Independent peptide research
What each compound is, what human research actually shows, and what is still only a claim — written plainly, graded honestly, updated as the studies land — with every grade traced to the trials behind it.
Pick one — the answer changes what we show you first.
What LYO does
One row per claimed effect, per compound. A grade from A to D based only on human evidence, an effect size, and the studies behind it.
See the evidence matrix →Six short lessons with a check at the end of each. By the third you can look at any peptide claim and know what it is worth.
Start the lessons →New trials and reviews, read in full and reduced to what changed. When a study moves a grade, the change is logged and dated.
This week's feed →Learn
Most bad decisions in this space come from not being able to read a study. A few minutes each, with a check at the end of every one. Progress is saved on this device.
Every protein in your body is a chain of amino acids folded into a shape. A peptide is the same thing, only short — the usual cut-off is about fifty amino acids, though the line is a convention rather than a law of nature. Insulin is 51 and gets called both.
The body uses peptides as messengers. GLP-1 is released by the gut after a meal and tells the pancreas and the brain what to do about it. Oxytocin, growth-hormone-releasing hormone, the endorphins — all peptides, all made in one place to carry a signal to another.
What separates a peptide from a protein?
Reference
One page per compound. A direct answer in the first paragraph, an evidence matrix, the regulatory picture, and every claim traced to a study you can open.
Synthetic pentadecapeptide from a gastric protein. Striking animal results; no controlled human trials.
GLP-1 receptor agonist. Among the most-studied molecules of the decade; approved for diabetes and obesity.
Triple GIP / GLP-1 / glucagon agonist. Two phase 3 obesity trials and a phase 3 diabetes trial are published; not approved anywhere.
Dual GIP / GLP-1 agonist. Approved; beat semaglutide in a direct head-to-head trial.
GHRH analogue. Approved for one narrow indication; the general-population claims are extrapolation.
Selective GH secretagogue. One real randomised trial in humans — and it was negative.
Copper-binding tripeptide. Modest human evidence for topical skin use; injected use has none.
Synthetic tetrapeptide from a Russian gerontology institute. The human studies mostly tested a different product.
Fragment of thymosin beta-4. Human trials exist only as an eye drop; the injectable use has none.
Long-acting amylin analogue. Strong phase 3 data in combination with semaglutide; not yet approved.
Russian-developed heptapeptide, registered in Russia, unstudied elsewhere.
Russian anxiolytic peptide, registered in Russia. Same evidence problem as Semax.
An approved medicine — for one indication, in one population, with a 40% nausea rate.
Was an approved medicine for diagnosing growth-hormone deficiency in children. Withdrawn commercially; adult use is off-label with thin data.
Long-acting GHRH analogue. Raises growth hormone for days; development stopped and no outcome trial exists.
A fat-loss peptide that was properly tested in humans — and failed.
Mitochondrial-derived peptide with striking mouse metabolic data and no human trial.
First-generation daily GLP-1 receptor agonist. Approved for diabetes and obesity; generics arriving.
GLP-1/glucagon dual agonist. Phase 3 published for obesity and liver disease; not yet approved.
GLP-1/glucagon dual agonist. Approved in China for obesity and type 2 diabetes; phase 2 elsewhere.
Reproductive-axis neuropeptide. Small academic trials in fertility and low sexual desire; not approved.
Immune-modulating thymic peptide. Approved in China and ~30 other countries; failed its largest trial.
The human cathelicidin antimicrobial peptide. Two randomised wound-healing trials; not approved.
Mitochondria-targeting tetrapeptide. Newly approved for one ultra-rare disease after a decade of failed trials in bigger ones.
A real hormone with a real injectable licence — and a nasal-spray reputation its trials do not support.
A tanning-and-erection peptide with real small-trial signals and real, documented harms.
A potent GH-releasing peptide whose effect fades with continued use — and that was never developed further.
A GH-releasing peptide approved in Japan as a diagnostic — not as the muscle drug it is sold as.
The original GH-releasing peptide — a good diagnostic tool, an unproven physique drug.
Not a peptide, but sold in the same market — supplements that raise a coenzyme, with clinical benefit still unproven.
An anti-inflammatory tripeptide with real mouse data and no human trial at all.
A mitochondria-encoded peptide with striking lab biology and only observational human data.
A potent synaptogenic molecule in rodents, and nothing at all in humans.
A metabolic enzyme inhibitor with mouse fat-loss data and no human trial.
The first oral small-molecule GLP-1 agonist, approved in the US for weight management in April 2026. A pill, not an injection — and not a peptide.
A GLP-1 and amylin agonist in one molecule, now named zenagamtide. The largest early-phase weight loss yet reported — in 125 people; two phase 2 diabetes trials published since.
An approved medicine for growth-hormone deficiency, and the original anti-ageing drug the evidence does not support.
A real fertility medicine, and the subject of one of the most thoroughly refuted weight-loss claims in medicine.
The body's main antioxidant, sold by injection for skin lightening on inconclusive evidence.
A laboratory reagent sold as a muscle drug. No human trial has ever tested it.
A sleep peptide that was actually tested in insomnia in the 1980s — and mostly did not work.
An oral, non-peptide growth-hormone secretagogue — the one with a year-long randomised trial, which found more lean mass and no less fat.
Once-weekly GLP-1 medicine approved since 2014. A cardiovascular-outcomes trial in nearly 10,000 people; weight loss well below the newer drugs.
The first GLP-1 medicine, from lizard venom, approved in 2005. Neutral in its cardiovascular trial; negative as a Parkinson's treatment.
The natural hormone at the top of the reproductive axis. A licensed diagnostic and fertility treatment — and a TRT-clinic add-on with no trial for that use.
The original tanning peptide — which became an approved medicine for a rare light-sensitivity disease, as a 60-day implant.
An approved obesity peptide — for a few thousand people worldwide with specific genetic defects, on trials of ten and eleven patients.
A myostatin-blocking protein sold as a muscle-building injection. The only human data is a six-patient gene-therapy trial — of a different thing.
A Russian dipeptide-derived nootropic, registered there since 2007, with one published trial — whose abstract is not even indexed.
A pig-brain peptide mixture given by infusion for stroke and dementia in some fifty countries. Cochrane: no effect on death after stroke, more serious adverse events.
The first medicine that builds bone rather than slowing its loss. Approved since 2002 on a trial that cut spinal fractures by two-thirds.
A GnRH analogue that switches the sex-hormone axis off. Forty years as a standard prostate-cancer treatment, and the drug behind most puberty suppression and IVF protocols.
The first amylin drug, approved in the US in 2005 for diabetes. Modest weight loss in trials — and the ancestor of cagrilintide.
A synthetic version of the hormone that switches growth hormone off. Approved since 1988 for acromegaly and hormone-secreting tumours — and it slows those tumours' growth.
A pegylated fragment of a muscle-made IGF-1 variant, sold for repair and growth. Never injected into a human in any study.
A peptide designed to kill the blood vessels that feed fat. Dramatic in obese monkeys, kidney-toxic in the same monkeys, and never reported in a human.
GHK-Cu's less-studied cousin, sold for hair growth. One laboratory study on plucked follicles; no human trial.
A 28-amino-acid gut and nerve hormone. Licensed in a few countries for erectile dysfunction; a 461-person COVID trial found nothing; sold as a nasal spray for 'mould illness' with no trial at all.
An 11-amino-acid fragment of erythropoietin that repairs without raising red cells. A real phase 2 in a rare neuropathy; sold online for everything else.
Cagrilintide and semaglutide in one pen. Two published phase 3 trials: 20.4% weight loss at 68 weeks — less than the company had hoped, more than either drug alone.
Lilly's selective amylin agonist. A published 48-week phase 2 found up to 20% weight loss with fatigue, not nausea, as the signature side effect.
Amgen's once-monthly antibody with two GLP-1 peptides attached. Up to 16% weight loss at 52 weeks in a published phase 2, with a GIP-blocking twist.
A Chinese once-weekly GLP-1 analogue with a published phase 3: 13.2% weight loss at 40 weeks. Approved nowhere outside China's review process.
The second bone-building peptide. An 18-month trial of 2,463 women cut new spinal fractures by more than 80%, with less high calcium than teriparatide.
A 32-amino-acid hormone that was an osteoporosis medicine for thirty years — until a cancer signal and a marginal fracture trial ended most of that use.
A nine-residue vasopressin analogue in use since the 1970s. Cuts night-time urination modestly; causes dangerous low sodium in about one older adult in thirteen. Banned in sport as a masking agent.
A calf-thymus peptide extract registered in Russia since the 1980s. Its longevity claim rests on one St Petersburg series without randomisation or blinding.
A three-amino-acid 'brain bioregulator' from the epitalon school. Cell and rat studies; the human reports are Russian, uncontrolled and unindexed.
The senolytic peptide. One landmark mouse paper in Cell, no human study of any kind, and a grey market that sells it as an anti-ageing injection.
Zealand and Roche's weekly amylin analogue. A published phase 2: 8–10% weight loss at 28 weeks in 493 adults, with nausea in one in five. Phase 3 started in September 2026.
A once-daily GLP-1 medicine with a neutral heart trial and one unexpected result: slower motor decline in early Parkinson's disease, in a 156-person phase 2.
The original peptide medicine, in use since 1922. Included as the reference point: what a century of evidence looks like, and what the sharpest peptide side effect looks like.
The best-selling peptide supplement. Pooled trials show better skin hydration and elasticity — until the trials are sorted by who paid for them, when the effect disappears.
A muscle dipeptide sold as a supplement for ageing and blood sugar. Eight small trials pooled show lower fasting glucose and HbA1c in prediabetes and diabetes.
A two-amino-acid 'thymic bioregulator' from the epitalon school. Mouse lifespan and tumour studies; no human trial.
A Russian brain-extract injection for stroke and cognitive decline, like cerebrolysin. One double-blind trial of 272 patients exists — with no numbers in its abstract.
Evidence matrix
The table every compound page is built around. One row per claimed effect; a grade, a size, and how many human studies stand behind it.
BPC-157 in one line: The most-discussed healing peptide on the internet has never been tested in a controlled human trial. Every outcome below is graded on animal data, which is why every grade is a D — not because the results are bad, but because they are not yet about people.
| Outcome | Grade | Effect size | Human studies | Where the evidence comes from |
|---|---|---|---|---|
| Tendon & ligament healing | DNo human evidence | Not measurable | 0 controlled | Rodent models of Achilles and MCL injury. Consistent in animals; no human trial has been run.12 |
| Gut mucosal healing | DNo human evidence | Not measurable | 0 controlled | Rat ulcer and colitis models. The original research programme; still preclinical.1 |
| Pain reduction | DNo human evidence | Not measurable | 1 case series (n=12) | A retrospective case report of 12 knee-pain patients given intra-articular injections; no control group. The only human study a 2025 systematic review could find.3 |
| Safety in humansadverse | DNo human evidence | Not measurable | 1 open-label pilot (n=2) | Two healthy adults given intravenous doses in an uncontrolled 2025 pilot showed no biomarker changes. That is the entire human safety literature.345 |
Several well-run randomised trials in people, pointing the same way. You can plan around this. It still says nothing about whether the effect is worth the side effects for you.
e.g. semaglutide → body weight
Research feed
New studies and reviews, reduced to what changed and whether a grade moved.
SYNCHRONIZE-2 (752 adults) is published in the New England Journal of Medicine: −8.2% on 3.6 mg and −9.8% on 6.0 mg against −3.9% on placebo, counting everyone whether or not they stayed on treatment. HbA1c fell 0.8–0.9 points. Gut side effects in 73–78% against 39%. Added as a new row at grade B.
Ten more lots of compounded glutathione injection were recalled in the US after two patients needed emergency care. Both had been given intravenously a product made for injection under the skin. It follows the FDA's August report of at least 30 endotoxin-type reactions and the September recall. There is still no efficacy trial for the injectable route.
ACHIEVE-4 followed 2,749 adults with type 2 diabetes and heart or kidney disease for a median of 2 years: a major cardiovascular event in 4.2% on the daily tablet against 5.0% on insulin glargine, hazard ratio 0.84 (95% CI 0.59–1.20). That confirms safety; it does not show the pill prevents heart attacks. The page now also records the US approval of 1 April 2026.
TRIUMPH-1 (2,339 adults without diabetes): −25.0% at 80 weeks on 12 mg against −3.9% on placebo. TRIUMPH-2 (1,152 adults with type 2 diabetes): −18.8% against −5.1%. Both papers count everyone randomised, so the figures sit below the 28.3% and 20.8% of the press releases. Still unapproved, and two trials are still unpublished: the overall grade stays B.
ZUPREME-1 randomised 493 adults with obesity and no diabetes. Weight fell 7.9–9.8% at 28 weeks against 1.7% on placebo, and up to 10.7% by week 42. Nausea 20% against 6%; vomiting and diarrhoea were no more common than on placebo. Phase 3, about 7,000 people, started on 22 September. Grade stays C for stage.
In force 1 January 2027. Semaglutide and tirzepatide remain permitted and stay on the Monitoring Program while WADA runs two studies in athletes. The section on non-approved substances now names peptides as examples. BPC-157, TB-500, the growth-hormone secretagogues, AOD-9604 and MOTS-c stay banned in the same sections as in 2026.
Regulation
The most-searched question in the category and the least honestly answered. Plain-language status, dated and sourced, for every compound in every jurisdiction.
Most research peptides have no marketing authorisation. Selling them for human use is a criminal offence for the seller; approved compounds are prescription-only.
United StatesApproved GLP-1s are prescription-only. Most other peptides are unapproved; the FDA reopened compounding of a dozen in 2026 and an advisory committee backed six in July, but none is on the list yet.
United KingdomUnlicensed medicines under the Human Medicines Regulations. The MHRA is prosecuting sellers, particularly of weight-loss injectables.
AustraliaThe TGA schedules most of these substances as prescription-only; personal importation is restricted and enforced at the border.
CanadaNo Health Canada authorisation for the research-market compounds; approved incretin drugs are prescription-only.
Sport (WADA)Growth-hormone secretagogues and releasing factors, BPC-157, TB-500 and most peptides discussed online are banned at all times, whatever their legal status.
Glossary
How this works
LYO answers one question: what does the human evidence actually say? The method is published, the studies behind each grade are listed by name, and the data is open — so you never have to take our word for any of it.
