What it is

DSIP, or emideltide, is a nine-amino-acid peptide (Trp-Ala-Gly-Gly-Asp-Ala-Ser-Gly-Glu) first found in rabbit cerebral venous blood in 1963 and synthesized in 1976 [1]. The Basel group that isolated it showed that intravenous DSIP produced sleep lasting hours in several animal species, and that immunoreactive DSIP-like material occurs in human plasma, urine, cerebrospinal fluid and milk [9]. In humans the endogenous level follows the sleep-wake circadian rhythm [1]. It is not an approved drug anywhere and has no USP or NF monograph [1]. Two pharmacies nominated it for 503A compounding as a subcutaneous injection at 1 mg/mL for opioid withdrawal, chronic insomnia and narcolepsy; both nominations were later withdrawn, but FDA chose to evaluate it anyway [1]. Compounded drugs are not FDA-approved [10].

How it works

FDA’s pharmacology review describes sleep-inducing properties that are conserved across species and mediated through opioid-dependent mechanisms: emideltide does not appear to bind opioid receptors directly, but it can stimulate calcium-dependent release of endorphins, which also accounts for its analgesic effects [1]. A rise in endogenous emideltide has been associated with suppression of both slow-wave and REM sleep [1]. The peptide crosses the blood-brain barrier [9]. In blood it is broken down quickly: multiple peptidases in human and rat blood cleave it from the N-terminal end, and the rapid disappearance of injected DSIP is due to degradation rather than protein binding [1][8]. Endogenous DSIP-like material, by contrast, is bound to a larger carrier protein that protects it from proteolysis [9].

What the evidence shows

Across the clinical studies FDA identified, intravenous emideltide at 25–150 nmol/kg was given to 209 subjects for 1 to 15 days; no study used the nominated subcutaneous route [1]. For chronic insomnia FDA judged the studies “inconclusive and at best preliminary,” with small samples and poor or absent control groups [1]. The largest double-blind trial, Bes et al. (1992), gave 16 chronic insomniacs either 25 nmol/kg intravenously or a glucose placebo on three afternoons; sleep efficiency and sleep latency improved slightly, but the authors judged the effects weak, partly explained by a change in the placebo group, and concluded that short-term DSIP treatment is not likely to be of major therapeutic benefit [7]. Narcolepsy evidence is a single case report, and opioid-withdrawal studies lacked blinding or randomization [1]. FDA-approved treatments exist for all three conditions [1].

Regulatory status

DSIP is not an FDA-approved drug [10]. In September 2023 FDA placed emideltide in 503A Category 2, the group it considers to carry significant safety risks; on Apr 15 2026 FDA said it would remove it from Category 2 because the nominations had been withdrawn, which is not a listing and does not place it in Category 1 [5]. FDA’s safety-risk page now lists emideltide as “nominated but withdrawn,” citing possible immunogenicity and no safety information for the proposed route [6]. FDA still brought it to the Pharmacy Compounding Advisory Committee on Jul 24 2026 and proposed that it not be included on the 503A bulks list [2][3]. The committee voted 6–7 with one abstention against recommending it; FDA staff said the peptide is not adequately characterized [4]. The vote is advisory; FDA decides [4]. As of Sep 24 2026 we found no final FDA action.

Half-life and what it means for a level curve

No primary human elimination study was located. FDA’s briefing states a plasma half-life of 8 minutes, degraded by aminopeptidases, but gives only a review as its source and notes there is no pharmacokinetic information for the subcutaneous route at all [1]. Animal values are similar: 3–6 min in dogs and 2–3 min in rats and a monkey after intravenous dosing, and 5–10 min when incubated in human serum [1][8]. STACK has no reference half-life for this compound because no human value has been published; you can enter your own estimate. Whatever value you use, a level reaches about 97% of steady state after roughly five half-lives; with a half-life measured in minutes, that means the estimated level returns to near zero within the hour after each dose. STACK’s curve is an estimate from this published half-life, not a blood level.

Reported side effects

Human safety data are limited to small intravenous studies. FDA searched its adverse event system through Mar 3 2024 and found zero reports for emideltide, and no case reports in the literature [1]. In the chronic-insomnia studies, intravenous emideltide was described as well tolerated without significant adverse events [1]. In a 1984 withdrawal study of 107 subjects, nine had minor transient effects (perspiration, headache, nausea, vertigo) and three had serious ones: two became hypotensive at the start of the first injection, one had repeated 15-minute episodes of discomfort, and one showed “progressive” hypotension after a second injection [1]. FDA concluded there are no safety data for the subcutaneous route, that safety for chronic intermittent use is insufficiently characterized, and that a nine-amino-acid peptide given by injection carries immunogenicity risk [1]. Call your prescriber about any reaction.

Storage and handling

FDA’s chemistry review reports that both emideltide free base and emideltide acetate are expected to be stable under storage below −20°C, that the free base has limited water solubility, and that peptides of this kind are sensitive to formulation and storage conditions that favor degradation products [1]. There is no approved label, so the compounding pharmacy’s label governs: follow its refrigeration and light-protection instructions and its beyond-use date. FDA does not verify the safety, effectiveness or quality of compounded drugs before they reach you [10]. If your product is a lyophilized powder, STACK’s reconstitution calculator handles the concentration math; it does not suggest what to take.

What STACK tracks for it

STACK records each DSIP dose as your prescriber wrote it, with the route, time of day and injection site. Because there is no published human half-life, the estimated-level curve is off unless you enter your own estimate. Sleep is the outcome most people care about, so STACK lets you log bedtime, wake time and a simple sleep-quality rating next to each dose, which you can export for your prescriber. Side-effect entries cover the reactions seen in studies: sweating, headache, nausea, dizziness and low blood pressure [1]. Vial inventory keeps the lot number and beyond-use date from the pharmacy label. Weight and other compounds in your plan sit on the same timeline.

The reference half-life in STACK

STACK ships DSIP without a default half-life: Only animal and in-vitro primaries; FDA cites 8 min without fetched primary. You enter your own estimate before a curve is drawn. Measured in: no primary human elimination study located; animal IV values 2–6 min.

DSIP

No default

No primary human half-life study located (animal intravenous values are 2–6 minutes); set your own estimate. Educational only — not medical or dosing advice.

Estimate only. Modelled from the published half-life you selected and the doses you logged — not a measured blood level. Do not use it to make medication decisions.

Free tools for DSIP

Sources

Every URL below was fetched and checked on 2026-09-24. Bracketed numbers in the text link here.

DSIP

What it is

How it works

What the evidence shows

Regulatory status

Half-life and what it means for a level curve

Reported side effects

Storage and handling

What STACK tracks for it

The reference half-life in STACK

DSIP

Free tools for DSIP

Read next

Sources