Psychiatric Comorbidity of Cannabis Use Disorder, David Gorelick, 2019

Psychiatric Comorbidity of Cannabis Use Disorder

David Gorelick

In book : “Cannabis Use Disorders”, January 2019

DOI : 10.1007/978-3-319-90365-1_13


Abstract :

Objective : This chapter reviews the epidemiology and treatment of cannabis use disorder (CUD) with psychiatric comorbidity.

Methods : We summarize the findings of English-language epidemiological studies reporting current (past-year) comorbidity and of controlled clinical trials of treatment in which the majority of participants had diagnosed CUD and a specific comorbid disorder.

Results : There is substantial CUD comorbidity among community-dwelling adults with major psychiatric disorders: 4–6% for depression, 14% for bipolar disorder, 5% for anxiety disorder, and 16% for schizophrenia. Conversely, there is substantial psychiatric comorbidity among community-dwelling adults with CUD: 18–32% for depression, 8–9% for bipolar disorder, and 23–40% for anxiety disorder. No treatment is proven effective for CUD comorbidity; small-scale trials suggest that combined motivational enhancement therapy/cognitive behavioral therapy focused on both CUD and the psychiatric disorder may be effective. Single, small-scale trials suggest that lithium may be effective for comorbid bipolar disorder and clozapine for comorbid schizophrenia.

Conclusions : CUD with comorbid psychiatric disorders is common, and some behavioral interventions appear efficacious; however, there are no proven effective pharmacological treatments for this disorder.

Keywords : Cannabis use disorder, Cannabis, Marijuana, Comorbidity, Schizophrenia, Depression, Bipolar disorder, Anxiety disorder, Personality disorder, Treatment



Cannabis use disorder (CUD) is among the most prevalent psychoactive substance use disorders (SUDs), with an estimated 13.1 million individuals worldwide having moderate-severe CUD (cannabis dependence in DSM-IV terms) in 2010 [1]. In the United States, an estimated 4.0 million community-dwelling residents had current (past-year) CUD in 2015, a prevalence rate of 1.5% [2].

Therefore, it is not surprising that CUD often co-occurs with other non-SUD psychiatric disorders [3]. For example, a 2007 nationally representative survey of 8,841 community-dwelling Australians 16–85 years old (2007 National Survey of Mental Health and Wellbeing [NSMHW]) found that 69.8% (standard error [SE] 6.5%) of respondents with current (past 12-month) CUD also had psychiatric comorbidity (affective [major depression, dysthymia, bipolar], anxiety [panic, agoraphobia, social phobia, generalized anxiety disorder, post-traumatic stress disorder, obsessivecompulsive disorder], and/or alcohol use disorder), compared with 37.8% (SE 3.5) of current cannabis users without CUD and 15.5% (SE 0.5) of current nonusers of cannabis [4]. The odds ratio (OR) for having any comorbid disorder was 3.8 (95% confidence interval [CI] 1.9–7.6) for current cannabis users with CUD vs. current users without CUD and 0.3 (95% CI 0.2–0.4) for current nonusers vs. current users without CUD [4]. A study of 15.1 million adult (18–65 years old) admissions to US non-federal, acute care community general hospitals between 2007 and 2011 (Nationwide Inpatient Sample [NIS] of the Healthcare Cost and Utilization Project [HCUP]) found that 62.05% (95% CI 60.66–63.43) of the 65,767 inpatients with CUD as their only diagnosed SUD also had another non-SUD psychiatric diagnosis (mood, anxiety, psychotic, adjustment, impulse control, personality, or attention-deficit disorder), compared with 27.28% (95% CI 26.90–27.66) of the 14.7 million inpatients without a CUD diagnosis [5]. A retrospective record review of all 1,814,830 patients admitted to 458 hospitals in New South Wales (NSW), Australia, between July 1, 2006, and June 30, 2007, identified 8,669 (5%) patients with a diagnosis of current CUD, of whom 53.8% had another major psychiatric diagnosis (major depressive disorder, bipolar disorder, anxiety disorder, schizophrenia, personality disorder, or “severe stress disorder”), compared with a 4.2% prevalence of these psychiatric disorders among all patients (OR 17.2, 95% CI 17.4–19.0) [6]. A study of 837 outpatients at Madrid mental health clinics found a 66.2% prevalence of current psychiatric comorbidity among the 135 outpatients with current CUD [7].

Whether this comorbidity is due to a direct causal relationship between disorders (in either direction), to the chance co-occurrence of two common disorders, or to the presence of antecedent risk factors that promote the development of both disorders is often unclear. Few epidemiological studies provide information that might allow causal inference, e.g., odds ratios, for occurrence compared to a relevant reference group that might isolate the influence of CUD itself, such as cannabis users without CUD. Even fewer studies adjust the ORs to account for likely confounding risk factors, e.g., other substance use or SUDs and sociodemographic characteristics. The temporal order of onset of the two disorders provides clinically useful information (e.g., distinguishing between primary and secondary disorders), but this information is rarely available from large-scale epidemiological studies. Genetic and twin studies (e.g., comparing concordance for comorbidity in monozygotic and dizygotic twin pairs) might also be informative, but such studies almost always focus on cannabis use, rather than CUD [8].

CUD psychiatric comorbidity is clinically relevant because its presence is often associated with a poorer prognosis for CUD, the other psychiatric disorder, or both [3]. Clinically significant adverse consequences, such as poor treatment adherence and retention, more severe symptoms, greater functional disability, longer duration of active illness, more frequent occurrence of acute exacerbations, and/or greater rates of hospitalization, have been shown for bipolar disorder [9, 10, 11, 12, 13], depression [14], PTSD [15], and schizophrenia [16, 17].

This chapter reviews the epidemiology and treatment of CUD occurring with comorbid psychiatric disorders (except for other SUDs). For epidemiologic data, we focus on recent large-scale, community-based epidemiological surveys, as these provide the most scientifically rigorous data. The majority of such studies are cross-sectional and so do not provide evidence regarding the causal relationship between the two comorbid disorders. When available, we also present data from large-scale, prospective longitudinal studies which provide information about the incidence of comorbid disorders over a defined period of time. We present data on current, rather than lifetime, diagnoses to minimize the influence of recall bias on findings. We distinguish epidemiologic studies along two dimensions. First, do they report prevalence of the psychiatric disorder among individuals with CUD or prevalence of CUD among individuals with the psychiatric disorder? Second, are study subjects selected because they live in the community (and, ideally, are selected to be representative of everyone living in that community), regardless of treatment or treatment seeking status, or are subjects selected because they are in treatment or seeking treatment (i.e., clinical populations)? The latter groups are likely to be enriched with individuals who have comorbid (i.e., two or more) disorders because those with multiple disorders are more likely to be in treatment (the so-called Berkson’s bias or paradox) [18]. For treatment data, we focus on controlled clinical trials in which the majority of participants have diagnosed CUD and another specific psychiatric disorder.

This chapter does not cover cannabis use (i.e., without CUD) in the context of other psychiatric disorders, a topic on which there is substantial published literature. CUD comorbidity with schizophrenia (i.e., non-affective psychosis) is covered only briefly, as this topic is dealt with in more detail in the chapter by Drs. Tikka and D’Souza.