Regular cannabis and alcohol use is associated with resting-state time course power spectra in incarcerated adolescents

Highlights

• Longer cannabis abuse is associated with decreased functional network connectivity.

• Adolescent cannabis use is associated with spectral power of several networks.

• Alcohol use duration is associated with decreased spectral power of several networks.

Abstract

Cannabis and alcohol are believed to have widespread effects on the brain. Although adolescents are at increased risk for substance use, the adolescent brain may also be particularly vulnerable to the effects of drug exposure due to its rapid maturation. Here, we examined the association between cannabis and alcohol use duration and resting-state functional connectivity in a large sample of male juvenile delinquents.

The present sample was drawn from the Southwest Advanced Neuroimaging Cohort, Youth sample, and from a youth detention facility in Wisconsin. All participants were scanned at the maximum-security facilities using The Mind Research Network’s 1.5T Avanto SQ Mobile MRI scanner. Information on cannabis and alcohol regular use duration was collected using self-report. Resting-state networks were computed using group independent component analysis in 201 participants. Associations with cannabis and alcohol use were assessed using Mancova analyses controlling for age, IQ, smoking and psychopathy scores in the complete case sample of 180 male juvenile delinquents.

No associations between alcohol or cannabis use and network spatial maps were found. Longer cannabis use was associated with decreased low frequency power of the default mode network, the executive control networks (ECNs), and several sensory networks, and with decreased functional network connectivity. Duration of alcohol use was associated with decreased low frequency power of the right frontoparietal network, salience network, dorsal attention network, and several sensory networks.

Our findings suggest that adolescent cannabis and alcohol use are associated with widespread differences in resting-state time course power spectra, which may persist even after abstinence.

Introduction

Adolescence is a period of significant growth and developmental change. Psychologically, one of the most marked changes may be the rise in risk-taking behavior, which places the adolescent at an increased risk for antisocial behaviors, such as substance use. In 2013, 34.9% of United States high-school students reported to have drunk alcohol in the last month, while 23.4% reported the use of cannabis (Kann et al., 2014). In adolescent detainees, these numbers are even higher, with 41–47% of youth reporting heavy alcohol use and 36–48% reporting use of cannabis (Ewing et al., 2015). Paradoxically, as the adolescent brain undergoes major changes in synaptic receptors density as well as in myelination (Crews et al., 2007), it may be particularly vulnerable to the effects of substance exposure. Elucidating the effects of substance use on the adolescent brain may provide important information for prevention, the development of intervention programs, and policy-making.

Both cannabis and alcohol are believed to have widespread effects on the brain. The main psychoactive component of cannabis, Δ9-tetrahydrocannabinol (THC), binds to endogenous cannabinoid receptor-1 (CB1). CB1 receptors are widely distributed throughout the brain, but are especially concentrated in the cerebellum, prefrontal cortex, striatum, amygdala, and hippocampus (Hirvonen et al., 2012), and play a role in neurotransmitter release and concentrations across neural systems. Cannabinoid receptor density is also greater in children and adolescents than in adults (Glass et al., 1997). By affecting glutamate and gamma-aminobutyric acid (GABA) systems, THC may interfere with brain maturational processes (Bossong and Niesink, 2010). Exposing the cellular and molecular targets for alcohol's actions has proven more challenging. However, like cannabis, alcohol is believed to affect many neurotransmitter systems in the brain (Paul, 2006). Three important neurotransmitter systems that undergo substantial changes during adolescence and are affected by alcohol consumption are dopamine, glutamine, and GABA (Hiller-Sturmhofel and Swartzwelder, 2004).

Cannabis and alcohol use during adolescence have been associated with both immediate and long-term outcomes that include disruptions in task-relevant neural activity and small or no effects on cognition (Gonzalez et al., 2012, Harvey et al., 2007, Jacobsen et al., 2007, Jager et al., 2010, Lane et al., 2007, Tapert et al., 2007, Tait et al., 2011). Structural magnetic resonance imaging (MRI) analyses suggest that adolescent cannabis use may be associated with decreased gray matter volume in widespread regions of the brain, such as the prefrontal cortex, amygdala, and hippocampus (Churchwell et al., 2010, Cousijn et al., 2012, Filbey et al., 2014, Yucel et al., 2006). Youth who drink alcohol show decreased brain volume or cortical thickness in the frontal, temporal, and parietal cortex, and hippocampus (Nagel et al., 2005, Luciana et al., 2013, Squeglia et al., 2014, Squeglia et al., 2015). Generally, earlier initiation of use and more frequent use have been associated with poorer outcomes (Buchy et al., 2015, Pope et al., 2003), and while some studies suggest that effects of alcohol and cannabis use decrease after prolonged periods of abstincence (Fortier et al., 2014, Hanson et al., 2010, Jacobus et al., 2012), other studies suggest long-lasting effects of adolescent alcohol and cannabis use (Ashtari et al., 2011, van Eijk et al., 2013). However, study results vary widely and await replication.

Functional connectivity, defined as the relation between the neuronal activation patterns of anatomically separated brain regions (Aertsen et al., 1989), describes the organization, inter-relationship and integrated performance of functionally coupled brain regions (Rogers et al., 2008). Most commonly, studies on functional connectivity describe the temporal correlation between two or more regions, or compare the spatial maps of resting-state networks. This latter measure examines the correspondence of the network’s time course and the time course of each voxel in a network (Balsters et al., 2013), thus providing a measure of a region’s strength of connectivity within a given network. The literature on cannabis use mostly describes increased functional connectivity in adult or adolescent cannabis users compared to controls in (regions of) the default mode network (DMN), salience network, and executive control network (ECN) (Cheng et al., 2014, Filbey et al., 2014, Harding et al., 2012, Houck et al., 2013, Pujol et al., 2014). However, more recently, several studies have reported negative associations between cannabis use and functional connectivity (Camchong et al., 2016, Orr et al., 2013, Peters et al., 2015).

Alcohol use has also been shown to be associated with functional connectivity. Both weaker and stronger functional connectivity have been reported in the DMN, salience network, subcortical reward network and ECN (Weiland et al., 2014, Zhu et al., 2017), as well as stronger functional connectivity in the left frontoparietal network (Jansen et al., 2015) basal ganglia network and primary visual network (Weilandt et al., 2014). Moreover, disturbances in frontoparietal connectivity have been observed even in substance-naïve youth with a family history of alcohol (Wetherill et al., 2012), which suggest that weaker frontoparietal connectivity may be a neurobiological marker for alcohol use disorders. Thus, both alcohol and cannabis use appear to be associated to functional connectivity, however, studies are inconclusive whether associations are positive or negative.

Although described less frequently, functional connectivity can also be examined by measuring time course power spectra. The MR signal is dominated by oscillations in the 0.0–0.1 Hz frequency band (Cordes et al., 2001). The magnitude of these oscillations may differ per brain region and per person, and thus can be examined as a marker of individual differences. The time course power spectra reflect the degree of fluctuation in amplitude of the intrinsic activity within the network (Calhoun et al., 2011). The time courses of the different networks can also be correlated, resulting in a measure called functional network connectivity (FNC, Arbabshirani et al., 2013). The present study will examine associations between alcohol and cannabis use and network spatial map, network time course power spectra and functional network connectivity.

Possibly due to relatively small sample sizes, studies examining substance use and functional connectivity mostly describe associations between a few regions or networks only rather than a comprehensive whole-brain analysis. Moreover, studies that do perform whole-brain analyses generally only assess differences in network spatial maps. The present study describes the results of a dose-response analysis of the association between duration of regular (comorbid) cannabis and alcohol use and whole-brain resting-state functional connectivity in a large sample of male juvenile delinquents, whom due to their imprisonment were abstinent. The majority of the sample has been dependent on cannabis or alcohol, many on both. We therefore expected that cannabis and alcohol use would be associated with widespread differences in functional connectivity despite current abstinence. In the spatial map domain, both cannabis and alcohol use are expected to affect the DMN, salience network, and ECN. Moreover, we hypothesize alcohol to be associated with frontoparietal network connectivity. However, as time course power spectra and between network connectivity are less frequently studied, we have no specific hypotheses regarding these domains.