Cannabidiol use and effectiveness: real-world evidence from a Canadian medical cannabis clinic Open Access This article is licensed under a Creative Commons Attribution 4.0 International The cannabis compound known as CBD is being touted as a treatment for a variety of conditions. But the substance’s uncertain legal status is stalling serious investigation.
Cannabidiol use and effectiveness: real-world evidence from a Canadian medical cannabis clinic
Open AccessThis article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Cannabidiol (CBD) is a primary component in the cannabis plant; however, in recent years, interest in CBD treatments has outpaced scientific research and regulatory advancement resulting in a confusing landscape of misinformation and unsubstantiated health claims. Within the limited results from randomized controlled trials, and lack of trust in product quality and known clinical guidelines and dosages, real-world evidence (RWE) from countries with robust regulatory frameworks may fill a critical need for patients and healthcare professionals. Despite growing evidence and interest, no real-world data (RWD) studies have yet investigated patients’ reports of CBD impact on symptom control in the common expression of pain, anxiety, depression, and poor wellbeing. The objective of this study is to assess the impact of CBD-rich treatment on symptom burden, as measured with a specific symptom assessment scale (ESAS-r).
This retrospective observational study examined pain, anxiety, depression symptoms, and wellbeing in 279 participants over 18 years old, prescribed with CBD-rich treatment at a network of clinics dedicated to medical cannabis in Quebec, Canada. Data were collected at baseline, 3 (FUP1), and 6 (FUP2) month after treatment initiation. Groups were formed based on symptom severity (mild vs moderate/severe) and based on changes to treatment plan at FUP1 (CBD vs THC:CBD). Two-way mixed ANOVAs were used to assess ESAS-r scores differences between groups and between visits.
This retrospective observational study suggests CBD-rich treatment has a beneficial impact on pain, anxiety, and depression symptoms as well as overall wellbeing only for patients with moderate to severe symptoms; however, no observed effect on mild symptoms. The results of this study contribute to address the myths and misinformation about CBD treatment and demand further investigation.
Cannabidiol (CBD) is one of the primary cannabinoids found in significant but variable concentrations in cannabinoid-based medicines (CBM). While structurally similar to Δ9-tetrahydrocannabinol (THC), CBD does not cause intoxication or euphoria (Russo 2017) and has showed considerable tolerability in humans with a low abuse potential (Chesney et al. 2020). This favorable safety profile has led to the recent mitigation of legal and regulatory barriers surrounding purified CBD products in several countries and recent increased interest in CBD treatments. While recent rulings clarified that CBD is not a drug under the 1961 United Nations as Single Convention on Narcotic Drugs, regulatory status in the USA remains extremely confusing. When derived from cannabis, CBD is a schedule 1 drug but when derived from “industrial hemp” plants it may be lawful federally (Corroon and Kight 2018; Corroon et al. 2020). In Canada, CBD is controlled under the Cannabis Act as are all cannabinoids, cannabis, and cannabis-derived products (Canada Go 2021). This regulatory status imparts restrictions and access obstacles for researchers.
CBD is widely touted as a panacea for a wide range of health problems and has been marketed as a dietary and “wellness” product (Russo 2017; Khalsa et al. 2020; Eisenstein 2019). CBD’s potential effects as an add-on therapy have been studied for social anxiety disorders, schizophrenia, non-motor symptoms in Parkinson’s disease, and substance use disorders (Bergamaschi et al. 2011; Crippa et al. 2019; McGuire et al. 2018; Millar et al. 2019; Prud’homme et al. 2015; Thiele et al. 2019; Leehey et al. 2020). However, the evidence of its effectiveness for indications other than drug-resistant pediatric epilepsy conditions remains very limited (Larsen and Shahinas 2020; Franco et al. 2020) and safety considerations such as drug-drug interactions associated with unsupervised use remain (Chesney et al. 2020; Freeman et al. 2019). Randomized controlled trials (RCTs) are limited in their rigorous design, population sample, and duration of observation making generalization of results and long-term data scarce. Therefore, real-world evidence (RWE) provides valuable insights and supplemental information about the use, safety, and effectiveness of CBD-based treatments (Graham et al. 2020).
RWE from retrospective analyses and patient registries shows that CBMs are used for pain (chronic, neuropathic), mental health conditions, cancer-related symptoms (nausea, fatigue, weakness), HIV/AIDS, and neurological conditions (Bonn-Miller et al. 2014; Gulbransen et al. 2020; Lintzeris et al. 2020; Lucas and Walsh 2017; Sexton et al. 2016; Waissengrin et al. 2015). Symptom control is the primary reason for use of CBM, with most patients looking to address unalleviated symptoms, perceived symptom intensity, and burden on health-related quality of life independently of primary diagnosis (Sexton et al. 2016; Waissengrin et al. 2015; Baron et al. 2018; Purcell et al. 2019; Swift et al. 2005; Webb and Webb 2014). The Edmonton Symptom Assessment Scale-revised version (ESAS-r) is a validated scale to assess symptom burden developed for use in oncology and palliative care (Hui and Bruera 2017), it has relevance to medical cannabis care as patients are often treated for similar symptom management (Good et al. 2019; Pawasarat et al. 2020). Specifically, studies showed self-perceived improvement in ESAS-r emotional symptoms (anxiety and depression) scores following CBM treatment in oncology patients, while pain and wellbeing symptoms showed no improvement (Good et al. 2019; Pawasarat et al. 2020). Yet, RWE on CBD-rich products is scarce (Goodman et al. 2020; Shannon et al. 2019). In addition, although careful titration and treatment adjustment after initiation is critical to symptom improvement and adverse effects care, current literature has failed to address this issue.
In this study, we investigated treatment with CBD-rich products within a dedicated clinical setting in Quebec, Canada, and the effects on a very common clinical symptom expression of pain and comorbid anxiety and depression symptoms, as well as the effect on overall wellbeing. We also examined the relevant clinical effects that were observed when CBD-rich treatments were replaced by THC:CBD-balanced products at subsequent follow-up visits.
This study is a retrospective examination of patients who were prescribed CBD-rich products by physicians at a clinic dedicated to CBM treatments operating at four locations across Quebec, Canada. All data are collected as part of standard clinical procedures during the initial visit and during 3 (FUP1) and 6 (FUP2) month follow-up visits and extracted from electronic medical records (EMR) (Prosk et al., 2021). All data were anonymized following extraction from the EMR and no identifiers linking to original data were maintained. A waiver of consent was required and approved by Advarra Ethics Committee, who also approved the study protocol, and by the provincial privacy commission (La commission d’accès à l’information du Quebec).
Adult patients, at least 18 years of age, who were initially treated exclusively with CBD-rich products from 1 October 2017 to 31 May 2019 and for whom outcome scores and product information were recorded at FUP1 were included in this study. Patients were generally referred by primary-care physicians and specialists for an assessment on the suitability of medical cannabis to treat refractory symptoms. A complete medical history, including primary and secondary diagnoses, was collected at baseline visit. Medical cannabis treatment decisions are determined at the discretion of a clinic physician according to a standardized clinical procedure, including symptom identification, selection of product format, cannabinoid profile, and dosage based on existing evidence (MacCallum and Russo 2018; Cyr et al. 2018), but also to minimize risk of adverse effects. Patient and physician preference may also indicate initiation with products that have higher CBD and lower THC concentration in order to limit use of THC and its inherent potential adverse events. The follow-up visits serve to assess treatment compliance, safety, and effectiveness.
CBD-rich products in Canada
CBD-rich products are administered in various methods and formats, but most commonly as oral plant-derived extracts or oils and as inhaled dried flowers. In the Canadian medical cannabis program, CBD-rich cannabis oils contain approximately 0.5–1 mg of THC/mL and 20–25 mg of CBD/mL depending on the product manufacturer. Table Table1 1 provides cannabinoid content and THC:CBD ratio for the three most common oil products (over 85% of patients) authorized at the clinic. Furthermore, product details in this study sample are described in Table Table3. 3 . The clinic procedure dictates that all products with a ratio of CBD (mg) to THC (mg) higher than 10 are considered CBD-rich products.
THC and CBD contents and associated THC:CBD ratio for the three most common oil products authorized at the clinic
|CBD-rich products at baseline||THC:CBD-balanced products at FUP1||THC-rich products at FUP1|
|Authorized dose range (in ml/intake)||0.1–2||0.05–3||0.2–1.5|
|Oil (mg/ml)||THC||CBD||Ratio THC:CBD||THC||CBD||Ratio THC:CBD||THC||CBD||Ratio THC:CBD|
|Product 1||1.2||24||1:25||9.5||12||10:13||27.5||< 1||30:1|
|Product 3||< 1||20||1:20||10||13.5||10:13||26.3||< 1||30:1|
The data is categorized by product category: CBD-rich products, THC:CBD-balanced products, and THC-rich products
CBD cannabidiol, THC Δ9-tetrahydrocannabinol, SD standard deviation
Details of the THC and CBD component of the CBD-rich, the THC:CBD 1:1, and the THC-rich formulations
|CBD-rich products at baseline (n = 279)||THC:CBD-balanced products at FUP1 (n = 104)||THC-rich products at FUP1 (n = 12)|
|Oil products (in mg/ml)||0.1–2.0||2.0–52.0||0.6–30||2.5–39||1.25–45||0–18|
|Dried flower (in % w/w)||0.7||17.0||3.7–9||7.7–13.4||13–27||0–0.5|
|Average daily dose (mg)||0.5||11.47||19.65||26.32||54.28||10.80|
|Standard deviation (mg)||0.43||10.21||5.80||9.12||29.65||7.64|
|Maximum daily dose (mg)||6||156||60||78||90||54|
Data comes from our sample of 279 patients
CBD cannabidiol, THC Δ9-tetrahydrocannabinol
Treatment adjustments occur at follow-up visits as a result of lack of effectiveness, presentation of adverse effects, or social or economic barriers. Adjustments may include a change of the recommended CBD-rich product, method of administration, dosage, or a change in product formulation such as the introduction of THC:CBD-balanced or THC-rich products. We investigated the change from CBD-rich to THC:CBD products during FUP1 by forming two groups based on their product adjustment at FUP1 (CBD-rich vs THC:CBD). Products at FUP1 reflect those recommended at the visit. Therefore, the adjusted treatment affects only the evaluation at FUP2.
Patients age, sex, and diagnosis were recorded at baseline. Patients completed the ESAS-r (Edmonton Symptom Assessment System-revised version) at each visit. The ESAS-r is a self-administered scale, rating the severity of symptoms from 0 (absence of symptom) to 10 (worst possible severity) at the time of assessment (Hui and Bruera 2017). Symptoms evaluated include six physical- (pain, tiredness, nausea, drowsiness, lack of appetite, and shortness of breath), two emotional- (depression, anxiety), and one overall wellbeing-related symptoms. ESAS scores can be categorized as mild (score 0 to 3) moderate (score 4 to 6) or high (score 7 and above) (Butt et al. 2008) and the threshold for clinically significant improvement is a decrease of 1 point (Hui et al. 2015). Since pain and mental health issues represent the most common symptoms for patients and physicians seeking medical cannabis treatments, we investigated effects on pain, depression, and anxiety symptoms as well as overall wellbeing. For each symptom, two groups of patients were formed: moderate-severe severity group in which a baseline score of 4 or more was recorded and a mild severity group with baseline score of 0 to 3.
Mean scores and standard deviation (SD), as well as percentage, where appropriate are presented for each variable. All analyses were performed on each ESAS-r symptom separately through the data analytics software R v4.0.2. An initial analysis compared the overall ESAS-r scores between each visit no matter the severity of the group, and looked at the role of product group (CBD/THC:CBD vs CBD/CBD group) (between-factor). Tukey HSD post hoc test was used to confirm where the differences occurred between groups.
To determine whether CBD-based treatments have different effectiveness based on the severity of patient symptoms, two-way mixed ANOVAs with severity group as between-factor and visit as a within-factor were conducted to assess the change in ESAS-r scores between visits. Paired t-tests were subsequently performed to assess the difference in mean scores within each severity group between baseline and FUP1. Significant p value was set at 0.05 and all analyses were two-tailed. Partial eta-squared (η 2 p) are reported to indicate magnitude of differences between groups.
A total of 1095 patients were seen at the four clinic sites during the study period. Out of those, 715 were eligible for the study (at least 18 years old and initially treated exclusively with CBD-rich products). A total of 279 patients with ESAS-r scores and product information at FUP1 were analyzed (190 (68%) female, mean age = 61.1, SD = 16.6). The analyzed sample did not differ from the study-eligible group in terms of age, sex, or THC and CBD initial doses (all ps > 0.4). Table Table2 2 outlines patient sample size and demographic information for each symptom and treatment group. Two hundred and ten (75%) patients were prescribed CBD-rich products to treat chronic pain, 19 (7%) for cancer-related symptoms, 21 (7.5%) to treat neurological disorders (Parkinson’s disease, multiple sclerosis, and drug-resistant epilepsy among others), 8 patients for inflammatory disease (arthritis), 10 for gastrointestinal disorders (Chron’s disease, inflammatory bowel syndrome, ulcerative colitis), 2 for anxiety, 1 for depression, 2 for headaches, and 6 unclassified. The chronic pain category included all medical indications for which pain was the main symptom such as but not limited to fibromyalgia, spinal stenosis, and chronic low back pain. Overall, 116 (41.6%) patients adjusted their prescription by adding THC at FUP1 (either to a THC:CBD-balanced combination or a THC-rich treatment). Two hundred and three (73%) patients had moderate/severe ESAS-r scores on at least 2 of the examined symptoms, 57 (20%) on three, and 75 (27%) on all four symptoms. Twenty-nine (10%) patients report no moderate/severe symptoms; these people may use CBD for other ESAS-r symptoms not examined here (shortness of breath, tiredness, nausea, drowsiness, appetite). There was no statistical difference on either age, sex, or THC and CBD initial doses between the patients who completed one FUP versus those who completed two FUP (all ps > 0.1).
Demographic characteristics of 279 medical cannabis patients, by symptom group
|Sample size (percentage)||Number of female patients (percentage)||Mean age (SD)|
|Overall sample||279||190 (68)||61.1 (16.6)|
|Moderate or severe pain symptom group||205 (73.5)||150 (73)||61.8 (15.9)|
|Moderate or severe anxiety symptom group||138 (48.5)||97 (70)||61.43 (16.3)|
|Moderate or severe depression symptom group||115 (41.2)||81 (70)||60.5 (15)|
|Moderate or severe wellbeing group||202 (72.4)||141 (70)||60.8 (16.1)|
|CBD/THC:CBD group||116 (41.6)||75 (65)||60.38 (14.4)|
The symptom groups are mild and moderate or severe. The table presents the moderate or severe demographic characteristics. The CBD/THC:CBD group is composed of patients who added THC to their CBD-rich prescription during FUP1
CBD cannabidiol, THC Δ9-tetrahydrocannabinol, SD standard deviation
CBD-rich products characteristics
The baseline average daily doses for CBD and THC are presented in Table Table3. 3 . The maximum initial CBD dose recorded (156 mg) was prescribed for the treatment of pain of one patient. The maximum THC dose recorded at FUP1 (90 mg) was prescribed for two patients for the treatment of pain.
Outcome of CBD treatment
Mean ESAS-r scores of pain, anxiety, depression symptoms, and overall wellbeing at baseline, FUP1, and FUP2 are described in Table Table4 4 and Fig. Fig.1 1 .
Mean and standard deviation (SD) scores of ESAS-r scales for each severity group (mild or moderate/severe) and for each product group (CBD/CBD or CBD/THC:CBD)
|Baseline (sample size)||277||270||272||268|
|Overall sample||5.14 (2.57)||3.86 (3.19)||3.16 (3.08)||5.34 (2.61)|
|Mild severity group||1.69 (1.1)||0.99 (1.15)||0.87 (1.18)||1.86 (1.18)|
|Moderate or severe severity group||6.34 (1.7)||6.61 (1.78)||6.3 (1.86)||6.47 (1.83)|
|CBD/CBD group||5.03 (2.66)||3.80 (3.21)||2.99 (3.04)||5.28 (2.72)|
|CBD/THC:CBD group||5.28 (2.45)||3.95 (3.17)||3.40 (3.13)||5.42 (2.46)|
|FUP1 (sample size)||262||261||261||254|
|Overall Sample||4.37 (2.73)||2.93 (2.95)||2.33 (2.79)||4.45 (2.6)|
|Mild severity group||2.3 (2.4)||1.62 (2.08)||1.12 (1.78)||3.73 (2.75)|
|Moderate or severe severity group||5.04 (2.49)||4.15 (3.09)||3.77 (3.07)||4.72 (2.5)|
|CBD/CBD group||4.09 (2.67)||2.74 (2.87)||2.23 (2.71)||4.43 (2.6)|
|CBD/THC:CBD group||4.75 (2.78)||3.2 (3.05)||2.47 (2.9)||4.49 (2.63)|
|FUP2 (sample size)||101||99||102||97|
|Overall Sample||4.7 (2.7)||2.85 (3.01)||2.67 (3.02)||4.57 (2.47)|
|Mild severity group||2.18 (2.43)||1.32 (1.89)||1.52 (2.31)||3.82 (2.81)|
|Moderate or severe severity group||5.2 (2.47)||3.96 (3.19)||3.74 (3.26)||4.93 (2.23)|
|CBD/CBD group||4.55 (2.6)||2.44 (2.68)||2.44 (2.82)||4.76 (2.22)|
|CBD/THC:CBD group||4.88 (2.81)||3.08 (3.07)||2.94 (3.24)||4.36 (2.73)|
The CBD/THC:CBD group is composed of patients who added THC to their CBD-rich prescription during FUP1. ESAS-r scores varied between 0 and 10 for all assessed symptoms and all visits except for the anxiety scale at FUP2 for which the maximum score was 9
CBD cannabidiol, FUP1 follow-up visit at 3 month, FUP2 follow-up visit at 6 month, THC Δ9-tetrahydrocannabinol
CBD-rich treatment effectiveness on pain, anxiety, depression symptoms, and on overall wellbeing in 279 patients. FUP1, follow-up visit at 3 month; FUP2, follow-up visit at 6 month. Mixed ANOVAs revealed a significant effect of visit on symptom reduction between baseline and FUP1 but not between FUP1 and FUP2
All average ESAS-r scores decreased between baseline and FUP1 and FUP2. This was further demonstrated by ANOVAs which revealed a significant effect of visit on mean ESAS-r scores for each symptom assessed (pain: F(2,634) = 4.9, p < 0.008; anxiety: F(2,624) = 8.36, p < 0.001, depression: F(2,629) = 5.36, p < 0.004; wellbeing: F(2,613) = 8.31, p < 0.001; all η 2 p between 0.008 and 0.02). In all assessed symptoms, no significant main effect of adding THC at FUP1, nor visit-by-product interaction, were observed (all ps > 0.2). Post hoc tests revealed ESAS-r mean scores significantly decreased between baseline and FUP1 (all ps < 0.003) for all symptoms, between baseline and FUP2 for anxiety and wellbeing (both ps < 0.03), but not between FUP1 and FUP2 for any symptoms (all ps >0.5). This suggests statistical improvement recorded at FUP1 is still present at FUP2 in all symptoms independently from treatment adjustment at FUP1.
CBD treatment impact according to symptom severity
From Table Table2, 2 , moderate or severe scores at baseline were most common for pain (205 patients, 73.5%) and poor wellbeing (202 patients, 72.4%).
Clinical effect (difference of 1.3 to 2.5 points) observed in all symptoms for patients with moderate/severe symptoms between baseline and FUP1; however, there was no clinical effect for patients with mild symptoms (from − 0.3 to − 1.8) (Fig. (Fig.2). 2 ). No clinical effect was observed in any symptoms between FUP1 and FUP2 for patients with moderate/severe symptoms (− 0.4 to 0.5) as well as for patients with mild symptoms (from − 0.7 to 0.4).
CBD-rich treatment effect according to symptom severity: mild or moderate/severe in 279 patients. FUP1, follow-up visit at 3 month; FUP2, follow-up visit at 6 month. a Mean ESAS-r scores for the pain symptom, b mean ESAS-r scores for the anxiety symptom, c mean ESAS-r scores for the depression symptom, and d mean ESAS-r scores for overall wellbeing. According to mixed ANOVAs, patients with moderate/severe symptoms reported symptom reduction whereas patients with mild symptoms reported symptom deterioration from baseline to FUP1. No effect was statistically significant between FUP1 and FUP2
The ANOVA revealed that all main and interaction effects were significant at the 0.001 level with effect sizes large for severity (η 2 p = 0.29), medium for visit (η 2 p = 0.06), and small for the interaction (η 2 p = 0.03). Post hoc tests revealed a significant score difference between baseline and FUP1 and FUP2 (both ps < 0.05) but not between FUP1 and FUP2 (p = 0.98). Patients with moderate/severe symptoms on pain experienced important improvement at FUP1 (t(194) = 7.61, p < 0.001) whereas ESAS-r scores for patients with mild symptoms actually increased (t(64) = − 2.03, p < 0.05) (Fig. (Fig.2 2 a).
There were significant effects of visit, severity group, and visit by group interaction (all ps < 0.001; η 2 p = 0.006, η 2 p = 0.4, η 2 p = 0.1, respectively). Post hoc tests revealed a significant score difference between baseline and FUP1 and FUP2 (both ps < 0.001) but not between FUP1 and FUP2 (p = 0.38). Although there was a large improvement for patients with moderate to severe anxiety symptoms (t(131) = 9.36, p < 0.001), the anxiety scores of patients with mild symptoms increased (t(119) = − 3.19, p < 0.01) from baseline to FUP1 (Fig. (Fig.2 2 b).
The ANOVA showed main effects of visit, severity group (both ps < 0.001 with η 2 p = 0.04 and η 2 p = 0.4, respectively) and a significant group-by-visit interaction (F(2,620) = 34.47, p < 0.001; η 2 p = 0.1). Post hoc tests revealed a significant score difference between baseline and FUP1 and FUP2 (both ps < 0.01) but not between FUP1 and FUP2 (p = 0.85). Specifically, the scores of moderate/severe group decreased notably (t(110) = 9.56, p < 0.001) between baseline and FUP1 but the scores of the group with mild depression symptoms did not (p = 0.07) (Fig. (Fig.2 2 c).
The ANOVA showed main effects of visit, severity group (both ps < 0.001 with η 2 p =0.04 and η 2 p =0.3 respectively) and a significant group-by-visit interaction (F(2,597) = 36.53, p < 0.001; η 2 p = 0.11). Post hoc tests revealed a significant main score difference between baseline and FUP1 and FUP2 (both ps < 0.01) but not between FUP1 and FUP2 (p = 0.89). Precisely, the scores of the group reporting good wellbeing increased (t(182) = 8.8, p < 0.001) whereas scores of patients with worst wellbeing notably decreased (t(59) = − 5.08, p < 0.001) between FUP1 and FUP2 (Fig. (Fig.2 2 d).
This retrospective study explored the use of CBD-rich products in a medical cannabis clinical setting in Canada and associated effectiveness on a common symptom cluster presentation of pain, anxiety, depression, and poor sense of wellbeing, as measured by ESAS-r.
Patients treated with CBD-rich products were mainly women in their sixties, seeking predominantly chronic pain management.
Our findings show that overall effectiveness of CBD treatment is primarily by patients with moderate to severe symptoms. A deficiency in the endocannabinoid system (ECS) may provide a possible explanation for this result (Russo 2016). The ECS could be more deficient in patients with moderate/severe symptoms compared to mild symptoms leading to increased improvement in the first group. The absence of significant improvement for patients with mild symptoms at baseline may be explained by a smaller margin for symptom improvement. In such patients, CBD treatments may have been targeted to other clinical symptoms not assessed in the current study. There is a probable placebo effect; however, there were no differences in initial CBD doses between the severity groups. Furthermore, associated placebo effect would likely be decreased by FUP3M, also considering the significant treatment cost. The distinct beneficial impact of CBD treatment observed for patients with moderate-severe symptoms could elucidate discrepancies found in the literature.
RCTs on CBM and pain symptoms provide inconclusive results; however, several report that treatments of THC and CBD have some benefit for pain management (Häuser et al. 2018; Russo 2008; Prosk et al. 2020). Our results are largely novel as research on the effect of CBD on pain control is very limited (Boyaji et al. 2020). The reduction in reported anxiety may also contribute to the improvement in pain perception.
Discrepancies still exist regarding the anxiolytic effect of CBD. Some RCTs indicate an anxiolytic effect of CBD upon experimentally induced scenarios (Bergamaschi et al. 2011; Zuardi et al. 2017; Bhattacharyya et al. 2010; Skelley et al. 2020); however, these findings are difficult to replicate (Larsen and Shahinas 2020; Hundal et al. 2018; Crippa et al. 2012). This reinforces our findings that CBD may have a differential effect depending on anxiety severity. Regarding the effects of CBD on depression symptoms, further research is required to draw conclusions (Khalsa et al. 2020; Schier et al. 2014; Turna et al. 2017).
The addition of THC to CBD during FUP1 did not produce any effect on ESAS-r scores at FUP2 in this analysis; however, the magnitude of the difference between groups is small. The examination of treatment regimen has been seldom addressed in the literature and further development is required to inform guidelines for prescription and refinement of clinical practice.
Furthermore, a significant discrepancy is observed between the recorded dosages of oral CBD in RCTs and dosages in real-world settings. The average daily CBD dosage authorized at our clinic (11.5 mg) is closer to other observational studies (Gulbransen et al. 2020) compared to what is seen in RCTs (up to 1000 mg for a single dose) (Larsen and Shahinas 2020). The presence of THC and other cannabinoids in CBD-rich products may affect the outcomes in this study. The majority of RCTs investigated single-dose administration of CBD making it difficult to compare observed treatment outcomes with chronic dosing clinical settings. Importantly, medical cannabis products are generally not covered by most insurers and patients rely on out-of-pocket payments. The cost of CBD remains very high globally, approximately $CAD 5–20 per 100 mg (Canada Go 2021; Eisenstein 2019; Canada 2020). Availability of reliable cannabinoid testing in certain international jurisdictions is also limited. The gap between effective doses demonstrated in RCTs and the actual affordable doses demonstrated by RWE mandate the need for a precise pricing and marketing strategy at the initiation of any drug development process.
Limitations are common in real-world data (RWD), especially in retrospective studies. In this study, with no control group, no causality effect can be drawn between CBD-rich treatment and symptom improvement. Most patients treated with CBM present with multiple severe symptoms and the analyses presented here are limited to identify the treatment outcomes for such patients. Further studies can investigate the use of CBD to treat several symptoms simultaneously.
The self-reported subjective assessment used may be biased by the patient’s positive expectation of treatment, which could lead to a possible placebo effect. This perceived effectiveness bias may also be increased by social and economic barriers. The current context of medical cannabis access, including social stigma, high cost, and lack of universal insurance coverage can increase the patient selection bias. Self-selection bias is increased by the significant patient interest in medical cannabis as these patients must be motivated to access the non-traditional medication system. This bias limits the generalizability of results but is common across international medical cannabis regimens and should not discount the observed results. The heterogeneity of the patient population with a variety of diagnoses and the diversity of medical cannabis preparations also affects the external validity of the study. However, clinical findings from within Canada’s controlled regulatory program do provide important models for international consideration. Future research is required in controlled clinical settings to examine these factors in order to provide a more complete account of CBD effectiveness.
Also, there was a large drop of sample size (53% loss) due to missing data. Additionally, there was an important loss to follow-up at the 6-month visit (FUP2) due to missed appointment and cost barriers, limiting the power of the findings. The total treatment cost has significant impact on treatment continuation. Improved patient retention and more robust, harmonized data collection methods will improve future observational studies and allow for long-term assessment. Collection of detailed, accurate product information is a challenge, especially with inhaled products (Corroon et al. 2020). There are opportunities for administration devices and other technology advancements to improve this limitation. Lastly, this study did not include safety data assessment, future studies should investigate safety considerations of CBD (Chesney et al. 2020). Collection of high-quality RWD will require improvements in patient retention, data monitoring, and more robust data collection methods within a controlled clinical setting.
This study on CBD-rich products demonstrates the potential of RWE for the advancement of medical cannabis research and practice guidelines, especially in a world where CBD use is exponentially increasing but scientific data are limited. It revealed that CBD-rich treatments have a beneficial impact on patients with self-reported moderate or severe symptoms of pain, anxiety, or depression and overall wellbeing but not in patients with mild symptoms. Further investigation is clearly required, but as of now the hyped, and often illegal, marketed claims of CBD as a wellness product are unsubstantiated. Our findings have important and novel implications to clinical practice, especially the examination of treatment plan adjustment during the first follow-up after initiation with CBD treatments. Improvements in access regimes, oversight, and clarification from regulatory agencies are also needed to improve the validity of RWE and assessment of the use of CBD-rich products.
The authors would like to thank the participants to this study. The authors would like to acknowledge Santé Cannabis co-founder Dr. Michael Dworkind and key clinical leaders Dr. Antonio Vigano, Dr. Howard Mitnick, Dr. Alain Watier, and Youri Drozd, clinical data assistant, for his contribution to data technical help.
|EMR||Electronic medical record|
|ESAS-r||Edmonton Symptom Assessment System-revised version|
All authors contributed to conception and design, interpretation of data, manuscript writing, and final approval. LR and RG conducted the analysis of data. All authors agreed to be accountable for their own contributions. The author(s) read and approved the final manuscript.
This research was funded internally by Santé Cannabis clinic.
Availability of data and materials
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
A waiver of consent was required and approved by Advarra Ethics Committee, who also approved the study protocol, and by the provincial privacy commission (La commission d’accès à l’information du Quebec).
L.R.: Clinical research associate, employee at Santé Cannabis.
R.G.: Epidemiologic and statistic consultant for Santé Cannabis.
C.EH.: Director of Research and Innovation, employee at Santé Cannabis.
MF.A.: Associate Research Director of Santé Cannabis.
E.P.: President and co-founder of Santé Cannabis.
Santé Cannabis is a medical clinic, research, and training center dedicated to medical cannabis. The views expressed are those of the authors. This is a retrospective, observational study which took place at Santé Cannabis; therefore, the design and conduct of the study was executed by Santé Cannabis clinic staff. C.EH. and E.P. had a supporting role, in the retrospective protocol development. The authors had no role in the conduct of the study and collection of data. None of the authors are involved in the care of patients or in treatment decisions. The authors acted independently, and Santé Cannabis had no role in the analysis of the study, nor the writing of the manuscript or decision to publish. There is no financial gain for Santé Cannabis or for the authors to publish. The authors, while connected to Santé Cannabis, do not have a financial or professional incentive for the decision to publish.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
The reality behind cannabidiol’s medical hype
The cannabis compound known as CBD is being touted as a treatment for a variety of conditions. But the substance’s uncertain legal status is stalling serious investigation.
- Michael Eisenstein0
Michael Eisenstein is a freelance writer in Philadelphia, Pennsylvania.
You can also search for this author in PubMed Google Scholar
Cannabidiol oil has purported health benefits, including helping to relieve chronic pain. Credit: Don Bartletti/Los Angeles Times/Getty
Cannabidiol (CBD) is an illegal drug with no redeeming value. It is also a useful prescription medicine for epilepsy, with considerable potential for treating numerous other conditions. And it is a natural dietary supplement or ‘nutraceutical’ with countless evangelists in the health and wellness community. Although contradictory, all three statements are true from different perspectives, and clinical researchers are frustrated.
“In New York City, you can go to a latte shop and get a CBD product, but if I want to do a clinical trial, I’ve got to get a 2,000-pound safe and go through six months of paperwork and licensing,” says Orrin Devinsky, director of the NYU Langone Comprehensive Epilepsy Center in New York City. Like the cannabis plant from which it is derived, CBD, a type of cannabinoid, is classified by the US Drug Enforcement Administration in the same way as are heroin and lysergic acid diethylamide (LSD) — schedule 1 substances with “high potential for abuse” and “no currently accepted medical use”.
Part of Nature Outlook: Cannabis
This flies in the face of current evidence. Numerous studies have shown that CBD is a safe and non-habit-forming substance that does not produce the ‘high’ associated with tetrahydrocannabinol (THC), the main psychoactive component of cannabis 1 . In 2018, the US Food and Drug Administration (FDA) determined that Epidiolex — a purified CBD product developed by GW Pharmaceuticals in Histon, UK — effectively reduces the frequency of seizures in certain rare forms of paediatric epilepsy. This approval has heartened the cannabinoid research community, which has long recognized the medicinal potential of CBD but come up against scepticism and regulatory constraints on the road to the clinic.
But at the same time, the many manufacturers that promote CBD-laden oils, lotions and foods as a panacea for various health issues, often with minimal regard for local laws or medical evidence, are putting CBD’s medical advocates in an uncomfortable position. “I get calls and e-mails all the time — not just from families, but from physicians who have no clue how to address the requests they get from patients,” says Yasmin Hurd, director of the Addiction Institute of Mount Sinai in New York City. “It’s a real problem.”
Stuck in the weeds
The breakthrough approval of Epidiolex was driven by strong investment from GW Pharmaceuticals, as well as vigorous advocacy from families of children with epilepsy who had heard tantalizing anecdotes about CBD’s effects from jurisdictions in which medical cannabis is legal. “About eight years ago, a patient’s father said he was hearing stories about families in Colorado and California who use high-CBD strains for their kids’ epilepsy,” says Devinsky. “He asked me to do a trial.” As a medical student, he had been taught the history of medicinal cannabis, including well-documented uses of the plant by nineteenth-century physicians to treat seizures. Indeed, cannabis has been part of the clinical armamentarium for epilepsy for more than 4,000 years.
Research on CBD in the 1970s and 1980s focused on its interplay with other cannabinoids, and particularly THC. “Whereas THC can induce psychotic symptoms, impair cognition and make people anxious, CBD appears to do the opposite,” says Philip McGuire, a psychiatrist at King’s College London.
The first clues that CBD might suppress epileptic episodes came from a small clinical trial 2 in 1980. It was led by Raphael Mechoulam, a chemist at the Hebrew University of Jerusalem, whose work on the synthesis and biochemical characterization of cannabinoids in the 1970s had led researchers to begin to explore the medicinal properties of CBD. A number of other trials that explored the compound’s pharmaceutical properties followed, although scientists conducting early forays into CBD clinical research faced an uphill battle. F. Markus Leweke, a psychiatrist who specializes in mental illness at Sydney Medical School, Australia, recalls struggling for seven years to publish findings from a randomized controlled trial that demonstrated that CBD might offer an effective treatment for psychotic symptoms in schizophrenia 3 . “We got about 15 rejection letters,” says Leweke. “And this is a paper that has since been cited almost 500 times.”
Claims about the health benefits of cannabis are often overstated and lack supporting evidence. Credit: Rodger Bosch/AFP/Getty
Forty years on from Mechoulam’s initial work, extensive randomized controlled trials have decisively shown that this purified cannabinoid can profoundly benefit children with certain epileptic disorders. “Over those trials, we saw about a 26–28% reduction in frequency over placebo in all convulsive seizures for Dravet syndrome and drop seizures for Lennox–Gastaut syndrome,” says Devinsky, who has led several such studies 4 , 5 . “Some of the patients became, and remain, seizure-free.”
Preclinical data from rodent and cell-culture studies have hinted at the possible benefits of using CBD to help treat disorders that range from Parkinson’s disease to chronic pain. The range of conditions in which CBD is being tested might seem diverse, but it is a compound with far-reaching, if poorly understood, physiological effects. Antonio Zuardi, a psychiatrist at the University of São Paulo in Brazil, notes that something on the order of 20 possible mechanisms of action have been described to date for CBD. “These multiple pharmacological effects may justify the wide range of possible therapeutic activities.”
The mechanism of CBD’s action on cannabinoid receptors, at least, is well understood. CBD can bind to the cannabinoid receptor CB1, which is the same receptor that THC seeks out in the brain. Unlike THC, however, CBD restrains rather than activates CB1 signalling, and therefore doesn’t induce the psychoactive effects of its cannabinoid cousin.
But CBD wears many hats. It seems to mediate its antiepileptic effects by binding to a protein called GPR55, which can otherwise trigger the onset of seizures by promoting the hyperactivation of neurons 6 . In addition, CBD acts on receptors that mediate pain signalling and inflammation, as well as at least one receptor for the neurotransmitter serotonin, 5-HT1A 7 . Gabriella Gobbi, a psychiatrist and neuroscientist at McGill University in Montreal, Canada, has found that CBD’s physiological effect on the brain resembles that of selective serotonin reuptake inhibitor (SSRI) drugs 8 , which are used to treat clinical depression. “After a few days, you get this desensitization of 5-HT1A, like you would with an SSRI, and increased serotonin signalling,” she says. Further experiments in rats failed to capture an antidepressant effect, but her team found that CBD-mediated modulation of 5-HT1A could relieve neuropathic pain in the animals.
Beyond epilepsy, clinical data to support the medicinal benefits of CBD are more limited, mainly due to the small scale and inconsistent design of trials. “We have very few double-blind, randomized placebo-controlled trials,” says Gobbi. But exciting progress is being made towards treating several conditions.
Psychosis — particularly in the context of schizophrenia — is one such area of promise. In 1995, Zuardi and Mechoulam reported the case of a person with schizophrenia who experienced meaningful relief from their symptoms when treated with high doses of CBD 9 . Several subsequent small-scale clinical studies detected similar hints of efficacy. In their groundbreaking trial 3 , Leweke and his colleagues put the compound through a particularly rigorous test by comparing its effects with those of amisulpride, a potent medication for schizophrenia. “We saw a significant decrease in symptoms over time for both compounds, and CBD beat amisulpride in terms of side effects, by far,” Leweke says. The team also found a clue to the mechanism by which CBD might exert its antipsychotic effects: treatment with CBD was associated with elevated levels of anandamide, a cannabinoid produced by the body that seems to offer protection from psychosis.
McGuire and his colleagues conducted a randomized controlled trial that showed that CBD can have an additive effect when used with conventional antipsychotic drugs 10 . Together, they were better able to control symptoms such as hallucinations and delusions than could conventional medication alone. His team has received funding for a large, international trial to test whether CBD can be developed as a licensed medicine for treating psychosis.
Anxiety disorders are another mental-health condition that CBD has been shown to help alleviate. Zuardi and his colleagues used a test that simulates speaking in public to show that pretreatment with a single dose of CBD can reduce the associated discomfort in people with social anxiety disorder 11 . A similar effect has been observed in healthy people in anxiety-inducing situations 12 , and several researchers are exploring CBD as a means of soothing social stress in people with autism spectrum disorder. Devinsky notes that many of his patients with epilepsy have also been diagnosed with autism spectrum disorder, and he is involved in two clinical trials that aim to test whether CBD can meaningfully reduce the irritability and anxiety of those with autism. “Many parents wanted to keep their children on it even if the seizures didn’t improve, because they’re calmer and sleeping better,” he says.
And although cannabis been demonized as a gateway to more dangerous substances, Hurd has found that it might actually contain an effective antidote for potentially deadly addictions. After observing that rats with a heroin addiction were less likely to seek out the opioid when treated with CBD, she began to investigate whether CBD might have the same effect on people with an opioid dependency. On the basis of an encouraging pilot study, Hurd and her team conducted a randomized controlled trial in 42 abstinent heroin users, who had avoided taking the drug for up to three months after years of routine or heavy use 13 . The researchers then exposed the participants to drug paraphernalia and videos that showed heroin use — cues that normally provoke strong cravings in people with a dependency — and then measured participant-reported responses and physiological indicators of stress and anxiety. “Cue-induced craving is associated with increased cortisol levels and increased heart-rate, and CBD reduced those,” she says. Participants receiving CBD also reported lower levels of drug craving and anxiety relative to placebo group, and Hurd notes that the beneficial effects persisted for a week after the final administration of CBD.
A difficult delivery
Despite its promise, CBD’s impact as a drug has been mixed. Importantly, it is relatively safe. The side effects most commonly associated with a high dose of Epidiolex include digestive problems, rash and drowsiness, as well as the potential for liver damage in patients taking certain other medications. For example, Devinsky notes that patients who are receiving valproic acid to treat seizures or migraines might be at an elevated risk. But in many of the CBD trials conducted so far — particularly in the realm of antipsychotic drugs, which are known for their strong side effects — CBD has proved more tolerable than existing alternatives. “The side effects weren’t significantly worse than with placebo,” says McGuire of his 2018 study of CBD in people with schizophrenia 10 .
This is important because people typically require large doses of the drug to experience a clinical benefit — in many studies, the doses used are as high as 1 gram or more. This is because CBD is poorly absorbed by the body, with most of every dose being excreted before it can take effect. “If you take it orally, the bioavailability is in the range of 4–6%, which is terrible,” says Devinsky. “If you take it after a fatty meal, you can get that up to 16–20%.” Zuardi notes that his group routinely observes a bell-shaped dose–response curve for CBD. For example, whereas 300 milligrams of CBD might reduce a person’s anxiety, the same person might not get any relief from a dose of either 100 milligrams or 900 milligrams. To complicate matters further, this sweet spot for CBD dosing can differ not only between symptoms, but also between patients.
Campaigners show support for legalizing cannabis for medical use in Atlanta, Georgia. Credit: Erik S. Lesser/EPA/Shutterstock
This is one of several reasons why researchers caution against self-medication with CBD products targeted at consumers. CBD is available in shops worldwide, but the legality of such sales varies widely. In Canada, selling cannabis and its derivatives is legal, whereas the European Union authorizes the sale of CBD derived from hemp (low-THC varieties of cannabis) but not from marijuana (high-THC cannabis). In the United States, the latest Farm Bill, which was enacted in 2018, potentially legalizes the production of CBD from hemp under certain conditions — although the sale of CBD products generally remains ostensibly illegal. Regardless of the legal situation at the federal level, CBD commercialization remains something of a free-for-all in the United States — individual states are making their own laws, and the FDA has taken only limited action to enforce federal laws on CBD. “They’ve sent some notices to companies that have made medical claims, but that’s about it,” says Marcel Bonn-Miller, a psychologist at the University of Pennsylvania, Philadelphia, and global scientific director at Canopy Growth Corporation, a cannabis company in Smiths Falls, Canada. (An FDA spokesperson responded that the agency “is working quickly to continue to clarify our regulatory authority over products containing cannabis and cannabis-derived compounds like CBD”.)
More from Nature Outlooks
Many such claims lie beyond the bounds of medical evidence — including that regarding CBD preparations that purport to prevent cancer or to treat Alzheimer’s disease. However, even products that make more modest claims could be problematic. In 2017, Bonn-Miller and his colleagues performed chemical analyses on 84 products purchased online from 31 companies, and found that only 31% were accurately labelled with regard to CBD content 14 . What’s more, many commercially available preparations have been found to be contaminated with intoxicating doses of THC, heavy metals and pesticides, as well as toxic solvents from the CBD extraction process. In a case reported by the US Centers for Disease Control and Prevention, up to 52 people in Utah became seriously ill or were hospitalized after using a CBD oil that contained an intoxicating synthetic cannabinoid drug. The possibility of such contamination is concerning to all potential users, and especially to people who are seeking relief from the effects of a health condition. “It’s one thing if you’ve got too much THC in gummy bears you’re using with friends, but something entirely different if it’s a kid you’re giving CBD for medical reasons,” says Bonn-Miller. “I don’t trust any CBD product until I’ve done the tests.”
Between two worlds
The regulatory disconnect that surrounds CBD creates an odd situation in which the public can self-medicate using a potentially questionable product, while scientists face a struggle to perform high-quality clinical trials. “The fact that CBD remains schedule 1 in the United States is unconscionable,” says Devinsky. That restrictive classification, he says, “is impairing research”.
Obtaining sufficient quantities of pharmaceutical-grade CBD to conduct a well-powered clinical trial is already difficult. “It’s extremely expensive,” says Leweke. “You need about one gram a day, and the list price is about 60 euros [US$67] per gram.” This is because the process of extracting CBD from the cannabis plant is complex and arduous — and when the goal is to obtain CBD for use in people, the substance must meet the high bar set for clinical-grade preparations, under which only minimal quantities of THC or other contaminants are permissible. Several companies have developed strategies for manufacturing fully synthetic CBD, an approach that essentially eliminates concerns about purity. But synthetic CBD still falls under the schedule 1 classification in the United States, which creates extra economic and bureaucratic hurdles for clinical trials. Even in Canada, where recreational cannabis has been legalized, Gobbi describes a complex application process and a more than six-month wait to obtain government authorization to conduct a CBD study in people or animals.
Unfortunately, if studies such as these are not done — or not done properly — then consumers will be left to fend for themselves in a poorly monitored marketplace. In that scenario, the signal of true clinical benefit would almost certainly be drowned out by the noise from personal anecdotes and the placebo effect, which could jeopardize the future of a potentially valuable medicine. “Humans are notoriously bad when they think they see patterns,” says Devinsky. “When everyone is convinced that they’re right with no data, I call that religion — and CBD is currently religion for the average person.”
Nature 572, S2-S4 (2019)
This article is part of Nature Outlook: Cannabis, an editorially independent supplement produced with the financial support of third parties. About this content.
Updates & Corrections
Correction 23 July 2020: An earlier version of this Outlook article misquoted Gabriella Gobbi. She said that 5-HT1A is desensitized in response to cannabidiol, not sensitized.
World Health Organization. Cannabidiol (CBD): Critical Review Report (World Health Organization, 2018).
Cunha, J. M. et al. Pharmacology 21, 175–185 (1980).
Leweke, F. M. et al. Transl. Psychiatry 2, e94 (2012).
Devinsky, O. et al. N. Engl. J. Med. 376, 2011–2020 (2017).
Devinsky, O. et al. N. Engl. J. Med. 378, 1888–1897 (2018).
Kaplan, J. S., Stella, N., Catterall, W. A. & Westenbroek, R. E. Proc. Natl Acad. Sci. USA 114, 11229–11234 (2017).
White, C. M. J. Clin. Pharmacol. 59, 923–934 (2019).
De Gregorio, D. et al. Pain 160, 136–150 (2019).
Zuardi, A. W., Morais, S. L., Guimarães, F. S. & Mechoulam, R. J. Clin. Psychiatry 56, 485–486 (1995).
McGuire, P. et al. Am. J. Psychiatry. 175, 225–231 (2018).
Bergamaschi, M. M. et al. Neuropsychopharmacology 36, 1219–1226 (2011).
Linares, I. M. et al. Braz. J. Psychiatry 41, 9–14 (2019).
Hurd, Y. L. et al. Am. J. Psychiatry https://doi.org/10.1176/appi.ajp.2019.18101191 (2019).
Bonn-Miller, M. O. et al. J. Am. Med. Assoc. 318, 1708–1709 (2017).