DSIP

In 1977 the Schoenenberger–Monnier group in Basel took venous blood from the brains of rabbits in induced sleep, isolated a nonapeptide from it, and named it delta sleep-inducing peptide [1]. That is a legitimate piece of classical neuroendocrinology, and it is why DSIP got a proper run of human trials while most of the peptides sold alongside it got none.

The human trials, in order. All used 25 nmol/kg intravenously.

• 1981, healthy volunteers. Six subjects, double-blind crossover, dosed in the morning. Subjects reported immediate sleep pressure and total sleep time over the next 130 minutes rose 59% against placebo; the following night had shorter sleep onset and better efficiency. EEG analysis showed no sedation in the classical pharmacological sense [2].
• 1981, insomniacs. Six middle-aged chronic insomniacs given a single dose slept longer and better, with slightly more REM and no daytime sedation. The effect appeared only in the second hour; the first hour showed a slight arousing effect [3].
• 1987, seven nights. Fourteen middle-aged chronic insomniacs, placebo-controlled and double-blind. Sleep improved substantially with the first dose and further with repeated ones, the improvement carried into a post-treatment placebo night, and daytime alertness and mental performance rose significantly [4].
• 1987, crossover. Four nights, double-blind crossover. Total and NREM sleep rose, and the rise was in stage 2 - slow-wave sleep, the thing the peptide is named after, did not change. The differences from placebo were also present at baseline. Conclusion: "sleep improvement under DSIP treatment is of little clinical significance" [5].
• 1992, parallel groups. Sixteen chronic insomniacs, double-blind, matched pairs, five laboratory nights. Sleep efficiency and latency favoured DSIP, but the effects were weak, partly attributable to an incidental change in the placebo group, and subjective sleep quality did not move at all. Conclusion: short-term treatment of chronic insomnia with DSIP "is not likely to be of major therapeutic benefit" [6].

The pattern is the familiar one: the tighter the design, the smaller the effect. It is worth saying that even the negative trials found a consistent direction, and that all of them found the peptide well tolerated.

Outside sleep. An uncontrolled pilot in seven patients with migraine, vasomotor headache, tinnitus or psychogenic pain reported significantly lower pain in six of them after a course of injections, along with reduced depressive symptoms [8]. Seven patients, no control group. Animal work continues in stroke recovery [9], memory at altitude using a phosphorylated analogue [10] and stress protection with the Russian preparation Deltaran [11], and DSIP has been proposed as a link between circadian mechanisms and obesity [12].

The unresolved part. No DSIP gene has been isolated. No precursor protein. No receptor. The 2006 review that surveyed the field concluded the sleep hypothesis is "extremely poorly documented and still weak", and proposed that the immunoreactivity and activity attributed to DSIP may belong to some other, related peptide that has not been identified [1]. Earlier reviews had raised the same problem [13][14][15], and a survey of neuropeptides and human sleep placed it among the substances with suggestive rather than established effects [16].

There is no established mechanism, because the molecular target has never been found [1]. What exists are observations.

DSIP-like immunoreactivity is distributed in neurosecretory hypothalamic nuclei across vertebrate species — but, as the 2006 review points out, in nuclei that are not particularly relevant to sleep regulation, which is one of the reasons for suspecting the immunoreactivity belongs to a different peptide [1]. The peptide crosses the blood-CSF barrier in dogs, and the degree of penetration tracks plasma concentration, plasma half-life and lipophilicity — so its analogues, which differ in those properties, get into the brain to different degrees [17][18]. Notably, some artificial DSIP structural analogues promote slow-wave sleep more convincingly than DSIP itself does [1].

The wider effects attributed to it — interaction with opioid-peptidergic systems, induction of brain MAO-A activity, effects on circadian locomotion, cortisol and stress responses — come from the older animal literature and were the rationale for trying it in pain and withdrawal [8]. A proposed link to glucocorticoid-induced leucine zipper connects it to circadian and metabolic regulation, but this is a hypothesis paper, not a finding [12].

One practical consequence of the pharmacokinetics deserves emphasis. With a plasma half-life of two to four minutes [7], almost nothing that DSIP does can be a direct effect of the circulating peptide hours later. The seven-night trial's persistence of benefit into a post-treatment placebo night [4] therefore requires some downstream, longer-lived change — which nobody has characterised.

The trials are small but consistent on tolerability. In the first human study, extensive behavioural, EEG, physiological and biochemical measurements found no side effects and no classical sedation [2]. The insomniac studies reported no daytime sedation [3] and no adverse events prominent enough to reach the abstracts [4][5][6]. The pain pilot reported a reduction in depressive symptoms rather than any new problem [8].

That is a reassuring record, and it comes with three large qualifications. The total number of people dosed across all published trials is under sixty. Every one of them received the peptide intravenously in a laboratory, at 25 nmol/kg, for at most seven nights — so nothing is known about repeated subcutaneous use over months. And there is no post-marketing surveillance of any kind, because DSIP has never been a licensed medicine.

The deeper safety unknown is the one the 2006 review identified: with no receptor and no gene, there is no way to reason about what else the peptide might act on, or about what chronic exposure would do [1].

No interaction studies exist. The older animal literature describes modulation of responses to systemically administered morphine and amphetamine, and that pharmacology is why DSIP was tried in opioid withdrawal and chronic pain [8]. Whether that translates into a meaningful interaction in people has never been tested.

Nothing in the literature documents an interaction with melatonin or any other sleep compound on this site.