Compound
GrantsAbout

2026 Compound

Compound
GrantsAbout
Back to Database
Ozempic for Sleep
Bio

Ozempic for Sleep

A sleep-focused therapeutics startup

Concept

A cutting edge drug discovery startup focused exclusively on sleep indications, ideally expanding from narrower ones to AD/PD or even to the relatively healthy population

Longer Description

Sleep has received shockingly little pharma attention despite the following:

  • Everyone spending 1/3 of their life doing it
  • 35% of adults report not getting the sleep they desire
  • There’s existence proof of highly efficient sleep, with a small proportion of the population needing <7hr to function normally (or even better) and no apparent adverse effects.
  • Deep sleep shows high variance within and amongst individuals. In population samples, conventionally scored SWS can vary from ~2% to ~30% of the night within age-matched older adults. Sleep deprivation yields deeper sleep both in terms of SWS and SWA. And, deep sleep wains with age.
  • A handful of sleep-related disease indications with unmet needs that have previously produced four blockbuster drugs plus several more doing $300M+ a year in sales. These include obstructive sleep apnea, insomnia, REM sleep behavior disorder, type 1 narcolepsy, restless leg syndrome
  • A clear path (that has been attempted before) to expand from initial sleep indication to very broad neurological diseases like Alzheimer’s, Parkinson’s, and Dementia given sleep’s core role in brain health, memory, etc. Those populations also have systematically worse sleep. It also plays a role in other large-scale diseases including obesity, diabetes, cardiometabolic risk, and psychiatric conditions from depression to schizophrenia.
  • A longer-term theoretical path to one of the only population-scale indications, including one that could in theory be taken by effectively every person every day for the rest of their lives

The major clinical trials to date are detailed below.

Thesis image

Pharmacological Approaches

Results have largely been mixed with only a few demonstrably successes including Ambien, Jazz’s sodium oxybate, and Takeda’s recent OX2R-selective agonist. The latter improved T1 narcolepsy patients to normal functioning at 12 weeks in PIII. Other than that, most core sleep-related diseases remain unsolved.

Moreover, almost all clinical attempts have clustered around the same select few targets of GABA, orexin, serotonin, and calcium-channels.

A cutting edge startup focused exclusively on the biology of sleep could develop new therapeutic approaches to these known targets and discover novel sleep-related targets.

The dream target would a readily druggable master regulator like GLP-1s.

Unfortunately for those of us looking for an instant gold rush, it appears that the gene variants linked to claimed “natural short sleepers” are candidate genes invalidated by modern population-scale GWAS studies that find no reason to believe they replicate or are meaningfully related to sleep. Moreover, studies of 400K+ adults found that of the few dozen loci associated with short sleep, the one with the largest single effect contributed just 2.4 minutes.

However, just because sleep is a polygenic, complex phenomena likely regulated by many circuits doesn’t mean there aren’t fulcrums with disproportionate leverage like GLPs.

There’s a handful of other known targets and therapeutic approaches we’ve come across that could make for appealing starting grounds:

  • NPSR1 and SIK3 may be the only short sleep pathways possibly worth pursuing, as no carriers of its variants were found in that 127K-person study and thus weren’t actively discredited.
  • If NPSR1 proves to be a hit, it could the world’s first anti-anxiety stimulant but would go against orexin antagonists which are already quite good drugs.
  • SIK3’s role as a regulator of sleep need has been replicated by multiple labs in different animal models, though it hasn’t been identified in human GWAS studies
  • Novel approaches to obstructive sleep apnea. Despite its huge patient population of 30M US adults and high unmet need (I can personally attest to the annoyance of strapping a CPAP machine to your face every night), OSA is close to getting its first ever therapeutic treatment. For a broad CPAP-intolerant population, AD109 delivers moderate-to-large average AHI reductions but just 1 in 5 reach normal sleep as measured by an AHI below 5.
  • A polygenic risk score for alleles contributing to extremely high or low sleep need or insomnia odds replicated across UK and Finland’s populations. Maybe studying the pathways that lead to the tails will reveal a master regulator.
  • Novel approaches to insomnia. While it’s a somewhat crowded field with generic versions of benzodiazepines, Z-drugs, and dual orexin antagonists, each has meaningful side effects for daily use and patients are often dissatisfied with existing options.
  • Some specific targets with compelling evidence include adenosine pathways (e.g. A1R/A2AR PAMS but they’ll likely still have a small therapeutic window), astrocyte Ca2+, BF glutamatergic neurons, T-type Ca²⁺ channels in thalamus/reticular nucleus, SST+ cortical interneurons, stress axis peptides like CRH/CRHR1, PGD2/DP1, and GIRK1/2
  • Epigenetic editing to cure narcolepsy as research indicated that it’s caused by epigenetic silencing
  • Targeting corticotropin‐releasing hormone neurons, which are reduced by ~88% in NT1 patients’ post-mortem hypothalamus
  • A GWAS study fine-mapped a missense variant (p.Tyr727Cys) in PDE11A as the likely causal allele for nocturnal sleep duration in the cAMP/cGMP pathway
  • A few dozen genes are well linked to circadian rhythms. Some groups have even found preclinical hits against them, including BMAL1, CRY / CRY1 / CRY2, and PER1 / PER2 / PER3. A drug candidate targeting CRY2 just became the first asset ever to reach clinical trials targeting a circadian rhythm TF, in this case for glioblastoma.

    Lastly, many animals with unusual sleep patterns could be relevant for studying and possibly even yield valuable biobanks, including African elephants only sleep 2-4 hours a day, hibernating animals, induced torpor states, those that sleep with one brain hemisphere awake at a time, and species that evolve apart based on differences in habitat to require differing levels of sleep

  • Thesis image

    Non-Invasive Approaches

    Hypnosis has been used to substantially boost SWS and reduce time spent awake in highly hypnotizable subjects, though it will likely require more rigorous trials to confirm effect.

    Meanwhile, several startups have pursued boosting slow waves via sound or electric waves.

    Thesis image

    The approach feels appealing because its non-invasive, many people already wear eyemasks, and modulating slow waves is a plausible biological hypothesis.

    Lucky for future startups, the first attempts haven’t been impressive.

    In theory they’re intended to detect and amplify each individual’s exact brain waves over time in a closed-loop neurofeedback set up. However, they often can’t even measure sleep stages accurately. Three companies have run small trials that are far from convincing in terms of both results and the weakness of their controls.

    A current MIT + Samsung effort aims to use the Samsung smartwatch to create the auditory stimulus for slow wave enhancement, but presumably will be even less precise than those above.

    Next-gen startups could possibly push on the materials science innovation to create a non-stick EEG interface with your skin for precise measurement.

    Future attempts should embrace the nuances of sleep quality

    Most just measure total slow wave sleep, but quality / restorative sleep is driven by:

    • Amplitude, frequency, and shape of waves. Higher wave activity / “delta power” during NREM sleep is associated with better sleep quality while the opposite holds for REM sleep. Steeper slopes reflect stronger synaptic synchronization and higher sleep pressure early in the night.
    • Next‑day cognition and “rested” feeling depend on how slow oscillations (∼<1 Hz) coordinate with sleep spindles (∼11–16 Hz)—a timing relationship that supports memory consolidation and sensory gating. Some SWS‑enhancing drugs alter sigma/spindle activity (e.g., gaboxadol increases low‑frequency power but attenuates spindles dose‑dependently), potentially trading depth for impaired coupling.
    • Cardiopulmonary coupling: “stable NREM” (high‑frequency coupling) vs “unstable NREM” (low‑frequency)

    Non-Dilutive Money

    One especially appealing aspect of boosting wakefulness and / or sleep efficiency is it fills a core need for defense. The drugs for soldiers angle will open up serious non-dilutive funding opportunities. E.g. DARPA gave Jonathan Rivnay $33M to develop a implantable chip that controls wake/sleep cycles and the Navy has an outstanding request for sleep/fatigue modulators.

    Comparable Companies

    Takeda Pharmaceuticals

    Jazz Pharmaceuticals

    Centessa Pharmaceuticals

    Apnimed

    Applied Cognition

    AstronauTx

    Avadel Pharmaceuticals

    Alkermes

    Eisai

    Idorsia Pharmaceuticals

    Elemind

    SleepSpace

    Somnee

    Beacon Biosignals

    Comprehensive list of companies that have developed sleep-related drugs

    Insomnia

  • Merck — suvorexant (Belsomra, DORA). Merck+1
  • Eisai — lemborexant (Dayvigo, DORA). Eisai+1
  • Idorsia — daridorexant (Quviviq, DORA). Idorsia+1
  • Taisho — vornorexant (DORA; Japan filing). Insights+1
  • Takeda — ramelteon (Rozerem, melatonin-receptor agonist). Takeda Pharmaceuticals+1
  • Somaxon → Pernix — doxepin (low-dose) (Silenor, H1-antagonist). FDA Access Data+1
  • Sanofi — zolpidem (Ambien, “Z-drug”). FDA Access Data
  • Sepracor → Sunovion — eszopiclone (Lunesta, “Z-drug”). FDA Access Data+1
  • Wyeth-Ayerst → Pfizer — zaleplon (Sonata, “Z-drug”). FDA Access Data+1
  • J&J + Minerva — seltorexant (selective OX2R antagonist; insomnia & MDD with insomnia sx). PMC+1
  • Narcolepsy (cataplexy/EDS) & Idiopathic Hypersomnia (IH)

  • Jazz Pharmaceuticals — Xyrem/Xywav (oxybates; narcolepsy & IH). Jazz Pharmaceuticals Investor Relations+1
  • Avadel — LUMRYZ (once-nightly sodium oxybate; narcolepsy). Reuters
  • Harmony Biosciences / Bioprojet — pitolisant (Wakix; H3 inverse agonist; narcolepsy; peds sNDA in 2024). Harmony Biosciences+1
  • Axsome — Sunosi (solriamfetol; EDS in narcolepsy/OSA) and AXS-12 (reboxetine; late-stage for narcolepsy). PMC+1
  • Takeda — oveporexton / TAK-861 (OX2R agonist; positive Ph 3 in NT1). Nature+1
  • Obstructive Sleep Apnea (pharmacologic approaches)

  • Apnimed — AD109 (aroxybutynin + atomoxetine; positive Ph 3 topline). Apnimed+1
  • Circadian rhythm & rare sleep-wake disorders

  • Vanda — tasimelteon (Hetlioz; Non-24 and SMS nighttime sleep disturbance). PR Newswire+1
  • Neurim — Circadin (prolonged-release melatonin; EU insomnia) and Slenyto (pediatric prolonged-release melatonin; EU). European Medicines Agency (EMA)+1
  • Restless Legs Syndrome (RLS)

  • UCB — rotigotine (Neupro patch). FDA Access Data
  • GlaxoSmithKline — ropinirole (Requip). FDA Access Data
  • Boehringer Ingelheim — pramipexole (Mirapex). FDA Access Data
  • XenoPort → Arbor — gabapentin enacarbil (Horizant)
  • Related Reading

    https://pmc.ncbi.nlm.nih.gov/articles/PMC9499244/

    https://pmc.ncbi.nlm.nih.gov/articles/PMC6451011/

    https://pmc.ncbi.nlm.nih.gov/articles/PMC6851116/

    https://www.wired.com/review/somnee-sleep-headband/

    https://www.biospace.com/press-releases/sleepspace-awarded-nih-grant-to-transform-smartphones-into-digital-therapeutics-for-insomnia-cognition-and-mental-health?utm_source=chatgpt.com

    https://news.northwestern.edu/stories/2021/05/implantable-living-pharmacy-could-control-bodys-sleepwake-cycles

    https://www.media.mit.edu/publications/slow-wave-stimulation-using-a-smartwatch-improves-sleep-quality/

    https://www.mdpi.com/2624-5175/5/1/2

    https://www.nature.com/articles/s41467-018-08259-7

    https://apnimed.com/article/ad109toplinephase3results/

    Related Theses

    Marketplaces Requiring Private Intelligence
    01
    AI/ML

    Marketplaces Requiring Private Intelligence

    For deals requiring human-like analysis and negotiation but can’t risk information leakage

    Read
    New bio data + AI businesses
    02
    Bio

    New bio data + AI businesses

    Full-stack business required when new data modalities come online

    Read
    Cancer Vaccines
    03
    Bio

    Cancer Vaccines

    Vaccinate yourself against your own tumor

    Read

    2026 Compound