You have seen Semax, Selank and Dihexa sold as nootropics and DSIP sold for sleep. Most of that rests on Russian nasal-drop studies in stroke and anxiety patients, or on rats in water mazes. This page ranks every one by what was shown in people, where each is legal, and which trials came back empty.
Sorted by evidence grade, then by how many human studies exist. The left column has at least some data in people. The right column has only rodents and cell cultures, however confident the marketing sounds.
Dozens of randomized trials in stroke, brain injury and dementia. Independent Cochrane reviews are less positive than the industry-funded ones.
One randomized trial in older adults found better executive function. It is approved only for belly fat in people with HIV.
FDA-approved for moderate to severe obstructive sleep apnea in adults with obesity.
Not a peptide. Deepened sleep in a tiny study; did nothing for Alzheimer's in a 563-person trial.
Approved for labor. Nasal spray trials for social behavior and autism did not beat placebo.
Named for sleep in 1977. Independent insomnia trials found little or no benefit over placebo.
Raises growth hormone in small trials of older adults. No sleep outcome is in its record.
Human data stop at hormone levels. No study has measured sleep.
No human evidence for sleep. Its one efficacy trial, for gut recovery after surgery, failed.
Russian nasal drops studied in stroke and cerebrovascular patients without blinding. Never tested for focus in healthy people.
Two Russian anxiety studies compared it with benzodiazepines. Neither was a blinded placebo trial.
Small Russian reports on melatonin output from the group that invented it. No randomized trial.
Neuron cultures and rats, plus small human reports from the developers measured with questionnaires.
Not a peptide. Precursor pills raise blood NAD+, but outcome trials are mostly null and IV drips have no outcome trial.
Animal or cell data only. A memory result in a mouse shows a mechanism worth testing, not a treatment.
Rat and mouse data only. Its prodrug failed two Alzheimer's trials and a key paper was retracted in 2025.
Mouse and rat Alzheimer's models, almost all from the lab that made it.
Antidepressant-like behavior in mice from one French research group. No human data.
No compounds match these filters.
The doses used in each published trial are listed on every peptide page for free members. Create a free account
BDNF, GABA, enkephalins, growth hormone: each claim online names a pathway. This map shows each one, the peptides linked to it, and whether that link was seen in a person, in a blood test, or only in a rat.
BDNF is a growth protein that helps neurons survive and build new connections. It acts through the TrkB receptor on the neuron surface.
Blood BDNF rose in 110 stroke patients given two courses. The rise in brain BDNF was shown in rats.
Raised brain BDNF in mouse and rat models. A 2024 study led by another group saw no BDNF rise in living mice.
Growth-factor-like effects described in animal studies. It is a mixture with no single receptor.
Numbers on the diagram match the buttons. An amber ring marks every pathway the chosen peptide is linked to. Schematic, not to scale.
Bar length is the number of people, on a log scale so a 16-person sleep study and a 1,773-person stroke review both show. Solid bars were randomized. Outlined bars were open-label, with no blinding.
Death from any cause. No change in deaths. Non-fatal serious adverse events were higher with Cerebrolysin. Source
Combined recovery outcome at 90 days. No difference from placebo on the main outcome. Source
Cognition and global function. Small gains, but every trial was rated very low quality with high risk of bias. Source
Cognition and daily function at 12 months. IGF-1 rose by about 73 percent. No difference from placebo on any cognitive scale. Source
Breathing interruptions per hour at 52 weeks. Fell by 25 to 29 an hour, against about 5 on placebo, from a starting point near 50. Source
Social withdrawal at 24 weeks. Improved 3.7 points on oxytocin and 3.5 on placebo. No real difference. Source
Composite cognition and function at 26 weeks. Did not beat placebo. More people quit for side effects, mostly injection-site reactions. Source
Executive function at 20 weeks. Better than placebo. Side effects were reported by 68 percent of people on tesamorelin against 36 percent on placebo. Source
Western Aphasia Battery at day 90. Gained 14.8 more points than placebo. Source
Daily-living recovery (Barthel index) and blood BDNF. Faster recovery and higher BDNF, but groups were not described as randomized or blinded. Source
Depression and anxiety rating scale. Phenazepam worked sooner with fewer side effects when Selank was added. No placebo arm. Source
Anxiety scores. Anxiety fell about as much as on the benzodiazepine. No placebo arm. Source
Brain connectivity on fMRI only. Amygdala connectivity shifted within 20 minutes. Anxiety was not measured. Source
Brain network activity on fMRI only. A frontal network looked larger after a dose. No memory or focus test was given. Source
Sleep efficiency and time to fall asleep. Slightly better numbers, but people did not feel they slept better. Authors judged it unlikely to help much. Source
Sleep stages in a sleep lab. Deep sleep rose by about half and REM by over 20 percent in young adults. Source
Several of these are real prescription drugs in Russia or Austria and research chemicals in the US. Switch the toggle to count only approvals for a brain, mood or sleep use. Tap any peptide for its US compounding status.
Not legal to compound in the US as of this review. Semax was placed in 503A Category 2 (significant safety risks) in FDA's September 2023 update; after the nominations were withdrawn, FDA announced on April 15, 2026 that it would take Semax off Category 2, and it now appears only on FDA's list of withdrawn nominations. FDA then took Semax to its Pharmacy Compounding Advisory Committee on July 24, 2026 for cerebral ischemia, migraine and trigeminal neuralgia. FDA's briefing document concluded the criteria weigh against adding Semax to the bulks list, citing unclear identity (free base vs acetate), immunogenicity risk, no human pharmacokinetic data and possible bleeding risk. RAPS reported the committee voted 8-5 in favor anyway. The vote is advisory: FDA must still decide and go through rulemaking before compounding is allowed, and FDA had posted no decision as of this review.
No controlled trial has shown that Semax, Selank, Dihexa, P21, PE-22-28 or Pinealon improve memory or focus in healthy people. Two healthy-volunteer studies of Semax and Selank only took fMRI scans.
The Selank anxiety studies compared it with benzodiazepines or added it to one. None had a placebo arm, and neither was published in full in English.
For Semax, Selank, Dihexa, P21 and PE-22-28 there is no human pharmacokinetic study showing how much reaches the brain, by nasal spray or by injection.
Russian courses ran 10 to 14 days. Nobody has studied months of use, dependence, or what long-term growth-factor signaling from Dihexa does to tumor risk.
P21, PE-22-28 and Pinealon each come almost entirely from the group that created them. Three of four flagged Dihexa papers have been retracted.
None of the studies tested powders sold online. Whether a vial labeled P21 is the capped molecule used in the mouse work is unknown.
Sleep peptides are sold on the idea of deeper sleep. Only one compound here has an approved sleep use, and it treats a breathing disorder.
Light, temperature, caffeine and a fixed wake time move sleep more than any compound here. See the sleep guide
| Peptide | Class | Evidence | Studied for | Legal status | Half-life |
|---|---|---|---|---|---|
| Cerebrolysin | Nootropic | A Evidence 60 human studies | Memory and cognition | Approved outside US | Not established |
| Tesamorelin | GHRH analog | A Evidence 30 human studies | Memory and cognition | FDA-approved WADA prohibited | About 11 min |
| Tirzepatide | GLP-1 / incretin agonist | A Evidence 150 human studies | Sleep | FDA-approved | About 5 days |
| MK-677 (Ibutamoren) | GH secretagogue | A Evidence 25 human studies | Sleep | In clinical trials WADA prohibited | Not established |
| Oxytocin | Reproductive hormone | A Evidence 1000 human studies | Anxiety and mood | FDA-approved | About 1 to 6 minutes in plasma |
| DSIP (Delta Sleep-Inducing Peptide) | Nootropic | B Evidence 15 human studies | Sleep | Approved outside US | About 8 minutes in plasma |
Both went before an FDA advisory committee in July 2026 and are waiting on a ruling. One email when a trial reads out, a label changes or the FDA acts on a peptide, with the source linked.