DSIP, or delta sleep inducing peptide, is a nine-amino-acid peptide first isolated from rabbit blood in the 1970s. Researchers named it for the deep-sleep brainwave pattern it appeared to trigger. Human evidence stayed small and mixed. In July 2026, an FDA advisory panel declined to recommend it. This page sorts what the studies actually found.
.
What DSIP actually is
Chemists can describe DSIP down to the last atom. Biologists still cannot explain it.
Property | Value |
|---|---|
Peptide class | Nonapeptide (9 amino acids) |
Sequence (three-letter) | Trp-Ala-Gly-Gly-Asp-Ala-Ser-Gly-Glu |
Sequence (one-letter) | WAGGDASGE |
Molecular formula | C35H48N10O15 |
Molecular weight | 848.81 g/mol |
CAS number | 62568-57-4 |
Formal designation | Emideltide |
Common aliases | DSIP, delta sleep peptide, sleep inducing peptide |
Three names circle the same molecule, and the overlap confuses people. DSIP and delta sleep inducing peptide are the same thing, one abbreviated. Emideltide is the formal designation the FDA used in its 2026 review documents. Anyone reading the regulatory record and the research literature side by side is reading about one peptide under two labels.
The discovery story gets told badly almost everywhere. Several sites date the find to 1974. That year has no primary paper behind it that holds up to checking. What the record does support is a longer arc. Michel Monnier identified a sleep-promoting factor in rabbit blood during work in the early 1960s. Characterization of the actual nonapeptide came later, published by Schoenenberger and Monnier in the Proceedings of the National Academy of Sciences in 1977. So the honest answer spans fourteen years, from a blood factor nobody could name to a sequence somebody could synthesize.
DSIP-like material has since turned up across the body. Studies using radioimmunoassay found it in brain tissue, peripheral organs and plasma across several mammal species. One caution belongs with that finding. These methods detect immunoreactivity, meaning something the antibody binds to. That is not proof the peptide itself sits in those tissues. A 2006 review argued the signal may come from a related but different molecule.
How researchers think it works
Start with the gap at the centre of the subject. Nobody has isolated a DSIP gene, a DSIP protein product, or a DSIP receptor. Every mechanism below is a proposal built on indirect observation.
Adrenergic modulation. Graf and Schoenenberger reported in 1987 that DSIP changed how rat pineal N-acetyltransferase responded to norepinephrine, acting through alpha-1 receptors. The work was done in vitro, in rat tissue. It suggests a route by which a peptide might touch sleep-related signalling. Nothing in it demonstrates that route operates in a person.
Stress hormone signalling. A 1985 rat study found that DSIP reduced corticosterone release triggered by corticotropin releasing factor. The effect appeared at the pituitary level and did not show up when researchers stimulated release with ACTH instead. Precision matters when this finding gets repeated online. The study measured corticosterone, the rodent equivalent of cortisol, under CRF stimulation in anaesthetised rats. It did not measure resting cortisol in humans.
Getting into the brain. Zlokovic and colleagues perfused guinea pig forebrain in 1989 and found a saturable, high-affinity transport mechanism carrying DSIP across the blood-brain barrier. A broader 1983 review placed DSIP among the peptides that reach cerebrospinal fluid intact to a slight or moderate degree. Both findings point the same direction. Some transport happens in animal models, and the amount reaching brain tissue looks modest.
Put together, these give a plausible sketch and no confirmed pathway.
What the research actually shows
Here is the full picture, graded by what kind of study produced it.
Study | Year | Model | n | What it found | Evidence grade | PMID |
|---|---|---|---|---|---|---|
Schneider-Helmert & Schoenenberger | 1981 | Human, chronic insomnia | 6 | Longer sleep, fewer interruptions, no daytime sedation | Controlled human trial | 7028502 |
Dick, Grandjean & Tissot | 1983 | Human, alcohol and opiate withdrawal | 67 (49 evaluable) | Benefit reported in 48 of 49 evaluable patients | Open-label human | 6328354 |
Schneider-Helmert | 1984 | Human, narcolepsy | 1 | Fewer sleep attacks, better daytime alertness | Case report | 6548968 |
Graf, Kastin, Coy & Fischman | 1985 | Rat | Not stated | Reduced CRF-stimulated corticosterone release | Animal | 2995861 |
Susić & Mašírević | 1985 | Cat, sleep-deprived | Not stated | Shifted light sleep toward deep slow-wave sleep | Animal | 2421663 |
Monti, Debellis, Alterwain et al. | 1987 | Human, chronic insomnia | Not stated | Changes seen, but no significant difference against baseline or placebo | Controlled human trial | 3583493 |
Schneider-Helmert | 1987 | Human, chronic insomnia | 14 | Improved sleep and daytime performance over 7 nights | Controlled human trial | 3622582 |
Zlokovic, Susic, Davson et al. | 1989 | Guinea pig brain, perfused | Not stated | Saturable high-affinity blood-brain barrier transport | Ex vivo animal | 2547200 |
van Kammen, Widerlöv, Neylan et al. | 1992 | Human, schizophrenia | 15 | CSF DSIP-like immunoreactivity correlated with slow-wave sleep | Human observational | 1475566 |
Vgontzas, Friedman, Chrousos et al. | 1995 | Human, apnea and narcolepsy | 30 | No significant plasma differences between patient groups and controls | Human observational | 8532601 |
Backmund, Meyer, Rothenhaeusler & Soyka | 1998 | Human, opioid detoxification | Not stated | Reported outcomes in an open trial | Open-label human | 9617990 |
A pattern jumps out once the studies sit in one place. The uncontrolled work agrees with itself, and the controlled work does not.
Look at the three controlled human sleep trials. Schneider-Helmert found benefit twice, once in six patients and once in fourteen. Monti and colleagues ran a double-blind crossover design, saw the same directional changes, then checked them against baseline and placebo nights. The differences were already there before treatment. Their conclusion was that the sleep improvement carried little clinical significance.
Scale deserves just as much attention. Fourteen people is the largest controlled human sleep trial DSIP has ever had. Modern sleep medicines reach approval on trials with hundreds or thousands of participants. The withdrawal work looks more impressive at 67 patients, until you notice it was open-label, with no placebo group and no blinding, in 1983.
Two human studies measured DSIP levels without giving anyone DSIP. Van Kammen found that cerebrospinal fluid DSIP-like immunoreactivity tracked with slow-wave sleep in fifteen men with schizophrenia. Vgontzas measured plasma levels across sleep apnea patients, narcolepsy patients and controls, and found no significant differences between the groups. One supports a link, one does not.
All human administration in this literature happened intravenously, in clinical settings, under physician supervision, decades ago. That context travels with the findings and cannot be separated from them.
What the evidence does not show
In 2006, two researchers published a review in the Journal of Neurochemistry that should be required reading for anyone evaluating DSIP. Kovalzon and Strekalova described the link between DSIP and sleep as never properly characterized. They noted the gene was never isolated, the protein was never isolated, and no receptor was ever found. Their verdict on the sleep-factor hypothesis was that it remains poorly documented and weak. They went further, proposing that a different, DSIP-like peptide might account for much of what researchers attributed to DSIP itself.
That assessment has not been overturned. Most pages covering DSIP ignore it completely.
Here is what remains genuinely unknown:
- No modern trials. The controlled human sleep work ended in the 1980s. Nothing at comparable scale has replaced it.
- No long-term data. Every human study ran for a single night or a handful of nights. Nobody has studied what repeated exposure does.
- No receptor. Without one, nobody can explain how DSIP would produce an effect.
- No pharmacokinetic picture in humans. Questions about half-life have no well-sourced answer.
Three claims that circulate as settled are better described as graded:
Does it cross the blood-brain barrier? Animal evidence supports limited, saturable transport. Nothing establishes efficient delivery to the human brain.
Does it lower cortisol? One rat study showed reduced corticosterone under CRF stimulation. No human cortisol data supports the broader claim.
Does it improve sleep? Controlled human trials disagree with each other, and the disagreement has never been resolved.
One adequately powered modern trial would settle most of this. No such trial is currently underway.
How DSIP compares to melatonin
People reach for this comparison because both get discussed around sleep. They sit in completely different evidence categories.
DSIP | Melatonin | |
|---|---|---|
What it is | Synthetic nonapeptide | Hormone produced by the pineal gland |
Primary research focus | Sleep depth and slow-wave architecture | Sleep onset and circadian timing |
Human trial volume | Three small controlled trials, all pre-1990 | Hundreds of trials, many modern |
Largest controlled sleep trial | 14 participants | Thousands across pooled analyses |
US regulatory status | Research use only, not FDA approved | Sold as a dietary supplement |
Mechanism understood | No receptor identified | MT1 and MT2 receptors characterized |
Main open question | Whether any effect is real | Optimal timing and formulation |
The asymmetry is the whole story. Melatonin has a known receptor system, a large modern literature and decades of consumer use. DSIP has a handful of small studies from the Reagan era and an unexplained mechanism.
Research interest also pointed in different directions. Melatonin work concentrated on falling asleep and shifting circadian rhythm. DSIP work concentrated on sleep architecture, specifically the delta-wave activity that gave the peptide its name. Comparing them as competing sleep aids misreads both literatures.
Where DSIP and Epitalon diverged
Two peptides went before the same FDA advisory committee on the same day, July 24, 2026. Both were reviewed partly for insomnia. The committee sent them in opposite directions.
Epitalon received a favourable recommendation, reported at 7-4 with one abstention. Emideltide was rejected, reported at 6-7 with one abstention.
What separated them was not a ruling on whether either peptide works. Advisory committee votes on the 503A Bulks List turn on whether the evidence supports compounding a substance. That judgment folds in safety data, the quality of existing studies and the strength of the case presented. An FDA reviewer at the meeting pointed to a lack of safety and efficacy data supporting Emideltide.
Readers following both compounds can see how the committee handled Epitalon in our dedicated explainer. The regulatory paths genuinely split that afternoon.
What the FDA panel decided in July 2026
This section separates three different kinds of statement, because they carry different weight.
Verified. The FDA's Pharmacy Compounding Advisory Committee met on July 23 and 24, 2026, at the agency's White Oak campus. Seven peptide substances were under review for the Section 503A Bulks List. Emideltide, the formal name for DSIP, was reviewed on the second day for opioid withdrawal, chronic insomnia and narcolepsy. The committee did not recommend it. Six other peptides, including BPC-157, TB-500, KPV, MOTS-c, Semax and Epitalon, received favourable recommendations.
Reported by law firms and trade press. The tally on Emideltide was 6 in favour, 7 against, with 1 abstention. The FDA did not publish an official vote count, so this figure comes from attorneys and journalists who attended. Multiple independent accounts agree on it.
Interpretation. Here is the detail most coverage missed. FDA's own scientific reviewers had posted briefing materials recommending against all seven peptides. The committee overrode its staff six times across two days. Emideltide was the only vote where the panel agreed with FDA's own reviewers, and it lost by a single vote.
That makes the Emideltide result different in kind from the other six. The committee showed plenty of willingness to disagree with agency scientists. On this one substance, it did not.
Nothing changed legally on July 24. A PCAC vote carries advisory weight only. Adding any substance to the 503A Bulks List requires formal notice-and-comment rulemaking, meaning a proposed rule, a public comment period and a final rule. As of this writing, FDA has taken no such action on Emideltide or on the six recommended peptides. The process has historically taken well over a year to reach even a proposed rule.
One framing point deserves particular emphasis here. A compounding-list vote asks whether pharmacies should be permitted to prepare a substance as an unapproved drug. It is a judgment about the evidence package, not a scientific verdict on whether a peptide does anything.
Safety, and what nobody can tell you
The honest safety position on DSIP is that there is no modern safety dataset.
Small trials from the 1980s reported no major side effects. That sentence gets quoted often and carries less than it seems to. Six people over four nights, or fourteen over seven, cannot detect uncommon harms. Absence of reported problems in a study that small is not a safety record. It is a sample too small to find much of anything.
Nobody has published data on repeated or extended exposure. Interactions with other compounds remain unstudied. Researchers never characterized what happens in people with liver or kidney impairment, or in anyone outside the narrow groups those old trials enrolled.
Purity testing cannot fill this gap, and the distinction trips people up constantly. A certificate of analysis tells you what is in a vial. It confirms identity and measures purity against the stated specification. What it cannot tell you is whether the substance inside is safe or effective in a living system. Our guide to reading a certificate of analysis covers what those documents actually establish. A 99% purity figure answers a chemistry question, not a biology one.
DSIP is sold for laboratory research. It is not approved by the FDA for any human use, and this article provides no protocol, administration method or usage guidance.
What people report online
Forum and Reddit discussion of DSIP exists in volume, and it is anecdote. Treat everything in this section as uncontrolled self-report.
Recurring themes show up across those threads. Some users describe deeper sleep without a sedative feeling. Others report feeling nothing at all. A smaller group mentions unusually vivid dreams. Reports of next-day grogginess appear, as do reports of its absence.
Why this cannot function as evidence is worth spelling out. Sleep quality is subjective, highly variable night to night, and strongly responsive to expectation. Nobody posting a report had a placebo condition. None of them were blinded either. People who notice an effect post more often than people who notice nothing.
The controlled trials are the relevant comparison. Three of them exist, and they disagreed with each other under conditions specifically designed to remove the biases that forum reports are full of.
The bottom line
DSIP occupies an unusual place in peptide research. Its chemistry is settled, down to the formula and CAS number. The biology is not, and that gap has lasted almost fifty years.
The research gives us a small, dated and internally contradictory body of work. Two controlled trials found sleep benefit, one found it clinically insignificant, and animal studies describe a mechanism nobody has confirmed in people. Missing from all of it is a receptor or a modern safety profile. The evidence also failed to persuade an FDA advisory committee in July 2026.
Anyone researching DSIP should know both halves of that picture.
DSIP sold by PrimeTime BioLabs is intended for laboratory research only. It is not for human or veterinary use, not for diagnostic or therapeutic application, and not approved by the FDA for any medical purpose. See our DSIP product page for specifications, and our guide to choosing a research peptide supplier for what to check before you buy.
