Cannabis dependence (now cannabis use disorder), like all drug addictions, is a chronic, relapsing disorder marked by compulsive drug-taking despite negative consequences. Cannabis use has been associated with a range of adverse effects, including negative health outcomes (Kalant, 2004), acute and chronic mental health problems (Moore et al., 2007), and psychosocial and cognitive impairments (Crean et al., 2011). According to a recent review (Batalla et al., 2013), adult chronic cannabis users exhibit altered brain activity compared to healthy controls during cognitive tasks that involve sustained attention (Chang et al., 2006; Abdullaev et al., 2010), working memory (Jager et al., 2006, 2007), inhibitory control (Hester et al., 2009), decision-making skills (Wesley et al., 2011), and psychomotor function (King et al., 2011). These alterations in neural function and associated behaviors may contribute to continued use and relapse. The compulsion to use cannabis is often driven by stress (Hyman and Sinha, 2009) and craving (Filbey and DeWitt, 2012), whereas cannabis use alleviates these negative feelings by reducing anxiety and craving (Fokos and Panagis, 2010). When cannabis is smoked, the main psychoactive component, Δ9-tetrahydrocannabinol (THC; Gaoni and Mechoulam, 1971), is absorbed through the lungs and rapidly enters the bloodstream and crosses the blood-brain barrier within minutes (Ashton, 2001). Once in the brain, THC acts as a partial agonist at the endogenous cannabinoid 1 (CB1) receptor and interferes with the endogenous cannabinoid system (Hoffman et al., 2007). CB1 receptors are located throughout the brain, but are mainly located in the hippocampus, amygdala, cerebellum, prefrontal cortex, and striatum (Lawston et al., 2000; Downer et al., 2001; Burns et al., 2007).

Preclinical studies have shown that THC exposure induces dose-dependent toxicity and structural changes in brain regions rich in CB1 receptors (Hoffman et al., 2007); however, it remains unclear whether these preclinical findings are transferable to humans. Human structural magnetic resonance imaging (sMRI) studies have explored the effects of cannabis use on brain morphology by comparing gray matter differences between cannabis users and non-using controls, and provide insight into the potential effects of THC exposure on the brain (for reviews see Lorenzetti et al., 2010, 2014; Batalla et al., 2013; Rocchetti et al., 2013). The most consistent findings have been smaller gray matter volume and lower gray matter density in the hippocampus/parahippocampus of cannabis users compared to non-using controls (Matochik et al., 2005; Yucel et al., 2008; Ashtari et al., 2011; Demirakca et al., 2011) with hippocampal gray matter volume in cannabis users inversely correlating with the total number of joints (Ashtari et al., 2011) and cones smoked (Yucel et al., 2008). Other gray matter volume findings include smaller amygdalar gray matter volume (Yucel et al., 2008) and larger anterior cerebellar gray matter volume (Cousijn et al., 2012) of cannabis users than controls; however, some studies report no structural alterations in cannabis users (Block et al., 2000; Tzilos et al., 2005; Yucel et al., 2008). Furthermore, Battistella et al. (2014) explored gray matter volume differences between occasional and regular cannabis users to better understand the long-term effects of cannabis use on brain structure and found that regular cannabis users exhibited smaller gray matter volume in the medial temporal cortex, temporal pole, parahippocampal gyrus, left insula, and the orbitofrontal cortex, and larger gray matter volume in areas of the cerebellum. The discrepancies across these studies have been attributed to several factors, such as heterogeneity in sample characteristics and methodological differences in data processing and analyses (Lorenzetti et al., 2010), but could also be related to co-occurring use of cannabis and tobacco in the study populations.

Globally, cannabis and tobacco are commonly mixed together and smoked concurrently in a practice called mulling (European Monitoring Centre for Drugs and Drug Addiction, 2011). A recent study in Australia reported that nearly two-thirds of cannabis users combine cannabis and tobacco for smoking (Banbury et al., 2013). In the United States, co-occurring use is also prevalent among individuals over the age of 18, with 36% of current cigarette smokers reporting cannabis use during the past 30 days and 64% of current cannabis users reporting cigarette use during the past 30 days (United States Department of Health and Human Services and Substance Abuse and Mental Health Services Administration, 2011). Although co-occurring use of cannabis and tobacco is prevalent, the extant literature examining the effects of cannabis and tobacco use on the brain focuses on their independent effects without sufficiently considering the effects of co-occurring use. Within the cigarette/tobacco smoking literature, sMRI studies indicate that smokers exhibit smaller gray matter volume and/or lower gray matter density than nonsmoking controls in the prefrontal, cingulate, insular, parietal, temporal, and occipital cortices, as well as in the thalamus and cerebellum (Brody et al., 2004; Gallinat et al., 2006; Almeida et al., 2008; Yu et al., 2011; Zhang et al., 2011a, 2011b; Liao et al., 2012; Morales et al., 2012; Franklin et al., 2014; Hanlon et al., 2014). Smokers have also been found to have larger gray matter volume/density than nonsmokers in several regions, including the insula (Zhang et al., 2011a), putamen, and parahippocampus (Franklin et al., 2014). Despite similarities in their route of administration, apparent effects on gray matter, and prevalence of co-occurring use, previous research has not fully assessed and/or controlled for cannabis use within the cigarette smoking studies. Similarly, only two cannabis-related studies statistically controlled for cigarette use (Jager et al., 2007; Gilman et al., 2014). Given the high rates of co-occurring cigarette and cannabis use and the potential effects of cannabis and cigarette smoking on brain structure, cigarette smoking and cannabis use should be considered in each of these disorders separately in order to ensure that gray matter differences are not incorrectly attributed to one drug or the other.

In the current study, we aimed to: (1) separate the effects of cigarette smoking from those of cannabis use and the effects of cannabis use from cigarette smoking on gray matter volume; (2) examine the effects of co-occurring cannabis and cigarette use on gray matter volume; (3) explore whether gray matter volume in cannabis users is associated with years of cannabis use; and (4) determine whether gray matter volume in cigarette smokers is associated with pack years, a measure of nicotine exposure. Unlike the majority of sMRI studies, which have primarily predefined regions of interest, we used a whole-brain voxel-wise approach and compared gray matter volumetric differences between cannabis-dependent individuals who have never smoked tobacco/cigarettes (Cs), cannabis-dependent individuals who smoke tobacco/cigarettes (CTs), nicotine-dependent, cannabis-naïve individuals (Ts), and healthy, non-using controls (HCs) using voxel-based morphometry (VBM). As mentioned above, the most consistent finding within the cannabis literature is smaller parahippocampal/hippocampal gray matter volume (Matochik et al., 2005; Yucel et al., 2008; Ashtari et al., 2011; Demirakca et al., 2011), so we hypothesized that Cs and CTs would exhibit lower gray matter volume in the parahippocampus/hippocampus than Ts and HCs. In a recent study, we found that nicotine-dependent cigarette smokers exhibited smaller gray matter volume in the thalamus and cerebellum, yet larger gray matter volume in the putamen and parahippocampus than controls (Franklin et al., 2014). We therefore hypothesized that Ts and CTs would show smaller gray matter volume in the thalamus and cerebellum and larger gray matter volume in the putamen compared to Cs and HCs.