DSIP: What the Evidence Actually Shows

Search "DSIP peptide" and it's usually pitched as the body's own natural sleep hormone — a gentle way to deepen sleep and blunt stress, sometimes with a claim that it's been "studied since the 1970s." That last part is true, which already makes DSIP an unusual case in this article series: it has real, decades-old human trial data, not just rodent studies. What most product pages don't mention is what those trials actually found when a real placebo group was involved, or that DSIP's own status as a discrete, well-characterized hormone has been directly questioned by the researchers who worked on it. This article traces both threads — the human evidence and the identity question — to their actual sources, plus the current FDA record.

What DSIP actually is — and why its own identity is contested

"DSIP" stands for delta sleep-inducing peptide, and it has an unusually specific origin story for something now sold in vials online: it was purified from rabbit cerebral venous blood in 1977 by a Basel research group, sequenced as a nine-amino-acid peptide (molecular weight 849), and named for the slow-wave ("delta") sleep it appeared to promote in the animals it was first tested in³. That much is accurate marketing shorthand.

What most product pages don't volunteer is that DSIP's status as an actual, discrete hormone has been directly and publicly questioned in the peer-reviewed literature — not by an outside skeptic, but by researchers working in the field. A 2006 review titled, in its own words, "Delta sleep-inducing peptide (DSIP): a still unresolved riddle" lays out the problem plainly: nearly 30 years after DSIP's isolation, no one had identified the gene that encodes it, the larger precursor protein it would be cleaved from, or a receptor it binds to. The review's own assessment of the sleep-hormone hypothesis that gave DSIP its name: "extremely poorly documented and still weak"¹. Its authors go further, proposing that the immunoassay used to detect "DSIP" in blood and tissue — which measures "DSIP-like immunoreactivity," not DSIP itself — may actually be picking up a related but different, still-unidentified peptide, meaning some published DSIP research may not be about the exact molecule its name implies. A 2011 follow-up confirms the precursor-gene gap was still open five years later: it proposes a specific human gene family (JMJD1B-type histone demethylases) as a "putative precursor of endogenous DSIP" — explicitly a hypothesis, tested only in a preliminary way, not a confirmed origin². No more recent identity-specific review was found in a live search of the literature. A compound whose own biosynthetic origin remains this unsettled, two decades after its heaviest research period, is not standard footing for a confidently marketed sleep hormone.

What the animal evidence actually shows

The most-cited English-language DSIP review describes the peptide as inducing "mainly delta-sleep in rabbits, rats, mice, and humans," with a different, more REM-weighted effect in cats, and a "U-shaped" dose-response curve where both very low and very high doses were reportedly less effective than doses in between³. Beyond sleep, the same review catalogs a wide, almost scattershot list of other reported effects across species — changes to brain electrophysiology, neurotransmitter levels, circadian and locomotor patterns, hormone levels, and the behavioral effects of other drugs including their withdrawal. That breadth is itself worth flagging: a peptide credibly reported to touch that many separate systems, without a confirmed receptor to mechanistically anchor any single one of them, is a harder case to interpret cleanly than a peptide with one specific, well-mapped effect.

What the human evidence actually shows

DSIP does have a genuine, decades-old human research record — this isn't a compound invented for the supplement market with zero clinical history. But reading the actual trials, rather than the marketing gloss, tells a more mixed story than "clinically studied for sleep" implies.

The best-controlled human sleep study is a 1992 double-blind, placebo-controlled trial in 16 chronic insomnia patients, given intravenous DSIP or a glucose placebo before three consecutive nights in a sleep laboratory, with outcomes tracked by polysomnography. DSIP did produce higher measured sleep efficiency and shorter sleep latency than placebo — but the authors' own conclusion is blunt: the effects were "weak" and possibly confounded by an unrelated shift in the placebo group's own data, no other measure (including subjective sleep quality) changed, and short-term DSIP treatment of chronic insomnia "is not likely to be of major therapeutic benefit"⁴. A separate 1987 double-blind crossover study in insomniacs found a similar pattern: total sleep time and stage-2 sleep increased under DSIP versus placebo, but that same difference already existed at baseline before treatment started, and the authors concluded sleep improvement under DSIP was "of little clinical significance"⁵.

Two much smaller, uncontrolled reports point somewhere more interesting, with far weaker methodology behind them. A single-patient case study of a 35-year-old man with narcolepsy found that repeated DSIP injections reduced sleep-attack frequency and increased daytime alertness, by both self-report and polysomnography — a real finding, but from one person, with no placebo arm⁶. A 1984 pilot study gave DSIP to seven patients with chronic migraine, vasomotor headache, tinnitus, or psychogenic pain, comparing symptom levels before and after treatment rather than against a separate placebo group: pain fell in six of seven patients, alongside a reduction in co-occurring depressive symptoms⁷. That's a real, published clinical observation — and also exactly the small, open-label, before-after design that regularly fails to replicate once a proper placebo control is added, which is what happened to DSIP's own insomnia results above.

DSIP's record also includes a specific 1998 report on opioid detoxification — "results of an open clinical trial," by the paper's own title, meaning unblinded and without a placebo comparison group⁸. The indexed record for that study doesn't carry the kind of detailed outcome data that would let a reader independently size the effect; what's independently confirmable is that FDA's own July 2026 committee review later listed "opioid withdrawal" as one of DSIP's three evaluated nominated uses, alongside chronic insomnia and narcolepsy.

Two other human studies are worth flagging because they're negative findings marketing tends to skip. A 1993 double-blind study found intravenous DSIP had no effect on growth hormone or prolactin secretion in eight healthy women, even at doses known to alter ECG patterns⁹. And a 1995 study comparing plasma DSIP-like immunoreactivity across sleep apnea patients, narcolepsy patients, and healthy controls found no significant differences between the groups — undermining any claim that DSIP levels reliably track with either condition¹⁰.

Put together, DSIP's human record is real but not what "clinically studied for sleep" usually implies when a product page uses that phrase: the two best-designed trials, run against an actual placebo, came back negative or clinically insignificant on precisely the outcome — chronic insomnia — DSIP is most often marketed for today.

The current FDA and compounding status

DSIP has never had an FDA-approved drug product. A direct, live check against DailyMed, the National Library of Medicine's official archive of FDA-approved drug labeling, returns zero results for both "DSIP" and "emideltide" — the formal name FDA's own nomination and committee materials use for the identical molecule.

Its compounding status moved twice in 2026, and the second move did not go DSIP's way. FDA placed DSIP/Emideltide in Category 2 of its interim 503A bulk-substances list — substances presenting significant safety risks that pharmacies should not compound — in 2023, then removed it from Category 2 on April 15, 2026 (effective April 22), alongside eleven other peptides, because the nomination behind that listing was withdrawn. That removal was a procedural change, not a safety finding, and did not by itself make compounding DSIP legal.

The substantive review came a few months later. On July 24, 2026, FDA's Pharmacy Compounding Advisory Committee formally reviewed "Emideltide-related bulk drug substances" for the 503A Bulks List. FDA's own "Uses evaluated" column, read directly from its meeting page, lists "Opioid withdrawal, chronic insomnia, and narcolepsy"¹¹ — the same three uses the small, mostly uncontrolled human literature above actually covers. Independently confirmed against a separate regulatory summary of the same two-day hearing (a law firm's client alert, not a peptide vendor, cross-checked further against several other 2026 sources reporting on the same meeting), the committee's vote was 6 yes, 7 no, 1 abstention — the motion to recommend Emideltide for the list failed¹². DSIP was the only one of the seven peptides FDA's committee reviewed that week — the others being BPC-157, TB-500, KPV, MOTS-c, Semax, and Epitalon — that did not receive a majority recommendation. Unlike those six, which are now recommended and simply awaiting formal FDA rulemaking, DSIP currently has no committee recommendation behind it at all.

What this means if you're considering a compounded DSIP product

Line up DSIP's record next to the peptides FDA's committee did recommend in July 2026, and it comes up short on more than one axis at once. Its identity as a discrete, well-characterized hormone is disputed in its own field's peer-reviewed literature — not a fringe objection, but the conclusion of researchers who worked directly on it, published in a mainstream neurochemistry journal. Its best-controlled human trials, in the exact population it's most often marketed to today (chronic insomniacs), came back negative or clinically insignificant against a real placebo. And its own FDA advisory committee, reviewing the same opioid-withdrawal, insomnia, and narcolepsy uses this molecule has actually been studied for, declined to recommend it — the only rejection among seven peptides reviewed at the same two-day hearing.

None of that proves DSIP does nothing; a handful of small, uncontrolled pilot studies reported real benefits for chronic pain and one narcolepsy case, and this article isn't claiming those findings were fabricated. What it means is that a reader weighing a compounded DSIP product against a product page's claims should know that the placebo-controlled trials on DSIP's central current marketing claim didn't hold up, and that the compound's basic biological premise — that there is a single, well-defined "DSIP hormone" to begin with — remains an open question in the literature it comes from. Our DSIP reconstitution calculator covers the current compounding-status caution in more detail and does the vial-to-syringe arithmetic; it doesn't, and can't, resolve either of those two open questions, because neither is currently resolved in the published record.