The use of non-medical grade cannabis for medical purposes is a common phenomenon but knowledge of this patient population is limited. This study aimed to investigate sociodemographic and user practice differences between Finnish participants who use cannabis mainly recreationally versus mainly medically. The pairwise comparisons indicated that mainly medical users are less educated, more likely not employed, more likely to use edible forms of cannabis, and estimate their cannabis products to have higher THC content than mainly recreational users (Table 1). Moreover, medical users were slightly older and used cannabis more often than recreational users (Table 2). The logistic regression analyses mainly corroborated these findings, except for THC content and employment status which became non-significant.

In the base model, female and other gender, older age, lower education level, and higher monthly use frequency predicted mainly medical cannabis use. These findings are in line with our previous study using a similar approach,¹⁵ except for city size and earlier onset age of initiating cannabis use, which were significant predictors in the previous study but not in the current one. This discrepancy could be due to different samples, as the previous study was aimed for more younger users as part of a national harm reduction project that developed a mini-intervention model and tools for professionals to discuss cannabis use with young people. The significant variables of the base model in this study remained significant in the extended model, indicating a stable association across models. The addition of variables in the extended model accounted for an additional 5% of the variance, increasing the total explained variance to 12%. Among the five new variables included in the extended model (i.e., employment status, severity of dependence, self-estimated THC and CBD content and the preferred type of cannabis usage), only the type of usage predicted medical cannabis use.

Our results are in line with previous studies which have also found that people who use and grow cannabis for medical motives are usually older, and female compared to recreational users.^{17,34,36,46,47} A possible explanation for this might be that women are more open to alternative treatments than men^48,49 and that in general recreational cannabis use is more prevalent among men.¹³ According to Finnish population studies, males have used cannabis three times more prevalently compared to females in the past year and in the past month with lifetime prevalence for males in 2022 being 36% and for females 21%.¹³ In relation to age, there is a higher risk for experiencing chronic pain and diseases as people get older⁵⁰ which could explain why medical users are on average older than recreational users. There are also several cannabis-specific factors in relation to age, like general health, risk of injuries and other medication use⁵¹ and relation to biological sex that show differences in physiology, use practices and experienced effects of cannabis use,⁵² but research findings on these factors are often inconsistent and require further inquiry.

The frequency of use is often found to be higher among medically motivated cannabis users,^{17,34,40,47,53} which was also found in our study even though there were no differences in the amount of cannabis used per session in the mainly medically motivated and mainly recreationally motivated groups. Most commonly reported amount across both groups was between 0.1 and 0.5 g, so our findings suggest similar usage patterns regarding the self-estimated quantity of cannabis consumed per session among both recreational and medical users (Table 1). The higher frequency of use among medically motivated users is understandable in the context of treating one’s perhaps chronic diseases. However, the medically motivated users did not on average use cannabis daily, perhaps taking breaks from their use to avoid building up tolerance for their self-medication. Future studies should look more closely at whether there are differences in terms of timing of use.⁴⁷ It can be speculated that, for example, recreational users concentrate their use more for the weekends, while medical users might take weekends off from their use to lower their tolerance levels and experience more noticeable medical effects during the week to function better. A recent study based on Finnish national survey data partly confirms this as it found that medically motivated users distributed their use more evenly throughout the week compared to recreational users who mostly consumed on the evenings and weekends.³⁸ Other differences were also found, including medical users using more often alone at home compared to recreational users, and recreational cannabis use occasions involved more often simultaneous alcohol use.³⁸

While in our study there was a statistically significant difference between the groups in relation to employment status in the pairwise comparison (Table 1), this was no longer significant in the logistic regression model. This implies that there were other variables that are more important predictors for medical cannabis use than employment status, such as education level (lower education level predicted medical cannabis use). Other studies have not found statistically significant differences in education levels between recreational and medical users,^34,41,46,47 which again could be due methodological differences, country-contexts, and populations. Based on national survey data, highly educated men have somewhat more prevalent past year cannabis use in Finland compared to middle and low education levels (9.1% vs. 6.8% and 6.5% respectively), but among women there are basically no educational differences.⁵⁴ In general, education level has not been found to be associated with cannabis experimentation and use when other social background factors have been taken into account.¹² Lower education level among medically motivated cannabis users in our sample could indicate for instance challenges in attaining desired education due to illness, or lower access to official health care services which could lead to increased self-medication with cannabis. However, these possibilities require more research.

To address the potential correlation between education and employment, we conducted additional analyses. Although education and employment were correlated, the correlation was weak to medium, suggesting that education is an independent predictor in our results.

Overall, pre-pandemic self-reported health was poorer among groups with lower education compared to groups with middle and higher education both in Finland⁵⁵ and elsewhere in Europe⁵⁶ but whether the association between lower education level and medical motive to use cannabis is due to poorer health status is speculative at this point. Besides education level there are several other factors involved when modeling the relation between socioeconomic status and health⁵⁷ and how these relate to motivations around cannabis use is still unclear.

In addition to the statistically significant differences between the two groups, our study shows that factors like onset age of cannabis use, size of the city of residence, employment status, amount of use per session, severity of dependence, and self-estimated amounts of THC and CBD concentrations showed no statistically significant differences. Some of the associations observed in the pairwise comparisons (e.g., self-estimated THC content) were not significant in the logistic regression analysis, suggesting that higher THC content preferred by medical users in pairwise comparison is explained by the other factors, for instance the need for higher doses to treat their medical condition.³⁹ Again, some of these associations, or their lack of, are found in other studies, while others are not. For instance, similar to the findings by Turna et al.,³⁴ mainly medical users in our study reported using higher THC-containing products compared to recreational users. The high THC/low CBD concentration in Turna et al. was more for medical users who also endorsed recreational use. While higher THC concentrations have been associated with increases in adverse events for medical users and dependence and mental health problems for recreational users,³⁹ medical users using higher THC products could also point to a therapeutic value of the psychoactive effects of cannabis, even though they are often considered side effects.^14,58 However, there is no clear scientific consensus on what constitutes a high or low amount of THC for instance, even though attempts to set up a “standard unit” have been proposed.⁵⁹ It is even more difficult to evaluate cannabinoid levels based on self-estimates of illegal products, as in the case of our study where over a quarter of the participants did not know the THC level of their most used cannabinoid product.

Regarding cannabis dependence, we did not find statistically significant difference between the groups on SDS scores. However, in the cross-country grower study by Snitzman et al.⁴² recreational only growers scored higher on SDS levels compared to the medical groups, with and without recreational motive. Sznitman⁴⁷ also found licensed medical cannabis users scoring lower on The Cannabis Abuse Screening Test (CAST) compared to unlicensed medical users and recreational users. However, there are also studies showing medical users, especially those also endorsing recreational use, scoring higher on the Cannabis Use Disorder Identification Test – Revised (CUDIT-R) compared to recreational users and medical only users.^34,46 As these different screening tests measure slightly different things it is difficult to draw direct conclusions from them, especially for people with medical motives as most current measurement tools are designed to screen problematic use.⁶⁰ Overall, the dependence scores in the present study were relatively low in both groups, medical users scoring 1.69 while recreational users scored 1.94 on the scale from 0 to 12 (Table 1). All the groups in the study by Sznitman et al.⁴² had higher SDS scores compared to ours, with a total mean of 5.22 (SD = 1.77). This difference could partly be due to differences in samples as the study by Sznitman focused on comparing cannabis growers in 18 countries, but could also relate to the validity of screening tools which may not work equally well for different motivations of use.^42,61 The issue of dependence is complex, especially regarding medical cannabis users who might depend on cannabis to treat their chronic disease, which does not equal problematic use or addiction.^39,60

These empirical results are important to consider also in the wider context of cannabis use, both medical and recreational, which have long and complicated histories⁶² and sociology⁶³ which vary depending on the local country-contexts. For instance, by the end of 2023 in the United States, 38 states, three territories, and the District of Columbia had allowed the medical use of cannabis products and 24 states had passed legislation to allow recreational adult use.⁶⁴ Longitudinal studies looking at pre- and post-legalization of cannabis in, for example, California has shown that legislative changes have some impact on young adults who reported using cannabis for medical reasons prior to recreational laws and some of that association included transitioning from medical use to recreational use but also declined cannabis and other drug use after legalisation.⁶⁵ In the USA between 1976 and 2016 reasons for cannabis use among high school students in general has changed from more social/recreational reasons toward coping with negative affects,⁶⁶ perhaps indicating the rapidly changing cultural environment young people are experiencing. While state policies have an impact on for instance where people source their cannabis (legal store vs. illegal dealer),⁶⁴ and potentially on the prevalence of self-reported use of cannabis for medical purposes,⁶⁷ the full extent of public health impacts of cannabis policy reforms are challenging to evaluate in the United States context as it remains federally illegal.⁶⁴

On the European level, these types of monitoring studies on cannabis use motivations are lacking as, for example, the European School Survey Project on Alcohol and Other Drugs (ESPAD) is focused on prevalence of use and perceived availability among adolescents. The European Web Survey on Drugs (EWSD) by the European Union Drugs Agency (EUDA, formerly known as European Monitoring Centre for Drugs and Drug Addiction, EMCDDA) is also more focused on reporting patterns of drug use and sources of supply than motivations.⁶⁸ As an increasing number of European countries are reforming their cannabis policies, longitudinal monitoring of the public health impact these policy changes have across the EU on different user groups is needed,²³ similar to the United States.⁶⁴ Also due to complex pharmacology of cannabis and cannabinoids^69,70 and the variety of products available⁶⁴ future research on cannabinoid-based medicines and use in general can benefit from real-world evidence^2,71 and from naturalistic studies with ecological validity to look at factors like reported motivations and experienced effects among different groups, alterations in consciousness, environment or setting of use and even placebo effects.^39,72,73

While people continue to use cannabis for different reasons experiencing both benefits and harms, it is important to acknowledge that globally cannabis and cannabinoids are primarily used outside of official clinical contexts. Therefore, it is vital to investigate various factors involved to guide appropriate harm reduction and benefit increasement practices also outside of clinical use and to assist in designing future studies. Global and local cannabis policy reforms, with either medical or recreational motives, would also benefit from real-world evidence and from considering politicogenic drug effects, i.e., effects originating from modern drug policing.²⁹ Current legislation in Finland for instance considers self-medicative use of cannabis as a criminal offense while there are notable challenges to access official medical-grade products via prescription despite it being legal for over 15 years. Whether medical intention to use cannabis is enough to make the practice medical is open for debate⁷³ but taking into account the perspectives of people who have lived experiences using cannabis and cannabinoids for a variety of purposes is vital in this regard.

It is also important to note that even under recreational motivation there are several motivating factors like enjoyment, conformity, coping, experimentation, boredom, alcohol-related use, celebration, altered perception, social anxiety, relative low risk, sleep, and availability,⁷⁴ which even among medical users are almost all positively inter-correlated with one another and affect, for instance, frequency of use.⁷⁵ In addition, medical users have a plethora of indication-specific motivations to use cannabis and cannabinoids in personalized ways that offer the best symptom relief. Further studies are needed to include different motivational factors when studying motives of cannabis use and perhaps to go beyond the medical versus recreational distinction toward improvement as in the therapy versus enhancement debate.⁷⁶ Studying how exogenous cannabinoids enter and affect the extended endocannabinoid system⁷⁷ and understanding the interactions that follow has come a long way since the isolation of 9-delta-THC over 60 years ago⁷⁸ but there remains a lot we do not yet know.