-141 (bremelanotide)
Vyleesi
Synthetic analog of α-MSHFDA-approved (2019)Evidence: Established (modest; premenopausal women) · Promising (men, older intranasal data)
What we know
Approved for low sexual desire in premenopausal women, based on two trials of about 1,250 women. The effect is statistically solid and clinically modest: desire scores rose 0.5–0.6 points versus 0.2 on placebo on a roughly five-point scale, and the number of satisfying sexual events did not improve over placebo. Tolerability is the real story — nausea in 40% (vs 1.3%), flushing 20%, and 18% quit the drug because of side effects. In men, the earlier intranasal program showed genuine erectile responses, including in men who didn't respond to sildenafil; the FDA delayed it in 2007 citing blood pressure as "its greatest safety concern," and the company re-engineered the drug as a lower-exposure injection. The compounded PT-141 nasal sprays sold today reproduce the exposure route FDA rejected.
Doses in the research
1.75 mg under the skin at least 45 minutes before activity, no more than one dose in 24 hours and no more than eight per month. Intranasal doses in the abandoned male program were 7–20 mg.
Contraindications and cautions
Founded Uncontrolled hypertension or known cardiovascular disease. Blood pressure rises about 6/3 mmHg for a few hours after each dose. Label contraindication.
Founded More than eight doses a month. Skin darkening (face, gums, breasts) in 1% at the approved frequency — but 38% after eight consecutive daily doses, worse in darker skin, and not always reversible.
Founded Oral naltrexone. Bremelanotide can substantially reduce naltrexone absorption — a real problem if you take it for alcohol or opioid use disorder. It also slows gastric emptying enough to affect other oral drugs.
Founded Pregnancy — stop if suspected. Severe kidney or liver disease — worse nausea.
Theoretical Melanoma history or many atypical moles — by analogy to melanotan II (same receptor family); no melanoma signal in the trials.
Status
FDA-approved (now owned by Cosette Pharmaceuticals; specialty-pharmacy only and hard to fill, but not discontinued). Not on FDA's safety-risk list; compounding an approved drug is separately restricted. Not WADA-listed.
My takeA real drug with a small effect and a lot of nausea. For women with low desire, it's one option among several and rarely the first. For men, the off-label case is plausible but the blood-pressure history is exactly why I'd want to know your numbers before your first dose, not after.
Key sources
Kisspeptin
Kisspeptin-10 · kisspeptin-54 · metastin
Endogenous hormone (both -10 and -54 forms)Evidence: Promising (acute IV infusion) · Unknown (how it's actually used)
What we know
Kisspeptin is the master switch above the reproductive axis — it's what tells your hypothalamus to release GnRH, which drives LH, FSH, and testosterone or estrogen. The human research is unusually good, because it comes from an academic group in London rather than a vendor. In placebo-controlled crossover trials, a 75-minute IV infusion in men and women with low sexual desire changed brain activity in sexual-processing regions and, in men, increased penile response by up to 56%. As a single injection to trigger egg maturation in IVF, it produced live births in 62% at the best dose with zero cases of moderate or severe ovarian hyperstimulation in 60 high-risk women. In men, even a 22-hour high-dose infusion raised testosterone only to the upper-normal range — it doesn't push you supraphysiologic. Across 200+ research subjects, no meaningful adverse effects; a 2025 pooled analysis of 95 people found no effect on anxiety, cortisol, or blood pressure. The catch: an 8-week study found that twice-daily dosing nearly abolished the response by day two, while twice-weekly dosing kept it. Frequent daily-type dosing — the online pattern — is the schedule closest to the one shown not to work.
Doses in the research
Kisspeptin-54: 1 nmol/kg/hour IV for 75 minutes (desire studies); 3.2–12.8 nmol/kg single injection (IVF); twice-weekly injections over 8 weeks. Kisspeptin-10: IV boluses up to 3 mcg/kg, peak effect at 1 mcg/kg. The online 100–500 mcg daily injection has no human study behind it.
Contraindications and cautions
Founded Frequent dosing causes tachyphylaxis — the axis stops listening. Not dangerous; just self-defeating.
Theoretical Hormone-sensitive cancers (prostate, breast). Interesting wrinkle: the gene was discovered as a metastasis suppressor; the concern is the downstream testosterone or estrogen, not the peptide.
Theoretical Pregnancy — levels are naturally very high in pregnancy; exogenous use untested. Men on testosterone therapy — it acts upstream, so it can't work through a suppressed axis.
Status
No approved product anywhere. Kisspeptin-10 remains on FDA's safety-risk list (not among the 2026 removals). Not named by WADA, but as an LH-releasing factor it very likely falls under the prohibited category for male athletes.
My takeOf everything on this page, this is the endogenous peptide whose human data I find most interesting — and whose real-world use is furthest from the evidence. Everything positive is a single IV infusion in a scanner; the frequent injections people are buying are the schedule the research says stops working within days. If low desire or low testosterone is the problem, there's a proper workup to do first.
Key sources
Oxytocin
Pitocin (injection) · compounded nasal spray and troches
Endogenous hormoneFDA-approved (obstetric injection only)Evidence: Established (labor) · Negative (autism, weight, sexual function) · Unknown (most else)
What we know
The hormone itself is fully natural and the synthetic version is identical. As an IV drug in labor it's essential and well understood. Everything else is a long list of well-run trials that came up empty: 290 children with autism, 24 weeks, up to 80 IU daily — no difference from placebo (NEJM 2021); 61 adults with obesity, 24 IU four times daily for 8 weeks — no weight loss; multiple trials in female sexual dysfunction — no advantage over placebo; social anxiety as an add-on to therapy — no clinical improvement. A migraine program reached Phase 2 and was shelved. The recurring problem in the whole field is that we don't actually know how much intranasal oxytocin reaches the brain, and effects shrink as trials get bigger.
Doses in the research
Intranasal 24 IU is the near-universal single dose; chronic regimens 24 IU two to four times daily; the autism trial titrated 8–80 IU/day.
Contraindications and cautions
Founded Pregnancy. It contracts the uterus. Outside supervised labor, this is the most important line on this entry, and it applies to anyone who could be pregnant.
Founded Hyponatremia. Oxytocin has weak antidiuretic activity; high IV doses with free water have caused seizures and coma. Intranasal doses are far lower, but this is why "more is better" is a bad instinct with this molecule — and why I'd be careful in anyone on thiazides, SSRIs, or carbamazepine.
Theoretical Certain psychiatric contexts — some data suggest it can increase in-group bias or negative social memory in some people. Mechanism-based, not established harm.
Status
Injection FDA-approved. No approved nasal product in the U.S.; compounded nasal forms sit on unusually solid legal ground (USP monograph, component of an approved drug, not on any safety-risk list) even though the formulations themselves are unstudied. Not WADA-prohibited.
My takeSafe, natural, legally available — and repeatedly shown not to do the things it's sold for. Low-risk enough that I don't object to a defined trial for a specific goal, but I'd want you to know going in that the best studies are negative, and to set a date to stop if nothing changes.
Key sources
Melanotan I and Melanotan II
Melanotan I = afamelanotide (Scenesse) · Melanotan II = MT-2, "the Barbie drug"
Synthetic analogs of α-MSHMT-I: FDA-approved implant for EPPMT-II: not approved anywhereEvidence: MT-I Established (narrow) · MT-II Unknown, with founded harms
What we know
These are two different drugs that share a marketing name. Afamelanotide selectively hits the pigment receptor (MC1R) and is an FDA-approved implant (16 mg every two months) for erythropoietic protoporphyria, where it increased pain-free sun exposure in two Phase 3 trials (64 vs 41 hours in the U.S. study; 6 vs under 1 hour in the European study, which counted exposure differently). Its side-effect profile is mild — implant-site reactions, nausea — and its label requires twice-yearly full-body skin exams because it darkens existing moles. A Phase 3 in vitiligo is underway. Melanotan II hits all the melanocortin receptors at once, which is why tanning comes packaged with erections, nausea, appetite loss, and blood-pressure effects — it's the same drug hitting different receptors. Its human efficacy database is about 20–40 men from the 1990s (erections in 17 of 20 at 0.025 mg/kg; severe nausea in 13%). Everything since is case reports of harm: melanoma and melanoma-in-situ in users, eruptive new dysplastic moles, rhabdomyolysis with kidney injury (CPK near 18,000), renal infarction after six months of use, priapism, and — in 2025 — oral mucosal melanoma in a 22-year-old after using the nasal spray. Australian regulators tested products and found doses inconsistent with labels. Interestingly, PT-141 is what your body turns melanotan II into.
Doses in the research
Afamelanotide: 16 mg implant every two months. Melanotan II: 0.025 mg/kg (about 2 mg) in the 1990s studies; online "loading" protocols of 250–500 mcg daily then weekly maintenance are not from research, and product content is unreliable.
Contraindications and cautions (melanotan II)
Founded Melanoma history, dysplastic nevus syndrome, many atypical moles. Multiple case reports, plus the practical problem that it darkens and changes moles, which is exactly what dermatologists watch for.
Founded Kidney disease (rhabdomyolysis, renal infarction reported) and cardiovascular disease or hypertension (the receptor raises blood pressure; FDA rejected the high-exposure version of its metabolite over exactly this).
Theoretical Pregnancy; eating disorders (appetite suppression); priapism risk with ED drugs or sickle cell.
Status
Afamelanotide: FDA-approved for EPP only. Melanotan II: not approved anywhere; on FDA's safety-risk list 2023–April 2026, removed, FDA review scheduled before February 2027; Australian and UK regulators actively warn against it. Neither WADA-listed.
My takeThere is no scenario where melanotan II is the right answer: for photoprotection there's an approved selective drug, and for desire there's an approved dosed one. It's the single agent on this page with the clearest pattern of serious harm in ordinary users, and I'd tell anyone using it to stop and get a skin exam.
Key sources
↑ Back to list
Mitochondria, aging, and cellular energy
MOTS-c, SS-31, epithalon, NAD+ and its precursors.
MOTS-c
Mitochondrial-derived peptide · "exercise mimetic"
Endogenous (encoded in mitochondrial DNA)Evidence: Plausible (animal) · Negative-leaning (only human analog trial)
What we know
A 16-amino-acid peptide your own mitochondria encode, discovered in 2015. It activates AMPK — the same energy sensor metformin and exercise engage — and in mice it improves insulin sensitivity, doubles running distance in old animals, and protects bone. In ten young men, a single bout of exercise raised muscle MOTS-c nearly 12-fold, which proves it's exercise-responsive in humans; it does not prove injecting it produces exercise benefits. The only human trial used a modified analog (CB4211): in 20 people with obesity and fatty liver, 25 mg daily for four weeks lowered liver enzymes and fasting glucose modestly but did not reduce liver fat (the primary endpoint) or weight beyond placebo. The company dropped the program. One practical note: MOTS-c is reported to degrade 85–90% within a few hours at room temperature once reconstituted, so an unrefrigerated gray-market vial may not contain what it says.
Doses in the research
Human (analog only): 25 mg daily for 28 days. Mouse: 0.5–15 mg/kg intraperitoneally. The online 5–10 mg/week regimen isn't from a human study.
Contraindications and cautions
Theoretical Hypoglycemia with insulin or sulfonylureas — it lowers glucose in animals and modestly in the human analog trial. Theoretical Pregnancy (no data). Active cancer: AMPK activation is generally growth-suppressive, so the mechanism points the reassuring direction, but there's no data. Founded for the analog: injection-site reactions in more than 10%.
Status
Not approved anywhere. Safety-risk list 2023–April 2026; advisory committee voted in July 2026 to recommend allowing compounding; rulemaking pending. Not WADA-named but falls under the non-approved-substance catch-all.
My takeElegant biology, truly your own molecule, and the one human trial of a close cousin was a miss. The thing that reliably raises your MOTS-c twelve-fold is exercise, and it's free.
Key sources
SS-31 (elamipretide)
Forzinity · Bendavia · MTP-131
Fully synthetic tetrapeptideFDA-approved (September 2025, Barth syndrome)Evidence: Established (one ultra-rare disease) · Negative (everything else tested)
What we know
Unlike most gray-market peptides, "SS-31" is the same molecule as an approved drug, so we have a large human dataset — and most of it is negative outside one disease. Elamipretide binds cardiolipin in the inner mitochondrial membrane and stabilizes the machinery that makes ATP. In Barth syndrome (about 150 people in the U.S.), a 12-person crossover trial flatly failed its blinded endpoints, but the open-label extension showed sustained gains in walking distance and a 42% increase in knee strength, and after a refusal-to-file, a rejection, and a third submission, FDA granted accelerated approval in September 2025 on that strength endpoint. In 218 people with primary mitochondrial myopathy, a properly powered Phase 3 missed both primary endpoints. In dry macular degeneration, a 176-person trial missed its primary endpoint. Earlier heart-attack and heart-failure programs didn't pan out. There is no human trial for fatigue, "mitochondrial optimization," performance, or aging.
Doses in the research
40 mg under the skin daily in every pivotal trial (4 weeks to 168 weeks); 20 mg in severe kidney impairment. Online "SS-31" protocols of 5–10 mg/day are below the studied dose and not derived from any trial.
Contraindications and cautions
Founded Injection-site reactions in essentially everyone (redness 100%, pain 75%, itching 67%); serious allergic reactions, sometimes months into treatment (labeled warning); eosinophilia peaking around day 90; dose reduction in severe kidney disease; contains benzyl alcohol (not for neonates). Theoretical Pregnancy (animal studies reassuring, no human data); cancer (no signal, no plausible pro-tumor mechanism).
Status
FDA-approved for Barth syndrome (patients ≥30 kg) only. As an approved drug, compounding copies is restricted. Not WADA-listed.
My takeA real drug that earned a narrow approval by the skin of its teeth and failed the two big trials that would have made it broadly useful. For someone chasing "mitochondrial health," the studied dose is four to eight times what's sold online, nearly everyone gets a reaction at the site, and there's no evidence it helps a person who doesn't have a cardiolipin defect.
Key sources
Epithalon (epitalon)
Ala-Glu-Asp-Gly · synthetic stand-in for epithalamin (a bovine pineal extract)
Fully synthetic tetrapeptideNot endogenousEvidence: Unknown (unreplicated, single-group)
What we know
Designed in St. Petersburg by Vladimir Khavinson as a short-peptide representative of a crude cow pineal extract. Almost every "clinical" claim — including the widely quoted 20% versus 82% six-year mortality figure — comes from Khavinson's own group, in Russian, for the extract rather than the tetrapeptide, in combination with a second extract, in a treated arm of 20 elderly women. No independent group has replicated any of it in over twenty years. The telomerase claim comes from a 2003 cell-culture study in fetal fibroblasts that reports activation without usable numbers; no human has ever had telomere length measured before and after epithalon. In mice given it lifelong, average lifespan didn't change, maximum lifespan rose about 12%, total tumor incidence didn't change, and leukemia fell. There is no Phase 1 safety study, no pharmacokinetics, no adverse-event rates in any usable form.
Doses in the research
Human cohorts used the extract: 10 mg intramuscular daily for 10 days, or 5 doses every 2–3 days, repeated once or twice a year. Online regimens of 5–10 mg daily for 10–20 days mimic that structure but have no trial of the synthetic peptide behind them.
Contraindications and cautions
Theoretical Cancer — and the theory cuts both ways. Anything that activates telomerase could in principle help tumor cells immortalize; the available mouse data show no increase in tumors and a decrease in leukemia — but from the same group that invented the compound. Unresolved; avoid in active malignancy. Theoretical Pregnancy (no data). Founded as a category: no pharmacopeial standard, so no way to verify what's in the vial.
Status
Not approved as a drug anywhere. Safety-risk list 2023–April 2026; advisory committee narrowly voted in July 2026 to recommend allowing compounding (for insomnia); rulemaking pending. Falls under WADA's non-approved-substance catch-all.
My takeI try not to dismiss things because they're Russian or old. But this is one research group, one crude extract, extraordinary numbers, and no one else has ever checked. That's not a reason to say it can't work; it's a reason to say we don't know, and to price it accordingly.
Key sources
NAD+, NMN, and NR
IV or injected NAD+ · nicotinamide mononucleotide · nicotinamide riboside (Niagen)
NAD+: endogenous coenzyme in every cellNot a peptideEvidence: Oral precursors Probable (biomarkers) / Promising (function) · IV NAD+ Unknown
What we know
NAD+ is not a peptide; it's on this page because it's injected in the same clinics. Three products get conflated and the evidence runs opposite to the marketing: the oral precursors have real trials, the IV infusion has almost none. NR at 1,000 mg/day raised blood NAD+ about 60% in older adults over six weeks (blood pressure trended down but wasn't significant); 3,000 mg/day for a month in Parkinson's patients caused only mild side effects; a 30-person Parkinson's study showed brain NAD+ rising with mild clinical improvement in responders; and a 400-person, one-year Phase 3 in Parkinson's finished in 2025 with results pending as of this writing — the most important outstanding data point in this field. NMN at 250 mg/day for 10 weeks improved muscle insulin sensitivity in prediabetic postmenopausal women (Science, 2021) without changing weight, A1c, or anything else clinical; 300–900 mg/day for 60 days improved six-minute walk distance in middle-aged adults. IV NAD+: the one pharmacokinetic study (about 750 mg over six hours) showed nothing detectable in blood for two hours and then a rise in urinary methylated breakdown product — much of the infused dose is being methylated and excreted. The only tolerability study (2026) found all six of six NAD+ recipients reported moderate-to-severe cramping, nausea, and chest pressure and had to slow their own drip to an average of about 97 minutes per 500 mg (versus 37 minutes for IV NR). The 10-day "brain restoration" addiction protocols have no controlled trial support. Injected NAD+ under the skin has no published human trial at all.
Doses in the research
NR: 1,000 mg/day oral (up to 3,000 mg/day for 4 weeks). NMN: 250–900 mg/day oral. IV NAD+: 500–750 mg per infusion, run slowly. Subcutaneous NAD+: no studied dose.
Contraindications and cautions
Theoretical Active cancer — the most substantive concern here. Tumors need NAD+; the enzyme that makes it is an active cancer drug target, and drugs are being developed to deplete tumor NAD+. In mice, NR uptake by triple-negative breast cancer cells was associated with more tumors and more brain metastases (2022 — often misattributed online to a 2024 Nature paper). No human data either way. I'd avoid NAD+ precursors in active malignancy, and I'd coordinate with oncology for anyone in treatment, since NAD+ also fuels the DNA repair that chemotherapy is trying to overwhelm.
Founded IV infusion reactions — rate-dependent, reproducible, self-limited. "Chest tightness is normal" is a real reaction, not a therapeutic effect; in someone with coronary disease, a drug that reliably causes chest pressure is a bad idea on principle.
Theoretical Methylation drain. Excess nicotinamide is cleared using your methyl donor (SAM); the IV study confirmed the pathway is active. No trial has shown clinical depletion; adequate folate/B12 is sensible, alarm is not.
Not a concern per data: diabetes (favorable), cardiovascular disease for oral forms (neutral to favorable).
Status
NAD+ is not an approved drug; it's in FDA's "under evaluation" category (Category 1), not the safety-risk list, so compounded injectable NAD+ is in a better legal position than most of this page. NR is a lawful supplement ingredient. NMN's supplement status was restored by FDA in September 2025. None WADA-prohibited.
My takeWe
offer NAD+ infusions in this office, so I have an interest here, and I'll say the same thing to you I'd say to myself: the doses with trial support are oral, the IV route has one PK study and one 14-person tolerability report, and the reproducible side effect is exactly why it has to be run slowly. If you want to raise NAD+, the evidence-backed path is 500–1,000 mg of NR a day. If you want an infusion, I'll tell you what it can and can't be expected to do before you sit down.
Key sources
↑ Back to list
Immune and brain
Thymosin alpha-1, selank, semax, DSIP, cerebrolysin.
Thymosin alpha-1
Thymalfasin · Zadaxin
Endogenous 28-amino-acid peptideApproved in 30+ countriesEvidence: Established (hepatitis B, historically) · Negative (sepsis mortality) · Probable (immune markers)
What we know
A naturally occurring thymic peptide, identical in its synthetic form, approved in China, Italy, and most of Asia and Latin America for chronic hepatitis B and C and as a chemotherapy and vaccine adjunct — more than twenty years of marketing with no chronic-toxicity signal. Its hepatitis B data are real but dated (delayed virologic response roughly tripled at follow-up in older meta-analyses) and have been superseded by antivirals. The most important recent trial is TESTS (BMJ, 2025): 1,106 Chinese adults with sepsis, 1.6 mg twice daily for a week — 28-day mortality 23.4% versus 24.1%, no difference. A 2025 meta-analysis found benefit only when low-quality trials were included. In COVID, retrospective studies looked promising and randomized data did not confirm a mortality benefit. In melanoma, a 488-patient randomized Phase 2 raised response rates alongside chemotherapy without a significant survival difference, and was never confirmed in Phase 3. It reliably improves vaccine seroconversion in dialysis and elderly patients. What it has never been studied for is "immune support" in healthy people.
Doses in the research
1.6 mg under the skin twice weekly for 6–12 months (hepatitis); 1.6 mg twice daily for 7 days (sepsis); 1.6 mg daily for 5–7 days (COVID); up to 6.4 mg four days a week (melanoma). Range across trials 1–16 mg/day with no dose-limiting toxicity. Online "immune" regimens of 0.75–1.6 mg two or three times a week are within the studied range but for an unstudied purpose.
Contraindications and cautions
Founded Deliberate immunosuppression — transplant and stem-cell recipients (label caution, supported by case reports of graft failure and immune hemolysis). Founded Hypersensitivity. Theoretical Autoimmune disease — a foreign-label caution with no controlled flare data. Theoretical Pregnancy. Not a cancer contraindication — cancer adjunct is a labeled use. Side effects in trials: injection-site discomfort, essentially nothing else above placebo.
Status
Not FDA-approved (orphan designations only). Safety-risk list from 2023; FDA reviewers and advisory committee recommended against compounding in December 2024; the nomination has since been withdrawn with no active review pathway. Legal status for U.S. compounding is contested and unsettled. Not WADA-named.
My takeThe strongest safety record of any injectable on this page, an approved drug in much of the world, and a large trial that just showed it doesn't save lives in sepsis. That combination — very safe, modest and situational benefit — makes it a reasonable adjunct in specific settings (a poor vaccine responder, an immunocompromised patient with a defined goal) and a weak buy as general "immune support," where nobody has shown it does anything.
Key sources
Selank
TP-7 · Russian anxiolytic nasal spray
Synthetic analog of tuftsin (an endogenous fragment)Approved in RussiaEvidence: Promising-to-Unknown (Russian trials only)
What we know
A natural immune fragment (tuftsin) with a synthetic tail, registered in Russia in 2009 as a 0.15% nasal solution for generalized anxiety. The main trial (62 patients, about two weeks) found it "clinically comparable" to the benzodiazepine medazepam with an additional anti-fatigue effect — without a placebo arm, without effect sizes in the abstract, in Russian, from the developer's institute, with blinding not described. Rodent work shows GABA modulation and enkephalinase inhibition. No Western trial, no placebo-controlled trial in English, no data on the injected form sold online, no data beyond 30 days. Reported side effects: nasal irritation, mild headache; no sedation, dependence, or withdrawal described.
Doses in the research
0.15% solution, 2–3 drops per nostril, 2–3 times daily (roughly 600–900 mcg/day) for 14–30 days. The online 200–500 mcg nasal or injected dose isn't from the trials; injection has no human data.
Contraindications and cautions
Russian label lists pregnancy, breastfeeding, and under 18 — all Theoretical. Additive effect with benzodiazepines is a rat finding. Nothing founded beyond the general unregulated-product concern.
Status
Not approved outside Russia. On FDA's safety-risk list from 2023; nomination since withdrawn with no review pathway; status contested. Not WADA-named.
My takePlausible, apparently gentle, and essentially unevaluated by anyone without a stake in it. For anxiety we have options with real trials; if someone wants to try selank as a low-risk nasal adjunct, I'd frame it as an experiment with a stop date, not a treatment.
Key sources
Semax
ACTH(4-7)-Pro-Gly-Pro · Russian nootropic and stroke drug
Synthetic analog of an ACTH fragmentApproved in RussiaEvidence: Unknown (open-label Russian data only)
What we know
A piece of ACTH with a synthetic tail that has no adrenal (cortisol) activity; approved in Russia since 1994 as nasal drops for stroke (1%) and cognitive complaints (0.1%), and on Russia's essential-drugs list. FDA's 2026 review could retrieve only two human studies in full. The stroke data are open-label and non-randomized (110 patients in the largest, no placebo, no effect sizes reported); a migraine/neuralgia study was a single open dose in 37 people with mixed results; a 24-person fMRI study in healthy volunteers showed brain-network changes with no cognitive outcome. There is no efficacy data for the way it's used in the West — as a nootropic in healthy adults, or for ADHD, depression, or anxiety. Rodent work shows BDNF induction and neuroprotection, and also potentiation of amphetamine-induced dopamine release, which FDA flagged.
Doses in the research
0.1%: 2–3 drops per nostril 2–3 times daily (200–2,000 mcg/day). 1%: 6–18 mg/day for 10 days in acute stroke. Online 200–600 mcg/day nasal; injection has no human data.
Contraindications and cautions
Founded (partially) Diabetes — elevated glucose in about 7% of diabetic patients in Russian data, low quality. Theoretical Bleeding disorders and anticoagulants (animal anticoagulant effect); acute psychosis, agitated anxiety, pregnancy (Russian label); substance-use history (dopaminergic potentiation); cardiovascular disease.
Status
Not approved outside Russia. Safety-risk list 2023–April 2026; advisory committee narrowly voted in July 2026 to recommend allowing compounding (for stroke, migraine, neuralgia), over FDA staff objection; rulemaking pending. Not WADA-named.
My takeA thirty-year Russian track record for a drug whose trials we can't evaluate. That's not nothing, and it's not evidence. For cognition, the honest ranking still starts with sleep, exercise, blood pressure, hearing, and mood — the things with outcome data — and semax would be a curiosity at the end of the list.
Key sources
DSIP
Delta sleep-inducing peptide · emideltide
Uncertain endogenous status (no human gene identified)Evidence: Negative (placebo-controlled insomnia trials)
What we know
Isolated from rabbit blood in 1977. Something DSIP-like is detectable in human fluids, but no gene or precursor has ever been found, so whether it's actually a human hormone is unresolved. The human trials are all from the 1980s–90s, all IV, all tiny. The placebo-controlled ones — crossover trials of 6, 6, and 16 insomnia patients — showed no difference from placebo. The uncontrolled ones (14–18 patients) showed better sleep onset and less wake time. An open study in 107 alcohol and opiate withdrawal patients reported reduced symptoms that relapsed within days; a 7-patient pain study was uncontrolled. Nothing published since 1998, nothing longer than seven nights, nothing by injection under the skin or by nose.
Doses in the research
25–35 nmol/kg IV (about 1.5–2 mg for an adult) nightly for 4–7 nights. The online 100–300 mcg subcutaneous bedtime dose has no human data.
Contraindications and cautions
Nothing founded. Theoretical Additive sedation with hypnotics, opioids, alcohol; cardiovascular disease (rodent hypotension with IV); pregnancy. Reported effects: transient headache, injection-site pain, occasional paradoxical sleep disruption.
Status
Not approved anywhere. Safety-risk list 2023–April 2026; the only peptide the July 2026 advisory committee voted against (7 to 6, one abstention), citing thirty-year-old low-quality evidence. Not compoundable. Not WADA-named.
My takeThe controlled trials were negative, the uncontrolled ones were positive, and the whole literature would fit in a folder. For sleep,
CBT-I has better data than anything injectable, and I'd start there.
Key sources
Cerebrolysin
Porcine brain peptide mixture · FPF-1070
Not a single peptide — pig-brain extractApproved in ~45 countriesEvidence: Probable (stroke rehab, small effect) · Negative (stroke mortality) · Unknown (dementia; very low certainty)
What we know
An enzymatically digested extract of pig brain — mostly free amino acids, about a fifth short peptides — made in Austria and approved across Eastern Europe, Asia, and Latin America for stroke, brain injury, and dementia. Not approved in the U.S., EU-wide, UK, Germany, or Canada. The best evidence is a Cochrane review of seven acute-stroke trials (1,773 patients): no effect on death, and a roughly 2.4-fold increase in non-fatal serious adverse events (moderate certainty). In stroke rehabilitation, an industry-sponsored 208-patient trial (30 mL IV daily for 21 days) showed clearly better arm function at 90 days; the 240-patient replication in Russia was null; pooled, the effect is small-to-moderate. Brain-injury trials show a similar small effect. For vascular dementia, six trials of 597 patients suggest a cognitive benefit that Cochrane rated "very low certainty" and "may be too small to be clinically meaningful." Four preclinical papers co-authored with the manufacturer were retracted in 2025–26 for image manipulation; no clinical trial has been.
Doses in the research
Stroke: 30 mL IV daily for 10–21 days. Brain injury: 50 mL/day for 10 days, then 10 mL/day. Dementia: 10–30 mL IV five days a week for four weeks in cycles. The online 5–10 mL a few times a week for "cognition" in healthy people has no data.
Contraindications and cautions
Founded Allergy to pork protein; infusion-rate effects (flushing, palpitations with rapid infusion); more non-fatal serious adverse events in acute stroke; fluid load in heart or kidney failure. Theoretical Epilepsy and severe renal failure (labeled, no excess events in trials); pregnancy; prion transmission (manufacturer-validated clearance, no case reported, no independent assessment); MAOI/antidepressant additive effects (manufacturer advises dose reduction).
Status
Not FDA-approved (orphan designation for frontotemporal dementia only). Animal-derived, never nominated for compounding; the only way it reaches U.S. patients is importation, which is not lawful for domestic use. Not WADA-listed.
My takeA legitimate drug with decades of use abroad, a real but small effect on recovery after stroke, a safety signal in the acute phase, and essentially nothing for the healthy person hoping to think faster. It's not a fraud; it's a modest neurorehabilitation adjunct being sold as a nootropic.
Key sources
↑ Back to list
Quick reference
Everything above, on one screen. "Longest human study" is the longest published controlled or labeled exposure I could find.
| Agent | Natural to the body? | Best human evidence | Longest human study | U.S. status (Sept 2026) | WADA |
|---|
| BPC-157 | No (synthetic) | 1 negative RCT (n=53); uncontrolled case series | 2 weeks | Advisory vote for compounding; not yet legal | Prohibited |
| TB-500 / Tβ4 | Tβ4 yes; fragment no | Tβ4 eye-drop Phase 3s (mixed); fragment: none | Tβ4: months; TB-500: none | Advisory vote for compounding; not yet legal | Prohibited |
| GHK-Cu | Yes | Topical wound & cosmetic RCTs; injected: none | 12 weeks (topical) | Topical legal; injectable under FDA review | Not listed |
| KPV | Fragment of α-MSH | None (mouse colitis only) | None | Advisory vote for compounding; not yet legal | Not listed |
| LL-37 | Yes | Topical ulcer Phase 2b (missed overall) | 13 weeks (topical) | Under FDA review | Not listed |
| Sermorelin | Yes (GHRH fragment) | Former FDA label (children); 16-wk adult biomarker study | ~1 year (children) | Compounded; legally gray | Prohibited |
| CJC-1295 | No (modified analog) | 1 PK/PD study; halted unpublished Phase 2 | 7 weeks | No compounding pathway | Prohibited |
| Ipamorelin | No (synthetic) | 1 negative RCT (n=114, IV) | 7 days | On FDA safety-risk list | Prohibited |
| Tesamorelin | No (modified analog) | FDA-approved; Phase 3 n=404 | 52 weeks + 15 yrs marketed | Approved (HIV lipodystrophy) | Prohibited |
| MK-677 | No (not a peptide) | 2-yr RCT; 563-pt AD trial (null); CHF signal | 2 years | On FDA safety-risk list | Prohibited |
| GHRP-2/6, hexarelin | No (synthetic) | Acute GH-release studies; GHRP-2 diagnostic in Japan | Weeks | Safety-risk list (GHRP-2/6) | Prohibited |
| Semaglutide | No (modified analog) | FDA-approved; CV outcomes n=17,604 | 4 years | Approved; compounding restricted | Not listed |
| Tirzepatide | No (synthetic) | FDA-approved; CV outcomes n=13,299 | 4 years | Approved; compounding restricted | Not listed |
| Retatrutide | No (synthetic) | Phase 3 topline (n≈2,300) | 2 years | Not approved; filing 2027; no compounding | Not listed |
| AOD-9604 | Modified hGH fragment | 6 trials, ~900 pts; Phase 2b failed | 24 weeks | No compounding pathway; GRAS as food | Prohibited |
| 5-Amino-1MQ | No (not a peptide) | None (mice only) | None | Research chemical only | Not listed |
| PT-141 | No (α-MSH analog) | FDA-approved; Phase 3 n≈1,250 | 52 weeks | Approved (female HSDD) | Not listed |
| Kisspeptin | Yes | Placebo-controlled IV infusion RCTs; IVF trigger | 8 weeks | On FDA safety-risk list | Likely prohibited (men) |
| Oxytocin | Yes | Approved (labor); large null RCTs for other uses | 24 weeks (nasal) | Injection approved; nasal compounded | Not listed |
| Melanotan I / II | No (α-MSH analogs) | MT-I approved (EPP); MT-II: 1990s n≈20 + harm reports | MT-I years; MT-II none | MT-I approved; MT-II under FDA review | Not listed |
| MOTS-c | Yes | Analog Phase 1b (n=20, primary endpoint missed) | 4 weeks (analog) | Advisory vote for compounding; not yet legal | Non-approved catch-all |
| SS-31 | No (synthetic) | FDA-approved (Barth); 2 large negative Phase 3s | 168 weeks | Approved (Barth syndrome) | Not listed |
| Epithalon | No (synthetic) | Single-group Russian cohorts of the parent extract | Unverifiable | Advisory vote for compounding; not yet legal | Non-approved catch-all |
| NAD+ / NMN / NR | Yes | Oral precursor RCTs; IV: 1 PK + 1 tolerability study | 52 weeks (NR, pending) | NAD+ "under evaluation"; NR/NMN supplements | Not listed |
| Thymosin α1 | Yes | Approved abroad; sepsis RCT n=1,106 (null) | 12 months + 20 yrs marketed abroad | Not approved; compounding contested | Not listed |
| Selank | Analog of tuftsin | Russian active-comparator trial (n=62) | 30 days | Not approved; compounding contested | Not listed |
| Semax | Analog of ACTH fragment | Russian open-label studies | Weeks | Advisory vote for compounding; not yet legal | Not listed |
| DSIP | Unresolved | Tiny placebo-controlled trials (null) | 7 nights | Advisory committee voted against | Not listed |
| Cerebrolysin | No (pig-brain extract) | Cochrane reviews; rehab RCTs | Months | Not approved; no lawful U.S. route | Not listed |
What I'd suggest next
If you've read this far, you've probably noticed the pattern. The agents with the strongest evidence are the ones that went through the full drug-development process, and they carry real, well-characterized costs. The agents with the most appealing stories — natural, gentle, regenerative — mostly haven't been tested in people at all, and the risk that actually shows up in the adverse-event reports is the unregulated vial, not the molecule.
None of that means "don't." It means the right question isn't "is this peptide good?" — it's "what is the most effective, defensible, affordable, and patient-aligned next step for the problem I actually have?" Sometimes a peptide is part of that answer. Often the higher-leverage move is something less exciting: finding the mechanical driver behind the tendon that won't heal, fixing the sleep or thyroid or insulin problem underneath the fatigue, or using an approved drug at a known dose with a monitoring plan instead of a research chemical at a guessed one.
That's the conversation I'd like to have with you. Bring what you've read, what you've tried, and what you're hoping to change. I'll tell you what I think, rank the options honestly, and if a peptide belongs in the plan we'll set it up properly — defined dose, defined duration, the labs that matter, and a date to decide whether it's working.
Let's figure out the right next step
A 30-second request is all it takes to start. No decisions to make yet — just tell me you'd like to talk.
Request an Appointment
This guide is general education from William D. Starsiak, DO, and reflects my reading of the published research and regulatory record as of September 2026; it is not a treatment plan for you specifically and does not create a physician-patient relationship. Doses cited are the doses used in the studies described, reported so you can see what was actually tested — not a recommendation to use them. Do not start, stop, or change any prescribed medication based on this page. Most agents here are not FDA-approved, and regulatory status is changing quickly; I revise this page as it does. I have no financial relationship with any peptide manufacturer or vendor. This office does offer NAD+ infusions, which I've noted in that section.