Sex-related differences in subjective, but not neural, cue-elicited craving response in heavy cannabis users

Highlights

• We examined sex-related differences in neural and subjective craving to cannabis.

• We used PCA to assess relative contributions of neural and subjective craving.

• There were no differences in neural response, which accounted for the most variance.

• There were differences in subjective craving.

• Accounting for these individual differences will increase efficacy of treatments.

Abstract

Background

Studies indicate that female cannabis users progress through the milestones of cannabis use disorder (CUD) more quickly than male users, likely due to greater subjective craving response in women relative to men. While studies have reported sex-related differences in subjective craving, differences in neural response and the relative contributions of neural and behavioral response remain unclear.

Methods

We examined sex-related differences in neural and behavioral response to cannabis cues and cannabis use measures in 112 heavy cannabis users (54 females). We used principal component analysis to determine the relative contributions of neural and behavioral response and cannabis use measures.

Results

We found that principal component (PC) 1, which accounts for the most variance in the dataset, was correlated with neural response to cannabis cues with no differences between male and female users (p = 0.21). PC2, which accounts for the second-most variance, was correlated with subjective craving such that female users exhibited greater subjective craving relative to male users (p = 0.003). We also found that CUD symptoms correlated with both PC1 and PC2, corroborating the relationship between craving and CUD severity.

Conclusions

These results indicate that neural activity primarily underlies response to cannabis cues and that a complex relationship characterizes a convergent neural response and a divergent subjective craving response that differs between the sexes. Accounting for these differences will increase efficacy of treatments through personalized approaches.

Introduction

Cannabis use incidence has generally been higher in men; however, recent trends suggest growing use in women (Center for Behavioral Health Statistics and Quality, 2017; Hasin, 2018; Lopez and Blanco, 2019; United Nations Office on Drugs and Crime, 2018). This trend is of particular importance given evidence for sex-related differences in substance use that are critical in continued use, development of abuse and dependence, and relapse (Becker and Koob, 2016; Bobzean et al., 2014; Chauchard et al., 2013; Hernandez-Avila et al., 2004; Perry et al., 2016). From a translational perspective, it is important to identify how sex-related differences affect factors that not only impact clinical interventions, but also any therapeutic effects of cannabis (Cooper and Craft, 2018).

Cannabis use disorders (CUDs) are difficult to treat (Sherman and McRae-Clark, 2016; Walker et al., 2015) and rates of treatment-seeking in both men and women are very low (Khan et al., 2013). Although women in general start using cannabis at a later age than men, they progress more quickly through the milestones of CUD and ultimately enter into treatment programs at a younger age, a phenomenon known as telescoping (Hernandez-Avila et al., 2004; Kerridge et al., 2018; Khan et al., 2013). Women also report higher abuse-related effects, suggesting increased sensitivity to cannabis that may promote its continued use (Cooper and Haney, 2014; Kerridge et al., 2018). In treatment-seeking individuals with CUD, women report greater withdrawal symptoms and higher negative impact of withdrawal (Herrmann et al., 2015; Sherman et al., 2017) with similar results in non-treatment-seeking cannabis users (Copersino et al., 2010; Schlienz et al., 2017). These considerations have important clinical relevance as studies report poorer outcomes for women in psychological and pharmacological treatments (Ali et al., 2015; Bassir Nia et al., 2018; DeVito et al., 2014; Ketcherside et al., 2016; Litt et al., 2015; McHugh et al., 2018; Sherman et al., 2017; Wetherill et al., 2015).

Cue-exposure elicits subjective craving and activates neural pathways related to reward processing (Filbey et al., 2009) that underlie the development of CUD and cannabis-related problems. While numerous studies have examined cue-induced craving in cannabis users (Cousijn et al., 2012; Filbey et al., 2016, 2014, 2009; Lundahl and Johanson, 2011; McRae-Clark et al., 2011), the few that have investigated sex-related differences have examined behavioral and neural response separately; thus, potential interactions between sex and measure of response are unknown. In a behavioral study, Lundahl and colleagues reported no significant sex-related differences in behavioral response during a cannabis cue-exposure task (Lundahl and Johanson, 2011). In a functional MRI (fMRI) study, Wetherill and colleagues also found no sex-related differences in neural response to cannabis cues in cannabis-dependent adults (Wetherill et al., 2015). However, women exhibited a positive correlation between neural response to cannabis cues in the bilateral insula and craving and a negative correlation in the left lateral orbitofrontal cortex (OFC), while men exhibited a positive correlation between cannabis craving and neural responses to cannabis cues in the striatum. The authors suggest that this difference may signify a top-down treatment approach in women with increased higher-order processing and that different treatment approaches can improve clinical outcomes in men and women.

We examined sex-related differences in brain perfusion and metabolism during rest in a subset of the data included in the present study and found that female cannabis users had increased global cerebral blood flow compared to male users. However, regionally, men had increased cerebral blood flow in the right insula and women had increased cerebral blood flow in the left posterior cingulate and bilateral precuneus (Filbey et al., 2018a). We also found that female users had increased cerebral metabolic rate of oxygen compared to male users. These differences, however, were found during rest and were not task-related. To our knowledge, no studies have examined sex-related differences in the combined and relative contributions of behavioral and neural response to cannabis cues in a cue-exposure task.

Evidence from animal models supports sex-related differences in the subjective effects of cannabis. The endocannabinoid system modulates the endocrine system (Battista et al., 2012) and cannabis use is postulated to interact with hormones in a bi-directional manner (Brown and Dobs, 2002; Gorzalka and Dang, 2012; Maria et al., 2014). Δ-9-tetrahydrocannabinol (THC) affects the hypothalamic pituitary gonadal (HPG) and hypothalamic pituitary adrenal (HPA) axes and is modulated by cannabinoid (CB1) receptors (Brown and Dobs, 2002; Crane et al., 2013; Ketcherside et al., 2016; López, 2010; Mendelson et al., 1984). THC also alters the release of gonadotropin-releasing hormone (GnRH) from the hypothalamus, leading to differential downstream effects in men and women (Battista et al., 2012; Brents, 2016). For example, similar to humans, female rats exhibit increased sensitivity to the acute effects of cannabis (Craft et al., 2013; Farquhar et al., 2019). Fattore and colleagues have found that ovarian hormones regulate reinforcement of cannabinoids such that female rats exhibited increased learned associations between cannabis-related stimuli and effects of cannabis (Fattore et al., 2007) and that ovariectomized female rodents displayed reduced cue-elicited cannabinoid-seeking behavior compared to intact female rodents (Fattore et al., 2010). A study in gonadectomized rats found that replaced testosterone reduced withdrawal symptoms, while both estradiol and progesterone increased withdrawal symptoms (Marusich et al., 2015). More recently, Farquhar and colleagues found that repeated exposure to THC increased downregulation and desensitization of CB1 receptors in female rats (Farquhar et al., 2019). These preclinical studies indicate that ovarian hormones influence cannabis reinforcement and provide potential mechanisms underlying sex differences (Crane et al., 2013; Lynch et al., 2002).

The aim of this study was to examine the interaction between sex and measures of craving and use in heavy cannabis users. We used principal component analysis (PCA), a multivariate approach, to determine whether the different behavioral, neural, and use variables may be combined based on common dimensions. Specifically, PCA aims to describe and summarize patterns of data by grouping together correlated variables into components that account for the most variance possible in the original variables. This approach also limits the number of variables to be analyzed in subsequent analyses (Tabachnik and Fidell, 2013). This data reduction is achieved by identifying the underlying structure in the data, plotting the data points in a multidimensional space, and creating components that can be thought of as lines in the multidimensional space in the directions where there is the most variance. In doing so, PCA provides a more nuanced approach to understanding sex-related differences compared to group comparisons via traditional statistical methods (Halai et al., 2017). We hypothesized comparable neural response to cannabis cues, but increased subjective craving in female cannabis users compared to male users.