Cannabis Pill Disrupts REM Sleep in Insomnia Patients but Leaves Next-Day Function Largely Intact

A small pilot trial finds a single dose of THC/CBD significantly suppresses REM sleep without measurable impairment to daytime alertness or driving performance nine or more hours later, but the study’s size and single-dose design preclude any conclusions about therapeutic efficacy or long-term safety.

Why This Matters

Insomnia is among the most commonly cited reasons patients seek cannabis-based treatments, yet rigorous clinical evidence for cannabinoid sleep therapies remains remarkably thin. Most existing data come from preclinical models, retrospective surveys, or trials without polysomnography. This pilot trial matters because it applies gold-standard sleep measurement tools, including high-density EEG, to a clinically diagnosed insomnia population under placebo-controlled conditions. At a time when regulatory bodies and patients alike are making decisions about cannabinoids and sleep, even small controlled studies that generate objective data fill a critical void.

Clinical Summary

Cannabis compounds, particularly delta-9-tetrahydrocannabinol, have long been observed to alter sleep architecture, with REM suppression being one of the most consistently documented effects in healthy volunteers. However, translating those observations into clinical insomnia populations has lacked rigor. This pilot randomized controlled trial, published in the journal Sleep (2024) by Suraev and colleagues through the Lambert Initiative at the University of Sydney, enrolled 20 adults with DSM-5-diagnosed insomnia disorder. Using a within-participant crossover design, each participant received both a single oral dose of 10 mg THC combined with 200 mg CBD and a matched placebo, separated by a washout period. The mechanistic premise rests on THC’s known interaction with CB1 receptors in sleep-regulatory circuits, particularly those governing REM generation, while CBD’s inclusion was intended to modulate THC’s psychoactive profile.

The primary findings confirmed substantial REM suppression under active treatment: REM sleep duration fell by 33.9 minutes compared to placebo, a large effect (Cohen’s d of negative 1.5, p less than 0.001), while REM latency increased by over 65 minutes (d = 0.7). Total sleep time was modestly reduced by 24.5 minutes. Crucially, objective next-day assessments conducted at least nine hours post-dose, including the Maintenance of Wakefulness Test, psychomotor vigilance testing, and simulated driving, showed no significant impairment, though self-reported sleepiness increased by a small but statistically significant margin. The authors are explicit that the pilot scale of 20 participants, the single-dose design, the combined THC/CBD formulation preventing attribution to either cannabinoid alone, and the predominantly female sample all constrain generalizability. They call for larger, longer trials before any clinical recommendations can be drawn.

Dr. Caplan’s Take

This study does something genuinely valuable: it brings polysomnography and high-density EEG into a space that has been dominated by patient self-report and marketing claims. The REM suppression finding is robust and consistent with what we know about THC. But I want to be clear with my patients who bring this up: reducing REM sleep is not the same as treating insomnia. Many people with insomnia have difficulty initiating or maintaining sleep, and this pill actually reduced total sleep time. The fact that next-day function was preserved at nine-plus hours is reassuring, but a single dose in 20 people does not tell us what happens after weeks of nightly use.

In practice, when patients ask me about cannabis for sleep, I start by ensuring they have been evaluated for underlying causes and have tried cognitive behavioral therapy for insomnia, which remains the first-line treatment with the strongest evidence base. For patients already using cannabis products for sleep, I discuss what we actually know versus what is assumed, and I counsel them that REM suppression over time may have consequences we do not yet understand. I do not currently recommend THC/CBD formulations as a sleep treatment based on available evidence, though I follow this research closely and adjust my guidance as better data emerge.

Clinical Perspective

This trial occupies a very early position in the research arc for cannabinoid-based insomnia interventions. It confirms THC’s REM-suppressing properties in a clinical population, extending prior findings from healthy-volunteer studies, and it provides the first high-density EEG spectral data in insomnia patients receiving a standardized oral cannabinoid formulation. However, it does not demonstrate therapeutic benefit: subjective sleep quality was unchanged, total sleep time decreased, and the study was not designed or powered to evaluate clinical efficacy. The absence of next-day impairment at nine or more hours is notable and relevant for regulatory and driving-safety discussions, but clinicians should not interpret this as evidence that the compound is safe for chronic use or that shorter post-dose windows would yield the same result.

The combined THC/CBD formulation means the individual contributions of each cannabinoid cannot be parsed. CBD at 200 mg may have attenuated some THC-related effects, making extrapolation to THC-dominant products inappropriate. From a pharmacological standpoint, THC’s interactions with sedative medications, anticoagulants, and CYP-metabolized drugs remain relevant considerations for any patient already on polypharmacy regimens. The most actionable takeaway for clinicians right now is this: when patients cite cannabis as a sleep aid, this study provides an evidence-based framework for explaining that while cannabinoids clearly alter sleep architecture, altering sleep architecture is not equivalent to treating insomnia, and the long-term implications of chronic REM suppression remain unknown.

Study at a Glance

What Kind of Evidence Is This

This is a pilot randomized controlled trial using a within-participant crossover design, published in a peer-reviewed sleep journal and prospectively registered. While RCTs sit near the top of the evidence hierarchy, the pilot designation is critical: with only 20 participants and a single-dose exposure, the study is explicitly designed to generate preliminary data and effect-size estimates rather than definitive conclusions. The most important inference constraint is that findings from a single acute dose cannot be generalized to the repeated nightly use that would characterize any real-world insomnia treatment.

How This Fits With the Broader Literature

The REM suppression finding is highly consistent with decades of research on THC’s effects in healthy volunteers, including early work by Pivik and colleagues and more recent controlled studies using lower THC doses. What this trial adds is confirmation that the same effect occurs in a clinically diagnosed insomnia population using a standardized oral formulation with concurrent CBD. The next-day functional preservation finding aligns with emerging data from driving-safety research on oral cannabinoids at longer post-dose intervals, though it contrasts with studies showing impairment at shorter windows.

Notably, this study diverges from the patient self-report literature, which commonly describes improved subjective sleep with cannabis use. Here, subjective sleep quality was unchanged despite dramatic changes in objective sleep architecture, a disconnect that underscores the limitations of relying on patient-reported outcomes alone in cannabinoid sleep research.

Common Misreadings

The most likely overinterpretation is to read this study as evidence that THC/CBD is an effective or safe treatment for insomnia. It is neither. The study demonstrates that a single dose alters sleep architecture, predominantly by suppressing REM, but it does not show improvement in any insomnia-relevant outcome: subjective sleep quality was unchanged and total sleep time actually decreased. A secondary misreading involves extrapolating the next-day safety data to shorter post-dose windows, different formulations, or chronic use. The nine-hour-plus assessment window was deliberately chosen and the findings cannot be assumed to hold at earlier time points or after repeated dosing.

Bottom Line

This well-designed pilot confirms that oral THC/CBD powerfully suppresses REM sleep in insomnia patients and provides reassuring, though preliminary, data on next-day functional safety at nine or more hours post-dose. It does not demonstrate therapeutic benefit for insomnia, cannot separate THC from CBD effects, and is far too small and brief to inform clinical practice. Its primary value is as a methodological foundation for the larger, longer trials that are genuinely needed before cannabinoids can be responsibly recommended for sleep disorders.