UvA-DARE (Digital Academic Repository)

UvA-DARE (Digital Academic Repository) Mental health differences between frequent cannabis users with and without dependence and the general population van der Pol, P.; Liebregts, N.; de Graaf, R.; ten Have, M.; Korf, D.J.; van den Brink, W.; van Laar, M. Published in: Addiction DOI: 10.1111/add.12196 Link to publication Citation for published version (APA): van der Pol, P., Liebregts, N., de Graaf, R., ten Have, M., Korf, D. J., van den Brink, W., & van Laar, M. (2013). Mental health differences between frequent cannabis users with and without dependence and the general population. Addiction, 108(8), 1459-1469. DOI: 10.1111/add.12196 General rights It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons). Disclaimer/Complaints regulations If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: http://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible. UvA-DARE is a service provided by the library of the University of Amsterdam (http://dare.uva.nl) Download date: 27 Jan 2018

bs_bs_banner RESEARCH REPORT doi:10.1111/add.12196 Mental health differences between frequent cannabis users with and without dependence and the general population Peggy van der Pol, 1 Nienke Liebregts, 2 Ron de Graaf, 1 Margreet ten Have, 1 Dirk J. Korf, 2 Wim van den Brink, 3 & Margriet van Laar 1 Netherlands Institute of Mental Health and Addiction, Utrecht, the Netherlands, 1 Bonger Institute of Criminology, Law Faculty, University of Amsterdam, Amsterdam, the Netherlands 2 and Department of Psychiatry, Academic Medical Centre, University of Amsterdam, Amsterdam, the Netherlands 3 ABSTRACT Aims To compare the prevalence of mental disorders between frequent cannabis users with and without dependence and the general population. Design Cross-sectional comparison of interview data. Setting Enriched community sample of frequent cannabis users and a representative sample of non-users and non-frequent users from the general population. Participants A total of 521 young adult (aged 18 30 years) frequent cannabis users, 252 of whom were with DSM-IV cannabis dependence (D+) and 269 without DSM-IV cannabis dependence (D-), and 1072 young adults from the general population. Measurements Multinomial logistic regression was used to compare groups regarding the presence of DSM-IV mental disorders. Detailed measures of cannabis use, childhood adversity and other substance use were considered confounders. Findings Compared with the general population, externalizing disorders were more prevalent in D- [odds ratio (OR) = 8.91, P < 0.001] and most prevalent in D+ (OR = 17.75, P < 0.001), but internalizing disorders were associated only with D+ (mood OR = 4.15, P < 0.001; anxiety OR = 2.20, P = 0.002). Associations were attenuated (and often became non-significant) after correction for childhood adversity and substance use other than cannabis. However, the prevalence of mental disorders remained higher in D+ compared with D- (OR = 2.40, P < 0.001), although cannabis use patterns were remarkably similar. Conclusions Cannabis use patterns, childhood adversity and the use of other substances are similar in dependent and non-dependent frequent cannabis users. With the exception of more externalizing disorders, the mental health condition of nondependent frequent cannabis users is similar to that of the general population, whereas it is worse in dependent frequent cannabis users. Keywords Dependence, frequent cannabis use, general population, mental disorders. Correspondence to: Peggy van der Pol, Netherlands Institute of Mental Health and Addiction, PO Box 725, 3500 AS Utrecht, the Netherlands. E-mail: ppol@trimbos.nl Submitted 6 June 2012; initial review completed 14 August 2012; final version accepted 18 March 2013 INTRODUCTION Both frequent (daily or nearly daily) cannabis use and cannabis dependence are associated with various mental health problems [1 7]. However, in most studies the overlap (an estimated 20 50% [8,9]) between dependent users and frequent users is ignored [1 5,10 16]. Associations between mental health problems and nondependent frequent use are therefore unknown. As a consequence, the mental health risks based on studies of all frequent users together may be overestimated, because the subgroup of dependent frequent users may magnify this association. Most information on the association between frequent or dependent cannabis use and mental disorders is derived from general population studies. These studies show higher rates of depression and anxiety in dependent versus non-dependent users and in heavy versus infrequent users [3,6,10 16]. However, other studies found similar rates of internalizing disorders in heavy and infrequent users [6,17 19]. On one hand, these inconsistent findings may be due to small sample size, imprecise and

2 Peggy van der Pol et al. inconsistent measurement of cannabis use or divergence in measures of mood and anxiety problems and disorders [10,11]. On the other hand, they may indicate a variation in the percentages of dependent users, suggesting that associations between internalizing disorders and heavy use without dependence are small. More robust associations were found between externalizing disorders and both cannabis use and dependence, especially with regard to attention deficit and hyperactivity disorder (ADHD) and conduct disorder (CD) [12,14,20,21]. Nevertheless, as general population studies either investigate frequency of use or dependence, a direct comparison between frequent cannabis users with and without dependence is lacking. Moreover, general population studies do not allow a comparison between non-dependent frequent users and the general population. Therefore, it remains unknown to what extent the presence of specific internalizing and externalizing mental disorders differs between dependent and non-dependent frequent users and the general population. To our knowledge, there is only one study on differences in mental health between dependent and nondependent frequent cannabis users [22]. In this convenience sample (using mailing lists of organizations committed to altering drug laws), dependent frequent cannabis users reported higher levels of depression than non-dependent frequent users. In addition, dependent users were significantly younger, less educated and, although both groups were daily users, dependent users reported feeling higher on average during cannabis consumption, used more joints weekly and more often used other substances. However, these results were derived from a web-based survey, using proxy measures of dependence and depression, rather than valid standardized instruments measuring (DSM-IV) diagnoses. Use of other drugs has been related to both mental health problems and cannabis use or dependence, and attenuates these associations [5,7,23 26]. This is also true for childhood adversities [27,28]. In particular, it has been suggested that the development of cannabis abuse and dependence symptoms are influenced by childhood maltreatment both directly and through the development of externalizing (but not internalizing) disorders [29]. Although these factors do not negate mental health differences, they should be considered as alternative explanatory factors for mental health differences between these groups. Moreover, some investigators have suggested that a detailed assessment of cannabis exposure could improve our understanding of its association with mental disorders [30 32]. Usage patterns may vary widely also among frequent users. Hence, exposure may vary from one joint several days per week to daily consumption of numerous joints with a high dose of very potent cannabis. However, most studies measure only frequency of use, sometimes complemented with the number of units (often joints) per day, and neglect differences in dose per joint, cannabis potency and type (marijuana or hashish). Other cannabis factors shown to differ between dependent and non-dependent users that may also be associated with mental health problems include setting of use (e.g. solitary use and time of use) [33] and motives for cannabis use [34,35]. In particular, using cannabis to cope with negative affect has been identified as a mediator between dependence and internalizing disorders [36]. General population studies mainly ignore the distinction between dependent and non-dependent frequent cannabis use. Therefore, the primary aim of the present study was to compare the presence of DSM-IV mental disorders in three groups: young adult dependent frequent cannabis users, young adult non-dependent frequent cannabis users and young adult subjects from the general population. In these respective populations, a decreasing gradient in the prevalence of mental disorders was hypothesized [5,37]. Childhood adversity and substance use were expected to attenuate the mental health differences between the three groups. The second aim was to investigate whether differences in cannabis use patterns and habits, assessed in detail in dependent and non-dependent frequent users, may explain mental health differences between both user groups. It was hypothesized that dependent users would exhibit higher levels of cannabis exposure and more risky consumption habits [22] and that these factors would explain, at least partially, the mental health differences between the two groups. METHODS Subjects Data on frequent dependent and non-dependent cannabis users were gathered from the baseline wave of a cohort study on predictors of transitions in frequent cannabis use and dependence (CanDep). For a detailed description of the study design see Van der Pol et al. [38]. In brief, 600 cannabis users were recruited from coffee shops and through chain referral between September 2008 and April 2009. Most Dutch cannabis users buy cannabis in designated coffee shops, where the sale and use of small amounts of cannabis are officially tolerated [39]. Inclusion was restricted to those aged 18 30 years who used cannabis on 3 or more days per week for the past 12 months. Dependent frequent cannabis users (n = 252) were distinguished from non-dependent frequent cannabis users (n = 348) based on the 12-month DSM-IV diagnosis of cannabis dependence at baseline,

(Dependent) heavy cannabis use and mental health 3 according to the Composite International Diagnostic Interview (CIDI) 3.0 (see below). Seventy-eight of the 348 frequent cannabis users without a 12-month diagnosis of cannabis dependence had a life-time diagnosis of dependence. They were excluded from the analyses because their inclusion might attenuate the differences between groups. Data on the general population were derived from a Dutch representative psychiatric epidemiological general population survey, NEMESIS-2 (Netherlands Mental Health Survey and Incidence Study-2) [38,40]. Through a multi-stage, stratified random sampling procedure, 6646 participants between the ages of 18 and 64 years were recruited between November 2007 and July 2009. There were 1118 respondents aged 18 30 years, of whom 24 were excluded because they had used cannabis frequently ( 3 days per week for the past 12 months), but lacked the detailed cannabis exposure assessment allowing them to be included in the CanDep sample. Twenty-two NEMESIS-2 respondents who completed an abbreviated interview without childhood and substance use variables were excluded, as well as one CanDep respondent with missing data regarding the presence of mental disorders. The total sample comprised 1593 participants: 521 from the CanDep study (252 dependent and 269 non-dependent frequent cannabis users) and 1072 non-users and non-frequent users from NEMESIS-2. Instruments and assessments Measures in the CanDep and NEMESIS-2 samples converged to a large extent: they included the CIDI 3.0 for diagnoses of mental disorders and additional scales, which are discussed in detail in Van der Pol et al. [38]. Data were collected through comprehensive face-to-face computer-assisted interviews conducted by trained lay interviewers. The interviews took place at the participant s home (29.4%), the research institute (20.7%), in a coffee shop (12.9%), a café (23.0%) or a public venue such as a park or library (14.0%). The choice of location was made to suit the participant, as long as a confidential conversation was feasible. The interview required the participants to be fluent in Dutch. Socio-demographic factors Socio-demographic variables shown empirically to influence cannabis dependence included the following: sex, age, ethnicity (western/non-western), educational level (primary or lower secondary/higher secondary/ higher professional/university), urbanicity (urban/ rural), employment situation (employed/unemployed or unable to work/student) and living arrangements (living alone/with others) [7,12,26]. Mental disorders The following DSM-IV 12-month diagnoses were assessed with the CIDI 3.0: cannabis dependence, major depression, bipolar disorder, dysthymia, social phobia, panic disorder, generalized anxiety disorder, agoraphobia and adult ADHD. In addition, antisocial personality disorder (ASP), childhood ADHD and CD were assessed with the CIDI 3.0. Childhood adversities Participants were asked if they had experienced the following life events before the age of 16 years: being from a single-parent family, parental divorce and parental death. Additionally, using a list of examples, participants were asked whether they had experienced emotional, psychological, physical and/or sexual abuse [27,41,42]. The examples included parental ignorance of one s problems and experiences, being subordinated to siblings, being beaten up or being touched or having to touch someone in a sexual way. This instrument has been employed in numerous studies [43 45]. As these experiences may not always occur in isolation, the cumulative number of these seven childhood adversities was also reported. Substance use For all participants, age of onset and use in the past 12 months were assessed for cannabis, ecstasy (XTC), cocaine and (weekly or more frequent) alcohol use. Cannabis use patterns and habits Detailed cannabis use was assessed in the CanDep sample and included the self-reported number of cannabis use days in the past 4 weeks, number of joints per typical cannabis use day, preferred cannabis type (marijuana/hashish/no preference), subjective estimates of preferred cannabis potency (low/middle/high), average level of intoxication during cannabis consumption on a 1 10 visual analogue scale, dosage of cannabis per joint based on the estimated number of joints made from 1 g of cannabis, cannabis use location (home/ coffee shop/non-selective) and solitary use most of the time (yes/no). Information on timing of use included a dichotomous assessment of using [also] on weekdays and [also] in daytime, and the estimated average number of hours high on a cannabis using day. Furthermore, age of onset of regular use (at least once monthly) was assessed. Motives for cannabis use In the CanDep sample, motives for cannabis use were assessed with the Marijuana Motives Measure (MMM)

4 Peggy van der Pol et al. [46]. This 25-item questionnaire consists of five internally consistent subscales: enhancement (e.g. because it is exciting ; Cronbach s a=0.66), conformity (e.g. to fit in with the group I like ; a=0.74), expansion (e.g. to expand my awareness ; a=0.85), coping (e.g. to forget my worries ; a=0.85) and social (e.g. because it makes social gatherings more fun ; a=0.75). Analysis Socio-demographic factors were compared for nondependent cannabis users, dependent cannabis users and the general population in a multi-variable multinomial logistic regression model with the grouping variable as dependent variable. The general population was used as the reference. To allow comparison between dependent and non-dependent frequent users, non-dependent users were also defined as the reference group in the same models. This multinomial logistic regression procedure was also used to compare 12-month mental disorders between the three groups. Similarly, childhood adversity and substance use were compared. Next, comparisons of mental health between all three groups were adjusted for the number of childhood adversity factors and current substance use (tobacco, alcohol, illicit substances) [23,42]. Similarly, dependent and non-dependent frequent cannabis users were compared regarding detailed measures of cannabis use (exposure, setting and motives for use) with logistic regression. Mental health in these two groups was then compared adjusted for cannabis use variables that differed between the CanDep samples. All analyses were computed in STATA version 11.1 and were adjusted for socio-demographic factors. 1 STATA produces relative risk ratios in multinomial regression analyses, which can be interpreted as odds ratios (ORs) [47]. The NEMESIS-2 data were weighted to be representative for the general Dutch population. Moreover, multiplication of participants with similar characteristics is a risk of both chain referral (CanDep) and stratified samples based on municipalities (NEMESIS-2). Therefore, analyses allowed for clustering within CanDep chains of referral and within NEMESIS-2 municipalities. To account for multiple comparisons, a P-value 0.005 was considered statistically significant. As missing data were limited, complete case analyses were employed (itemized missing values are reported in footnotes to the Tables). RESULTS Socio-demographics No statistically significant (multivariate) differences in socio-demographic variables were observed between dependent and non-dependent users (Table 1) [38]. 2 However, both groups were very different from the general population sample, except for living alone. Mental disorders Compared to the general population sample, both dependent and non-dependent frequent cannabis users were at higher risk of having any mental disorder (Table 1). However, the risk was significantly and substantially higher for dependent (OR = 14.97, P < 0.001) than non-dependent users (OR = 5.35, P < 0.001). Analysis of specific disorders showed that non-dependent frequent users were only at significantly higher risk of having externalizing, not internalizing disorders. In contrast, dependent cannabis users showed higher risk of having both internalizing and externalizing disorders. Besides the higher risk of ADHD in frequent cannabis users compared to the general population, additional analyses showed that frequent cannabis users also had higher persistency rates into adulthood (91% versus 70% persistence; OR = 10.1, P = 0.007). Dependent frequent users had mental disorders more often than non-dependent cannabis users, both internalizing as well as (childhood and adult) externalizing disorders (Table 1). CD and ASP were the most prevalent disorders in both groups but, overall, ORs were larger for internalizing compared to externalizing disorders. Childhood and substance use Several potential confounders for mental health differences were considered. Generally, both frequently using groups experienced more childhood adversity and used more substances than the general population sample. However, the loss of parent(s) and childhood sexual abuse did not differ (Table 1). Moreover, frequent cannabis users reported less weekly (or more frequent) alcohol use than the general population sample. Remarkably, the frequently using groups were mostly similar, although dependent users reported use of XTC or cocaine and a history of childhood physical abuse more often compared to non-dependent frequent cannabis users. After adjustment for these variables, patterns in associated mental disorders changed considerably in the 1 Matching was considered, but socio-demographic characteristics of the samples did not allow this strategy. As the CanDep sample was young and mainly male, there were insufficient males in their early 20s in the NEMESIS-2 sample. Therefore, all analyses were adjusted statistically for socio-demographic variables. 2 Odds ratios of continuous variables represent the percentage increase in the odds of being in one versus the reference population: e.g. for every year a frequent user has used cannabis, the probability to be dependent increases with 11% (Table 3).

(Dependent) heavy cannabis use and mental health 5 Table 1 Comparison between dependent and non-dependent frequent cannabis users and the general population: socio-demographic variables, 12-month disorders, childhood adversity, and substance use. N2 a D- D+ D- versus N2 D+ versus N2 D+ versus D- n % n % n % OR P OR P OR P Male 464 48.7 206 76.6 200 79.4 4.04 <0.001 4.66 <0.001 1.15 0.55 Mean age (SE) 24.0 (0.16) 21.9 (0.23) 22.1 (0.24) 0.78 <0.001 0.81 <0.001 1.04 0.29 Western (versus non-western) 973 88.0 184 68.4 188 74.6 0.40 0.005 0.57 0.08 1.43 0.08 Education primary/lower secondary 208 34.0 81 30.1 67 26.6 1 1 1 Higher secondary 441 42.6 104 38.7 95 37.7 1.40 0.21 1.49 0.14 1.06 0.81 Higher professional 423 23.4 84 31.2 90 35.7 3.06 0.001 3.33 <0.001 1.09 0.74 Employment employed 732 65.4 122 45.4 88 34.9 1 1 1 Student 268 27.3 114 42.4 121 48.0 0.66 0.13 1.07 0.79 1.62 0.03 Unemployed/unable to work 72 7.3 33 12.3 43 17.1 2.74 0.001 5.60 <0.001 2.05 0.03 Living alone (versus with others) 273 18.3 54 20.1 53 21.0 1.22 0.43 1.19 0.51 0.98 0.92 Urban (versus town/village) 542 55.2 247 91.8 236 93.7 8.02 <0.001 10.42 <0.001 1.30 0.42 Any 12-month disorder 184 19.7 146 54.3 189 75.0 5.35 <0.001 14.97 <0.001 2.80 <0.001 Any mood disorder 72 8.0 17 6.3 46 18.3 1.18 0.67 4.15 <0.001 3.52 <0.001 Major depressive disorder 57 5.6 11 4.1 30 11.9 1.43 0.39 4.55 <0.001 3.17 0.006 Bipolar disorder 14 2.3 6 2.2 16 6.4 0.90 0.87 2.96 0.02 3.28 0.02 Dysthymia 8 0.9 2 0.7 7 2.8 1.67 0.57 5.62 0.008 3.37 0.13 Any anxiety disorder 72 7.1 11 4.1 38 15.1 0.58 0.13 2.20 0.002 3.81 <0.001 Social phobia 45 4.5 7 2.6 27 10.7 0.71 0.47 3.35 0.001 4.72 0.002 Panic disorder 14 1.1 2 0.7 12 4.8 2.23 0.37 15.11 <0.001 6.77 0.004 General anxiety disorder 25 2.4 2 0.7 8 3.2 0.24 0.10 1.24 0.71 5.06 0.07 Agoraphobia 3 0.2 0 0 4 1.6 Adult ADHD 22 2.2 18 6.7 32 12.7 3.49 <0.001 7.26 <0.001 2.08 0.004 ASP 34 4.5 43 16.0 84 33.3 3.33 <0.001 9.98 <0.001 3.00 <0.001 Any childhood disorder 73 9.3 120 44.6 147 58.3 8.91 <0.001 17.75 <0.001 1.99 <0.001 Childhood ADHD 30 3.2 20 7.4 35 13.9 2.52 0.005 5.13 <0.001 2.03 0.005 Conduct disorder 54 7.9 113 42.0 142 56.4 9.47 <0.001 19.99 <0.001 2.11 <0.001 Number childhood adversities (SE) 1.52 (0.04) 2.07 (0.08) 2.40 (0.10) 1.66 <0.001 2.11 <0.001 1.27 0.006 Not raised by both parents 90 9.5 68 25.3 68 27.0 2.80 0.003 3.24 0.001 1.16 0.49 Parents divorced 184 19.2 97 36.3 96 38.1 1.97 0.002 2.20 <0.001 1.12 0.55 Loss of parent(s) 38 3.8 22 8.2 23 9.1 1.86 0.12 2.16 0.05 1.16 0.66 Emotional abuse 124 11.9 73 27.1 96 38.1 2.90 <0.001 5.15 <0.001 1.78 0.02 Psychological abuse 174 14.8 105 39.0 119 47.2 4.31 <0.001 6.32 <0.001 1.47 0.04 Physical abuse 73 6.4 40 14.9 68 27.0 2.74 <0.001 6.65 <0.001 2.43 0.004 Sexual abuse 61 5.6 17 6.3 19 7.5 1.67 0.26 2.05 0.14 1.23 0.60 Last month tobacco use 362 34.4 211 78.4 211 83.7 9.17 <0.001 13.64 <0.001 1.49 0.05 12-month weekly daily alcohol use 662 63.1 129 48.0 128 50.8 0.40 <0.001 0.45 <0.001 1.11 0.54 12-month cocaine or XTC use 39 3.8 112 41.6 149 59.1 22.54 <0.001 44.67 <0.001 1.98 <0.001 D+ frequent cannabis users with 12-month cannabis dependence (n = 252); D- frequent cannabis users who have never fulfilled criteria for cannabis dependence (n = 269); N2 non-users and non-frequent users in the general population (n = 1072). a N2 percentages weighted to be representative for the general Dutch population. Odds ratios (OR) are adjusted for the socio-demographic factors. In bold type, P < 0.005. Two cases had missing values for both last month tobacco use and parental divorce. 2 XTC: ecstasy.

6 Peggy van der Pol et al. Table 2 Adjusted comparison of 12-month disorders between dependent and non-dependent frequent cannabis users and the general population. Model 1: childhood and substance use Model 2: cannabis D- versus N2 D+ versus N2 D+ versus D- D+ versus D- OR P OR P OR P OR P Any 12-month disorder 2.21 <0.001 5.30 <0.001 2.40 <0.001 2.22 <0.001 Any mood disorder 0.55 0.26 1.62 0.34 2.93 0.002 2.67 0.003 Major depressive disorder 1.09 0.87 2.81 0.08 2.58 0.04 2.22 0.08 Bipolar disorder 0.28 0.06 0.80 0.68 2.85 0.06 2.99 0.04 Dysthymia 7.23 0.05 19.81 0.01 2.74 0.28 2.70 0.18 Any anxiety disorder 0.32 0.009 1.12 0.79 3.55 <0.001 2.95 0.004 Social phobia 0.38 0.11 1.53 0.49 4.00 0.007 3.25 0.03 Panic disorder 1.48 0.56 9.13 <0.001 6.16 0.009 4.87 0.01 General anxiety disorder 0.16 0.03 0.58 0.54 3.61 0.20 2.84 0.28 Agoraphobia Adult ADHD 1.81 0.27 3.34 0.03 1.85 0.03 1.70 0.05 ASP 1.29 0.58 3.13 0.009 2.43 <0.001 2.49 <0.001 Any childhood disorder 3.90 <0.001 6.86 <0.001 1.76 0.002 1.72 0.008 Childhood ADHD 0.97 0.95 1.76 0.28 1.82 0.03 1.75 0.03 Conduct disorder 4.10 <0.001 7.60 <0.001 1.85 0.001 1.86 0.003 D+ frequent cannabis users with 12-month diagnosis of cannabis dependence (n = 252); D- frequent cannabis users who have never fulfilled criteria for cannabis dependence (n = 269); N2 non-users and non-frequent users in the general population (n = 1072). In bold type, P < 0.005. Model 1 was adjusted for socio-demographic factors, the number of childhood adversity factors, past month tobacco use, 12-month weekly+ alcohol use and 12-month cocaine or ecstasy use. The two cases with missing values (tobacco/divorce) were excluded. Model 2 included socio-demographic factors and statistically significant cannabis use differences between D+ and D- (Table 3: solitary use, expansion and coping motive). ADHD: attention deficit hyperactivity disorder; OR: odds ratio. comparison with the general population (Table 2). Only the risk of having CD remained significantly higher for both dependent and non-dependent users compared to the general population sample, and dependent users had a higher risk of panic disorder. The confounders (childhood adversities and substance use) contributed equally to these attenuating effects (data not shown). By contrast, patterns in the associations remained similar in the adjusted comparison of frequent users with and without dependence, although ORs were attenuated (Table 2). Significant differences included the major diagnostic categories (any disorder, any mood disorder, any anxiety disorder, any childhood disorder), ASP and CD. Cannabis use in dependent and non-dependent frequent users Cannabis use patterns and habits were assessed in detail to investigate whether mental health differences between dependent and non-dependent frequent users could be attributed to differences in cannabis use characteristics. However, both groups were unexpectedly similar regarding cannabis use (Table 3). The only differences were more solitary cannabis use, expansion and coping motives in dependent users. Post-hoc analyses of daily cannabis users only, in accordance with Looby et al. [22], revealed very similar results. Moreover, when all motivation subscales were included in a single multivariate logistic regression model to identify the unique contribution of each motive, only the coping motive remained significant [OR = 1.14, 99.5% confidence interval (CI): 1.06 1.24, P < 0.001]. Similarly, after adjustment for solitary use and motives for usage (coping, expansion), dependent frequent users remained at significantly higher risk than non-dependent users of meeting criteria for any mental disorder, any mood disorder, any anxiety disorder, ASP and CD. DISCUSSION This is, to our knowledge, the first study to consider the distinction between dependent and non-dependent frequent cannabis users in a comparison of DSM-IV mental disorders with the general population. The mental health of non-dependent frequent cannabis users was better than expected from studies of all (including dependent) frequent users [3,10 16,20]: although mental disorders were more common in non-dependent users than in the general population (OR = 5.35, P < 0.001), this was largely attributable to (childhood) externalizing disorders (OR = 8.91, P < 0.001), whereas there was no difference in the presence of internalizing disorders. Dependent

(Dependent) heavy cannabis use and mental health 7 Table 3 Cannabis use by frequent users with and without dependence. D- D+ D+ versus D- n % n % OR P History Mean age of onset 1st use (SE) 14.4 (0.15) 14.0 (0.19) 0.92 0.08 Mean age of onset regular use (SE) 15.5 (0.14) 15.5 (0.14) 0.99 0.69 Mean years of use (SE) 6.7 (0.25) 7.4 (0.28) 1.11 0.02 Last year use 3 4 days/week 87 32.3 72 28.6 1 5 6 days/week 93 34.6 102 40.5 1.41 0.23 daily 89 33.1 78 31.0 1.21 0.52 Current Mean last month days of use (SE) 22.3 (0.43) 22.9 (0.46) 1.03 0.18 Mean number of joints per day (SE) 3.3 (0.14) 3.5 (0.14) 1.04 0.41 Mean dosage a (SE) 0.33 (0.01) 0.33 (0.01) 1.16 0.87 Preferred type marijuana 157 58.4 171 67.9 1 Resin 89 33.1 64 25.4 0.59 0.02 Both 23 8.6 17 6.8 0.62 0.09 Preferred potency low 28 10.5 17 6.8 1 Middle 87 32.7 78 31.1 1.58 0.25 High 151 56.8 156 62.2 1.83 0.08 Setting Solitary use 42 15.6 69 27.4 2.08 0.001 Location non-selective 50 18.6 44 17.5 1 Home settings 156 58.0 179 71.0 1.30 0.34 Coffee shops 63 23.4 29 11.5 0.53 0.09 Using (also) on weekdays 220 81.8 221 87.7 1.59 0.10 Using (also) in daytime 112 41.6 96 38.1 0.89 0.60 Hours per day high (SE) 5.5 (0.22) 5.8 (0.24) 1.03 0.29 Level of intoxication (SE) 6.2 (0.10) 6.4 (0.10) 1.06 0.29 Motives Mean enhancement (SE) 14.7 (0.27) 15.6 (0.20) 1.07 0.008 Mean conformity (SE) 5.4 (0.06) 5.8 (0.11) 1.35 0.006 Mean expansion (SE) 8.5 (0.25) 9.8 (0.30) 1.08 <0.001 Mean coping (SE) 7.1 (0.21) 8.9 (0.26) 1.17 <0.001 Mean social (SE) 11.2 (0.26) 12.1 (0.28) 1.06 0.03 D+ frequent cannabis users with 12-month cannabis dependence (n = 252); D- frequent cannabis users who never fulfilled criteria for cannabis dependence (n = 269). a Dosage: inverse of number of joints from 1 g of cannabis. Odds ratios (OR) of cannabis-related variables are adjusted for all socio-demographic factors in Table 1. In bold type, P < 0.005. Missing values: dose per joint n = 13, preferred potency n = 4, hours high n = 1, level of intoxication n = 1. SE: standard error. 2 frequent cannabis users also had more mental disorders than the general population (OR = 14.97, P < 0.001), but this was true for both internalizing disorders (mood OR = 4.15, P < 0.001; anxiety OR = 2.20, P = 0.002) and externalizing disorders (any childhood OR = 17.75, P < 0.001; adult ADHD OR = 7.26, P < 0.001; ASP OR = 9.98, P < 0.001). Moreover, cannabis use patterns and usage habits were compared thoroughly for frequent users with and without dependence and were found to be strikingly similar, with little effect on mental health differences. Limitations While an enriched population sample was recruited from coffee shops and through chain referral to achieve a large sample of frequent dependent and non-dependent cannabis users, this design cannot guarantee full representation of this population. Nonetheless, this does not necessarily mean that the sample is highly biased. After all, most frequent cannabis users in the Netherlands visit coffee shops, which is why recruitment occurred at these sites. Another issue that should be noted when interpreting the results is that all measures were based on selfreport. While self-report has been accepted as a suitable method for obtaining information about population behaviours, over- or under-reporting due to deliberate concealment or unconscious memory effects cannot be precluded [38,48,49]. Mental disorders Roughly half the frequent users had an externalizing disorder. Compared to the general population, associations were strong in both groups, consistent with previous studies on externalizing disorders [12,14,21]. However, the association between CD and dependence was much stronger than reported previously (OR = 19.99,

8 Peggy van der Pol et al. P < 0.001), due possibly to the higher frequency of cannabis use in the present study. CD and childhood ADHD usually precede cannabis initiation, and both disorders predict cannabis use and cannabis dependence [50,51]. Therefore, this finding emphasizes the potential of targeted prevention in this high-risk group, as well as the importance of treating these childhood disorders to reduce the development of both frequent cannabis use and dependence. This need is supported further by the greater persistence of ADHD into adulthood in frequent (dependent) cannabis users compared to the general population. Internalizing disorders were associated only with dependent frequent use (adjusted for socio-demographic factors). This supports findings of a higher prevalence of internalizing disorders in dependent compared to nondependent users in populations of daily cannabis users [22] and among less frequent ( 5 times last year) users [5]. However, in the latter study prevalence rates of mood disorders were comparable for dependent and nondependent users, but higher than in non-users (<5 times) [5]. Overall, clinical practice should be alert to the comorbidity of cannabis dependence and internalizing disorders. Dependent cannabis users should be monitored for emerging symptoms of an internalizing disorder and vice versa. Timely intervention could prevent the onset of a secondary disorder [52], which is often associated with higher treatment costs and special treatment needs [53]. Childhood adversity and substance use Frequent users with and without dependence were very similar with regard to childhood adversity and substance use, and both user groups scored much higher on these variables than the general population. These considerable differences between frequent cannabis users and the general population were responsible for the attenuation of most of the mental health differences between the user groups and the general population sample, most prominently regarding the high prevalence of internalizing disorders in dependent frequent users. Based on the assumed chronology, one may suggest that childhood adversity is partly responsible for the co-occurrence of cannabis dependence and mental disorders later in life [29]. However, compared to the general population, nondependent users also had more childhood adversity without a higher risk of internalizing disorders. Therefore, other unmeasured factors are probably involved in the mental health differences between frequent users with and without dependence, such as personality, genetic vulnerability or other environmental factors [e.g. 54,55]. The same reasoning applies to the current use of other substances, but implications are less straightforward because the chronology assumption is much weaker. Therefore, the actual direction of associations remains unknown in this cross-sectional study. Cannabis use in dependent and non-dependent frequent users Remarkably, frequent users with and without dependence did not differ in cannabis exposure. Age of onset of (regular) cannabis use, frequency of cannabis use, preferred potency, number of joints, number of hours high on a cannabis-using day, dosage per joint, potency preference or subjective achieved high were all similar. This is inconsistent with the only other study comparing nondependent and dependent frequent cannabis users, which showed higher levels of cannabis exposure in dependent daily users [22]. The fact that CanDep included respondents with a wider range in frequency of use cannot explain these contrasting results, as no difference in cannabis use patterns was observed in the daily cannabis users in our sample. However, the larger difference in number of joints per day between dependent and non-dependent users found by Looby et al. (0.35 joints compared to 0.2 joints in the current study) may actually correspond to similar delta9-tetrahydrocannabinol (THC) exposure, considering that the concentration of THC in Dutch marijuana is, on average, twice as high as Australian marijuana [56,57]). An alternative explanation could be that the differences in cannabis use between the non-dependent and dependent daily cannabis users in the study by Looby et al. were significant, but that the effect sizes were all somewhat small (standardized effect sizes ranging from 0.17 to 0.27), and that our study was not able to detect these small (clinically irrelevant) differences. In contrast to the findings for cannabis use patterns, solitary use as well as (especially coping) motives for cannabis use discriminated between dependent and nondependent frequent users. This finding is consistent with previous research in smaller samples of less frequent cannabis users [34,35,36,58]. These studies also suggested that coping motives moderate the association between internalizing disorders and cannabis dependence. This issue will be addressed in more detail in future studies of the CanDep cohort. However, after controlling for these differences in consumption habits and motives, dependent frequent cannabis users still showed a significantly higher prevalence of mental disorders compared to nondependent frequent cannabis users, suggesting that other (vulnerability) factors are responsible for these mental health differences between the two cannabis user groups. These results also suggest that, rather than cannabis exposure, factors discriminating dependent from nondependent frequent cannabis users include the presence of (pre-existing) mental disorders, solitary use and (coping) motives for cannabis use. This runs contrary to

(Dependent) heavy cannabis use and mental health 9 the conclusion of Looby et al., that [daily] cannabis users should be informed that increased cannabis consumption is likely to produce or exacerbate many problems and undesirable mood states. CONCLUSIONS Our data suggest that the prevalence of internalizing disorders is no higher in non-dependent frequent cannabis users than in the general population, suggesting that the mental health risks of frequent non-dependent cannabis use may have been overestimated in previous studies. This observation may lead to an important nuance for public debate, policy discussions and prevention messages. However, non-dependent users had more externalizing disorders than the general population (but less than dependent users), both in childhood and adulthood. As these disorders are likely to precede cannabis use, this finding points to the potential of targeted prevention, early detection and treatment of children with externalizing disorders. It should be noted, however, that nondependent users are still at risk of becoming dependent, and may therefore have a higher risk of internalizing disorders. Future studies should appreciate the distinction between dependent and non-dependent frequent users to investigate further whether the overlap between dependent users and frequent cannabis users explains inconsistent findings regarding internalizing disorders. Finally, frequent users with and without dependence had very similar cannabis use patterns and habits. Moreover, adjusting for the few differences in solitary use and motives for use did not eliminate the mental health differences between these groups of frequent cannabis users. Therefore, other unmeasured factors besides the currently studied cannabis use patterns and habits, childhood adversity and other substance use need to be examined to understand more clearly the reasons for the higher prevalence of mental disorders among dependent compared to non-dependent frequent cannabis users. Declarations of interest None. Acknowledgement This study was sponsored by the Netherlands organization for health research and development; grant number: 31160009. References 1. Hall W. The adverse health effects of cannabis use: what are they, and what are their implications for policy? IntJDrug Policy 2009; 20: 458 66. 2. Fergusson D. M., Boden J. M. Cannabis use and later life outcomes. Addiction 2008; 103: 969 76. 3. Cheung J. T., Mann R. E., Ialomiteanu A., Stoduto G., Chan V., Ala-Leppilampi K. et al. Anxiety and mood disorders and cannabis use. Am J Drug Alcohol Abuse 2010; 36: 118 22. 4. Mathews R. R., Hall W. D., Gartner C. E. Depression and psychological distress in tobacco smokers and people with cannabis dependence in the National Survey of Mental Health and Wellbeing. Med J Aust 2011; 195: S12 S15. 5. Swift W., Hall W., Teesson M. Cannabis use and dependence among Australian adults: results from the National Survey of Mental Health and Wellbeing. Addiction 2001; 96: 737 48. 6. Degenhardt L., Coffey C., Romaniuk H., Swift W., Carlin J. B., Hall W. D. et al. The persistence of the association between adolescent cannabis use and common mental disorders into young adulthood. Addiction 2013; 108: 124 33. 7. Grant B. F., Pickering R. The relationship between cannabis use and DSM-IV cannabis abuse and dependence: results from the National Longitudinal Alcohol Epidemiologic Survey. J Subst Abuse 1998; 10: 255 64. 8. Coffey C., Carlin J. B., Degenhardt L., Lynskey M., Sanci L., Patton G. C. Cannabis dependence in young adults: an Australian population study. Addiction 2002; 97: 187 94. 9. European Monitoring Centre for Drugs and Drug Addiction (EMCDDA). Annual Report: The State of the Drugs Problem in Europe. Lisbon: EMCDDA; 2009, p. 38 47. 10. Degenhardt L., Hall W., Lynskey M. Exploring the association between cannabis use and depression. Addiction 2003; 98: 1493 504. 11. Crippa J. A., Zuardi A. W., Martin-Santos R., Bhattacharyya S., Atakan Z., McGuire P. et al. Cannabis and anxiety: a critical review of the evidence. Hum Psychopharmacol 2009; 24: 515 23. 12. Stinson F. S., Ruan W. J., Pickering R., Grant B. F. Cannabis use disorders in the USA: prevalence, correlates and co-morbidity. Psychol Med 2006; 36: 1447 60. 13. Bovasso G. B. Cannabis abuse as a risk factor for depressive symptoms. Am J Psychiatry 2001; 158: 2033 7. 14. Agosti V., Nunes E., Levin F. Rates of psychiatric comorbidity among U.S. residents with lifetime cannabis dependence. Am J Drug Alcohol Abuse 2002; 28: 643 52. 15. Chen C. Y., Wagner F. A., Anthony J. C. Marijuana use and the risk of major depressive episode. Epidemiological evidence from the United States National Comorbidity Survey. Soc Psychiatry Psychiatr Epidemiol 2002; 37: 199 206. 16. Conway K. P., Compton W., Stinson F. S., Grant B. F. Lifetime comorbidity of DSM-IV mood and anxiety disorders and specific drug use disorders: results from the National Epidemiologic Survey on Alcohol and Related Conditions. J Clin Psychiatry 2006; 67: 247 57. 17. Zvolensky M. J., Bernstein A., Sachs-Ericsson N., Schmidt N. B., Buckner J. D., Bonn-Miller M. O. Lifetime associations between cannabis, use, abuse, and dependence and panic attacks in a representative sample. J Psychiatr Res 2006; 40: 477 86. 18. Kouri E., Pope H. G., Jr, Yurgelun-Todd D., Gruber S. Attributes of heavy vs. occasional marijuana smokers in a college population. Biol Psychiatry 1995; 38: 475 81. 19. Moore T. H., Zammit S., Lingford-Hughes A., Barnes T. R., Jones P. B., Burke M. et al. Cannabis use and risk of psychotic or affective mental health outcomes: a systematic review. Lancet 2007; 370: 319 28.

10 Peggy van der Pol et al. 20. Lee S. S., Humphreys K. L., Flory K., Liu R., Glass K. Prospective association of childhood attention-deficit/ hyperactivity disorder (ADHD) and substance use and abuse/dependence: a meta-analytic review. Clin Psychol Rev 2011; 31: 328 41. 21. Pingault J. B., Cote S. M., Galera C., Genolini C., Falissard B., Vitaro F. et al. Childhood trajectories of inattention, hyperactivity and oppositional behaviors and prediction of substance abuse/dependence: a 15-year longitudinal population-based study. Mol Psychiatry 2012. doi: 10.1038/mp.2012.87. 22. Looby A., Earleywine M. Negative consequences associated with dependence in daily cannabis users. Subst Abuse Treat Prev Policy 2007; 2: 3. 23. Degenhardt L., Hall W., Lynskey M. The relationship between cannabis use, depression and anxiety among Australian adults: findings from the National Survey of Mental Health and Well-Being. Soc Psychiatry Psychiatr Epidemiol 2001; 36: 219 27. 24. Palmer R. H., Young S. E., Hopfer C. J., Corley R. P., Stallings M. C., Crowley T. J. et al. Developmental epidemiology of drug use and abuse in adolescence and young adulthood: evidence of generalized risk. Drug Alcohol Depend 2009; 102: 78 87. 25. Wittchen H. U., Frohlich C., Behrendt S., Gunther A., Rehm J., Zimmermann P. et al. Cannabis use and cannabis use disorders and their relationship to mental disorders: a 10-year prospective-longitudinal community study in adolescents. Drug Alcohol Depend 2007; 88 (Suppl 1): S60 S70. 26. Lopez-Quintero C., Cobos J. P., Hasin D. S., Okuda M., Wang S., Grant B. F. et al. Probability and predictors of transition from first use to dependence on nicotine, alcohol, cannabis, and cocaine: results of the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC). Drug Alcohol Depend 2011; 115: 120 30. 27. Reigstad B., Jorgensen K., Wichstrom L. Diagnosed and selfreported childhood abuse in national and regional samples of child and adolescent psychiatric patients: prevalences and correlates. Nord J Psychiatry 2006; 60: 58 66. 28. Hayatbakhsh M. R., Najman J. M., Jamrozik K., Mamun A. A., Alati R. Do parents marital circumstances predict young adults DSM-IV cannabis use disorders? A prospective study. Addiction 2006; 101: 1778 86. 29. Oshri A., Rogosch F. A., Burnette M. L., Cicchetti D. Developmental pathways to adolescent cannabis abuse and dependence: child maltreatment, emerging personality, and internalizing versus externalizing psychopathology. Psychol Addict Behav 2011; 25: 634 44. 30. Noack R., Höfler M., Gründler R., Becker A., Schulz R., Paul K. Consumption patterns as covariates of DSM-IV cannabis dependence. Sucht 2009; 55: 137 47. 31. Temple E. C., Brown R. F., Hine D. W. The grass ceiling : limitations in the literature hinder our understanding of cannabis use and its consequences. Addiction 2011; 106: 238 44. 32. Walden N., Earleywine M. How high: quantity as a predictor of cannabis-related problems. Harm Reduct J 2008; 5: 20. 33. Noack R., Hofler M., Lueken U. Cannabis use patterns and their association with DSM-IV cannabis dependence and gender. Eur Addict Res 2011; 17: 321 8. 34. Bonn-Miller M. O., Zvolensky M. J. An evaluation of the nature of marijuana use and its motives among young adult active users. Am J Addict 2009; 18: 409 16. 35. Chabrol H., Duconge E., Casas C., Roura C., Carey K. B. Relations between cannabis use and dependence, motives for cannabis use and anxious, depressive and borderline symptomatology. Addict Behav 2005; 30: 829 40. 36. Johnson K., Mullin J. L., Marshall E. C., Bonn-Miller M. O., Zvolensky M. Exploring the mediational role of coping motives for marijuana use in terms of the relation between anxiety sensitivity and marijuana dependence. Am J Addict 2010; 19: 277 82. 37. Milich R., Lynam D., Zimmerman R., Logan T. K., Martin C., Leukefeld C. et al. Differences in young adult psychopathology among drug abstainers, experimenters, and frequent users. J Subst Abuse 2000; 11: 69 88. 38. Van der Pol P., Liebregts N., De Graaf R., Korf D. J., Van den Brink W., van Laar M. The Dutch Cannabis Dependence (CanDep) study on the course of frequent cannabis use and dependence: objectives, methods and sample characteristics. Int J Methods Psychiatr Res 2011; 20: 169 81. 39. Wouters M., Korf D. Access to licensed cannabis supply and the separation of markets policy in the Netherlands. JDrug Issues 2009; 39: 627 51. 40. De Graaf R., Ten Have M., Van Dorsselaer S. The Netherlands Mental Health Survey and Incidence Study 2 (NEMESIS-2): design and methods. Int J Methods Psychiatr Res 2010; 19: 125 41. 41. Von Sydow K., Lieb R., Pfister H., Hofler M., Wittchen H. U. What predicts incident use of cannabis and progression to abuse and dependence? A 4-year prospective examination of risk factors in a community sample of adolescents and young adults. Drug Alcohol Depend 2002; 68: 49 64. 42. De Graaf R., Bijl R. V., Ten Have M., Beekman A. T., Vollebergh W. A. Pathways to comorbidity: the transition of pure mood, anxiety and substance use disorders into comorbid conditions in a longitudinal population-based study. J Affect Disord 2004; 82: 461 7. 43. Hovens J. G., Wiersma J. E., Giltay E. J., van Oppen P., Spinhoven P., Penninx B. W. et al. Childhood life events and childhood trauma in adult patients with depressive, anxiety and comorbid disorders vs. controls. Acta Psychiatr Scand 2010; 122: 66 74. 44. Janssen I., Krabbendam L., Bak M., Hanssen M., Vollebergh W., de Graaf R. et al. Childhood abuse as a risk factor for psychotic experiences. Acta Psychiatr Scand 2004; 109: 38 45. 45. Konings M., Stefanis N., Kuepper R., de Graaf R., Ten H. M., vanosj. et al. Replication in two independent populationbased samples that childhood maltreatment and cannabis use synergistically impact on psychosis risk. Psychol Med 2012; 42: 149 59. 46. Simons J. S., Gaher R. M., Correia C. J., Hansen C. L., Christopher M. S. An affective motivational model of marijuana and alcohol problems among college students. Psychol Addict Behav 2005; 19: 326 34. 47. Gutierres R. G. Use of the term relative risk ratio. Available at: http://www.stata.com/statalist/archive/2005-04/ msg00678.html Archived by WebCite at http://www. webcitation.org/6bg3pltak Statacorp. 2005 (accessed 8 October 2012). 48. Harrison L. The validity of self-reported drug use in survey research: an overview and critique of research methods. NIDA Res Monogr 1997; 167: 17 36. 49. Zaldivar Basurto F., Garcia Montes J. M., Flores Cubos P., Sanchez Santed F., Lopez Rios F., Molina Moreno A. Validity