Cannabis and sleep disorders: not ready for prime time? A qualitative scoping review

Abstract

Study Objectives:

To perform a qualitative scoping literature review for studies involving the effects of cannabis on sleep and sleep disorders.

Methods:

Two electronic databases, MEDLINE and EMBASE, searched for comprehensive published abstracted studies that involved human participants. Inclusion criteria were article of any type, published in English, a target population of cannabis users, and reported data on cannabis effect on sleep and sleep disorders. The Joanna Briggs Institute’s (JBI) approach was elected as the methodology framework guidance in the scoping review process.

Results:

A total of 40 unique publications were found. The majority (82.5%) were from the Americas with 60% published in the last decade. Of the 40 studies, only 25% were randomized control trials, and the sleep outcome measurements were similar and comparable in only 20%. Cannabis users studied were reported either 73% frequent users or 27% sporadic users. The utilization of cannabis showed improved sleep (21%), worse sleep (48%), mixed results (14%), or no impact at all (17%) in the studies published in the last 5 decades.

Conclusions:

Our findings summarize the lack of robust evidence to support the use of cannabis for sleep disorders. The varied cannabis user-related characteristics may account for the inconsistent results identified. Further studies assessing cannabis and sleep are needed to discern what works in what context, how it works, and for whom.

Citation:

Amaral C, Carvalho C, Scaranelo A, Chapman K, Chatkin J, Ferreira I. Cannabis and sleep disorders: not ready for prime time? A qualitative scoping review. J Clin Sleep Med. 2023;19(5):975–990.

INTRODUCTION

Cannabis sativa L is a plant found worldwide with more than 100 distinct derivatives described,^1–3 where the most prevalent ones that are used in products for consumers are the cannabinoids Δ-9-tetrahydrocannabinol (THC) and cannabidiol (CBD).^4–6 An estimated 200 million people used cannabis at least once in 2019. This represents almost 4% of the world’s population, with the highest consumption in the Americas (59 million recent users) followed by Africa (47 million recent users) in 2019.⁷

The acute exposure to THC (short-term use) decreases sleep onset latency, increases slow wave sleep (SWS), and decreases wake after sleep onset and rapid eye movement (REM) sleep. In contrast, long-term use has been shown to decrease SWS. Studies in humans are lacking to evaluate sleep-related effects of CBD. An increase in total percentage of sleep in rats injected with CBD is available where higher doses showed an increase in REM latency and lower doses showed a decrease.⁸

In Canada, the use of cannabis for medical reasons has been allowed since 1999. In October 2018, Canada became the second country in the world to legalize cannabis for nonmedical use, authorizing the drug’s sale and possession, trying to minimize harms associated with criminalization.⁹ There was an increase in cannabis users, from 9.1% of the country’s population in 2011 to 14.7% in 2015.¹⁰ Around 6.2 million people aged 15 or older (20% of Canadians in that age group) reported by the end of 2020 having used cannabis in the past 3 months. This was higher than both the population reporting use prior to legalization and during the first months after the Cannabis Act was enacted.¹¹ Cannabis remains a controlled substance in the United States at the federal level, while the illegal sale of cannabis continues to rise there.^12,13 This observation coincides with a significant increase in cannabis-related adverse events, including emergency room visits, hospital admissions, driving under the influence of cannabis,¹⁴ and traffic-related deaths.¹²

In addition to the rising number of recreational consumers, medicinal users of cannabis are also increasing in number worldwide. In a recent Canadian survey of cannabis users, 62% reported using it to treat medical conditions such as pain, anxiety, depression, sleep disorders, and epilepsy.^15,16 Although the therapeutic effects of cannabis remain uncertain for most disorders,^17,18 significant adverse effects of its use, such as the risk of addiction and mental illness, have been well described.¹⁹ There is limited evidence from well powered clinical trials that cannabinoids are effective therapy for sleep disorders, including insomnia and sleep apnea. A possible role of cannabis as an alternative for opioids has been suggested.^20,21 The evidence of benefit is weak for some clinical conditions, and some disorders may be worsened by cannabis use.^22,23 Given this background, we undertook a scoping review of the available literature to assess cannabis use and sleep, as this association seems to be poorly understood. This review has not examined the legal framework of medicinal and recreational use of cannabis nor the effect of cannabis on other organs and systems. The goal of this study is to perform a qualitative scoping literature review for studies involving the effects of cannabis on sleep and sleep disorders.

METHODS

The conduct of this scoping review was based on Chapter 11 of the Joanna Brigg’s Institute (JBI) Manual for Evidence Synthesis, where the framework and principles were reported by Peters et al²⁴ and included the following 5 key phases:

• Stage 1: Identifying the research question

• Stage 2: Identifying relevant studies

• Stage 3: Study selection

• Stage 4: Charting the data

• Stage 5: Collating, summarizing, and reporting the results

Search strategy

Two databases (MEDLINE and EMBASE) were searched using entry terms, like “cannabis” and “sleep disorders”, etc up to December 2020 (Table 1) structured according to the Problem/Population, Concept, and Context frame for scoping reviews as a knowledge representation for the clinical question posed in a natural language by the practicing physician. The search topic formulated was the impact of cannabis use on sleep. The Population was defined as cannabis users, the Concept was defined as the cannabis utilization compared to placebo groups or not, and the Context was defined as the impact on sleep parameters. The standards of quality for reporting meta-analyses and systematic reviews of observational studies were reviewed during the planning, conduct, and reporting of this study.³ We limited the review to publications in English and to human studies. The reference lists of the selected studies were combed for additional relevant publications.

Table 1.

Details of search strategy related to the concepts.

MEDLINEa (1946–Dec 2020)	EMBASEb (1947–Dec 2020)
Concepts Associated to Users	Concepts Associated to Users
cannabidiol, cannabinoids, cannabinol, Cannabis, dronabinol, marijuana abuse, marijuana smoking, medical marijuana, tetrahydrocannabinol	1 deoxy 3 (1,1 dimethylbutyl), delta-8 tetrahydrocannabinol, cannabidiol, cannabidiol derivative, cannabidiol dimethyl ether, cannabidiol methyl ether, cannabinoid, cannabinol, cannabinol derivative, Cannabis, cannabis addiction, cannabis derivative, Cannabis sativa, Cannabis sativa subsp. indica, cannabis smoking, cannabis use, delta-8 tetrahydrocannabinol, dexanabinol, dronabinol, lenabasum, medical cannabis, nabiximols, tetrahydrocannabinol, tetrahydrocannabinolic acid

Concepts Associated to Measures	Concepts Associated to Measures
polysomnography	apnea hypopnea index, polysomnography, respiratory disturbance index, sleep spindle, sleep quality, slow wave sleep, stage 1 sleep

Concepts Associated to Outcomes	Concepts Associated to Outcomes
dyssomnias; insomnia; parasomnias; REM sleep behavior disorder; restless legs; sleep; sleep apnea, central; sleep apnea, obstructive; sleep apnea, syndromes; sleep bruxism; sleep disorders, circadian rhythm; sleep wake disorders; sleep, REM	central sleep apnea, circadian rhythm sleep disorder, insomnia, narcolepsy, parasomnia, REM sleep, REM sleep deprivation, sleep quality, sleep bruxism, sleep deprivation, sleep disorder, sleep disordered breathing

^a Medical Subject Headings (MeSH) terms and accompanying entry terms were exploded in search. ^bEmtree-preferred terminology; all terms exploded in search. REM = rapid eye movement.

Eligibility criteria

Study inclusion criteria are shown in Table 2. Initially, titles and abstracts were screened independently by 2 reviewers (C.P.A. and C.G.C.). Selected articles from this screening underwent subsequent independent full-text reviews.

Table 2.

Study inclusion criteria.

1) Number of patients are cannabis users. Studies focused on cannabis withdrawal or designed to assess variables other than sleep were not included.

2) Cannabis users were adults (defined as at age 18 or higher)

3) Sleep characteristics or sleep disorders described among cannabis users

4) Review articles, posters, opinion pieces, and published abstracts were not included.

Quality assessment and data retrieval

Following the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) statement^25,26 recommendations, methodological quality assessment was conducted in consensus by the 2 reviewers with duplicates removed (Figure 1). A measurement tool to appraise systematic reviews of both randomized and nonrandomized studies was used and clinical trials, cross sectional studies, longitudinal surveys, and retrospective chart reviews describing either sleep characteristics or sleep disorders among cannabis users could be included in the analysis. The AMSTAR2 tool²⁷ is a 16-point assessment instrument to facilitate the development of high-quality reviews. We defined a representative spectrum of patients as current cannabis users (age 18 or older) to reflect a representative population. Data extractions also included type of study design: descriptive, eg, cross-sectional survey, or analytic, eg, experimental (clinical trials), or observational, the sampled population, cannabis use, and sleep characteristics. Data extraction was performed independently by 2 reviewers (C.P.A. and C.G.C.). Discrepancies were resolved by consensus.

Figure 1. Selection of studies for this review.

A trial flow diagram shows the number of identified, screened, and included studies.

Sleep analysis and sleep disorders

Studies varied in their use of either objective polysomnographic findings or patient-reported outcomes. The parameters for sleep analysis included reduced or increased sleep latency, decreased time spent awake after falling asleep, increased or decreased total sleep time, and changes in stage 3 or reported time in rapid eye movement sleep. The studies were categorized relative to the overall quality of sleep as improved sleep, worse sleep, or mixed results. The data available for sleep disorders insomnia, obstructive sleep apnea (OSA), and restless legs syndrome (RLS) were recorded.

Insomnia was defined as the self-reported perception of difficulty with sleep initiation, duration, consolidation, or quality that occurred despite adequate opportunity for sleep, and resulting in some form of daytime impairment.²⁷ OSA was defined as recurring episodes of cessation (apnea) or reduction (hypopnea) in airflow during sleep caused by obstruction of the upper airway.^28–39 RLS was defined as movement disorder characterized by an urge to move the legs or arms, commonly in response to uncomfortable dysesthesia.³⁰ RLS can impact sleep quality and sleep quantity³² as it may cause involuntary movements during sleep, denominated periodic leg movements of sleep.

Statistical analysis

Descriptive statistics were used. Contingency tables were created using a spreadsheet (Microsoft Excel, 2018). Continuous variables were described using mean plus or minus standard deviation or median and range and categorical variables, using frequency and percentage.

RESULTS

The initial search dated from January 1946 to December 2020 using 2 databases (MEDLINE and EMBASE) identified 389 articles. Of those, 3 articles were duplicated and excluded. After a manual bibliography search from the remaining 386 nonduplicated studies, 4 more articles were included. From the total of 390 articles, 258 were excluded upon review of the title or abstract based on inclusion and exclusion criteria. After full-text review, 92 additional articles were excluded. Reasons for further exclusion were described in Figure 1. This study cohort had 40 articles, where 28 examined the effects of cannabis on sleep,^33–60 4 evaluated on both sleep and chronic pain,^61–64 3 described the effects of cannabis on sleep and posttraumatic stress disorders,^65–67 3 examined the effects of cannabis solely on OSA,^68–70 and 2 studies described cannabis for RLS.^31,71,72

Of the 40 articles, only 25% (10 of 40) were randomized clinical trials (RCT). The majority 82.5% (33 of 40) were performed in the Americas, 15% (6 of 40) in Europe, and 2.5% (1 of 40) in the Oceania. No studies were found in Africa or Asia. The distribution within the Americas shows 94% of the studies from North America and 6% equally distributed between Central (3%) and South America (3%). Among the North American countries, no studies were from Mexico, 20% from Canada and 80% from the United States. The 40 studies^33–72 described sleep in 8,434 patients with a median age of 32 years, standard deviation 10.86, where 61% were males and 39% females. Details on the consumption of cannabis were reported for 5,484 patients in 21^{33–43,45,46,53,54,57,60} of the 40 studies, with 41% categorized as daily users, 32% as weekly users, and 27% as sporadic users.

There were 11 (39%) experimental clinical trial studies (Table 3A) and 17 (61%) observational investigations (Table 3B) evaluating the effects of cannabis on overall sleep, with most employing self-reported questionnaires (surveys). Those 28 studies^37–64 reported improved sleep (18%), worse sleep (50%), mixed results (14%), or no impact at all (18%).

Table 3A.

Experimental studies investigating the effect of cannabis on sleep in chronological order of their publication.

First Author, Publication Year	Country	Study Description	Sleep Measure	Control Group	Participants* Total Number (male %); Average Age*	Cannabis Frequency*	Sleep Outcome Reported
Pivik 197234	USA	Double-blind study, oral THC (4.3–17 mg) or synhexl administered to undisturbed and REM deprived volunteers. During the basal study there were 3–6 baseline nights and 1–4 recovery nights, with 1 drug night for each participant.	PSG	Placebo	4 (100%); N/A	Casual users or naïve participants	Increments in stage 4 sleep and total sleep time, with a slight decrement in time in REM sleep. Among the undisturbed group, reduction in stage 1 and time awake after sleep onset were observed at the highest THC dose level.
Cousens 197335	USA	Randomized, double-blind, single-dose oral THC, 10 mg, 20 mg or 30 mg in inducing and maintaining sleep in patients with insomnia symptoms tested 1×/wk for a 6-week period	Questionnaire; direct participant observation every 15 min throughout the night	Placebo	9 (100%); 21-40 years	Presumed occasional users	THC decreased “total time to fall asleep” (P < .05). Once asleep, interruptions of sleep were not significantly altered over the night (P > .05). Next-day perceptual disruption and thought alterations with 30 mg THC (“hangover effect”).
Halikas 198538	USA	Longitudinal survey analyzing cannabis effects, over follow-up of 6–8 years	Questionnaire	N/A	97 (61%); 22 years	Regular users	Desirable after-effects (restful sleep; awake refreshed; more sleep) initially reported as more than occasional, decreased over time.
Chait 199039	USA	Longitudinal assessment of smoked THC 2.1% (or placebo) over 2 weekends of sleep in a clinical setting	LSEQ; Expired air carbon monoxide	Placebo	12 (75%); 21 years	Regular cannabis smokers	Easier to get to sleep.
Chait 199440	USA	Longitudinal assessment of smoked THC 3.1% (or placebo) over 5 nights of sleep in a clinical setting	LSEQ	Placebo	14 (60%); 24 years	Current cannabis and alcohol users	Easier to fall asleep; suggestive of decreased sleep quality if concomitant use of alcohol.
Nicholson 200441	UK	Double-blind study with a 4-way crossover design of oromucosal spray (placebo, THC 15 mg, THC 5 mg combined with CBD 5 mg, and 15 mg THC combined with CBD 15 mg) administered at least a week apart	PSG; SSS; Sleep Latency Test	Placebo	8 (50%); 21.8 years (females); 28.8 years (males).	Occasional users, abstinent for at least 30 days	Concomitant THC and CBD decreased stage 3 sleep while higher dose combination increased wakefulness. THC 15 mg was associated with a decrease in sleep latency, and increased sleepiness in the next day.
Bedi 201043	USA	Double-blind study assessing dronabinol 10 mg for two 16-day inpatient stays in a residential laboratory, separated by a 5- to 15-day outpatient phase, in patients with HIV	Sleep monitor; self-reported sleep quality scale	Placebo	7 (100%); 37 years	Regular cannabis smokers	Sleep efficiency improved in the first 1–8 days only, due to an increase in time spent in non-REM sleep and a decrease in time spent awake. Greater satisfaction with sleep and fewer awakenings on days 1–8.
Ware 201044	Canada	RCT comparing nabilone (0.5–1.0 mg) to amitriptyline (10–20 mg) administered for 2 weeks in patients with fibromyalgia and chronic insomnia	LSEQ; ISI	Active-control, equivalency crossover	31 (16%); 49.5 years	Negative urine screen for cannabinoids at the baseline visit	Nabilone and amitriptyline improved sleep, but nabilone was superior on the ISI (adjusted difference-3.25%; 95%CI-5.26 to -1.24). On restfulness, nabilone was marginally better.
Gorelick 201346	USA	Longitudinal evaluation of around‐the‐clock dosing with oral dronabinol 40–120 mg daily for 7 days	SMQ; Plasma THC and 11‐OH‐THC	N/A	13 (100%); 25 years	Daily smokers	Shorter sleep latency, more daytime sleep the following day. Slight decrease in nighttime sleep over the study.
Belendiuk 201548	USA	Cross-sectional survey assessing the preceding 30 days of adults purchasing cannabis for a health condition (49% for insomnia and 9% for nightmares) at a retail store	PSQI	N/A	163 (N/A); 40 years	Regular users	No association between frequency of cannabis use and nightmares or insomnia.
Whitehurst 201550	USA	Longitudinal study of cannabis users assessed by questionnaire and actigraph for 3 weeks	MEQ; actigraph; urine sample	Nonusers	30 (43%); 21 years	Daily cannabis users	No differences on sleep/rest parameters or sleep disruption.
Linares 201854	Brazil	Double-blind RCT studying CBD 300 mg administered to patients who spent 2 nights in the sleep lab (crossover)	PSG, PSQI, ESS	Placebo	26 (46%); 29 years	Infrequent users (abstinent > 1 year)	No differences in polysomnographic findings and self-reported measures of sleep.
Shannon 201959	USA	Case series, CBD (mostly 25 mg/d); some were given 50 or 75 mg/d; 1 was given 175 mg/d for 3 months, for anxiety or sleep, at a psychiatric outpatient clinic	PSQI	N/A	72 (60%); 34 years	N/A	Sleep scores improved in 66% of patients within the first month of starting CBD, while 25% experienced worsening of symptoms.

^*For each study the percentage of male participants, the averaged population age in years, and frequency of cannabis use status are presented when they were provided. CBD = cannabidiol, ESS = Epworth Sleepiness Scale, FSS = Fatigue Severity Scale, HIV = human immunodeficiency virus, ISI = Insomnia Severity Index, LSEQ = Leeds Sleep Evaluation Questionnaire, MEQ = Morningness–Eveningness Questionnaire, N/A = not applicable, PSG = polysomnography, PSQI = Pittsburgh Sleep Quality Index, RCT = randomized controlled trial, REM = rapid eye movement, SMQ = St. Mary’s Hospital Sleep Questionnaire, SSS = Stanford Sleepiness Scale, THC = tetrahydrocannabinol, UK = United Kingdom, USA = United States of America.

Table 3B.

Observational studies investigating the effect of cannabis on sleep in chronological order of their publication.

First Author, Publication Year	Country	Study Description	Sleep Measure	Control Group	Participants* Total Number (male %); Average Age*	Cannabis Frequency*	Sleep Outcome Reported
Tart 197033	USA	Cross sectional survey about cannabis use over the preceding 6 months	Questionnaire	None	150 (67%); 87% under age 30 years	42% ≥ 1×/wk; 39% < 1×/wk; 19% almost every day.	Easier to fall asleep, drowsiness, refreshing sleep, vivid dreams.
Feinberg 197636	USA	Longitudinal study of patients who agreed to be hospitalized on a psychiatric ward for 21–32 days and who were administered cannabis extract (29% THC, 1.5% cannabinol, and 2.8% CBD) or THC (70–210 mg)	PSG	Placebo	4 (100%); 22–31 years	Experienced users	Both drugs reduced eye movement density with some tolerance developing to this effect. Stage 4 tended to increase.
Karacan 197637	Costa Rica/Mexico	Longitudinal assessment of cannabis smokers who spent 8 consecutive nights of sleep in the lab (usual routine during the day)	PSG	Non-users	32 (100%); 30 years	Regular cannabis smokers	Slight increase in sleep latency, REM percentage, and REM period length.
Halikas 198538	USA	Longitudinal survey analyzing cannabis effects, over follow-up of 6–8 years	Questionnaire	N/A	97 (61%); 22 years	Regular users	Desirable after-effects (restful sleep, awake refreshed, more sleep) initially reported as more than occasional, decreased over time.
Fisk. 200942	UK	Cross-sectional survey assessing ecstasy and cannabis users	ESS; KSS	Drug naïve	227 (42%); 21 years	51% ecstasy/polydrug users; 23% cannabis-only users; 26% drug naïve	Ecstasy/polydrug users reported poorer sleep quality than drug-naïve individuals.
Babson 201345	USA	Cross-sectional analysis of adults prescribed medical cannabis for anxiety (60%), chronic pain (59%), insomnia (48%), and depression (43%)	PSQI	N/A	162 (78%); 42 years	Regular users	Depression symptoms related to cannabis use disorder were apparently moderated by perceived quality of sleep.
Ly 201347	USA	Cross-sectional analysis of the preceding 30 days of patients with Attention-Deficit/Hyperactivity Disorder	PSQI	N/A	76 (74%); 27 years	Cannabis users and nonusers	Cannabis was associated with decreased sleep quality in women but not in men.
Belendiuk 201548	USA	Cross-sectional survey assessing the preceding 30 days of adults purchasing cannabis for a health condition (49% for insomnia and 9% for nightmares) at a retail store	PSQI	N/A	163 (N/A); 40 years	Regular users	No association between frequency of cannabis use and nightmares or insomnia.
Ogeil 201549	Australia	Cross-sectional online survey	PSQI, ESS	Self-reported “non-risky” alcohol and cannabis use	248 (47%); 26 years	34% “risky” cannabis users; 61% “risky” alcohol users; 53% both cannabis and alcohol “risky users”	“Risky” alcohol and cannabis users had poorer sleep quality than those with no “risky use”.
Conroy 201651	USA	Cross-sectional survey assessing the effects of cannabis in the preceding 4 weeks	PSQI, ISI, ESS, MEQ, urine sample	Non-users	98 (46%); 22 years	50% daily users; 30% nondaily cannabis users; 20% nonusers.	PSQI scores were higher in daily cannabis users than nondaily users.
Maple 201652	USA	Cross-sectional survey assessing the effects of cannabis in the preceding year	PSQI	N/A	41 (63%); 18–25 years	Cannabis users abstinent for 7 days prior to the study sessions	Dose-dependent relationship between increased past-year cannabis use and greater past-month self-reported sleep problems.
Pacek 201753	USA	Cross-sectional evaluation of individuals seeking treatment for cannabis use.	Home PSG; PSQI; ISI; DBAS-10	N/A	87 (62%); 31 years	Frequent cannabis users	Overall, disordered sleep and poor sleep quality were frequent. No significant differences in PSG or self-reported sleep quality.
Altman 201955	USA	Cross-sectional online survey	PSQI	None	311 (55%); 42 years	Cannabis users	Sleep-problems correlated with amount of cannabis used. Positive cannabis-induced sleep expectancies were associated with decreased cannabis-related problems.
Bachhuber 201956	USA	Cross-sectional electronic survey with customers at cannabis retail stores taking it for sleep (74%) and pain (65%)	Questionnaire	Not controlled	1,000 (58%); 50% under age 50 years	Cannabis users	Easier to fall asleep; reduced the number of medications used to sleep.
Casarett 201957	Canada	Retrospective cohort study of patients in palliative care prescribed vaporized THC and CBD over about 1 year	Questionnaire	N/A	2,431 (43%); 34 years	Regular users for palliative care	Improvement of insomnia symptoms with increased THC:CBD ratio.
Goodhines 201958	USA	Longitudinal 14-day assessment of college students using cannabis and/or alcohol as sleep aid	PSQI; FSS; Insomnia status; Expanded Consensus Sleep Dairy for morning.	N/A	217 (24%); 19 years	31% cannabis sleep aid users	Longer same-night sleep duration, shorter same-night wake time after sleep onset, greater next-day daytime fatigue.
Winiger 201960	USA	Analysis of the influence of retrospectively assessed age of onset of regular cannabis use on adult sleep duration	Jessor Health Questionnaire	Occasional cannabis use	1,656 twins (44%); 26 years	Regular users	Earlier age of onset for regular cannabis use was significantly associated with shorter adult sleep duration.

^*For each study the percentage of male participants, the averaged population age in years, and frequency of cannabis use status are presented when they were provided. CBD = cannabidiol, DBAS-10 Dysfunctional Beliefs and Attitudes about Sleep–10, ESS = Epworth Sleepiness Scale, FSS = Fatigue Severity Scale, ISI = Insomnia Severity Index, LSEQ = Leeds Sleep Evaluation Questionnaire, KSS = Karolinska Sleepiness Scale, MEQ = Morningness–Eveningness Questionnaire, N/A = not applicable, PSG = polysomnography, PSQI = Pittsburgh Sleep Quality Index, REM = rapid eye movement, THC = tetrahydrocannabinol, UK = United Kingdom, USA = United States of America

Table 4 shows 75% RCT and 25% cross-sectional design studies of cannabis investigated as an alternative medicine for patients with insomnia and chronic pain distributed equally between North America and Europe. Of the 4 studies, 3 reported improvement in sleep^61–63 with the remaining study reporting⁶⁴ mixed results.

Table 4.

Selected characteristics, study designs, and major findings of the included studies in this scope review about cannabis effects in participants with chronic pain and their sleep.

First Author, Publication Year	Country	Study Description	Sleep Measure	Control Group	Participants* Total Number (male %); Average Age*	Cannabis Frequency*	Sleep Outcome Reported
Notcutt 200461	UK	N-of-1 RCT comparing sublingual spray THC, CBD and 1:1 mixture of them (crossover) for 12 weeks, for patients with chronic, stable pain poorly responsive other treatments	Self-reported duration of sleep and quality of sleep (Good, Fair, Poor) for each night	Placebo	34 (22%); 47 years	Current or previous experience of using cannabinoids. 47% multiple sclerosis	Little effect on the recorded number of hours of sleep; high prevalence of self-reported change in quality of sleep from “poor” or “fair” to “good”.
Rog 200562	UK	RCT comparing follow-up of patients at 5 weeks undergoing oromucosal spray THC:CBD as an adjunctive analgesic treatment	11-point numerical rating scale recording sleep disturbance due to neuropathic pain rating	Placebo	66 (22%); 49 years	18% used cannabis within 3 months of study entry; multiple sclerosis	Cannabis reduced pain and self-reported sleep disturbance.
Ware 201063	Canada	RCT comparing follow-up of patients at 14 days undergoing smoked THC for neuropathic pain (crossover)	Leeds Sleep Evaluation Questionnaire	Placebo	23 (48%); 45 years	82% reported previous cannabis use	THC was associated with drowsiness, being easier and faster to fall asleep, and with fewer periods of wakefulness.
Cranford 201764	USA	Cross-sectional survey examining adults seeking medical cannabis for severe chronic pain	Jenkins Sleep Problems Questionnaire	None	801 (52%); 46 years	92% used cannabis in the last 6 months	Cannabis improved self-reported sleep among patients with chronic pain. The frequency of cannabis use to help sleep was associated with a higher risk of sleep problems.

^*For each study the percentage of male participants, participant age (mean) in years, and cannabis use status are shown where provided. CBD = cannabidiol, RCT = randomized controlled trial; THC = Tetrahydrocannabinol; UK = United Kingdom; USA = United States of America.

The effect of cannabis on sleep and post-traumatic stress disorder are included in the Table 5. Three Canadian studies (2 RCTs) with predominantly male populations were conducted showing either more refreshing sleep^65,66 or not⁶³; with⁶⁶ or without^65,67 increments in total sleep and with^65,66 or without⁶⁷ decrease of nightmares in individuals using nabilone.

Table 5.

Studies evaluating cannabis effects in the sleep of participants with posttraumatic stress disorders.

First Author, Publication Year	Country	Study Description	Sleep Measure	Control Group	Participants* Total Number (male %); Average Age*	Cannabis Frequency*	Sleep Outcome Reported
Fraser 200965	Canada	Open label clinical trial with oral nabilone (0.5–4 mg) for treatment-resistant nightmares in patients with PTSD	Questionnaire	None	47 (43%); 44 years	N/A	Nightmares stopped or significantly reduced in intensity.
Cameron 201466	Canada	Retrospective chart review of inmate patients with PTSD who were prescribed oral nabilone (0.5–6 mg) for an average of 11 weeks	Sleep hours per night, nights with nightmares per week	N/A	104 (100%); 33 years	Nabilone non-user, 91% had prior cannabis use disorder	Increase in the average number of hours slept. Improvement in insomnia and nightmares.
Jetly 201467	Canada	Double-blind trial (crossover) with oral nabilone (0.5–3 mg) or placebo for nightmare suppression, for 7 weeks, in military personnel with PTSD	Clinician-administered PTSD Scale, Recurring and Distressing Dream Scores	Placebo	10 (100%); 44 years	Cannabis non-users	Nabilone reduced frequency and intensity of distressing dreams. No effect on self-reported sleep quality and quantity.

^*For each study the percentage of male participants, participant age (mean) in years, and cannabis use status are shown where provided. N/A = not available, PTSD = posttraumatic stress disorder.

Cannabis has been explored as an option for treating patients with OSA (Table 6) and used in patients with RLS to improve their sleep quality.^71,72 In clinical practice, patients with OSA frequently ask if there is an “oral pill” to treat their apnea instead of continuous positive airway pressure. In 2013, a proof-of-concept study⁶⁸ evaluated the dronabinol (an oral THC) safety in 17 patients with OSA. A significant reduction in in the apnea-hypopnea index (AHI) was noted in 15 participants who completed 21 days of treatment with dronabinol as per protocol, where the change in the AHI at week 3 of dronabinol treatment compared to baseline was reported as mean ± standard deviation: −14.1 ± 17.5, P = 0.003 and with 95% confidence interval of −23.8, −4.4. There was no significant change in sleep efficiency, arousal index, and duration of REM or slow wave compared to baseline. No degradation of sleep architecture or serious adverse events were noted. Prasad et al⁶⁸ reported the Stanford Sleepiness Scale scores of 3.4 ± 1.6 (mean ± standard deviation) at baseline, which improved to 2.3 ± 1.6 (P = .05). Despite the improvements in AHI, dronabinol did not lead to numerical improvement in the nadir oxygen saturation or in the time of oxygen saturation below 85%. Further analysis of this population investigated by Prasad et al in 2013 was published by Farabi et al⁶⁹ in 2014 focusing on the effects of dronabinol on quantitative electroencephalogram markers of the sleep process, including power distribution and ultradian cycling. This 2014 study⁶⁹ suggests that in patients with OSA, dronabinol could yield a shift in electroencephalogram power toward delta and theta frequencies and a strengthening of ultradian rhythms in the sleep electroencephalogram. It is possible that dronabinol could have improved restorative sleep.

Table 6.

Studies investigating cannabis effects in participants with obstructive sleep apnea.

First Author, Publication Year	Country	Study Description	Sleep Measure	Control Group	Participants* Total Number (male %); Average Age*	Sleep Outcome Reported
Prasad 201368	USA	Open label, oral dronabinol (2.5–10 mg) administered for 3 weeks for patients with moderate to severe OSA	Polysomnography; Stanford Sleepiness Scale	None (compared to baseline)	17 (35%); 52 years	AHI decreased from baseline to night 21 (−14.1 events/h; P = .003). Sleep architecture not disturbed. Oxygen saturation unchanged.
Farabi 201469	USA	Open label, oral dronabinol (2.5–10 mg/d) administered for 3 weeks to patients with moderate to severe OSA	Polysomnography	None (compared to baseline)	15 (40%); 52 years	Shift in EEG power toward delta and theta frequencies (P = .002) and a strengthening of ultradian rhythms, suggesting that dronabinol may have improved restorative sleep.
Carley 201870	USA	Double-blind RCT of dronabinol (2.5 or 10 mg/d) for 6 weeks for patients with moderate or severe OSA	Polysomnography; Epworth Sleepiness Scale Score; Maintenance of Wakefulness Test	Placebo	73 (71%); 55 years	AHI decreased, self-reported sleepiness improvement, and greater overall treatment satisfaction. Objective sleepiness did not change from baseline in any treatment group.

^*For each study the percentage of male participants, participant age (mean) in years, and cannabis use status are shown where provided. AHI = apnea-hypopnea index, EEG = electroencephalogram, OSA = obstructive sleep apnea, RCT = randomized controlled trial, UK = United Kingdom, USA = United States of America.

A group of researchers conducted in 2018 the PACE (Pharmacotherapy of Apnea by Cannabimimetic Enhancement) study, a phase II, placebo-controlled trial, with fully blinded parallel groups, evaluating dronabinol in patients with moderate to severe OSA.⁷⁰ In comparison to placebo, dronabinol reduced AHI by 10.7 ± 4.4 (P = .02) and 12.9 ± 4.3 (P = .003) events/h at doses of 2.5 and 10 mg/d, respectively. Dronabinol at 10 mg/d reduced Stanford Sleepiness Scale score by −3.8 ± 0.8 points from baseline (P < .0001) and by −2.3 ± 1.2 points in comparison to placebo (P = .05). The maintenance of wakefulness test sleep latencies, gross sleep architecture, and overnight oxygenation parameters were unchanged from baseline in any treatment group.⁷⁰ However, there was an increase in the number of events from baseline in the placebo group, accounting somewhat for the difference between groups. There was an improvement in self-reported but not objective daytime sleepiness. Despite suggesting clinically meaningful and statistically significant effects of dronabinol in reducing AHI and increasing daytime alertness, Carley et al⁷⁰ observed considerable interparticipant variability in the magnitude of these effects.

Moreover, when considering the use of cannabis for OSA treatment, it is necessary to take into account the studies that indicate an association of car accidents with dronabinol⁸⁹ and smoked cannabis.⁹⁰ Nighttime administration of THC alone or in combination with CBD is associated with increased sleepiness, memory impairment, and mood change in the morning post-dose.^44,87,90 In this regard, continuous positive airway pressure is known as an essential tool to reduce motor vehicle accidents in patients with OSA.^91,92 Another issue associated with cannabis use is the risk of hyperphagia,⁷⁰ since weight gain is directly linked to OSA worsening. The PACE trial⁷⁰ suggested that participants had no significant weight gain in 6 weeks, but long-term studies would be necessary to confirm these results.

Cannabis use has been reported to be associated with the treatment of refractory RLS.^71,72 All patients (n = 6) in the first publication⁷¹ complained of poor efficacy and/or poor tolerance to their current RLS medication and self-reported that the use of cannabis (eg, recreational cannabis smoking or sublingual administration of CBD) improved RLS symptoms and sleep quality. Nausea was also associated with cannabis use.⁷¹ A short communication⁷² showed total relief of patient’s RSL symptoms while smoking cannabis in 92% (11 of 12 patients) of their cohort. However, as pointed out by French authors,^71,72 the self-reported evaluation of the benefit of cannabis use could have been distorted by the psychoactive properties of cannabis. We believe that the effectiveness and safety of cannabis to treat patients with RLS and periodic leg movements have yet to be studied in clinical trials designed with adequate power to detect statistical significance.

DISCUSSION

We identified 40 studies addressing the effects of cannabis use on sleep published between 1970 and 2020 in this scoping review, and they demonstrated a paucity of robust evidence to support that cannabis should be prescribed for use in patients with sleep disorders. Our findings suggest conflicting results as there were studies demonstrating cannabis without significant effect on sleep parameters^48,50,53,54 or associated with a dose-dependent effect on sleep with either increased sleep latency³⁷ or reduced sleep latency.^{33,35,39–41,46,56} Studies showed decreased time spent awake after falling sleep^37,47,62 with either increased sleep time^38,62 or decreased sleep time.⁶⁴ Studies showed either increased time in REM sleep³⁷ or decreased time in REM sleep.³⁴ As well, studies reported greater satisfaction with sleep^43,59 or overall worse quality of sleep^{47,51,52,55,59} and decreased time in stage 3 sleep.⁴¹ Interestingly, some of the desirable effects sought by cannabis users, such as longer sleep time and refreshing sleep, have decreased over time.^38,45 Increased sleepiness, drowsiness, and fatigue were described by cannabis users in the day after its use.^33,41,46,58 The association between cannabis and other substances was analyzed and showed worse sleep quality among those consuming concomitant alcohol^40,42,49 and ecstasy.⁴²

Since the discussion toward cannabis legalization started to emerge in Canada, clinicians increasingly faced questions from patients if they could use cannabis to treat sleep disorders. Studies on Table 4 (75% RCT) investigating cannabis as an alternative medicine for patients with insomnia and chronic pain were equally distributed between North America and Europe. Those RCTs^61–63 included a predominantly female population (74%) with an average age of 47 years that ranged from 23 to 66 individuals in the 3 trials assessing tetrahydrocannabinol and cannabidiol for a maximum of 12 weeks for neuropathic pain, mainly among patients with multiple sclerosis. These studies suggest a significant self-reported improvement in sleep quality and time spent to fall asleep. A cross-sectional survey study included 801 participants with severe chronic pain, 92% of individuals with previous use of cannabis in the last 6 months.⁶⁴ That survey’s results showed cannabis improving self-reported sleep quality; however, habitual cannabis use is more often associated with a higher risk of sleep problems. The design of the studies does not allow one to determine if the relationship is causative or confounded by indication. Some patients are unaware of possible adverse effects and have a misperception that cannabis is totally harmless. Accordingly, it is crucial for clinicians to acknowledge what is available in the literature to manage these patients according to the current clinical evidence.

Cannabis has attracted media raising the expectation of both patients and physicians for its use in insomnia. However, its current use in insomnia is largely based on clinical experience; scientific evidence (including RCT) is scarce and the literature is conflicting.^73–75 Linares et al⁵⁴ observed the acute effects of CBD on sleep and did not find significant differences in the polysomnographic parameters compared to placebo. This perspective raises the unanswered question of whether cannabis would have less effect on sleep architecture than usual medications used for insomnia (such as benzodiazepines and selective serotonin reuptake inhibitors).^54,76–78 A meta-analysis published in 2015 found moderate quality evidence to support the use of cannabinoids for treatment of chronic pain and spasticity but low/very low-quality evidence that cannabinoids may improve sleep disorders.⁷⁹ That review included 2 studies that evaluated cannabinoids (nabilone) specifically for sleep disorders and 19 studies that used cannabinoids for other indications (chronic pain and multiple sclerosis) besides sleep.⁷⁸ In 2019, Kuhathasan et al⁸⁰ published a review evaluating the use of cannabis for sleep. They found studies suggesting that nabiximol (1 THC:1 CBD) would improve self-reported sleep.^81–86 However, most of these studies did not employ a validated sleep measure or use objective techniques, and many of them were of small sample size.⁸⁰

While THC has been associated with decreased sleep latency in some studies,^41,48 chronic use of THC can lead to tolerance.^37,86 This could explain the lack of improved sleep if the amount of drug is used equal to or lower than the amount customarily consumed.⁸⁸ The available literature suggests that the use of cannabis for insomnia may offer questionable benefit, but the current use is still based on low-quality evidence, small sample sizes, and sleep used as a secondary outcome.⁷⁹ Its use should be considered carefully and case by case. The decision must also be shared with the patient, and the risks that cannabis use may entail should be explained. Side effects associated with cannabis use should be compared with other available medications for insomnia. It is worth mentioning that insomnia symptoms are commonly reported among cannabis users who abruptly stop or decrease their intake.⁸⁷

In June 2018, the Minnesota Department of Health included OSA as a possible disease treatable with cannabis derivatives.⁹³ As a response, the American Academy of Sleep Medicine (AASM) published a statement advising against the use of cannabis to treat OSA, also informing state legislators, regulators, and health departments that OSA should not be included as an indication for their medical cannabis programs.⁹⁴ It triggered a backlash from some authors voicing their concerns regarding the AASM’s conservative position and strongly criticizing the AASM’s statement. Subsequently, the AASM issued a new response article cautioning against the use of cannabis for OSA, stating that the current evidence is not yet sufficient to treat this disease. The subject is still under discussion,^95–98 and more studies are required.^22,99

The present scoping review shed light on the lack of concordance among the current evidence about cannabis effects on sleep. Moreover, a wide variety of research methods investigating the effects of cannabis on sleep was observed. The conflicting findings observed among the 40 investigations reviewed can be associated with several factors: design, drug-related, user-related, etc. Study designs were variable, ranging from case reports with very small sample size to power-adequate RCT, with consequent implications as to their scientific relevance. The cannabis dose was not standard among the studies, eg, THC:CBC ratio, as well the route and length of administration described. Cannabis use was reported as widely variable, eg, previous exposure to cannabis vs first time, as was the purpose of use, recreational vs medicinal. The individual effect and tolerance to cannabis was also not standard. Not all studies reported the sex of the cohort or presence of comorbidities. The sleep measurements (eg, questionnaires, polysomnography, etc) were variable. The environment described in the studies, eg, cannabis used alone vs in-group was also not standardized. Despite the extensive changes in policy and the rapid rise in the use of cannabis both for medical purposes and for recreational use, conclusive evidence regarding the short- and long-term health effects (harms and benefits) of cannabis use remains elusive.^22,100 Scarce prospective, randomized, scientific research has resulted in a shortage of information on the health implications of cannabis use, which is a significant public health concern for vulnerable populations.¹⁰¹ Although small studies and retrospective analysis have straightened out the risks of cannabis use,^{19,22,23,87,100,102,103,104,105,106} there is virtually no multicenter, robust randomized control trial assessing individuals with chronic disorders, such as chronic obstructive pulmonary disease (COPD) and cardiovascular diseases in Canada. Despite that, a retrospective population-based study with large cohort (n = 185,876) in Ontario²³ showed patients with COPD receiving a high-dose of cannabinoids had increased rates of all-cause mortality (HR 3.31; 95% CO 1.30–9.51) as well hospitalization for COPD and pneumonia (HR 2.78; 95% CI 1.17–7.09). It is expected that cannabis use (as already demonstrated in the general population in Canada and the United States¹⁰⁷) will increase among patients with chronic disorders, possibly adding more risks to their current diseases. Large RCTs addressing potential confounders and clinical implications are necessary.

This article has limitations as the specific statistical tests commonly used for a systematic (meta-analysis) review were not addressed in this scoping review. We believe that well-designed and robust RCTs evaluating the effect of cannabis on sleep are urgently needed. Moreover, despite strict inclusion and exclusion criteria, some studies present in this analysis applied self-reported parameters of sleep. We understand that these investigations contributed to the current knowledge of cannabis and sleep besides functioning as hypothesis generating research. Nevertheless, this scoping review shows several strengths. This is one of the largest literature reviews assessing the effects of cannabis on sleep that has been systematically performed. First, we specified the frameworks that we followed, depicting inclusion and exclusion criteria. Second, we limited our analysis to studies with well-reasoned methodology and excluded review articles. A small cohort study was accepted for RLS as there was nothing else available. Third, we excluded studies that were designed to assess other outcomes rather than sleep. And last, we excluded in vitro and animal investigations, eliminating the speculation that experimental observations might not be reproducible in humans. We believe that our rigorous methodology allowed us to refine our search, leading to a greater quality of findings, and a superior understanding of results. This study delivers to the sleep disorder researcher the most comprehensive literature review to inform an exploration into possible projects to undertake.

In summary, we found inconsistent results associated with the effects of cannabis on sleep. It is suspected that cannabis prescription to treat some sleep disorders may potentially have some application in the near future, but more research is still necessary to provide accurate conclusions. The current evidence is not enough to confidently prescribe cannabis for sleep disorders. This is in accordance with a more recent metanalyses of the treatment of insomnia,¹⁰⁸ medical cannabis used for impaired sleep,¹⁰⁹ and a narrative review.⁸ The response to cannabis use varies widely, and more extensive, power-adequate, randomized placebo-controlled trials are needed to clarify the role of medical cannabis in the sleep disorders.

DISCLOSURE STATEMENT

All authors have seen and approved the manuscript. Institution where work was performed: University of Toronto. The authors report no conflicts of interest.

ABBREVIATIONS

AASM

American Academy of Sleep Medicine

AHI

apnea-hypopnea index

CBD

cannabidiol

OSA

obstructive sleep apnea

RCT

randomized controlled trial

REM

rapid eye movement

RLS

restless legs syndrome

THC

tetrahydrocannabinol