Prevalence of cannabis use for pain management in Quebec: A post-legalization estimate among generations living with chronic pain

ABSTRACT Background Medical cannabis has been legal in Canada since 2001, and recreational cannabis was legalized in October 2018, which has led to a widespread increase in the accessibility of cannabis products. Aims This study aimed to estimate the prevalence of cannabis use among adults living with chronic pain (CP) and investigate the relationship between age and cannabis use for CP management. Methods A cross-sectional analysis of the COPE Cohort data set, a large Quebec sample of 1935 adults living with CP, was conducted. Participants completed a web-based questionnaire in 2019 that contained three yes/no questions about past-year use of cannabis (i.e., for pain management, management of other health-related conditions, recreational purposes). Results Among the 1344 participants who completed the cannabis use section of the questionnaire, the overall prevalence of cannabis use for pain management was 30.1% (95% confidence interval 27.7–32.7). Differences were found between age groups, with the highest prevalence among participants aged ≤26 years (36.5%) and lowest for those aged ≥74 years (8.8%). A multivariable logistic model revealed that age, region of residence, generalized pain, use of medications or nonpharmacological approaches for pain management, alcohol/drug consumption, and smoking were associated with the likelihood of using cannabis for pain management. Conclusions Cannabis is a common treatment for the management of CP, especially in younger generations. The high prevalence of use emphasizes the importance of better knowledge translation for people living with CP, rapidly generating evidence regarding the safety and efficacy of cannabis, and clinicians’ involvement in supporting people who use cannabis for pain management.


Introduction
Despite limited evidence of the safety and efficacy of cannabis as a treatment, chronic pain (CP) is among the most frequent health-related conditions for which cannabis is used. [1][2][3] Medical cannabis has been legal in Canada since 2001, and recreational use was legalized toward the end of October 2018. 4 Following this new legislation, social acceptability has increased 5 and consumers have experienced improved access to both recreational and medical cannabis. 6 Evidence regarding the benefits of using cannabis for CP management is limited and contradictory. Though some studies suggested advantages, a recent metaanalysis reported small to very small improvements in terms of pain, functioning, and sleep. 7 Physical and mental harms are also present, such as dizziness, nausea, vomiting, drowsiness, impaired attention, and transient cognitive impairment. 7 Due to significant gaps in research regarding the safety and efficacy of cannabis and cannabinoids for the management of pain, the International Association for the Study of Pain Presidential Task Force 8,9 does not currently support its use. Their work highlights preliminary evidence for cannabinoid-induced analgesia in preclinical investigations but emphasizes that no evidence of sufficient quality currently exists in clinical populations. 8 This echoes the position statement from the Canadian Rheumatology Association 10 and UK National Institute for Clinical Excellence. 11 A number of factors complicate the review of the current literature or conducting new investigations on the use of cannabis and cannabinoids for the management of CP, including the great variability of types of cannabinoids used, route of administration, dosage, types of populations, types of pain, as well as the quality of the research currently published. 12 Despite published position statements and concerns, the attitudes of patients with CP toward cannabis products appear to be highly favorable. 13,14 An international panel including representatives from relevant medical and nonmedical specialties in addition to methodologists and patients recently recommend that if standard care and other treatment options are not sufficient, a trial of noninhaled medical cannabis or cannabinoids could be suggested to patients (with the knowledge that the improvements in pain and sleep will probably be small). They were in favor of cannabis use for CP management with some careful follow-up (e.g., start with low doses, carefully monitor adverse events, avoid driving, etc.). 15 The pre-legalization prevalence of cannabis use in Canadians living with CP was estimated to be 10% or less (studies conducted before October 2018). 16,17 In Ware and colleagues' study, 10% of participants reported current use of cannabis for pain relief, and results further highlighted a wide range of frequency of use, dosage, mode of administration, and type of cannabis product. 16 A few years prior to the legalization of recreational cannabis, Ste-Marie and colleagues' study of individuals living with rheumatic diseases reported that 4.3% of participants had ever used medical cannabis, with 2.8% reporting continuing use. 17 They repeated the same study method post-legalization and found a prevalence of 12.6% of individuals who had ever used medical cannabis, with 6.5% reporting current medical use. 18 Their results suggested a twofold increase in prevalence of medical use of cannabis following the legalization of recreational cannabis. However, it is relevant to expand investigations about the current prevalence of cannabis use in large and diversified samples of Canadians living with CP in this post-legalization environment.
In order to have a more complete picture, particular attention should also be given to sociodemographic factors that may predict cannabis use among people living with CP. According to some studies conducted in the general population and published before the legalization of recreational use, age appeared to be a factor influencing the use of cannabis, alongside other sociodemographic factors, such as sex, tobacco consumption, race/ethnicity, marital status, and other socioeconomic factors. 19 Though some Canadian general population studies showed no substantial variation across age groups, 20 a 2015 report observed a lower prevalence of cannabis use among older people. 1 It should be noted that limited post-legalization data specific to Canadian CP populations is available regarding the influence of age and other sociodemographic factors on cannabis use.
The goals of the present study were to (1) measure the post-legalization prevalence of cannabis use among Canadians living with CP and explore the reasons for use, (2) compare the prevalence of cannabis use for CP management among generations/age groups, and (3) investigate the relationship between age and cannabis use for CP management adjusted by several sociodemographic and clinical factors.

Data Source
The present study utilized data from the ChrOnic Pain trEatment (COPE) Cohort, 21 which is a data set designed to further our understanding of real-world usage of pharmacological as well as of physical and psychological treatments among people living with CP. The COPE Cohort includes 1935 French-speaking adults from the province of Quebec (Canada) who self-reported living with CP (persistent or recurrent pain for more than three months 22 ). Between June and October 2019, participants completed a web-based questionnaire that included previously used items and validated composite scales. All indicators identified as a minimum data set by the Canadian Registry Working Group of the Strategy for Patient-Oriented Research Chronic Pain Network 23 were included: pain location, circumstances surrounding its onset, duration, frequency, intensity, neuropathic component, interference, physical function, anxiety and depressive symptoms, age, sex, gender, and employment status. Item selection was also guided by the core outcome domains and measures by the Initiative on Methods, Measurement and Pain Assessment in Clinical Trials, 24,25 items of the Canadian minimum data set for chronic low back pain research, 26 and variables assessed in the Quebec Pain Registry. 27 Selfreported data were also intended to be linked to longitudinal administrative data (medical and prescription claims). The complete methodology of the COPE cohort implementation is described elsewhere. 21 In terms of representativeness, online self-reported data collection enabled the research team to reach many participants from all regions of the province of Quebec (including several remote understudied regions), reduce social desirability bias, 28 as well as minimize data entry errors. COPE cohort participants' pain characteristics, age, employment status, and level of education have been found to be representative of other large random samples of adults living with CP in Canada and elsewhere. [29][30][31][32][33][34] However, the COPE cohort was found to overrepresent individuals having access to the Internet, women, and users of pain medications. 21 The online recruitment strategy and questionnaire administration could explain the oversampling of women, because they use Facebook 35 and work in online environments more often than men. 36 Women also use more prescribed medications. 37 Nevertheless, the COPE cohort still includes a diverse spectrum of profiles, allowing the assessment of valid multivariable associations. For prevalence estimates, calculating genderstratified measures reduces the possibility of bias that may arise from the overrepresentation of women.

Informed Consent Statement
The present study received ethics approval from the Université du Québec en Abitibi-Témiscamingue's research ethics committee (#2018-05, Lacasse, A.). Participants completed informed consent for participation online and were informed that they would not be identifiable from the published results.

Selection Criteria
This study was conducted using self-reported data among the sample of participants who completed the cannabis use section of the questionnaire (n = 1344/ 1935). Participants included in the study were clinically comparable to those not included (n = 591/1935) in terms of mean age (49.72 versus 52.3 years old), proportion of women (83.5% versus 86.3%), and proportion of individuals reporting moderate to severe pain in the past 7 days (67.7% versus 70.9%).

Study Variables
Past-year use of cannabis. The cannabis use section of the questionnaire was made up of three yes/no questions about past-year use of cannabis (i.e., for pain management, for the management of other health-related conditions, for recreational purposes; non-mutually exclusive prevalence estimates). Overall prevalence of use was computed by calculating the percentage of participants who answered "yes" to at least one of these three questions.
Age groups. Age groups were formed based on recognized generational groups in Canada 38 applied to the year of study recruitment (2019): (1) Generation Z (≤26 years old), (2) children of baby boomers (27-47 years old, which include millennials/Generation Y), (3) baby busters (48-53 years old, which includes Generation X), (4) baby boomers (54-73 years old), and (5) parents of baby boomers (≥74 years old, also known as traditionalists or the silent generation). This categorization was chosen because we felt that it would reflect different social norms that might influence cannabis use.
Covariates. We considered the following covariates: sociodemographic characteristics (gender identity, gender roles, country of birth, Indigenous identity, employment status, disability, education level, region of residence), CP characteristics (generalized pain, multisite pain, pain frequency, duration,  intensity, tendency toward pain catastrophizing, neuropathic component, pain interference), pain management (current use of prescribed pain medications, over-the-counter/nonprescribed pain medications, or nonpharmacological treatments; access to a trusted health care professional for pain management), and health profile and lifestyle variables (psychological distress, perceived general health, number of medications currently used, physical functioning, alcohol or drug use, cigarette smoking). All COPE self-reported measurements and validated composite scales are described elsewhere. 21

Statistical Analysis
The characteristics of the study population and cannabis use (overall, for pain management, for other health-related conditions, for recreational purposes) were summarized using descriptive statistics. Prevalence of cannabis use (overall and for pain management) was then compared across the abovementioned age groups. As per best sexand gender-based analysis practices guidelines, 39 prevalence estimates were also stratified across gender identity subgroups (women, men, nonbinary). Chi-square and Fisher's exact tests, in addition to post hoc multiple comparisons (Tukey-style multiple comparisons of proportions), were conducted.
A multivariable logistic regression model was used to investigate the relationship between age and cannabis use for CP management. The COPE cohort variables that could potentially be related to age and/or cannabis use were identified a priori and were included in the regression analysis. These variables were chosen based on existing literature and clinical considerations. Our substantial sample size allowed us to favor this approach versus other criticized covariate selection techniques (e.g., relying on bivariate regression analyses, P values, 40 or computer algorithms 41 ). Variance inflation factors were below 2.5 for all variables included in the multivariable model (variance inflation factors <5 or 10 are often suggested to detect multicollinearity 42 ). The Hosmer-Lemeshow test confirmed a well-fitting model (chi-square = 7.2; P = 0.5190). Regression results are presented as adjusted odds ratios (ORs) along with their 95% confidence intervals (CIs). Because interpreting ORs quantitatively can be misleading when the underlying outcome is common and when effect sizes are modest to strong, 43 they were interpreted qualitatively (i.e., presence of a statistically significant association and its direction rather than its magnitude).
A sensitivity analysis was also carried out to assess the impact of missing data imputation on conclusions. Multiple regression analyses were used to estimate missing values 41 (five different imputations) and our model was run again using those data sets as a sensitivity analysis. Analyses were conducted using SAS version 9.4 (SAS Institute, Cary, NC, USA). Table 1 presents the characteristics of the sample. Among the 1344 participants who were included in the present study, 83.5% self-identified as women, 16.2% as men and 0.3% as nonbinary (n = 4). The average age 49.7 (SD ±13.2 years, range 18-88 years) and our sample was included mostly children of baby boomers (27-47 years,  (Figure 1). Prevalence of cannabis use included 30.1% (95% CI 27.7-32.7) of participants who reported using cannabis to manage pain, 9.1% (95% CI 7.6-10.8) to manage other health-related conditions, and 12.7% (95% CI 11.0-14.6) for recreational purposes (non-mutually exclusive groups, because participants could check more than one reason). In addition, 55.4% checked only one reason for cannabis use, 31.8% checked two reasons, and 12.8% checked all three reasons (pain, other health-related conditions, and recreational purposes). Only 3.0% used cannabis for recreational purposes without selecting the pain management option. Although cannabis use was assessed using standardized questionnaire items, 56 participants also reported cannabis use when asked about nonpharmacologic treatments (in answer to a semi-open-ended question), allowing us to draw certain conclusions regarding how cannabis is represented among participants. Figure 2 presents the prevalence of cannabis use overall and for pain management across generations. The proportion of participants reporting cannabis use for pain management significantly varied across age groups (P < 0.0001): 36.5% for Generation Z (≤26 years), 36.6% for children of baby boomers (27-47 years), 30.7% for baby busters (48-53 years), 24.4% for baby boomers (54-73 years), and 8.8% for the parents of baby boomers (≥74 years). Post hoc multiple comparisons revealed differences between (1) parents of baby boomers versus Generation Z, (2) parents of baby boomers versus children of baby boomers, (3) parents of baby boomers versus baby busters, and (4) baby boomers versus children of baby boomers. The prevalence of cannabis use for the management of pain was not found to significantly vary across gender identity groups (P = 0.4253; see Figure 3). Table 2 shows the estimates of the multivariable model used to investigate the relationship between age and cannabis use for CP management. Independent of other sociodemographic and clinical factors included in the model, being older was associated with a decreased likelihood of using cannabis to manage CP symptoms; Compared to Generation Z, baby busters (OR = 0.4, 95% CI 0.2-0.9), baby boomers (OR = 0.3, 95% CI 0.1-0.6), and parents of baby boomers (OR = 0.1, 95% CI 0.01-0.8) were less likely to use cannabis for CP management. Neither missing data imputation nor entering age in the model as a continuous versus categorical variable changed that conclusion.

Discussion
The present study took advantage of the large COPE cohort data set that includes more than one thousand adults from the province of Quebec (Canada) who are living with CP to  estimate the post-legalization prevalence of cannabis use. The present results indicate an overall prevalence of cannabis use among Canadians living with CP of 33.3%, followed closely by a prevalence of 30.1% for pain management. Based on studies conducted before the legalization of recreational cannabis, 16,17 the prevalence of cannabis use estimated in the present study indicates a threefold increase in reported usage. Cannabis use for CP management was especially frequent in younger generations, but no gender differences were found.

Prevalence of Cannabis Use
Few cannabis use prevalence studies applied the same methodology among CP populations before and after legalization of recreational cannabis in Canada. 17,18 In the general population, however, Statistics Canada reported an increase in cannabis use from approximatively 15% to 17% between 2018 and 2019. 44 It was also reported in the Canadian Cannabis Survey 2020 that 27% of Canadians had used cannabis in the past 12 months (compared to 22% in 2018). 5 An international study showed that frequency of adults reporting having ever used cannabis for medical purposes in the United States in 2018 was 34% in states where recrational use is legal, 23% in states where it is illegal, 25% in states where medical use only is legal; in Canada, the frequency was 25%. 45 In a study published post-legalization and conducted among people living with fibromyalgia, Fitzcharles and colleagues 46 found a prevalence of cannabis use of 23.9%. The prevalence of cannabis use was higher in our study (33.3% in general and 30.1% for pain management in 2019). Because people living with CP have reported using even more cannabis during the first wave of the COVID-19 pandemic, 47 it seems reasonable to expect the prevalence to be even higher today. Our results suggest that cannabis is a common treatment reported by people living with CP and underscore the importance of rapidly generating more evidence on the safety and efficacy of cannabis. In addition, some participants described cannabis use when asked about nonpharmacologic treatments, which suggests that the pharmaceutical properties of cannabis are underestimated. In fact, cannabis is often considered natural and safe and not a drug. 48,49 In essence, there is an urgent need for better knowledge translation for people living with CP and engagement of communitybased clinicians in supporting people who use medical as well as recreational cannabis for pain management. Above all, generating high-quality evidence on the safety and efficacy of cannabis use among people living with CP is a priority. Legalization of recreational cannabis in Canada has created a situation where it can be used for treatment without having gone through rigorous drug approval process. Security is a challenge because patients can selfmedicate without proper advice from a health care professional, employees of government-operated recreational cannabis are not authorized to give advice on medicinal use, and cannabis use is not systematically recorded in medical and pharmacy charts. Even in regions where cannabis is legal, people do not always disclose cannabis use to health care professionals for a variety of reasons, including a perceived stigma, unfavorable attitudes of health care professionals, not wanting cannabis in their medical chart, fear of losing professional status, and fear of losing health insurance. 50 This self-medication raises major concerns regarding drug-drug interactions and management of adverse effects. In terms of support, it has been argued that pharmacists should be providing counsel to medicinal cannabis users 51,52 and potentially should be the dispensers of medical cannabis in Canada. 53 A recent survey of individuals living with CP suggested that more than half would prefer their treating physician to be involved in the prescription of cannabis products. 13 More studies about the expected versus actual versus desired role of primary care health care professionals should be conducted. Given the most recent position statements from the International Association for the Study of Pain Presidential Task Force 8 and the Canadian Rheumatology Association, 10 which do not currently support the use of cannabis and cannabinoids for the purpose of pain management, it is not likely that people will disclose their use of cannabis to their health care professionals. This lack of communication can lead to unintended consequences. For instance, cannabis use in people living with CP who are prescribed a chronic opioid therapy has been associated with opioid misuse. 54 Therefore, if professionals were better informed of patients' use of cannabis, they could work with patients appropriately by providing sound advice regarding pain management. It is thus essential to rapidly put in place awareness and education campaigns for patients and health care professionals and to encourage professionals to have an open and honest discussion with their patients regarding their use of cannabis products.

Reasons of Consumption
Though 30.1% of our participants with CP used cannabis for pain management and 12.7% reported using it for recreational purposes, these categories were not mutually exclusive, and only few participants used cannabis for recreational purposes without pain management goals (3%). To our knowledge, no other study specifically evaluated recreational cannabis use post-legalization in Canadians living with CP. Our results further emphasize the importance of better support for people living with CP. Because the reasons for use and the rationale for self-medication may vary compared to the general population or people using cannabis for other medical reasons, we suggest that efforts be tailored to the CP population. Specific "other health-related conditions" were not listed by participants in our study, but results of recent Canadian studies among medical and recreational cannabis users suggest that, in addition to pain, anxiety, insomnia, and depression are common conditions for which cannabis is used. 3,55

Differences across Age Groups
One aim of the present study was to explore whether there were differences in cannabis use for CP management between age groups. In fact, new legislation and age, alongside sex, ethnicity, and socioeconomic status, can have an influence on cannabis use trends 19 and are important to take into consideration to understand the real context of cannabis use in Canada. In our study, a higher prevalence of cannabis use for CP management was observed among younger generations, whereas people aged 74 years or older (parents of baby boomers) reported the lowest use. Baby boomers did not stand out in terms of cannabis use, although cannabis use was more prevalent among this generation in the 1960s and 1970s, which could be a factor influencing openness toward cannabis use. 56 Our results are in line with other studies suggesting that cannabis use is higher in younger age groups and not often use in those aged 75 years and older (both in CP populations 10,57 and in the general Canadian population 58 ). The age groups with the highest cannabis use were those ≤26 and 27-47 years old, in accordance with previous investigations. This higher prevalence of cannabis use among youth could be partly explained by the fact that younger generations often have fewer responsibilities (cannabis use decreases with additional responsibilities such as work and family 59 ) or because of peer influence. 59,60 It could also be related to vaping habits among youth. 61 Because cannabis use at a younger age has been associated with multiple subsequent adverse health and social effects, 62 it is concerning to see such a high prevalence of cannabis use among the younger generations in Canada. Globally, knowledge translation implementation and messages for people living with CP should be tailored according to age.

Absence of Gender Differences
Some studies of people living with CP showed a lower prevalence of cannabis use among women. 57 Others found that women were more likely to substitute cannabis for analgesics. 63 Differences have also been noted in the general Canadian population. In fact, in 2013 the prevalence of cannabis use was nearly double among men compared to women (13.9% versus 7.4%). 64 Similar differences between men and women were found in 2020 (31% versus 23%). 5 In contrast to these reports, our study revealed that prevalence of cannabis use for pain management was not statistically different between men (32.3%) and women (29.7%). Multivariable analyses supported the absence of gender identity and gender role differences. Comprehensive sex-and gender-based analyses 65-67 of cannabis use among people living with CP could complete our findings.

Other Factors Associated with Cannabis Use for Pain Management
It is beyond the scope of this article to discuss all of the factors associated with cannabis use for pain management. However, some of these factors (Table 2) are worth mentioning and should be addressed by future studies: residing in a remote region (associated with decreased use) and using nonpharmacological treatments for pain management (associated with increased use).

Strengths and Limitations
The present study has a number of important strengths, such as the use of a large community sample of adults living with CP, which enhances statistical power of the analyses, and the inclusion of a broad set of variables relevant to the CP experience. Some limitations must nonetheless be highlighted. First, participants were selfselected and all data were self-reported. Although cannabis use has become more socially acceptable, it is possible that some participants did not disclose their use due to social desirability concerns. This may be especially true in older age groups. It is thus possible that prevalence of cannabis use among persons living with CP is even higher that that estimated in our study. However, this limitation is not unique to the present prevalence study and thus does not undermine its ability to suggest an increase in self-reported prevalence of cannabis use for recreational and/or pain management reasons. Second, the COPE database was not specifically designed for cannabis research. With the available data, it was thus not possible to differentiate between medical cannabis and recreational cannabis users or type of product, mode of administration, and directions for use. Such differentiation would have been interesting, and we acknowledge that prevalence of use could vary according to such cannabis product characteristics. However, this does not affect the relevance of our recommendations, which are applicable to cannabis regardless of the channel through which it is obtained. Third, we found no gender identity differences in the proportion of cannabis users, suggesting valid prevalence estimates. Finally, data were cross-sectional, which limits our ability to make causal assumptions, detect changes in cannabis use in this specific population, or identify predictors of increased/decreased use. Future longitudinal research would be important to study trajectories of cannabis use over time.

Conclusions
The present study suggests a higher prevalence of cannabis use for pain management among people living with CP since its legalization for recreational purposes in 2018. Our results also indicate a higher prevalence of cannabis use for pain management among younger generations, whereas people aged 74 years or older reported significantly lower use. Cannabis is thus a common treatment reported in people living with CP, and its legalization in Canada has created a situation where it can be used for self-medication without having gone through rigorous drug approval processes and without proper guidance from patients' health care teams. Our study re-emphasizes the importance of rapidly generating evidence on the safety and effectiveness of cannabis, in addition to age-tailored education and awareness efforts among people living with CP.

Disclosure statement
Marimée Godbout-Parent holds a Canadian Institutes of Health Research (CIHR) master's degree scholarship. Adriana Angarita-Fonseca holds an FRQS postdoctoral scholarship. M. Gabrielle Pagé has a Junior 1 research scholarship from the FRQS and received honoraria from Canopy Growth for work unrelated to this study. Lucie Blais received research grants from AstraZeneca, TEVA, and Genentech, as well as consultation fees from AstraZeneca, TEVA, and Genentech for projects unrelated to this study. Anaïs Lacasse holds a Junior 2 research scholarship from the FRQS in partnership with the Quebec SUPPORT Unit (Support for People and Patient-Oriented Research and Trials). She leads the Chronic Pain Epidemiology Laboratory, which is funded by the Fondation de l'Université du Québec en Abitibi-Témiscamingue (FUQAT), in partnership with local businesses: the Pharmacie Jean-Coutu de Rouyn-Noranda (community pharmacy) and Glencore Fonderie Horne (copper smelter). All other authors have no conflicts of interest to report.

Funding
This work was supported by the Quebec Network on Drug Research and the Quebec Pain Research Network, two thematic networks of the Fonds de recherche du Québec-Santé (FRQS).

Data availability statement
The data supporting the findings of this study are available from the corresponding author, A. Lacasse, upon reasonable request and conditional to proper ethics approval for a secondary data analysis. The data are not publicly available because participants did not provide consent to open data.