An unfinished drug story
DSIP, still on the bench
A rabbit experiment, a formal drug name, a handful of human results. I expected delta sleep-inducing peptide to have a tidier history.
The candidate with two names
DSIP is delta sleep-inducing peptide, a chain of nine amino acids. It began with a 1977 experiment in rabbits: a substance isolated from their cerebral venous blood enhanced delta and spindle activity in recordings of the brain. The rabbits had been put into an electrically induced sleep state. Even the origin is more peculiar than the reassuring name suggests.
Then there is Emideltide, DSIP's International Nonproprietary Name. That sounds like the next chapter in a drug's development. In this case the chapter remains unfinished: DSIP has no FDA approval for any indication, and no approved medicine has emerged under the Emideltide name.
I thought sleep scores would settle the argument. Instead, the review by Kovalzon and Strekalova leaves much earlier questions open: the receptor, the gene and the precursor protein from which the body might make DSIP. For a would-be medicine, that is an unusually incomplete identity.
Six sleepers and an awkward first hour
The 1981 human sleep experiment deserves more than a dismissive wave. Six middle-aged people with chronic insomnia slept longer, had fewer interruptions and slightly more REM sleep after intravenous synthetic DSIP. The report described no daytime sedation.
I underestimated the timing. Slight arousal came first; the sleep-promoting effect appeared in the second hour. That sequence complicates the idea of a substance that simply switches sleep on. It also makes a promise of immediate drowsiness sound especially careless.
I credit the small experiment with opening a worthwhile question. Its size rules out treatment-level confidence for me, particularly beside the later review of weak and inconsistent sleep evidence. The human research page follows that experiment in detail.
Motor recovery after stroke
Some of the most interesting DSIP work wandered away from sleep. In adult male rats after focal stroke, motor recovery improved by day 7. Meanwhile, infarct volume—the amount of damaged brain tissue—showed no statistically significant difference between groups.
The discordance between movement and tissue loss makes this, to me, an unusually useful animal result. Movement and tissue damage are different outcomes; in this experiment they told different stories. The benefits page puts the actual measurements side by side, alongside the separate animal work on seizures and low oxygen.
Newer sleep work has also changed the substance itself. A 2024 mouse experiment with an engineered DSIP fusion peptide reported stronger effects than unmodified DSIP. The engineered candidate earns further testing, to my mind. Crediting all of that result to ordinary DSIP skips the very modification the researchers were testing.
The remaining human work
The original peptide leads back to rabbits and to some surprisingly ambiguous tissue tests. DSIP peptide benefits follows the sleep, stress and recovery claims. Reports and side effects gets closer to the ordinary question of how people say they felt, including the frequent answer that nothing happened.
I favor a less glamorous next step: a convincing human trial with a clearly identified substance, useful sleep outcomes and adequate safety follow-up. The animal leads make DSIP interesting enough to study. The missing human work is why the story still feels suspended.