Same-Day Sedative and Night-Time Sleep Effects Following Combined Cannabinoid Formulations: A Randomised-Controlled Trial
https://ror.org/031rekg67grid.1027.40000 0004 0409 2862Centre for Mental Health and Brain Sciences, Swinburne University of Technology, Hawthorn, VIC Australia
https://ror.org/010mv7n52grid.414094.c0000 0001 0162 7225Institute for Breathing and Sleep, Austin Hospital, Melbourne, VIC Australia
International Council of Alcohol, Drugs, and Traffic Safety (ICADTS), Melbourne, VIC Australia
Abstract
Background and objectives
Cannabinoid treatments are commonly used for sleep conditions, but the direct sedating effects of daytime treatment consumption and indirect effects on night-time sleep are unclear. This study measures the direct effects of low-dose cannabinoid treatments on daytime sleepiness and potential indirect night-time sleep effects in healthy adult, novice cannabis users.
Methods
Using a double-blind, randomised, placebo-controlled cross-over design, participants were orally administered a standardised dose of 1 mL oil containing THC:CBD ratios of either 1:1, 1:16 or a placebo over five weekly in-lab visits. Daytime sleepiness was measured at 40, 135 and 265 min post-dosing using the Karolinska Sleepiness Scale (KSS). Indirect night-time sleep effects on total sleep time (TST), sleep-onset latency (SOL), and number of awakenings after onset were measured using daily wrist-actigraphy and sleep-diary entries during the 7-day washout period between treatments.
Results
Final analyses (N = 20) showed subjective sleepiness (KSS score) significantly increased (mean difference = 1.9, SE 0.25) from 40 min to 265 min post-treatment (p < 0.001). No significant differences were observed between treatments for KSS. Indirect sleep measures (TST, SOL, number of awakenings) showed no differences between treatments or over time (all p > 0.05).
Conclusion
Daytime consumption of low-dose cannabinoid oils did not induce direct sleepiness or indirect night-time effects post-dosing among adults. Future studies would benefit from exploring pharmacokinetics and the possibility of treatment amplification of daytime fatigue, mood and cognitive changes to assist the development of therapeutic guidelines for safe daytime medical cannabis use.
ANCTR Trial Registration Number
ACTRN12622001539729, 13 December 2022, prospectively registered.
Key Points
| Low-dose cannabinoid treatments did not directly induce daytime sleepiness in healthy adults. |
| The time it took to get to sleep, the amount of night-time sleep, and the number of awakenings after sleep did not change as an indirect effect of day-time treatment. |
| Results may support daytime use without night-time impairment, but future studies are necessary to explore different doses and administration routes and its impact on subjective states and performance to determine safe use. |
Introduction
Medical cannabis is widely used to manage a broad range of refractory conditions, including chronic pain and anxiety, which are often associated with sleep disruptions through direct or indirect mechanisms [1]. To effectively manage symptoms, individuals with medical conditions are often required to administer their prescribed medical cannabis treatments during the day. Despite this, it remains unclear whether cannabinoids taken for these non-sleep applications can induce daytime drowsiness, consequently impacting performance on every-day tasks. This concern is particularly relevant given that more than a third (35%) of patients have reported performing safety-sensitive tasks, such as driving, within 3 h of consuming their medicinal cannabis treatment [2]. Furthermore, it is not yet understood if specific daytime dosing regimens taken for non-sleep applications might contribute to secondary nocturnal sleep-related effects such as longer and undisturbed sleep.
Medical cannabis formulations, often distinguished by their combined and varying ratios and doses of cannabidiol (CBD) and delta-9-tetrahydrocannabinol (THC), are shown to stimulate sleep-promoting regions such as the basal forebrain, in addition to orexin-secreting neurons which are responsible for maintaining wakefulness through modulation of the endocannabinoid and neuroendocrine systems; however, the exact mechanism driving this effect remains uncertain [3]. Pharmacokinetic studies report substantial variance in the time to peak plasma concentrations for oral formulations of CBD, with the onset, extent and length of effects highly dependent on both external factors such as dose and administration route, and individual factors such as past cannabis use and health [4, 5, 6]. The sedative and sleep-inducing effects of cannabinoids alone or in combination are hypothesised to occur either directly or indirectly with little scientific consensus on therapeutic guidelines for safe and effective use [7]. The direct and indirect sleep-based effects of cannabinoids can vary between individuals and medical conditions, with dosages and ratios suggested to play large determining factors in these effects. Whilst the safety of a single dose of up to 6000 mg CBD alone is generally well tolerated, clinical research on doses ranging from 5 mg to 600 mg report inconsistent sedative and nocturnal sleep effects in both sleep and non-sleep applications such as pain [7, 8]. Due to the psychoactive nature of THC, its therapeutic dose is often titrated up from 2–2.5 mg, with an upper limit of 40 mg to reduce negative side effects including paranoia and tolerance effects in both combined and THC-alone treatments [10, 11, 12].
The combination of THC and CBD is thought to mitigate the negative psychoactive and panic-inducing effects of THC alone whilst enhancing the potential therapeutic benefits of both cannabinoids. These combinations are typically provided in a range of formulations from balanced THC/CBD (1:1), to CBD-dominant (e.g., 1:20 THC/CBD) and THC-dominant preparations (e.g., 10:1 THC/CBD) [13]. THC-dominant formulations are more commonly prescribed for conditions associated with severe pain such as cancer [14]. These formulations can indirectly improve sleep quality ratings and the time taken to fall asleep or sleep-onset latency (SOL) through their prominent analgesic effects [15]. However, higher doses of THC in these formulations can have counterintuitive sleep-impairing effects for people with chronic insomnia, generally lengthening SOL and increasing daytime sleepiness [16–18]. Instead, CBD-dominant and balanced variations are more frequently prescribed for sleep. A single oral dose of 10 mg THC/200 mg CBD (1:20 THC/CBD) showed a significant reduction in nocturnal total sleep time (TST) among people with insomnia, with no reported next day effects on cognitive performance [20]. In contrast, healthy adults with no reported sleep problems reported increased nocturnal wakefulness following oral consumption of 15 mg THC/15 mg CBD compared to 5 mg THC/5 mg CBD [21]. In the same study, 15 mg THC alone was suggested to have a residual sedative effect the next day, further suggesting its combination with CBD may have some dose-dependent effects on daytime alertness [21]. CBD-dominant and balanced formulations may also indirectly improve subjective sleep quality ratings through pain management in other chronic conditions such as fibromyalgia and multiple sclerosis [22, 23], and have shown efficacy when used for mood conditions such as anxiety and depression [19]. These formulations are thought to improve affect via endocannabinoid and neuroendocrine system modulation which may indirectly improve sleep quality ratings without inducing next-day sleepiness [7, 19].
The broad range of heterogenous cannabinoid dosages and ratios across diverse patient populations makes it increasingly difficult to delineate the magnitude and extent of medicinal cannabis sleep-based effects for doses taken at night and those taken during the day for non-sleep applications. Concerningly, these direct and indirect effects alongside the potential impact on performance in safety-sensitive tasks remain underexplored. Therefore, this study primarily aimed to explore the direct effects of four different low doses of balanced THC/CBD (1:1) and CBD-dominant (1:16) cannabinoid formulations on daytime sleepiness in healthy novice cannabis users, relative to placebo. It secondarily assessed the indirect impact of these cannabinoid formulations on night-time sleep. It was expected that CBD-dominant formulations would induce daytime sleepiness and produce greater night-time sleep-enhancing effects relative to equivalent CBD:THC ratios (1:1) and placebo based on past studies [7, 24], with findings intended to assist in outlining the magnitude and duration of daytime and nocturnal sleep-based effects for varied cannabinoid treatments. Such data are critical to aid in identifying potential treatments for disrupted sleep and assist in developing guidelines for safe daytime use.
Methods
Study Design
A 5-week double-blind, placebo-controlled, cross-over, counterbalanced design was utilised to examine the direct effects of four standardised 1 mL oral doses of either 1:1 or 1:16 ratio THC:CBD oil on daytime sleepiness and secondary night-time sleep effects on total sleep time (TST), sleep-onset latency (SOL) and number of awakenings after sleep onset in healthy novice cannabis users relative to placebo. This trial was prospectively registered with the Australia and New Zealand Clinical Trials Registry (ACTRN ID: 12622001539729) and approved by the Swinburne University of Technology’s Human Research Ethics Committee (approval granted: 22/12/2022; Ref: 20236915-16331). All data were collected in adherence with Good Clinical Practice guidelines and the ethical standards of the Declaration of Helsinki and stored at the Centre for Mental Health and Brain Sciences at Swinburne University of Technology in Hawthorn, Melbourne, Australia. Written informed consent was obtained prior to participant enrolment with ongoing eligibility and consent re-confirmed prior to treatment administration at each testing session.
Participants
Participants were eligible if they were aged between 21 and 60 years, able to speak and read English, had a previous history (at least once in lifetime) of cannabis use with no self-reported adverse effects, reported no current or history of psychiatric or physiological illnesses, were not taking psychoactive medications that could impact study measures, and were within the normal body mass index (BMI) range (18.5–30.0) to ensure that active doses were similarly metabolised. Eligibility criteria specified the exclusion of participants if they were unable to understand or comply with testing procedures, were taking any form of medication within 1 week of admission (with exceptions for routine medications to treat benign conditions, prophylactics, antibiotics or contraceptive pills), had moderate-severe current depression (determined by a Beck Depression Index score of > 20) or severe anxiety (determined by a Beck Anxiety Index score of > 16) [25, 26], were under legal supervision, had current or historical drug abuse or dependence, were pregnant, planning pregnancy or breastfeeding, or had participated in another trial involving consumption of an investigational product within the past 30 days.
Procedure
Recruitment, Screening and Practice Visit (V0)
Participants were recruited through online advertisements on social media and word of mouth. Each participant was required to read the Participant Information and Consent Form prior to completing an online pre-screening questionnaire on a secured online platform (Qualtrics, Utah, USA). Participants were asked to provide information about their cannabis use including lifetime history, recent use, frequency, typical amount consumed, administration route and future plans to consume cannabis in addition to any current medical conditions, current medication usage, drivers licence status, and demographic information to confirm eligibility prior to being asked to attend Swinburne University of Technology for an in-person screening visit (V0). During this in-person visit, participants were made aware of the requirements of the study, provided written informed consent, and were advised of their ability to withdraw from the study at any time. Eligibility was confirmed by the research nurse prior to enrolment and allocation of a randomisation identification code. Participants were scheduled to attend five individual testing sessions (approximately 5-h duration each) over a minimum of 5 weeks, with at least a 1-week washout period between each visit to mitigate potential carryover effects.
Testing Visits (V1–V5)
Participants were requested to restrict their consumption of alcohol or nicotine for 24 h, caffeine for 12 h, and food or drinks other than water for 2 h prior to each testing session. Participants were also advised to not consume any illicit substances for 2 weeks prior to and throughout the study duration. Female participants provided a urine sample which was screened for pregnancy prior to each testing visit and study restriction adherence was checked. Participants’ saliva was screened for recent use of drugs using Securatec DrugWipe 6S devices (amphetamine/d-methamphetamine, 3,4-methylenedioxymethamphetamine (MDMA), cocaine, cannabis (delta-9- tetrahydrocannabinol) and opiates) and alcohol (breathalyser, Blood Alcohol Concentration, BAC > 0.00%). If the presence of either was confirmed, participants were withdrawn from that visit with an opportunity to participate in the future if study procedures were adhered to (Fig. 1).
Participants were required to attend Swinburne University between 9:00 a.m. and 12:00 p.m. to start their testing session, with all proceeding visits scheduled at the same time. Participants broke the required 2-h fast with a standard high-fat/calorie breakfast (approximately 700 kcal) consisting of two slices of whole-wheat bread, a spread of their choice (butter, nut-butter, or vegemite) and a high-calorie drink (three scoops of Sustagen mixed with 200 mL water). Following pre-dosing ongoing eligibility checks, participants consumed a single randomised treatment prior to a 40-min absorption period after which participants completed the first round of questionnaires (Karolinska Sleepiness Scale, KSS) [27]. The same questionnaires were then completed twice, at 135 min and 265 min post-dosing (Fig. 1). The timepoints were chosen to measure treatment effects on sleepiness across expected onset, peak and residual stages of cannabinoid plasma concentrations [28]. A standard lunch was provided at approximately 145 min post-dosing and consisted of two slices of toast with the same spread and a coconut yoghurt (110 g). Once other measures pertaining to the study (see trial protocol registration for more details, ACTRN12622001539729) were completed, participants were provided with an actigraphy watch and a sleep diary to be completed daily throughout the week prior to the next testing visit with additional instructions to wear the actigraphy watch daily whilst maintaining their normal routines. At the end of each testing visit, the nurse reviewed the participant before they were allowed to leave the testing site. Participants were instructed to abstain from driving or using heavy machinery, use of illicit drugs, medication and alcohol 12 h after testing, and provided with a taxi voucher (AUD$50) to get home. Participants were compensated after each visit (AUD$100 gift voucher) and received an additional bonus upon successful completion of all testing visits (AUD$200 gift voucher).
Investigational Product
Raw treatments (one 30 mL bottle of T26 – 26 mg/mL to provide 780 mg THC; one 60 mL bottle of CBD 100 – 100 mg/mL to provide 6,000 mg CBD) were donated by CANN group Ltd. (Melbourne, Australia) and were diluted into separate 30 mL bottles with medium-chain triglyceride (MCT) oil by MyCompounding Pharmacy, Bayswater, Melbourne, Australia. The prepared bottles consisted of combined THC/CBD treatments at ratios of 1:1 (2 mg THC/2 mg CBD; 5 mg THC/5 mg CBD) or 1:16 ratios (2 mg THC/32 mg CBD; 5 mg THC/80 mg CBD). Placebo treatments contained MCT oil only. All treatments were flavoured with peppermint oil for masking. Treatment codes (A, B, C, D, E/1,2,3,4,5) were determined by a disinterested third party, and each participant was assigned five treatments according to a randomisation schedule kept by lab staff independent of the trial over the 5-week dosing period. Two independent staff members (including a research nurse) dispensed, sighted and confirmed each administered treatment to ensure blinding. Participants received one 1 mL oil treatment via a single use, opaque disposable syringe at each testing visit.
Measures
Karolina Sleepiness Scale (KSS)
Subjective daytime sleepiness was assessed using the KSS. The KSS is a 9-point scale ranging from extremely alert (1) to very sleepy, great effort to keep alert, and fighting sleep (9) [27]. It allows for a time-efficient measure of situational sleepiness and is commonly used to assess sleepiness/fatigue in response to drug effects [29]. The scale has high reliability and validity and is sensitive to fluctuations in sleepiness, with outcomes closely linked to electroencephalogram and behavioural markers of sleepiness [29]. Participants completed these questionnaires at three timepoints, at approximately 40 min (T1), 135 min (T2) and 265 min post-dosing (T3) (Fig. 1).
Actigraphy
Objective sleep data were collected during the 1-week wash-out period following each of the five testing visits using wrist-mounted actigraphy (GENEActiv, version 1.1, Activinsights, Kimbolton, UK) (Fig. 1). Collected data were measured at a sampling rate of 100 Hz. Watches were set to default sensitivity (12 bit [3.9 g] resolution, range of ± 8 g) and analysed using 5-s epochs using the GGIR package for R Software (version 4.2.2; R Foundation for Statistical Computing, Vienna, Austria) [30]. The default GGIR algorithm was used to determine sleep periods (5 min of inactivity, ± 5 degrees) and the number of awakenings after sleep onset, classifying an awakening as movement during the sleep period between periods of sustained inactivity [30].
A daily minimum of 16 valid hours of actigraphy data was necessary to be considered for analyses (within 24 h measured from noon to noon and consisting of the sleep window). Data were excluded if they did not meet these criteria. Sleep diary entries were used to derive sleep outcomes of TST (time spent in bed subtracted from time spent awake), SOL (time to bed subtracted from actigraphy derived time of sleep onset), and number of awakenings after sleep onset using the GGIR package to clean and produce data for analyses [30] (Fig. 1).
Sleep Diary
Daily sleep diaries (pen and paper) were completed to enhance objective actigraphy data [31]. Data pertaining to time to bed, time out of bed, an estimated number of nightly awakenings, and total time in bed were extracted. Data on self-reported daily exercise (over 20 min), daily number of caffeinated beverages, and nightly habits within 2 h prior to sleep (alcohol, caffeine, heavy meal or not applicable) were also extracted to inform analyses (Table 1).Baseline characteristics n % Gender Female 8 40 Male 12 60 Height, cm, mean (SD) 171.14 (12.68) Weight, kg, mean (SD) 67.15 (9.18) Age, y, mean (SD) 31.40 (7.13) Handedness Left 3 15 Right 16 80 Ambidextrous 1 5 Total years education, mean (SD) 16.06 (3.63) Highest educational level Secondary 3 15 Tertiary 12 60 Postgraduate 5 25 Employment Full-time 8 40 Part-time 8 40 Studying 1 5 Studying and working 2 10 Unemployed 1 5 Ethnicity European/European descent 14 70 Asian 5 25 South American 1 5 First language English 18 90 Other 2 10 Caffeinated beverages 15 75 Mean per week (SD) = 1.62 (1.23) Exercise (20 min, at least once throughout study period) 14 70 Consumption before bed (alcohol/caffeine/heavy meal, at least once throughout study period)) 14 70
Statistical Analysis
Separate linear mixed effects models (LMMs) with restricted maximum likelihood estimation were used to assess treatment-specific changes across time for KSS scores and night-time sleep outcomes of TST, SOL and number of awakenings after sleep onset. Compound symmetry was determined to be the best fit for variance structure based on the likelihood ratio statistic. For each model, treatment and time were entered as repeated and fixed effects, and participants were added as random effects (subject groupings). Post hoc paired t-tests with Bonferroni correction for multiple comparisons were used to examine any significant main effects (p < 0.05). Log transformations were used if the original data was not normally distributed. The day of the week (labelled as either a weekday or weekend) was used to adjust for variability in night-time sleep outcomes between weekdays and weekends for each measure of night-time sleep [32]. Daily exercise, daily number of caffeinated drinks consumed, and nightly habits consisting of the consumption of caffeine, heavy meal or alcohol within 2 h of sleep were included as separate covariates for each night-time sleep outcome, resulting in each night-time sleep outcome having five individual tests in total to assess the differences between treatments without weekday/weekend adjustments and covariates, when adjusting for weekday/weekend, daily exercise, caffeinated drinks or nightly habits. All randomised participants were included in the final analysis (N = 20). Sleep-based outcomes were included in the study design as secondary measures and were not the primary driver of the sample size calculation. A total of 30 participants was powered to detect a moderate effect (f = 0.25) for driving performance outcomes between treatments, approximating the clinically relevant difference of 2.4 cm in standard deviation of lane position (equivalent to 0.05% BAC) compared to placebo (see registration for more details, ACTRN ID: 12622001539729). Based on an LMM, a two-tailed analysis with α = 0.05, β = 0.80, Cohen’s f = 0.5 and a critical F of 3.17, 28 participants were required to detect a moderate effect size in primary and secondary driving simulator outcome variables between treatments and when compared to a placebo. Therefore, the study aimed to reach sufficient power for the driving-related analyses. The study concluded due to time-based constraints. All analyses were conducted using SPSS Statistic software (version 29).
Results
Recruitment and Sample Characteristics
At the start of the trial (17 March 2022), 74 participants had registered and completed online screening (Fig. 2). Of these, 23 participants attended V0, with three participants excluded due to non-adherence, medical reasons or declined participation due to time constraints. The remaining 20 participants were randomised. The data of the 18 participants who successfully completed all five visits (V1–V5) and two participants that partially completed the study (V1–V4 only) were included in the analyses.
Table 1 shows the demographic characteristics of the participants included in the analysis (N = 20). It includes biological details including participant sex, age (range, mean, ± SD), height and weight (mean, ± SD for both), in addition to key demographic characteristic such as handedness, highest level of education completed (primary, tertiary, postgraduate), number of years of study, ethnicity, language spoken, and employment status (full-time, part-time, studying, unemployed). It also includes reported consumption of caffeinated beverages (mean, ± SD), the number of participants reporting daily exercise (of 20 min or more per day) at least once and nightly habits including the consumption of either a heavy meal, caffeine or alcohol within 2 h of sleep at least once.
Daytime Sleepiness—KSS
A main effect of time was observed for KSS scores [F(2,190.1) = 40.60, p < 0.001], and post hoc analyses showed significant increases in sleepiness across time after doses of 2 mg THC/2 mg CBD (T1–T2 mean difference =1.75, SE 0.50, p = 0.002, [confidence interval (CI) 0.53, 2.98]); 5 mg THC/5 mg CBD (T1–T3 mean difference = 2.11, SE 0.61, p < 0.001, [CI 0.65, 3.56]); 2 mg THC/32 mg CBD THC (T1–T3 mean difference = 2.36, SE 0.61, p < 0.001, [CI 0.88, 3.83]); and 5 mg THC/80 mg CBD THC from T1 (to T2 and T3 only) (T1–T3 mean difference = 2.42, SE 0.61, p < 0.001, [CI 0.96, 3.89]). No significant changes were observed for placebo at any time point (p > 0.05 for all). No significant main effects were observed for treatment or the interaction of treatment and time on KSS scores (all p > 0.05). Post hoc, pairwise comparisons of each treatment at each timepoint showed no significant differences between active treatments and placebo at any timepoint (all p > 0.05), revealing no time-dependent or condition-specific effects. Figure 3 shows the changes over time in addition to the mean scores of KSS outcomes presented in Table 2.Treatment Time (post-dosing) F-value p-value T1 (40 mins) T2 (135 mins) T3 (265 mins) KSS, mean (SD) 2 mg THC/2 mg CBD 4.35 (1.73) 6.10 (2.15) 5.75 (2.22) 0.83(8, 229.02) 0.58 5 mg THC/5 mg CBD 3.95 (1.78) 5.80 (2.14) 5.95 (2.14) 2 mg THC/32 mg CBD 3.68 (1.89) 5.79 (2.25) 6.00 (2.45) 5 mg THC/80 mg CBD 3.85 (1.60) 5.21 (2.35) 6.26 (2.21) Placebo 4.16 (1.83) 5.42 (2.12) 5.21 (2.18)
Night-Time Sleep—Total Sleep Time (TST), Sleep-Onset Latency (SOL) and Number of Awakenings
A main effect of treatment was observed for TST [F(4,342) = 9.27, p < 0.01]. Post hoc analyses showed no significant changes for any treatment throughout the trial period. No other main effects for time or the interaction of treatment and time were observed for TST. Post hoc, pairwise comparisons of each treatment at each timepoint showed no significant differences between active treatments and placebo at any timepoint (all p > 0.05), revealing no time-dependent or condition-specific effects (see Table 3).Treatment Time (nights (N)) F-value p-value N1 N2 N3 N4 N5 N6 N7 TST (h), mean (SD) 2 mg THC/2 mg CBD 6.12 (1.89) 5.69 (0.72) 5.43 (2.63) 5.75 (1.44) 6.35 (0.76) 5.11 (1.84) 5.85 (1.77) 0.32 (24,340) 1.00 5 mg THC/5 mg CBD 7.03 (1.74) 6.21 (1.33) 5.72 (1.37) 7.12 (1.45) 6.08 (1.00) 5.30 (2.01) 5.81 (1.18) 2 mg THC/32 mg CBD 6.48 (2.00) 5.68 (1.25) 6.08 (2.01) 6.31 (0.45) 6.15 (0.50) 5.47 (1.51) 5.93 (0.67) 5 mg THC/80 mg CBD 6.06 (2.12) 5.39 (1.82) 5.85 (1.60) 6.44 (1.41) 5.98 (2.27) 6.06 (1.79) 6.00 (1.29) Placebo 6.65 (1.51) 6.93 (1.38) 5.80 (1.94) 6.46 (2.05) 6.21 (1.74) 6.60 (1.42) 6.45 (1.74)
A main effect of treatment was observed for number of awakenings [F(4,360) = 9.67, p < 0.01]. Post hoc analyses showed no significant changes for any treatment throughout the trial period. No other main effects for time or the interaction of treatment and time were observed. Post hoc, pairwise comparisons of each treatment at each timepoint showed no significant differences between active treatments and placebo at any timepoint (all p > 0.05), revealing no time-dependent or condition-specific effects (see Table 4).Treatment Time (nights (N)) F-value p-value N1 N2 N3 N4 N5 N6 N7 SOL (h), mean (SD) 2 mg THC/2 mg CBD 0.38 (0.32) 0.84 (1.17) 0.38 (0.36) 0.46 (0.51) 0.34 (0.21) 0.70 (0.69) 0.34 (0.41) 0.64(24,224.50) 0.90 5 mg THC/5 mg CBD 0.32 (0.43) 0.17 (0.13) 0.29 (0.20) 0.26 (0.22) 0.16 (0.14) 0.43 (0.43) 0.47 (0.31) 2 mg THC/32 mg CBD 0.32 (0.30) 0.48 (1.01) 0.24 (0.19) 0.21 (0.20) 0.27 (0.28) 0.77 (0.79) 0.38 (0.24) 5 mg THC/80 mg CBD 0.29 (0.25) 0.29 (0.20) 0.32 (0.27) 0.44 (0.58) 0.49 (0.43) 0.57 (0.46) 0.34 (0.25) Placebo 0.20 (0.17) 0.20 (0.25) 0.41 (0.86) 0.22 (0.14) 0.15 (0.14) 0.25 (0.19) 0.27 (0.32)
No main effects of treatment, time or its interaction were noted for SOL throughout the trial (all p > 0.05). Adjusting for day of the week, daily exercise, daily number of caffeinated drinks consumed and nightly habits (heavy meal, caffeine or alcohol) within 2 h of sleep did not influence the outcomes for any measures of night-time sleep (see Table 5).Time (nights (N)) F-value p-value Treatment N1 N2 N3 N4 N5 N6 N7 Awakenings, mean (SD) 2 mg THC/2 mg CBD 15.10 (7.72) 14.80 (4.61) 12.00 (7.76) 14.78 (4.41) 15.70 (6.06) 13.00 (7.28) 12.38 (8.58) 0.01(24,358.39) 1.00 5 mg THC/5 mg CBD 16.80 (6.76) 15.11 (6.01) 15.44 (9.67) 16.11 (7.74) 12.25 (5.44) 10.78 (7.92) 16.40 (3.72) 2 mg THC/32 mg CBD 16.18 (4.47) 14.33 (5.48) 16.50 (7.47) 15.22 (6.40) 15.63 (2.88) 13.89 (7.36) 14.44 (6.60) 5 mg THC/80 mg CBD 14.82 (9.15) 13.40 (5.60) 13.55 (8.72) 13.56 (8.40) 13.73 (7.38) 15.30 (7.39) 16.11 (4.46) Placebo 18.30 (6.77) 17.45 (6.20) 14.73 (5.71) 16.45 (6.61) 15.73 (5.20) 16.82 (6.71) 16.00 (5.40)
Discussion
The study’s findings indicate that low-dose cannabinoid treatments administered during the daytime do not directly increase daytime sleepiness or exert residual or indirect effects on night-time sleep measures of TST, SOL and number of awakenings in healthy, infrequent cannabis users. Daytime sleepiness was noted to increase irrespective of treatment and was likely due to external factors such as testing procedures and duration. Contrary to expectations, CBD-dominant doses did not induce greater daytime sleepiness than 1:1 ratios and placebo. Participants generally reported moderate alertness 40 min post-dosing, with slight increases in sleepiness observed over the subsequent testing period, persisting for up to 265 min post-dosing. This suggests the involvement of factors likely unrelated to the cannabinoid treatments. Instead, it is likely that factors related to the research environment, including the complexity and duration of testing procedures, may have contributed to the observed increases in sleepiness. Furthermore, night-time sleep outcomes of TST, SOL and the number of awakenings remained unaffected across all treatment conditions, suggesting future studies to assess sleep in greater detail using methods such as electroencephalography to capture potential changes.
Nicholson et al. [21] observed morning sleepiness 9 h after combined treatment administration (15 mg THC/15 mg CBD) in healthy adults, whereas Suraev et al. [33] noted a small increase in self-reported sedation scores without any changes to ratings of alertness or sleepiness up to 10 h after administration in adults with insomnia after 10 mg THC/200 mg CBD dose. These outcomes were supported by changes to nocturnal sleep after night-time administration of treatments in both studies. This suggests cannabinoid treatments could amplify habitual sleep and circadian rhythms, with the slow absorption of oil-based treatments lengthening effects and further leading to differences in carry-over effects the next day [34, 35]. Conversely, we report that normal night-time sleep measures remained unchanged following daytime use. This highlights the distinct differences in effects based on dose, administration time, and potentially the purpose of treatment. In addition to this, active treatments could have mildly amplified daytime sleepiness outcomes through an indirect influence on sedation or mood; however, this cannot be concluded from this study. Nonetheless, outcomes suggest that doses taken during the day for non-sleep applications show no significant residual treatment effects on nocturnal sleep. This is further supported by the lack of change in measures of daytime sleepiness following active treatments, suggesting complex interactions between cannabinoid treatment timing, their intended application, and mechanisms of action in healthy participants. Investigating these factors, alongside greater emphasis on the pharmacokinetics of different administration routes, will be crucial to better understand the therapeutic and safety profiles of cannabinoid products. Pharmacokinetic factors have been observed to have vast differences based on administration routes, resulting in different times for onset, peak and residual phases of treatment effects and plasma levels. Though ingested routes are noted to take more time for absorption than vaporization, the effects are lengthened due differences in metabolic processing. Therefore, it is imperative to understand the extent of therapeutic effects and safety profiles of similar treatments in different forms.
The protective effect of CBD in combined THC/CBD formulations may not be consistent across different populations [36], and is unlikely to be ubiquitous for all doses or preparations. In comparison to Nicholson et al. [21], where a 15 mg THC/15 mg CBD dose produced observable changes to subjective night-time sleep, no significant differences were noted in the present study following administration of either 1:1 or higher CBD-dominant doses. Previous research cites that CBD doses up to 150 mg have no measurable impact on night-time sleep even in sleep-disordered populations [9], with therapeutic benefits proposed to follow an inverted U-shaped dose-response curve, likely peaking at doses between 300 mg and 400 mg [7]. Taken together, these findings suggest that the CBD doses examined here may have been too low to induce changes to sleepiness or night-time sleep effects after a single administration. Although the doses in this study were not intended to replicate those commonly used as sleep aids, the data provide valuable insights into potential daytime and residual nocturnal effects for individuals taking low doses for other therapeutic purposes. If THC-induced sleepiness and CBD-induced arousal effects, as observed by Nicholson, et al. [21], were present in this study, they were not readily apparent in either the balanced or the CBD-dominant treatments. It is acknowledged that the low doses used make it challenging to determine whether arousal or sedative effects were masked by interactions between the cannabinoids, or whether these effects were simply absent at these doses.
The sedative effects of single doses of CBD, THC and their combination have primarily been studied in the context of night-time administration and their subsequent impact on sleep [9, 21, 37]. Assessments of daytime sleepiness have typically focussed on waking states the morning following a night-time dose intended for sleep-enhancing effects [21]. In contrast, the present study investigated the effects of daytime cannabinoid dosing on sleepiness, contributing novel insights to this relatively underexplored area. Measurable treatment-induced changes have been observed for comparable doses and ratios of combined treatments in the past [16–20, 33]. Yet, the time of treatment administration and the direct and indirect sleep-based effects measured differs between studies, generally being dependent on the type of condition being treated. In populations with disrupted sleep such as insomnia, nocturnal treatment administration and the measurement of direct treatment effects were on sleep-based measures and next-day residual effects [20, 33]. In non-sleep applications such as in populations with pain, treatment administration was often during the day, with indirect nocturnal sleep-based effects measured secondarily [22, 23]. Though results were mixed for both sleep and non-sleep applications, these outcomes fail to inform the possibility of hazardous daytime sleepiness when administered during the day for non-sleep applications. This study differed from others by measuring direct effects of cannabinoid formulations for non-sleep applications, providing insight into the temporal onset of daytime sleepiness during in-lab testing sessions that studies with nocturnal administration often do not assess. Moreover, comparing the effects of morning versus night-time dosing is inherently challenging due to differences in purpose, populations, physiological state and circadian phase. Night-time dosing conditions, including reduced light, melatonin production and sleep pressure, likely play a significant role in facilitating sleep [38]. These factors are absent or diminished during the day, potentially moderating the treatment’s sedative-type effects. To further understand the functional implications of the daytime sleepiness observed in this study, cognitive performance and daily tasks, such as driving, should be assessed alongside sleepiness measures. This approach would help determine whether the observed changes in sleepiness correlate with performance impairment within the context of cannabinoid treatments.
The strengths of this study included stringent screening procedures, the inclusion of multiple doses of active treatments across two ratios compared to placebo, and the repeated measurement of daytime sleepiness at time-critical points intended to align with the pharmacokinetic phase of orally consumed cannabis. Administering treatments in a fed state ensured consistent baseline conditions across participants, and dosing and assessments were conducted at consistent intervals to reduce possible circadian-phase intrusion effects. Fed states specifically consisting of a high-fat meal are observed to enhance time to peak plasma concentration when compared to less pronounced effects in fasted states for ingested cannabinoid treatments [39, 40]. This is related to the lipophilic nature of CBD and THC and may therefore influence the onset, peak and residual effects of cannabinoid treatments [39, 40]. Measuring night-time sleep at home via wrist-mounted actigraphy provided a more naturalistic context than typically seen in laboratory-based studies [41] whilst collecting objective data and the addition of a sleep diary enabled for detailed analysis of sleep parameters. Nonetheless, differences in the sleep environment and routines, coupled with known limitations of actigraphy-derived (e.g., poorer detection of wake after sleep) and self-reported sleep data (e.g., recall bias) [41] as well as no baseline sleep data being collected prior to the start of the study may have reduced the precision and generalisability of these outcomes. This could be better controlled in laboratory settings using measures such as polysomnography, but may risk disrupting naturalistic sleep habits due to setting. Regardless, the use of at-home actigraphy and sleep diaries provide greater detail into nocturnal sleep parameters than the subjective self-reported sleep measures used in past studies. Bonferroni corrections used for post hoc analyses are suggested to reduce Type I error [42]. However, the sample size was smaller than intended and the analyses used may have reduced statistical power and increased the risk of Type II error [42]. Therefore, the results should be interpreted cautiously. Further research is needed with larger and more diverse populations to improve the replication and generalisability of current observations.
Conclusion
As the use of cannabinoid products continues to rise, it is critical to expand our understanding of their effects, particularly when used during the daytime. This study indicates that single low doses of CBD-dominant and balanced CBD/THC oil treatments consumed during the day do not significantly induce daytime sleepiness, nor do they indirectly alter night-time sleep among healthy participants compared to placebo. The potential effects of higher doses, varying cannabinoid ratios and combinations, however, remain largely underexplored. Investigating these factors, alongside greater emphasis on the pharmacokinetics of different administration routes, will be crucial to better understand the therapeutic and safety profiles of cannabinoid products. Future research should also assess the effects of repeated dosing on day and night-time sleep outcomes to validate these findings and refine safety and therapeutic guidelines. These efforts will help facilitate the responsible and effective use of cannabinoids in both clinical and everyday settings.
Acknowledgements
Centre for Mental Health and Brain Sciences, Swinburne University of Technology, Hawthorn, Australia; Institute for Breathing and Sleep, Austin Hospital, Melbourne, Australia. We thank the participants for their time and commitment.
Declarations
Funding
Open Access funding enabled and organized by CAUL and its Member Institutions. ACH is supported by a Rebecca L Cooper Al and Val Rosenstrauss Fellowship (GNT:F2021894). This study was funded by a grant to ACH from the Australian Office of Road Safety as part of the Road Safety Innovation Fund initiative (RSIF2-51). The sponsors provided funding for the study only. The study treatments were provided by CANN group Ltd. (Melbourne, Australia).
Conflicts of Interest
AN, BM, BA, LAD and ACH have no conflicts of interests to declare that go beyond financial interests that could impart bias on the work submitted for publication such as professional interests, personal relationships or personal beliefs (amongst others).
Ethics Approval
Swinburne University of Technology’s Human Research Ethics Committee (approval granted: 22/12/2022; Ref: 20236915-16331). This trial was performed in accordance with Good Clinical Practice guidelines and the ethical standards of the Declaration of Helsinki.
Consent to Participate
Written informed consent was obtained prior screening and the enrolment of each participant with ongoing eligibility and consent re-confirmed prior to treatment administration at each testing session.
Consent for Publication
Not applicable.
Code Availability
Not applicable.
Availability of Data and Material
Data supporting findings in this study are available within the list of tables presented.