DSIP — delta sleep-inducing peptide, international name emideltide — is a nine-amino-acid peptide first described in 1977 from the blood of rabbits and studied as a sleep aid and a withdrawal treatment through the 1980s. It was never approved anywhere. In 2026 it was nominated for US compounding, which meant FDA's reviewers had to read every human safety record they could find. Their briefing document is the most complete account of DSIP's side effects that exists, and it is short. This page reports what it found, adds the numbers it implies, and states the gaps. What doses the old studies used, and why they will not convert to a vial, is on our DSIP dosage page.
Where the record comes from
The source is FDA's briefing document on emideltide-related bulk drug substances (free base and acetate) for the Pharmacy Compounding Advisory Committee meeting of July 23–24, 2026 (FDA briefing document, evaluation dated 2026-05-11). Its reviewers searched PubMed, Embase, the Cochrane database, FAERS and ClinicalTrials.gov, plus the nominator's submissions.
What they found, in one table:
| What FDA looked for | What it found |
|---|---|
| Human studies, any route | IV emideltide, 25–150 nmol/kg, given to 209 subjects for 1 to 15 days |
| Human studies, subcutaneous (the nominated route) | None |
| FAERS adverse-event reports, through 2024-03-03 | Zero |
| Published case reports | None relevant |
| Pharmacokinetics | IV only; plasma half-life about 8 minutes |
| Acute, repeat-dose, genotoxicity, carcinogenicity, reproductive toxicity studies | None submitted or found |
Our own checks on 2026-10-09 agree: ClinicalTrials.gov returned no study with emideltide as an intervention (a search for the full name returned one unrelated L-carnitine study), and an openFDA adverse-event query for emideltide returned no matches.
The insomnia studies: no significant adverse events
FDA summarises the insomnia studies as finding intravenous emideltide "well tolerated and not associated with significant AEs". They were small — the first-in-human study gave it to six healthy middle-aged volunteers — and short. One detail from that study is a reported effect rather than a side effect: five of the six described a "feeling of pressure to sleep" immediately after the infusion, which had gone two hours later without measured sleepiness or loss of vigilance. The longest human exposure in the sleep work was a few consecutive nights.
The withdrawal study: where the adverse events are
Almost everything recorded as a side effect comes from one paper: Dick and colleagues, 1984, who gave intravenous DSIP to 107 inpatients in alcohol (47) or opiate (60) withdrawal (PMID 6548969), with up to six injections a day for up to six days. The paper's own abstract says tolerance was good "aside from headaches reported by a few patients". FDA's reading of the full paper is more specific:
| Outcome, as counted in FDA's briefing | Patients | Share of 107 |
|---|---|---|
| No significant adverse event | 95 | 88.8% |
| Minor, transient effects — perspiration, headache, nausea, vertigo | 9 | 8.4% |
| Serious reaction | 3 | 2.8% |
| — hypotension at the start of the first injection | 2 | |
| — repeated 15-minute episodes of discomfort with sweating and nausea | 1 |
The percentages are our arithmetic from FDA's counts. One of the patients with a serious reaction received a second injection and developed what the paper called "progressive hypotension"; FDA notes there is no further information about that patient. Some opiate patients who responded relapsed into anxiety with marked insomnia 24–72 hours after treatment stopped. And the authors attributed some side effects to "solubility difficulties" with some of the vials, which came from Hoffmann-La Roche — a remark FDA says it cannot interpret.
FDA's caution about this table is the right one to repeat: sweating, nausea, headache and blood-pressure swings are also symptoms of withdrawal itself, so the adverse effects "may have been confounded" by the condition being treated. There was no placebo group to separate them.
What FDA flagged beyond the trials
- No toxicity testing of any kind. FDA found no acute, repeat-dose, genotoxicity, carcinogenicity or reproductive studies. The one Russian finding that a DSIP-containing product (Deltaran) reduced chromosome damage and tumours in mice could not be read as evidence of safety, FDA wrote, because the product also contained glycine, which has its own reported anti-mutagenic effects.
- Heart rate in rats. In the one acute animal study FDA found (Yehuda 1988), a single injection did not change blood pressure but lowered heart rate by about 3% in daytime and about 10% at night, and shifted the rats' temperature rhythm. FDA states it is unknown whether this relates to sleep induction.
- Immunogenicity and aggregation. A nine-amino-acid peptide injected under the skin can provoke antibodies, and peptides can clump into aggregates that make that more likely. FDA found nothing showing emideltide does not carry these risks.
- Characterisation and endotoxin. FDA found the substances poorly characterised, with no endotoxin testing for an injectable form — and limited water solubility that made the nominated 1,000 mcg/mL injection hard to formulate without a co-solvent.
FDA proposed not adding emideltide to the 503A bulks list. At the meeting on July 24, 2026 the committee voted 6 to 7 against, with one abstention, according to trade-press reports — the only one of the seven peptides reviewed that it did not recommend. How that list works, and where the other six stand, is on our 503A category 2 page.
What the record does not contain
- Any human data for subcutaneous DSIP — the route nominated for compounding and the way research vials are used. The 209 people in the record all received it into a vein.
- Any exposure beyond 15 days.
- Any controlled comparison in the study that produced the adverse events.
- Anything about the vials sold as research chemicals, which carry none of the controls FDA found missing even for the pharmacy-grade substance — see what research-use-only labelling means.
The 1993 endocrine paper discussed on the dosage page described the dose range studied as one "known to modify ECG patterns"; nothing since has measured that. An empty FAERS record is not a safety finding — our peptide side-effects overview explains why. Anyone connecting a symptom to DSIP, or weighing its use alongside sleep or blood-pressure medicines, should take the question to a clinician. Peptide Lexicon's DSIP entry covers what the peptide is.
Sources and dates
- FDA. FDA Briefing Document, Pharmacy Compounding Advisory Committee, July 23–24, 2026: Emideltide-related bulk drug substances (emideltide free base and emideltide acetate). Evaluation dated 2026-05-11. https://www.fda.gov/media/193344/download — sections II.D.1 (nonclinical) and II.D.2 (human safety) and the conclusions. Read 2026-10-09.
- Dick P, Costa C, Fayolle K, Grandjean ME, Khoshbeen A, Tissot R. DSIP in the treatment of withdrawal syndromes from alcohol and opiates. Eur Neurol. 1984;23(5):364-71. PMID 6548969.
- Schneider-Helmert D et al. First-in-human study of DSIP in six healthy volunteers, 1981, as summarised in FDA's briefing.
- ClinicalTrials.gov API v2, intervention "emideltide" (0 studies) and "delta sleep-inducing peptide" (1 unrelated study), queried 2026-10-09.
- openFDA drug adverse event endpoint, medicinal product "emideltide": no matches, queried 2026-10-09 (dataset updated 2026-07-30).
- Committee vote: trade-press reports of the July 24, 2026 meeting (American Med Spa Association, 2026-07-24), read 2026-10-09; FDA's official minutes were not yet posted.