Examination of Gamma-Aminobutyric Acid and Glutamate-Glutamine Levels in Association with Impulsive Behavior in Adolescent Marijuana Users

Abstract

Background:

Adolescent marijuana (MJ) use has been associated with alterations in brain structure and function as well as behavior. Examination of neurochemical correlates such as GABA (gamma-aminobutyric acid) and Glx (glutamate + glutamine) in MJ users remains limited. Impulsivity, identified as a risk factor and consequence of MJ use, has been associated with GABA and Glx levels in healthy and clinical populations. However, this relationship has not been investigated in MJ users. In this study, we examined levels of GABA and Glx in the anterior cingulate cortex (ACC) and its relationship with impulsive behavior in MJ-using adolescents and healthy controls.

Methods:

Healthy control subjects (HC; N=21) and MJ-using adolescents (N=18) completed a metabolite-edited ¹H MRS exam to measure ACC GABA and Glx levels, a structured clinical interview to assess MJ use, and the Barratt Impulsivity Scale (BIS-11) to evaluate impulsive behavior.

Results:

Adolescent MJ users had significantly lower tissue-corrected GABA (with macromolecules; GABA+) levels (p=0.029) compared to HC’s. No significant between-group differences were observed in ACC Glx levels. Assessment of impulsive behavior demonstrated no significant between-group differences in motor, non-planning, attention, and total impulsivity scores. Additionally, impulsivity measures and tissue-corrected GABA+ or Glx levels were not significantly correlated in either group.

Conclusion:

Lower GABA levels in MJ users may indicate alterations in excitatory-inhibitory mechanisms critical for neurodevelopment. Although no significant relationships were observed between impulsive measures and GABA or Glx levels in both groups, further investigations are needed examining the relationship between neurochemical correlates, behavior, and adolescent MJ use.

1. Introduction

Marijuana (MJ) is one of the most commonly used drugs among adolescents. According to the 2019 National Survey on Drug Use and Health (NSDUH), an estimated 43.5 million people reported MJ use in 2018, out of which 14.9 million were adolescents and young adults between the ages of 12 – 25 years (SAMHSA, 2019). Estimates from the 2019 Monitoring the Future (MTF) study indicate that levels of daily MJ use among 8^th, 10^th, and 12^th graders were either at or close to the highest levels observed since 1991 (Johnston et al., 2020). Recent legalization of MJ use for recreational and/or medical purposes has resulted in a lower perception of risk and increased MJ use among adolescents (Cerda et al., 2017; Rusby et al., 2018). Combined with the inherent risk-taking and impulsive nature observed during this developmental period, adolescents are especially vulnerable toward initiation and continued use of substances such as MJ (Churchwell and Yurgelun-Todd, 2011). The implications of adolescent MJ use are an important area of study due to the significant neurodevelopmental changes that occur during adolescence. These changes are essential for the advancement of cognitive, emotional, and behavioral processes making the adolescent brain particularly susceptible to perturbations caused by exogenous substances such as MJ (Crews et al., 2007). While the association between adolescent MJ use and brain development continues to be widely investigated, there is still a paucity of research, particularly in the examination of neurochemical correlates that play a critical role in brain developmental processes and its relationship with behaviors that are associated with MJ use such as impulsivity.

Delta-9-tetrahydrocannabinol (THC), the main psychoactive component found in MJ, acts on the brain by disrupting endogenous signaling mechanisms regulated by the endocannabinoid system (ECS) (Mechoulam and Parker, 2013). The ECS has been shown to play an integral role in neurodevelopment throughout childhood and adolescence and is comprised of cannabinoid receptors, endogenous cannabinoid (endocannabinoid) ligands, and enzymes involved in the synthesis and breakdown of endocannabinoids (Meyer et al., 2018). The cannabinoid 1 (CB1) receptor is one of the most abundantly expressed receptors in the brain and is primarily located on presynaptic regions of inhibitory (gamma-aminobutyric acid (GABA)) and excitatory (glutamatergic) neurons (Mechoulam and Parker, 2013). Retrograde activation of these receptors by postsynaptically released endocannabinoids leads to the inhibition of neurotransmitter release. This mechanism plays an important role in maintaining the excitatory-inhibitory balance critical for the regulation of a wide range of neurodevelopmental and behavioral systems (Galve-Roperh et al., 2009; Mechoulam and Parker, 2013). Consequently, exposure to THC, which acts as a partial agonist at CB1 receptors can lead to dysregulation of these neurotransmitter signaling mechanisms and subsequently to downstream effects on cognitive functioning, emotion regulation, and behavior (Trezza et al., 2008). This highlights the importance of studying potential perturbations in neurotransmitter levels associated with adolescent MJ use.

Proton magnetic resonance spectroscopy (¹H MRS) is a non-invasive technique that enables the detection of a range of neurometabolites and neurotransmitters in vivo. Several ¹H MRS studies have examined levels of glutamate (Glu), glutamate + glutamine (Glx), as well as GABA in both adolescent and adult MJ users with mixed findings. For instance, lower Glu levels have been found in the anterior cingulate cortex (ACC) (Prescot et al., 2011; Prescot et al., 2013) and basal ganglia (Chang et al., 2006) of MJ users compared to healthy controls. Contrastingly, studies have also reported increases in striatal Glu and no differences in ACC Glu levels following administration of THC (Mason et al., 2019) as well as no differences in hippocampal (Blest-Hopley et al., 2020) and ACC Glu (Watts et al., 2020) levels between chronic MJ users and healthy controls. Similarly, inconsistent findings have been observed in Glx levels which is often examined as a derivative of glutamate due to the significant spectral overlap observed between glutamate and glutamine signals. While Watts et al. (2020) demonstrated no significant differences in ACC Glx levels between chronic MJ users and healthy controls, in a separate study, a sex interaction effect was observed with female MJ users between the ages of 18 – 21 years showing lower Glx levels in the dorsal striatum compared to male MJ users and control participants (Muetzel et al., 2013). To our knowledge, only two studies have explored the relationship between MJ use and GABA levels with one study finding lower ACC GABA levels in adolescent MJ users compared with control participants (Prescot et al., 2013) and another finding no significant difference in GABA levels between MJ users and controls following administration of THC (Mason et al., 2019). While findings from these studies have been equivocal, potentially due to variations in regions examined, methodologies used, and populations studied, there is clear indication towards alterations in neurotransmitter levels that needs to be explored further.

A key behavioral trait that has been linked to substance use during adolescence is impulsivity. Impulsive behavior has been identified as both a risk factor and consequence of using substances such as MJ (Ansell et al., 2015; Moeller and Dougherty, 2002). Research into the neurochemical correlates of impulsive behavior has primarily focused on monoaminergic neurotransmitters such as dopamine, serotonin and noradrenaline which are critically regulated by glutamatergic and GABAergic inputs from the prefrontal cortex (PFC) (Del Arco and Mora, 2009; Mitchell and Potenza, 2014). However, in the past decade, ¹H MRS studies have been increasingly utilized to examine the association between Glu, Glx or GABA levels with impulsivity in both healthy (Boy et al., 2011; Silveri et al., 2013) and clinical populations (Bauer et al., 2018; Li et al., 2020). In a review examining the relationship between Glu and GABA with neuropsychological assessments, it was noted that Glu or Glx levels were positively associated, and GABA levels were negatively associated with impulsivity measures (Ende, 2015). Despite significant overlap between MJ use, impulsive behaviors, and Glu, Glx, and GABA levels, understanding of the relationship between Glu, Glx or GABA levels and impulsive behavior in adolescent MJ users remains limited.

In this study, we applied edited ¹H-MRS acquisition methods to examine levels of GABA and Glx in the ACC of adolescent MJ users and healthy controls. We also investigated the relationship between these neurometabolites and impulsive behavior. The ACC was chosen as the region of interest for several reasons. First, the ACC is part of the cognitive control network and is uniquely positioned to receive and project information to frontal and limbic regions of the brain in order to regulate behavioral responses (Cole and Schneider, 2007; Stevens et al., 2011). Additionally, the ACC has high concentrations of CB1 receptors (Svizenska et al., 2008) and MJ use has been associated with alterations in ACC structure (Maple et al., 2019; Rapp et al., 2013) and function (Gruber and Yurgelun-Todd, 2005). These alterations have been associated with impairments in cognitive functions (Gruber and Yurgelun-Todd, 2005; Hester et al., 2009) and affective processing (Maple et al., 2019). Moreover, the ACC has consistently been shown to be a key area involved in the regulation of impulsive behavior independent of MJ use (Castellanos-Ryan and Séguin, 2016; Golchert et al., 2017).

Based on prior findings, we hypothesized that adolescent MJ users would exhibit lower GABA and Glx levels in the ACC compared to non-users and that adolescent MJ users would demonstrate higher impulsivity scores compared to healthy controls. Furthermore, we hypothesized that lower ACC GABA levels and higher Glx levels would be associated with increased impulsivity in MJ users.

2. Methods

2.1. Participants

Study protocols were approved by the University of Utah Institutional Review Board (IRB). Participants between the ages of 14 to 21 years were recruited from the Salt Lake City community using advertisements and word of mouth. Twenty-one healthy control (HC) and 20 MJ-using adolescents were enrolled in the study. Prior to enrollment, interested participants completed a phone screen to determine study eligibility. Inclusion criteria for the HC group included no Axis 1 diagnosis based on the Diagnostic and Statistical Manual of Mental Disorder IV (DSM-IV). Participants in the MJ group were required to have a self-report of chronic MJ use and must have used MJ a minimum of 100 times in the past year. Exclusion criteria for all participants included an estimated full-scale IQ < 70; history of claustrophobia, autism, schizophrenia, anorexia nervosa, or bulimia; major sensorimotor handicaps; current medical or neurological disease; metal fragments or implants that are non-MRI compatible; history of electroconvulsive therapy; history of head injury with loss of consciousness ≥ 30 minutes; and current pregnancy or lactation. Dependence on alcohol or other drug use in the last 12 months and lifetime use of illicit drugs other than MJ ≥ 20 times was exclusionary. Participants in the MJ group were also requested to abstain from MJ use 24 hours prior to their study visit. Prior to participation in the study, written informed consent was obtained from participants ≥ 18 years of age. For participants < 18 years of age, written informed assent and parental permission were obtained.

A semi-structured clinical and diagnostic interview was administered to all participants. Participants under the age of 18 were administered a web-based computerized version of the Kiddie Schedule for Affective Disorders and Schizophrenia (KSADS-COMP) (Townsend et al., 2020) and participants 18 years and older were administered the Structured Clinical Interview for DSM-IV Disorders (SCID-I) (First et al., 2002). An expanded assessment of participants’ MJ use history was also obtained. This included information regarding age of onset of MJ use, age of regular use (defined as MJ use ≥ 2 times per week) as well as duration and frequency of MJ use. Based on the information obtained, an estimated total lifetime use, specified as number of MJ use events was calculated. Participants were also asked about the frequency of using other substances including alcohol and nicotine, which was quantified as number of uses per week. A urine sample was also obtained from all participants on the day of their visit and a 5-panel drug test was administered testing for the presence of THC, cocaine, opiates, benzodiazepines and amphetamines.

2.2. Impulsivity measure

Participants completed the Barratt Impulsivity Scale – version 11 (BIS-11), a 30-item questionnaire designed to examine different constructs of impulsivity (Patton et al., 1995). Specifically, motor impulsivity (BIS-M), non-planning behavior (BIS-NP) and inattention (BIS-A) were evaluated. For each question, participants were asked to respond on a 4-point Likert scale: rarely/never, occasionally, often, and almost always/always. A total composite score was obtained from these subscales to assess total impulsivity (BIS-Total). Higher scores on the BIS-M, BIS-NP, BIS-A, and BIS-Total are indicative of increased impulsive behavior.

2.3. MR data acquisition

Magnetic resonance imaging (MRI) acquisitions were performed using a 3 Tesla (T) Siemens Prisma (Erlangen, Germany) whole-body scanner. Radiofrequency (RF) transmission was performed using a circularly polarized whole-body coil and a 32-channel receive-only head coil was used for signal reception.

2.4. Structural Data Acquisition

A three-dimensional (3D), T₁-weighted magnetization prepared rapid gradient echo (MPRAGE) sequence was used to obtain high-resolution, high-contrast images to assist with MRS voxel positioning and tissue segmentation. The acquisition parameters were as follows: Repetition time (TR) = 2500ms; Echo time (TE) = 2.80ms; Inversion time (TI) = 1060ms; Flip angle = 8°; Number of slices = 176; Matrix size = 256 × 256; Field of view (FOV) = 256 × 256; Slice thickness = 1.0mm.

2.5. MRS Voxel Placement

The MRS voxel, measuring 30mm × 25mm × 25mm, was obliquely positioned along the sagittal plane and was placed bilaterally to encompass predominantly gray matter within the ACC. The smallest and largest dimensions of the voxel were oriented along the anterior-posterior and superior-inferior axis and the anterior ventral edge of the voxel was aligned with the centroid of the genu of the corpus callosum (Figure 1). Voxel positioning was based on prior MRS studies that have specified the ACC as a region of interest (Prescot et al., 2011; Prescot et al., 2013).

Figure 1:

Axial, sagittal and coronal views of anterior cingulate cortex (ACC) voxel position from a study participant. The MRS voxel is represented by the white highlighted box.

2.6. ¹H MRS Data Acquisition

Within-voxel static magnetic field (B₀) shimming was performed using the FAST(EST)MAP (Gruetter, 1993; Gruetter and Tkac, 2000) method along with automated and manual shimming methods to ensure an unsuppressed water signal linewidth ≤ 10 Hz (real component). Six saturation bands positioned approximately 1 cm away from the MRS voxel was used for outer-volume suppression. Edited ¹H MRS data for five participants in the study (HC group) were acquired using the Meshcher-Garwood Point Resolved Spectroscopy (MEGA-PRESS; Mescher et al. (1998)) technique with the following acquisition parameters: TR = 2000ms, TE = 68ms, number of averages (N_avg) = 512. An unsuppressed water reference spectrum was also acquired (TR = 2000ms, TE = 68ms, N_avg = 32). The rest of the participants (N = 36) were scanned using the Hadamard Encoding and Reconstruction of Mega-Edited Spectroscopy (HERMES; Saleh et al. (2016)) sequence (TR = 2000ms, TE = 80ms, N_avg = 320). The parameters for the unsuppressed water reference scan were TR = 2000ms, TE = 80ms, and N_avg = 16. The GABA editing pulses for both MEGA-PRESS and HERMES were placed at 1.9ppm for the “ON” scans and at 7.46ppm (MEGA-PRESS) and 7.22ppm (HERMES) for the “OFF” scans. The edited (DIFF) spectrum was obtained by subtracting the “OFF” scans from the “ON” scans.

2.7. Data processing

Spectral fitting and quantification for MRS data were performed using Gannet version 3.0 with SPM12 integrated within the Gannet Toolkit (Edden et al., 2014). This allows for the segmentation of the T1-weighted image and MRS voxel to obtain gray matter (GM), white matter (WM), and cerebrospinal fluid (CSF) content within the voxel. Gray matter fraction was calculated as the ratio of total brain matter (100 X GM/(WM+GM)). Importantly, Gannet also incorporates the ability to differentiate between MEGA-PRESS and HERMES sequences to be processed accordingly. Since the GABA edited signal at 3.0 ppm also includes contributions from macromolecules (Mullins et al., 2014), GABA levels will be reported as GABA+. GABA+ (3 ppm) and Glx (3.75 ppm) signals from the edited spectrum were quantified in institutional units (i.u) and normalized to the unsuppressed water signal. Additionally, tissue-correction accounting for variations in metabolite and water T2-relaxation times were applied for both GABA+ and Glx values which were also incorporated in Gannet. Spectral quality was assessed using the GABA+ fit error (%) and Glx fit error (%) values determined using Gannet (Edden et al., 2014). Based on previous studies, a threshold of <15% was used as a measure of good spectra quality (Cao et al., 2018; Saleh et al., 2020; Wenneberg et al., 2020).

2.8. Statistical analysis

Data were analyzed using IBM SPSS Statistics for Macintosh, Version 25.0 (SPSS Inc., Chicago, IL, USA). Prior to analysis, data were examined using the Shapiro-Wilk’s test for normality to determine use of parametric or non-parametric tests. Between-group differences for ¹H MRS and BIS-11 variables were examined using independent samples t-test for normally distributed data and non-normal data was examined using the Mann-Whitney U test. One-way analysis of covariance (ANCOVA) was used to adjust for age and sex as well as nicotine and alcohol use. Pearson’s or Spearman’s correlations were used to assess the relationship between ¹H MRS data and MJ use variables as well as the relationship between GABA and Glx with BIS-11 scores. Additionally, partial correlations were run with age and sex as well as frequency of nicotine and alcohol use as covariates. For all analyses, alpha (α) value was set at p < 0.05.

3. Results

3.1. Demographic and clinical variables

Twenty MJ-using adolescents and 21 HC participants were enrolled in the study. Two participants in the MJ group were unable to complete the scanning protocol. Therefore, present analysis included 18 MJ-users (12 male; mean age: 20.39 ± 0.99) and 21 HC participants (7 male; mean age: 18.43 ± 2.57). Demographic variables are summarized in Table 1.

Table 1.

Summary of demographic and BIS-11 variables for HC and MJ cohorts.

	HC (N = 21)	MJ (N = 18)	Statistics
Demographic Variables			t	df	p	Cohen’s d
Sex (number of males)	7	12	-	-	0.038	-
Age (years), mean (SD)	18.43 (2.57)	20.39 (0.99)	−3.230	26.64	0.003	1.0
Age (years), [range]	[14–21]	[18–21]	-	-	-	-
BIS-11
BIS-NP, mean (SD)	22.29 (4.84)	24.83 (4.63)	−1.672	37	0.103	0.5
BIS-M, mean (SD)	20.62 (4.74)	22.33 (3.53)	−1.263	37	0.215	0.4
BIS-A, mean (SD)	17.29 (3.38)	18.61 (3.47)	−1.207	37	0.235	0.4
BIS-Total, mean (SD)	60.19 (11.03)	65.78 (8.56)	−1.745	37	0.089	0.6

Abbreviations: HC: Healthy control; MJ: Marijuana; BIS-11: Barratt Impulsivity Subscale (version 11); BIS-NP: BIS-Non-planning; BIS-M: BIS-Motor; BIS-A: BIS-Attention

For the MJ group, mean age of initial MJ use was 15.51 ± 1.99 years and mean age of onset of regular MJ use was 16.74 ± 1.37 years. On average, the estimated total lifetime MJ use was 1432.14 ± 1224.30 events. Participants in the HC group reported using alcohol approximately 0.19 times/week whereas MJ-users reported an average alcohol use of 1.25 times/week. Regarding nicotine use, no participants in the HC group reported using nicotine and 9 MJ users reported nicotine use at an average of 2.81 times/week. In terms of use of other substances, specifically cocaine, speed, heroin/opioids, hallucinogens, steroids, and prescription drugs, all HC participants reported no lifetime use. MJ participants who reported using other substances had a lifetime use of < 20 times and tested negative for other drugs except MJ on the day of their study visit. Substance use measures are summarized in Table 2.

Table 2:

Summary of MJ use and other substance use measures for HC and MJ groups

MJ Use Measures	HC (N = 21)	MJ (N = 18)
Age of onset, mean (SD)	-	15.51 (1.99)
Age of onset, [range]	-	[12–18]
Age of regular use, mean (SD)	-	16.74 (1.37)
Age of regular use, [range]	-	[14–19]
Estimated lifetime use (events), mean (SD)	-	1432.14 (1224.30)
Estimated lifetime use (events), [range]	-	[153–5065]
Last use (hours), mean (SD)	-	16.03 (18.95)
Last use (hours), [range]	-	[1 –72]
Measures of other substances
Alcohol, N (%)	4 (19)	14 (77)
Alcohol (events/week), mean (SD)	0.19 (0.55)	1.25 (1.09)
Alcohol (events/week), [range]	[0–2.5]	[0–3.5]
Nicotine, N (%)	-	9 (50)
Nicotine (events/week), mean (SD)	-	2.81 (3.45)
Nicotine (event/week), [range]	-	[0–7]
Cocaine (lifetime use), N (%)	-	6 (33)
Cocaine (past month), N (%)	-	1 (5)
Heroin/Opioids (lifetime use), N (%)	-	3 (16)
Heroin/Opioids (past month), N (%)	-	0 (0)
Hallucinogens (lifetime use), N (%)	-	13 (72)
Hallucinogens (past month), N (%)	-	2 (11)
Prescription drugs (lifetime use), N (%)	-	11 (61)
Prescription drugs (past month), N (%)	-	3 (16)

Abbreviations: HC: Healthy control; MJ: Marijuana

Impulsivity as measured by the BIS-11 showed no significant differences on the BIS-M (p = 0.215), BIS-A (p = 0.235) and BIS-NP (p = 0.103) subscales as well as BIS-Total (p = 0.089) scores between MJ users and HC’s. BIS-11 variables are summarized in Table 1.

3.2. Tissue segmentation

Figure 1 demonstrates the axial, sagittal and coronal tissue segmented slices from a participant in the study. The MRS voxel positioned within the ACC is represented by the white highlighted box. No significant between group differences were observed for within-voxel tissue fractions for GM and WM in the ACC. However, MJ users demonstrated significantly higher CSF content compared to HC’s (p < 0.001). Table 3 summarizes within-voxel tissue segmentation data for both MJ and HC groups.

Table 3:

Mean and standard deviations for percentage tissue fractions, tissue-corrected GABA+ and Glx concentrations, and data quality metrics in the ACC

	HC (N = 21)	MJ (N = 18)	Statistics
ACC % Tissue Fractions	Mean (SD)	Mean (SD)	t	df	p	Cohen’s d
Gray Matter	78.41 (4.34)	78.77 (3.84)	−0.011	37	0.794	< 0.1
Cerebrospinal Fluid	12.87 (2.09)	15.89 (3.14)	−3.530	37	0.001	1.1
Metabolite	Mean (SD)	Mean (SD)	t	df	p	Cohen’s d
Tissue-corrected GABA+	2.59 (0.32)	2.34 (0.35)	2.269	37	0.029	0.7
Tissue-corrected Glx	9.67 (0.95)	9.17 (0.81)	1.739	37	0.090	0.6
1H MRS Data Quality Metrics	Mean (SD)	Mean (SD)	t	df	p	Cohen’s d
GABA+ fit error (%)	5.63 (1.47)	7.16 (1.54)	−3.167	37	0.003	1.0
Glx fit error (%)	2.40 (0.36)	2.84 (0.41)	−3.573	37	0.001	1.1

Abbreviations: HC: Healthy control; MJ: Marijuana; GABA+: Gamma-aminobutyric acid (with co-edited macromolecules); Glx: Glutamate + glutamine; ACC: Anterior cingulate cortex

3.3. ¹H MRS: GABA+ and Glx

Spectral quality determined using fit errors for GABA+ and Glx were < 15% for all participants. However, MJ-using participants demonstrated higher fit errors for both GABA+ and Glx measurements compared to HC’s (Table 3). A representative ¹H MRS spectra fitted for GABA+ and Glx using Gannet is shown in Figure 2. Tissue-corrected GABA+ levels were significantly lower in MJ-using adolescents compared with HC participants (Table 3; p = 0.029). Findings showed a trend towards significance when covaried for age (F_(1,36) = 3.859, p = 0.057, partial η² = 0.097) and remained significant when sex (F_(1,36) = 5.526, p = 0.024, partial η² = 0.133) was included as a covariate. Similarly, when age and sex were included as covariates in a single analysis, tissue-corrected GABA+ remained significantly lower (F_(1,35) = 4.190, p = 0.048, partial η² = 0.107). No significant between group differences were observed for tissue-corrected Glx levels (Table 3; p = 0.090) even when age (F_(1,36) = 1.220, p = 0.277, partial η² = 0.033) and sex (F_(1,36) = 2.295, p = 0.139, partial η² = 0.060) were included as covariates separately or in a single analysis (F_(1,35) = 0.338, p = 0.564, partial η² = 0.010). Furthermore, no significant group × age interaction was observed for GABA+ (F_(1,35) = 1.209, p = 0.279, partial η² = 0.033) or Glx ((F_(1,35) = 0.091, p = 0.765, partial η² = 0.003) levels. We also found no significant interaction effect between group and sex (GABA+: (F_(1,35) = 0.008, p = 0.929, partial η² < 0.001); Glx: (F_(1,35) = 0.533, p = 0.470, partial η² = 0.015).

Figure 2:

Representative spectra (Gannet output) from a study participant. GABA+ and Glx signal is visible at 3ppm and 3.75ppm, respectively. The plot in the top left corner represents the GABA-edited spectrum expanded across a more limited ppm range.

3.4. Association between GABA+ and Glx levels with MJ use measures

For the MJ group, no significant correlations were observed between tissue-corrected GABA+ levels and age of initial MJ use (r = 0.097, p = 0.702), age of onset of regular MJ use (r = −0.145, p = 0.566) or estimated total lifetime MJ use (r = 0.051, p = 0.841). Similarly, no significant correlations were noted between tissue-corrected Glx levels and age of initial MJ use (r = −0.068, p = 0.790), age of onset of regular MJ use (r = 0.191, p = 0.447) or estimated total lifetime MJ use (r = 0.009, p = 0.971). Correlations between last MJ use, measured in hours, and Glx levels were not significant (Spearman’s rho = 0.304, p = 0.235) whereas GABA levels were marginally correlated with last MJ use (Spearman’s rho = −0.459, p = 0.064).

Given the differences in frequency of alcohol and nicotine use, tissue-corrected GABA+ and Glx levels were examined with frequency of alcohol and nicotine use included as covariates. Tissue-corrected GABA+ levels remained significant (F_(1,35) = 4.233, p = 0.047, partial η² = 0.108) and Glx levels showed a trend toward significance when frequency of alcohol and nicotine use were included as covariates (F_(1,35) = 3.711, p = 0.062, partial η² = 0.096).

3.5. GABA+ and Glx correlations with BIS-11

No significant correlations were observed between tissue-corrected GABA+ levels and BIS-11 variables or between tissue-corrected Glx levels and BIS-11 variables in either MJ or HC participants (Table 4). Partial correlation analysis controlling for age and sex, or frequency of alcohol and nicotine use showed no significant associations between tissue-corrected GABA+ or Glx levels with BIS-11 variables (all p>0.05).

Table 4:

Pearson correlations between GABA+ and Glx with BIS-11 variables in both HC and MJ using groups

	HC				MJ
	GABA+		Glx		GABA+		Glx
	r	p	r	p	r	p	r	p
BIS-M	−0.087	0.709	0.192	0.404	−0.05	0.842	−0.205	0.416
BIS-A	0.003	0.989	0.172	0.456	−0.256	0.305	0.136	0.590
BIS-NP	−0.133	0.567	0.422	0.057	−0.073	0.773	0.292	0.240
BIS-Total	−0.094	0.684	0.320	0.157	−0.164	0.515	0.129	0.611

Abbreviations: HC: Healthy control; MJ: Marijuana; BIS-11: Barratt Impulsivity Subscale (version 11); BIS-NP: BIS-Non-planning; BIS-M: BIS-Motor; BIS-A: BIS-Attention; GABA+: Gamma-aminobutyric acid (with co-edited macromolecules); Glx: Glutamate + glutamine

4. Discussion

In this study, we examined levels of GABA and Glx in the ACC of MJ-using adolescents and HCs. As hypothesized, significantly lower tissue-corrected GABA+ levels were observed in adolescent MJ users compared to control participants. No significant between-group differences were observed in ACC Glx levels. The relationship between these neurometabolites and MJ use measures such as age of first use, age of regular use, estimated total lifetime use, and hours since last MJ use were also examined. No significant correlations were observed although a trend towards significance was observed for the correlation between hours since last MJ use and tissue-corrected GABA+ levels. Additionally, we assessed impulsive behavior using the BIS-11 as well as the relationship between GABA and Glx levels with impulsive behavior in both adolescent MJ users and HCs. No significant between-group differences were found for motor, attention, non-planning, and total impulsivity scores on the BIS-11. Additionally, correlations between impulsivity measures and tissue-corrected GABA+ or Glx levels did not reach statistical significance in either participant group.

Findings of lower ACC GABA+ levels in MJ-using adolescents are consistent with a previous study from our group that reported lower ACC GABA+ levels in a separate MJ-using adolescent cohort compared to control participants (Prescot et al., 2013). Interestingly, a recent study found that acute administration of THC in occasional MJ users, either at a single full dose of 300 μg/kg or divided over 3 successive doses of 100 μg/kg did not reduce GABA levels in the ACC or striatum (Mason et al., 2019). Discrepancy in GABA findings between these two studies could be due to differences in methods of MJ/THC administration as well as history of MJ use (heavy versus occasional users) among participant groups. It is important to take into consideration that although THC is the main psychoactive component in MJ, presence of other phytocannabinoids in MJ might lead to different neuropharmacological effects as opposed to administration of THC alone (Bloomfield et al., 2019). Furthermore, preclinical and neuroimaging studies have demonstrated that acute versus chronic MJ use may have differential implications on neurodevelopmental processes and cognitive functioning (Crean et al., 2011; Jager and Ramsey, 2008). In this study, we found no significant association between the different MJ use measures assessed, specifically between estimated total lifetime use, an indicator of chronic use, and GABA and Glx levels. Similarly, no significant association was found between last MJ use, a measure of acute exposure, and Glx levels. Higher GABA levels were found to be marginally correlated with more recent use. Findings from a review by Loeber and Yurgelun-Todd (1999) suggested that chronic exposure to MJ results in changes in cannabinoid receptor levels which impacts levels of neurotransmitters such as dopamine and subsequently lower brain metabolism in frontal and cerebellar regions. Interestingly, alterations in brain metabolism were found to be reversed by acute MJ exposure. Therefore, acute versus chronic MJ use could have different associations with specific neurochemicals which will need to be examined in future studies.

Preclinical studies have also demonstrated lower GABA levels following exposure to THC. For example, Pistis et al. (2002) showed that administration of THC in male rats caused decreases in extracellular GABA levels in the rat PFC. Our findings of lower ACC GABA levels in adolescent MJ users could indicate alterations in different components involved in regulating GABAergic processes. This view is supported by preclinical studies that have reported lower GABA levels which were associated with alterations in GABAergic systems following chronic exposure to THC during adolescence. For instance, adult female rats exposed to THC during adolescent stages showed lower GABA levels in the PFC as well as reduced levels of GAD67, an enzyme that catalyzes the synthesis of GABA (Zamberletti et al., 2014). Chronic administration of THC in adolescent rats was also found to induce long-lasting impairments in inhibitory activity mediated by GABAergic neurons in the medial PFC evidenced by downregulation of GABAergic protein markers, increased neuronal output and impaired gamma oscillatory activity (Renard et al., 2017).

Contrary to our hypothesis, no significant between group differences were observed in ACC Glx levels. This is consistent with another study which also reported no significant differences in ACC Glx levels between chronic MJ users and HCs (Watts et al., 2020). Glx is a composite measure for glutamate and glutamine and is often used as a proxy for glutamatergic processes. Previous examinations of glutamate levels in adolescent MJ users have shown lower Glu levels in the ACC of MJ-users compared to HCs (Prescot et al., 2011; Prescot et al., 2013). In contrast, studies have reported no differences in ACC Glu levels following administration of THC (Mason et al., 2019) or between chronic MJ users and HCs (Watts et al., 2020). To our knowledge, no study has examined levels of glutamine in MJ users, potentially due to the inability to reliably measure and quantify glutamine levels at lower magnetic field strengths. Therefore, lack of difference in ACC Glx levels between MJ users and controls in this study could potentially mask specific changes in glutamate or glutamine levels. Nonetheless, taking into consideration our findings of lower GABA+ levels in the ACC, these findings could indicate alterations in excitatory-inhibitory mechanisms in adolescent MJ users.

Several studies have indicated that the relationship between MJ use and neurochemical levels in the brain may be sex specific. In this study, when sex was included as a covariate, significance of Glx findings were reduced and GABA findings remained significant. Additionally, no significant group by sex interactions were observed for both GABA and Glx levels. In comparison, Muetzel et al. (2013) found a significant group by sex interaction with female MJ users demonstrating lower levels of Glx in the dorsal striatum compared to male MJ users and healthy controls. Similarly, in a separate study, although Glu levels in the dorsal ACC were not significantly different between chronic MJ users and controls, regression analysis in the MJ group showed a trend towards significance implying sex may modulate the relationship between monthly MJ use and Glu levels (Newman et al., 2019). Additionally, findings of lower Glu levels in the ACC of adolescent MJ users were found to be more significant when co-varied for sex (Prescot et al., 2011; Prescot et al., 2013). With regard to the interaction between sex and GABA levels in MJ users, Prescot et al. (2013) found that co-varying for sex reduced the significance of lower GABA level findings in the ACC of MJ users. The inconsistencies in results regarding the effect of sex on neurochemical levels in MJ users could potentially be due to limited sample sizes, differing analysis methods and study population. Preclinical and human studies have shown differential impacts on mood, cognition and neural correlates related to MJ use during adolescence in males and females (Rubino and Parolaro, 2015). Furthermore, sex-related changes in Glu have been reported in healthy participants whereby higher Glu levels were found in female participants compared to male participants. However, these findings were specific to the hippocampus and not the ACC (Hadel et al., 2013). In another study, higher levels of GABA, Glx, and Glu were reported in the dorsolateral PFC of males compared to females (O’Gorman et al., 2011). Moreover, in females, GABA levels have been shown to vary with the menstrual cycle. Specifically, reduction in GABA levels have been observed during the menstrual cycle (Epperson et al., 2002). Therefore, it is important that future studies examining the relationship between MJ use and neurochemicals consider the potential impact of sex on these measures.

To better understand the relationship between impulsivity and MJ use as well as the association between impulsivity and GABA and Glx levels, we examined impulsive behavior using the BIS-11 in both adolescent MJ users and HCs. Overall, similar levels of impulsivity were noted on the motor, non-planning, and attention subscales as well as total impulsivity between MJ users and controls. Previous studies evaluating the relationship between adolescent MJ use and impulsivity using the BIS-11 have reported mixed results. For example, Churchwell and colleagues reported higher non-planning behavior (BIS-NP) in adolescent MJ users compared to controls but found no significant differences in BIS-M, BIS-A or BIS-Total (Churchwell et al., 2010). In contrast, a separate study showed higher BIS-NP, BIS-A and BIS-M scores in MJ-using adolescents compared to controls. Total impulsivity scores were not reported (Dougherty et al., 2013). In line with our findings, lack of differences in BIS-11 measures of impulsivity between current and former MJ dependent users as well as healthy controls have also been reported, although current MJ dependent users demonstrated higher scores on the impulsiveness subscale of the Eysenck Impulsiveness-Venturesome-Empathy questionnaire (Johnson et al., 2010). Impulsivity is a multidimensional construct with various domains that could be differentially associated with neurometabolites and MJ use. Therefore, the BIS-11 scale, designed to assess motor, attention, and non-planning components of impulsivity might not have captured impulsivity domains that are potentially more sensitive to MJ use which might be captured by other measures of impulsivity. Furthermore, in a meta-analysis examining the relationship between MJ use and impulsivity traits using the UPPS-P framework, it has been suggested that impulsivity may be more strongly linked to the negative consequences of MJ use such as difficulties at work or home compared to MJ use itself (VanderVeen et al., 2016). Self-report measures of impulsive behavior were not significantly associated with tissue-corrected GABA+ or Glx levels in either group. In a previous study examining this relationship between cigarette smokers, polysubstance users and healthy controls, no correlation was observed between BIS-11 impulsivity measures and GABA or Glx levels in the dorsal ACC region (Schulte et al., 2017). In this study, although participants were chronic MJ users, no significant impairments regarding comorbid diagnoses or daily functioning were reported which might explain the lack of differences observed in impulsivity measures as well as lack of association with neurometabolite levels.

One of the major strengths of this study was the application of state-of-the-art MRS methodologies to examine GABA levels in the brain. Furthermore, recruitment of participants with low comorbid use of other substances such as alcohol, nicotine, and other illicit drugs which have independently been shown to be associated with altered neurometabolite levels (Chang et al., 2003; Li et al., 2020; Licata and Renshaw, 2010) is another notable strength. Additionally, participants in this study had no current diagnosis of DSM-IV Axis I disorders, as determined using the SCID and KSADS-COMP. This eliminated participants with current medication use and reduced potential confounding effects associated with comorbidities, lending further support that our findings are specific to MJ use.

Several limitations need to be considered when interpreting the results of this study. First, the sample size in this study was relatively small. Therefore, study findings will need to be replicated with larger sample sizes which will allow for rigorous and multiple testing procedures. Second, the cross-sectional design precludes us from determining the causal nature of our findings. Specifically, it cannot be ascertained whether lower GABA levels were a precursor or consequence of MJ use. Future studies with longitudinal designs could provide important insights regarding the trajectory of neurometabolite changes in association with MJ use. The Glx findings should also be interpreted with caution due to methodological limitations. In this study, edited ¹H MRS techniques optimized for the acquisition of GABA were utilized. While this also allowed for the quantification of Glx during data processing, a recent study has indicated that Glx levels quantified from edited ¹H MRS MEGA-PRESS acquisitions showed poor agreement with Glx levels quantified from standard ¹H MRS PRESS acquisitions (Bell et al., 2020). However, the authors indicated that due to the lack of information regarding basal Glx levels, it is difficult to determine which acquisition methods provided an accurate reflection of Glx levels. Another important consideration in interpreting our results is the differing methods of MJ use among participants. Although participants reported that their primary or preferred method of MJ use was smoking (joint, blunt, pipe, etc), this was not exclusive. Participants also indicated using MJ in other forms such as edibles, dabs and vape pens. In recent years, a variety of products such as edibles, dabs, and shatter have been formulated that have a THC concentration of up to 95% (Stuyt, 2018). Furthermore, in a study examining concentration of THC in MJ over a period of 20 years (1995 – 2014), a progressive increase in THC concentration (4% – 12%) was observed (ElSohly et al., 2016). The distinctive effects of THC in these various forms and at different doses are still relatively unknown highlighting an increased need for studies to examine these factors and its impact on the brain. We also cannot overlook the potential acute versus chronic effects of MJ use on neurometabolite levels as well as impulsive behavior. Although participants were requested to abstain from MJ use 24 hours prior to their study visits, this request was not strictly adhered to as participants on average reported using MJ 16 hours prior to the visit.

4.1. Conclusion

In conclusion, the current study findings demonstrated lower ACC GABA+ levels in adolescent MJ users compared to healthy controls, which is consistent with previous examinations of GABA levels in MJ-using adolescents. Although no significant differences were found in ACC Glx levels, our findings indicate potential changes in neurotransmitter levels that might impact excitatory-inhibitory balance critical for neurodevelopmental processes in MJ users. Furthermore, no between-group differences were observed in self-reported impulsive measures and no significant associations were found between impulsive measures and ACC GABA+ or Glx levels in either group. However, given the multidimensional nature of impulsivity and related behaviors, further investigations into the different domains of impulsivity and its relationship with neurochemical correlates in association with MJ use are warranted. Moreover, this study demonstrates the need to further investigate the relationships between neurochemical correlates, behavioral outcomes, and MJ use.

Highlights.

• Adolescent MJ users showed lower GABA levels in the ACC compared to controls

• No significant between-group differences were observed in ACC Glx levels

• MJ users and control participants reported similar levels of impulsive behavior

• GABA or Glx levels were not significantly correlated with impulsivity in either group

• Changes in GABA levels might impact excitatory-inhibitory balance in MJ users