Illicit Cannabis Use to Self-Treat Chronic Health Conditions: A Cross-Sectional Study from the United Kingdom

Background: In 2019, it was estimated that approximately 1.4 million adults in the United Kingdom (UK) purchase illicit cannabis to self-treat chronic physical and mental health conditions. This analysis was conducted following the rescheduling of cannabis-based medicinal products (CBMPs) in the UK, but before the first specialist clinics had started treating patients. Objective: The aim of this study was to assess the prevalence of illicit cannabis consumption to treat a medically diagnosed condition, following the introduction of specialist clinics who could prescribe legal CBMPs in the UK. Methods: Adults over the age of 18 in the UK were invited to participate in a cross-sectional survey through YouGov® between 22 nd and 29th September 2022. A series of questions were asked about respondents’ medical diagnoses, illicit cannabis use, cost of purchasing illicit cannabis per month, and basic demographics. The responding sample was weighted to generate a sample representative of the adult population of the UK. Modelling of population size was conducted based on an adult (≥ 18 years) population of 53,369,083 according to 2021 national census data. Results: There were 10,965 respondents to the questionnaire, to which weighting was applied. 5,700 (51.98%) respondents indicated that they were affected by a chronic health condition. The most reported condition was anxiety (n = 1588; 14.48%). Of those suffering with health conditions, 364 (6.38%) purchased illicit cannabis to self-treat health conditions. Based on survey responses, it was modelled that 1,770,627 (95% confidence interval: 1,073,791 – 2,467,001) individuals consume illicit cannabis for health conditions across the United Kingdom. On multivariable logistic regression, the following were associated with increased likelihood of reporting illicit cannabis use for health reasons: chronic pain, fibromyalgia, post-traumatic stress disorder, multiple sclerosis, other mental health disorders, male gender, younger age, living in London, being unemployed or not working for other reasons, and working part-time (p<0.050).


Introduction
In November 2018, the United Kingdom (UK) Home Office rescheduled cannabis-based medicinal products (CBMPs), allowing them to be initiated by consultant physicians on the General Medical Council's specialist register for individuals who failed to achieve sufficient benefit from licensed therapies [1].This was in response to commentary provided by the then Chief Medical Officer to suggest that there was conclusive evidence of the medicinal value of CBMPs [1].At the end of 2022 it was estimated that 32,000 patients were now being prescribed CBMPs [2].The most common conditions for which they are now prescribed include chronic pain, anxiety, and fibromyalgia [3].However, there are still several barriers to eligible patients accessing CBMPs, including cost, perceived stigmatisation, and a lack of high-quality randomised controlled trials [4][5][6][7].Therefore, whilst there has been significant growth in the number of patients being prescribed CBMPs, this is surpassed by the most recent estimates of illicit cannabis use, including those who are using cannabis to self-treat diagnosed health conditions [8,9].
Cannabis is one of the most used drugs globally and the UK is among the top 10 highest consumers of cannabis in Europe [8,[10][11][12][13].The incidence of cannabis use in the UK is continuing to rise in line with other countries [8,[10][11][12][13][14].In March 2013, the past-year prevalence of illicit cannabis use was 6.3% [8].It has since risen to 7.6% [8].Whilst the perception of risk associated with cannabis is low and assessment of longitudinal registry data suggests CBMPs are largely well-tolerated [3,[15][16][17], there are inherent personal and societal harms that may be associated with illicit cannabis use, even when intended for symptom management or control.
Illegal cannabis does not have to meet any regulatory standards to ensure consistency or absence of harmful contaminants.Several potentially pathogenic bacterial species have been identified on cannabis, including Acinetobacter baumannii, Escherichia coli, Pseudomonas aeruginosa, and Clostridium botulinum [18].Moreover, several fungal species identified from dried cannabis flower, including Penicillium spp., Aspergillus spp., and Fusarium spp.
commonly cause invasive infections in the immunocompromised [18].The true incidence of adverse effects due to exposure to potential pathogens in illicit cannabis is not well characterised.Case reports indicate that the greatest risk is risk to immunocompromised individuals or those with underlying lung disease [18].Yet, there are reports of invasive infections secondary to contaminated illicit cannabis in otherwise healthy individuals [19,20].Moreover, the use of tainted fertilisers or phosphate-heavy fertilisers can lead to heavy metal contamination, specifically cadmium and arsenic, at levels exceeding those which are considered safe [18].Inappropriate use of pesticides may also expose individuals to harmful compounds, including carcinogens [21].
There are other risks associated with illicit cannabis use.Between 2010 and 2020 there were 162,000 convictions in English and Welsh courts where drug possession was the most significant offence [22].Over-policing of drug possession disproportionately affects Black communities [22].Beyond harms to the individual, illicit drug markets, including cannabis, actively contribute to the sustenance of organised crime groups and their exploitation of vulnerable individuals, including women, children, and refugees [23][24][25][26].
There is a paucity of high-quality randomised controlled trials to inform the evidence base on CBMPs [27,28].Consequently, whilst there is promising evidence of its medicinal effects, this is currently insufficient to recommend their use on a population basis, except for a few specific indications for which there are licensed CBMPs [28,29].There is therefore limited access to CBMPs in the UK.A cross-sectional, nationally representative survey on the prevalence of self-treating health conditions with illicit cannabis was conducted in October 2019 [9].It estimated that 1.4 million individuals were consuming illicit cannabis for health reasons [9].Whilst that analysis was conducted following the rescheduling of CBMPs, it was completed before the first clinic meeting regulatory standards, Curaleaf Clinic (formerly known as Sapphire Medical Clinics), began seeing patients [30].Consequently, the impact of rescheduling of CBMPs will not be incorporated in that study's findings.Moreover, the report by Couch and colleagues has not undergone peer review [9].This study therefore primarily aimed to assess the prevalence of cannabis use for health conditions, as the effect of access to CBMPs on illicit cannabis consumption is unknown.This study also aimed to assess which demographic factors are associated with an increased likelihood of consuming illicit cannabis for health reasons.

Study Design
A cross-sectional survey was administered to adults (aged ≥ 18 years old) residing in the UK between 22 nd and 29 th September 2022.The survey was distributed to a nationally representative sample by YouGov® (YouGov PLC, London, United Kingdom).
Participants were recruited utilising active sampling by YouGov® from a panel of more than 800,000 individuals [31].This method was utilised to generate a nationally representative sample of UK adults

Ethical considerations
Participants provide YouGov® with consent to be contacted via email and participate in questionnaires.YouGov® is a member of the British Polling Council and the European Society for Opinion and Marketing Research.It is also registered with the Information Commissioner.Participants are rewarded for taking part in surveys by receiving points, which can be converted to financial compensation.As all data was anonymised, there was no need for participants to provide further consent, beyond the implicit consent by completing the survey.The survey was developed by the study authors and reviewed by YouGov® to ensure compliance with their Global Code of Ethics.The questionnaire was distributed on behalf of the authors by YouGov®.The were subsequently provided access to anonymised data following completion of the study period.Ethical approval was obtained from the University of the West of Scotland School of Education and Social Sciences Ethics Committee (reference: # 2024-21236-17820).

Study Overview
The survey was developed utilising the questionnaire developed previously in a report by Couch and colleagues [9].This was to allow for direct comparison of prevalence between each analysis.Changes were made to the questionnaire, to account for the differences between October 2019 and September 2022 in access to specialist medical cannabis clinics who could prescribe CBMPs.Questions were delivered in series with branching logic applied between questions one and two, removing respondents who reported they did not have any diagnosed health conditions.Branching logic was also applied between question two and the rest of the survey, removing participants who did not use illicit cannabis for their health condition or were not prepared to disclose their use.The questions are detailed in full in Table 1.

GP -general practitioner; N/A -not applicable
In addition, the following demographic data was recorded: age, gender, geographical region, government region, working status, marital status, number of children in the household, parent or guardian status, and use of social media or messaging services in the last month.The social grade was also recorded as either ABC1 (middle class) or C2DE (working class or non-working individuals) as defined by National Readership Survey (NRS) social grade classifications [32].

Data Collection
In accordance with YouGov® active sampling methodology, participants were selected at random from the base sample of over 800,000 individuals.These emails are generic and do not alert the participant to the subject matter prior to engaging with the questions.The invitation link was active for this survey throughout the data collection period.This methodology, when used in conjunction with proportional weighting to a population matched sample has been shown to create a sample which is representative of the UK adult population [31].This has been demonstrated in predicting public opinion on political and social issues [31].

Statistical Analysis
The responses to the questionnaire underwent proportional weighting to ensure they were

Results
There were 10,965 respondents to the questionnaire.After weighting was applied, 5,700 (51.98%) participants reported experiencing any diagnosed health conditions (Figure 1).The number of UK adults estimated to be affected by a diagnosed health condition was subsequently modelled as 27,741,361 (95% CI: 27,242,290-28,240,433) (Supplementary Table 1).The most reported diagnoses were the groups Other physical condition (24.58%; 95% CI:   Multivariable logistic regression analysis was conducted including variables that were statistically significant on univariable analysis.The following were identified as having an association with consuming illicit cannabis for a diagnosed health condition: chronic pain, fibromyalgia, post-traumatic stress disorder (PTSD), multiple sclerosis, other mental health conditions, gender, age category, geographic region, employment status, and number of children in household (p<0.050)(Table 4).

Reasons for Consuming Cannabis Illicitly
On questioning as to why the participants chose to consume cannabis illicitly, the most common response was that they presumed legal access was very difficult (n=148; 40.75%) (Figure 4; Supplementary Table 5).Participants could select more than one answer and other responses included that they presumed legal access was expensive (n=105; 28.87%), they wanted to treat their condition quickly (n=103; 28.35%), or that they were unaware it was legal (n=88; 24.15%).When asked to consider their discussions with either a general practitioner (GP) or specialist doctor, 48.11% (n=175) said they had never discussed it (Figure 5; Supplementary Table 6).
11.86% (n=43) of respondents said they had discussed CBMPs with either a GP or specialist doctor, but they knew nothing about them.Considering those participants who had discussed it with their physician, 11.86% (n=43) said their doctor knew nothing about them, 9.71% (n=35) are exploring the option further or have explored it, 5.18% (n=19) decided against it in collaboration with their doctor, and 8.40% (n=31) were advised against CBMPs by their doctor.

Discussion
This nationally representative survey study estimates that 6.38% of individuals with a diagnosed health condition consume cannabis illicitly as a component of self-treating that health condition.Utilising census data this estimates that 1.77 million UK adults are using illicit cannabis for this reason.The health conditions with the strongest association for cannabis use on multivariate analysis were multiple sclerosis, chronic pain, and PTSD.The demographic factors with the strongest association with illicit cannabis use for health reasons included male gender, younger age categories, living in London, and being unemployed.The most common reported reason for using cannabis illicitly, rather than opting for legally prescribed CBMPs was due to presumed difficulties in accessing CBMPs.
One in four illicit cannabis users were unaware that CBMPs had been rescheduled and could be legally prescribed in the UK.Moreover, almost half or respondents who used illicit cannabis in this way had never discussed whether CBMPs may be an option for them.
The most striking finding from this study is that a 1.77 million people were estimated to use illicit cannabis to treat their diagnosed health conditions based upon best available survey data.This is an increase from the only previous nationally representative study which has sought to quantify the population of UK adults who use illicit cannabis for this reason [9].That report, which was conducted in 2019 but has not undergone peer review, before the introduction of specialist medical cannabis clinics, but after the rescheduling of CBMPs in the UK, estimated the figure to be 1.4 million [9].Whilst this previous study did not present the 95% CIs for this estimate it is important to note that the lower bound of the figure derived in the present study is 1.07 million, which may therefore reflect that there has been no change.The difference could therefore simply represent natural variance in repeated surveying of similar populations.Whilst the present study utilised very similar methodology to the study conducted in 2019, there are some key differences, which may also be reflected in the modelled population estimate.The present study utilised terminology to capture individuals with any diagnosed health condition incorporating variables, such as Other mental health condition, Other physical condition, and Other, not described.The 2019 study, in comparison utilised a longer list of specific diagnoses, but could not be exhaustive due to limitations of a survey study, and did not utilise a catch-all term [9].Consequently, the prevalence of any diagnosed health condition was 51.98% in the present study, compared to 46.37% in the 2019 report [9].Both figures are similar to the estimated prevalence derived from the UK sample of the European Health Interview Survey for 2019-2020 (48.06%) [33].The 95% CIs of estimated illicit cannabis use for health reasons overlap between 2019 and 2022, suggesting there was no change to the overall proportion of people who consume cannabis.This is supported by a study by Waldron and colleagues which found perception of risk towards both CBMPs, and recreational cannabis is unchanged despite the rescheduling of CBMPs [34].
Considering the potential health and societal harms that may be associated with illicit cannabis [18][19][20][21][23][24][25][26], irrespective of potential medicinal value, it is important to consider policy interventions which may facilitate the transition of patients from illicit cannabis to legal CBMPs with clinical oversight.Despite rescheduling there may be many factors which mean participants continue to consume illicit cannabis.The study highlights a general lack of awareness of the rescheduling of CBMPs, with one in four participants being unaware of their legal status.Many participants also highlighted that they thought access to CBMPs may be difficult, expensive, or not appropriate to get timely treatment of their condition.
Almost half of all individuals using illicit cannabis for self-treating their health condition had not talked about CBMPs with their GP or specialist.This may be reflective of the high levels of perceived stigma among medical cannabis patients [4].One in five individuals reported that their doctor had either advised against CBMPs or did not know enough about these medications.This is supported by data from the Primary Care Cannabis Network which suggests 72% and 68% of GPs are concerned about the unlicensed nature of most CBMPs and lack of efficacy respectively [35].There may be supplementary barriers to accessing CBMPs which are not assessed in the present study.Most care for individuals prescribed CBMPs is provided in the private sector [36,37].The associated costs of this care may mean that it is not accessible to all.In addition, patients must meet national criteria for eligibility for CBMPs [38].Therefore, patients who have not had a sufficient trial of licensed therapies will not be able to access CBMPs.Policy interventions, specifically targeted at overcoming these barriers to access may have positive implications with respect to harm reduction.Implementation of National Health Service (NHS) provisions to access CBMPs and care for individuals who report positive impacts on their diagnosed health conditions from illicit cannabis, and otherwise meet relevant eligibility criteria [38], may have positive effects at an individual and population basis.Ultimately improved quality and quantity of clinical research will be required to truly address barriers to accessing CBMPs.At present there is insufficient evidence to support national prescribing via the NHS [29].Research with CBMPs that demonstrates cost-effectiveness in appropriate conditions will help reduce financial barriers, increase healthcare practitioner education, and help reduce stigma.
This study highlights differences between groups which may influence their likelihood of self-treating their health conditions with illicit cannabis.Multiple sclerosis had the strongest associatiation with illicit cannabis on multivariate logistic regression.Patients with multiple sclerosis were also more likely to report higher levels of expenditure on illicit cannabis.
Multiple sclerosis was also the single condition that was most likely to be aware that CBMPs were available legally on prescription and to have discussed its use with a doctor.These findings may be secondary to awareness of nabiximols, a licensed CBMP for spasticity in adults with multiple sclerosis [29].Whilst this may serve to increase awareness of CBMPs as a treatment class, nabiximols is only available in restricted settings [29].Chronic pain is the most common reason why CBMPs are prescribed in the UK and is the most common indication for symptomatic treatment in multiple sclerosis, but nabiximols is not available in this setting [39].
Observational or real-world evidence has played a crucial role in advancing the field of cannabis science in the absence of a sufficient number of high-quality randomised controlled trials.In the UK, for example, the efficacy of CBMPs in treating rare, treatment-resistant forms of epilepsy in select individuals was an important factor in the rescheduling of CBMPs [1].As this study uncovers, 1.77 million people in the UK are estimated to consume illicit cannabis for health reasons.This is observational evidence of the potential therapeutic value of CBMPs, but insufficient to support wider access to CBMPs for individuals who are cannabis naïve.It does, however, support the need for further funding for randomised controlled trials of CBMPs in conditions such as chronic pain and anxiety which are estimated in the present study to affect 3.99 and 7.73 million UK adults respectively.
Considering the inherent challenges in conducting randomised controlled trials with CBMPs [40], novel approaches to incorporating and analysing real-world evidence should also be considered.Large patient registries, such as the UK Medical Cannabis Registry, may be utilised in increasingly novel and innovative ways to further understand the clinical efficacy of CBMPs, beyond the preliminary data which has been published on chronic pain, anxiety, fibromyalgia, and PTSD so far [41][42][43][44][45][46][47].
Despite the utilisation of a sampling and weighting methodology to derive a nationally representative population, this study is subject to inherent limitations.Responses to the survey may be affected by a social desirability bias [48].Whilst all responses were anonymous, it is still well-known that participants in research are more likely to provide responses which are deemed acceptable.This may therefore lead to a reduction in declared illicit cannabis use for health reasons.YouGov® utilises an online sampling methodology, which may inappropriately exclude individuals who cannot engage with digital technology.This may disproportionately affect certain members of society and therefore the representativeness of the survey [49].Whilst the weighting of the survey is adjusted to account for this sampling bias, there may still be characteristics of those who lack digital inclusivity which are unable to be accounted for by statistical weighting.The weighting of YouGov data, also does not account for ethnicity or race and therefore information about this variable is not included in this analysis.Considering how over-policing of cannabis possession disproportionately affects Black communities [22], further information on ethnicity would have been beneficial.Another limitation is the self-reporting of conditions.Whilst efforts were made to word the first question appropriately to specifically ask about conditions diagnosed by an appropriately trained healthcare professional, without confirmation from a healthcare professional of the diagnostic accuracy it may lead to inappropriate recording of conditions.In addition, the study was limited to asking about five specific conditions, rather than having a more discrete list of other conditions which people may self-treat with illicit cannabis.Further conditions were not added due to costconstraints.This limits the additional analysis that can be assessed, such as the more comprehensive list published by Ware and colleagues [50].Whilst this reduces the granularity of available data, by including terms to capture any other medical conditions this is likely to improve the accuracy of the estimated total population of UK adults who use illicit cannabis for health reasons.
In conclusion, this study estimates that 1.77 million UK adults are consuming illicit cannabis for the purpose of managing their health conditions based upon nationally representative survey data.This number has not materially changed since 2019, when it was estimated to be 1.4 million [9].This is despite the introduction of specialist medical cannabis clinics, who provide clinical care to an estimated 32,000 individuals in the UK [2].To address the potential public health and societal problems this creates, despite any therapeutic value derived from the illicit cannabis, it is important to prioritise polices which help reduce the barriers to accessing CBMPs.This is particularly important for the estimated 1.77 million UK adults who are consuming illicit cannabis for health reasons.Beyond this, it is important to prioritise funding and the adoption of novel research methodologies to establish the efficacy of CBMPs and the role they should play in the treatment of chronic health conditions for all individuals.

Figure 2 .
Figure 2. Prevalence (%) of illicit cannabis consumption reported by respondents to a nationally representative survey for medical conditions diagnosed by a healthcare professional.

Figure 3 .
Figure 3. Proportional spend on illicit cannabis for a diagnosed health condition by respondents reported as a proportion of patients declaring illicit cannabis use to self-treat health conditions.

Figure 4 .
Figure 4. Reasons why respondents consumed illicit cannabis to self-treat their health conditions reported as a proportion of patients declaring illicit cannabis use to self-treat health conditions.

Figure 5 .
Figure 5. Outcome of discussions with general practitioner or specialist regarding cannabisbased medicinal products reported as a proportion of patients declaring illicit cannabis use to self-treat health conditions.

Table 1 .
Questionnaire and available responses administered to a nationally representative sample via YouGov® to adults (aged ≥ 18 years old) residing in the UK between 22 nd and 29 th September 2022.
Approximately, how much money would you say you personally spend on cannabis to manage or treat some symptoms or side effects brought on by the treatment of your condition(s), in an average month?As a reminder, your answers will always be treated anonymously and will never be analysed individually.Which, if any, of the following are reasons why you obtain cannabis for your condition(s) in the way you currently do?(Please select all that apply.If any of your answers don't appear in the list below, please type them in the 'Other' box) I was not aware that it is available legally to manage/ Thinking about speaking to a GP or specialist about medical cannabis being used to manage/ treat your current physical/ mental health condition(s)...Which ONE, if any of the following statements BEST applies to you? (If you have spoken to a GP or specialist about medical cannabis more than once, please think about the most recent time)

Table 2 .
Figure The condition with the highest proportional prevalence of illicit cannabis use was multiple sclerosis (38.31%; 95% CI: 23.19-53.43%).The specific condition with the largest estimated population self-treating with illicit cannabis in the UK was anxiety (n = 775,782; 95% CI: The full demographics of individuals who consume illicit cannabis for health reasons is contained within Supplementary Table3. 1. Prevalence (%) of medical conditions diagnosed by a healthcare professional reported by respondents to a nationally representative survey.Mental Health -Other mental health condition; Other Conditions -Other, not described; Physical Condition -Other physical condition; PTSD -post-traumatic stress disorder Reported Use of Illicit Cannabis for Health Conditions condition (Figure2).The total population of UK adults who use illicit cannabis to manage health conditions was estimated as 1,770,627 (95% CI: 1,073,791-2,467,001) (Table2).

Table 2 .
Illicit cannabis consumption reported by survey respondents to self-treat medical conditions diagnosed by a healthcare professional.

Table 4 .
Multivariable logistic regression analysis of the relationship between independent variables and the likelihood of consuming illicit cannabis to self-manage a diagnosed health condition.