DSIP is a nine-amino-acid peptide isolated from rabbit brain in the 1970s and named for the delta-wave sleep it was associated with. Its human literature is unusually clean in one respect and unusually awkward in another: the doses are stated precisely, in nanomoles per kilogram, and they are all intravenous. The research market sells a lyophilized powder for subcutaneous use in amounts that do not match. This page converts the trial figures into micrograms and syringe units, reports what each study found, and states plainly where the circulating figures depart from the published ones. It reports and does not recommend.
The DSIP calculator runs the same sums for any vial and volume; the dosage hub indexes every compound here, and the all-compounds chart carries the one-line version beside every other compound.
Why this compound's doses need converting at all
Almost every dosage figure on this site arrives in micrograms or milligrams, because that is how labels and vendor pages state them. DSIP's arrive in nanomoles per kilogram, because its human studies were pharmacological experiments rather than product trials, and a molar dose is the natural unit when the question is receptor occupancy rather than a package insert. A nanomole is a quantity of molecules, not a mass, so converting it needs the molecular weight: 848.8 g/mol for DSIP (PubChem CID 68816). Twenty-five nanomoles of a substance weighing 848.8 grams per mole is 21,220 nanograms — 21.2 micrograms. Per kilogram of body weight, that is the figure below.
| Trial dose | Micrograms per kg | 60 kg | 70 kg | 90 kg |
|---|---|---|---|---|
| 25 nmol/kg | 21.2 mcg/kg | 1.27 mg | 1.49 mg | 1.91 mg |
| 50 nmol/kg | 42.4 mcg/kg | 2.55 mg | 2.97 mg | 3.82 mg |
| 100 nmol/kg | 84.9 mcg/kg | 5.09 mg | 5.94 mg | 7.64 mg |
A single 100 nmol/kg dose for a 90 kg adult is 7.64 mg, which is more than a whole 5 mg vial. That is worth stating because it shows how far the published figures sit from the way the compound is packaged: the vial sizes on the market suit the circulating microgram figures, not the trial ones.
What the trials state
| Study | Subjects | Dose and route | What it reported | Opened |
|---|---|---|---|---|
| Lancet, PMID 6112579 (1981) | Insomniacs | Not in the abstract | Synthetic DSIP improves sleep in insomniacs | 2026-09-19 |
| Neuropsychobiology, PMID 6689058 (1983) | Healthy subjects | 25 nmol/kg intravenously, four nights | Multifunctional psychophysiological effects besides sleep induction | 2026-09-19 |
| Eur Neurol, PMID 3622582 (1987) | 14 middle-aged chronic insomniacs | Seven successive nights, dose not in the abstract | Sleep efficiency and daytime alertness reached normal-control levels; effects persisted into a placebo night | 2026-09-19 |
| Psychoneuroendocrinology, PMID 2554357 (1989) | 11 healthy men, 25 to 39 | Single 25 nmol/kg intravenous dose | Reduced immunoreactive ACTH in plasma | 2026-09-19 |
| Neuropsychobiology, PMID 1299794 (1992) | 16 chronic insomniacs | 25 nmol/kg intravenously on three afternoons | Higher sleep efficiency and shorter latency, but effects judged weak; unlikely to be of major therapeutic benefit | 2026-09-19 |
| Psychoneuroendocrinology, PMID 7777652 (1995) | Healthy young men, n = 5 per condition | Total doses of 3 mg and 4 mg by intravenous infusion | ACTH and cortisol responses almost identical to placebo | 2026-09-19 |
| Eur J Anaesthesiol, PMID 19142086 (2009) | 24 women, ASA I or II | Intravenous bolus at 25, 50 or 100 nmol/kg | Increased heart rate, decreased heart-rate variability, reduced delta rhythm and lightened anaesthetic depth | 2026-09-19 |
Seventeen records in PubMed carry the clinical-trial tag; the seven above are the ones whose abstracts state a dose, a design or a result specific enough to report. The set spans thirty-three years and two research questions — sleep, and the hypothalamic-pituitary axis — and it contains one clear reversal: the 2009 study set out to test DSIP as an anaesthetic adjunct on the hypothesis that a natural hypnotic would deepen anaesthesia, and measured the opposite.
The arithmetic per vial
The calculator's presets are the 5 mg and 10 mg vials. One U-100 unit is a hundredth of a millilitre (MedlinePlus, "Giving an insulin injection", reviewed 2024-07-21). The columns are the two figures that circulate for the vial, the 25 nmol/kg trial dose worked at 70 kg, and the 1995 study's 3 mg infusion total.
| Vial and water | mcg/mL | mcg per unit | 100 mcg | 300 mcg | 1,490 mcg | 3,000 mcg |
|---|---|---|---|---|---|---|
| 5 mg in 1 mL | 5,000 | 50 | 2 units | 6 | 29.8 | 60 |
| 5 mg in 2 mL | 2,500 | 25 | 4 | 12 | 59.6 | 120 |
| 5 mg in 3 mL | 1,667 | 16.7 | 6 | 18 | 89.4 | 180 |
| 10 mg in 1 mL | 10,000 | 100 | 1 | 3 | 14.9 | 30 |
| 10 mg in 2 mL | 5,000 | 50 | 2 | 6 | 29.8 | 60 |
| 10 mg in 3 mL | 3,333 | 33.3 | 3 | 9 | 44.7 | 90 |
Two rows exceed 100 units, which is a full 1 mL insulin syringe: the trial-scale amounts do not fit the small vials at ordinary water volumes. At the other end, a 10 mg vial in 1 mL puts 100 mcg at a single unit, where a half-unit reading error is half the amount. The circulating microgram figures and the published milligram-scale ones are not two versions of the same number; they are different quantities that happen to share a compound.
Where the circulating figures depart from the published ones
Three departures, each checkable. The route: every human dose in the literature was intravenous, including the sleep studies, and the figures stated for the vial are subcutaneous — a route with its own absorption profile for which no DSIP pharmacokinetic study was found. The amount: 100 to 300 mcg is five to fifteen times less than the 1.49 mg that 25 nmol/kg comes to for a 70 kg adult. And the schedule: the trials gave single doses, three afternoons, or four consecutive nights under polysomnography, while the circulating pattern is a nightly evening injection continued indefinitely. Peptifact's DSIP dosage page examines the provenance of those circulating figures in more detail.
What the figures omit
The trial figures carry populations — chronic insomniacs, healthy young men, surgical patients — and settings, a sleep laboratory or an operating theatre, with continuous monitoring. They carry a route that puts the whole dose into the bloodstream at once. They carry a literature that disagrees with itself about whether the sleep effect is real and a 2009 result that runs against the compound's name. What they do not carry is any subcutaneous data, any long-term safety data, any modern registered trial, and any approved product anywhere. The mechanism and the evidence file are covered in Medibact's DSIP guide; the vial arithmetic on its own, for any volume, is in the DSIP calculator.
What this page does not do
It does not state an amount of DSIP to use, by any route, or a frequency, or a course. It converts nanomoles per kilogram into micrograms using a published molecular weight, converts micrograms into syringe units at each concentration, and reports which study each figure came from and what that study concluded. Decisions about administering anything to a person belong with a clinician.
Sources and dates
Opened 2026-09-19: PubMed through the NCBI E-utilities for delta sleep-inducing peptide in title or abstract (407 records) and with the clinical-trial filter (17 records), with the abstracts of PMIDs 6112579, 6689058, 3622582, 2554357, 1299794, 7777652 and 19142086 read; PubChem compound record 68816 for the sequence and molecular weight; the DailyMed SPL search API for delta sleep-inducing peptide (zero records); the ClinicalTrials.gov v2 API for delta sleep-inducing peptide as an intervention (one record, unrelated); MedlinePlus, "Giving an insulin injection" (reviewed 2024-07-21). The circulating subcutaneous figures are as recorded in this site's dosage chart on 2026-09-18. Corrections go to the contact page.