The association between cannabis use and mood disorders : A longitudinal study
Daniel Feingold, Mark Weiser, Jürgen Rehme, Shaul Lev-Ran
Journal of Affective Disorders, 2015, 172, 211–218.
doi : 10.1016/j.jad.2014.10.006
a b s t r a c t
Background : The association between cannabis use and mood disorders is well documented, yet evidence regarding causality is conflicting. This study explored the association between cannabis use, major depressive disorder (MDD) and bipolar disorder (BPD) in a 3-year prospective study.
Methods : Data was drawn from waves 1 and 2 of the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC). MDD and BPD were controlled at baseline and defined as meeting full criteria in the 12 months prior to the follow-up. Initiation of cannabis use was defined as any cannabis used by former lifetime abstainers in the time period between baseline and follow-up.
Results : Cannabis use was not significantly associated with increased incidence of MDD (Adjusted Odds Ratio (AOR) for daily use¼0.58(0.22–1.51)). Weekly to almost daily cannabis use was associated with increased incidence of BPD ((AOR for weekly to daily use¼2.47(1.03–5.92)); daily use was not (AOR¼0.52(0.17–1.55)). Baseline MDD was associated with initiation of cannabis use (AOR¼1.72(1.1– 2.69)). A crude association between baseline BPD and incidence of cannabis use was not maintained in adjusted models (AOR¼0.61(0.36–1.04)).
Limitations : Lack of information regarding frequency of cannabis use at follow-up and limitations regarding generalization of the results.
Conclusions : Our findings do not support a longitudinal association between cannabis use and incidence of MDD. Results regarding the association between cannabis use and incidence of BPD are conflicting and require further investigation. Baseline MDD, but not BPD, may be associated with future initiation of cannabis use. This may have implications for clinical, social and legislative aspects of cannabis use.
Keywords : Cannabis, Major depressive disorder, Bipolar disorder, Longitudinal association
Cannabis is the most widely used illicit substance (United Nations Office on Drugs and Crime, 2012), with reports estimating 2.8–4.5% annual prevalence of cannabis use worldwide (Degenhardt et al., 2011; Degenhardt and Hall, 2012). In recent years research has pointed out the extensive influence of the human endo-cannabinoid system on various neurological aspects (Di Marzo et al., 2011), including psychiatric disorders (Carvalho and Van Bockstaele, 2012).
The association between cannabis use and psychotic disorders has been studied extensively, with evidence suggesting that cannabis users, especially frequent users and those who initiated cannabis use at young age, are at increased risk of developing psychotic disorder (Moore et al., 2007). Yet the association between cannabis use and mood disorders is less clear (Hall and Degenhardt, 2009; Lev-Ran et al., 2013).
Depression is one of the most common psychiatric disorders, with a vast effect on global burden of disease (Whiteford et al., 2013). Co-morbidity of depression and cannabis use has been reported by numerous cross-sectional studies (Chen et al., 2002; Grant, 1995), indicating high prevalence of depression among cannabis users and vice-versa. Longitudinal studies have reported conflicting evidence regarding the association between cannabis use and depression. Several studies reported of a significant association between cannabis use at baseline and future occurrence of depression (Bovasso, 2001; Fergusson and Horwood, 1997), while others suggested that cannabis users and nonusers are equally prone to develop MDD at follow-up (Brook et al., 2002; Degenhardt et al., 2013). In a recent meta-analysis exploring the association between cannabis use and depression (Lev-Ran et al., 2013), the authors concluded that cannabis use, particularly heavy use, may be associated with an increased risk for developing depression, but that confounding factors limit conclusions concerning causality.
Though different research designs were used, including the use of multiple waves, prolonged follow-ups and meta-analysis; evidence regarding the longitudinal association between cannabis use and MDD are still controversial. Notably, many longitudinal studies exploring the association between cannabis use and MDD used unclear categories for defining frequency of cannabis use and did not account for the frequency of cannabis use in statistical analyses (Lev-Ran et al., 2013). In order to establish longitudinal association between cannabis use and MDD, greater exposure to cannabis is expected to lead to greater incidence of MDD at follow-up. Some evidence point to the inverse direction of causality, indicating that depressive symptoms may lead to cannabis use as a means of “selfmedication” (Gruber et al., 1997). Yet this notion did not receive much attention on empirical platforms, and scarce findings from large-scale longitudinal studies did not support a causal relationship between MDD and BPD and the initiation of cannabis use or increased frequency of use (Bolton et al., 2009; Degenhardt et al., 2003). Bipolar disorder is a severe psychiatric disorder that affects approximately 3% of the population worldwide (Grant et al., 2005; Whiteford et al., 2013). Compared to MDD, the association between cannabis use and BPD was studied to a much lesser extent, and research in the field suffers from lack of terminological standardization and scarce use of diverse methodological designs. Crosssectional studies have shown strong association between BPD and cannabis use (Cerullo and Strakowski, 2007; Etain et al., 2012).
Several longitudinal studies have indicated a link between baseline cannabis use and BPD at follow-up (Baethge et al., 2008; Henquet et al., 2006; Tijssen et al., 2010), yet these studies included subclinical manic and hypomanic symptoms as outcome measures and thus pose methodological shortcomings. Research has also shown that individuals suffering from BPD may use cannabis in order to alleviate mood symptoms related to both manic/hypomanic and depressive stages of the disorder (Bizzarri et al., 2007, 2009), yet these findings are not sufficiently supported by findings from longitudinal studies (Bolton et al., 2009).
In this study we sought to explore the association between cannabis use, MDD and BPD. As the directionality of any association between cannabis use and mood disorders is unclear, we explored this association using two longitudinal designs: the association of cannabis use and consequent incidence of mood disorders, as well as the association between mood disorders and consequent initiation of cannabis use. Furthermore, in order to overcome a previous methodological shortcoming and address possible dose-related effects of cannabis use on mood disorders, we explored this association differentially by frequency of cannabis use. Based on the above findings we hypothesized that there is a bidirectional association between cannabis use and mood disorders, i.e. that cannabis use at baseline will be associated with higher prevalence of MDD and BPD at follow-up and that incidence of mood disorder at baseline will be associated with higher prevalence of initiating cannabis use.