Nationwide Study of Postlegalization Marijuana Use Among Patients With Epilepsy in Canada

Methods

Results

We obtained 395 responses from 9 provinces and 1 territory of Canada, with most of them from Ontario (n = 222; 78.20%) and Alberta (n = 32; 8.51%). A proportion (n = 296; 74.92%) reported having ever used marijuana, which was the subgroup used for further analysis. As summarized in Table 1, most of the participants completed the survey themselves (n = 261; 88.18%). Surveys were completed in full by 256 respondents (80.50%). Female individuals accounted for 70.41% (n = 207), and the median age was 32 years (IQR = 26–41). Participants who experienced seizures for more than 10 years accounted for 50.68% (n = 148) of respondents. Most of them (n = 165; 56.89%) reported having seizures at least monthly over the past year. Generalized seizure was the most common type reported by patients (n = 169; 57.10%).

Three or more ASMs had been tried by 52.03% (n = 154) of respondents. The median number of ASMs used was 3 (IQR = 1–5). The number of participants who had tried 3 or more ASMs was associated with an increasing duration since diagnosis of epilepsy (p < 0.001). Those patients were also more likely to have started using marijuana due to drug-resistant epilepsy (DRE) (p < 0.001) and were more likely to have experienced side effects from ASMs (p = 0.002). Side effects of ASM use were reported by 76.27% (n = 180), with the most common side effects overall being impaired thinking (n = 40; 17.17%), anxiety (n = 37; 15.74%), and altered hunger (n = 36; 15.32%). Different therapeutic strategies were tried in addition to ASMs including vagus nerve stimulation (VNS) in 8.11% (n = 24), ketogenic diet in 15.87% (n = 47), and resective epilepsy surgery in 13.18% (n = 39). Participants who had tried these alternative treatments were more likely to start using marijuana for DRE (p < 0.001). The participants who had tried alternative treatments were also more likely to experience side effects from ASMs (p = 0.027). The most common mental health comorbidity seen in this population was anxiety in 41.90% (n = 124), followed by depression in 40.54% (n = 120) (Table 2).

As seen in Figure 2, the participants found marijuana to be “somewhat effective” overall for helping their seizures (median 58.00; IQR 31.75–82.00), reducing the number of seizures (median 50.00; IQR 14.00–80.00), the severity of seizures (median 50.00; IQR 4.25–76.00), and ASM side effects (median 50.00; IQR 23.00–84.00). Most of the participants found that marijuana had a positive impact on sleep (median 80.00; IQR 62.75–96.00) and stress levels (median 78.00; IQR 67.00–90.00), as seen in Figure 3. There was no difference in the perceived effectiveness of marijuana between groups that had tried 3 or more ASMs or alternative treatments (p = 0.940 and p = 0.588, respectively).

Marijuana use in the past 12 months was reported by 90.95% (n = 221), and use at least once per month was reported by 83.68% of patients (n = 200). First marijuana use over 10 years ago was stated by 45.53% (n = 112). Most of them (55.42%; n = 133) used marijuana before their first seizure. Most participants (n = 152; 51.35%) started using marijuana for recreational purposes (Figure 4A), though 47.54% (n = 116) used marijuana to treat their epilepsy. A significant proportion of the participants used marijuana additionally to manage anxiety and depression (n = 134; 45.27%), sleep disturbances (n = 125; 42.23%), and pain (n = 85; 28.72%). Most of them (n = 199; 93.87%) had not changed the dose of their ASM because of marijuana use.

The most widely used method of obtaining marijuana was at a local dispensary without a prescription (44.93%; n = 133. Figure 4B). Smoking was both the most preferred method (n = 83; 34.73%) and the most tried method (n = 193; 65.20%) of marijuana consumption (Figure 4C). Use at least once daily was reported by 70.29% (n = 168), and most use occurred in the evening (n = 153; 51.69%) or before bed (n = 144; 48.65%). The median amount of marijuana used per week was 5.00 g (IQR = 1–10). Top preferences for marijuana were high cannabidiol (CBD) n = 106; 35.81%), high delta-9-tetrahydrocannabinol (THC) (n = 104; 35.14%), and combined CBD and THC (n = 99; 33.45%). A favorite brand or producer of marijuana was denied by 77.25% (n = 180). More than CAD$200 per month on marijuana (not covered by insurance) was spent by 19.13% (n = 44).

Using a logistic regression analysis, tobacco, alcohol, and street drug use were found to be predictive of marijuana use (OR: 10.8; 95% CI 4.55–25.71; p < 0.001; OR: 3.38; 95% CI 2.10–5.54; p < 0.001; OR: 54.34; 95% CI 7.44–396.75; p < 0.001). However, no significant association was observed with sex, age, duration of epilepsy, employment, or frequency of seizures (OR: 1.00; 95% CI 0.99–1.02; p = 0.70; OR: 1.01; 95% CI 0.99–1.03; p = 0.41; OR: 0.98; 95% CI 0.76–1.3; p = 0.98; OR: 1.00; 95% CI 0.99–1.00; p = 0.52; OR: 1.13; 95% CI 0.92–1.34; p = 0.25). An increased usage of marijuana since legalization was reported by 50.67% (n = 150) with the most common reason being increased accessibility (n = 63; 21.28%), as shown in Figure 2D.

Discussion

This is the first survey to evaluate marijuana use in patients with epilepsy after legalization of the drug. We found a high prevalence of marijuana use among Canadian patients with epilepsy. Approximately 74.92% of the participants reported ever using marijuana with most of them (70.29%) reporting use at least once daily, though it is difficult to draw conclusions from these percentages based on selection bias of respondents. Before legalization, studies reported rates of patients with epilepsy using marijuana as 20%–36%.^7,8 However, it is previously well-established that marijuana use has become widespread and has been increasing in recent years, in both populations with epilepsy and the general public.^10,12 A remaining major limitation of any self-report study, especially the one broadly distributed through hospital centers, word of mouth, and epilepsy community organizations, is a selection bias for individuals interested in the effects of marijuana. Despite this, our data provide valuable insight into the demographic data of those who do use marijuana and how they are using marijuana as an adjuvant treatment for epilepsy.

Before legalization, the Canadian Tobacco, Alcohol and Drugs Survey in 2017 found that 15% of Canadians aged 15 years or older used marijuana in the previous 12 months.¹² After the legalization of marijuana in October 2018, Canadians reported an increased usage of marijuana; up to 25% of Canadians in 2019 used marijuana during the 12 months before.¹² A study on marijuana use in the Canadian epilepsy population before legalization found that 57% of the cohort had tried marijuana, whereas 36% had used marijuana in the previous 12 months.¹⁰ More than half of the users in our study reported that their usage increased after legalization. The participants cited increased accessibility, more control over types of marijuana, and increased safety as key reasons for increased usage postlegalization. A study by Massot-Tarrus et al. before legalization showed that male sex and age younger than 30 years were predictors of marijuana use. However, our results showed no significant association between marijuana use and either age or sex, with the median age of marijuana users who responded to our survey being 32 years of age.¹⁰ More than two-thirds of the overall survey response rate were females, which may be related to the well-known effect that more female individuals complete surveys compared with male individuals.¹³ Similar to Massot-Tarrus et al., tobacco use was found to be a predictor of marijuana use. We found that alcohol and street drug use were also predictive, and these variables were not shown to have an association in this previous study.¹⁰

Roughly half of the respondents endorsed current marijuana use because of their epilepsy. This is consistent with past studies that have shown that patients may use marijuana as an alternative treatment for epilepsy, particularly when epilepsy is drug resistant.¹⁰ However, this finding is also subject to social desirability bias because marijuana popularity has been growing in recent years. Patients are considered to experience DRE if they have trialed 2 ASMs at appropriate doses for an adequate time period without success in managing seizures. However, for our analysis, we considered patients to be drug resistant if they had trialed 3 or more ASMs. This was to increase the probability of this population being truly drug resistant, as opposed to those who may have discontinued an ASM because of other reasons, such as due to side effects or after an inadequate trial. In our study, 52% experienced DRE, which points to the use of marijuana in this population with more severe epilepsy. Previously, Epidiolex received US FDA approval after multiple randomized clinical trials showed a reduction in seizure frequency among patients with DRE syndromes (Dravet or Lennox-Gastaut).⁸ In addition, 84% of the participants in a study by Porter et al. reported a decrease in seizure frequency with marijuana use, while a systematic review by Stockings et al. found that an estimated 48.5% of patients achieved a >50% reduction in seizures with CBD being more likely to produce this outcome.^5,9 Although both are biologically active cannabinoids contained naturally in cannabis, CBD, unlike THC, does not have psychoactive effects.⁶ Similarly, we found that the participants who reported a previous trial of 3 or more ASMs were more likely to start using marijuana for their DRE. A large proportion of the participants (60.60%) also found marijuana to be “somewhat effective” (n = 72; 18.95%) to “very effective” (n = 65; 17.11%) in helping their seizures overall, particularly for reducing seizure frequency, although this may be an incorrect perception, and this cannot verified with other objective methods. In general, our study supports previous literature indicating that there may be a benefit of marijuana on seizure frequency, particularly in a drug-resistant population. However, further clinical trials are needed to corroborate our observational results.

Although only containing CBD, Epidiolex was approved by the US FDA for up to 20 mg/kg daily dose.¹⁴ The mean daily dose reported by the participants in the study by Massot-Tarrus et al. before legalization was 1 g.¹⁰ In this study, the median amount of marijuana consumed was 5 g per week (the average amount of marijuana in a rolled joint being 0.32 g), and most of the respondents (70.29%) consumed marijuana products at least once a day, indicating that many participants regularly use large quantities of marijuana.¹⁵ Porter et al. reported the side effects of cannabidiol-enriched marijuana to be drowsiness, fatigue, and decreased appetite, while the most common side effects of marijuana reported in our survey were impaired thinking, anxiety, and altered hunger.⁵ Conversely, many patients also reported using marijuana to manage comorbid anxiety and depression, and many marijuana users are known to seek impaired thinking and alterations in hunger as desired effects.⁶ This indicates that the adverse effect profile may differ not just due to differences in experienced side effects of marijuana but also based on its intended use. While 36.91% reported unwanted side effects from marijuana use, 76.27% of respondents reported side effects from ASMs, and a large proportion found that marijuana was at least somewhat effective in managing ASM side effects. Physicians must be aware, however, that a small subgroup will experience serious side effects such as abnormalities in brain development and psychosis.⁷

The top concerns surrounding marijuana use were cost, a lack of recommendation from a doctor, and a lack of information about marijuana, though most of the participants (76.99%) had discussed use with their doctor. This points to a need for doctors to use available evidence to counsel patients appropriately on marijuana use. For example, smoking was the most used method of consumption, indicating a need for harm reduction strategies. Despite our results indicating a decrease in stigma since legalization, patients continue to feel stigma surrounding their marijuana usage. Our results also confirmed results of past studies that have shown patients to be more comfortable using marijuana because it is considered a “natural product” though the importance of a substance being “natural” is debated.^16,17

There are several limitations of self-reported surveys. A major limitation of this survey is the inherent self-report bias and the difficulty of validating the results. It should be noted that this study was designed to elicit opinions and perceptions of the participants and not necessarily the true effects of marijuana. It is also not possible to verify that all participants have a clinical diagnosis of a seizure disorder. However, most of the participants (76.35%) report being referred to the study as an epilepsy inpatient, outpatient, or from an epilepsy organization, increasing confidence that these results are representative of a true population with epilepsy. Furthermore, if the structure of the study represents a potentially heterogeneous population, this is also more representative of a real-life population seen in a neurology clinic setting. The study also lacks a control group for comparison, and there is an inherent recall bias in self-report surveys. In addition, this study was limited to English-speaking Canadians, which account for roughly 75.5% of the Canadian population.¹⁸ Furthermore, because this study was performed in Canada, the ability to extrapolate the results to other countries such as the United States is limited.

In conclusion, our study presents novel results of a nationwide survey of patients with epilepsy that revealed a high prevalence of marijuana usage for epilepsy management after legalization, especially among those with drug-resistant epilepsy. This indicates a need for physicians to be aware of usage habits and prepared to counsel their patients accordingly.

Study Funding

The authors report no targeted funding.

Disclosure

The authors report no relevant disclosures. Full disclosure form information provided by the authors is available with the full text of this article at Neurology.org/cp.