We found no significant differences in polysomnography results following the administration of CBD and placebo to healthy volunteers. Likewise, there were no statistically significant changes in the subjective and cognitive measures collected during the two nights of polysomnographic exams.

Several parameters were recorded during polysomnography, considering that the essential tests for sleep staging are electroencephalogram, electrooculogram, and electromyogram. Given the lack of studies on the effect of CBD on human polysomnography-monitored sleep, other parameters were selected based on studies that tested the effect of other drugs in healthy volunteers (Orr et al., 2012; Yadollahi et al., 2014). When comparing our polysomnographic data with results from other studies that used placebo in healthy volunteers, similar findings were observed (Buysse et al., 1989; Sabbatini et al., 2005; Fidan et al., 2011; Feld et al., 2013; Wilson et al., 2015).

No statistically significant changes were found between the three different time points in the four factors evaluated by the VAMS and, as well as in the STAI. These results suggest that none of the different moments of the exams were subjectively rated as anxiogenic, sedative, uncomfortable or as producing cognitive impairment. It should be noted here that, unlike other medications, the anxiolytic effect of CBD is only observed when given to subjects in obviously anxiogenic situations (Zuardi et al., 1993, 2017; Bergamaschi et al., 2011a; Crippa et al., 2010).

In the present study, we found no residual effects of CBD on cognitive or psychomotor functions compared to placebo, as measured by the Digit Symbol Substitution and Symbol Copying subscales of the WAIS, which have been described as sensitive measures of residual drug effects (Garber et al., 2004; Rosenberg et al., 2007).

Although no previous study on sleep and CBD applied these specific measures, our findings are consistent with a study on multiple sclerosis that used the digits test to assess possible changes in disease status following the administration of CBD associated with THC, in which no significant change was recorded (Vaney et al., 2004).

It is known that lack of sleep can interfere with certain aspects of cognitive functioning, such as attentional levels (Goel et al., 2009) and PVT, which has a high sensitivity to measure responses that require selective attention (Basner and Dinges, 2011). However, the results of the present study did not show any significant impairment in either the reaction time or number of errors measured by the PVT, suggesting that the attention levels of the volunteers were preserved in the morning after the sleep assessment, regardless of the administration of CBD or placebo. Not having administered the PVT test before CBD and placebo administration does not significantly affect the conclusions once the study does not intend to assess the effect of CBD on baseline vigilance (which would require comparison with baseline PVT results), but to rather evaluate if CBD may be safely administered to patients without affecting their vigilance state overall, such that the patients may safely conduct every-day tasks, like for example driving.

Earlier preclinical studies have suggested that the therapeutic effects of CBD might depend on the presence of specific clinical conditions. As an example, Campos et al. (2013) showed that the chronic use of CBD for 2 weeks, while not directly increasing hippocampal neurogenesis, prevented its decrease by unpredictable chronic stress. Thus, the absence of changes in the sleep of healthy volunteers treated with CDB in our study should not be considered as a final indication that CBD could not have positive effects in patients with sleep disorders.

It is known that a major problem of several medications used in the treatment of clinical anxiety and depression is their effect on sleep architecture. Benzodiazepines are an example, since despite the rapid onset of their anxiolytic action, these drugs may produce undesirable side effects such as the increase in non-REM stage 2 sleep and reduction of SWS (Borbély et al., 1985). Long-term use of benzodiazepines may also cause reduction of SWS, loss of efficacy in the treatment of insomnia, alterations in electroencephalogram results during sleep (Poyares et al., 2004) and cognitive dysfunction, even after drug discontinuation (Stewart, 2005).

Likewise, selective serotonin reuptake inhibitors (SSRIs) and selective serotonin and norepinephrine reuptake inhibitors (SNRIs) may interfere with sleep architecture and decrease restorative sleep, leading to increased awakenings, reduced REM sleep, increased REM latency, as well as increased periodic limb movement during sleep (Feige et al., 2002). In addition, SSRIs and SNRIs have been associated with REM sleep without atonia, characterized by increased tonic or phasic motor activity in electromyographic channels during REM sleep (Schenck et al., 1992; American Academy of Sleep Medicine, 2014; Lee et al., 2016).

Cannabidiol’s anti-anxiety (Zuardi et al., 1993, 2017; Crippa et al., 2009; Bergamaschi et al., 2011b) and antidepressant (Saito et al., 2010; Zanelati et al., 2010) potential seems to differ from other drugs with effects on the central nervous system, since we found no alterations in sleep architecture. Additionally, studies on the anxiolytic, antipsychotic and antiparkinson effects of CBD described no sedation or drowsiness side effects in their volunteers (Zuardi et al., 1993; Crippa et al., 2004; Fusar-Poli et al., 2009; Chagas et al., 2014a). These findings complement the literature on the few significant side effects resulting from the administration of CBD to humans in a wide range of doses, administered chronically or acutely (Bergamaschi et al., 2011b; Kerstin and Grotenhermen, 2017). It seems, therefore, that CBD has an adequate safety profile with good tolerability and does not affect psychomotricity or cognition (Hayakawa et al., 2007; Crippa et al., 2010; Bergamaschi et al., 2011b; Kerstin and Grotenhermen, 2017). This is particularly important in Parkinson’s disease, where motor and cognitive symptoms play a central role.

The relative representativeness of the small sample size and the use of a single dose of CBD can perhaps be regarded as a limitation of our study, as it does not allow the assessment of the effects of chronic treatment with CBD on sleep. In the study by Chagas et al. (2014b), for example, CBD was chronically administered for 6 weeks to patients with Parkinson’s disease and REM sleep behavior disorder. Since the effects of CBD are biphasic (Zuardi et al., 2017), the use of a single dose also limits the interpretation of the present findings. Moreover, monitoring changes in sleep using a conventional polysomnography presents some intrinsic limitations, as it is insufficient alone to detect drug-induced changes of the sleep EEG. For this purpose, a spectral analysis or a similar procedure is also needed. Conversely, the use of preclinical polysomnography to characterize drug-induced sleep disturbances has been increasingly recommended in the regulatory context (Authier et al., 2016). Finally, it is essential to evaluate the effects of CBD in a larger sample and in individuals diagnosed with sleep disorders in addition to healthy volunteers.

Despite these limitations, this is the first controlled study to evaluate the effects of CBD on sleep architecture using polysomnography. Although the absence of interference with the sleep cycle is not sufficient for concluding that sleep is not affected, the results obtained contribute for the understanding of the effects of CBD in the modulation of sleep in humans.