DSIP stands for delta sleep-inducing peptide. The name promises a lot. Deep, restorative sleep from a small injection. In July 2026 it became the one peptide out of seven that an FDA advisory panel voted against. The reasons why tell you most of what you need to know about it.
Weak and old. Most of the human research is from the 1980s, all of it small, and all of it used IV dosing, not the under-the-skin injections people use today. The better-designed sleep trials found little or no effect. Nearly fifty years after it was described, nobody has found the gene that makes it, the protein it comes from, or the receptor it acts on.
DSIP is a chain of nine amino acids. Its official drug name is emideltide. It was isolated by Swiss researchers who collected blood draining from the brains of rabbits while the animals were put into deep sleep with electrical stimulation. When that blood was given to other rabbits, it seemed to increase deep sleep. The team characterized the peptide in 1977.
You will sometimes read that it was “discovered in rabbit brain.” It was found in blood leaving the brain, not brain tissue.
Here is the odd part. Scientists have measured DSIP-like material in the body, but no one has ever found a gene or precursor protein for it. A 2006 review called the sleep-factor idea “extremely poorly documented.” The authors suggested that what researchers measured may have been a related, still unidentified molecule.
Nobody really knows. In animals, DSIP increased deep sleep in some labs but not in others. One well-known 1980 rat study failed to reproduce the effect. Some rat work suggests it may act indirectly through the body’s opioid system, since an opioid blocker cancelled out the sleep effect. Other claims about stress, seizures and pain come from rodents only.
There is also a delivery question. Small amounts appear to cross into the brain, but the peptide is reportedly broken down in the blood within minutes. How much of an injection reaches the brain is uncertain.
Insomnia. Early studies by one Swiss researcher reported benefits, but they were very small and often compared patients to outside groups. A 1987 randomized, double-blind, placebo-controlled crossover study of six people found no significant effect. A 1992 double-blind study of 16 people found “at best some weak effects.” A 1986 review concluded the evidence was not enough.
Narcolepsy. One case report, one man.
Withdrawal. A 1984 study gave DSIP to opioid and alcohol withdrawal patients with no control group. A 1998 study enrolled seven patients, and only two finished.
Those studies used about 25 nanomoles per kilogram by IV, roughly 1.5 mg for a 70 kg adult.
Sleep, mostly. Also stress and recovery. It is sold as injectable vials and nasal sprays through gray-market sellers, clinics and med spas.
There is no long-term safety data, and the FDA said it found no safety information at all for the under-the-skin route people actually use. Its main concern was immune reactions to the peptide or its impurities. The FDA also noted the original nominations were not even consistent about which chemical form they meant. Gray-market vials come with no guarantee of what is inside, how much, or whether it is sterile.
Reporting, not a recommendation. Community sources describe 100 to 300 micrograms, sometimes up to 500, injected under the skin before bed in short cycles. That is a different route from every human study, and several times lower than the IV doses tested. No human data supports these regimens. We do not give mixing or injection steps.
Not FDA-approved. The FDA placed DSIP in Category 2 in 2023, which blocked compounding. It was removed in April 2026 after the nominations were withdrawn, and the FDA then reviewed it on its own.
At the July 24, 2026 Pharmacy Compounding Advisory Committee meeting, it was reviewed for opioid withdrawal, chronic insomnia and narcolepsy. FDA staff recommended against it. They pointed to poor characterization, no data for the under-the-skin route, conflicting insomnia results, a single narcolepsy case, uncontrolled withdrawal studies, and the fact that approved treatments already exist for all three. The panel agreed this time. The vote was 6 yes, 7 no, with 1 abstention. It was the only one of seven peptides not recommended. You can read how that meeting played out in our post on what happens after the FDA panel vote.
DSIP is not on the compounding list, so pharmacies cannot lawfully compound it. It is not named on the WADA list, but it almost certainly falls under the “non-approved substances” category, which is banned at all times.
Is DSIP made naturally in the body? DSIP-like material has been measured, but no gene or precursor has ever been found.
Does it work for insomnia? The controlled trials are tiny and conflicting. The better-designed ones found little or no benefit.
Is it FDA-approved? No. And the FDA’s advisory panel voted against allowing pharmacies to compound it.
Can a compounding pharmacy legally make it? No. It is not on the approved compounding list.
Is there an established dose? No. Studies used IV dosing by body weight. Community doses under the skin are untested.
Is it safe? Unknown. There is no long-term or under-the-skin safety data.
Is it banned in sport? Effectively yes, under WADA’s non-approved substances category.
Plain-language updates on new research and the shifting rules. No hype, unsubscribe anytime.