The therapeutic role of Cannabidiol in mental health: a systematic review
The therapeutic application of cannabidiol (CBD) is gaining interest due to expanding evidence for its use.
To summarize the clinical outcomes, study designs and limitations for the use of CBD and nabiximols (whole plant extract from Cannabis sativa L. that has been purified into 1:1 ratio of CBD and delta-9-tetrahydrocannabinol) in the treatment of psychiatric disorders.
Materials and method
A systematic review was conducted including case reports, case series, open-label trials, non-randomized and randomized controlled trials (RCTs). The search resulted in 23 relevant studies on CBD and nabiximols in the treatment of a wide range of psychiatric disorders. The quality of evidence was judged by using the Oxford Centre for Evidence-Based Medicine 2011 Levels of Evidence that ranges from Level 1 to Level 5 based on the quality and study design. These levels of evidence help in grading the recommendations, including Grade A (strong), Grade B (moderate), Grade C (weak), and Grade D (weakest).
CBD and CBD-containing compounds such as nabiximols were helpful in alleviating psychotic symptoms and cognitive impairment in patients with a variety of conditions, and several studies provided evidence of effectiveness in the treatment of cannabis withdrawal and moderate to severe cannabis use disorder with Grade B recommendation. There is Grade B recommendation supporting the use of CBD for the treatment of schizophrenia, social anxiety disorder and autism spectrum disorder (ASD), and attention deficit hyperactivity disorder (ADHD). Grade C recommendation exists for insomnia, anxiety, bipolar disorder, posttraumatic stress disorder, and Tourette syndrome. These recommendations should be considered in the context of limited number of available studies.
CBD and CBD-containing compounds such as nabiximols were helpful in alleviating symptoms of cannabis-related disorders, schizophrenia, social anxiety disorder, and comorbidities of ASD, and ADHD with moderate recommendation. However, there is weaker evidence for insomnia, anxiety, bipolar disorder, posttraumatic stress disorder, and Tourette syndrome. The evidence for the use of CBD and CBD-containing compounds for psychiatric disorders needs to be explored in future studies, especially large-scale and well-designed RCTs.
Cannabis sativa, a species of cannabis plant, is well known to humankind, with its earliest use in ancient Chinese culture dating as far back as 2700 B.C. (Zuardi, 2006). The use of medical cannabis in China was reported in the world’s oldest pharmacopoeia (Martin et al., 1999). However, interest in the role of cannabis flourished in the late twentieth century after the recognition of an endogenous cannabinoid system in the brain (Zuardi, 2006; Martin et al., 1999). More recently, research has centered on the description and cloning of specific receptors and the therapeutic effects of medical cannabis, and different cannabinoids in the cannabis plant have gained interest (Martin et al., 1999). Recent studies have focused on the therapeutic role of medical cannabis in different disorders. As a result, there is a growing need to summarize and review the evidence for its therapeutic and adverse effects as an aid to public health policy development, and to provide direction and impetus to pharmaceutical research in this field.
The cannabis plant has more than 140 cannabinoid compounds, with Δ9-tetrahydrocannabinol (Δ9-THC) and cannabidiol (CBD) attracting significant interest (Citti et al., 2018). Δ9-THC is the primary psychoactive ingredient, and CBD is a non-intoxicating ingredient (Zuardi, 2006; Citti et al., 2018). Evidence from preclinical studies suggested that CBD had potential therapeutic benefits ranging from antiinflammatory to neuroprotective, antipsychotic, analgesic, anticonvulsant, antiemetic, antioxidant, antiarthritic, and antineoplastic properties; for a review, see (Pertwee, 2006). CBD has several receptors and molecular targets. This compound antagonizes the action of CB1 and CB2 receptor agonist (Blessing et al., 2015; Peres et al., 2018). The CB1 and CB2 receptors are coupled negatively through G-proteins to adenylate cyclase and positively to mitogen-activated protein kinase (Pertwee, 2006). In addition to CB1 and CB2 receptor activity, CBD is an agonist of vanilloid receptor TRPV1. It also acts as an agonist of serotonin receptor 5-hydroxytryptamine (5-HT1A), an antagonist of G-protein-coupled receptor GPR55, and an inverse agonist of GPR3, GPR6, and GPR12 (Peres et al., 2018). Data from single-photon emission computed tomography showed CBD to exert anxiolytic effects by acting on paralimbic and limbic pathways (Crippa et al., 2011). The agonist effect of CBD on 5-HT1A also supports its anxiolytic and antidepressant properties (Russo et al., 2005). CBD inhibits enzymatic hydrolysis and anandamide uptake through its agonist action on CB1, CB2, and TRPV1 receptors (Peres et al., 2018). In addition, CBD indirectly enhances endogenous anandamide signaling by inhibiting the intercellular degradation of anandamide (Leweke et al., 2012). This endogenous neurotransmitter exerts antipsychotic effects in patients with schizophrenia (Leweke et al., 2012).
The pharmacokinetic profile of CBD has been extensively explored in the existing literature. A recently published systematic review of the pharmacokinetics of CBD found that the area under curve (AUC0 − t) and maximum serum concentration (Cmax) occurs between 1 and 4 h (Millar et al., 2018). The AUC0 − t and Cmax reach maximum values faster after smoking or inhalation compared to oral or oromucosal routes. Bioavailability was 31% after smoking, but no other studies reported the absolute bioavailability of CBD after other routes in humans. The half-life of CBD ranges between 1.4 and 10.9 h after oromucosal spray and 2–5 days after chronic oral administration (Millar et al., 2018). Fed states and lipid formulations increase Cmax (Millar et al., 2018). The bioavailability of oral CBD ranges between 11 and 13%, compared to 11 to 45% (mean 31%) via inhalation (Scuderi et al., 2009). CBD is well-tolerated, yet despite a relatively lower risk of drug–drug interactions, it should be used cautiously in combination with drugs metabolized by the CYP3A4 and CYP2C19 pathways, and the substrates of UDP-glucuronosyltransferases UGT1A9 and UGT2B7 (Millar et al., 2018). The clinical relevance of these interactions needs to be explored in future studies (Brown & Winterstein, 2019).
Dronabinol and nabilone are synthetic in origin, whereas nabiximols is plant-based (Papaseit et al., 2018). The percentage of THC and its ratio to CBD (THC/CBD ratio) defines the potency and psychoactive effects of a given formulation (Papaseit et al., 2018). Those with higher CBD/Δ9-THC ratios have euphoric, anxiolytic, and relaxing effects, whereas lower CBD/Δ9-THC ratios have sedative properties (Papaseit et al., 2018). Nabiximols, a CBD-containing compound, contains Δ9-THC and CBD at a 1:1 ratio (Papaseit et al., 2018). The Food and Drug Administration has approved Epidiolex® (an oral formulation of CBD) for two forms of childhood seizures (Lennox–Gastaut syndrome and Dravet syndrome) in children 2 years of age and older (Papaseit et al., 2018).
Previous efforts to synthesize the evidence for medical cannabis use in patients with psychiatric disorders have been published (Hoch et al., 2019; Lowe et al., 2019). For example, Hoch et al. conducted an excellent systematic review that summarized four systematic reviews and 14 randomized controlled trials (RCTs), but did not consider non-clinical trial evidence (case reports and case series) (Hoch et al., 2019). A review by Mandolini et al. recently summarized the clinical findings from 14 studies of psychiatric disorders, but these authors did not provide information about nabiximols (Mandolini et al., 2018). In contrast to the review articles noted above, the present article aims to provide a more comprehensive review of the use of CBD and CBD-containing compounds such as nabiximols to treat psychiatric disorders. The present review included studies focused on schizophrenia, cannabis-related disorders, attention deficit hyperactivity disorder (ADHD), comorbidities in autism spectrum disorder (ASD), social anxiety disorder (SAD), other anxiety disorders, insomnia, bipolar disorder, post-traumatic stress disorder (PTSD), psychosis in Parkinson’s disease, and Tourette syndrome. This article broadly reviews the efficacy, safety, and psychiatric benefits of CBD and CBD-containing compounds (nabiximols). We distinguish clearly here between the clinical findings for CBD and nabiximols, as the latter also contains THC.
The main inclusion criterion was studies of the psychiatric use of CBD and CBD-containing compounds such as nabiximols. Only case reports, case series, retrospective chart reviews, open-label trials, and RCTs were considered. All books, conference papers, theses, editorials, review articles, metaanalyses, in-vitro studies, laboratory studies, animal studies, studies of participants without psychiatric disorders, and abstract-only articles were excluded. No restrictions on language, country, publication year, or patients’ age, gender, or ethnicity were applied.
Eight electronic databases were searched on October 28th, 2018: PubMed, Scopus, Web of Science, POPLINE, New York Academy of Medicine Grey Literature Report, PsycINFO, Psycarticles, and CINAHL. The following search strategy was used in all cases: (CBD OR Cannabi* OR nabiximols) AND (psychiat* OR Depress* OR Anxiety OR Psycho* OR schizo* OR Bipolar OR Substance OR ADHD OR Attention OR Autism) AND (treatment). The manual search of references of included studies was performed by four independent reviewers.
The search results from the eight databases were imported to Endnote v. 7 (Thompson Reuters, CA, USA) to remove any duplicates. Four independent reviewers (RK, NM, AF, MAF) screened the titles and abstracts (when available), followed by full-text screening of each included article with the predetermined eligibility criteria. All articles included after full-text screening were then searched manually. Discrepancies were resolved by consensus through discussion among reviewers, or with guidance from a third reviewer (SN).
Data extraction and grading
The data were extracted independently by the authors, and were cross-checked by discussion among the four reviewers (RK, NM, AF, MAF), with guidance from the senior author (SN) in case of discrepancy. The data were categorized as pertaining to target diagnosis, study design, sample size, duration of the trial, age range, dose ranges, measurement scales, clinical outcomes, study limitations, and common side effects.
The Oxford Centre for Evidence-Based Medicine 2011 Levels of Evidence was used to grade the quality of evidence (OCEBM, 2019). Level 1 evidence is for systematic review of RCTs or individual RCT of narrow confidence interval, Level 2 for cohort studies or systematic review of cohort studies, Level 3 for case-control studies or systematic review of case-control studies, and Level 4 for case-series for studies focused on therapy, prevention, etiology and harm (OCEBM, 2019). These levels of evidence are used to generate Grades of Recommendation. Grade A is for consistent level 1 studies, Grade B for consistent level 2 or 3 studies or extrapolations from level 1 studies, and Grade C for level 4 studies or extrapolations from level 2 or 3 studies. Grade D is ranked for level 5 evidence or inconsistent or inclusive studies of any level (OCEBM, 2019).
Results & discussion
The search of eight electronic databases and our manual screening method generated 511 results. After the removal of duplicates, titles and abstracts were screened, resulting in the exclusion of 459 articles. Full-text screening of 52 articles was performed, and 23 articles meeting the inclusion criteria were analyzed. Figure 1 summarizes the screening process.
PRISMA Flow Diagram
Of the 23 articles, there were eight RCTs, one clinical trial, four open-label trials, one retrospective chart review, seven case reports, and two case series, comprising a total patient population of 526. The studies focused on CBD and nabiximols use in the treatment of schizophrenia, cannabis-related disorders, ADHD, ASD and comorbidities, anxiety, insomnia, SAD, bipolar disorder, PTSD, psychosis in Parkinson’s disease, and Tourette syndrome. No studies of substance use disorders other than cannabis use were identified. In this review article, the authors have used DSM-5 terminologies for most of the disorders except for DSM-IV-Text Revised terminology of substance dependence. A comparable DSM-5 terminology of moderate-severe substance use disorder was used in this case.
Qualitative synthesis of eligible studies
Schizophrenia and psychosis in Parkinson’s disease
There were three RCTs (164 patients), one clinical trial (27 patients), one case series (three patients), one case report for schizophrenia, and one open-label trial (six patients) for psychosis in Parkinson’s disease (Table 1) (Leweke et al., 2012; Hallak et al., 2010; Boggs et al., 2018; McGuire et al., 2018; Zuardi et al., 2006; Zuardi et al., 1995; Zuardi et al., 2009). Of the seven studies, level 2 evidence was found in three RCTs, level 3 evidence in two clinical trial, and level 4 evidence in one case report and one case series (OCEBM, 2019). Since most of the studies were from level 2 and level 3 evidence, there is Grade B recommendation for schizophrenia. The dose of CBD in these studies ranged from 200 to 1500 mg daily. The highest dose was titrated to 1500 mg daily as reported by Zuardi and colleagues (Zuardi et al., 1995). Irrespective of the study design, three studies reported that CBD alleviated psychotic symptoms and cognitive impairment in patients with chronic cannabis use and Parkinson’s disease (Leweke et al., 2012; Zuardi et al., 1995; Zuardi et al., 2009), while only two RCTs and one clinical trial provided evidence for the effectiveness of CBD among patients with schizophrenia, albeit with mixed results (Leweke et al., 2012; McGuire et al., 2018; Zuardi et al., 2009).
Table 1 Studies of CBD use in the treatment of schizophrenia and psychosis in Parkinson’s disease and levels of evidence (1 to 5)*
In a clinical trial, Hallak and colleagues suggested an improvement in schizophrenia-associated cognitive impairment with a CBD dose of 300 mg/day, while no significant improvement was seen at a CBD dose of 600 mg/day (Hallak et al., 2010). In another RCT, McGuire and colleagues found that CBD (1000 mg/day) improved positive psychotic symptoms, but failed to improve negative symptoms and general psychopathology associated with this illness (McGuire et al., 2018). In another RCT, Boggs and colleagues found that CBD (600 mg/day) failed to improve outcomes pertaining to reasoning and problem-solving domains (Boggs et al., 2018).
In a comparison of CBD with amisulpride, Leweke and colleagues reported similar improvements in patients taking CBD 800 mg/day and those taking amisulpride (Leweke et al., 2012). This study also indicated an increase in intrinsic anandamide signaling, an effect that explained the antipsychotic properties of CBD (Leweke et al., 2012). Moreover, CBD treatment was associated with a lower risk of extrapyramidal symptoms, less weight gain, and a lower increase in prolactin, which is a predictor of galactorrhea and sexual dysfunction (Leweke et al., 2012). An open-label study of CBD to treat psychosis in Parkinson’s disease also suggested promising results at a dose of 400 mg daily; however, there was a strong risk of bias because of inadequate blinding of participants, personnel and outcome assessors (Zuardi et al., 2009).
The remaining evidence comprised two minimal quality case reports and case series. Zuardi and colleagues were the first to report favorable findings for CBD in patients with schizophrenia (Zuardi et al., 1995). The dose of CBD ranged from 600 to 1500 mg daily in schizophrenia studies. A case series of three patients with treatment-resistant schizophrenia found improvement in only one patient (Zuardi et al., 2006). In the first case, there was an improvement in psychotic symptoms with CBD at 1280 mg/day; however, the symptoms worsened after CBD was discontinued. In second case, CBD was ineffective for the symptoms. Patient had an improvement in symptoms with clozapine. In the third case, no improvement with CBD and partial improvement with olanzapine were observed, although clozapine was subsequently required. In case 3, mild improvement was reported with CBD in a patient who had previously failed to respond to olanzapine, clozapine, or haloperidol decanoate. These results suggest a limited role of CBD in treatment-resistant schizophrenia (Zuardi et al., 2006). The dose were not individually mentioned for case 1 and 2.
Four of the included studies did not report any adverse effects of CBD among patients with psychosis. CBD was well-tolerated in these patients except for mild transient sedation, hyperlipidemia, and gastrointestinal distress. Patients with schizophrenia had fewer instances of extrapyramidal symptoms, less weight gain, and a lower increase in prolactin levels.
CBD is postulated to improve cognitive performance in psychosis through the mediation of CB1 and CB2 receptor agonism at lower concentrations (Hallak et al., 2010; Solowij et al., 2018; Manseau & Goff, 2015). This cognitive improvement has been hypothesized due to the higher concentration of cannabinoid receptors in the hypothalamus, suggesting a role in superior cognitive functioning (Hallak et al., 2010). Naturalistic studies of CBD report better cognitive performance including memory, increased grey matter in the hippocampus, and fewer psychotic symptoms in patients given higher doses of CBD (Solowij et al., 2018).
The therapeutic benefits for psychosis is hypothesized due to the inhibition of anandamide re-uptake and degradation, resulting in increased anandamide levels in the brain (Manseau & Goff, 2015). Increased anandamide levels and improvements in the symptoms of psychosis were reported in another 4-week-long RCT comparing the efficacy of CBD to amisulpride for the treatment of schizophrenia (Leweke et al., 2012). Interestingly, anandamide levels were elevated in patients with acute schizophrenia compared to chronic schizophrenia, indicating a compensatory increase in an acute state (Giuffrida et al., 2004).
The present review included three RCTs (107 patients), two open-label trials (28 patients), one case series of four patients, and two case reports for cannabis-related disorders as summarized in Table 2 (Solowij et al., 2018; Crippa et al., 2013; Trigo et al., 2016a; Trigo et al., 2018; Trigo et al., 2016b; Allsop et al., 2014; Pokorski et al., 2017; Shannon & Opila-Lehman, 2015). Of the eight studies, level 2 evidence was found in three RCTs, level 3 evidence in two clinical trial, and level 4 evidence in two case reports and one case series (OCEBM, 2019). For cannabis-related disorders, there is Grade B recommendation based on majority of studies ranked at the level 2 and level 3 of evidence.
Table 2 Studies of the use of CBD and CBD-containing compounds such as nabiximols in the treatment of cannabis-related disorders and levels of evidence (1–5)
Four of these studies evaluated the efficacy of nabiximols, and four others reported the use of CBD. The doses tested ranged from 20 mg CBD to a maximum of 1200 mg/day. Nabiximols was used in spray form at doses ranging from an average of 28.9 sprays/day (equivalent to 77.5 mg THC or 71.7 mg CBD) to 40 sprays/day (equivalent to 108 mg THC or 100 mg CBD). In CBD-only studies the dose of CBD ranged from 200 to 600 mg/day in divided doses. All three RCTs in this section provided evidence for the use of nabiximols for moderate to severe cannabis use disorder. These trials tested different doses of nabiximols ranging from 21.6 mg THC and 20 mg CBD (twice a day) to 113.4 mg THC or 105 mg CBD per day. All trials reported lower withdrawal rates, better tolerance, and retention rates in the experimental group. Moreover, no serious adverse effects were reported in any of these studies. In one RCT, nabiximols (total dose of 21.6 mg THC and 20 mg CBD at 4 and 10 in evening and night, respectively) was associated with marked improvement in cannabis withdrawal symptoms, leading to shorter withdrawal times and higher retention rates (Allsop et al., 2014). In a second RCT, a fixed dose of nabiximols produced more positive results compared to self-titrated administration (Trigo et al., 2016a). Patients in the fixed-dose group had four sprays of medications every hour compared to four sprays as needed every hour in self-titrated dose group. The maximum dose was 40 sprays/day in the self-titrated dose group. Medication intake was higher with fixed doses, which were associated with fewer withdrawal symptoms compared to the self-titrated regimen (Trigo et al., 2016a). In another RCT, the efficacy and safety of nabiximols were compared to a placebo while all participants also received weekly motivational enhancement therapy (MET) and cognitive–behavioral therapy (CBT) (Trigo et al., 2018). The dose range of 4.1 to 12.8 sprays/day was reported among nabiximols group. The withdrawal scores in this study were similar in both groups (Trigo et al., 2018). Only one of the studies reported decreased appetite, while the number and severity of adverse effects were not reported or observed in the other two RCTs.
Two open-label studies testing the effectiveness of two different concentrations of CBD (200 mg/day and 600–1200 mg/day) obtained positive outcomes with doses as low as 600 mg/day (Hallak et al., 2010; Pokorski et al., 2017). These studies had a small sample size of eight (Solowij et al., 2018) and 20 (Pokorski et al., 2017) participants, respectively. In the former open-label trial with eight participants, a dose of 600 mg/day was tested, and two out of five participants completed the 7-day inpatient treatment. These two participants reported abstinence at follow-up (day 28), and the remaining three participants reported decreased use of cannabis, confirmed by blood and urine analysis. In the second group, participants took 600 mg twice a day. Two out of three participants reported abstinence and in the remaining one, cannabis use had decreased, as confirmed by blood and urine analysis. All participants showed a decrease in Cannabis Withdrawal Scale scores. The second open-label trial tested the effectiveness of 200 mg CBD in divided doses in improving cognition and depressive symptomatology among patients with chronic cannabis use, and found improvement in severity of depression, verbal learning, and memory performance, and decreased frequency of positive psychotic-like symptoms and level of distress from baseline to endpoint (Solowij et al., 2018). State anxiety increased with no change in trait anxiety, functional impairment, or accuracy on cognitive tests (Solowij et al., 2018).
The remaining studies were either case series or case reports; all found positive outcomes in withdrawal and cannabis-dependence symptoms (Crippa et al., 2013; Trigo et al., 2016b; Shannon & Opila-Lehman, 2015). Mean age in the case series was 35 years, although the first participant was 19 years old and the second was 27 years old. The case series used self-titrated nabiximols at a dose of 77.5–113.4 mg THC and 71.5–105.0 mg CBD (Trigo et al., 2016b). Moreover, all participants reported a significant reduction in craving (Crippa et al., 2013; Trigo et al., 2016b; Shannon & Opila-Lehman, 2015), quicker relief (Crippa et al., 2013), lower anxiety, and an improved sleep schedule (Shannon & Opila-Lehman, 2015). However, the case series reported increased craving scores during the first 2 weeks with a subsequent reduction in craving at week 9. CBD was well-tolerated in this patient population, except for decreased appetite reported in one study (Trigo et al., 2016b). For patients receiving nabiximols or CBD, treatment should be augmented with psychotherapeutic modalities considering the positive evidence for an effect on cravings.
The effectiveness and tolerability of CBD and nabiximols for moderate to severe cannabis use disorder was reported in several studies. The efficacy may also be due to the synergetic or additive benefits of Δ9-THC and CBD rather than CBD alone. The Δ9-THC component of nabiximols decreases the severity of withdrawal symptoms, lowering the risk of relapse (Trigo et al., 2016a). However, there is mixed evidence for the role of nabiximols in cannabis-related craving (Trigo et al., 2016a; Trigo et al., 2018; Trigo et al., 2016b). Studies that included combined motivation enhancement and behavioral response prevention strategies suggested a reduction in craving (Trigo et al., 2016a; Trigo et al., 2018). CBD is thought to modulate the euphoric, anxiogenic, psychological, and physiological effects of Δ9-THC (Crippa et al., 2013). However, these benefits of CBD alone and in combination with THC need to be explored in head-to-head studies.
The present review included two RCTs (54 patients), one open-label trial (53 patients), one retrospective chart review (72 patients), and four case reports for CBD and nabiximols use in the treatment of other psychiatric disorders. Specifically, this review looked at ADHD (one RCT), comorbidities in ASD (one open-label trial), anxiety and sleep problems (one retrospective chart review), SAD (one clinical trial), bipolar disorder (one case report), PTSD (one case report), and Tourette syndrome (two case reports), as summarized in Table 3 (Cooper et al., 2017; Barchel et al., 2018; Bergamaschi et al., 2011; Shannon et al., 2019; Zuardi et al., 2010; Shannon & Opila-Lehman, 2016; Trainor et al., 2016; Pichler et al., 2019). Of the nine studies, level 2 evidence was found in two RCTs, level 3 evidence in one clinical trial, and level 4 evidence in one retrospective chart review, four case reports (OCEBM, 2019). There is Grade B recommendation for comorbidities in patients with ASD, anxiety disorders including SAD and sleep problems, and ADHD where as bipolar disorder, PTSD and Tourette Syndrome has Grade C recommendation. However, this should be considered in the context of fewer studies of each these diagnoses.
Table 3 Studies of the use of CBD and CBD-containing compounds such as nabiximols in the treatment of other psychiatric disorders and levels of evidence (1–5)*
The oromucosal nabiximols spray was tested to evaluate its effects on cognitive performance, hyperactivity, inattention, and emotional lability in 15 participants in a placebo-controlled RCT (Cooper et al., 2017). The mean dose of nabiximols was 4.7 sprays per day (2.7 mg Δ9-THC and 2.5 mg CBD). Although an improvement in these symptoms was observed in the intervention group, it failed to reach statistical significance (Cooper et al., 2017). However, this result may not be valid or reliable due to the low power of the study.
One case report on the use of CBD by two patients with bipolar disorder showed limited to no improvement with doses of 600–1200 mg for bipolar mania in one of the patients (Shannon et al., 2019). The second patient was prescribed CBD 600 mg (5–9 days) and olanzapine (10–15 mg), followed by CBD 900–1200 mg (20–33 days), and showed improvement on the Brief Psychiatric Rating Scale (37% reduction) and Young Mania Rating Scale (33% reduction) with CBD and olanzapine, but no additional improvement with CBD monotherapy (Shannon et al., 2019). This effect was consistent with results from animal studies that modeled acute mania with dextroamphetamine (Shannon et al., 2019). The lack of effectiveness can be attributed to the shorter duration of treatment in both cases. This evidence from studies of bipolar mania should be considered in the context of different pharmacological agents responding differently to certain episodes of bipolar disorder. In animal studies, CBD induced a rapid, persistent antidepressant response by increasing brain-derived neurotrophic factor in the prefrontal cortex (Shannon et al., 2019). Given its possible antidepressant benefits, the role of CBD should be explored in unipolar and bipolar depression.
In an open-label trial involving children with ASD, Barchel and colleagues reported that a solution of CBD and Δ9-THC (1,20 ratio) was effective for hyperactivity, insomnia, self-injurious behaviors, and anxiety (Barchel et al., 2018). The median dose was 90 mg with an interquartile range (IQR) of 45–143 mg for CBD whereas The medical dose was 7 mg with IQR of 4–11 mg. In this cohort of 53 patients, 74.5% showed improvement in their comorbid symptoms, 68.4% in hyperactivity, 67.6% in self-injurious behaviors, 71.4% in sleep problems, and 47.1% in anxiety symptoms. This treatment regimen lasted for a median of 66 days. However, Salgado and Castellanos suggested guiding principles for the use of CBD in this population, including a better clinical understanding of CBD, open discussion with parents and patients, addressing their perceptions, promoting informed consent, and exercising caution in the use of CBD (Salgado & Castellanos, 2018). Patients with ASD make up a heterogeneous group of individuals with different comorbidities that should be considered.
The efficacy of CBD for SAD and PTSD was explored in three studies including one RCT, one case report, and one chart review. The RCT reported the results of a simulated public speaking test among 12 healthy control participants and 24 patients with SAD who received a single dose of CBD 600 mg or a placebo before the test. This study reported that pretreatment with CBD resulted in less anxiety, cognitive impairment, and discomfort during their speaking performance. It also resulted in a significant reduction in alertness in their anticipatory speech compared to the placebo group (Bergamaschi et al., 2011).
In a 10-year-old patient, 5 months of treatment with CBD oil (25 mg) and liquid CBD (6–12 mg) in a sublingual spray as needed was associated with less anxiety and better sleep quality, with no adverse effects (Shannon & Opila-Lehman, 2016). These results were replicated for anxiety in a recently published chart review of 72 adult patients with insomnia and anxiety (Shannon et al., 2019). Most patients in this group were given 25 mg CBD/day, while a few patients were given 50 or 75 mg/day, and one patient with schizoaffective disorder and trauma was given up to 175 mg/day. All patients showed less anxiety and improved sleep, with reductions of 65–80% in the Hamilton Anxiety Rating Scale and Pittsburgh Sleep Quality Index scores.
Nabiximols produced improvements in patients with Tourette syndrome at a much lower dose than what was used for cannabis-related disorders (Trainor et al., 2016; Pichler et al., 2019). These case reports tested two oromucosal nabiximols sprays used twice a day (total dose 10.8 mg Δ9-THC and 10 mg CBD per day) (Trainor et al., 2016), and the second also tested cannabis tincture (34 drops three times a day (Pichler et al., 2019). Both case reports found improvements in tic frequency (Trainor et al., 2016; Pichler et al., 2019), severity (Trainor et al., 2016; Pichler et al., 2019), quality of life, and social activity (Trainor et al., 2016). These treatments regimens were used for 4 weeks with the oromucosal spray form (Trainor et al., 2016) and 8 weeks for cannabis tincture (Pichler et al., 2019). The therapeutic benefits can be attributed to the anxiolytic and sleep-inducing properties of CBD (Trainor et al., 2016). It is difficult to ascertain whether these improvements were due to due to CBD, Δ9-THC, additive, or synergetic effects. The anxiolytic properties of CBD explain the attenuation of anxiety associated with the onset of tics, and the improvement in tics with a combination of Δ9-THC and CBD (Trainor et al., 2016; Pichler et al., 2019).
Adverse effects were reported in four of the studies, and included muscular seizures and spasms (Cooper et al., 2017), somnolence and changes in appetite (Barchel et al., 2018), fatigue, and sexually inappropriate behavior in a patient with developmental disorder (Shannon et al., 2019), mild sedation (Zuardi et al., 2010), and mild xerostomia (Pichler et al., 2019).
Summary of evidence
The present article provides a comprehensive review of the evidence supporting the use of CBD and CBD-containing compounds such as nabiximols to treat psychiatric disorders. CBD and nabiximols were effective in cannabis use-related disorders, and preliminary evidence was found in support of their use for other psychiatric disorders. Of the 23 studies reviewed here, level 2 evidence was found in eight RCTs, level 3 evidence in four open-label trials and one clinical trial, and level 4 evidence in one retrospective chart review, seven case reports, and two case series, according to the Oxford Centre for Evidence-Based Medicine 2011 Levels of Evidence (OCEBM, 2019). This review covers the evidence for different routes of administration, e.g. oral, inhalation spray, and sublingual. The bioavailability of these routes (11–13% for oral vs. 11–43% for inhalation) varies significantly – a factor that can impact the efficacy of different formulations.
Their antipsychotic, neuroprotective, anxiolytic, and sedating properties suggest a potential therapeutic role of CBD and nabiximols to treat various psychiatric disorders. The use of CBD at higher doses (above 1200 mg per day) showed promising results in case studies of schizophrenia and psychosis in patients with Parkinson’s disease, except in treatment-resistant cases. Regarding the use of CBD to treat anxiety disorders, its anxiolytic effect can help patients with PTSD-related and social performance-related anxiety, and nabiximols can reduce the anxiety associated with the onset of tics. There is also favorable evidence in patients with ASD for reducing hyperactivity, self-injurious behaviors, anxiety, and insomnia. Nabiximols showed no credible effect in the treatment of ADHD, while CBD was also found to be ineffective for bipolar disorder. Of all the cases examined, the strongest evidence was found for the treatment of cannabis-related disorders. The use of nabiximols yielded positive results in multiple studies of moderate to severe cannabis use disorder; however, the use of CBD alone has not been adequately documented outside a few cases and case series. Notably, CBD compounds were helpful in alleviating psychotic symptoms and improving cognitive impairment in patients across a variety of conditions.
Recommendations for future research
This review found low-level evidence for the use of cannabis and nabiximols in a variety of disorders. Despite our comprehensive literature search, only a few RCTs related to the disorders of interest were found. These RCTs were marred by a number of limitations, most importantly failure to blind the outcome assessor, participants, and research personnel (in the open-label trials). In addition, most RCTs had a small sample size, critically reducing the power of the study to draw robust conclusions. The findings of the RCTs reviewed here need to be validated via a series of larger, well planned, randomized, double-blinded, and placebo-controlled studies. The present report can be used to design and plan further studies; however, at present the use of CBD and nabiximols in clinical practice cannot be recommended with confidence due to the drawbacks noted above.
The evidence from studies included in this review can guide future trials by providing information pertaining to the dosages, formulations and routes of administration of CBD and nabiximols. Moreover, future studies should investigate different routes of administration in light of the differences in bioavailability. In view of the (albeit limited) evidence for treatment-resistant schizophrenia, the role of CBD should be explored in the early stages of psychosis or as an adjunct medication. Although CBD was ineffective for bipolar mania, its possible efficacy as an antidepressant should be assessed in studies focused on bipolar depression. Nabiximols has been helpful in cannabis-related disorder and Tourette syndrome, owing to the synergetic benefits of CBD and THC. Future studies designed to explore the comparative benefits of these treatments can shed further light on their clinical potential. Future RCTs should also consider adding first-line treatment agents as comparison arms, to ascertain the comparative efficacy of CBD in different mental disorders. Although fewer side effects were reported overall by patients in the studies reviewed here, the vulnerability to addiction to cannabinoids should not be ignored.
Limitations of the review
This review article has several limitations that should be considered. This review article provides evidence for CBD and CBD-containing nabiximols are two different pharmacological agents. Nabiximols has two active compounds and included studies do not consider the separate effects of THC VS CBD. There is need for future analyses to carefully consider their benefits individually. Only one-third of studies (8/23) in this review article are RCTs and most of these RCTs had a small sample size decreasing the power of the study to draw robust conclusions.
The evidence reviewed here favors CBD use for patients with schizophrenia and psychosis in Parkinson’s disease in four out of seven studies, except in treatment-resistant cases. There is a Grade B recommendation this diagnosis based on the levels of evidence. Nabiximols and CBD were beneficial in cannabis-related disorders in almost all studies with Grade B recommendation, resulting in a decreased risk of withdrawal symptoms and dependence among participants. The effect on cannabis-related craving was pronounced, with an additive benefit from the use of psychotherapeutic options such as MET or CBT. One open-label trial suggested favorable evidence for the use of cannabinoids CBD and Δ9-THC for hyperactivity, self-injurious behaviors, and anxiety symptoms in patients with ASD with Grade B recommendation. CBD was helpful in patients with anxiety and insomnia related to SAD and PTSD in one chart review. Nabiximols was found to be effective in reducing the frequency and severity of tics and improving the quality of life in patients with Tourette syndrome according to case reports. There was no firm evidence to support CBD to treat bipolar mania (one case report) or nabiximols (one RCT) to treat ADHD. There is Grade B (moderate) recommendation for attention deficit hyperactivity disorder. Grade C recommendation (weaker) exists for insomnia, anxiety, bipolar disorder, posttraumatic stress disorder, and Tourette syndrome. These recommendations should be considered in the context of limited number of available studies. The authors recommend well-planned randomized controlled trials to further study the benefits of CBD and CBD-containing options such as nabiximols in patients with psychiatric disorders. It is also important to assess the individual pharmacodynamic and pharmacokinetic effects of CBD and Δ9-THC in different treatments.
Medicinal cannabis for psychiatric disorders: a clinically-focused systematic review
Medicinal cannabis has received increased research attention over recent years due to loosening global regulatory changes. Medicinal cannabis has been reported to have potential efficacy in reducing pain, muscle spasticity, chemotherapy-induced nausea and vomiting, and intractable childhood epilepsy. Yet its potential application in the field of psychiatry is lesser known.
The first clinically-focused systematic review on the emerging medical application of cannabis across all major psychiatric disorders was conducted. Current evidence regarding whole plant formulations and plant-derived cannabinoid isolates in mood, anxiety, sleep, psychotic disorders and attention deficit/hyperactivity disorder (ADHD) is discussed; while also detailing clinical prescription considerations (including pharmacogenomics), occupational and public health elements, and future research recommendations. The systematic review of the literature was conducted during 2019, assessing the data from all case studies and clinical trials involving medicinal cannabis or plant-derived isolates for all major psychiatric disorders (neurological conditions and pain were omitted).
The present evidence in the emerging field of cannabinoid therapeutics in psychiatry is nascent, and thereby it is currently premature to recommend cannabinoid-based interventions. Isolated positive studies have, however, revealed tentative support for cannabinoids (namely cannabidiol; CBD) for reducing social anxiety; with mixed (mainly positive) evidence for adjunctive use in schizophrenia. Case studies suggest that medicinal cannabis may be beneficial for improving sleep and post-traumatic stress disorder, however evidence is currently weak. Preliminary research findings indicate no benefit for depression from high delta-9 tetrahydrocannabinol (THC) therapeutics, or for CBD in mania. One isolated study indicates some potential efficacy for an oral cannabinoid/terpene combination in ADHD. Clinical prescriptive consideration involves caution in the use of high-THC formulations (avoidance in youth, and in people with anxiety or psychotic disorders), gradual titration, regular assessment, and caution in cardiovascular and respiratory disorders, pregnancy and breast-feeding.
There is currently encouraging, albeit embryonic, evidence for medicinal cannabis in the treatment of a range of psychiatric disorders. Supportive findings are emerging for some key isolates, however, clinicians need to be mindful of a range of prescriptive and occupational safety considerations, especially if initiating higher dose THC formulas.
The Cannabaceae family is a comparatively small family of flowering plants encompassing 11 genera and approximately 170 different species, a small number of which elicit a range of varying psychoactive effects . Several medical applications have been studied over the past decades, with the National Academies of Sciences, Engineering and Medicine (NASEM) recently holding the position that cannabis and cannabinoids demonstrate conclusive or substantial evidence for chronic pain in adults, chemotherapy-induced nausea and vomiting and spasticity in multiple sclerosis, with limited evidence for use in increasing appetite in HIV/AIDS patients and improving symptoms of post-traumatic stress disorder (PTSD) .
While there is increasing psychiatric interest (and debate) regarding the potential mental health applications (in concert with concerns over the potential for triggering latent psychosis), historical evidence for the use of cannabis in mental health conditions is remarkably ancient. For instance, the Shen-nung Pen-tsao Ching (Divine Husbandman’s Materia Medica) described its benefit as an anti-senility agent [3, 4], while in the Assyrian culture, cannabis was indicated as a drug for grief and sorrow [5, 6]. Sections of the Indian Atharva Veda (1500 BCE) suggest bhanga (Cannabis) exerted anxiolytic effects [5, 7], while in 1563, Da Orta  described cannabis as allaying anxiety and engendering laughter. With respect to modern use, contemporary consumers of cannabis report (as assessed via meta-analysis of patient usage data) that pain (64%), anxiety (50%), and depression/mood (34%) are the most common reasons .
Increasing scientific research, conducted over recent years, has seen the regulatory pendulum swinging away from the United Nations Single Convention on Narcotic Drugs in 1961 (which recommends enforcement of cannabis use as illegal) , towards consideration of its potential use in medical conditions. Recent scientific evidence ascribes anxiolytic, neuroprotective, antioxidant, anti-inflammatory, antidepressant, anti-psychotic and hypnotic pharmacological actions due to several phytochemicals commonly found in the cannabis genus [11, 12].
While Δ 9 -tetrahydrocannabinol (THC) is considered the main psychoactive constituent, other cannabinoids have also revealed less potent psychotropic effects. These include cannabidiol (CBD) , Δ 8 -tetrahydrocannabinol , and other less-studied cannabinoids including cannabinol (CBN) and Δ 8 tetrahydrocannabivarin (THCV) . Further, many other constituents such as the terpenes (i.e. volatile organic compounds found mainly as essential oils in many plants), also provoke a range of biological effects, and produce the characteristic aroma of the plant . The hundreds of cannabis chemovars or varieties (commonly referred to as strains) developed over millennia have unique and complex constituent profiles, of which each may provide targeted therapeutic usage due to the unique synergistic combination of plant chemicals. Some pharmaceutical preparations have attempted to isolate the key constituents (there are over 140 phytocannabinoids ) to provide standardised formulas that may harness this ‘entourage effect’ [16, 18], while being able to provide batch-to-batch assurance of the medicine.
While other reviews have covered cannabis’ use in a range of conditions (cf. Whiting et al. 2015  for a general review of evidence for medicinal cannabis), none to date have provided both a systematic and ‘clinically-focused’ review on psychiatric disorders. As the focus was on emerging data for the use of mental health disorders, we omitted addiction and neurological disorders, which have been extensively covered elsewhere [10, 19], cf. pain [20,21,22], cf. epilepsy [23, 24], cf. movement disorders . A further motive for this paper focusing solely on psychiatric disorders, concerns cannabis users noting that self-reported anxiety, insomnia, and depression symptoms are amongst the most common reasons for usage .
Thus, the primary purpose of this paper is to provide a systematic review of the current state of evidence in the emerging field of cannabinoid therapies for psychiatric disorders (PTSD, generalised anxiety disorder, social anxiety, insomnia, psychotic disorders, and attention-deficit hyperactivity disorder: ADHD). In addition, this review provides clinical prescriptive guidelines and consideration of both safety and occupational public health issues. We also provide discussion on considerations for future research in the field. Our intention was to provide a review of the extant literature to inform a discussion with clinical context and appropriate recommendations.
Due to the field still being in its infancy, a broad inclusion criteria was applied to the available data. The purpose was to locate human studies involving whole cannabis plant medicines and cannabis-derived isolates (singularly or in combinations) for the treatment of major psychiatric disorders or mental health symptoms. Synthetic cannabinoid analogues (e.g. nabilone) and THC isomers (e.g. dronabinol) were omitted as these fall under the auspices of a pharmaceutical-focused review (as these are classified as pharmaceutical ‘drugs’).
Major electronic databases including OVID MEDLINE, Cochrane Central Register of Controlled Trials, Health Technology Assessment Database, Allied and Complementary Medicine and PsychINFO were accessed for data up to July 2019. Initially, data were sought for meta-analytic or systematic review level epidemiological evidence (as there is sufficient data available) on the cross-sectional or longitudinal association of cannabis use and individual psychiatric disorders or symptoms. This was undertaken to assess any deleterious relationship between cannabis and psychiatric disorders. We then specifically sought any literature involving interventional human trials and observational studies, including case studies (due to deficient randomised controlled trials [RCTs] in this emerging area). We included studies with any sample size or age or gender, which used either inhalant, oral, or transdermal administration of medicinal cannabis or cannabis-derived isolates. All studies in English were assessed for inclusion (see supplementary data for the PRISMA flow chart for the number of human clinical trials or case studies excluded/included). The results are presented to firstly cover the major current epidemiological evidence, and then next all available clinical trial or case study data.
The following search terms were used to locate human studies or case report publications:
TITLE: cannab* OR THC OR tetrahydrocannabinol OR canab*
TITLE: depression or depressive or mental illness* or mental disorder* or mental health or mood disorder* or affective disorder* or anxi* or panic disorder or obsessive compulsive or adhd or attention deficit or phobi* or bipolar or psychiat* or psychological or psychosis or psychotic or schizophr* severe mental* or serious mental* or antidepress* or antipsychotic* or post traumatic* or personality disorder* or stress
In summation, 481 articles were located, which was reduced to 310 after duplicates were removed. Of these, 13 studies fitted the eligibility criteria as clinical studies of cannabis-based treatments for symptoms of psychiatric disorders. The full search and screening process is displayed in the supplementary data. There were insufficient homogenous studies to perform a meta-analysis.
The endocannabinoid system has been found to be a modulator of anxiety and mood, with recent data showing that cannabinoids or substances which target this system may interact with specific brain regions, including the medial prefrontal cortex, amygdaloid complex, bed nucleus of stria terminalis, and hippocampus . Interaction with the CB1 receptor has a modulating effect on GABAergic and Glutamatergic transmission , while also influencing the hypothalamic pituitary adrenal (HPA) axis, immune system activation, and neuroplastic mechanisms. In respect to specific psychotropic mechanisms of action, the anxiolytic (and antidepressant effects) may also in part be mediated via CBD’s serotonergic effects via 5-HT1A receptor activation , and THC’s CB1 receptor agonism [30, 31]. It is worth noting that studies have demonstrated that CBD may partially inhibit the psychoactive effects of THC, with CBD and THC having demonstrated differing symptomatic and behavioural effects on regional brain function [32,33,34,35].
As in the case of certain other psychiatric symptoms and disorders, epidemiological evidence indicates that there is a relationship between cannabis use and anxiety symptom levels. This association (assessed by Kelzior and colleagues  via meta-analysis of 31 studies) has to date only been found to be weak, and based largely on cross-sectional data. Thus, it may be that those with anxiety seek cannabis treatment, rather than a causal effect occurring from cannabis use. Longitudinal data is also not convincing due to the bias of one study with a large odds ratio included in their meta-analysis. However, a stronger positive association was revealed between anxiety and cannabis use disorder. Other longitudinal data involving the USA-based National Epidemiologic survey on Alcohol and Related Conditions  confirms there is no obvious causal inference. The study included individuals with a diagnosis of any anxiety disorder during the initial 4-year data collection period, comparing cannabis nonusers to users, and also individuals with cannabis use disorder at a later time point on a range of psychosocial measures. Results revealed that, when controlling for baseline confounders, no significant relationship was found with cannabis use and a greater frequency of anxiety.
While to date no human trials could be located for treatment of Generalised Anxiety Disorder using whole cannabis plant extracts or combined isolates, there was one study identified testing CBD (Table 1) for Social Anxiety Disorder. One small preliminary double-blind RCT compared the effects of a simulated public speaking test on treatment-naïve patients with social anxiety (n = 24) versus healthy control participants (n = 12) . Each group received a single acute oral dose of CBD (600 mg) 1.5 h before the test, or matching placebo. Results revealed that pre-treatment with CBD significantly reduced anxiety, cognitive impairment and discomfort in the social anxiety group’s speech performance, and significantly decreased hyper-alertness in their anticipatory speech compared to the placebo group (which presented higher anxiety, cognitive impairment, discomfort, and higher alertness levels). Neuroimaging research has also revealed that in individuals diagnosed with social anxiety, cerebral blood flow may be altered via CBD. One study employed fMRI in 10 treatment-naïve patients with social anxiety who were given 400 mg of oral CBD or placebo in a double-blinded crossover manner. Relative to placebo, 400 mg of CBD was associated with significantly decreased subjective anxiety, with blood flow being modulated in the left parahippocampal gyrus, hippocampus, and inferior temporal gyrus, and the right posterior cingulate gyrus . This suggests that CBD’s activity may occur via interaction with the limbic and paralimbic brain areas.
Due to the small sample sizes, the above data needs to be considered with caution. Further, appropriate and considered treatment of anxiety disorders with cannabinoid therapies is crucial due to the complex relationship with substance use disorders, often requiring a more complex biopsychosocial approach . With this context in mind, CBD (being a non-intoxicant compared to THC) may be a more preferable option, having also shown anxiolytic effects in preclinical studies .
In respect to planned or ongoing research, one study in Colorado USA has just commenced and is exploring the anxiolytic effects of vaporised or ingested THC/CBD in differing ratios (1:0, 1:1, 0:1) in people with mild-moderate anxiety . Another study is assessing the effect of CBD on reducing symptoms of anxiety disorders in a youth cohort (12–25 years old). The Australian-based study is a 12-week open-label pilot, which aims to see if 200 mg–800 mg of oral CBD (titrated depending on age, tolerability, and efficacy) is safe and effective for a youth population .
Post-traumatic stress disorder (PTSD)
Whole plant cannabis use for the management of PTSD symptoms has been identified in usage analyses , and in particular in returned armed services veterans . The proposed neurobiological mechanisms by which medicinal cannabis may assist with PTSD are varied and mostly derived from animal research. There are high concentrations of endocannabinoid receptors in the prefrontal cortex, amygdala and hippocampus , having a role in fear acquisition and extinction . There is strong evidence revealing that a disruption of the endocannabinoid system impairs fear extinction in CB1 knockout mice, suggestive of a critical role of CB1 receptors (and thereby potentially THC) being related to the extinction of fear [48,49,50].
One survey involving a convenience sample of 170 patients via a medical cannabis dispensary in California evaluated a range of health elements, the frequency of cannabis use, and general mental health . Results revealed that those with high PTSD scores (assessed via The PTSD Checklist-Civilian Version) were more likely to use cannabis to assist with mental health coping, in addition to improving sleep, when compared with those with low PTSD scores. In particular, cannabis use frequency was greater among those with high PTSD scores who often used this for improving sleep. While there is increased use of cannabis in those with PTSD, there is currently no firmly supportive epidemiological data. A cross-sectional case control study of veterans showed that regular users do not have lower PTSD symptoms than non-users .
A recent open label retrospective analysis of case study data from 11 adults with PTSD assessed the patients over 8 weeks of CBD treatment (capsule or spray; mean dosage at week-8 of 49 mg) . Results revealed that a reduction in mean PTSD symptoms occurred in 28% of the sample, as assessed on the PTSD checklist for DSM-5 (PCL-5). Statistical data analysis was not conducted, however, and thereby it is not possible to draw firm conclusions. Another retrospective study analysing PTSD symptoms collected during 80 psychiatric evaluations of patients applying to the New Mexico Medical Cannabis Program during 2009 to 2011 , revealed more supportive findings. The data identified a greater than 75% reduction in Clinician Administered Posttraumatic Scale for DSM-IV (CAPS) symptom scores when patients with PTSD were using cannabis compared to when they were not. While this study had a small sample, and is a retrospective analysis that has some methodological weaknesses, a 75% reduction on the CAPS is a compelling result, and has spurred recent RCTs which are currently in recruitment [55, 56].
Phytocannabinoids and terpenes have a potential application for modulation of the endocannabinoid system and the 5HT1A receptor to provide an antidepressant effect . No RCTs to date have been conducted on the primary outcome of depression. Three studies assessing oral-administered nabiximols (i.e. botanically derived preparation containing standardised levels of THC, CBD, terpenes and flavonoids from cannabis) for other conditions (multiple sclerosis and cannabis withdrawal) found no significant effect on the secondary outcome of depression [57,58,59]. It is worth noting that one study involving cancer patients using nabiximols showed a significant reduction in mood occurred for those who used the highest dose (11–16 sprays per day) compared to the placebo . Further, some epidemiological evidence has revealed a greater level of depressive symptoms in heavy cannabis users compared to light-users and non-users . Due to this, higher dose THC should be avoided in people with major depressive disorder (MDD) or low mood. However, a cross-sectional survey on patterns of use and perceived efficacy suggested that in over 1429 participants identified as medical cannabis users, over 50% reported using medicinal cannabis specifically for depression .
Anecdotal survey evidence abounds for the soporific effect of cannabis, with sufferers of a range of conditions including pain, anxiety and PTSD reporting that it assists in the management of insomnia [51, 63,64,65,66,67,68]. While this may commonly take the form of whole plant cannabis being administered via vaporised inhalation, isolated CBD may also be of benefit. An example case study detailed in the literature concerns a 10-year-old girl with prior early childhood trauma . A trial of oral CBD oil (25 mg) resulted in a decrease in this patient’s anxiety, and improvement in the quality and quantity of her sleep. A more substantial retrospective case series of 72 adults given CBD for anxiety and sleep complaints at a psychiatric clinic (as an adjunct to usual treatment) assessed patient data monthly over 12 weeks . Anxiety scores on the Hamilton Anxiety rating Scale (HAMA) decreased within the first month in 79% of the sample and remained low during the study duration. The Pittsburgh Sleep Quality Index score also improved within the first month in 67% of the sample, but fluctuated over time. It should be noted that the data were not analysed for statistical significance, and it appeared that the sub-sample presenting primarily for anxiety treatment did not fare as well as the cohort presenting primarily with sleep issues.
A study by Johnson et al.  tested the long-term safety and tolerability of a THC/CBD spray and a THC spray in relieving pain in patients with advanced cancer. A total of 43 patients were continued on a previous three-arm RCT involving an open label administration of a self-titrated THC/CBD spray (n = 39) or THC spray (n = 4) (2.7 mg) assessed over a 5-week period. While results revealed a consistent reduction in perceived pain, participants also reported a decrease in their insomnia, which also reflected less fatigue. Cannabinoids may have a dual effect of lessening pain (which makes it easier to sleep), in addition to their direct soporific and anxiolytic effects being mediated in part via serotonergic activity.
As detailed above, the evidence for this use is currently very weak, and to date no RCTs were located in the literature specifically assessing cannabinoid isolates or whole plant formulas. As of late 2018, there is however, a clinical trial taking place in Australia assessing cannabinoid treatment in chronic insomnia . The study, based in Western Australia, is aiming to enrol 24 participants aged (25–70 years) who have insomnia (defined as difficulty initiating or maintaining sleep for 3 or more nights per week for at least 3 months). The intervention involves an oral MC extract (ZTL-101) or placebo given in a cross-over manner for a study period of approximately 2 months. Participants will be assessed via the clinically-validated insomnia scales, an actigraph watch, and will be assessed in a sleep centre after 2-weeks.
Consistent evidence has shown that there is a relationship between schizophrenia and cannabis use [73,74,75]. Heavy cannabis use may proceed to a diagnosis of the disorder, however, increased use may also result from ‘self-medication’. Cannabis use is cross-sectionally associated with more severe symptoms of psychosis in young people who do not meet the threshold for schizophrenia, and appears to be one high-risk component for the tumescence of the disorder . More importantly, there is also longitudinal data to support a causal relationship [77,78,79]. A 2016 meta-analysis showed that while general lifetime use is not cross-sectionally associated with increased risk of psychosis, there is a robust relationship demonstrated in recent or current use in ultra-high-risk (UHR) adolescents with a DSM-diagnosed cannabis use disorder . A recent prominent study has corroborated this finding. Data from 11 sites across Europe and Brazil involving patients with first-episode psychosis versus population controls, revealed that daily cannabis use was associated with increased odds of a psychotic disorder occurring compared with never-users, with nearly five-times increased odds for daily use of high-potency THC types of cannabis . Several academics [82,83,84,85,86] have disputed these findings and comment that while there is a relationship, cannabis use is not causally related to increased psychosis risk (potentially due to a range of confounders e.g. correlated genetic liabilities or indirect and bidirectional processes). However, di Forti and colleagues (the study authors)  maintain that the data does indeed support this causal association, and that other research has flawed elements (e.g. previous Mendelian Randomisation studies using imprecise measurements of cannabis use).
It is of note that schizophrenia risk alleles are linked to cannabis use in a general population . Regardless, the transition rates from a general population of cannabis users to schizophrenia is very low and can be considered to be part of a constellation of various potential gene-environment interactions. Several key genes have been implicated as potentially modulating the risk of developing schizophrenia after early cannabis use: BDNF, CNR1, COMT, AKT1, and DRD2 genes [88, 89]. There is also a likely increased susceptibility when a combination of these at-risk alleles from these single nucleotide polymorphisms (SNPs) are combined with childhood trauma [90, 91].
The apprehension regarding the promotion of psychotic symptoms are primarily based on the THC constituent of cannabis, a CB1 receptor agonist, which is the primary psychoactive phytochemical. This effect has been shown to be more prominent in users of high-THC cannabis, or in chronic heavy users . Thus, THC should be avoided in people with or at risk of schizophrenia. Exposure to THC increases extracellular dopamine and glutamate and decreases GABA concentrations in the prefrontal cortex . A recent double-blind crossover RCT investigated whether altered striatal glutamate (measured via proton magnetic resonance spectroscopy) was a mediating biomarker from intravenously administered THC in 16 healthy participants . Results revealed that that an increase in striatal glutamate levels may underlie acute cannabis-induced psychosis, while lower baseline levels may provide a valid biomarker of greater sensitivity to its acute psychotomimetic effects.
The psychotropic effects of THC may mimic the presentation of psychotic symptoms, including paranoia, sensory alteration, euphoria, and hallucinations . In laboratory-based research, people with schizophrenia appear to be even more sensitive to the psychosis-inducing effects of THC than healthy controls . In contrast to THC, as mentioned in the introduction, CBD may in fact provide an opposing effect to THC albeit more research into this mechanism is required. Additional effects include the inhibition of anandamide breakdown via fatty acid amide hydrolase (FAAH) blocking effects, and anti-inflammatory effects [97, 98].
CBD is well-tolerated with minimal deleterious psychoactive effects (although some psychological effects are evident due to modulation of the 5HT1A receptor and enhanced anandamide signalling) . Due to this, studies have primarily employed isolated CBD, however this work could potentially be extended to formulas from whole-plant strains which are high in CBD (> 10 mg/g) and lower in THC (< 4 mg/g). These preparations may also contain other yet-to-be-studied compounds from the plant which may be beneficial for the positive or negative symptoms of schizophrenia.
In respect to current research, aside from an initial index case study conducted by Zuardi, Morais  in 1995, who showed that 1500 mg of CBD administered for 26 days was beneficial for treatment-resistant schizophrenia, three clinical studies exist to date. A study by Leweke, Piomelli  tested in a double-blind, RCT design 600–800 mg/day of oral CBD vs the antipsychotic amisulpride over 4 weeks in 42 patients. While both treatments were safe and led to significant non-differential clinical improvements, the CBD arm had a superior side-effect profile. CBD also significantly increased anandamide levels, which was associated with clinical improvement. Another double-blind parallel-group trial, involving 88 patients with schizophrenia who were given either oral CBD (1000 mg/day) or placebo adjunctively to existing antipsychotic medication revealed after 6 weeks of treatment that the CBD group had lower levels of positive psychotic symptoms on the Positive and Negative Syndrome Scale (PANSS), and were more likely to have been rated as improved via clinician-ratings . While these studies were supportive of CBD, a recent double-blind RCT by Boggs, Surti  found no benefit for 600 mg/day of CBD in comparison to placebo. The 6-week study involving 36 patients with schizophrenia revealed that both placebo and CBD PANSS scores improved, but no Group × Time effect was evident. The CBD was well-tolerated, however, and more sedation was evident in the CBD group compared to placebo.
Furthermore, CBD may confer some protective effects in young people at clinical high-risk for psychosis (n = 33), as a recent single-dose RCT found that 600 mg of CBD temporarily normalised aberrant brain activity in the parahippocampal, striatal, and midbrain areas, which is associated with increased psychosis risk . Currently, an ongoing clinical trial in the United Kingdom is assessing the efficacy of 600 mg of CBD per day for reducing symptoms of psychosis in young people at clinical high-risk for psychosis .
To date no clinical trial has assessed cannabinoids for the treatment of bipolar disorder (in respect to maintaining euthymia, or as a treatment of hypo/mania or depression), although there is a potential role of the endocannabinoid system in the disorder, as detailed above. Initial case reports contend this approach may not however be of benefit . Two patients diagnosed with DSM-IV Bipolar type I disorder, and presenting with mania, were provided adjunctive CBD (titrated to 1200 mg per day) after receiving placebo for an initial five-day period. On Day 31, CBD treatment was discontinued and replaced by placebo for five days. While the first patient showed symptom improvement while on olanzapine plus CBD, she showed no additional improvement during CBD monotherapy, while the second patient had no symptom improvement with any dose of CBD during the trial. Both patients tolerated CBD very well and no side-effects were reported, despite no obvious effect on reducing mania.
Evidence has revealed that adults with ADHD may self-medicate with cannabis as a coping strategy for a range of potential effects . Off-label use in the US for this application has been noted despite a relative deficit of evidence for this use . One study was located, the “Experimental Medicine in ADHD-Cannabinoids” pilot RCT, using nabiximol (cannabinoid/terpene combination) oromucosal spray in 30 adults with ADHD for 6 weeks . The primary outcome was cognitive performance and activity level (as measured by head movements) using the Quantitative Behavioural Test. Secondary outcomes included ADHD and emotional lability symptoms. While a trend towards significance occurred in favour of nabiximols, no significant difference was revealed on the primary outcomes. Notably, the use of nabiximols did not impair cognition. For secondary outcomes, the combination of note was associated with a nominally significant improvement in hyperactivity/impulsivity scores on the investigator-rated Conners Adult ADHD Rating Scale. The combination was well-tolerated, however, a serious adverse event involving muscular seizures/spasms occurred in the active group. While not definitive, this study provides preliminary evidence supporting the self-medication theory of cannabis use in ADHD and the need for further studies of the endocannabinoid system in ADHD. Results, however, did not meet significance following adjustment for multiple testing, and it should be recognised that the sample size was small, thus a more robust sample would be better placed to determine the true effect.
As the present data indicates, the current field of cannabinoid therapeutics in psychiatry currently provides no convincing evidentiary support for use in any mental health application. More research is urgently needed, and many RCTs are currently being undertaken; thereby the landscape will change rapidly over the next several years. Currently, the most promising (although inconclusive) evidence is for CBD as an adjunctive treatment in schizophrenia, with an additional isolated study showing efficacy in social anxiety, and weak data suggesting a potential effect for ADHD symptoms. The evidence also tentatively suggests that a role exists for cannabinoids in PTSD, and also in reducing insomnia, which may also commonly occur in chronic pain. For other plant-derived cannabinoid therapy applications for psychiatric symptoms/disorders (e.g. several affective disorders) no firm conclusions can currently be drawn.
Clinical prescriptive considerations
It initially should be recognised by clinicians that, as detailed above, weak evidence currently exists in the field, thus this prescriptive advice should be taken in the context of evolving research. The first consideration faced by a clinician (in a legal jurisdiction) with a patient who is interested (or for clinician-initiated prescription) in using cannabis medicinally, is whether this is medically appropriate for them. A thorough screening firstly needs to occur, with Canadian British Columbian Physician guidelines  suggesting that clinicians initially assess:
Age – higher-dose THC forms not advised in people < 25 years of age;
If a personal history or family history of psychosis is present, and if so, no THC is advised;
Any current or past drug or alcohol misuse or dependence (avoid especially in individuals with cannabis dependence or misuse);
Cardiovascular or respiratory diseases (avoid or use caution);
Current medications which may interact with cannabis; and
Pregnant or planning or conceive or breastfeeding (avoid).
Next, if no contraindication is apparent, medical consideration can be given to what potential clinical application the MC may present for, and the cannabis formulation or isolate/s that may be appropriate for them. Given the complexity of MC whole-plant formulas (and the current challenge to standardise for batch-to-batch consistency), companies have primarily tested cannabinoid isolates and analogues. The most studied including nabiximols (Sativex), nabilone (Cesamet), and dronabinol (Marinol). While this may provide more pharmacological assurity, such an approach also negates the potential of unique genetic chemovars of cannabis which may provide specific therapeutic activity due to a complex synergistic interaction of constituents (known as the entourage effect). Patient preference may also be towards vaporisation of dried raw material . To this end, specific prescriptive considerations need to be adopted, including:
Determining patient preference regarding administration – vaporisation (via specific devices), inhalation via traditional smoking apparatuses, oral dose (i.e. capsule, oil or in some cases food product), sublingual via lozenges or sprays. Note that each has a different onset of action and half-life. Inhalants will provide a more instantaneous effect (due to the alacrity of THC decarboxylation), whereas oral forms will take longer e.g. 45–90 min to take effect. Both forms of administration may be advisable to provide flexible symptom management;
Patient’s personality in terms of the effects of higher THC formulas. Avoidance of higher THC formulas should occur in youth and in those with paranoid personalities;
Potential for abuse (with greater theoretical potential in vaporised/inhaled forms [which also carries additional general health consequences]);
When the application should occur with respect to occupational and carer responsibilities and driving. Note that there is the potential to prescribe different cannabis preparations which contain differing levels of THC and CBD, with higher dose THC applications being applied preferentially in the evening.
CBD dosage (based on current evidence) varies according to disorder, age, weight, and potentially pharmacogenetic differences. Most research tends to focus on a range of 200 mg–800 mg per day . In respect to THC-containing formulas, it has been advised to be cautious exceeding 20 mg per day due to potential side effects , and people may find a psychotropic effect with as little as 1 mg–2.5 mg per dose.
As mentioned above, there are a myriad of potential cannabis chemovars that can be developed, each with unique medicinal applications. However, to maintain pharmacodynamic/kinetic consistency, at present, the two major constituents commonly standardised for are THC and CBD (in some cases select terpenes are also included). THC provides, as indicated above, the primary psychotropic effect, and higher doses may be preferable for the amelioration of pain and inducement of sleep . Further, it may provide an acute mood elevating effect in some people, however as mentioned, this may also elicit symptoms of paranoia, anxiety, and cognitive impairment (and in higher doses may actually impair mood). This effect may potentially be opposed by CBD (and/or other less studied cannabinoids), however data is mixed as to this effect. Additionally, the findings are not clear cut, with users of cannabis (to treat anxiety) having a statistically significant preference for higher THC/lower CBD containing cannabis cultivars .
Clinicians needs to be aware that cannabinoid therapies may elicit a range of side effects. In respect to potentially expected side-effects from cannabinoid interventions, occasional adverse effects revealed in clinical trials include co-ordination problems, dizziness, disorientation, euphoria, drowsiness or fatigue, dry mouth, nausea and gastrointestinal upsets . Due to this, regular monitoring is advised, especially when commencing treatment in cannabis-naïve patients.
The previously cited report conducted by the National Academy of Sciences  on the health effects of cannabis and cannabinoids cites limited evidence that cannabis use increases the rates of initiation of other psychoactive drugs. Additionally, while there are concerns over the relationship with schizophrenia, no firm evidence shows any association between cannabis use and the likelihood of developing bipolar disorder. Further evidence is suggestive that smoking cannabis on a regular basis is associated with cough and phlegm production, while limited evidence exists suggestive of a statistical association between cannabis use and ischaemic stroke and/or acute myocardial infarction. Evidence is noted to exist for the association between increased cannabis use frequency and progression to developing problematic cannabis use , as well as potential respiratory infections/disorders (especially in the use of poor-quality raw material). Clinicians need to balance these concerns together with the potential benefits, especially regarding the potential for lesser harm from other prescriptive or illicit options in patients managing psychiatric and pain conditions.
Increased attention to the influence of pharmacogenetics factors is advised, with several genes being identified that may differentially affect cannabinoid pharmacokinetics and pharmacodynamics. A recent review led by Hryhorowicz  characterised pertinent genes with relevant interaction with cannabis into three broad categories: Receptor genes (CNR1, CNR2, TRPV1, and GPR55), transporter genes (ABCB1, ABCG2, SLC6A) and pharmacokinetic/metabolism (CYP3A4, CYP2C19, CYP2C9, CYP2A6, CYP1A1, COMT, FAAH, COX2, ABHD6, ABHD12). Research into the pharmacogenomic influence is however nascent, with most of the focus being on the relationship with cannabis dependence (e.g. CNR1 receptor SNPs which shows no obvious association), or schizophrenia (COMT, DRD2 SNPs showing a stronger correlation) [81, 114]. Further exploration of FAAH SNPs differentially affecting people’s response to CBD is also of value (given its important role in inhibiting the degradation of anandamide).
Occupational and public health considerations
Occupational health and safety issues also exist in consideration with medicinal cannabis users. Workplace safety concerns have been raised in relation to the potential for medicinal cannabis use to impair judgement and psychomotor skills, especially in relation to motor vehicle use, operation of fixed and mobile plants particularly heavy industrial machinery, and the potential for risk-taking behaviors and those working in safety sensitive positions [115, 116]. Employers have a ‘duty of care’ to provide safe and healthy workplaces, which includes the management of alcohol and drug use and their potential to create unsafe workplaces or practices. Workplace drug testing (WDT) is common in some industries including mining, transportation and correctional services . Employees in building, transportation, maritime and mining operations cannot use drugs, legal or illegal, if they could impair their ability to safely undertake their duties . However, the presence of a drug, or its metabolite, in a person’s system is not always proportional to cognitive impairment . In addition, WDT does not discriminate between recreational or medicinal use and could place medicinal cannabis patients at risk of discrimination or unfair dismissal. Implementation of WDT should be balanced with a greater knowledge on the dose response relationship between cannabis-based medicines and their potential side effects.
Medicinal cannabis patients may also be subject to mobile drug testing in jurisdictions such as Australia. The salivary testing process is inefficient for assessment of cannabis related driving impairment because the tests can trace THC in saliva for days after consumption, long after any cognitive impairment has abated. The potential impact of medicinal cannabis on function will vary with dose, the length of usage (tolerance), route of administration (oral versus smoking),  and saliva THC levels are not direct measures of cognitive status. Further, the concentration of THC in urine does not correlate with cognitive function . Conversely, in the US, a whole blood THC level of 5 ng/mL has been established as a legal limit for driving in states where cannabis has been legalised . Guidelines and strategies for the specific risk management of cannabis in the workplace have been published in North America [120,121,122,123]. However, Australia is yet to publish its own risk management guidelines relating to medicinal cannabis in the workplace, although generic workplace alcohol and drug risk management guidelines could be adapted in the interim .
Future research considerations
While research is rapidly advancing, there is a challenge regarding the adequate blinding of medicinal cannabis studies (due to the obvious psychotropic effect, and lack thereof in cannabinoid-removed controls) . This may be addressed via cannabis naïve participants with psychomimetic controls (e.g. atropine; these however have the innate challenges of being biologically active themselves); adequate assessment of un/blinding; and use of varying levels of THC within the study. It still should, however, be taken in the context of other psychiatric or neurological research, with opioids and benzodiazepines also eliciting an obvious psychophysiological effect, and the acceptance of the research demonstrating analgesic and anxiolytic effects, respectively.
Currently the evidence is nascent and too weak to recommend cannabinoid-based interventions for a range of psychiatric disorders. While encouraging, research is only just beginning to determine whether cannabis or its isolates may or may not be effective for this application, and clinicians need to be mindful of several safety considerations (as articulated above). The most promising (although inconclusive) evidence is for CBD as an adjunctive treatment in schizophrenia, with an additional isolated study showing efficacy in social anxiety, and some data suggesting a potential effect for PTSD and ADHD symptoms. The data also tentatively suggests that a role exists for cannabinoids in reducing insomnia, which may also commonly occur in chronic pain. Given the generally favorable safety profile of cannabinoids observed across the observational studies and clinical trials conducted to date, there is clearly a strong case for encouraging further research.