DSIP (emideltide): the monograph

DSIP is a chain of nine amino acids, Trp-Ala-Gly-Gly-Asp-Ala-Ser-Gly-Glu, with molecular formula C35H48N10O15 and a molecular weight of 848.8 ⁷. The initials stand for delta sleep-inducing peptide, the name its discoverers gave it in 1977 after infusing it into the brain ventricles of rabbits and watching the delta-wave and spindle activity of their EEGs increase ^12,13. Its international nonproprietary name, assigned in 1993, is emideltide, and that is the name under which FDA evaluated it in 2026 ³.

Those two names have almost nothing to do with each other in practice. The entire clinical literature exists under DSIP; exactly one indexed paper has ever used the word emideltide, and it is a 2020 chemistry methods study that borrows the peptide as a test molecule ⁹. A buyer searching the sold name and a regulator searching the official name are reading two nearly disjoint literatures about the same molecule ^9,3.

In the United States DSIP is not a medicine. No approved product contains it, no USP or NF monograph applies to it, FDA lists it among bulk substances that may present significant safety risks in compounding, and in July 2026 FDA proposed that it not be allowed into compounded drugs ^5,1,2. It reaches American buyers as a research chemical, mostly as vials for injection ³.

This monograph does something the name of this compound makes unusually necessary: it grades every claim by the species it was tested in and reads the human sleep-trial record at its actual design strength. The name promises delta sleep. The record, as the sections below document, could not reliably find it.

DSIP was isolated from rabbit blood in 1977 and named for a rabbit EEG effect ¹². It has been infused intravenously into about 209 people across the small clinical studies FDA could identify, most of them between 1981 and 1998 ³.

The indexed literature holds 419 records; their median publication year is 1992, and only 7 have appeared since 2016 ⁹. Five records carry a randomized-controlled-trial tag, and the five together enrolled 72 people ⁹.

The only double-blind parallel-group placebo-controlled insomnia trial (16 patients, 1992) concluded DSIP "is not likely to be of major therapeutic benefit" ²⁵. The most recent human study (24 patients, 2009) found DSIP "paradoxically, significantly reduced delta rhythm" ²⁷.

The peptide's own gene, precursor protein and receptor have never been found; a 2006 review calls the sleep-factor hypothesis "extremely poorly documented and still weak" ¹⁴.

FDA proposed in July 2026 that emideltide not be included on the 503A compounding list, citing characterization, effectiveness and safety gaps ^2,3. Trade press reported the advisory committee voted against it as well, alone among the seven peptides reviewed ¹¹.

There is no established human dose, no human pharmacokinetic study, no data for the sold subcutaneous route, no registered trial, and no published analysis of what commercial DSIP vials contain ^3,6,9.

Delta sleep-inducing peptide is the only compound in this fleet whose marketing copy is contained entirely inside its name. Vendors do not need to make a claim; the initials make it for them. So it matters where the name came from: it was coined in 1977 because infusing the peptide into rabbit brain ventricles "consistently increased the power of electroencephalographic (EEG) delta waves and spindles," as FDA's 2026 review puts it ^3,12. The name records a rabbit experiment. It was never a summary of human results.

Read against the human record, the name performs badly. In the only parallel-group double-blind placebo-controlled insomnia trial ever run, the authors concluded that short-term DSIP treatment "is not likely to be of major therapeutic benefit" ²⁵. In the crossover trial by an independent group, no significant differences from placebo were found and the authors judged the improvement "of little clinical significance" ²⁴. And in the most recent human study of the compound, a 2009 randomized anesthesia study, DSIP "paradoxically, significantly reduced delta rhythm" ²⁷. The peptide named for inducing delta sleep, in the last controlled look anyone took, reduced the delta rhythm it is named for.

The positive results in the record come almost entirely from one research line: the Basel-linked group of Schneider-Helmert and Schoenenberger, whose studies reported normalized sleep in small numbers of insomniacs but, as FDA documents, compared treated patients to external groups of healthy people rather than to a placebo arm ^23,22,3. FDA's one-line verdict on the whole file: "the results appear inconclusive and at best preliminary" ³.

Even the endogenous story behind the name is unresolved. The gene that would encode this peptide has never been found, no precursor protein has been isolated, and no receptor has been identified; the most recent dedicated review calls the sleep-factor hypothesis "extremely poorly documented and still weak" ¹⁴. This page and everything else on this site treat the name as what it is: a 49-year-old hypothesis that the human trial record failed to confirm.

DSIP (emideltide) is not FDA approved and no human efficacy has been established. In July 2026 FDA proposed not to allow it for pharmacy compounding. That is this site's required status line, and each half of it is independently checkable: Drugs@FDA returns no product containing emideltide under any name ⁵, and FDA's July 2026 briefing proposed excluding both the free base and the acetate from the 503A Bulks List ².

FDA's compounding safety-risks page carries this entry for Emideltide (DSIP), in full: "Compounded drugs containing emideltide may pose risk for immunogenicity for certain routes of administration and may have complexities with regard to peptide-related impurities and API characterization. FDA has not identified safety-related information regarding emideltide for the proposed route of administration. Therefore, the agency lacks sufficient information to know whether the drug would cause harm if administered to humans." ¹

No USP or NF monograph exists for either form ³. FDA's historical-use review found evidence of compounding use only since about 2018, with injection and intranasal products reaching buyers through integrative neurology clinics, medical concierge services, medical spas, wellness clinics, and online retailers ³.

None of this is a scheduling question: DSIP is not a controlled substance, and this site makes no claim about the legality of possession. The regulatory fact that matters to a buyer is simpler: nothing about the vial in the mail has undergone any regulatory review ^1,3.

On July 24, 2026, the Pharmacy Compounding Advisory Committee took up emideltide, one of seven peptides reviewed across the two-day meeting. FDA's briefing document states its position as points 9 and 10, verbatim: "FDA is proposing that Emideltide (free base) NOT be included on the 503A Bulks List" and "FDA is proposing that Emideltide acetate NOT be included on the 503A Bulks List" ².

The strange procedural detail: nobody was still asking. Both nominations, from LDT Health Solutions on behalf of the International Peptide Society and from Wells Pharmacy Network, had been withdrawn before the meeting, and FDA elected to present emideltide anyway, evaluating it on its own initiative ^2,3.

The evaluation behind the proposal is the most thorough document ever written about this compound: a roughly 100-page review signed by eight named FDA reviewers, covering characterization, historical use, every clinical study, and safety ³. Its conclusion, verbatim: "These substances are not well characterized from a physical and chemical characterization perspective, and endotoxin testing for injectable ROA is lacking," and, on the clinical record, "there is insufficient information concerning effectiveness to support the use" for chronic insomnia, narcolepsy, or opioid withdrawal. It ends: "Accordingly, we propose not adding emideltide (free base) or emideltide acetate to the 503A Bulks List" ³.

What the committee then did is, as of this writing, documented only in trade press: Regulatory Focus (RAPS) reported that "the committee voted against recommending emideltide for inclusion on the 503A Bulks List in a narrow 6-7 vote with one abstention," while voting in favor of the other six peptides reviewed, BPC-157, KPV, TB-500, MOTS-c, epitalon and semax ¹¹. By that report, DSIP was the only compound at the meeting where the committee and FDA agreed, and the only one of the seven rejected. FDA has not posted vote tallies, so this site handles the vote as press reporting, not as an FDA record ¹¹.

None of it is final. A committee recommendation is advisory, FDA states it will not issue a final determination until all reviews are complete, and the rulemaking is pending ³. What is already true regardless of outcome: even inclusion on the 503A list would not have made DSIP an FDA-approved drug; it would only have defined what compounding pharmacies may lawfully use ³.

Every peptide hormone you have heard of has a gene, a precursor protein and a receptor. DSIP, 49 years after its isolation, has none of the three. The 2006 Journal of Neurochemistry review by Kovalzon and Strekalova states it plainly: "the link between DSIP and sleep has never been further characterized, in part because of the lack of isolation of the DSIP gene, protein and possible related receptor. Thus the hypothesis regarding DSIP as a sleep factor is extremely poorly documented and still weak" ¹⁴.

The same review notes that DSIP's structure "is different from any other known representative of the various peptide families," that its "natural occurrence and biological activity still remains obscure," and hypothesizes that unidentified DSIP-like peptides, not DSIP itself, may be responsible for the immunoreactivity measured in tissues for two decades ¹⁴. In other words, the field's own specialists are not certain the endogenous substance the assays detected was ever this exact nonapeptide.

The measurement record fits that doubt. A 1995 study measuring DSIP-like immunoreactivity in the blood of narcolepsy patients, sleep apnea patients and healthy controls found no significant differences between the groups ³⁴. And the synthetic peptide itself is unstable in blood, degrading rapidly in vitro, which the investigators studied precisely because inactivation rates might explain its "variable effects in vivo" ⁴⁰.

None of this proves DSIP does nothing; the rabbit and rat findings are real experiments with real EEG changes ^12,38. It means something stranger: the compound sold in vials as a sleep hormone has never been confirmed to be a human hormone at all.

This site ran one census query against Europe PMC and archived every record it returned: title or abstract containing the full name delta sleep-inducing peptide (any hyphenation) or emideltide. The result is 419 records ⁹. A separate acronym-only query adds 63 records that mix genuine peptide papers with false positives from other fields; they are archived and disclosed, not counted ¹⁰.

The shape of the corpus is the finding. Median publication year: 1992. Records published before 1996: 272 of 419, or 64.9 percent. The most productive decade was the 1980s with 152 records; the 2020s have produced 4. Since 2016, the entire world has published 7 indexed papers about this peptide ⁹.

Within those 419 records, exactly 5 carry the Randomized Controlled Trial publication tag, and the five together enrolled 72 people: two insomnia trials (6 and 16 patients), two endocrine-mechanism studies (11 and 15 men), and one anesthesia study (24 patients) ^{9,18,25,31,32,27}. The corpus contains zero meta-analyses and zero records tagged as systematic reviews ⁹.

The language split matters less than for our sibling compounds but exists: 290 records are English and 115 are Russian, reflecting a substantial Russian-sphere research line including the Deltaran preparation literature ^9,42.

One more census fact worth sitting with: the name FDA evaluated this substance under, emideltide, appears in exactly one of the 419 records, a 2020 protein cross-linker methods paper ⁹. The regulatory record and the scientific record use different names for the same molecule, and buyers searching either one will never see the whole picture. This page is built to show both.

In rabbits, the founding species, intraventricular DSIP enhanced delta-wave and spindle EEG activity under double-blind conditions across 58 animals in the 1977 characterization, and the synthetic peptide matched the isolated original ^12,13. In rats, an early study reported increased sleep duration ³⁵, and the strongest mechanistic result is indirect: antiserum against DSIP, injected into the third ventricle, blocked the slow-wave-sleep and growth-hormone rises that normally follow sleep deprivation ³⁸.

In cats the story runs backward. Intraperitoneal DSIP reduced sleep, cut REM sleep and delayed REM onset in one study ³⁶, while an intracerebroventricular study in sleep-deprived cats found no change in slow-wave sleep, paradoxical sleep or total sleep time at all, only a shift between sleep stages ³⁷. FDA's nonclinical reviewers credit the peptide with sleep-inducing properties "conserved across species" mediated by opioid-dependent endorphin release ³; the cat literature is a standing asterisk on that sentence.

In humans, the subject of the next three sections, roughly 209 people have received intravenous DSIP in identified studies, in doses of 25 to 150 nmol/kg for 1 to 15 days ³. What those studies found, and what they could not find, is the heart of this monograph.

The insomnia record is seven small studies, 1981 to 1992, totaling under 80 patients, split cleanly by design. The studies with placebo controls found little; the studies without them found normalization ³.

The controlled side. The 1981 Basel crossover (6 insomniacs, randomized, double-blind): sleep-promoting effects in hours two through six of the night, but no effect on falling asleep, and results FDA judged impossible to interpret from the reported data ^18,3. The 1987 Montevideo crossover by an independent group (6 insomniacs, randomized, double-blind): "no significant differences ... in comparison to baseline or to double-blind placebo nights," improvement "of little clinical significance" ²⁴. The 1992 Amsterdam parallel-group trial, the only one of its design ever run (16 patients): weak effects the authors thought partly attributable to "an incidental change in the placebo group," and the conclusion that DSIP "is not likely to be of major therapeutic benefit" ²⁵.

The uncontrolled side. Schneider-Helmert's 1986 study (18 insomniacs) and 1987 study (14 insomniacs) reported sleep restored to normal values, with sleep-onset latency halved and total sleep time up more than an hour ^22,23. But FDA's review documents the flaw both share: they were described as placebo-controlled, and neither had a placebo arm. Treated insomniacs were compared to external groups of healthy sleepers, a design FDA calls "challenging to interpret due to inherent limitations" ³. Kaeser's 1984 open series (7 patients, 10 injections, no control, no blinding) reported sleep normalized in six of seven for months ²⁶. Open series produce results like that for many things that later fail controlled trials; that is why the grade column on this site exists.

FDA's synthesis of all of it, verbatim: "Although some studies showed that short-term IV emideltide appeared to exert an effect on disturbed sleep, the results appear inconclusive and at best preliminary" ³. The 1986 review from Graf and Kastin, the field's own chroniclers, had already said the same: FDA records their finding that the results were preliminary and there was insufficient evidence that DSIP reliably induced or maintained sleep ^17,3. Nothing published in the four decades since has changed either sentence.

The DSIP literature is not a case of good early results that failed to replicate. It is a case of results that disagreed from the beginning, inside and outside the founding laboratory, in patterns the field itself documented.

The dose-response was strange from the start. Graf and Kastin's 1984 review describes "a U-shaped activity curve ... for the dose as well as for the time of infusion" ¹⁶. In the founding group's own healthy-volunteer work, the identical 25 nmol/kg dose increased total sleep time by 30 minutes when infused over 2.5 minutes, by an hour when infused over 7.5 minutes, and did nothing when infused over 1 minute ^3,20. A drug whose effect appears and disappears with infusion speed at the same dose is a drug whose trials will disagree with each other, and they did.

The animal record disagreed too: increased sleep in rats, reduced sleep in cats in one study, no sleep change in cats in another ^35,36,37. The peptide degrades in minutes in blood, and the investigators who measured that flagged it explicitly as a candidate explanation for the "variable effects in vivo" ^40,39.

FDA's summary names the other driver: the positive human studies "had poor study methodologies (i.e., lack of or poor selection of control groups)," small samples, differing baselines and regimens, and no long-term data ³. Put simply: where the design could not manufacture an effect, the effect mostly was not there, and where it was reported, the design could not rule out that it was manufactured. Both halves of that sentence are documented, and neither supports buying a vial.

The second life of DSIP was as a withdrawal treatment, and it is the indication FDA led with in 2026: the uses evaluated were "opioid withdrawal, chronic insomnia, and narcolepsy" ⁴. The hypothesis was mechanistic: DSIP appeared to act on opioid systems indirectly, stimulating endorphin release without binding opioid receptors ^3,28.

The clinical record is three open-label, uncontrolled studies. Geneva, 1983: 67 inpatients in alcohol or opiate withdrawal, DSIP as sole treatment; of 49 evaluable patients the authors reported benefit in 48, with rapid onset ²⁸. Geneva, 1984: 107 inpatients; the authors claimed marked rapid improvement in 97 percent of opiate and 87 percent of alcohol patients, while conceding that "nothing allows to draw any conclusion as to a potential usefulness of a maintenance treatment," and noting relapse of anxiety and insomnia within 24 to 72 hours of stopping in many patients ²⁹. Munich, 1998: 7 opioid-dependent subjects on a fixed schedule; withdrawal scores halved after the first injection, but only 2 of 7 completed the protocol and detoxified, and later injections mostly failed to suppress recurring symptoms ^30,3.

No blinding, no randomization, no control group exists anywhere in that record, and FDA's conclusion reflects it: results "should be interpreted cautiously due to small number of published studies, small sample sizes, different treatment regimen and several limitations owing to nonrandomized and uncontrolled study design" ³.

FDA's review adds a warning unique to this indication: because emideltide stimulates the opioid system, it "could lead to development of addiction," and no study of its addictive potential exists ³. The withdrawal chapter also supplied the only serious adverse events in the whole record, described in the side-effects section below ³.

The entire narcolepsy literature for DSIP is a single 1984 case report: one 35-year-old man, a week each of evening and morning intravenous dosing, self-reports and sleep-lab measures, fewer sleep attacks and more daytime alertness reported ²¹.

FDA's review of that report is a catalog of why single cases cannot carry an indication: "insufficient methodological rigor, potential recall bias, researcher subjectivity and issues of reliability and external validity" ³. No second narcolepsy patient has ever been published, and measured blood DSIP-like immunoreactivity does not distinguish narcolepsy patients from controls ³⁴.

Narcolepsy is a serious, chronic condition with FDA-approved treatments of established efficacy; the practice guidelines for it do not mention this peptide ³.

Small randomized studies probed DSIP's endocrine effects in healthy men: a 1989 crossover in 11 volunteers found reduced plasma ACTH-like immunoreactivity for at least three hours without cortisol change ³¹; a 1994 study in 15 men found no effect on vasopressin ³²; a 1995 study in 20 men found no effect on CRH- or meal-stimulated ACTH and cortisol, undercutting the proposed corticotropin-inhibiting role ³³. Mechanism studies, not benefits: none measured a health outcome.

The 2009 Manchester anesthesia study deserves its own sentence because it is the last human data anyone collected: 24 surgical patients, randomized, DSIP added to isoflurane anaesthesia on the hypothesis that a natural hypnotic would deepen it. It did the opposite: heart rate rose, heart-rate variability fell, and delta rhythm, the peptide's namesake, was "paradoxically, significantly reduced" while measured anaesthetic depth lightened ²⁷. The first human study of DSIP increased sleep; the last one lightened anaesthesia. The record closes on the inversion of its own premise ^19,27.

In Russia, a preparation called Deltaran, described in the literature as emideltide free base with glycine, generated a research line of its own, seven records in our corpus, mostly Russian-language mouse work on stress protection and aging; in one study it reduced spontaneous tumor incidence in female SHR mice, a finding FDA declined to read as reassurance because glycine itself has antimutagenic and anticarcinogenic reports ^41,42,3. This site could not verify any marketing authorization for Deltaran and asserts none.

Scattered animal work reports analgesic effects in mice, seizure modulation, and stress-protection endpoints; all of it is preclinical, none of it has a human counterpart, and FDA identified no nonclinical toxicology program at all ³. An early healthy-volunteer finding is worth preserving for honesty in both directions: the 1981 first-in-human study reported a 59 percent increase in total sleep time in the two hours after morning infusion, without classic sedation, and delayed effects on the following night ¹⁹. It was never followed by a confirmatory program.

There is no human pharmacokinetic profile of DSIP to summarize: no published human half-life, no bioavailability figure for any route, no Cmax, no exposure curve ^9,3. What exists is animal and in vitro work, and it points one direction: this peptide barely survives in blood.

In dogs, IV DSIP disappeared with a mean half-life of 4.0 +/- 0.7 minutes and a metabolic clearance rate of 30.7 +/- 2.5 mL/kg per minute ³⁹. In vitro, DSIP degrades rapidly in human and rat blood, while iodinated analogs formed aggregates; the authors proposed inactivation kinetics as an explanation for the compound's variable in vivo effects ⁴⁰. Older dog work reported blood-brain-barrier passage of labeled peptide, a finding the 2001 anaesthesiology review summarizes ¹⁵.

The practical meaning is blunt. Every trial that reported effects used slow intravenous infusion under supervision, with effect strength dependent on infusion rate ³. The products sold today are subcutaneous or intranasal, routes for which FDA identified no clinical data of any kind ³. A four-minute half-life measured in dogs is not a human number, but it is the only number, and it is not an encouraging one for a peptide sold to be injected under the skin once before bed.

In the clinical literature DSIP was given one way: slow intravenous infusion, typically 25 nmol/kg (about 21 micrograms per kilogram), in supervised settings ³. The 1983 dose-response work found the effect depended on infusion lasting several minutes; a one-minute push did nothing ^20,3.

What is sold today is different: lyophilized vials marketed for subcutaneous injection, and intranasal products, offered through clinics, spas and online retailers ³. For the subcutaneous route, FDA "did not identify any data to support the effectiveness" for any evaluated use, and no clinical safety data exist either ³. For intranasal use the indexed record is silent entirely.

The nominated compounded product was a 1,000 mcg/mL injectable solution, and FDA's chemists flagged a problem with the vial itself: emideltide free base dissolves in water at only about 0.5 mg/mL, so it is "unclear how it would be possible to formulate the proposed injectable dosage form with concentration of 1,000 mcg/ml without co-solvent used" ³. A route nobody studied, at a concentration that may not dissolve: that is the product as actually sold ³.

Across every clinical study FDA identified, intravenous doses of 25 to 150 nmol/kg were given to a total of 209 subjects for 1 to 15 days ³. The workhorse regimen was 25 nmol/kg, about 21 micrograms per kilogram, infused over roughly 4 minutes; the withdrawal studies used up to six injections a day for up to six days, and the 1998 Munich study used 35 nmol/kg on a fixed taper ^3,30.

None of this translates to the products sold. The trial route was intravenous; the sold route is subcutaneous, with no data ³. The trial effect depended on infusion rate, something a subcutaneous injection cannot reproduce ^3,20. And the dose-response was U-shaped, meaning more was not better even in the original laboratory ¹⁶.

Our study-dose explorer lists every regimen in the record attached to the study that used it, its design, and what it actually found, including the null results. It computes nothing that resembles a recommendation, because there is no established human dose of DSIP and the literature does not contain the information a dose would require ³.

Has anyone run a dedicated human safety study? No. FDA identified none, and also no nonclinical toxicology program: "the nominator did not submit, and FDA did not identify nonclinical toxicity studies to inform safety considerations" ³.

Is there long-term human data? No, at any dose, by any route ³. The longest exposures in the record are 15 days ³.

What does FDA actually say? That it "lacks sufficient information to know whether the drug would cause harm if administered to humans" ¹. That a nine-amino-acid peptide injected subcutaneously "may pose a significant risk for immunogenicity, potentially amplified by aggregation as well as potential peptide-related impurities," and that nobody has provided information suggesting emideltide does not present these risks ³.

Were the trials reassuring? Within their tiny size and short duration, the insomnia studies reported good tolerance ^3,19. The withdrawal studies did not: they produced the record's serious adverse events, including progressive hypotension, detailed below ³.

Is there an addiction concern? FDA raises one on mechanism: emideltide stimulates the opioid system indirectly, "could lead to development of addiction," and no study of its addictive potential exists ³. That is a hypothesis flagged by a regulator, not an observed harm; on this compound, both halves of that sentence deserve equal weight.

In the insomnia and healthy-volunteer studies, no significant adverse effects were reported; the 1981 first-in-human study called the compound well tolerated ^3,19. Those studies totaled a few dozen people dosed for days.

The withdrawal series carried the real signal. In the 1984 Geneva study of 107 patients, nine had transient side effects (perspiration, headaches, nausea, vertigo) and three had serious adverse effects: two with hypotension at the start of the first injection, and one who received a second injection and experienced "progressive hypotension" ^3,29. The study authors attributed some side effects to "solubility difficulties" of some vials, which had come from Hoffman-La Roche; FDA notes it is unclear what that meant, and that no other publication reported it ³.

FDA's FAERS database contains zero adverse event reports for emideltide through March 3, 2024 ³. On an unapproved research chemical, that is a fact about reporting channels, not about safety: there is no product label, no prescriber, and no manufacturer obligated to file ^3,1.

No drug-interaction study of DSIP exists in the indexed record, and FDA's evaluation identifies none ^3,9. The mechanistic overlap FDA flags, indirect stimulation of opioid signaling, sits uncomfortably close to the very drugs the withdrawal studies combined it with, and nobody has characterized that interaction ³.

The trial populations were narrow: middle-aged insomniacs, healthy young men, inpatients in withdrawal, one narcoleptic, and 24 surgical patients ³. Never studied at all: anyone pregnant or breastfeeding, children, the elderly beyond one 9-person subgroup, anyone with cardiovascular disease (a pointed gap given the hypotension cases and the 2009 heart-rate findings), and anyone taking sleep medications, the population most likely to buy it ^3,27,22.

The 2009 anesthesia study is the one direct caution in the record about combining DSIP with CNS depressants: added to isoflurane, it raised heart rate, reduced heart-rate variability and lightened anaesthesia, none of which was predicted ²⁷.

Nobody has published an answer. Our census of the entire 419-record literature finds zero analyses of commercial DSIP products: no purity survey, no identity testing of sold vials, nothing ⁹. Every statement on a vendor certificate of analysis is, as far as the indexed scientific record is concerned, unaudited ⁹.

What the record does contain is FDA's dissection of the one CoA offered in the nomination process: it "lacked specific tests (including impurities, aggregates, and endotoxins)"; for the free base, no bioburden or endotoxin test and no residual solvent testing appeared at all ³. Endotoxin testing is how you find out whether an injectable will cause a febrile reaction; it was simply absent ³.

Two details bracket the problem historically. In 1984, even trial vials supplied by Hoffman-La Roche had "solubility difficulties" that study authors blamed for side effects ^3,29. In 2026, FDA noted that emideltide free base dissolves at only about 0.5 mg/mL in water, making the 1 mg/mL concentration on sellers' labels chemically questionable without a co-solvent nobody discloses ³. If pharmaceutical vials struggled in 1984 and the advertised concentration may not dissolve in 2026, the gray vial in the mail sits somewhere between those two facts, unexamined ³.

FDA's characterization verdict covers the substance itself, not just the paperwork: emideltide is "not well characterized from a physical and chemical characterization perspective," with naming inconsistencies that create a risk of buyers receiving a different substance than intended ^3,1.

FDA's own market survey is the best description available: emideltide injection and intranasal products reach US buyers through "integrative neurology clinics, medical concierge services, medical spas, wellness clinics, and online retailers," with evidence of compounding use only since about 2018, and it is unclear whether the products sold are compounded at all ³.

No approved manufacturer exists anywhere in the chain: no approved product, no USP monograph to test against, and, if FDA's proposal is finalized, no lawful path through compounding pharmacies either ^3,5. What remains is the research-chemical channel, whose vials no published study has ever analyzed ⁹.

This site sells nothing and recommends no vendor. The sourcing fact a buyer most needs is the one in the section above: for this compound, the contents of the vial are an open question at every level, from identity to endotoxins ³.

The stability facts in the record: emideltide free base is a white to off-white lyophilized powder, reported stable stored below -20 C, with water solubility around 0.5 mg/mL ³. In blood it degrades in minutes, and analogs aggregate ^40,39.

FDA additionally notes that for peptides like emideltide, "the stability, pharmacological activity, and immunogenic properties ... are highly sensitive to the manufacturing process and quality attributes of the compounded/finished drug product" ³. Storage advice on vendor sites assumes a verified starting material, which is the thing nobody has verified ^9,3.

See the evidence-limits block on this page for the structured version. The five defining gaps: (1) no reliable efficacy evidence for any use, with the only parallel-group placebo-controlled trial negative; (2) no human pharmacokinetics for any route and zero data for the routes actually sold; (3) no dedicated human safety study, no toxicology program, no long-term data, and an unstudied addiction-potential question raised by FDA; (4) no analysis of what commercial vials contain, against a solubility ceiling that makes the advertised concentration questionable; (5) no gene, precursor or receptor for the peptide itself, leaving even its biological identity unresolved. Each is documented in the sections above with its source ^3,25,9,14,6.

Citations and FAQ render from the numbered manifests on this page.

Citations and FAQ render from the numbered manifests on this page.