While THC/CBD -based medicinal cannabis is being increasingly prescribed off-label for insomnia, a new Australian study suggests it may actually impair sleep rather than improve it.
A randomised, placebo-controlled trial led by the Woolcock Institute of Medical Research has found that a single oral dose of THC/CBD – a combination of 10 mg of tetrahydrocannabinol (THC) and 200 mg of cannabidiol (CBD) – reduced total sleep time by an average of 25 minutes in adults with diagnosed insomnia.
“People assume cannabis helps with sleep, but the reality is we have very little rigorous data to back that up,” lead researcher Dr Camilla Hoyos told the limbic.
The trial, one of only a handful worldwide to objectively investigate THC’s effects on sleep, involved 20 adults with chronic insomnia. Each participant spent two nights in a sleep lab – one after taking the THC/CBD oil and another after a placebo – spaced at least one week apart to allow for a full washout period. Researchers monitored their sleep using high-density EEG and polysomnography.
Contrary to popular belief, the cannabis-based medication did not improve sleep onset or reduce night-time awakenings. Instead, it disrupted sleep architecture by suppressing rapid eye movement (REM) sleep by 34 minutes and delaying its onset by more than an hour.
“There’s very limited high-quality trial data, especially using objective sleep measures,” Dr Hoyos said. “Our findings show that this combination of THC and CBD changed the structure of sleep – reducing both total sleep and REM – which is not what you’d hope to see from a treatment for insomnia.”
The study, published in The Journal of Sleep Research [link here], also revealed brain activity changes associated with poorer sleep quality. While some cortical arousal was reduced during lighter sleep stages, the deeper stages and REM sleep were disrupted.
“We were able to observe changes in different regions of the brain during sleep thanks to high-density EEG,” Dr Hoyos said. “We saw reduced delta activity in deep sleep and increased arousal during REM, but we still need more data to fully understand the implications of that.”
Interestingly, participants did not report major changes in their subjective experience of sleep.
“We didn’t see big differences in how participants felt after taking the medication – but this was a one-night study,” she explained. “Most symptom scales in insomnia research are taken over a two-week period, so it’s not surprising we didn’t see strong subjective effects.”
The study also touches on challenges around study design in cannabis research.
“These medicines can produce noticeable effects, so in crossover trials like ours, there’s always the risk participants can tell what they’ve taken,” she said. “That can influence how they report their symptoms, even if the objective brain data tells a different story.”
Investigators also examined next-day safety outcomes. Despite the disruption to sleep, there was no measurable impairment in driving, cognitive performance, or psychomotor function nine hours after the dose.
“We designed a related study where participants completed a driving simulation the next morning, mimicking a commute to work,” Dr Hoyos said. “We didn’t find any difference in performance between the THC/CBD and placebo nights.”
However, she noted that despite this, roadside drug testing remains a legal issue for patients prescribed THC products.
“Even if cognitive performance is unaffected, people may still test positive for THC under roadside screening laws – and that has implications for driving.”
Regulators have also moved to tighten oversight. The Australian Health Practitioner Regulation Agency recently issued guidance warning prescribers about unsafe or inappropriate use of cannabinoid products – including for sleep – pushing the need for stronger data to support clinical decision‑making.
The findings come amid a sharp rise in prescriptions for medicinal cannabis, particularly among Australians over 45. Sleep is now the third most common reason for a prescription, according to the federal health department’s medicinal cannabis dashboard. Yet the Therapeutic Goods Administration does not list insomnia as an approved indication.
Looking at the broader evidence base, Dr Hoyos urges caution.
“If you think about what people wait for before prescribing other drugs, it’s usually large phase 3 trials published in major journals,” she said. “Right now, for insomnia and THC, we’re not there yet. Most studies have fewer than 30 participants. Ours was just one night – the longest I’m aware of is two weeks.”
Dr Hoyos and colleagues are calling for longer-term, repeated dosing studies to determine whether tolerance develops or if withdrawal effects, such as REM rebound and vivid dreaming, may exacerbate sleep problems.
“It’s crucial to build a stronger evidence base before widely prescribing THC/CBD for sleep disorders,” she said.