DSIP for sleep: what happened when two other teams repeated the experiment

The short answer

• DSIP, delta sleep-inducing peptide, is a synthetic nine-amino-acid peptide first isolated in 1977 from the cerebral venous blood of rabbits whose thalamus had been electrically stimulated to induce sleep. The name comes from the EEG delta activity it enhanced in recipient rabbits.
• Evidence tier for chronic insomnia: Early human data, and conflicting. The Swiss group that discovered it reported normalised sleep in several small intravenous studies. Two independent double-blind studies, in Uruguay in 1987 and the Netherlands in 1992, found little or no clinically meaningful effect.
• Evidence tier for the subcutaneous injections and nasal sprays actually sold today: Preclinical only. There are no human data of any kind for those routes in sleep.
• No DSIP gene, precursor protein or receptor has ever been identified, nearly 50 years after the peptide was described. A 2006 review in the Journal of Neurochemistry called it a still unresolved riddle.
• Every controlled human sleep study used slow, weight-based intravenous injection over three to seven nights. The products in circulation use a different route, a flat dose rather than a weight-based one, and open-ended duration.
• As of September 2026 DSIP is not approved anywhere that could be verified, and no study of it is registered on ClinicalTrials.gov.

What the positive studies actually found

The original clinical work came from Schoenenberger and Schneider-Helmert in Switzerland, the same group that isolated and sequenced the peptide.

A 1981 randomised double-blind crossover report in the Lancet gave six people with chronic insomnia 25 nmol/kg by intravenous injection for three nights. Awakenings fell and sleep efficiency rose. Notably, sleep onset did not change, which is worth holding onto given that DSIP is now marketed largely as a falling-asleep aid.

A 1986 study in 18 people with chronic insomnia used 30 nmol/kg intravenously on six occasions and reported normalised sleep, with a larger initial effect in older patients. A 1987 placebo-controlled study in 14 people with severe chronic insomnia used the same dose for seven nights and reported improvement in both night sleep and daytime alertness, though it compared results against external healthy controls rather than a matched group.

These are small studies, and all of them come from the group with the strongest interest in the result. That is not an accusation, it is the ordinary reason replication exists.

What happened when other teams tried

Two independent groups ran similarly designed double-blind studies, and this is the part that usually gets left out of the marketing.

Monti and colleagues, in 1987, ran a randomised double-blind crossover study in people with chronic insomnia at 25 nmol/kg intravenously for four nights. The differences they found between DSIP and placebo were already present at baseline, and they described the effect as of little clinical significance.

Bes and colleagues, in 1992, studied 16 people with chronic insomnia double-blind at 25 nmol/kg intravenously for three nights. The effects they measured were weak and partly explained by a change in the placebo group. Their conclusion was that DSIP was not likely to be of major therapeutic benefit.

The precise way to describe this is not that DSIP was never replicated. Replication was attempted, twice, by different groups using a comparable design and the same dose, and it failed. Nothing published since has resolved the disagreement.

The biology that was never found

Most peptides sold today have at least a receptor. DSIP does not.

Despite decades of searching, no gene, no precursor protein and no receptor have been identified. The 2006 review that called DSIP an unresolved riddle proposed that the DSIP-like immunoreactivity measured in tissue may actually belong to related, unidentified peptides. If you want the general picture of how a peptide is supposed to produce an effect, see how peptides work in the body.

There is also a timing problem nobody has explained. DSIP has a blood half-life of under ten minutes and is rapidly chopped up by plasma peptidases into fragments including free tryptophan. Yet the human studies reported sleep effects lasting many hours, with a latency of roughly an hour after injection. The original investigators framed this as a “programming” effect. That is a description, not a mechanism, and it remains open.

Radiolabelled DSIP does cross the blood-brain barrier in rats by a non-competitive mechanism, so brain access is not the objection, though rapid peripheral degradation limits how much intact peptide arrives.

The claims that human data contradict

Two marketing claims can be checked directly against published human studies, and both fail.

DSIP is frequently sold as a growth-hormone-supporting sleep peptide. Intravenous DSIP did not change growth hormone or prolactin secretion in healthy women. It also did not affect CRH-stimulated or meal-induced ACTH and cortisol. The endocrine story has been tested and did not hold.

The broader claim that deep sleep peptides raise slow-wave sleep as a class is not supportable either. Among growth hormone secretagogues the effects run in different directions: GHRH promotes slow-wave sleep, intravenous GHRP-6 increased stage 2 but not slow-wave sleep, and intravenous hexarelin actually decreased stage 4 sleep and delta power. There is no class effect to borrow.

Route, dose and regulatory status as of September 2026

Every controlled human sleep study used slow intravenous injection at a weight-based dose: 25 nmol/kg in the 1981, 1987 and 1992 studies, roughly 21 micrograms per kilogram, and 30 nmol/kg in the 1986 and 1987 Swiss studies. Slow injection was described as essential. Studies ran three to seven nights. No study tested subcutaneous dosing, flat doses, long-term nightly use, or the nasal route for sleep.

What circulates now is flat-dose subcutaneous injection and nasal spray. The compounded product nominated to FDA was a 1,000 microgram per millilitre subcutaneous injection. That differs from the studied protocols in route, in dose basis and in duration, which is three separate reasons the trial data do not transfer.

On regulation, FDA has not approved DSIP, which it calls emideltide. The substance appears on FDA’s list of bulk drug substances whose compounding nominations were withdrawn, a page current as of 22 April 2026, and removal from that list’s earlier safety category is not approval. At the Pharmacy Compounding Advisory Committee meeting of 23 to 24 July 2026, FDA proposed that emideltide free base and acetate not be added to the 503A bulks list, citing that both forms are not well characterised, with no impurity, aggregation, bioburden or endotoxin data, plus immunogenicity concerns for injected peptide and insufficient effectiveness and safety data. The committee voted against inclusion, 6 in favour, 7 against, 1 abstention.

In Canada there is no DIN for DSIP. It was named among unauthorised injectable peptides seized from an online seller in a Health Canada advisory of 1 August 2025, and again in the advisory of 9 April 2026 warning against injecting peptides bought online. That April 2026 advisory also establishes that a “research use only” label confers no legal status of its own.

DSIP is not named on the 2026 WADA Prohibited List, but S0 covers substances with no governmental approval for human therapeutic use anywhere, which is FDA’s finding. A tested athlete should treat it as prohibited. For the compound page, see DSIP.

Safety, as far as it is known

The small intravenous insomnia studies described DSIP as well tolerated. The largest series, an open-label 1984 study of 107 people in alcohol and opioid withdrawal, is more informative: 9 had minor effects including sweating, headache, nausea and vertigo, and 3 had serious events. Two became hypotensive on first injection, one of them progressively so when re-dosed, and one had repeated discomfort with sweating and nausea. Some patients had a relapse of anxiety with marked insomnia after stopping.

FDA’s adverse event reporting system contains no DSIP reports, and FDA itself notes that this does not imply safety for a product sold outside the regulated system. There is no chronic toxicology and no carcinogenicity data. The opioid-system involvement seen in animals raises an addiction-liability question for which FDA found no nonclinical data at all.

For context on what chronic insomnia treatment actually rests on: cognitive behavioural therapy for insomnia is first-line in both American College of Physicians and American Academy of Sleep Medicine guidance, and DSIP does not appear in the AASM guideline on pharmacologic treatment.

The bottom line

DSIP deserves more respect than most grey-market sleep peptides, because someone did the work: real double-blind studies, a real dose rationale, published in indexed journals. It also deserves an honest verdict, which is that the discovering group’s positive results did not survive two independent attempts to reproduce them, no receptor was ever found, and the route people inject today has never been tested in a human sleep study. That is a closed loop, not a promising lead awaiting funding.

Education only. This page is educational. Most compounds referenced are not approved for human use, and a “research use only” label carries no legal status of its own. Nothing here constitutes medical advice, diagnosis, or treatment.

Sources

• Schoenenberger GA, et al. Characterization of a delta-electroencephalogram (-sleep)-inducing peptide. Proc Natl Acad Sci U S A, 1977.
• Schneider-Helmert D, et al. Synthetic delta-sleep-inducing peptide improves sleep in insomniacs. Lancet, 1981.
• Monti JM, et al. Study of delta sleep-inducing peptide efficacy in improving sleep on short-term administration to chronic insomniacs. Int J Clin Pharmacol Res, 1987.
• Bes F, et al. Effects of delta sleep-inducing peptide on sleep of chronic insomniac patients. A double-blind study. Neuropsychobiology, 1992.
• Kovalzon VM, et al. Delta sleep-inducing peptide (DSIP): a still unresolved riddle. J Neurochem, 2006.
• Graf MV, et al. Degradation and aggregation of delta sleep-inducing peptide (DSIP) and two analogs in plasma and serum. Peptides, 1987.
• Giusti M, et al. Delta sleep-inducing peptide administration does not influence growth hormone and prolactin secretion in normal women. Psychoneuroendocrinology, 1993.
• Qaseem A, et al. Management of Chronic Insomnia Disorder in Adults: A Clinical Practice Guideline From the American College of Physicians. Ann Intern Med, 2016.