ESCAPE Study: Cannabidiol Use in Patients Treated for Substance Use Disorders, Prevalence of Use, and Characteristics of Users

Cannabidiol (CBD) is one of the two primary phytocannabinoids found in cannabis (Cannabis Sativa) alongside Δ-9-tetrahydrocannabinol (THC). Although discovered in 1940, interest in CBD has only surged relatively recently, becoming widely popular and used on a large scale over the past decade. This increase in popularity is linked to studies suggesting its potential benefits for a wide range of medical conditions. While THC is well-known for its psychotropic effects due to its partial agonism on cannabinoid receptors type 1 (CB1) and 2 (CB2), CBD was initially mistakenly believed to be non-psychoactive. However, CBD acts as an antagonist of CB1 receptors [1‒4], which prevents it from inducing psychotic symptoms or transient euphoria. Recently the World Health Organization’s (WHO) Expert Committee on Drug Dependence (ECDD) evaluated CBD and concluded that it has “no abuse or dependence potential” [5]. In contrast, CBD exhibits pleiotropic effects, acting as a partial agonist on serotonin 5-HT1A receptors and dopamine D2 receptors [6]. It also enhances endocannabinoid signaling by inhibiting the hydrolysis of anandamide [7‒9]. These pharmacological properties suggest that CBD may have anxiolytic, antipsychotic, antidepressant, and analgesic effects and could potentially be used to treat substance abuse disorders [7‒23]. Although these effects have been demonstrated in preclinical studies, there is still a lack of robust clinical evidence for CBD’s effectiveness in broader populations. Additionally, CBD inhibits voltage-dependent sodium and calcium channels, which contributes to its anti-epileptic effects [24, 25]. An oral solution containing CBD, Epidyolex^®, has been approved by both the European Medicines Agency (EMA) and the Food and Drug Administration (FDA) as a treatment for severe forms of juvenile epilepsy [25]. Despite this progress, the full range of CBD’s mechanisms of action remains not fully elucidated, highlighting the complexity of its pharmacological profile.

CBD currently has an unclear regulatory status in many countries. In some, CBD products are legal with varying maximum THC levels, while in others, they are banned outright [26, 27]. In the USA, marijuana and its derivatives are classified as Schedule I substances under the Controlled Substance Act, meaning they are considered to have a high potential for use disorders. In December 2018, hemp was removed from the definition of marijuana under the Controlled Substance Act, making it legal to cultivate, market, and consume products containing CBD derived from hemp [28, 29]. In Europe, the sale of CBD-containing products has also seen dramatic growth, both online and in specialty stores. Since late 2013, the cultivation and sale of these products can be subsidized by the European Union, provided the THC content of the initial plant does not exceed 0.2%. Notably, there is no requirement to control the THC level in the finished product [30, 31]. Conversely, in France, since 1990, only CBD products with zero THC content have been permitted, excluding flowers and leaves. However, the legislation was so ambiguous that many people purchased and sold CBD without regard to THC content, relying on European Union regulations. France eventually revised its laws in November 2020 to align with European legislation, allowing the sale of naturally derived CBD with a THC level of less than 0.2% in the original plant, but still excluding CBD flowers. This ban was lifted by the Council of State in January 2022 [32]. On the other hand, products containing CBD cannot be classified as drugs, with the exception of Epidyolex^®, due to the limited evidence of CBD’s clinical effectiveness. As a result, these products are currently marketed under the broad concept of “wellness.”

Faced with this exponential increase in CBD circulation, it is crucial to monitor the frequency of its use within the population and understand its general usage patterns. Currently, there are limited data in the literature on the prevalence of CBD use and even less on the profile and objectives of CBD users. A few recent studies, published since 2019, have estimated the prevalence of CBD use in various subpopulations, such as patients seeking orthopedic care or treatment for chronic pain [33‒35]. Additionally, only one US study has explored CBD use among substance users [36]. This group seems particularly likely to consume CBD, especially due to its similarities in presentation to cannabis. This hypothesis is further supported by observations from substance use disorders specialists, who report an increasing trend in CBD consumption among their patients. These professionals also question the relationship between CBD use and other known psychoactive substances, as well as the potential for CBD to be addictive. Therefore, it seems important to investigate CBD use within this specific population.

The primary objective of our study was to estimate the prevalence of CBD use among patients being treated for substance use disorders. The secondary objectives were (i) to characterize CBD use by describing usage patterns, assessing patient satisfaction with the product, and examining associated substance use and (ii) to describe the profile of CBD users.

A multicenter cross-sectional observational study was conducted at the Nantes University Hospital and the Vendée Departmental Hospital Center, in western France. All patients who had received an outpatient consultation for substance use disorder or had been hospitalized in the substance use disorder department at these two hospitals between May 6, 2020, and July 23, 2021, were invited to participate in the study. Exclusion criteria included being a minor, being under legal protection, or having insufficient proficiency in French. The study was approved by the French Committee for the Protection of Individuals on March 13, 2020 (file number 19.12.27.48510). In accordance with current French law, oral consent was obtained from all participants. The study was registered prior to patient enrollment (NCT04405414).

All data were collected using a questionnaire specifically designed for the study by a multidisciplinary steering committee consisting of substance use disorders specialists, psychiatrists, and pharmacists. Initially, sociodemographic data, reasons for specialized care, and information on analgesic and psychiatric comorbidities were collected by the referring physician at the time of patient inclusion. In the second phase, a clinical research technician gathered data on current medication use and the consumption of psychoactive substances over the past 12 months through an interview conducted 1 month after the inclusion date. To avoid impacting the therapeutic alliance and to minimize reporting bias, these consumption data were not shared with the patient’s referring physician.

The primary outcome was the proportion of patients who had used CBD at least once in the past 12 months compared to the total number of patients in the study.

To describe the patterns of CBD use, we collected data on the following aspects: the source of discovery, the source of supply, the form, the type of administration, the physician’s awareness of the patient’s use, the duration and frequency of use, and the effects sought from CBD. Patient satisfaction with the product was assessed using a visual analog scale, ranging from 0 (minimal satisfaction) to 10 points (maximum satisfaction).

To investigate the association between CBD use and other substance use, we analyzed the consumption of other psychoactive substances (alcohol, tobacco, e-cigarettes, cannabis, CBD, cocaine, heroin, and ecstasy/MDMA) using multiple correspondence analysis (MCA). The distinction between tobacco and e-cigarettes is justified by the strong enzyme induction potential of tobacco polycyclic hydrocarbons on CYP 1A2 (not nicotine, therefore absent in e-cigarettes) [37].

We collected patients’ sociodemographic characteristics (age, sex, profession), reasons for admission, analgesic and psychiatric comorbidities, and psychoactive substance use, and integrated these variables into both bivariate and multivariate analysis.

The prevalence of CBD use was estimated along with its 95% confidence interval (CI). Quantitative variables were described using their mean and standard deviation, while qualitative variables were described using their numbers and percentages. The mean level of patient satisfaction with CBD was calculated on the average score from a visual analog scale administered to each patient. MCA, a graphical representation method, was used to explore the correlation between two or more categorical variables. MCA summarizes information about the individuals in a multidimensional graph with only the two main dimensions representing most of the data variability displayed on the axes: the horizontal axis for the first dimension and the vertical axis for the second. This graph allows for a visual assessment of the proximities between the modalities of qualitative variables, revealing the correlations among them.

To identify a typical profile of CBD users, we compared CBD users to non-users using t tests for quantitative variables and χ² tests or Fisher’s exact tests for qualitative variables (bivariate analysis). Variables with a p value of less than 0.20 were included in a stepwise backward logistic regression analysis [38, 39]. The variables tested in the logistic regression included age, sex, profession, reasons for admission, comorbidities (both analgesic and psychiatric), and use of psychoactive substances (alcohol, tobacco, e-cigarettes, cannabis, cocaine, heroin, MDMA/ecstasy). Medication use variables were excluded from the logistic regression to avoid overfitting as these were correlated with comorbidities. All the analyses were conducted using R version 4.02.

Of the 201 patients initially included in the study, 18 could not be interviewed by the clinical research technician for various reasons, such as refusal or being unreachable despite multiple attempts. Consequently, a total of 183 patients were included in the analysis (Table 1).

The majority of patients were men (71%), with a mean age of 41 years, and most were not engaged in professional activities (64%). The three primary reasons for seeking specialized care for substance use disorders were alcohol use disorder (37%), opiate use disorder (28%), and multiple substance use disorder (13%). More than half of the patients had psychiatric comorbidities (59%), mainly mood disorders and anxiety disorders. The most commonly used psychiatric medications among the patients were benzodiazepines (41%), antidepressants (40%), and antipsychotics (28%). Nearly one-third of the patients were using opiate substitution treatments.

A large number of psychoactive substances were reported by patients in our study (Table 1). The vast majority of patients indicated using alcohol, tobacco, and cannabis. More than a third reported using e-cigarettes and cocaine. The frequency of psychoactive substance use is detailed in Figure 1. Tobacco and e-cigarettes were predominantly used on a daily basis. Alcohol was mainly consumed frequently, either daily or weekly. Approximately half of the patients used cannabis daily, while the other half used it less regularly, ranging from once a year to several times a week. Regarding heroin use, two distinct patterns emerged: one-third of the patients used it daily, while another third (31%) used it occasionally, once or several times a year. Cocaine and the stimulants MDMA/ecstasy were mostly used occasionally.

The prevalence of CBD use in the past 12 months was 38.2% (31.3%–45.7%).

Of the 70 patients who reported using CBD, only 26 continued to use it (37%) (Table 2). The primary reasons for discontinuation were the ineffectiveness of the product (59%) and its cost (18%). Two-thirds of the patients discovered CBD through a relative. CBD was predominantly consumed by inhalation, either by smoking the plant or using e-cigarette liquid. In about half of the cases, at least one physician involved in the patient’s care was aware of CBD use.

Regarding the duration of use, approximately half of the patients used CBD for a short period (4 weeks or less), while the other half used it for a longer duration (several months or more). In terms of frequency, about one-third of the patients used CBD daily, 19% used it regularly (weekly or monthly) and the remaining half used it more occasionally (a few times a year or for experimentation). The two main effects sought from CBD were recreational effects (39%) and substitution for cannabis (30%).

The average satisfaction level with CBD among these 70 patients was 5.1 out of 10 (SD: 2.95) (n = 69, with 1 missing data). Among the 27 patients who primarily sought a recreational effect, the mean satisfaction level was 4.3 (SD 3.25), whereas among the 21 patients who primarily sought cannabis substitution, it was 5.6 (SD 2.75) (p = 0.191).

According to the results of the MCA, 47% of the variability in substance use data could be explained by two dimensions. The primary dimension accounted for 32% of the variability, while the second dimension explained 14% (online suppl. material 1; for all online suppl. material, see https://doi.org/10.1159/000541887 [online suppl. S1] and online suppl. material 2 [online suppl. S2]). Cocaine, cannabis, and MDMA/ecstasy use were the main contributors to the primary dimension of the MCA (online suppl. material 3 [online suppl. S3]). E-cigarette use along with alcohol use was the main contributors to the second dimension.

The MCA indicates that CBD use is not strongly correlated with the use of other substances. However, it appears to be more closely correlated with sedative substances, such as heroin and cannabis, rather than with stimulant substances, such as cocaine and MDMA/ecstasy. This suggests that, although there is no typical profile for CBD users, they are more likely to use sedative substances than stimulants.

The association between CBD use and other factors in the bivariate analysis is detailed in online supplementary material 4 (online suppl. S4). Factors with a p value <0.20 associated with CBD use included age (p < 0.001), psychiatric comorbidities (p = 0.017), and the use of non-alcohol substance use (p = 0.009), including tobacco (p = 0.011), e-cigarette (p = 0.001), cannabis (p = <0.001), cocaine (p = <0.001), heroin (p = 0.021), and MDMA/ecstasy (p = 0.003) use.

To meet the validity conditions for logistic regression, a composite variable was created that included all psychoactive substance uses except alcohol. After adjusting for factors associated with CBD use identified in the bivariate analysis, the multivariate analysis revealed two factors significantly associated with CBD use (Table 3): age and the presence of psychiatric comorbidities. CBD users were more likely to be younger and to have psychiatric comorbidities.

The prevalence of CBD use among our 183 patients is notably high (38%). This finding aligns with the overall high consumption of psychoactive substances observed in our study. While recent research has estimated CBD use prevalence in various populations, our study is the first to assess this within a population of individuals with substance use disorders. Three American studies have estimated CBD use prevalence among patients with chronic pain at 24%, 25%, and 19% [40‒42]. Additionally, a study of English professional rugby players also explored CBD use [43]. Moreover, two studies, one from Germany and one from the USA, estimated CBD use prevalence in the general population at 4% and 40%, respectively [44, 45]. These variations in prevalence can be attributed to several factors, including cultural and regulatory differences between countries, participant age, and population diversity. The higher prevalence of CBD use in our study compared to existing literature is likely due to our specific study population, which consists of heavy users and experimenters of substances, contrasting with populations treated for chronic pain.

The legal ambiguity surrounding the sale and consumption of CBD in France at the time of the study partly explains its usage patterns. Due to limited advertising and information about the product, most patients discovered CBD through relatives, with very few finding it via the Internet. Additionally, a significant portion obtained CBD from family and friends, making these individuals the primary source of both information and supply.

The MCA indicates that there is no typical profile for CBD users among psychoactive substance users. However, CBD appears to be more commonly used by individuals who consume sedative substances such as cannabis or heroin, rather than those who use stimulant substances such as cocaine, ecstasy or MDMA. This finding aligns with CBD’s known anxiolytic properties, which are attributed to its action on 5-HT1A receptors [10‒15]. Three main factors may explain the high prevalence of CBD use among cannabis users. First, a third of the patients in our study reported using CBD to aid in cannabis withdrawal. Second, the similarity in appearance and method of administration (same aspect and type of administration – inhalation – for CBD in plant/leaves and same smell) may mimic patients’ behavioral dependence on cannabis. Third, potential factor is a low but variable rate of THC in products marketed as CBD. Indeed, it has been shown that CBD products purchased on the Internet or on the black market can have higher THC levels than advertised, sometimes well above the 0.2% concentration [46]. These two last factors may also contribute to the higher satisfaction reported by patients seeking cannabis substitution (mean satisfaction: 5.6/10) compared to those seeking recreational effects (mean satisfaction 4.3/10). Additionally, while hypnotic and anxiolytic effects were not the primary effects sought by patients in our study, it is likely that they are also sought by patients using CBD as a substitute for cannabis as a main effect sought, and therefore, their proportion is underestimated in our study.

All these results must be interpreted within the context of the French legal framework in place at the time of patient inclusion. At that time, the trade and consumption of CBD were only authorized if the finished product contained no THC, and sales of CBD flowers and leaves were prohibited. This ban on flowers and leaves was lifted by the Council of State in January 2022, allowing for the consumption of CBD in all its forms [32]. It is expected that the prevalence of CBD use has increased since this study was conducted. Similarly, it is anticipated that both the modalities of CBD use and the profile of users have evolved: (1) the proportion of CBD consumed in plant form, which was already high in our study (79%), has likely increased (2) purchases from specialized stores (40%) or online (21%) are probably become more common.

Furthermore, the current requirement for zero THC concentration in finished products may have impacted satisfaction among consumers seeking CBD as a substitute for cannabis (30% of patients in our study, with a mean satisfaction rating of 5.6/10). This impact is particularly plausible given that the quality of CBD products sold online and on the black market before the legalization of CBD flowers and leaves was variable, with some products containing THC levels significantly higher than advertised [46]. Consequently, the primary effect sought from CBD might now lean more toward recreational use than cannabis substitution. These hypotheses should be confirmed through future studies that explore CBD in this specific population.

In our study, many patients used CBD in search of therapeutic effects, including cannabis substitution (30%) and anxiolysis (13%). For half of them, this consumption corresponded to a self-medication since they did not inform their doctor(s) of this consumption. This practice of self-medication introduces several risks: (1) potential ineffectiveness of CBD in achieving the desired effects, possibly worsening initial symptoms (2) the risk of interactions with the patient’s existing medications.

First, among those seeking cannabis substitution, many might have expected psychoactive effects similar to cannabis. However, CBD acts as an antagonist of CB1 receptors, lacking the psychoactive effects associated with cannabis, such as psychotic symptoms or euphoria [1‒4]. This may lead to dissatisfaction with CBD and a continued use of cannabis. On the other hand, CBD’s anxiolytic effects, mediated through its action on serotonin 5-HT1A receptors, might have provided some benefit, particularly during cannabis withdrawal. Despite this, the anxiolytic and analgesic effects of CBD are not yet fully supported by clinical evidence, making it an unreliable sole treatment for these conditions.

Second, CBD’s pleiotropic action on various receptors and enzymes raises concerns about potential pharmacokinetic and pharmacodynamic drug-drug interactions. Among the potential pharmacokinetic interactions relevant to patients treated with psychotropic medications, a publication [47], highlighted the case report of a 13-year-old girl from whom cannabis led to a marked elevation in serum methadone and clinically significant signs of overdose likely due to CBD inhibitory action on cytochromes 3A4 (CYP 3A4) and 2C19 (CYP 2C19) [48]. This interaction is particularly relevant given that 32% of patients in our study are under opioid substitution treatment, including methadone. Additionally, this inhibitory action on the large family of cytochromes P450 is also the cause of potential interactions with antipsychotics such as loxapine, aripiprazole, quetiapine, and risperidone [48, 49], and with selective serotonin reuptake inhibitors, serotonin-norepinephrine reuptake inhibitors and tricyclic antidepressants [49‒52]. The pharmacodynamic effect of CBD on 5-HT1A serotonin receptors also plays a role in its interaction with antidepressants. These interactions should be considered carefully because of the high proportion of patients treated with antipsychotics and antidepressants, respectively, 28% and 25%, in our study. CBD has such a complex phamacodynamics that the many potential drug interactions cannot be all addressed in this article. These multiple drug interactions with CBD highlight the risk of consuming it as self-medication, without the advice of a doctor. Added to this is the great variability in the composition of CBD on the market [30], but also the current lack of knowledge and training of doctors on CBD and the lack of clear recommendations on the precautions of use and prescription of CBD.

Our study has several strengths. First, it is the only French study to evaluate the prevalence of CBD use and the first to our knowledge to assess this prevalence in patients with substance use disorders. Additionally, the systematic inclusion of all patients presenting for consultation or hospitalization during the study period minimized selection bias. Moreover, data on substance use were collected anonymously by a clinical research technician independent of the referring physician. This approach helped reduce under-reporting and non-responses by patients, thus minimizing missing data.

However, our study has two main limitations. First, recruitment was restricted to hospital settings, which may introduce a recruitment bias. The rate of prescribing of psychotropic medications and opioid substitution treatments is higher in our population compared to the general population of patients with substance use disorders. Second, the long duration of the inclusion period is responsible for a large variability in the data, including in CBD use patterns. Finally, our final sample size is relatively small due to the limited number of patients treated for substance use disorders at the two participating hospitals. However, it is probably representative of the population of psychoactive substance users receiving hospital-based treatment for substance use disorders because the Nantes University Hospital is a leading hospital in a large French region.

The prevalence of CBD use was notably high among our patients treated for substance use disorders. Our study showed that there was no typical profile of CBD users among substance users, although young people with psychiatric comorbidities were more likely to use CBD. Given these findings, it is crucial for clinicians to systematically inquire about CBD use in these populations. The risks associated with CBD use, whether for self-therapeutic or recreational purposes, seem to be underestimated. Special attention should be given to patients on multiple medications due to the potential for drug interactions, particularly with psychotropic drugs. Additionally, patients should be informed about the significant variability in THC content in CBD products, which can lead to unexpected and potentially harmful effects.

The authors would like to thank all the physicians who included subjects in the addictology and algology departments of the Nantes University Hospital and the Vendée Hospital.

The study was approved by the French Committee for the Protection of Individuals on March 13th 2020 (File No. 19.12.27.48510). In accordance with current French law, oral consent of all participants was received. The study was registered on clinicaltrials.gov before patient enrollment under the number NCT04405414.

The authors have no conflicts of interest to declare.

The ESCAPE study was funded by a grant from the Nantes University Hospital and the CHD de Vendée (joint internal call for tenders Nantes - La Roche sur Yon; RC19_0405). The funder had no role in the design of the study, the collection and analysis of the data, the decision to publish, or the preparation of the manuscript.

Claire Mauriat and Caroline Victorri-Vigneau acquired the funding. Marion Istvan, Morgane Rousselet, Claire Mauriat, and Caroline Victorri-Vigneau conceptualized the study and validated the study. Project supervision was conducted by Marion Istvan, Morgane Rousselet, and Caroline Victorri-Vigneau. Mélanie Duval and Marion Istvan conducted the data curation and formal analysis. Investigation was conducted by Marion Istvan, Morgane Rousselet, Edouard-Jules Laforgue, Marie Grall-Bronnec, and Claire Mauriat. The original draft of the manuscript was prepared by Mélanie Duval, and all authors have approved the final article.

Caroline Victorri-Vigneau and Claire Mauriat are co-last authors.

All data are available from the corresponding author upon request as the informed consent forms completed by all participants did not ask for permission to share the data via a data repository. Further inquiries can be directed to the corresponding author.