What this peptide is and what studies showed about it
What it is
A short chain of nine links that was called the sleep peptide.
- a small molecule of the protein type
- the name comes from the first experiments of the 1970s
- it has an official name: emideltide
Where the name comes from
The name promises deep sleep, but in the studies the picture is more modest.
- in one controlled study only light sleep increased
- deep sleep and dreaming sleep stayed the same
- the name turned out louder than the result
What was measured in insomnia
Sleep was recorded with instruments, and people were asked themselves.
- how quickly they fall asleep
- how long they sleep and how often they wake
- how the person rates their own sleep
What came out
How people felt barely changed in the controlled studies.
- in the placebo-controlled study sleep was rated the same as before
- the positive numbers came from studies without a placebo
- almost all of them were done by one group of authors
What was not found
No daytime heaviness was noted after a night with the substance.
- in one small study daytime performance did not decline
- it was not compared with sleeping pills, though
- its effect on night sleep is modest
Stress and cortisol
The stress hormone did not change with the substance.
- a pituitary hormone fell in one study
- cortisol itself stayed the same
- in the second study nothing changed
Pain and withdrawal
Here the studies are very small.
- for pain there is one small open study
- in opiate and alcohol withdrawal the study was open, with no comparison
- a repeat check by another group showed less
How it was given in studies
In the work on sleep, pain and withdrawal: into a vein, slowly.
- the infusion took several minutes
- with a very fast infusion there was no effect
- the course lasted from one day to two weeks
An injection under the skin
This is how the substance is sold, but it was not given this way in clinical studies.
- the studies were done differently
- the safety of subcutaneous injections is not described in the studies
- a nasal spray was used in two small studies
What the regulator says
It is not a registered medicine.
- the American regulator assessed the substance in 2026
- it proposed not adding it to the list for pharmacies
- the quality of the forms on sale is not confirmed
Who should see a doctor first
Insomnia, pain and dependence have different causes.
- blood pressure: some drugs change how the substance is broken down
- pregnancy and breastfeeding
- any insomnia lasting longer than a few weeks
Important. Such a substance is not prescribed without an examination and a look at the cause of the sleep problem.
The main point
A substance with a nice name and modest results.
- night sleep changed only slightly
- in a small study no daytime heaviness was noted
- the decision is made by a doctor after finding the cause
The main point
DSIP (from the English delta sleep-inducing peptide) consists of 9 amino acids. Under this name it is sold as a remedy for sleep, stress, pain and recovery. The official international name of the substance is emideltide. According to a document of the U.S. Food and Drug Administration (FDA) of 11.05.2026, it is not an ingredient of any registered drug, and in total 209 people have received it intravenously. The most reassuring finding: after a week-long course, daytime performance was checked with tests and no decline was found on any measure (14 people, 1 study without a placebo group). The most surprising: in a controlled study with sleep recording, deep sleep did not increase (6 people, 1 study).
| Substance and form | What it does | What the studies showed | Who it is discussed for |
|---|---|---|---|
| DSIP (emideltide), slow intravenous infusion | in humans it follows the sleep and wake cycle | sleep in insomnia: 16 people, 1 double-blind placebo-controlled study, the authors do not expect substantial benefit | people with chronic insomnia |
| DSIP, intravenous, course of up to 6 days | by design, eases withdrawal symptoms | opiate and alcohol withdrawal: 107 people, 1 open uncontrolled study | people in withdrawal |
| DSIP, intravenous, a week in the morning and a week in the evening | changes the number of sleep attacks | narcolepsy: 1 person, a case report | one patient |
How it works
The substance was first described in the 1970s and was made synthetically in 1976–1977. From the name it seems to deepen delta sleep, the deepest stage; in the studies with sleep recording the picture is different, and it is described below. The body's own DSIP rises and falls with the sleep and wake cycle, and its rise, according to data cited by the FDA, is linked to suppression of both deep sleep and rapid sleep (REM, the dreaming phase). The plasma half-life in humans is 8 minutes: aminopeptidases break it down in the blood. In people with major depressive disorder the plasma level was higher (34 people after a suicide attempt and matched healthy people, 1 study; Westrin A Biol Psychiatry 1998. PMID 9606527[1]). This is an association, not a cause.
Chronic insomnia: what was measured and what came out
The largest placebo-controlled study included 16 people with chronic insomnia, 8 received DSIP and 8 glucose. The substance was given intravenously at 25 nmol/kg (about 1.5 mg for 70 kg) before each of three nights in a sleep laboratory, and sleep was recorded by polysomnography. Sleep efficiency and time to fall asleep were better in those who received DSIP, but according to the authors these effects are weak and partly explained by a chance change in the placebo group. Subjective sleep quality did not change. Other sleep recording measures did not differ between the groups or within the groups (Bes F Neuropsychobiology 1992. PMID 1299794[2]).
In a crossover study, 6 people with severe insomnia received 25 nmol/kg over 4 minutes. Total sleep time rose by 20 minutes on average. The number of awakenings, wake time after falling asleep and the time to slow-wave sleep decreased, but did not differ significantly from placebo or from the baseline nights. Of the stages, only the second, light one rose, and the same difference from placebo was already present on the baseline nights, while the third and fourth (deep sleep) and REM stayed the same. The authors judged the improvement to be of little clinical significance (Monti JM Int J Clin Pharmacol Res 1987. PMID 3583493[3]; the number of participants and the gain are from the FDA review, p. 29).
In the first crossover study by another group of authors, 6 middle-aged people received 25 nmol/kg right before sleep. In the first hour after the injection a weak alerting effect was noted, and from the second hour to the sixth, sleep was calmer, with fewer interruptions. Time to fall asleep and time of final awakening did not differ from placebo. The authors noted no daytime sleepiness after the injections. The authors described "trends" without numbers (Schneider-Helmert D Experientia 1981. PMID 7028502[4]; the comparison with placebo is from the FDA review, p. 25).
The largest numbers on the topic came from work without a placebo group: 14 middle-aged people received 30 nmol/kg intravenously for 7 evenings in a row. The same people received placebo on the adaptation night and on the night after the course, and the outside comparison was a group of healthy people matched for sex and age, so the study has no figures for a placebo group (FDA review, p. 26–27). Time to fall asleep fell from 58.4 to 27.6 minutes, wake time after falling asleep from 120.8 to 65.6 minutes, total sleep time rose from 313 to 385 minutes, and sleep efficiency by 27 % of baseline. The main practical observation: daytime state and mental performance were tested in detail before and after six injections, and no decline was found on any measure in the interval from 1.5 to 10 hours after morning awakening (Schneider-Helmert D Eur Neurol 1987. PMID 3622582[5]). That is, in these 14 people no daytime heaviness the next day was noted; nobody was compared with sleeping pills here.
Another study by this author included 18 people: 9 aged 29–59 and 9 aged 60–83, six intravenous injections of 30 nmol/kg over a week, without a placebo group. In the older group the time to fall asleep, initially 1–2 hours, shortened to 22–30 minutes, and total sleep time rose by 97 minutes versus 57 in the younger group. The effect was larger the more disturbed sleep was before treatment (Schneider-Helmert D Eur Neurol 1986. PMID 3792404[6]; numbers from the FDA review, p. 28). An open series of 7 people with severe insomnia, 10 injections each, showed normalization of sleep in 6 of 7 over a follow-up of 3–7 months (Kaeser HE Eur Neurol 1984. PMID 6391926[7]).
The large positive sleep numbers come from the work of one group of authors, without a placebo group; two independent placebo-controlled studies are described above, and their result is more modest. Experiments on infusion speed are described separately: with a 1-minute infusion there was no effect, with 2.5 minutes total sleep time rose by 30 minutes, with 7.5 minutes by 1 hour, and the authors named 4–8 minutes as the optimum (Schneider-Helmert D Neuropsychobiology 1983. PMID 6689058[8]; numbers from the FDA review, p. 28). For comparison: in chronic insomnia, the American College of Physicians guideline names cognitive behavioral therapy as the first line. For more on different approaches, see the articles 5 drug-free insomnia remedies, Insomnia: GABA, glycine, magnesium and Daridorexant.
Opiate and alcohol withdrawal
The largest study on the topic included 107 people, 47 with alcohol withdrawal and 60 with opiate withdrawal. DSIP was given intravenously at 25 nmol/kg, up to 6 injections a day and up to 6 days. There was no randomization, no placebo group and no standard withdrawal scales; the rating was given by doctors and nurses. 13 % in the alcohol group and 22 % in the opiate group dropped out of the assessment, and among those assessed, symptoms disappeared or clearly decreased in 97 % of the opiate and 87 % of the alcohol patients. Anxiety went away more slowly, over hours and after several injections. In some patients anxiety and insomnia returned 24–72 hours after the end of treatment, and 13 people remained without insomnia. The authors wrote that the result does not allow a conclusion about the benefit of maintenance treatment (Dick P Eur Neurol 1984. PMID 6548969[9]; the return of symptoms and the authors' conclusion are from the FDA review, p. 38–39). An earlier study from the same clinic had 67 people, with benefit in 48 of 49 assessed, and 27 % dropped out of the assessment (Dick P Neuropsychobiology 1983. PMID 6328354[10]). Both studies were done by one group in one clinic, and there is no independent replication for withdrawal either in PubMed or in the FDA review.
Another group of authors repeated the regimen in 7 people with opioid dependence, 35 nmol/kg intravenously, on a tapering schedule over 7 days. The mean score on the withdrawal scale SOWS (Subjective Opiate Withdrawal Scale) was 34.4 an hour before the first injection and 17.8 after the first. 2 of 7 people completed the full regimen, and both needed a sleeping medicine on the third day. The return of symptoms after the second and subsequent injections, except in one person, was not suppressed (Backmund M J Clin Psychopharmacol 1998. PMID 9617990[11], a letter without an abstract; numbers from the FDA review, p. 39). The ASAM and SAMHSA clinical guidelines on opioid dependence do not mention DSIP, and the drug approved in the U.S. for this is called lofexidine.
Pain, narcolepsy and stress
In chronic pain there is one open pilot study from 1984: 7 people with migraine, tinnitus and psychogenic pain attacks, intravenously for 5 days in a row and another 5 injections every 48–72 hours. Pain, by the authors' assessment, decreased in 6 of 7, and depressive states decreased at the same time. The abstract has no numerical pain values (Larbig W Eur Neurol 1984. PMID 6548970[12]). In narcolepsy, one 35-year-old man is described: sleep attacks fell from 6 to 3 a week, and daytime sleep went from daily to once every 2–3 days. Morning injections worked better than evening ones, and the injections did not cause cataplexy or sleep paralysis (Schneider-Helmert D Eur Neurol 1984. PMID 6548968[13]; numbers from the FDA review, p. 33).
The claim "lowers cortisol and protects against stress" was tested twice. In the first study, 11 healthy men received 25 nmol/kg intravenously: plasma ACTH (adrenocorticotropic hormone of the pituitary) fell for at least 3 hours compared with control, while cortisol followed its usual daily decline and did not change (Bjartell A Psychoneuroendocrinology 1989. PMID 2554357[14]). In the second, 20 healthy men received 3–4 mg of DSIP, that is, more than twice as much: the ACTH and cortisol responses to corticotropin-releasing hormone and to a meal were almost the same with DSIP and with placebo (Späth-Schwalbe E Psychoneuroendocrinology 1995. PMID 7777652[15]). In the only series that wrote about stress tolerance, two psychotherapists rated improvement in two patients, and no scales were used (PMID 6689058[8]; FDA review, p. 28).
A 2014 study: 15 people with diabetic retinopathy received 300 mcg of DSIP a day in the nose for 2 months. The recovery time of the visual evoked potential after light stress fell from 137.2 to 95.1 seconds (73.5 seconds in healthy people). There was no comparison group within the treatment, and what was measured was the electrical response of the retina and cortex, not vision (Kresyun N.V. Eksp Klin Farmakol 2014. PMID 25739189[16]).
How it is used in studies
In all the work on sleep, pain and withdrawal, DSIP was given intravenously and slowly: in the insomnia studies the infusion took 4 minutes, and the authors named 4–8 minutes as the optimum. Most often the dose was 25 nmol/kg (about 1.5 mg for 70 kg), less often 30 nmol/kg (about 1.8 mg) and 35 nmol/kg (about 2.1 mg). The range across all clinical studies was 25–150 nmol/kg, for 1 to 15 days. In insomnia it was given before sleep or an hour before bed; morning injections raised daytime activity and still improved night sleep, while two doses a day did not give such an effect (Schneider-Helmert D Eur Neurol 1984. PMID 6391925[17]). Subcutaneous administration, the way the substance is sold today, was not used in clinical studies: for each indication the FDA separately notes that it found data only on the intravenous route. DSIP was given intranasally in a 2001 letter with an electrophysiological outcome (Hruz P J Clin Psychopharmacol 2001. PMID 11763019[18]) and in the 2014 retinopathy study described above.
What the regulator says
A document of the FDA for the Pharmacy Compounding Advisory Committee (PCAC) of 11.05.2026 states: for emideltide there is no applicable monograph of the United States Pharmacopeia (USP) and the National Formulary (NF), and it is not an ingredient of any FDA-approved drug. It is not recognized by the European, Japanese or International Pharmacopoeias and is not a registered medicine in any country checked. Two pharmacy chains proposed in 2015 and 2018 to include it in the list for pharmacy compounding, both requests were withdrawn, and the FDA proposed not adding emideltide to this list. The committee met on 24.07.2026, and according to specialist press reports of the meeting, emideltide was the only one of the peptides reviewed that the committee did not support; the outcome and the minutes were not posted on the FDA website at the time of checking, so the vote split is not given here. The committee's vote does not change the law: for a substance to get on the list, the FDA must issue a draft rule and hold a public comment period.
DSIP is sold for subcutaneous injection and nasal spray, in 5 mg vials and 2 mg/mL solutions, whereas in the studies a single dose was about 1.5–2.1 mg intravenously. The FDA notes that different salts and derivatives are sold under one name, and in two reviewed requests the certificates of analysis did not match the declared substance; no tests for impurities, aggregates and endotoxins were found for either form. In the FDA adverse event reporting system (FAERS), there were no complaints about DSIP up to 03.03.2024.
Who should see a doctor first
The substance has no registered instructions, so contraindications are not listed. From the FDA document and the studies, the following is known. In alcohol and opiate withdrawal in 107 people, 95 had no significant adverse events, 9 had sweating, headache, nausea and dizziness, and 3 had serious ones: two had hypotension at the start of the first injection, and the third had repeated episodes of feeling unwell with sweating and nausea lasting 15 minutes each (FDA review, p. 51). The only interaction named by the regulator: dipeptidase inhibitors, for example captopril (a group of blood pressure drugs), may slow the breakdown of DSIP and raise its concentration. A subcutaneous peptide of nine amino acids carries, according to the FDA, a risk of an immune response, up to antibodies to the person's own peptide. There are no studies in children, pregnant or breastfeeding women, and the longest administration in the studies was 2 months in the nose in 15 people. For insomnia, pain, anxiety and dependence, a doctor and an examination come first: the causes of these conditions differ, and treatment is chosen for the cause.
What the studies showed
| What was studied | Who and how many | Result | Source |
|---|---|---|---|
| Insomnia, double-blind placebo-controlled | 16 people (8 and 8), 1 study, 3 nights | Time to fall asleep and sleep efficiency slightly better, subjective sleep quality unchanged | Bes F Neuropsychobiology 1992. PMID 1299794[2] |
| Insomnia, placebo-controlled crossover | 6 people, 1 study | Total sleep time +20 minutes, little clinical significance according to the authors; deep sleep did not rise, only stage 2 rose | Monti JM Int J Clin Pharmacol Res 1987. PMID 3583493[3] |
| Insomnia, placebo-controlled crossover | 6 people, 1 study | Better sleep from the 2nd hour to the 6th; time to fall asleep no different from placebo | Schneider-Helmert D Experientia 1981. PMID 7028502[4] |
| Infusion speed and series III–V | 6, 4 and 2 people, 1 publication, 3 series | 1-minute infusion without effect, optimum 4–8 minutes; sleep numbers without a norm group | Schneider-Helmert D Neuropsychobiology 1983. PMID 6689058[8] |
| Insomnia, no placebo | 14 people, 1 study, 7 evenings | Falling asleep 58.4 to 27.6 min; total sleep time 313 to 385 min; performance did not decline | Schneider-Helmert D Eur Neurol 1987. PMID 3622582[5] |
| Insomnia by age, no placebo | 18 people (9 and 9), 1 study, 6 injections in a week | Total sleep time +97 min at 60–83 years and +57 min at 29–59 years | Schneider-Helmert D Eur Neurol 1986. PMID 3792404[6] |
| Severe insomnia, open | 7 people, 1 study | Sleep normalized in 6 of 7, follow-up 3–7 months | Kaeser HE Eur Neurol 1984. PMID 6391926[7] |
| Opiate and alcohol withdrawal, open | 107 people (47 and 60), 1 study | Improvement in 97 % of the opiate and 87 % of the alcohol patients assessed; anxiety and insomnia returned after 1–3 days | Dick P Eur Neurol 1984. PMID 6548969[9] |
| Withdrawal, earlier study | 67 people, 1 study | Benefit in 48 of 49 assessed; 27 % dropped out of the assessment | Dick P Neuropsychobiology 1983. PMID 6328354[10] |
| Opiate withdrawal, repeat check | 7 people, 1 study | SOWS scale 34.4 to 17.8; 2 of 7 completed the regimen | Backmund M J Clin Psychopharmacol 1998. PMID 9617990[11] |
| Chronic pain, pilot | 7 people, 1 study | Pain decreased in 6 of 7, no numbers | Larbig W Eur Neurol 1984. PMID 6548970[12] |
| Narcolepsy, one case | 1 person, 1 report | Sleep attacks 6 to 3 a week | Schneider-Helmert D Eur Neurol 1984. PMID 6548968[13] |
| ACTH and cortisol in healthy people | 11 people, 1 crossover study | ACTH lower for at least 3 hours, cortisol unchanged | Bjartell A Psychoneuroendocrinology 1989. PMID 2554357[14] |
| ACTH and cortisol at 3–4 mg | 20 people, 2 experiments | ACTH and cortisol responses almost the same with DSIP and placebo | Späth-Schwalbe E Psychoneuroendocrinology 1995. PMID 7777652[15] |
| Retinopathy, in the nose for 2 months | 15 people, 1 uncontrolled study | Potential recovery 137.2 to 95.1 s | Kresyun N.V. Eksp Klin Farmakol 2014. PMID 25739189[16] |
| Benzodiazepine withdrawal, one case | 1 woman aged 47, 1 report | Sleep phase shifted by 5 hours, drug withdrawn | Schneider-Helmert D Dtsch Med Wochenschr 1987. PMID 3582201[19] |
Summary
DSIP was studied in small groups: in the controlled crossover study with sleep recording (6 people) only the second, light stage rose, and cortisol did not change in two studies in healthy people. The large numbers for falling asleep and for withdrawal came from series by one group of authors without a placebo group, and the same series noted no decline in daytime performance. Everything known about sleep, pain and withdrawal was obtained with slow intravenous administration; the substance was given intranasally in two studies with electrophysiological measures, and the FDA found no clinical studies with subcutaneous administration, the way DSIP is sold. The forms on sale are not characterized, according to the same FDA document. With problems of sleep, pain or dependence, it is more reliable to start by finding the cause. The situation can be reviewed at a consultation.
References
- Westrin A Biol Psychiatry 1998. PMID 9606527
- Bes F Neuropsychobiology 1992. PMID 1299794
- Monti JM Int J Clin Pharmacol Res 1987. PMID 3583493
- Schneider-Helmert D Experientia 1981. PMID 7028502
- Schneider-Helmert D Eur Neurol 1987. PMID 3622582
- Schneider-Helmert D Eur Neurol 1986. PMID 3792404
- Kaeser HE Eur Neurol 1984. PMID 6391926
- Schneider-Helmert D Neuropsychobiology 1983. PMID 6689058
- Dick P Eur Neurol 1984. PMID 6548969
- Dick P Neuropsychobiology 1983. PMID 6328354
- Backmund M J Clin Psychopharmacol 1998. PMID 9617990
- Larbig W Eur Neurol 1984. PMID 6548970
- Schneider-Helmert D Eur Neurol 1984. PMID 6548968
- Bjartell A Psychoneuroendocrinology 1989. PMID 2554357
- Späth-Schwalbe E Psychoneuroendocrinology 1995. PMID 7777652
- Kresyun N.V. Eksp Klin Farmakol 2014. PMID 25739189
- Schneider-Helmert D Eur Neurol 1984. PMID 6391925
- Hruz P J Clin Psychopharmacol 2001. PMID 11763019
- PMID 3582201
Key facts
- Emideltide (DSIP) is not an ingredient of any FDA-approved drug and is not recognized by the European, Japanese or International Pharmacopoeias (FDA document of 11.05.2026).
- In total, 209 people received the substance intravenously, at doses of 25–150 nmol/kg, for 1 to 15 days (FDA document, p. 51).
- In a double-blind placebo-controlled study in chronic insomnia (16 people, 1 study), sleep efficiency and time to fall asleep were slightly better, subjective sleep quality did not change, and the authors do not expect substantial benefit (Bes F Neuropsychobiology 1992. PMID 1299794).
- In a placebo-controlled crossover study (6 people, 1 study), total sleep time rose by 20 minutes on average, and the gain came from the second, light stage: stage 1, deep sleep (stages 3 and 4) and REM did not change (Monti JM 1987. PMID 3583493; the size of the gain is from the FDA review, p. 29).
- Cortisol did not change either at 25 nmol/kg (11 people, 1 study) or at 3–4 mg (20 people, 2 experiments) (PMID 2554357; PMID 7777652).
- All the work on sleep, pain and withdrawal used intravenous administration; the authors called slow infusion essential, and with a 1-minute infusion there was no effect (PMID 6689058). The FDA found no clinical studies with subcutaneous administration (FDA document, p. 31, 36, 39).




