Abstract
Introduction
The number of people using medical cannabis has increased in the United States, but little is known about the prevalence or risk factors for Cannabis Use Disorder (CUD) in this population. We sought to estimate the prevalence of CUD and to identify individual characteristics and cannabis use patterns associated with CUD in a statewide sample of adults certified to use medical cannabis.
Methods
The sample included 632 adults (median age 45, 63% female, 75% non-Hispanic White) certified and currently using medical cannabis in Florida. Participants were recruited from medical cannabis clinics and completed an online cross-sectional survey in 2022. The primary outcome, based on the 8-item Cannabis Use Disorder Identification Test-Revised (CUDIT-R), was categorized as possible CUD (CUDIT-R score ≥12), hazardous use (score ≥8), or non-hazardous use (<8). Independent variables included sociodemographics; mental health diagnoses; past-5-year substance use; cannabis use history, motives, and patterns; and concern about addiction or dependence. Bivariate analyses and multivariable logistic regression examined associations with possible CUD, with secondary analysis examining associations with any hazardous use.
Results
Overall, 20% (95% CI: 17%–24%) had possible CUD) and 54% (95% CI: 50%–58%) had hazardous use. Possible CUD was significantly associated with younger age (aOR 0.98, 95% CI: 0.96–1.0), >1 year of daily marijuana use (1–5 years aOR 2.2 95% CI: 1.2–5.2; 6+ years aOR 3.1 95% CI: 1.5–7.2), lower preference for oral concentrates (aOR 0.29, 95% CI: 0.10–0.70), and self-reported recreational reasons for marijuana use (aOR 5.1, 95% 1.5–17.3). Possible CUD was not associated with sex, race/ethnicity, mental health conditions, or preferred THC percentage. Hazardous use was associated with >1 year of daily marijuana use, recreational reasons for use, younger age at marijuana use, greater preference for marijuana flower, and anxiety diagnosis. Few (4%) reported concern about addiction of dependence.
Discussion
About 1 in 5 adults certified to use medical cannabis in Florida had possible CUD, and over half (54%) had hazardous use. Possible CUD was associated with younger age, prolonged daily use, lower preference for oral concentrates, and recreational motives for marijuana use.
Keywords: Medical cannabis, Cannabis use disorder, Cannabis use patterns, CUDIT-R, Risk factors
Introduction
Medical cannabis (MC) use has been on the rise in the United States; the number of registered patients who use cannabis medicinally increased nationally from 678,408 to 2,974,433 between 2016 and 2020 in 26 states and Washington, DC [1]. Florida has one of the largest MC programs in the USA [2], with 924,473 certified patients as of September 2025 [3]. With the rise of MC use, there is also concern about Cannabis Use Disorder (CUD), a condition in which cannabis use is associated with significant impairment in health or life activities. A clinical diagnosis of CUD is broadly defined by the Diagnostic Statistical Manual of Mental Disorders (DSM-5) as the presence of clinically significant impairment with at least 2 out of 11 specific criteria being manifested [4]. Because a formal CUD diagnosis typically requires a clinical evaluation or interview, screening tests are often used to estimate the prevalence of CUD in a population. One of the more commonly used screening tests is the Cannabis Use Disorder Identification Test – Revised (CUDIT-R), an 8-item measure that can create a score that is often categorized as probable CUD (score ≥12), or hazardous use (score ≥8) [5–7].
The prevalence of CUD among past-year cannabis users in the US, estimated from formal criteria, CUDIT-R scores, and meta-analyses, ranges from 21% to 31% (8–13). To our knowledge, no recent study included a sample of adults who were recruited from medical cannabis clinics, and who could be using cannabis exclusively for medical indications or a combination of medical and recreational intentions. At least one study found that persons using cannabis for both medicinal and recreational reasons had a higher rate of CUD in comparison to those utilizing cannabis for strictly recreational uses [8].
Several previous studies have identified possible factors associated with CUD. These include sociodemographic characteristics (younger age, male gender, lower educational attainment, unstable housing, lower income), use of other substances (e.g., smoking, vaping, nicotine, use of other drugs), and co-morbid mental health conditions (e.g., depression, anxiety, post-traumatic stress disorder [PTSD]) [8–11]. CUD has also been associated with different cannabis use patterns, including history of cannabis use (e.g., age of onset [11, 12], duration of cannabis use [10], and motivation for use [medical, recreational or both]) [8, 11]. Availability of medical cannabis dispensaries also provides access to a growing diversity of cannabis product types and to cannabis flower with higher overall tetrahydrocannabinol (THC) content, but much less is known about the associations of preferred products or percent THC with the prevalence of CUD.
Because research investigating the factors associated with CUD among individuals registered in a MC program is limited, the objectives of the current study were: (1) to estimate the prevalence of CUD and hazardous cannabis use in a cross-sectional sample of adults receiving MC in Florida, and (2) to identify individual and cannabis-related factors associated with CUD in this sample.
Methods
Study Design
We conducted a cross-sectional analysis of data from the Medical Marijuana and Me (M3) Study [2]. M3 provides survey data from a sample of individuals who were currently registered in the MC program in Florida, USA. Approval for the parent study and current analyses was obtained from the University of Florida (UF) Institutional Review Board (IRB202200068). Our study is reported according to the STROBE guidelines [13], and the STROBE checklist is included as online supplementary material (for all online suppl. material, see https://doi.org/10.1159/000550537) linked to this article.
Participants
Participants were eligible if they were 18 years and older, able to speak and write in English, able to provide informed written consent, and were currently certified to obtain cannabis from medical dispensaries in Florida. A full description of the M3 study methods has been published previously [2], and the study questionnaire is available online (https://mmjoutcomes.org/m3study/). In brief, participants represented a convenience sample recruited from nine major MC clinics across Florida between May 2022 and December 2022. After consent, the participants completed an online questionnaire that included questions about demographics, physical and mental health, other substance use, and cannabis use patterns. All participants received a $20 gift card upon survey completion.
Measures
The primary outcome measure in this study was possible CUD, defined as a CUDIT-R score of ≥12 [5]. The CUDIT-R is an 8-item screening test with a 93% sensitivity and 88% specificity for CUD and has been validated for both psychiatric populations and non-clinical samples of adults who use cannabis medicinally [6, 7]. Participants answered the standardized CUDIT-R questions in relation to their cannabis use in the past 6 months, with response options for each item ranging from 0 to 4, and a total score calculated from the sum of the 8 items [5]. We also created a second dichotomous category with a cut point of ≥8 (any hazardous use) [6, 7, 10, 11], and a 3-category outcome based on the overall CUDIT-R score as low risk (<8), hazardous use (8–11), and possible CUD (≥12).
Independent variables included socio-demographic characteristics, self-reported mental health and substance use, previous use of marijuana, and current marijuana use patterns and history, based on existing literature that supports these factors being related to CUD [10, 11, 14].
Socio-Demographic Characteristics
Self-reported socio-demographic data collected included age (years), race/ethnicity (non-Hispanic White, non-Hispanic Black, Hispanic, or other), sex (male, female), education (high school or GED or below, some college, college graduate, graduate degree), employment (currently working, unemployed, retired, other including student), and income (less than $40,000, $40,000–$79,999, more than $80,000, don’t know/don’t want to answer).
Self-Reported Mental Health and Substance Use
Participants were provided with a list of mental health conditions and asked to check all the conditions they had been diagnosed with by a healthcare professional. Based on these responses, each participant was categorized as “yes” or “no” for anxiety, depression, PTSD, Attention-Deficit/Hyperactivity Disorder (ADHD), bipolar disorder, and schizophrenia. Participants were also asked to indicate whether they have used any of the following substances in the past 5 years, followed by a list of potential substances including alcohol, tobacco/cigarettes, nicotine vapes, cocaine, hallucinogens, club drugs (e.g., MDMA/Ecstasy/Molly), synthetic marijuana, kratom, opioids with and without a prescription, and amphetamines with and without a prescription. Persons were categorized “yes” or “no” to each substance used in the past 5 years. We also created a variable for “any drug use” that included any past 5-year use of cocaine, hallucinogens, club drugs, opioids without a prescription, and/or amphetamines without a prescription.
Cannabis Use History
We assessed the age of first marijuanause by asking “How old were you the first time you used marijuana of any kind?” Years of lifetime daily marijuana use was assessed by first asking participants whether they have ever used marijuana daily (Yes/No), and if they selected yes, they were then asked, “How many years of your life have you used marijuana daily?” with 9 categorical answer options ranging from less than 1 year to more than 50 years. These were recoded into three categories, <1 year, 1–5 years, and >5 years, with persons who never used marijuana daily categorized as <1 year.
Current Cannabis Use
Preferred method of MC use was assessed by asking participants “Have you found a method of using and/or a product type or strain that you prefer overall?”, and of those who indicated yes were asked “What method of using medical marijuana do you prefer? Please check all that apply”, followed by a list of 7 modes of consumption (flower, vape cartridges or pens, concentrates (for vaping or smoking), oral tinctures (with a dropper), topicals, oral concentrates, and oral capsules or edibles (lozenges, chews, chocolates, or gels). Participants were categorized as yes or no regarding whether they preferred each of these products, and individuals who indicated no overall preference were coded as “no” for each specific preference category. Motive for using marijuana was assessed by asking, “Which of the following best describes how much of your marijuana use is for recreational reasons vs. medical reasons?” Followed by 5 response options ranging from completely medical to completely recreational. Preferred potency of marijuana flower was assessed among people who used any marijuana flower by asking, “On average, what is the THC concentration of the flower you typically use?”, with 7 response options ranging from <5% to >35%. Participants also had the option to report “don’t know/not sure”, and we present responses from all who selected a specific % range. Perceptions of high levels of THC were also assessed by asking whether people agree or disagree that “Marijuana products with high THC content are more effective for my conditions or symptoms”, using a 5-point Likert scale (strongly disagree to strongly agree).
To assess Concerns about cannabis addiction, participants were asked, “How concerned are you about being addicted or dependent on medical marijuana?”, with five response options in a Likert scale (very concerned, concerned, neither concerned nor unconcerned, unconcerned, and very unconcerned).
Statistical Methods
We used the full sample of 632 persons who completed the M3 cross-sectional survey. The original study sample size was selected based on feasibility and financial resources rather than any specific research hypothesis. All analyses were conducted in R software.
We performed bivariate analysis to assess the relationships between each independent variable and the primary dichotomous outcome (CUDIT-R score ≥12 vs. <12), and also present findings with a different dichotomous cut point (CUDIT-R score ≥8 vs. <8) and with a 3-category outcome (CUDIT-R score ≥12, 8–11, or <8). Continuous variables such as age and age at first cannabis use were not normally distributed, so statistical tests included the Wilcoxon Rank Sum test (2 category outcome) and Kruskal-Wallis’s rank sum test (3-categories). Categorical independent variables were analyzed using Fisher’s exact tests and Pearson’s Chi-squared tests as appropriate. In general, we report in the text the actual p values and refer to findings with a p < 0.05 as “statistically significant”. However, testing multiple independent variables raises the potential for false-positive findings. Therefore, the Table with bivariate analyses also presents a p value adjusted for multiple hypothesis testing, using the Holm-Bonferroni method [15].
For multivariable analysis, we used logistic regression to identify variables associated with the primary outcome (possible CUD) and the secondary outcome (hazardous cannabis use). Fewer than 1% of responses were missing for any individual item, and therefore, we used complete case analysis (n = 622). Variable selection for the regression models was based on a-priori decisions to include age and sex, and up to 2 additional variables, based on findings from the bivariate analyses, from each of the following categories: other sociodemographic characteristics, mental health history, substance use, cannabis use history, and current cannabis use preferences. The final multivariable model included age, race/ethnicity, sex, income, any drug use (cocaine, hallucinogens, club drugs, opioids without prescription, amphetamines without prescription), age of first marijuana use, years of daily use, preferred flower, preferred oral concentrate, and motive for marijuana use. The model with the outcome of any hazardous use (CUDIT ≥8) also included history of anxiety disorder. For each of these models, the Hosmer-Lemeshow test showed a good fit (p > 0.05), and the Variance Inflation Factor did not suggest multicollinearity.
Results
Participants
All 632 participants who completed the survey were included in the analysis. Overall characteristics of participants are shown in the first column of Table 1. Participants age ranged from 19 to 85 years, with a median of 45 years. The majority were white (75%), female (63%), and educated (83% had at least some college). A high proportion reported a diagnosis of anxiety (62%) or depression (43%) and had consumed alcohol (66%) and/or other substances in the past 5 years. The age when participants first used any marijuana ranged from 5 to 79 years, with a median age of 16 at first use. Participants also reported an extensive experience with marijuana use; the majority (81%) reported using marijuana daily for at least a year, with 46% of the total sample having used marijuana daily for five or more years. Among those reporting a preferred product (n = 554), flower was the most frequently reported preferred product (57%), followed by vaping (41%) and edibles/capsules (34%). Nearly all (98%) indicated that their marijuana use was mostly or equally for medical reasons, with only 2% reporting their use was mostly for recreational reasons. Overall, nearly all (84%) reported being either very unconcerned or unconcerned about being addicted or dependent on medical marijuana, while only 4% reported either concerned or very concerned.
Table 1.
Associations of demographic, mental health, and marijuana use history and patterns with CUDIT-R Score Categories among 632 adults using medical cannabis in Florida: Bivariate analyses
| | | High risk for CUD | Any hazardous use | Low, medium and high risk for CUD | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Sample size Percent | total N = 632 (100.0%) | CUDIT <12 n = 504 (79.7%) | CUDIT ≥ 12 n = 128 (20.3%) | raw p value | adj p value | CUDIT <8 N = 290 (45.9%) | CUDIT ≥ 8 N = 342 (54.1%) | raw p value | adj p value | CUDIT <8 n = 290 (45.9%) | CUDIT 8–11 n = 214 (33.9%) | CUDIT ≥12 n = 128 (20.3%) | raw p value | adj p value |
| 95% CI | [76.4–82.7] | [17.3–23.6] | [42.0–49.8] | [50.2–58.0] | [42.0–49.8] | [30.3–37.6] | [17.3–23.6] | |||||||
| Individual factors | ||||||||||||||
| Age (median, IQR) | 45.0 [35.0; 58.0] | 47.0 [37.0; 59.0] | 39.5 [29.0; 55.2] | 0.0001 | 0.005 | 49.0 [38.0; 60.0] | 42.0 [32.0; 57.0] | 0.0001 | 0.0033 | 49.0 [38.0; 60.0] | 43.0 [34.0; 57.0] | 39.5 [29.0; 55.2] | 0.0000 | 0.001 |
| Race/Ethnicity (n, %) | | | | 0.35 | 1.0 | | | 0.064 | 1.0 | | | | 0.26 | 1.0 |
| Non-Hispanic White | 471 (74.5%) | 381 (75.6%) | 90 (70.3%) | | | 226 (77.9%) | 245 (71.6%) | | | 226 (77.9%) | 155 (72.4%) | 90 (70.3%) | | |
| Non-Hispanic Black | 35 (5.5%) | 29 (5.8%) | 6 (4.7%) | | | 18 (6.2%) | 17 (5.0%) | | | 18 (6.2%) | 11 (5.1%) | 6 (4.7%) | | |
| Hispanic | 93 (14.7%) | 71 (14.1%) | 22 (17.2%) | | | 37 (12.8%) | 56 (16.4%) | | | 37 (12.8%) | 34 (15.9%) | 22 (17.2%) | | |
| Other | 33 (5.2%) | 23 (4.6%) | 10 (7.8%) | | | 9 (3.1%) | 24 (7.0%) | | | 9 (3.1%) | 14 (6.5%) | 10 (7.8%) | | |
| Sex (n, %) | | | | 0.39 | 1.0 | | | 0.23 | 1.0 | | | | 0.46 | 1.0 |
| Male | 236 (37.3%) | 184 (36.5%) | 52 (40.6%) | | | 101 (34.8%) | 135 (39.5%) | | | 101 (34.8%) | 83 (38.8%) | 52 (40.6%) | | |
| Female | 396 (62.7%) | 320 (63.5%) | 76 (59.4%) | | | 189 (65.2%) | 207 (60.5%) | | | 189 (65.2%) | 131 (61.2%) | 76 (59.4%) | | |
| Education (n, %) | | | | 0.16 | 1.0 | | | 0.27 | 1.0 | | | | 0.27 | 1.0 |
| High school or below | 108 (17.1%) | 80 (15.9%) | 28 (21.9%) | | | 42 (14.5%) | 66 (19.3%) | | | 42 (14.5%) | 38 (17.8%) | 28 (21.9%) | | |
| Any college | 408 (64.6%) | 334 (66.3%) | 74 (57.8%) | | | 192 (66.2%) | 216 (63.2%) | | | 192 (66.2%) | 142 (66.4%) | 74 (57.8%) | | |
| Graduate degree | 116 (18.4%) | 90 (17.9%) | 26 (20.3%) | | | 56 (19.3%) | 60 (17.5%) | | | 56 (19.3%) | 34 (15.9%) | 26 (20.3%) | | |
| Employment (n, %) | | | | 0.34 | 1.0 | | | 0.50 | 1.0 | | | | 0.66 | 1.0 |
| Currently working | 369 (58.9%) | 288 (57.7%) | 81 (63.3%) | | | 165 (57.7%) | 204 (59.8%) | | | 165 (57.7%) | 123 (57.7%) | 81 (63.3%) | | |
| Unemployed | 121 (19.3%) | 98 (19.6%) | 23 (18.0%) | | | 56 (19.6%) | 65 (19.1%) | | | 56 (19.6%) | 42 (19.7%) | 23 (18.0%) | | |
| Retired | 102 (16.3%) | 87 (17.4%) | 15 (11.7%) | | | 52 (18.2%) | 50 (14.7%) | | | 52 (18.2%) | 35 (16.4%) | 15 (11.7%) | | |
| Other/student) | 35 (5.6%) | 26 (5.2%) | 9 (7.0%) | | | 13 (4.5%) | 22 (6.5%) | | | 13 (4.5%) | 13 (6.1%) | 9 (7.0%) | | |
| Income (n, %) | | | | 0.019 | 0.51 | | | 0.10 | 1.0 | | | | 0.035 | 0.81 |
| Less than $39,999 | 214 (34.2%) | 160 (32.1%) | 54 (42.2%) | | | 83 (29.1%) | 131 (38.4%) | | | 83 (29.1%) | 77 (36.2%) | 54 (42.2%) | | |
| $40,000–79,999 | 206 (32.9%) | 160 (32.1%) | 46 (35.9%) | | | 99 (34.7%) | 107 (31.4%) | | | 99 (34.7%) | 61 (28.6%) | 46 (35.9%) | | |
| More than $80,000 | 181 (28.9%) | 155 (31.1%) | 26 (20.3%) | | | 91 (31.9%) | 90 (26.4%) | | | 91 (31.9%) | 64 (30.0%) | 26 (20.3%) | | |
| Don't know/no answer | 25 (4.0%) | 23 (4.6%) | 2 (1.6%) | | | 12 (4.2%) | 13 (3.8%) | | | 12 (4.2%) | 11 (5.2%) | 2 (1.6%) | | |
| Mental Health Conditions | ||||||||||||||
| Anxiety | 451 (71.4%) | 358 (71.0%) | 93 (72.7%) | 0.72 | 1.0 | 190 (65.5%) | 261 (76.3%) | 0.0028 | 0.078 | 190 (65.5%) | 168 (78.5%) | 93 (72.7%) | 0.006 | 0.81 |
| Depression | 392 (62.0%) | 305 (60.5%) | 87 (68.0%) | 0.12 | 1.0 | 172 (59.3%) | 220 (64.3%) | 0.20 | 1.0 | 172 (59.3%) | 133 (62.1%) | 87 (68.0%) | 0.24 | 1.0 |
| PTSD | 270 (42.7%) | 221 (43.8%) | 49 (38.3%) | 0.26 | 1.0 | 116 (40.0%) | 154 (45.0%) | 0.20 | 1.0 | 116 (40.0%) | 105 (49.1%) | 49 (38.3%) | 0.07 | 1.0 |
| ADHD | 123 (19.5%) | 95 (18.8%) | 28 (21.9%) | 0.44 | 1.0 | 51 (17.6%) | 72 (21.1%) | 0.28 | 1.0 | 51 (17.6%) | 44 (20.6%) | 28 (21.9%) | 0.52 | 1.0 |
| Bipolar | 89 (14.1%) | 71 (14.1%) | 18 (14.1%) | 0.99 | 1.0 | 35 (12.1%) | 54 (15.8%) | 0.18 | 1.0 | 35 (12.1%) | 36 (16.8%) | 18 (14.1%) | 0.32 | 1.0 |
| Substance use (past 5 yrs) | ||||||||||||||
| Alcohol | 419 (66.3%) | 326 (64.7%) | 93 (72.7%) | 0.089 | 1.0 | 193 (66.6%) | 226 (66.1%) | 0.90 | 1.0 | 193 (66.6%) | 133 (62.1%) | 93 (72.7%) | 0.14 | 1.0 |
| Tobacco/Cigarettes | 224 (35.4%) | 177 (35.1%) | 47 (36.7%) | 0.74 | 1.0 | 97 (33.4%) | 127 (37.1%) | 0.33 | 1.0 | 97 (33.4%) | 80 (37.4%) | 47 (36.7%) | 0.62 | 1.0 |
| Nicotine Vape | 123 (19.5%) | 98 (19.4%) | 25 (19.5%) | 0.98 | 1.0 | 56 (19.3%) | 67 (19.6%) | 0.93 | 1.0 | 56 (19.3%) | 42 (19.6%) | 25 (19.5%) | 0.99 | 1.0 |
| Cocaine | 24 (3.8%) | 13 (2.6%) | 11 (8.6%) | 0.0035 | 0.1054 | 3 (1.0%) | 21 (6.1%) | 0.0008 | 0.026 | 3 (1.0%) | 10 (4.7%) | 11 (8.6%) | 0.0004 | 0.012 |
| Hallucinogen | 59 (9.3%) | 36 (7.1%) | 23 (18.0%) | 0.0002 | 0.0056 | 16 (5.5%) | 43 (12.6%) | 0.0024 | 0.071 | 16 (5.5%) | 20 (9.3%) | 23 (18.0%) | 0.0003 | 0.009 |
| Club Drugs | 23 (3.6%) | 13 (2.6%) | 10 (7.8%) | 0.014 | 0.38 | 6 (2.1%) | 17 (5.0%) | 0.052 | 1.0 | 6 (2.1%) | 7 (3.3%) | 10 (7.8%) | 0.022 | 0.53 |
| Synthetic marijuana | 1 (0.2%) | 1 (0.2%) | 0 (0%) | 1.0 | 1.0 | 1 (0.3%) | 0 (0%) | 0.46 | 1.0 | 1 (0.3%) | 0 (0%) | 0 (0%) | 1.0 | 1.0 |
| Kratom | 35 (5.5%) | 26 (5.2%) | 9 (7.0%) | 0.41 | 1.0 | 11 (3.8%) | 24 (7.0%) | 0.08 | 1.0 | 11 (3.8%) | 15 (7.0%) | 9 (7.0%) | 0.21 | 1.0 |
| Opioid (w/o prescription) | 24 (3.8%) | 13 (2.6%) | 11 (8.6%) | 0.0035 | 0.11 | 7 (2.4%) | 17 (5.0%) | 0.094 | 1.0 | 7 (2.4%) | 6 (2.8%) | 11 (8.6%) | 0.013 | 0.32 |
| Amphetamine | 17 (2.7%) | 11 (2.2%) | 6 (4.7%) | 0.13 | 1.0 | 3 (1.0%) | 14 (4.1%) | 0.018 | 0.45 | 3 (1.0%) | 8 (3.7%) | 6 (4.7%) | 0.038 | 0.83 |
| Any Drug Use* | 88 (13.9%) | 55 (10.9%) | 33 (25.8% | 0.00 | 0.0005 | 23 (7.9%) | 65 (19.0%) | 0.0001 | 0.0021 | 23 (7.9%) | 32 (15.0%) | 33 (25.8%) | 0.00 | 0.0002 |
| Cannabis-related factors | ||||||||||||||
| Age first use (median, IQR) | 16.0 [14.0; 20.0] | 17.0 [15.0; 20.0] | 16.0 [14.0; 18.0] | 0.003 | 0.10 | 17.0 [15.0; 23.0] | 16.0 [14.0; 18.0] | 0.00 | 0.0008 | 17.0 [15.0; 23.0] | 16.0 [14.0; 19.0] | 16.0 [14.0; 18.0] | 0.0001 | 0.0024 |
| Number years used daily | | | | 0.0001 | 0.002 | | | 0.00 | 0.00 | | | | 0.00 | 0.00 |
| Zero or <1 year | 120 (19.0%) | 110 (21.8%) | 10 (7.8%) | | | 85 (29.3%) | 35 (10.2%) | | | 85 (29.3%) | 25 (11.7%) | 10 (7.8%) | | |
| 1–5 years | 223 (35.3%) | 183 (36.3%) | 40 (31.2%) | | | 110 (37.9%) | 113 (33.0%) | | | 110 (37.9%) | 73 (34.1%) | 40 (31.2%) | | |
| 5 or more years | 289 (45.7%) | 211 (41.9%) | 78 (60.9%) | | | 95 (32.8%) | 194 (56.7%) | | | 95 (32.8%) | 116 (54.2%) | 78 (60.9%) | | |
| Preferred method of use** | ||||||||||||||
| Flower | 359 (56.8%) | 271 (53.8%) | 88 (68.8%) | 0.002 | 0.072 | 132 (53.4%) | 227 (73.9%) | 0.00 | 0.00 | 132 (53.4%) | 139 (70.6%) | 88 (80.0%) | 0.00 | 0.00 |
| Vape | 256 (40.5%) | 208 (41.3%) | 48 (37.5%) | 0.44 | 1.0 | 107 (43.3%) | 149 (48.5%) | 0.089 | 1.0 | 107 (43.3%) | 101 (51.3%) | 48 (43.6%) | 0.049 | 1.0 |
| Inhaled Concentrates | 109 (17.2%) | 84 (16.7%) | 25 (19.5%) | 0.44 | 1.0 | 36 (14.6%) | 73 (23.8%) | 0.003 | 0.083 | 36 (14.6%) | 48 (24.4%) | 25 (22.7%) | 0.01 | 0.26 |
| Tincture | 80 (12.7%) | 65 (12.9%) | 15 (11.7%) | 0.72 | 1.0 | 49 (19.8%) | 31 (10.1%) | 0.003 | 0.083 | 49 (19.8%) | 16 (8.1%) | 15 (13.6%) | 0.007 | 0.19 |
| Topical | 47 (7.4%) | 36 (7.1%) | 11 (8.6%) | 0.58 | 1.0 | 25 (10.1%) | 22 (7.2%) | 0.30 | 1.0 | 25 (10.1%) | 11 (5.6%) | 11 (10.0%) | 0.29 | 1.0 |
| Oral Concentrates | 56 (8.9%) | 51 (10.1%) | 5 (3.9%) | 0.027 | 0.68 | 26 (10.5%) | 30 (9.8%) | 0.93 | 1.0 | 26 (10.5%) | 25 (12.7%) | 5 (4.5%) | 0.05 | 1.0 |
| Edibles/Tablets/Capsules | 217 (34.3%) | 184 (36.5%) | 33 (25.8%) | 0.023 | 0.58 | 109 (44.1%) | 108 (35.2%) | 0.11 | 1.0 | 109 (44.1%) | 75 (38.1%) | 33 (30.0%) | 0.062 | 1.0 |
| Other | 8 (1.3%) | 8 (1.6%) | 0 (0%) | 0.37 | 1.0 | 2 (0.8%) | 6 (2.0%) | 0.30 | 1.0 | 2 (0.8%) | 6 (3.0%) | 0 (0%) | 0.057 | 1.0 |
| Motive for using marijuana | | | | 0.00 | 0.0007 | | | 0.00 | 0.00 | | | | 0.0001 | 0.003 |
| Completely/mostly med | 439 (69.7%) | 370 (73.7%) | 69 (53.9%) | | | 231 (80.2%) | 208 (60.8%) | | | 231 (80.2%) | 139 (65.0%) | 69 (53.9%) | | |
| Equal med/rec | 177 (28.1%) | 125 (24.9%) | 52 (40.6%) | | | 55 (19.1%) | 122 (35.7%) | | | 55 (19.1%) | 70 (32.7%) | 52 (40.6%) | | |
| Completely/mostly rec | 14 (2.2%) | 7 (1.4%) | 7 (5.5%) | | | 2 (0.7%) | 12 (3.5%) | | | 2 (0.7%) | 5 (2.3%) | 7 (5.5%) | | |
| Average THC concentration in Flower (n = 438) | | | | 0.19 | 1.0 | | | 0.078 | 1.0 | | | | 0.11 | 1.0 |
| <15% | 32 (7.3%) | 28 (8.5%) | 4 (3.7%) | | | 17 (9.6%) | 15 (5.7%) | | | 17 (9.6%) | 11 (7.2%) | 4 (3.7%) | | |
| 15%–20% | 150 (34.2%) | 116 (35.2%) | 34 (31.5%) | | | 67 (37.9%) | 83 (31.8%) | | | 67 (37.9%) | 49 (32.0%) | 34 (31.5%) | | |
| 20%–25% | 173 (39.5%) | 121 (36.7%) | 52 (48.1%) | | | 66 (37.3%) | 107 (41.0%) | | | 66 (37.3%) | 55 (35.9%) | 52 (48.1%) | | |
| 25%–30% | 53 (12.1%) | 39 (11.8%) | 14 (13.0%) | | | 14 (7.9%) | 39 (14.9%) | | | 14 (7.9%) | 25 (16.3%) | 14 (13.0%) | | |
| 30% or more | 30 (6.9%) | 26 (7.9%) | 4 (3.7%) | | | 13 (7.4%) | 17 (6.5%) | | | 13 (7.4%) | 13 (8.5%) | 4 (3.7%) | | |
| High THC content is more effective for me | | | | 0.21 | 1.0 | | | 0.96 | 1.0 | | | | 0.44 | 1.0 |
| Strongly Disagree | 46 (7.3%) | 34 (6.8%) | 12 (9.4%) | | | 20 (7.0%) | 26 (7.6%) | | | 20 (7.0%) | 14 (6.5%) | 12 (9.4%) | | |
| Disagree | 38 (6.0%) | 32 (6.4%) | 6 (4.7%) | | | 17 (5.9%) | 21 (6.1%) | | | 17 (5.9%) | 15 (7.0%) | 6 (4.7%) | | |
| Neutral | 104 (16.5%) | 85 (17.0%) | 19 (14.8%) | | | 51 (17.8%) | 53 (15.5%) | | | 51 (17.8%) | 34 (15.9%) | 19 (14.8%) | | |
| Agree | 182 (28.9%) | 136 (27.1%) | 46 (35.9%) | | | 83 (28.9%) | 99 (28.9%) | | | 83 (28.9%) | 53 (24.8%) | 46 (35.9%) | | |
| Strongly Agree | 259 (41.2%) | 214 (42.7%) | 45 (35.2%) | | | 116 (40.4%) | 143 (41.8%) | | | 116 (40.4%) | 98 (45.8%) | 45 (35.2%) | | |
| Concern about addiction/dependence | | | | 0.00 | 0.00 | | | 0.0008 | 0.026 | | | | 0.0001 | 0.003 |
| Very unconcerned | 400 (63.5%) | 341 (67.9%) | 59 (46.1%) | | | 206 (71.5%) | 194 (56.7%) | | | 206 (71.5%) | 135 (63.1%) | 59 (46.1%) | | |
| Unconcerned | 128 (20.3%) | 98 (19.5%) | 30 (23.4%) | | | 50 (17.4%) | 78 (22.8%) | | | 50 (17.4%) | 48 (22.4%) | 30 (23.4%) | | |
| Neutral | 78 (12.4%) | 54 (10.8%) | 24 (18.8%) | | | 27 (9.4%) | 51 (14.9%) | | | 27 (9.4%) | 27 (12.6%) | 24 (18.8%) | | |
| Concerned | 18 (2.9%) | 6 (1.2%) | 12 (9.4%) | | | 3 (1.0%) | 15 (4.4%) | | | 3 (1.0%) | 3 (1.4%) | 12 (9.4%) | | |
| Very concerned | 6 (1.0%) | 3 (0.6%) | 3 (2.3%) | | | 2 (0.7%) | 4 (1.2%) | | | 2 (0.7%) | 1 (0.5%) | 3 (2.3%) | | |
Mental health concerns–self-reported having been diagnosed by a clinician.
*Any drug use - any past 5 years use of cocaine, hallucinogens, club drugs, opioids without a prescription, or amphetamines without a prescription.
**Preferred method of use–persons could choose more than one category; some had no preferred method and were coded as “no” for each option.
Self-reported motives for use Med, medical use; rec, recreational use.
For the primary outcome (CUDIT-R score), among the 632 study participants, 128 (20.3% 95% CIs [17.3%–23.6%]) had possible CUD (≥12), 214 (33.9% [30.3%; 37.6%]) had a score consistent with hazardous use [16–19], and the remaining 290 (45.9% [42.0%; 49.8%]) had a lower risk CUDIT-R score (<8) (Fig. 1). Thus, one in five participants from our sample had a possible CUD, and more than half (54%) had a CUDIT-R score consistent with hazardous use or possible CUD.
Fig. 1.
CUDIT-R score distribution among 632 persons receiving medical cannabis in Florida, 2022.
Bivariate Analyses: Factors Associated with Possible CUD (CUDIT-R Score ≥12)
Table 1 presents comparisons of the demographic characteristics, mental health conditions, substance use history, cannabis use history, and cannabis use preferences between those with and without possible CUD. Those with possible CUD were younger (median age and interquartile range of 39.5 years [22.0–55.2] vs. 47.0 years [37.0–59.0]), and reported lower incomes, compared to those without possible CUD. However, there were no significant differences by possible CUD status in race/ethnicity, sex, education, or employment status (Table 1).
None of the self-reported mental health conditions (anxiety, depression, PTSD, ADHD, bipolar) were significantly associated with possible CUD (See Table 1). Persons with possible CUD were significantly more likely to report past-5-year use of cocaine, hallucinogens, club drugs, and opioid use without a prescription. However, neither alcohol, tobacco, nicotine products, or amphetamine use without a prescription were associated with possible CUD.
Compared to those without possible CUD, people with a possible CUD had a younger median age of first marijuana use (16.0 vs. 17.0 years) and a greater proportion of those with possible CUD reported ≥5 years of lifetime daily marijuana use (61% vs. 42%). Among those who identified one or more preferred methods of use (n = 553), a greater proportion of those with possible CUD indicated a preference for using flower (69% vs. 54%), and a lower proportion preferred using either oral concentrates (4% vs. 10%) or oral edibles/capsules (26% vs. 37%). Those with possible CUD were less likely to report using marijuana completely or mostly for medical reasons (54% vs. 74%). Among the subset of persons who used marijuana flower and knew the average THC concentration in the flower they used, there were no significant differences in preferred THC % in flower among those with or without possible CUD (Table 1). There were also no differences between those with or without possible CUD in the proportion who “agreed/strongly agreed” that marijuana products with a higher THC content are more effective (71% vs. 70%).
While only 4% of the entire sample reported that they were “concerned” or “very concerned” about being addicted or dependent on marijuana, those with possible CUD were more likely to report any concern and less likely to report being “very unconcerned” (see Table 1).
Bivariate Analyses: Secondary Outcomes
Table 1 also presents comparisons of demographic characteristics, mental health and substance use behavior, and cannabis use history and patterns across two groups at the CUDIT cut-point of ≥8 (any hazardous use). Results were generally similar to the primary outcome (CUDIT-R score ≥12) with a few exceptions. At this cut point, some differences in preferred mode of marijuana consumption were even more prominent; those with any hazardous cannabis use were more to report preferences for flower and for smoked concentrates, and less likely to report preferences for oral tinctures or oral edibles/capsules (Table 1). Also, those with any hazardous use were more likely to report a diagnosis of anxiety.
Table 1 also shows the distribution across 3 groups: participants with CUDIT scores of ≥12 (possible CUD), 8–11 (hazardous use), and <8 (non-hazardous use). The proportions of participants with various characteristics in the middle (hazardous use) were mostly in-between those with possible CUD and those with non-hazardous use, with a few exceptions. For example, the middle group (hazardous use) was most likely to report anxiety compared to either those with possible CUD or those with non-hazardous use. Also, those with hazardous use were most likely to report preferences for smoked concentrates and least likely to report preference for oral tinctures compared to the other two groups. Very few of those in the hazardous group expressed any concern about possible addiction or dependence (see Table 1).
Multivariable Analyses
In the multivariable analyses of factors associated with having a possible CUD (Table 2), four variables remained statistically significant: younger age (aOR 0.98, 95% CIs [0.96–1.0] for each increased year), number of years of daily marijuana use vs. <1 year (1–5 years aOR 2.25 95% CIs [1.25–5.20]; 5+ years aOR 3.14 95% CIs [1.49–7.22]), overall preference for oral concentrates as a mode of consumption (OR 0.29, 95% CIs [0.10–0.70]), and self-reported reason for marijuana use (completely/mostly recreational vs. completely/mostly medical (OR 5.09, 95% CI: [1.53–17.27]).
Table 2.
Factors associated with possible CUD (CUDIT-R score ≥12) and any hazardous use (CUDIT-R score ≥8) among 622 adults using medical cannabis in Florida: Multivariable analysis
| | Possible CUD (CUDIT-R ≥12) | Any hazardous use (CUDIT-R ≥8) | ||||
|---|---|---|---|---|---|---|
| Variable | OR | 95% CI | p value | OR | 95% CI | p value |
| Age | 0.98 | [0.96, 1.00] | 0.016 | 0.99 | [0.98, 1.00] | 0.21 |
| Race/Ethnicity | ||||||
| Non-Hispanic White | – | – | – | – | – | – |
| Non-Hispanic Black | 0.72 | [0.25, 1.80] | 0.51 | 0.82 | [0.38, 1.77] | 0.61 |
| Hispanic | 1.13 | [0.62, 2.01] | 0.68 | 1.26 | [0.75, 2.13] | 0.38 |
| Other | 1.29 | [0.54, 2.93] | 0.55 | 2.12 | [0.93, 5.19] | 0.08 |
| Sex | ||||||
| Female vs. Male | 0.85 | [0.55, 1.32] | 0.46 | 0.81 | [0.55, 1.18] | 0.27 |
| Income | ||||||
| Less than $39,999 | – | – | – | – | – | – |
| $40,000–79,999 | 1.05 | [0.64, 1.70] | 0.85 | 0.79 | [0.52, 1.22] | 0.29 |
| More than $80,000 | 0.59 | [0.33, 1.04] | 0.07 | 0.81 | [0.52, 1.28] | 0.38 |
| Don’t know/Don’t want to answer | 0.30 | [0.04, 1.19] | 0.13 | 1.29 | [0.49, 3.45] | 0.60 |
| Age when used marijuana for first time | 0.98 | [0.94, 1.00] | 0.12 | 0.97 | [0.95, 0.99] | 0.006 |
| Number of years used marijuana daily | ||||||
| Less than 1 year | – | – | – | – | – | – |
| 1–5 years | 2.25 | [1.05, 5.20] | 0.04 | 2.34 | [1.39, 3.98] | 0.001 |
| 5+ or more years | 3.14 | [1.49, 7.22] | 0.004 | 3.61 | [2.13, 6.22] | <0.001 |
| Any Drug Use* (Yes vs. No) | 1.71 | [0.98, 2.95] | 0.054 | 1.52 | [0.87, 2.73] | 0.15 |
| Overall Preferred Method of Medical Marijuana | ||||||
| Flower (Yes vs. No) | 1.22 | [0.77, 1.93] | 0.40 | 1.48 | [1.02, 2.14] | 0.039 |
| Oral Concentrate (Yes vs. No) | 0.29 | [0.10, 0.70] | 0.012 | 0.85 | [0.46, 1.58] | 0.61 |
| Motive For Using Medical Marijuana | ||||||
| Completely/Mostly Medical | – | – | – | – | – | – |
| Equally Recreational and Medical | 1.51 | [0.96, 2.37] | 0.07 | 1.70 | [1.13, 2.59] | 0.01 |
| Completely/Mostly Recreational | 5.09 | [1.53, 17.3] | 0.007 | 6.35 | [1.58, 42.9] | 0.02 |
| Anxiety Diagnosed by a Health Professional (Yes vs. No) | – | – | – | 1.72 | [1.14, 2.59] | 0.009 |
*Any drug use - any past 5 years use of cocaine, hallucinogens, club drugs, opioids without a prescription, or amphetamines without a prescription.
In the multivariable analyses of factors associated with any hazardous use (Table 2), ≥1 year of daily marijuana use and self-reported recreational reasons for marijuana use were similarly associated with the higher risk category. However, at this cut point, age and preference for oral concentrates were no longer associated with the outcome, whereas younger age of first marijuana use, marijuana flower as a preferred product, and self-reported anxiety diagnosis were significantly associated with hazardous cannabis use (see Table 2 for specific odds ratios and 95% confidence intervals).
Discussion
We sought to estimate the prevalence of cannabis use disorder (CUD) and to identify factors associated with possible CUD among adults who were certified to obtain medical cannabis in Florida. In this sample of 632 adults recruited from medical cannabis providers and clinics from across the state, 20% had a CUDIT score of 12 or more, indicating possible CUD. An additional 34% had a CUDIT score of 8–11 (hazardous use), such that more than half of the sample (54%) had either hazardous use or possible CUD. The prevalence of possible CUD in our sample is consistent with other studies that measured or estimated CUD among adults who use marijuana, which range from about 10% to 30% or more [8, 11, 16–18, 20]. Taken together, these findings suggest a need for additional resources to help reduce the associated public health and individual consequences that may occur among persons who use medical cannabis to manage health conditions.
In our sample, CUD was associated with younger age, which is consistent with other studies [8, 11, 21]. Other demographic factors such as sex and race/ethnicity were not significantly associated with CUD in this sample. None of the individual self-reported diagnoses of mental health were associated with CUD in this sample, although self-reported anxiety disorders were associated with hazardous cannabis use (CUDIT-R score ≥8). Some previous studies have found mental health conditions to be more strongly associated with CUD [8–11]. In our sample, some types of drug use were more strongly associated with CUD than others, including the use of cocaine, hallucinogens, club drugs, or opioids without a prescription. Notably, past 5 years use of alcohol, tobacco, or nicotine vape products were not significantly associated with CUD, which is different from some other studies finding tobacco use to be associated with CUD [22]. The substance use measures in our study were very general (any use in the past 5 years), and more detailed assessments of substance use severity might have different associations with CUD.
Several factors associated with marijuana use history and current patterns of use were associated with CUD. Unlike some other studies [12], we did not find younger age of first cannabis use to be associated with CUD, although it was associated with any hazardous cannabis use. However, CUD risk was greater in those with one or more years of daily cannabis use. Notably, 81% of this sample reported one or more years of using cannabis daily in their lifetime, suggesting that daily use may be more of the norm for persons engaged in a medical cannabis system.
Access to dispensaries in the US has also provided access to a wider range of products, some of which provide more immediate effects (especially inhaled products), whereas oral products tend to have a slower onset and longer effect. Among those who expressed a preferred type of product, we found a higher risk of hazardous use among those who preferred marijuana flower (which is usually inhaled), and a lower risk of CUD among those who preferred oral concentrates. It is possible that being able to obtain a more immediate effect could support addictive tendencies, and at least one other study found that inhaled methods of cannabis administration were associated with a higher risk of CUD [22].
Although all participants in this sample had received a certification to obtain medical cannabis, about 30% of them indicated that their reason for use was at least equally for recreational reasons. Self-reported recreational motivations for use were associated with a greater risk for CUD in this sample. However, other studies have sampled persons from the general population and found that use for self-reported medical reasons were more likely to have a CUD compared to those using for recreational reasons [8, 11]. One potential limitation of the CUDIT-R is that it includes an item about the “frequency” of cannabis use, and persons who use every day or nearly every day will have higher CUDIT-R scores. Increased frequency of use could be expected when using cannabis as a medicinal treatment [6].
The percentage of THC in flower products has increased substantially over time, and some have expressed concern that higher levels of THC could result in more problematic use [10]. However, in our sample, neither possible CUD or hazardous cannabis use were associated with a preference for a higher THC percentage in flower (among persons who use flower), nor with the perception that cannabis products with higher concentrations of THC are more effective.
The study has several important limitations. This was a cross-sectional analysis of individuals who currently use medicinal cannabis in the state of Florida, which limits the ability to assess temporality between the various exposures and the outcome of current CUD. Also, while we recruited persons from many types of medical settings across the state, it was not a formal systematic sample, and it is unknown whether study participants are more or less likely to have CUD compared to the total population. This study did not include a formal substance abuse assessment, and the CUDIT-R is only an estimate of the true prevalence. As noted above, medical cannabis patients often use cannabis daily to treat chronic conditions and therefore will score higher on the CUDIT-R item assessing frequency of consumption. The study sought to explore a number of factors that may be associated with possible CUD, which resulted in multiple statistical comparisons in the bivariate analyses. Because multiple statistical comparisons increase the possibility that some statistically significant findings in the bivariate analyses could be due to chance, we also provided results on adjusted p values that account for multiple comparisons.
Despite the limitations, the study has several strengths. It is among the first to analyze the prevalence of possible CUD among a statewide sample of adults with access to medical cannabis, and it included items on specific cannabis products that are used and preferred by participants. Our data suggest that approximately 1 in 5 persons in the medical cannabis system could have a CUD, which is significantly greater than the 4% of the sample who reported having concern about being addicted to cannabis. Given the large number of adults who are participating in the medical cannabis program, there could be a need to provide more screening for CUD and to provide resources for treatment for those who are interested.
Acknowledgments
The authors acknowledge the contributions of the Planning Committee to the M3 study design and surveys: Dinese C. Vidot, PhD, Dushyantha T. Jayaweera, MD, Jason Ford, MA, PhD, Jennifer Jean-Jacques, MPH, Jessica Walters, John M. Crump, MD, Jonathan Quinonez, DO, Justin Davis, MD, Martha Rosenthal, PhD, Michelle Weiner, DO, Michelle Wilson, MBA, Patricia A. Green-Powell, PhD, and MC clinic networks in Florida, USA, who helped with patient recruitment: Affordable Marijuana License, Dr. Bob's Compassion Clinic, CannaMD, DocMJ, Green Health & Wellness, John M. Crump, MD, with Releafe Now, the practice of Justin Davis, MD, the practice of Kelly King, MD, the practice of Melanie Bone, MD, and Michelle Weiner, DO, with Spine and Wellness Centers of America.
Statement of Ethics
This study was approved by the University of Florida Institutional Review Board (IRB202002925). All study participants provided written informed consent.
Conflict of Interest Statement
Dr. Wang was a member of the journal’s Editorial Board at the time of submission. The other authors have no conflicts of interest to declare.
Funding Sources
This study was funded by the Consortium for Medical Marijuana Clinical Outcomes Research, a state-funded research consortium of 11 universities in the state of Florida.
Author Contributions
Robert L. Cook, Sophie Maloney, Abigail Gracy, Myooran Sivarupan, Ruba Sajdeya, Mahmudul Hasan, and Yan Wang conceptualized the manuscript. Hanzhi Gao did the data analysis with input from co-authors. Sophie Maloney prepared the first draft. All authors revised, edited, and approved the final draft. Robert L. Cook and Yan Wang developed the initial survey, and Ruba Sajdeya and Hannah Jury contributed substantively to survey development.
Funding Statement
This study was funded by the Consortium for Medical Marijuana Clinical Outcomes Research, a state-funded research consortium of 11 universities in the state of Florida.
Data Availability Statement
Study surveys, procedures to request data, and contact information regarding the M3 study are available online (https://mmjoutcomes.org/m3study/). Data are available to interested researchers upon application and approval by the M3 Databank. Further enquiries can be directed to the corresponding author.
Supplementary Material.
References
- 1. Boehnke KF, Dean O, Haffajee RL, Hosanagar A. U.S. trends in registration for medical cannabis and reasons for use from 2016 to 2020: an observational study. Ann Intern Med. 2022;175(7):945–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Sajdeya R, Fechtel HJ, Spandau G, Goodin AJ, Brown JD, Jugl S, et al. Protocol of a combined cohort and cross-sectional study of persons receiving medical cannabis in Florida, USA: the Medical Marijuana and Me (M3) study. Med Cannabis Cannabinoids. 2023;6(1):46–57. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Office of Medical Marijuana . Weekly updates. Available from: https://knowthefactsmmj.com/about/weekly-updates/
- 4. Patel J, Marwaha R. Cannabis use disorder. Statpearls Publ [Internet]. 2024; Available from: https://www.ncbi.nlm.nih.gov/books/NBK538131/ [Google Scholar]
- 5. Adamson SJ, Kay-Lambkin FJ, Baker AL, Lewin TJ, Thornton L, Kelly BJ, et al. An improved brief measure of cannabis misuse: the Cannabis Use Disorders Identification Test-Revised (CUDIT-R). Drug Alcohol Depend. 2010;110(1–2):137–43. [DOI] [PubMed] [Google Scholar]
- 6. Loflin M, Babson K, Browne K, Bonn-Miller M. Assessment of the validity of the CUDIT-R in a subpopulation of cannabis users. Am J Drug Alcohol Abuse. 2018;44(1):19–23. [DOI] [PubMed] [Google Scholar]
- 7. Turna J, Balodis I, Munn C, Van Ameringen M, Busse J, MacKillop J. Overlapping patterns of recreational and medical cannabis use in a large community sample of cannabis users. Compr Psychiatry. 2020;102:152188. [DOI] [PubMed] [Google Scholar]
- 8. Rubin-Kahana DS, Hassan AN, Sanches M, Le Foll B. Medical cannabis and past-year cannabis use disorder among adult recreational users in the United States: results from a nationally representative sample. Front Psychiatry. 2022;13:836908. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Bonn-Miller MO, Boden MT, Bucossi MM, Babson KA. Self-reported cannabis use characteristics, patterns and helpfulness among medical cannabis users. Am J Drug Alcohol Abuse. 2014;40(1):23–30. [DOI] [PubMed] [Google Scholar]
- 10. Gorelick DA. Cannabis-related disorders and toxic effects. Hardin CC, editor. N Engl J Med. 2023;389(24):2267–75. [DOI] [PubMed] [Google Scholar]
- 11. Mills L, Lintzeris N, O’Malley M, Arnold JC, McGregor IS. Prevalence and correlates of cannabis use disorder among Australians using cannabis products to treat a medical condition. Drug Alcohol Rev. 2022;41(5):1095–108. [DOI] [PubMed] [Google Scholar]
- 12. Millar SR, Mongan D, Smyth BP, Perry IJ, Galvin B. Relationships between age at first substance use and persistence of cannabis use and cannabis use disorder. BMC Public Health. 2021;21(1):997. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol. 2008;61(4):344–9. [DOI] [PubMed] [Google Scholar]
- 14. Jeffers AM, Glantz S, Byers A, Keyhani S. Sociodemographic characteristics associated with and prevalence and frequency of cannabis use among adults in the US. JAMA Netw Open. 2021;4(11):e2136571. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Holm S. A simple sequentially rejective multiple test procedure. Scand J Stat. 1979;6(2):65–70. [Google Scholar]
- 16. Dawson D, Stjepanović D, Lorenzetti V, Cheung C, Hall W, Leung J. The prevalence of cannabis use disorders in people who use medicinal cannabis: a systematic review and meta-analysis. Drug Alcohol Depend. 2024;257:111263. [DOI] [PubMed] [Google Scholar]
- 17. Kritikos AF, Taylor B, Lamuda P, Pollack H, Schneider JA. Patterns of past month cannabis consumption and cannabis use disorder-insights from a nationally representative survey. Drug Alcohol Depend. 2025;272:112680. [DOI] [PubMed] [Google Scholar]
- 18. Lapham GT, Matson TE, Bobb JF, Luce C, Oliver MM, Hamilton LK, et al. Prevalence of cannabis use disorder and reasons for use among adults in a US state where recreational cannabis use is legal. JAMA Netw Open. 2023;6(8):e2328934. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Leung J, Chan GCK, Hides L, Hall WD. What is the prevalence and risk of cannabis use disorders among people who use cannabis? A systematic review and meta-analysis. Addict Behav. 2020;109:106479. [DOI] [PubMed] [Google Scholar]
- 20. Substance Abuse and Mental Health Services Administration . Treating substance use disorder in older adults. Treatment Improvement Protocol (TIP) series No. 26, SAMHSA publication No. PEP20-02-01-011. Rockville, MD: Substance Abuse and Mental Health Services Administration; 2020. [Google Scholar]
- 21. Haug NA, Padula CB, Sottile JE, Vandrey R, Heinz AJ, Bonn-Miller MO. Cannabis use patterns and motives: a comparison of younger, middle-aged, and older medical cannabis dispensary patients. Addict Behav. 2017;72:14–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Mills L, Arnold JC, Mcgregor IS, Lintzeris N. Factors associated with cannabis use disorder among Australians using prescribed and illicitly-sourced medical cannabis. Drug Alcohol Depend Rep. 2025;16:100362. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Study surveys, procedures to request data, and contact information regarding the M3 study are available online (https://mmjoutcomes.org/m3study/). Data are available to interested researchers upon application and approval by the M3 Databank. Further enquiries can be directed to the corresponding author.