Peptides for perimenopause sleep: what the human trials actually show

If you are awake at 3am with the duvet off and a full day ahead, the peptide pitch will find you. It is written for exactly this moment: a molecule your body already makes, a mechanism that sounds like biology rather than sedation, and deep sleep without a prescription or a hangover. We read the trials. All of them. Here is what is in there, and what is not.

What is actually waking you

Perimenopausal sleep does not break in one way. It breaks in several at once, and they need different answers. In the Study of Women's Health Across the Nation, 82% of 3,082 women reported a sleep disturbance at baseline.

• Night sweats. A vasomotor event during sleep wakes you, often fully. Women in the MsFLASH trials averaged eight hot flushes a day; the nocturnal ones cost you sleep.
• Anxiety and the early-hours cortisol rise. The 3am wake with a racing mind is a different animal from the 1am wake soaked through.
• Falling progesterone. Its metabolites are sedating, and as cycles become erratic the luteal-phase calm becomes unreliable.
• Undiagnosed sleep-disordered breathing. In the Wisconsin Sleep Cohort, postmenopausal women had 3.5 times the odds of moderate sleep apnoea versus premenopausal women, adjusted for age and body habitus. If you snore or feel unrefreshed however long you were in bed, ask for a sleep study before you ask for a peptide.

None of this is what the peptide vendors address. They address a fifth mechanism, and it is a real one.

The honest mechanism, and the part they leave out

Growth hormone comes in pulses, and the largest of the day is tied to the first episode of slow-wave sleep after you fall asleep. Reviewing the evidence in 2000, Van Cauter and Copinschi described a linear relationship between the amount of slow-wave sleep and the amount of concurrent GH secretion; with age, both decline exponentially and on the same timetable. That is real physiology, and it is why the marketing works.

A plausible mechanism is not a trial. When researchers gave GH-releasing drugs to women and recorded their sleep, the results did not go the way the mechanism predicted.

The compounds sold for sleep, compared

DSIP: the one peptide the FDA panel said no to

Delta sleep-inducing peptide is in every competitor's comparison table and explained in none of them. At the Pharmacy Compounding Advisory Committee meeting on 23–24 July 2026, the committee voted on seven peptides. FDA's own career scientists recommended against all seven. The committee voted in favour of six anyway. Emideltide — the formal name for DSIP — was the only one that failed: six in favour, seven against, one abstention.

What FDA reviewers actually wrote

The agency's briefing document is public. On effectiveness for insomnia, reviewers wrote that results "appear inconclusive and at best preliminary," with "conflicting study outcomes." On the route people are actually buying, they were blunter: there is "no study evaluating effectiveness to support use of emideltide via the nominated SC route."

Every human sleep trial of DSIP used intravenous infusion. Every vial sold to women for sleep is subcutaneous. Nobody has tested the thing being sold, by the route it is sold in. Reviewers also noted that the nominator's certificate of analysis mentioned no bioburden or endotoxin test, and that missing information on impurities and peptide aggregation carries an immunogenicity risk that matters more when a product is injected.

The actual trial record

DSIP was isolated in Basel in the 1970s. The human sleep work runs from 1981 to 1992, then stops. Six trials, all intravenous, nearly all at 25 nmol/kg. Two — Experientia 1981 (n=6) and Eur Neurol 1987 (n=14), both in middle-aged chronic insomniacs — reported longer sleep, fewer interruptions and better daytime alertness. Both came from the group that discovered the molecule.

The two controlled studies run by other teams did not agree. A double-blind crossover in Montevideo (Int J Clin Pharmacol Res, 1987) found wakings and waking time fell, but not significantly against baseline or placebo; the improvement was "of little clinical significance." A double-blind study of 16 chronic insomniacs in Amsterdam (Neuropsychobiology, 1992) found higher sleep efficiency and shorter latency, but statistically weak: short-term use in chronic insomnia was "not likely to be of major therapeutic benefit."

The largest all-female DSIP study found the opposite

In 2009, anaesthetists at the University of Manchester gave DSIP to 24 women undergoing surgery, testing whether a "natural hypnotic" would deepen anaesthesia. It did not. DSIP significantly reduced delta rhythm and raised the bispectral index — anaesthesia got lighter, not deeper — while increasing heart rate and decreasing heart rate variability.

That is the biggest study of DSIP in women ever published, and the delta-wave effect ran backwards. Different setting, different question, and it does not prove DSIP disturbs ordinary sleep — but it is the only substantial female dataset there is, and it points the wrong way. For the separate claim that DSIP raises growth hormone: in 1993, endocrinologists in Genoa gave it to eight healthy women and nothing moved — not GH, not prolactin, not their circadian rhythms. There are zero registered interventional studies of DSIP on ClinicalTrials.gov.

CJC-1295 and ipamorelin: no sleep trial exists

This is the pairing most clinics put on a sleep protocol, dosed at bedtime on the explanation that it amplifies your natural night-time GH pulse. There is no human trial of CJC-1295 or ipamorelin with a sleep endpoint, in either sex. Not a small one, not an unpublished one. The claim has never been tested.

The molecule studied is not the molecule sold. The single human study of CJC-1295 — Teichman and colleagues in JCEM, 2006, in healthy adults aged 21 to 61 — used CJC-1295 with DAC, half-life 5.8 to 8.1 days. Much of what clinics dispense is CJC-1295 without DAC, also called mod GRF 1-29: half-life around thirty minutes, no published human trials whatsoever. Different drugs, and most vials do not tell you which you have.

The only efficacy trial was terminated. NCT00267527, a Phase 2 in HIV-associated visceral obesity with a planned enrolment of 120, was terminated in 2006. No reason posted, never published. That is the only registered efficacy trial CJC-1295 has ever had.

Ipamorelin's dosing numbers come from eight men. The only human pharmacokinetic study used eight healthy males per dose level; every half-life and clearance figure in clinic protocols traces to them. Two ipamorelin Phase 2 trials completed in 2009 and 2014, enrolling 117 and 320, and neither has posted results.

What happened when researchers gave women GH secretagogues

This is the part of the literature that decides the question, and the part nobody in this niche has read. The Max Planck Institute of Psychiatry in Munich spent two decades giving GH-releasing drugs to people overnight with EEG electrodes attached, and found a consistent sex difference. It is not in your favour.

Sermorelin, CJC-1295 and tesamorelin are all GHRH analogues. That is the class the finding applies to. The ghrelin-mimicking side of the family, which is what ipamorelin is, looks no better:

• Ghrelin increases slow-wave sleep in young men. Given to ten healthy young women on the same protocol it produced no significant effect on sleep at all, while raising both GH and cortisol.
• Hexarelin, a secretagogue in the same family as ipamorelin, decreased stage 4 sleep and EEG delta power while raising ACTH and cortisol.

Among drugs that stimulate GH secretion, then, the human sleep data show: better in men, neutral or worse in women, several raising cortisol at exactly the hour you want it lowest.

The trial that would have answered this, and never reported

In June 1996 the University of Washington registered NCT00000380: GHRH or placebo for age-related sleep impairment in 40 older men and 40 older women on estrogen replacement therapy — the closest thing to a GHRH-for-sleep trial in postmenopausal women that has ever existed. It completed in July 2007. No results have ever been posted. The team published a cognitive outcome in 2006; the sleep outcome, the study's stated purpose, does not appear in the literature.

If you are on HRT, the route of your estrogen matters here

This is the most consequential thing on this page for anyone already on hormone therapy, and no clinic protocol or product label discloses it. Oral estrogen blunts growth hormone action — not GH secretion, GH action, at the liver. It induces SOCS-2, which inhibits GH's JAK/STAT signalling. The result is lower IGF-1 and reduced fat oxidation, and in postmenopausal women a loss of lean tissue and gain of body fat. Transdermal estrogen does not do this to anything like the same degree, because it bypasses first-pass hepatic metabolism.

You can see it directly in women. In a University of Washington study, 24 postmenopausal women on oral estrogen for a mean of sixteen years were compared with 40 women who were not, blood drawn every twenty minutes for 24 hours. The oral-estrogen group had significantly higher 24-hour GH and higher GH during sleep, and significantly lower IGF-1 — the signature of a liver not listening.

What does have evidence for perimenopausal sleep

CBT-I, delivered by telephone Supported

In a MsFLASH randomised trial, 106 perimenopausal and postmenopausal women aged 40 to 65 with moderate insomnia and at least two hot flushes a day got six telephone sessions over eight weeks — either cognitive behavioural therapy for insomnia or a menopause education control. Insomnia Severity Index scores fell 9.9 points in the CBT-I group against 4.7 in controls. At 24 weeks, 84% of the CBT-I group scored in the no-insomnia range against 43% of controls. That is a larger and longer-lasting effect than anything else in this article, from six phone calls.

Addressing the night sweats Mixed

In another MsFLASH trial, 339 women were randomised to low-dose oral estradiol 0.5 mg, venlafaxine XR 75 mg, or placebo for eight weeks. Insomnia scores fell 4.1 points on estradiol, 5.0 on venlafaxine and 3.0 on placebo. Against placebo, venlafaxine reached significance for insomnia symptoms (p=0.007) and estradiol did not (p=0.09); on sleep quality estradiol did (p=0.04) and venlafaxine narrowly did not.

We will not overstate this. Hormone therapy's effect on midlife sleep in a randomised trial was real but modest, and it works largely by reducing what wakes you rather than acting on sleep directly. It is still better evidenced than any peptide here, and it comes with a prescriber, a dose and a known safety profile.

Micronized progesterone Thin

Eight postmenopausal women took 300 mg of micronized progesterone or placebo nightly for three weeks. On an undisturbed night it did nothing; on a deliberately disturbed night, wake-after-sleep-onset was 53% lower and slow-wave sleep about 50% higher. Eight women is eight women — but it is a randomised, double-blind, polysomnographic study of a drug your doctor can already prescribe.

The 30% problem

Everything above assumes the vial contains what the label says. Independent testing by BioTools found problems in nearly 30% of peptide vials tested — mislabelled, underdosed, overdosed, or contaminated. Co-founder Rina Dukor: "Among the thousands of tests we are doing, we do find either live or dead bacteria in a lot of the vials."

That makes every dosing discussion in the category partly hypothetical. If a vendor cannot show you a certificate of analysis for the batch you are buying, you do not know what you are injecting at midnight.

Where the law actually stands

A lot of sites are currently writing that the FDA has approved these peptides, or that they are no longer banned. Both are false, and the error matters.

In April 2026 the FDA removed a group of peptides — DSIP, BPC-157, TB-500, Epitalon, MOTS-c and others — from Category 2, the list of substances flagged as presenting significant safety risks. That does not make them legal to compound. And a July 2026 committee vote is advisory only: to become lawfully compoundable, a substance must go through notice-and-comment rulemaking — proposed rule, comment period, final rule — and the comparable previous process took more than two years. Until a final rule issues, none of these satisfies any of the three conditions in section 503A: a USP or NF monograph, being a component of an FDA-approved drug, or appearing on the 503A Bulks List. And DSIP did not even get the advisory nod.

If you try something anyway

• Sort the obvious things first: night sweats, apnoea, alcohol, a thyroid check. Thyroid disease is common at this age and routinely mistaken for menopause.
• Know which molecule you have. If it says CJC-1295, ask whether it is with or without DAC. If the vendor cannot tell you, that answers a different question.
• Doses here are flat and male-derived. Published pharmacology is weight-normalised; clinic protocols are not. A 55 kg woman on the same 250 mcg as a 95 kg man gets roughly 1.7 times the exposure.
• Set a stop date before you start, with a sleep diary and an Insomnia Severity Index score at baseline and six weeks.
• You can still get pregnant in perimenopause, and none of these compounds has reproductive safety data.
• If you have had breast cancer, carry a BRCA variant, or take tamoxifen or an aromatase inhibitor, note that GH secretagogues raise IGF-1 — a discussion for your oncologist, not a forum.

The same evidence problems recur across the category — see our guides to energy and brain fog, weight in perimenopause, libido and mood, muscle and bone and skin, hair and collagen.

The bottom line

DSIP has six small human sleep trials, all intravenous, all finished before 1993 — the positive ones from the laboratory that discovered it, the independent ones concluding the effect was too small to matter. The route it is sold in has never been tested, and the FDA's advisory committee declined to recommend it. CJC-1295 and ipamorelin have no sleep trial at all, and when researchers gave drugs from those classes to women overnight, sleep got no better and sometimes worse, while cortisol went up.

Meanwhile six telephone calls of CBT-I moved insomnia severity by ten points and held it for six months in exactly your population. We are not telling you what to take. We are telling you the evidence is not where the marketing says it is, and that on this symptom the well-supported options are unglamorous, prescribable and considerably cheaper.