Medical Cannabis Use among Adults in the Southeastern United States.

This was a one-time, cross-sectional, anonymous survey conducted on Amazon’s Mechanical Turk (MTurk). MTurk is a crowdsourcing marketplace that has been a successful recruitment platform in past studies of substance use ( Huhn, Tompkins, & Dunn, 2017 ; Peters, Rosenberry, Schauer, O’Grady, & Johnson, 2017 ; Strickland & Stoops, 2015 ). To recruit participants, an MTurk advertisement was posted and specifically asked if MTurk workers had ever used marijuana and other drugs (so to not disclose inclusion criteria). MTurk workers who were 18 years or older, had used cannabis on 20 of the past 30 days, and resided in South Atlantic region of the US were eligible to participate. As defined by the US census, states in the South Atlantic region include: Alabama, Delaware, District of Columbia, Florida, Georgia, Kentucky, Maryland, Mississippi, North Carolina, South Carolina, Tennessee, Virginia, and West Virginia. At the time of survey administration (June through August 2017), the District of Columbia and Delaware had legalized recreational and medical cannabis, respectively. All other states either had medical cannabis legislation that had passed, but was not operational, was limited to a few specific medical conditions, or was CBD-specific. Interested participants completed an eligibility screener through the web-based Research Electronic Data Capture (REDCap) ( Harris et al., 2009 ). Participants who were eligible for the study were linked immediately to the full survey, which took approximately 20–30 minutes to complete. Participants were compensated $1.00 for survey completion and a $1.50 bonus if they answered attention check questions correctly. This compensation amount is appropriate for MTurk studies and is similar to another study that successfully recruited heavy cannabis users from this platform ( Peters et al., 2017 ). Of the workers who saw the advertisement, 10,175 workers completed the eligibility screener, and 472 (4.6%) were eligible. Among those responding to the eligibility survey, approximately 71% resided outside of the southeastern US and were not eligible. Additionally, 93% did not meet the eligibility criteria for heaviness of cannabis use (20 out of the past 30 days). Of the 472 participants who were deemed eligible, 50 (11%) participants were excluded due to incomplete responses or having incongruent answers between the screener and full survey. The final sample included 432 adult cannabis users. The vast majority of survey participants came from states without wide-spread medical cannabis access (428 out of 432). All study procedures were approved by the institutional review board at the author’s academic institution. Demographic characteristics collected included age, gender, race, residing state, ethnicity, marital status, employment status, current annual household income, and level of education. Participants were asked about their cannabis use, in terms of frequency, amount, different preparations of use, and cannabis use history. Participants completed the Cannabis Use Disorder Identification Test – Revised (CUDIT-R) ( Adamson et al., 2010 ). To assess the reasons for cannabis use, participants were asked if they used the drug for medical reasons (i.e., to treat or decrease symptom(s) of a health condition), recreational/non-medical reasons (i.e., getting pleasure or satisfaction, to fit in with a group, or just for fun), or both medical and recreational reasons. If cannabis was endorsed for medical reasons, participants were asked to self-report the reason(s) for their use (open-ended text field). Participants had the ability to list as many medical conditions as they choose. The open-ended answers were categorized into general systems terms. When responses did not include sufficient detail, they were categorized in an “Other” category or in a “General” category within the broader system. Participants were also asked if they had a referral card for their medical cannabis use. Tobacco co-use was also assessed. Current tobacco users were asked about frequency and amount, history with tobacco, and completed the Fagerström Test for Nicotine Dependence ( Heatherton, Kozlowski, Frecker, & Fagerstrom, 1991 ). Aside from the assessment of cannabis use disorder and nicotine dependence, assessments used in this study were either locally developed or came from another MTurk survey study focused on cannabis and tobacco co-use ( Peters et al., 2017 ). Sample descriptive statistics were calculated and separated by 1) medical only, 2) recreational or medical and 3) recreational only cannabis use status. Chi-square analysis and one-way analysis of variance (ANOVA) were conducted to analyze differences between recreational only users, medical only users, and medical and recreational users for several demographic, cannabis, and tobacco use variables. When p <0.05, post-hoc comparisons were conducted among the three groups. The Bonferroni procedure (correction) was conducted for continuous variables given multiple comparisons. For ordinal/nominal variables, column proportions were compared between groups and z-scores were calculated, while adjusting p-values using the Bonferroni method. All analyses were conducted using IBM SPSS Statistics Version 22.0 and an alpha level of p <0.05 was used for all statistical analyses.

Demographic information for the overall sample is shown in Table 1 and is also separated by self-reported medical, recreational, or combined (medical and recreational) cannabis use. Among all study participants (N=432), 60.2% were female, 73.8% identified as Caucasian/White, and 16.4% identified as African American/Black. The mean age was 32.9 ( SD =10.2; Range=18–76). There was sizeable representation from Florida ( n =137; 31.7%), Georgia ( n =63; 14.6%), and North Carolina ( n =50; 11.6%). All other states were represented with a range of 2 to 35 participants. The majority of participants (n= 249; 57.6%) lived in states with CBD-only legislation, while 181 (41.9%) lived in states that legalized medical cannabis (though dispensaries with either not operational or medical use was limited), and two participants (0.5%) lived in DC. Given low numbers of participants from DC and Delaware ( n= 4), those participants were retained in this analysis. Fifty participants (11.6%) reported using cannabis in the past 30 days for medical reasons only, 180 (41.7%) reported recreational/non-medical reasons for use, and 202 (46.8%) reported using cannabis for both medical and recreational reasons. The only statistically significant demographic difference between the groups was on employment ( p =0.044), in which higher numbers of medical only cannabis users endorsed “other” employment compared to other groups. The other category included; student, retired, stay at home parent, or unable to work and the difference is likely driven by those in the medical group who are unable to work (8% compared to 0.6% and 2% in the other groups). Cannabis and tobacco use characteristics for the entire sample and separated by medical vs. recreational use are shown in Table 2 . Overall, participants used cannabis an average of 27.3 ( SD= 3.6) out of the last 30 days and were using an average of 4.1 ( SD= 3.5) times per day. The primary method of cannabis use was to smoke it, though 5.6% reported vaporizer use primarily, and 2.8% reported using dabs (concentrated THC preparation). Statistically significant differences between groups are shown in Table 2 . Generally, the medical only group appeared to vary in several ways from the recreational only group and the medical and recreational group. Groups varied in terms of age of first cannabis use ( p =0.004) and age of regular use ( p =0.010) with the medical only group trying cannabis and initiating regular use at an older age than the other groups. The medical cannabis group had a higher proportion of participants knowing the percentage of THC in their last cannabis purchase and appeared to purchase higher potency cannabis ( p =0.023). A greater proportion of medical users reported using a vaporizer as their primary method of cannabis use, compared to smoking cannabis ( p <0.001). Grams of cannabis used per pipe/bowl (most common method of use) was higher among medical users ( p =0.013). CUDIT-R scores tended to be lower among medical only users ( p =0.046). Tobacco use status was statistically different between groups ( p =0.002), with less current daily smoking occurring in the medical cannabis group, while no other tobacco characteristics were statistically different. The number and type of medical conditions endorsed by participants who reported any medical cannabis use to treat or reduce symptoms is shown in Table 3 (by general terms). Among 252 participants (out of 432; 58%) endorsing cannabis for medical use, over 60 unique medical conditions were listed. The mean number of medical conditions listed among participants was 1.8 ( SD =1.0; Median=1; Range=1–6) with a total of 445 conditions listed. The most commonly reported reason for medical cannabis use was for a psychiatric symptom or disorder (67.9%), followed by musculoskeletal and connective tissue disorders (29.4%), and gastrointestinal disorders (9.9%). A detailed list of specific conditions and their frequency of endorsement by participants is shown in Appendix Table A . The most commonly cited psychiatric disorders listed were anxiety (or anxiety-related conditions) and depression. Other common reasons for medical use included insomnia, post-traumatic stress disorder, pain, and headaches/migraine.

The results of this study found that within a region of the US where medical cannabis is not widely accessible, medical cannabis use is still occurring in a substantial portion of cannabis users. Of our sample, 11.6% of heavy cannabis users were using cannabis medically, and 46.8% endorsed using it for both medical and recreational reasons. Despite federal and state laws and sufficient scientific evidence to support the use of medical cannabis for certain conditions, some continue to use cannabis medically for a range of medical issues to alleviate their symptoms. While some participants have most likely attempted to treat their conditions with other strategies or under a physician’s care, it is likely that some in this group have not sought medical care for their condition and may not experience long-term benefit from their medical cannabis use. This is an important area of future research to determine if medical cannabis users are generally seeking other treatment options first or if they are not utilizing healthcare services in favor of cannabis use. Self-medicating with cannabis is an important topic for future research and a recent study found differences among those who are licensed vs. unlicensed users of medical cannabis ( Sznitman, 2017 ). The self-reported reasons for medical cannabis use were largely due to psychiatric conditions or symptoms, with anxiety being the most common condition listed by participants. This is consistent with past literature finding similar reasons for medical use among dispensary consumers ( Bonn-Miller, Babson, et al., 2014 ; Bonn-Miller, Boden, et al., 2014 ). Another study showed that medical cannabis users are more likely to report psychiatric comorbidity compared to recreational cannabis users ( Martins et al., 2015 ). The range of other conditions listed in this study is of interest, especially given that many of these have insufficient or no evidence to support the use of medical cannabis as a treatment strategy. The results of the current study found few demographic differences between medical and recreational users, which is consistent with previous research ( Lin et al., 2016 ), though our study also found few differences regarding current cannabis frequency between groups of users. Medical cannabis users did tend to differ from the recreational group and the medical/recreational group in terms of later age of cannabis use, more knowledge about THC content in recent purchases, more vaporizer use as their primary method, lower reported problematic cannabis use as determined by the CUDIT-R, and lower rates of current and ever tobacco use. This study had several limitations. First, data were self-reported, so validity may have been compromised if participants over or under estimated cannabis use, had difficulty with retrospection, or otherwise were poor reporters. Those using for medical reasons may have wanted to avoid any indication of dependence or problems associated with use, which could have affected responses. Also, there is no way of confirming diagnosis of the medical conditions listed by participants, which may not have been formal diagnoses. Second, the sample used in this study was majority female (60.2%). Rates of cannabis use tend to be higher among men in the US ( Substance Abuse and Mental Health Services Administration, 2014 ). Having more representation among female cannabis users may be a limitation, but when compared to previous cannabis studies, it contributes to a better understanding of traditionally underrepresented participants. Overall, despite similar frequency of cannabis use, medical only cannabis users, at least per their own report, appeared to be a lower risk group since they reported fewer dependence symptoms, were less susceptible to harms from smoked cannabis, started using cannabis at a later age, and did not engage as frequently in co-occurring tobacco use. In fact, avoiding combustible cannabis and the use of cannabis in adolescence are two recommended ways to reduce risk associated with cannabis use ( Fischer et al., 2017 ). In contrast, medical only users tended to report use of higher THC products (when the percentage of THC purchased was known), which is thought to be a riskier practice ( Fischer et al., 2017 ). As cannabis use increases, particularly medical cannabis use, it will become essential that our measures of problematic and/or hazardous use accurately distinguish between problematic, non-problematic, and therapeutic use ( Sznitman & Room, 2018 ). For example, it may be difficult to disentangle symptoms of withdrawal from a return to an impaired pre-cannabis functioning state. Further, treatments for cannabis use disorder will need to take into account an individual’s real and perceived benefits of cannabis. If participants perceive medical benefit from their cannabis use, it is unlikely that they will be motivated to quit or even reduce. Instead, it may be beneficial to implement the aforementioned harm reduction efforts with these individuals through information delivered via healthcare providers or dispensary staff. The issue of medical cannabis is clearly complicated and requires a great deal of further research to provide clarity on health benefits, harms, and education for healthcare providers, patients, and legislators.

Cannabis is a commonly used drug in the United States (US) ( Substance Abuse and Mental Health Services Administration, 2017 ), with legality status that is rapidly changing at a state-by-state level. Past-month cannabis use continues to rise in the US, especially among adults aged 26 and older ( Substance Abuse and Mental Health Services Administration, 2017 ). A recent nationally representative study showed that the risk of cannabis use and cannabis use disorders has increased, especially in states where medical cannabis laws have been passed ( Hasin et al., 2017 ). The use of cannabis to treat medical issues and symptomology is also increasing in the US ( Hill, 2015 ) and globally ( Fitzcharles & Eisenberg, 2018 ), and the rates at which people are using medical cannabis have increased in states where medical cannabis laws have been passed ( Cerda, Wall, Keyes, Galea, & Hasin, 2012 ). In California alone, which legalized medical cannabis in 1996, there were two million individuals in 2012 that were using cannabis medically ( Ryan-Ibarra, Induni, & Ewing, 2015 ). There are also high rates of individuals self-medicating in locations without legalized medical cannabis ( Park & Wu, 2017 ). As of February 2018, 29 states in the US, including the District of Columbia, permit the medical use of cannabis, and another 17 states permit limited access of low Δ9-tetrahydrocannnabinol (THC) and high-cannabidiol (CBD) products ( National Conference of State Legislatures, 2018 ). In the southeastern US specifically, most of the legislation passed focuses on CBD products only and among states with passed legislation, many do not have operational systems of dissemination in place. There is a lack of data on medical cannabis use in states without passed legislation and especially within the southeastern US, which is a region where movement towards legal recreational and medical cannabis has not moved as quickly as in the western and northeastern US. There is an increasing body of research on the potential medical benefits of THC/CBD products. Synthetic cannabinoid drugs (dronabinol and nabilone) have been approved in the US to treat symptoms associated with chemotherapy-induced nausea and vomiting ( Todaro, 2012 ) and are frequently used in patients with HIV to prevent weight loss and increase appetite ( Badowski & Perez, 2016 ). A recent summary of the evidence supporting the use of cannabis to treat medical disorders concluded that its use may be supported for certain conditions (i.e., chronic pain for adults, nausea and vomiting for chemotherapy patients, and for treatment of multiple sclerosis spasticity), but there is insufficient evidence to support its use for other conditions at this time ( National Academies of Sciences, 2017 ). Despite the lack of scientific evidence supporting the use of THC or CBD for certain conditions, many are using cannabis to treat a variety of medical symptomology. Cannabinoids have been proposed as a means to manage chronic pain and alleviate reliance on prescription opioids;( Choo, Feldstein Ewing, & Lovejoy, 2016 ; Lucas & Walsh, 2017 ; Sohler et al., 2018 ) however, double-blind, randomized clinical trials testing the efficacy of cannabinoid usage on reducing chronic pain have not been completed. Muscle-related pain and arthritis are frequently endorsed conditions that medical cannabis are used to treat ( Park & Wu, 2017 ; Ryan-Ibarra et al., 2015 ; Ste-Marie et al., 2016 ), and cannabis users have self-reported that cannabis use is helpful in alleviating their psychological symptoms and social anxiety ( Bonn-Miller, Babson, & Vandrey, 2014 ; Bonn-Miller, Boden, Bucossi, & Babson, 2014 ). However, there remains insufficient research on the short- and long-term health effects of THC and CBD products, especially in light of novel products and preparations that have higher percentages of THC. In states with legal medical cannabis, 17% of cannabis users reported cannabis use for medical reasons, as recommended by a physician ( Lin, Ilgen, Jannausch, & Bohnert, 2016 ). Though many patients experience benefits from medical cannabis use, they may be using it to treat conditions in an unapproved manner, without the recommendation of a physician, or in place of other, potentially efficacious treatment strategies. Even in states with medical cannabis dispensaries, not all advice being given by trained dispensary staff is consistently evidence-based or medically guided ( Haug et al., 2016 ; Linares et al., 2016 ). In states where medical and recreational cannabis is not legal or is restricted, little is known about its medicinal uses, though data suggest that medical cannabis use is still prevalent regardless of legal status ( Compton, Han, Hughes, Jones, & Blanco, 2017 ). Understanding medical cannabis use, especially in states where medical use is restricted and patients are likely self-medicating, is particularly necessary. This information may help inform physicians as to the conditions and symptoms underlying medical cannabis use and could lead to tailored education in states with limited medical cannabis or CBD availability. In addition to understanding the reasons for medical cannabis use in states without wide-spread medical cannabis availability, it’s also important to identify meaningful differences among those using cannabis for medical versus recreational reasons. Previous studies have found that those who used cannabis medically were more likely to report poorer health, more frequent daily use, lower levels of other drug use ( Lin et al., 2016 ; Roy-Byrne et al., 2015 ), and less risky drug use ( Richmond et al., 2015 ). This cross-sectional survey aimed to examine the reasons for medical cannabis use among thirteen Southeastern US states. This region of the US has more restrictive medical cannabis legislation and limited distribution compared to other regions of the US, making it a generally understudied area for medical cannabis use. This study had two aims: 1) compare differences in demographics and cannabis use characteristics in recreational and medical cannabis users, and 2) assess the reasons for medical cannabis use.