Research Article 51 Ved ABSTRACT Objective: Recreational and medical cannabis use has increased, particularly among young adults, but little is known regarding who uses for these purposes or how purpose of use is associated with problematic use. Method: We analyzed Fall 2019 survey data among 1,083 US young adults (ages 18-34) reporting past 6- month cannabis use. Multivariable regression analyses examined: 1) characteristics of those using for only/primarily medical purposes, primarily recreationally, and only recreationally vs. equally for medical and recreational purposes (referent; multinomial logistic); and 2) reasons for use in relation to cannabis use disorder symptoms (linear) and driving under the influence of cannabis (DUIC; binary logistic). Results: 37.1% used only recreationally, 23.5% primarily recreationally, 21.5% equally for both, and 17.8% medically. Compared to those using equally for medical and recreational purposes, those using only/primarily medically had fewer friends who used cannabis; those using primarily recreationally were younger, more educated, less likely used tobacco, and reported fewer ACEs. Those using only recreationally were younger, more likely male, less likely to report an ADHD diagnosis or past-month alcohol or tobacco use, and reported fewer friends who used cannabis, ACEs, and depressive symptoms. Using equally for medical and recreational purposes (vs. all other cannabis use subgroups) correlated with greater use disorder symptoms and DUIC. Conclusions: Using cannabis equally for medical and recreational purposes may pose particularly high-risk, given the association with greater mental health concerns and problematic use. Understanding use profiles and how young adults interpret and distinguish medical and recreational use is critical. Key words: = cannabis use; medical and recreational cannabis; cannabis use characteristics; risk factors; young adults Cannabis is the most commonly used federally illicit drug in the US. In 2021, past-year cannabis use prevalence was 18.7% among US individuals ages 12 or older, which was highest among those ages 18-25 (35.4%), 26 and older (17.2%), and 12- 17 (10.5%) (SAMHSA, 2021). Although cannabis is federally prohibited, as of November 2022, 21 states and 3 territories (including the District of Priyanka Sridharan1, Katelyn F. Romm2,3, & Carla J. Berg4,5 1Department of Epidemiology, Milken Institute School of Public Health, George Washington University 2TSET Health Promotion Research Center, Stephenson Cancer Center, University of Oklahoma Health Sciences Center 3Department of Pediatrics, College of Medicine, University of Oklahoma Health Sciences Center 4Department of Prevention and Community Health, Milken Institute School of Public Health, George Washington University 5George Washington Cancer Center, George Washington University Cannabis 2024, Volume 7 (2) © Author(s) 2024 researchmj.org 10.26828/cannabis/2024/000216 Use of Cannabis for Medical or Recreational Purposes Among US Young Adults: Correlates and Implications for Problematic Use and Interest in Quitting Corresponding Author: Katelyn Romm, PhD, TSET Health Promotion Research Center, Department of Pediatrics, Stephenson Cancer Center, University of Oklahoma Health Sciences Center, 655 Research Pkwy #400, Oklahoma City, OK 73104. Email: katelyn-romm@ouhsc.edu. Cannabis, A Publication of the Research Society on Marijuana 52 Columbia) have legalized recreational use, and ~40 states and 4 territories have legalized medical use (Hansen et al., 2022). Despite some potential medical benefits of cannabis use (e.g., epilepsy, multiple sclerosis, chronic pain; Banerjee & McCormack, 2019; Bilbao & Spanagel, 2022), cannabis use poses potential negative consequences, especially for young people (Hall & Lynskey, 2020), including impaired memory and attention, decreased motivation and productivity, mental health problems (e.g., anxiety, depression, psychosis), increased risk of addiction (Stuyt, 2018), and driving under the influence of cannabis (DUIC) and related motor vehicle accidents (Azofeifa et al., 2019). As legalized recreational and medical cannabis has expanded in the US, the use of cannabis for both recreational and medical purposes has become increasingly prevalent among young adults (Schauer, 2021). Although cannabis is predominantly used recreationally, it is increasingly used for relief from various physical and mental health conditions (Leung et al., 2022; Lin et al., 2016; Paul et al., 2020). For example, a 2016 analysis of data from a nationally-representative US sample found that, in states with medical cannabis legislation, 17% of individuals used cannabis for medical reasons, while 83% used it recreationally (Lin et al., 2016). A 2018 analysis of nationally-representative data from adults in the US and Canada indicated that 27% had ever used cannabis for medical purposes (Leung et al., 2022). It is important to understand who uses cannabis for medical vs. recreational purposes and the potential profiles of use that might entail the highest risk, for example, in terms of addiction, long-term chronic use (vs. quitting), and high-risk behaviors like driving under the influence. In terms of correlates of use for medical or recreational purposes, the existing literature is limited. One prior study using nationally- representative data indicated greater odds of medical use among men vs. women and those ages 26-35 vs. other age groups (Leung et al., 2022). Related to markers of addiction, although the majority of those using cannabis do not experience signs of addiction, recent epidemiological patterns indicate a growing population living with some form of cannabis use disorder (Compton et al., 2019). One study found that, among US adults using cannabis in a nationally-representative sample, past-year prevalence of cannabis use disorder was ~10% (Compton et al., 2019). Interestingly, while one national study found that individuals who used cannabis for medical reasons had a higher prevalence of daily or almost daily use (33%) compared to those who used cannabis recreationally (11%; Lin et al., 2016), other research has found that those using cannabis medically (vs. recreationally) have lower rates of problematic cannabis use and related harm (Connor et al., 2021). A particular concern is DUIC, as cannabis use increases the risk of being involved in a motor vehicle crash (Asbridge et al., 2012). In 2018, 12 million (4.7%) US adults reported DUIC in the past year, which was more prevalent among those ages 16-34 and men (Azofeifa et al., 2019). Relevant to the current study, prior analyses of nationally-representative data have documented that DUIC is more prevalent among those who use cannabis for medical vs. recreational purposes (27% vs. 9.7%; Wickens et al., 2022) and among individuals with symptoms of cannabis use disorder (Salas-Wright et al., 2021). Interest in quitting cannabis use is also an important area of investigation, as lack of interest may have implications for ongoing, chronic, and potentially escalating use. Many people who use cannabis on a regular basis are interested in reducing or quitting their use (Masters et al., 2018; McClure et al., 2019; Zvolensky et al., 2018). For example, one study of young adults using cannabis documented that 22% reported quit attempts in the past 4 months and 19% reported readiness to quit in the next month (Masters et al., 2018). Another study found that 16% of those co-using cannabis and tobacco had attempted to stop using cannabis in the previous year, and 11% intended to in the next month (McClure et al., 2019). However, cannabis cessation-related outcomes have not been investigated in relation to primary purposes for use. Notably, most of the prior literature has documented whether people have used medically or recreationally, with very little research examining how individuals describe the reasons for their current use in terms of how often they use medically vs. recreationally. Further, there is limited research characterizing individuals who use cannabis medically vs. recreationally or how reasons for use are associated with indicators of problematic use. This study aims to advance the Medical or Recreational Cannabis Use in Young Adults 53 literature by addressing these gaps. Specifically, this study examined: 1) correlates (i.e., sociodemographics, psychosocial factors, use characteristics) of young adults’ reasons for using cannabis (i.e., recreational, medical, or both); and 2) indicators of problematic use (i.e., levels of use, symptoms of dependence, DUIC, considering quitting) in relation to young adults’ reasons for use (i.e., recreational, medical, or both). METHODS Study Design The current study is an analysis of survey data among 3,006 young adults (aged 18-34) participating in a 2-year, 5-wave longitudinal cohort study, the Vape shop Advertising, Place characteristics and Effects Surveillance (VAPES) study. VAPES examines the vape retail environment and its impact on substance use, drawing participants from 6 metropolitan statistical areas MSAs (Atlanta, Boston, Minneapolis, Oklahoma City, San Diego, Seattle), selected for their variation in state tobacco control and cannabis retail legislation. This study, detailed elsewhere (Berg et al., 2020), involved survey data collection launched in Fall 2018 with assessments every 6 months for 2 years during Fall and Spring. This study was approved by the George Washington University Institutional Review Board. Participants & Recruitment Advertisements posted on Facebook and Reddit targeted eligible individuals (18-34 years old, living in one of the 6 MSAs, English speaking) using imagery, taglines, and interests that appeal to young adults. Individuals who clicked on ads were directed to a webpage with a study description and consent form, screened for eligibility, and then administered the baseline survey. Purposive sampling was used to ensure sufficient proportions of the sample represented those using e-cigarettes and cigarettes (roughly 1/3 each), both sexes, and racial/ethnic minorities. Subgroup enrollment was capped by MSA. Participants received an email 7 days after completing the baseline survey asking them to confirm their participation by clicking a “confirm” button included in an email. After confirming, participants were enrolled and emailed their first incentive (a $10 e-gift card). The duration of recruitment ranged from 87 to 104 days across MSAs. Overall, 65,843 Facebook/Reddit users viewed study ads, 10,433 clicked on ads, 9,847 consented, and 7,096 were eligible. Additionally, 2,751 were not allowed to advance to the baseline survey, with 1,427 ineligible and 1,279 not enrolled in order to reach recruitment targets of other demographics. The baseline survey was completed by 3,460 (48.8%; 51.2% partial completes, n = 3,636); 3,006 (87%) confirmed participation. The current analyses focused on Fall 2019 data (i.e., one year post baseline; n = 2,375, 79.0% response rate; compensation of a $20 e-gift card). Attrition analyses indicated that participants who did not (vs. did) complete the follow-up survey were younger, more likely male, and more likely to report past-month cannabis use at baseline (Berg et al., 2020). Measures Sociodemographic covariates. We coded MSA of residence and whether it was in a state where cannabis retail was legal (California, Massachusetts, Washington) or was not (Georgia, Minnesota, Oklahoma). Other sociodemographics included age, sex, sexual orientation, race, ethnicity, and highest level of educational attainment. Cannabis use characteristics. Participants were asked to report the number of days used in the past 6 months; those reporting any use were asked to report the number of days used in the past 30 days. Among those who reported any past 6-month use, we asked, “Do you use marijuana for medical or recreational purposes – or both: only for medical purposes, primarily for medical purposes, equally for both, primarily for recreational purposes, only for recreational purposes, I’m not sure.” Based on the distributions and limited variability, those reporting only or primarily for medical purposes were collapsed into a single category. Among those reporting past 6-month use, we also assessed age of first use (to operationalize early onset use; i.e., before age 18), number of times used per day, and whether participants held a medical cannabis card. We also asked participants how they use cannabis most of the Cannabis, A Publication of the Research Society on Marijuana 54 time: smoked (in a joint or bowl, rolled in cigar papers with or without tobacco); vaped (with a vaporizer with or without tobacco); pipe/bong (in a waterpipe or bong with our without tobacco); ingested (with or without food, drank); and other (including tinctures, dabs, etc.; Fong et al., 2006). We also administered the Cannabis Use Disorder Identification Test – Revised (CUDIT-R), an 8-item scale assessing hazardous use, with scores ranging from 0-32 with higher scores indicating more hazardous use (Adamson et al., 2010). One item from the CUDIT-R that assesses interest in quitting was also used separately (“Have you ever thought about cutting down, or stopping your use of marijuana? never; yes, but not in the past 6 months; or yes, during the past 6 months”). We recategorized participants as considered quitting (or reducing) in the past 6 months vs. others. We also asked, “During the past 30 days, how many times did you ride in a car or other vehicle driven by someone who had been using marijuana? 0, 1, 2-3, 4-5, 6 or more, or prefer not to answer.” This item was categorized as 0 vs. ≥1 time. Other substance use. Participants were asked to report number of days in the past 30 days they used: alcohol, cigarettes, e-cigarettes, little cigars/cigarillos, large cigars, hookah/waterpipe, and smokeless tobacco (NIH, 2020). Alcohol use was used as a continuous variable; use status for each tobacco product was operationalized as any vs. no use in the past 30 days and as a single aggregate variable as any vs. no use of any tobacco product in the past 30 days. Psychosocial factors. Participants were asked if a parental figure uses/used cannabis (yes/no) and how many of their 5 closest friends use cannabis (Berg et al., 2015). Depressive symptoms were assessed using the Patient Health Questionnaire – 2 item (PHQ-2; Kroenke et al., 2003), which assesses feeling down/depressed and little interest in doing things in the past 2 weeks (0 = not at all to 3 = nearly every day; summed scores of 0-6; Cronbach’s α = .87). The ACEs-10 item scale assessed maltreatment and household challenges before age 18 (0 = no, 1 = yes; range 0-10; α = .81; Felitti et al., 1998). Finally, participants were asked whether they had ever been diagnosed with ADHD. Data Analysis The current study analyzed data from 1,083 participants who reported any cannabis use in the past 6 months (i.e., since the last assessment). Participant characteristics were summarized using descriptive statistics. Chi-square and one- way ANOVA tests were used to explore differences in participant characteristics in relation to reasons for use. Then multinomial logistic regression was used to examine correlates of reasons for use, using “equally both” as our referent group. We included other substance use and psychosocial factors, as well as sociodemographic covariates that were significant in bivariate analyses (i.e., age, sex, race/ethnicity, education level). Finally, regression models were used to examine reasons for use in relation to: 1) number of days used (linear regression); 2) CUDIT scores (linear regression); 3) DUIC (binary logistic regression); and 4) recently considering quitting (binary logistic regression). In these models, we accounted for age, sex, race/ethnicity, and education level. Regression analyses were also conducted using multilevel modeling to account for the hierarchical structure of the data (i.e., young adults at the individual level nested within MSA; Aveyard, Markham, & Cheng, 2004; Aveyard, Markham, Lancashire, et al., 2004). However, all intra-class correlations were approximately .01, and findings were not significantly different. All analyses were conducted using SPSS (version 26.0) and alpha set at .05. RESULTS Participant Characteristics In this sample of 1,083 participants who reported past 6-month cannabis use, the average age was 24.46 (SD = 4.64), 57.1% lived in states with legalized recreational cannabis, 41.9% were male, 40.3% were sexual minorities, 24.1% were non-White, 12.7% were Hispanic, and 72.9% possessed at least a Bachelor’s degree. In this sample, 37.1% used only for recreational purposes, 23.5% for primarily recreational purposes, 21.5% equally for medical and recreational purposes, and 17.8% for only/primarily medical purposes. Of the 1,083 young adults reporting past 6-month use, 79.3% (n = 859) also reported using in the past 30 days (M days of use = 10.59, SD = 11.38), and the average CUDIT score was 7.45 (SD = 5.68). Overall, 24.3% reported past 30-day DUIC, and Medical or Recreational Cannabis Use in Young Adults 55 30.9% reported considering past 6-month quit attempts. Correlates of Reasons for Use Bivariate analyses (Table 1) indicated that there were differences in the proportions of young adults using cannabis who reported different purposes of use in relation to MSA, age, sex, race, ethnicity, and education level (p’s < .05; see Table 1 for significant post-hoc differences). Regarding cannabis use characteristics, those who used equally for medical and recreational purposes reported the greatest number of days used and times used per day; those using only for recreational purposes reported the least (p’s < .001). Those using only for recreational purposes were also the least likely to report early onset use (p’s < .001) but the most likely to report never trying to quit (p < .001); they also reported the greatest number of days of alcohol use (p = .006). Those using only for medical purposes were most likely to report having a medical cannabis card; those using only recreational were the least likely (p < .001). Those using only for recreational purposes were the least likely to report using via pipe/bong, but were the most likely to report ingesting cannabis as their most common mode of use (p < .001). Those using primarily for recreational purposes were the most likely to report parental use of cannabis and lifetime diagnosis of ADHD (p’s < .001). Multinomial logistic regression analyses (Table 2) indicated that compared to those who used equally for medical and recreational purposes (referent group), those who reported using for only/primarily medical purposes were less likely to live in Boston, Minneapolis, Seattle, or “other” MSA (vs. Oklahoma City, p’s < .05) and had fewer friends who used cannabis (p = .003). Those who reported using primarily for recreational purposes (vs. equally for medical and recreational purposes) were more likely to live in any other of the MSAs (except Boston) vs. Oklahoma City (p’s < .05), were younger (p < .001), more likely to have at least a bachelor’s degree (p = .042), reported fewer ACEs (p = .011), and were less likely to report past-month tobacco use (p < .001). Those who reported using only for recreational purposes (vs. equally for medical and recreational purposes) were more likely to live in Atlanta, Boston, or Minneapolis (vs. Oklahoma City, p’s < .05), were younger (p < .001), were more likely male (p = .018), had fewer friends who used cannabis (p < .001), reported fewer ACEs (p < .001) and fewer depressive symptoms (p = .008), were less likely to report an ADHD diagnosis (p = .010), used alcohol on more days in the past month (p = .015), and were more likely to report past- month tobacco use (p < .001). Reasons for Use in Relation to Use Frequency, Dependency, DUIC and Interest in Quitting Regression models examining reasons for cannabis use in relation to the number of days used, CUDIT scores, DUIC, and interest in quitting in the past 6 months among those reporting past 6-month cannabis use are shown in Table 3. Using equally for medical and recreational purposes (vs. all other subgroups of cannabis use) correlated with more days of use, greater CUDIT scores, and greater odds of DUIC (p’s < .011). Additional correlates included: being male (p = .027) and lower education (p < .001) for days of use; living in any other of the MSAs except Atlanta or Boston (vs. Oklahoma City, p’s < .05), being younger (p = .042), being male (p < .001), and lower education (p < .001) for CUDIT scores; and being male (p < .001) or White (vs. Black; p = .027) for DUIC. Correlates of considering quitting cannabis included living in San Diego or Seattle (vs. Oklahoma City, p’s < .05), being younger (p < .001), female (p = .002), and using equally for medical and recreational purposes vs. only recreational purposes (p < .001). DISCUSSION In this sample of US young adults ages 18-34 reporting past 6-month cannabis use, over one- third (~37%) used only recreationally, while only about one-fifth used primarily recreationally (~23%), equally for both (~21%), and only/primarily medically (~18%). Cannabis, A Publication of the Research Society on Marijuana 56 Table 1. Correlates of reasons for cannabis use among young adults using cannabis within the past 6 monthsin Fall 2019, N=1,083 * Total N=1,083 (100%) Only/primarily medical N=193 (17.8%) Equally both N=233 (21.5%) Primarily recreational N=255 (23.5%) Only recreational N=402 (37.1%) Variable N (%) or M (SD) N (%) or M (SD) N (%) or M (SD) N (%) or M (SD) N (%) or M (SD) p MSA, N (%) <.001 Atlanta 163 (15.1) 33 (17.1) 25 (10.7) 39 (15.3) 66 (16.4) Boston 227 (21.0) 32 (16.6) 46 (19.7) 51 (20.0) 98 (24.4) Minneapolis-St. Paul 191 (17.6) 17 (8.8)a 45 (19.3)b 46 (18.0)b 83 (20.6)b Oklahoma City (ref) 90 (8.3) 42 (21.8)a 22 (9.4)b 8 (3.1)c 18 (4.5)c,d San Diego 161 (14.9) 37 (19.2) 31 (13.3) 43 (16.9) 50 (12.4) Seattle 213 (19.7) 25 (13.0)a 53 (22.7)a,b 61 (23.9)b 74 (18.4)a,b Other 38 (3.5) 7 (3.6) 11 (4.7) 7 (2.7) 13 (3.2) Cannabis retail law, N (%) .155 Legalized 617 (57.1) 99 (51.3) 134 (57.5) 157 (62.1) 227 (56.6) Not legalized 463 (42.9) 94 (48.7) 99 (42.5) 96 (37.9) 174 (43.4) Sociodemographics Age, M (SD) 24.46 (4.64) 26.51 (4.82)a 25.35 (4.82)b 23.67 (4.38)c 23.45 (4.18)c <.001 Male, N (%)** 439 (41.9) 62 (34.1)a 91 (41.4)a,b 96 (38.6)a,b 190 (47.9)b .009 Sexual minority, N (%) 436 (40.3) 80 (41.5) 97 (41.6) 113 (44.3) 146 (36.3) .201 Race, N (%) .048 White 813 (75.1) 144 (74.6) 185 (79.4) 189 (74.1) 295 (73.4) Black 39 (3.6) 7 (3.6) 8 (3.4) 7 (2.7) 17 (4.2) Asian 98 (9.0) 13 (6.7)a,b 9 (3.9)b 28 (11)a 48 (11.9)a Other 133 (12.3) 29 (15.0) 31 (13.3) 31 (12.2) 42 (10.4) Hispanic, N (%) 138 (12.7) 26 (13.5) 34 (14.6) 32 (12.5) 46 (11.4) .699 ≥Bachelor’s degree, N (%) 789 (72.9) 121 (62.7)a 153 (65.7)a,b 195 (76.5)b,c 320 (79.6)c <.001 Cannabis use characteristics Number of days used, past 30 days, M (SD) 10.59 (11.38) 12.36 (11.88)a 17.15 (11.78)b 11.79 (10.97)a,c 5.19 (8.30)d <.001 Early onset use (<18), N (%) 555 (55.7) 104 (59.8)a 147 (69.0)a 142 (58.2)a 162 (44.4)b <.001 Times used per day, M (SD) 2.43 (2.69) 2.87 (3.21)a 3.90 (3.84)b 2.13 (1.91)c 1.57 (1.31)d <.001 Has medical card, N (%) 122 (11.7) 57 (32.6)a 46 (20.5)b 13 (5.2)c 6 (1.5)d <.001 Most common mode of use, N (%) <.001 Smoked 456 (42.3) 67 (35.1) 108 (46.4) 103 (40.4) 178 (44.7) Vaped 242 (22.5) 48 (25.1) 52 (22.3) 57 (22.4) 85 (21.4) Pipe/bong 173 (16.1) 34 (17.8)a 47 (20.2)a 57 (22.4)a 35 (8.8)b Ingested 182 (16.9) 36 (18.8)a,b 20 (8.6)c 35 (13.7)b,c 91 (22.9)a Other 24 (2.2) 6 (3.1) 6 (2.6) 3 (1.2) 9 (2.3) CUDIT score, M (SD) 7.45 (5.68) 7.33 (5.45)a 10.03 (5.72)b 8.68 (5.73)c 5.24 (4.86)d <.001 Medical or Recreational Cannabis Use in Young Adults 57 Drove under influence, N (%) 255 (24.3) 45 (24.3)a 95 (43.8)b 74 (29.7)a 41 (10.3)c <.001 Considered cutting down or quitting, past 6 months, N (%) <.001 Never 559 (53.2) 99 (52.7)a 100 (44.6)a,b 97 (38.8)b 263 (67.6)c Yes, but not in the past 6 months 167 (15.9) 45 (23.9)a 40 (17.9)a 45 (18.0)a 37 (9.5)b Yes, in the past 6 months 325 (30.9) 44 (23.4)a 84 (37.5)b 108 (43.2)b 89 (22.9)a Other substance use Number of days of alcohol use, past 30 days, M (SD) 6.83 (6.54) 5.60 (6.72)a 6.65 (7.03)a,b 6.72 (6.13)a,b 7.59 (6.33)b .006 Past-month tobacco use, N (%) Cigarettes 363 (33.5) 80 (41.5)a,b 106 (45.5)b 79 (31.0)a,c 98 (24.4)c <.001 E-cigarettes 506 (46.7) 110 (57.0)a,b 138 (59.2)b 114 (44.7)a,c 144 (35.8)c <.001 Little cigars/cigarillos 152 (14.0) 44 (22.8)a 48 (20.6)a 42 (16.5)a 18 (4.5)b <.001 Large cigars 98 (9.0) 25 (13.0) 24 (10.3) 16 (6.3) 33 (8.2) .080 Hookah 125 (11.5) 31 (16.1) 26 (11.2) 32 (12.5) 36 (9.0) .079 Smokeless tobacco 40 (3.7) 10 (5.2) 12 (5.2) 7 (2.7) 11 (2.7) .234 Any tobacco 664 (61.3) 136 (70.5)a 175 (75.1)a 147 (57.6)b 206 (51.2)b <.001 Psychosocial factors Parental use of cannabis, N (%) 264 (24.4) 60 (31.1)a 72 (30.9)a 74 (29.0)a 58 (14.4)b <.001 Number of friends using cannabis, M (SD) 3.20 (1.49) 3.06 (1.58)a 3.60 (1.41)b 3.46 (1.34)b,c 2.86 (1.49)a,d <.001 Depressive symptoms, M (SD) 1.78 (1.74) 1.96 (1.85) a 2.15 (1.98)a 1.92 (1.65) a 1.40 (1.50)b <.001 ACEs, M (SD) 2.50 (2.48) 3.60 (2.77)a 3.24 (2.53)a 2.37 (2.34)b 1.62 (2.01)c <.001 Lifetime diagnosis of ADHD, N (%) 144 (13.3) 30 (15.5)a 39 (16.7)a 43 (16.9)a 32 (8.0)b .001 Note. p-values indicate omnibus tests (per ANOVA and Chi-Square) across modes of use. Bolded and italicized values indicate statistical significance at p < .05. Different superscripts denote statistically significant differences between groups at p < .05. * Excluding those who report “not sure” for purpose of use (N = 15). ** 87 reported “other” sex. Cannabis, A Publication of the Research Society on Marijuana 58 Table 2. Multinomial logistic regression examining correlates of reasons for cannabis use among young adults using cannabis in the past 6 months (referrent: use cannabis equally for medical and recreational purposes) * Only or primarily medical Primarily recreational Only recreational Variable aOR CI p aOR CI p aOR CI p MSA (ref: Oklahoma City) Atlanta 0.73 0.34-1.60 .435 3.35 1.23-9.07 .018 2.64 1.11-6.24 .027 Boston 0.39 0.19-0.81 .012 2.58 1.00-6.63 .050 2.50 1.12-5.58 .025 Minneapolis 0.22 0.10-0.50 <.001 2.63 1.02-6.81 .046 2.31 1.02-5.21 .044 San Diego 0.70 0.33-1.48 .346 3.43 1.29-9.16 .014 1.69 0.71-4.02 .233 Seattle 0.24 0.11-0.51 <.001 3.12 1.23-7.95 .017 1.92 0.85-4.30 .115 Other 0.29 0.09-0.98 .046 1.43 0.39-5.23 .587 1.14 0.36-3.61 .824 Sociodemographics Age 1.04 0.99-1.09 .057 0.90 0.87-0.94 <.001 0.88 0.84-0.91 <.001 Female (ref: male)** 1.48 0.95-2.31 .081 1.00 0.67-1.50 .995 0.63 0.43-0.92 .018 Race (ref: White) Black 0.97 0.31-2.99 .955 1.15 0.36-3.70 .812 2.04 0.75-5.59 .164 Asian 2.21 0.87-5.61 .097 2.15 0.95-4.89 .067 2.07 0.93-4.58 .074 Another race 1.13 0.61-2.09 .710 1.01 0.56-1.81 .976 0.90 0.50-1.60 .715 Hispanic (ref: non-Hispanic) 0.87 0.47-1.63 .670 0.82 0.46-1.48 .512 0.90 0.51-1.60 .721 ≥Bachelor’s degree (ref: