Workers Compensation
Sleep is one of the most common reasons medicinal cannabis is prescribed to injured people, and "I sleep better on it" is the most common reason the prescription continues. The objective sleep evidence tells a more complicated story, and it matters for anyone funding long-term THC in a personal injury claim.
Why does feeling asleep differ from sleeping well?
THC is a CB1 receptor agonist with sedative, anxiolytic and analgesic effects. Take it before bed and most people will feel more relaxed, fall asleep faster and report that their sleep has improved. That perception is real. It is also exactly what you would expect from any sedating drug, including the benzodiazepines and Z-drugs the sector has spent a decade trying to deprescribe.
The problem is that sleep is not simply a block of hours spent unconscious. Restorative sleep depends on moving through the normal cycle of light sleep, deep slow wave sleep and REM sleep, with few awakenings and a nervous system that shifts into parasympathetic recovery mode overnight. A drug can improve the feeling of sleep while disrupting several of those things at once.
This disconnect between subjective and objective sleep is one of the most consistent findings in the cannabinoid sleep literature, and it is precisely the gap that claims decisions tend to fall into.
What do objective sleep studies show about THC?
Polysomnography (PSG) measures brain activity, eye movement, muscle tone and heart rhythm across a night's sleep. It is the reference standard for sleep architecture and it does not care how the patient felt in the morning. The PSG evidence on THC breaks into four areas.
Deep sleep: an early gain that does not last. Short-term THC administration increases slow wave (deep) sleep and reduces sleep onset latency. This is the honeymoon effect that drives early patient satisfaction. The review literature notes the effect is not persistent: with chronic administration, slow wave sleep decreases, which points to tolerance. Regular users then show longer sleep onset, more wake after sleep onset and reduced total sleep time compared with controls.
REM sleep: suppressed in older studies, mixed at therapeutic doses. Early controlled studies, mostly using higher THC doses, reported delayed REM onset and reduced REM sleep. A 2025 systematic review and meta-analysis in Sleep Medicine Reviews found that more recent studies with larger samples and lower therapeutic doses show mixed, and often no, REM suppression. The honest position is that REM suppression is dose-dependent and inconsistent. What is consistent is what happens on withdrawal: REM rebound, vivid dreams, prolonged sleep onset and reduced total sleep time. A drug that produces rebound when stopped has been suppressing something while it was taken.
Sleep continuity: fragmentation with chronic use. The largest objective dataset comes from a December 2025 cross-sectional PSG analysis of 1,449 adult sleep clinic patients, comparing 151 chronic cannabis users with 1,298 never-users. Chronic users had 21 per cent more wake after sleep onset (roughly 17 minutes), sleep efficiency 3.8 per cent lower, and a higher proportion of light N1 sleep. REM and N3 percentages did not differ significantly. The authors describe this pattern as more fragmented sleep, in a population where most participants were using cannabis with the intention of sleeping better. The cohort had a high rate of sleep apnoea, so the finding should be read with that in mind, but the direction is consistent with the tolerance data above.
Overnight HRV: reduced parasympathetic recovery. Heart rate variability (HRV) during sleep is a marker of vagal, or parasympathetic, control of the heart and a widely used proxy for overnight recovery. In a placebo-controlled study, 10 mg of THC taken one hour before bed reduced RMSSD, a standard HRV measure, by 22 per cent in cannabis-naive participants and 23 per cent in regular users, with corresponding reductions in high-frequency power. The researchers describe this as reduced vagal-cardiac modulation and possible increased cardiovascular stress during sleep. Notably, sleep staging did not change on the same night. Large consumer wearable datasets show the same direction: lower HRV, higher resting heart rate, more light sleep and slightly lower sleep efficiency on nights cannabis is logged, despite total sleep being marginally longer.
| Measure | What the patient reports | What objective studies show |
|---|---|---|
| Falling asleep | Faster, easier | Shorter sleep onset latency acutely; longer with chronic use and on withdrawal |
| Deep sleep | "Deeper" sleep | Increased initially; decreases with regular use (tolerance) |
| REM sleep | Fewer dreams, fewer nightmares | Reduced in older, higher-dose studies; mixed at therapeutic doses; rebound on cessation |
| Sleep continuity | "Slept through" | More wake after sleep onset, lower efficiency and more light sleep in chronic users |
| Overnight recovery | Feels rested | Reduced HRV and parasympathetic activity on THC nights |
Doesn't the Australian trial evidence show benefit?
Some of it does, and it is worth being precise about what it showed. The best-known Australian study, a 2021 randomised crossover trial of a THC-dominant sublingual oil (20 mg/mL THC with small amounts of CBN and CBD) in 23 adults with chronic insomnia, found a clinically meaningful drop in Insomnia Severity Index scores and modest objective gains on wrist actigraphy over two weeks: about 33 minutes more total sleep, 10 minutes less wake after sleep onset and a 2.9 per cent rise in sleep efficiency. Single-night PSG in the same trial showed minimal treatment effect.
Three things follow. First, this was a two-week trial, so it tells us nothing about the tolerance that the chronic-use literature describes. Second, actigraphy measures movement, not sleep stages, so it cannot see the architecture changes PSG detects. Third, a 2.9 per cent efficiency gain over a fortnight is not the same claim as "the patient's sleep is better and will stay better", which is the claim most long-term funding decisions rest on. The trial supports short-term symptomatic benefit in a selected group. It does not support indefinite prescribing without objective review.
What does this mean for medicinal cannabis in Australian claims?
Insomnia sits alongside chronic pain and anxiety as one of the three dominant indications for medicinal cannabis prescribing in Australia, and almost all of it occurs through the TGA's unapproved access pathways. No product is registered by the TGA for insomnia. In July 2025, Ahpra and the National Boards issued updated prescribing guidance stating there is little evidence to support medicinal cannabis for insomnia, anxiety or chronic pain, that it should not be prescribed first line for those conditions, and that prescribers must document an evidence-based indication, a mental health assessment, a monitoring plan and an exit strategy. That guidance followed enforcement action against more than 50 practitioners and cases of individual prescribers issuing more than 10,000 scripts in six months.
For a claims manager, the practical consequence is this. A THC prescription for sleep in a personal injury claim is an unapproved product being used for an indication the regulator regards as weakly evidenced, on the strength of a subjective report that objective studies say is unreliable, in a drug class where tolerance and dose escalation are documented. That is not an argument to decline every request. It is an argument that the usual claims test, whether treatment is reasonably necessary and producing a demonstrable benefit, cannot be satisfied by "the claimant says they sleep better" alone.
The same logic already applies to temazepam, zolpidem and zopiclone. The insurer cannot direct the prescriber, so the lever is what the insurer funds and what evidence it asks for before funding continues.
How should THC prescribed for sleep be monitored?
Objective monitoring does not require a sleep laboratory for every claimant. It requires that the treatment goal is written down, that something other than the claimant's impression is used to measure it, and that a review actually happens. A reasonable framework for a claims manager or independent reviewer looks like this.
| Question to ask | What good looks like | Red flag |
|---|---|---|
| What is the documented indication? | Insomnia with a stated cause, prior non-drug and conventional treatment recorded, sleep apnoea excluded | "For sleep" with no history, no CBT-I offered, no apnoea screen |
| What objective measure is used? | Validated sleep diary plus actigraphy or wearable data, or PSG where apnoea or fragmentation is suspected; Insomnia Severity Index at baseline and review | Benefit recorded only as the claimant's verbal report |
| What is happening to the dose? | Stable THC dose, oral or sublingual product, no vaporised flower | Escalating THC, switch to inhaled products, multiple prescribers on RTPM |
| Is function improving? | Daytime function, work capacity and psychological measures tracked alongside sleep | Better sleep reported but no change in capacity, mood or participation |
| What is the exit strategy? | Review date set, tapering plan documented, withdrawal insomnia anticipated and managed | Open-ended repeat prescribing; treatment continued because stopping made sleep worse |
The last row deserves emphasis. Because THC withdrawal reliably produces rebound insomnia and vivid dreams, a claimant who tries to stop and sleeps badly will often conclude, and tell their prescriber, that the medication was working. That is the tolerance and rebound cycle that keeps sedatives on claims for years. It should be anticipated and managed, not treated as proof of benefit.
Key Takeaways
- THC produces sedation and can shorten time to fall asleep in the short term. Sedation is not the same as restorative sleep.
- Objective studies show the early gain in deep sleep fades with regular use, chronic users have more fragmented and less efficient sleep, and THC before bed lowers overnight heart rate variability.
- REM suppression is real in older, higher-dose studies but inconsistent at therapeutic doses. REM rebound and worse sleep on stopping are consistent, and are a sign the drug has been altering normal sleep regulation.
- Ahpra's 2025 prescribing guidance says there is little evidence for medicinal cannabis in insomnia, and no Australian product is registered for it.
- Where THC is funded for sleep, pain or psychological symptoms, benefit should be measured objectively and reviewed on a schedule, not inferred from the claimant's perception alone.
- Rebound insomnia on stopping THC is a withdrawal effect, not proof the medication was working.
Frequently Asked Questions
Does THC improve sleep quality?
THC reliably produces sedation and can shorten the time taken to fall asleep in the short term. Objective sleep studies are far less positive on quality: deep sleep gains fade with tolerance, chronic use is associated with more fragmented and less efficient sleep, and overnight heart rate variability falls. Feeling sedated is not the same as sleeping restoratively.
Does THC suppress REM sleep?
Older, higher-dose studies found THC delayed REM onset and reduced REM sleep. A 2025 meta-analysis found the effect is mixed and often absent at lower therapeutic doses. What is consistent is REM rebound and disturbed sleep when THC is stopped after regular use, which is a sign the drug has been altering normal sleep regulation.
Is medicinal cannabis an evidence-based treatment for insomnia in Australia?
No Australian medicinal cannabis product is TGA-registered for insomnia. Ahpra's July 2025 prescribing guidance states there is little evidence to support medicinal cannabis for insomnia, anxiety or chronic pain, and that it should not be used first line for those conditions. Prescribing for sleep occurs almost entirely through unapproved access pathways.
How should a claims manager assess THC prescribed for sleep?
Ask for the treatment goal, the objective measure being used to judge it, the planned review date and the exit strategy. A claim where the only evidence of benefit is the claimant saying they sleep better, while the dose creeps up and function does not improve, does not meet a reasonably necessary test on its own. An independent clinical pharmacy review can establish whether the medication is doing what it is funded to do.
Primary sources: Kaul, Zee and Sahni, Neurotherapeutics 2021; Velzeboer et al, Sleep Medicine Reviews 2025 (meta-analysis) and SLEEP 2025 (sleep-clinic PSG cohort); Gonzalez et al, THC before bedtime and nocturnal cardiac autonomic activity, SLEEP 2023 abstract and Journal of Sleep Research 2026; Walsh et al, SLEEP 2021; Ahpra and National Boards medicinal cannabis prescribing guidance, July 2025; TGA medicinal cannabis access pathway guidance, 2026.
