Decreased Anterior Cingulate Cortex γ-Aminobutyric Acid in Youth With Tourette’s Disorder

Abstract

BACKGROUND:

γ-Aminobutyric acid has been implicated in the pathophysiology of Tourette’s disorder. The present study primarily sought to examine in vivo γ-aminobutyric acid levels in the anterior cingulate cortex in psychotropic medication-free adolescents and young adults. Secondarily, we sought to determine associations between γ-aminobutyric acid in the anterior cingulate cortex and measures of tic severity, tic-related impairment, and anxiety and depression symptoms.

METHODS:

γ-Aminobutyric acid levels were measured using proton magnetic resonance spectroscopy. Analysis of covariance compared γ-aminobutyric acid levels in 15 youth with Tourette’s disorder (mean age = 15.0, S.D. = 2.7) and 36 healthy comparison subjects (mean age = 15.9, S.D. = 2.1). Within the Tourette disorder group, we examined correlations between γ-aminobutyric acid levels and tic severity and tic-related impairment, as well as anxiety and depression severity.

RESULTS:

Anterior cingulate cortex γ-aminobutyric acid levels were lower in participants with Tourette’s disorder compared with control subjects. Within the Tourette disorder group, γ-aminobutyric acid levels did not correlate with any clinical measures.

CONCLUSIONS:

Our findings support a role for γ-aminobutyric acid in Tourette’s disorder. Larger prospective studies will further elucidate this role.

Introduction

Tourette’s disorder (TD), also known as Gilles de la Tourette syndrome or Tourette syndrome, is a neurodevelopmental disorder characterized by multiple motor and one or more vocal tics present for at least one year. TD has an estimated prevalence of 1% and is associated with substantial psychiatric comorbidity and impairment.¹ Converging data derived from pharmacologic, model systems, and neuroimaging studies have implicated dopaminergic neurotransmission within frontostriatal neurocircuitry in TD.² However, the direct mechanisms are not fully understood. Over the past decade, the role of γ-aminobutyric acid (GABA), the primary inhibitory neurotransmitter, has been increasingly recognized in the pathophysiology of TD as playing a regulatory role contributing to other monoamine dysfunction.³ For example, clinical evidence suggests that anxiety, known to be associated with GABA deficits,⁴ is frequently present in patients with TD and is often a precipitant of tic exacerbation.¹ Similarly, anxiolytic medications such as benzodiazepines, which increase GABA neurotransmission,⁵ have demonstrated therapeutic effects on tics.⁶ Decreased GABA in TD was supported by a transcranial magnetic stimulation study, which documented decreased cortical inhibition in TD patients that indirectly points to GABA deficiencies.⁷ Furthermore, positron emission tomography imaging studies have confirmed altered binding of GABA receptors in basal ganglia-thalamocortical circuits in TD, with decreased and increased binding in subcortical and cortical brain regions.⁸ Similarly, postmortem studies documented both decreased and increased GABAergic interneurons containing parvalbumin in different regions within the basal ganglia of TD patients.^9,10 These studies point to the complexity of GABA’s role in TD, with different brain regions having opposite GABA binding capacity and neuron number.

Extending the work mentioned previously, here we sought to examine GABA levels in youth with TD using proton magnetic resonance spectroscopy (¹H MRS), which enabled the measurement of in vivo brain GABA non-invasively. There have been only a few studies using ¹H MRS in TD to date. Focusing on the right sensorimotor brain region, Puts et al.¹¹ reported lower GABA in the right sensorimotor cortex of children with TD, with GABA in this region correlated inversely with motor tic severity. On the other hand, Draper et al.¹² observed higher GABA levels in the supplementary motor area for the TD group but no differences for the primary sensorimotor cortex. In adults, Tinaz et al. ¹³ reported no differences in GABA levels in the sensorimotor cortex in TD and healthy control participants. These inconsistencies may be attributed to age, the different brain regions examined, and psychotropic medication status, which can affect GABA levels in the brain.²

Building on these findings, we aimed to examine cortical GABA levels in psychotropic medication-free youth with TD using ¹H MRS. Our region of interest was the anterior cingulate cortex (ACC), responsible for emotion and motor functions, in light of its documented role in TD.^14,15 For example, neuroimaging studies have shown that ACC activity is correlated with tic behavior.^16–18 In addition, electrical stimulation of, or seizure activity in, this region produces involuntary vocalization and behaviors that resemble the symptoms of TD.¹⁵ Our hypotheses were that, compared with healthy control subjects, participants with TD would have lower ACC GABA levels. We also explored associations between ACC GABA levels and measures of tic severity and tic-related impairment, as well as clinical measures of anxiety and depression severity, given the potential role of GABA in these clinical phenotypes.^4,19

Materials and Methods

Participants and procedures

The sample included 15 psychotropic medication-free participants with TD and 36 healthy control participants, aged 12 to 21 years. TD participants were recruited from the Tics and Tourette’s Clinical and Research Programs at the New York University (NYU) Child Study Center and Icahn School of Medicine at Mount Sinai (ISSM). Control subjects were recruited through local advertisements in the greater New York metropolitan area and families of NYU and ISSM staff, as part of our ongoing studies in pediatric psychopathology. The study was approved by Institutional Review Boards of NYU School of Medicine, ISSM, and Weill Cornell Medicine, where the neuroimaging assessments were performed.

Exclusion criteria for all subjects were as follows: psychotropic medication use; the presence of a significant medical or neurological disorder; intelligence quotient less than 80; claustrophobia; magnetic resonance imaging (MRI) contraindication; positive urine toxicology test; and, in females, a positive pregnancy test. Exclusionary diagnoses for participants with TD were bipolar disorder, major depressive disorder, pervasive developmental disorder, psychotic disorder, substance-related disorder in the past 12 months, and current suicidal ideation with intent or plan. Given that attention deficit hyperactivity disorder, obsessive-compulsive disorder, social phobia, and generalized anxiety disorder are highly common in TD,¹ these disorders were not exclusionary. Control participants were excluded if they met criteria for any current or past DSM-IV disorder.

Two board-certified child and adolescent psychiatrists with expertise in diagnosis and treatment of TD interviewed all participants and their parents using the Schedule for Affective Disorders and Schizophrenia–Present and Lifetime Version for Children.²⁰ Fulfillment of DSM-IV criteria was required for the diagnosis of TD.

Clinical measures

TD severity

Tic symptom severity and tic-related impairment were quantified by the Yale Global Tic Severity Scale (YGTSS),²¹ a clinician-administered instrument comprised five domains of tic severity for motor and vocal tics: number, frequency, intensity, complexity, and interference. The total tic score is derived from the sum of the motor and vocal tic scores (range 0 to 50). In addition, an impairment score is measured by the quantitative impact of tics on school, social and family functioning, and self-esteem (range 0 to 50).

Anxiety severity

Anxiety severity was assessed using the Multidimensional Anxiety Scale for Children (MASC),²² a 39-item self-report scale distributed across four major factors: physical symptoms, social anxiety, harm avoidance, and separation anxiety. These subscales yield a total anxiety score, used for the current analyses.

Depression severity

Although a diagnosis of major depressive disorder was exclusionary, depression severity was assessed for all participants using the clinician-administered Children’s Depression Rating Scale–Revised.²³ A raw score of ≥40 has often been used as indicator of significant depressive symptomatology.

Magnetic resonance neuroimaging procedures

All neuroimaging studies, which included limited structural brain MRI examination and two single-voxel ¹H MRS scans, were conducted on a research-dedicated 3.0 T GE MR system with an eight-channel phased-array head coil at the Citigroup Biomedical Imaging Center of Weill Cornell Medicine.

Structural MRI

A three-plane, low-resolution, high-speed scout imaging series was obtained, followed by a series of high-resolution scans, consisting of standard axial, coronal, and sagittal T1-, T2-, and spin density-weighted scans that were appropriately obliqued for prescribing the ¹H MRS voxels of interest. In addition, a T1-weighted spoiled gradient-recalled echo volumetric scan and an axial fast fluid-attenuated inversion recovery scan were performed for brain tissue segmentation and detection of exclusionary focal brain lesions, respectively.

¹H MRS

The GABA-edited ¹H MRS data were acquired using the standard J-edited spin echo difference method²⁴ and then processed as illustrated in Fig 1B and fully described recently.^25,26 Each spectrum was recorded from a single 2.5 × 2.5 × 3.0 cm³ voxel prescribed in the ACC (Fig 1A).

FIGURE 1.

(A) Sagittal (top) and axial localizer images showing the size and location of the ACC voxel of interest. (B) Demonstration of in vivo human brain GABA and Glx detection by ¹H MRS in the ACC: (a) and (b), single-voxel subspectra acquired in 13.4 minutes with the editing pulse on and off and 256 (512 total) interleaved averages; spectrum (c), difference between spectra (a) and (b) showing the edited brain GABA and Glx resonances; spectrum (d), model-fitting of the experimental spectrum (c) to obtain the GABA and Glx peak areas; spectrum (e), individual components of the fits; spectrum (f), residual of the difference between spectra (c) and (d). ACC, anterior cingulate cortex; GABA, γ-aminobutyric acid; Glx, glutamate plus glutamine; MRS, magnetic resonance spectroscopy; NAA, N-acetylaspartate; tCho, total choline; tCr, total creatine.

¹H MRS data processing and quantification

Details of the MRS data quality assessment criteria and procedures used in this study to retain or reject spectra for inclusion in group analyses are provided in the online supplement for Shungu et al.²⁵ For all patients who fulfilled quality assessment criteria, spectral peak areas, which are proportional to the concentrations of the associated metabolites, were obtained as illustrated in Fig 1B (traces [a-f]). Briefly, the GABA and the glutamate plus glutamine resonance, referred to as Glx and coedits within the J-edited difference spectra, were modeled as a linear combination of pseudo-Voigt lineshape functions and then fitted in the frequency domain using a robust and highly optimized public-domain Levenberg–Marquardt nonlinear least-squares minimization routine.²⁷ For normalization across subjects, the GABA and Glx levels were expressed as ratios of peak areas relative to the area of the synchronously acquired and similarly fitted unsuppressed voxel water signal (W).

Assessment of voxel tissue heterogeneity

To estimate the proportions of gray matter, white matter, and cerebrospinal fluid (CSF) contained in each voxel of interest, MEDx software (Medical Numerics, Germantown, MD) was used to segment the brain tissue based on the signal-intensity histogram of each subject’s volumetric (spoiled gradient-recalled echo) MRI. In-house software developed in MATLAB (MathWorks, Natick, MA) was then implemented to generate a segmentation mask for each voxel, from which the proportions of gray matter, white matter, and CSF were determined. These were then compared between the groups, and in case of statistically significant (P ≤ 0.05) or trending (P ≤ 0.10) differences, included in the statistical model as covariates.

Statistical methods

Statistical analyses were performed using Statistical Package for the Social Sciences (SPSS), version 22. Analysis of covariance compared mean GABA/W levels between participant groups. We further examined group differences in other neurometabolite levels (i.e., Glx, choline, creatine, and N-acetylaspartate) in the ACC that were simultaneously detected with ¹H MRS to provide a contrast for examination of GABA. These analyses controlled for any demographic variables showing statistically significant (P ≤ 0.05) or trending (P ≤ 0.10) group differences, as well as any differences in tissue proportions of gray matter, white matter, and CSF, as described previously.

Next, we examined correlations of GABA/W with tic severity (YGTSS motor, vocal, total tic, and tic-related impairment scores) within the TD group alone. Associations between clinical measures of anxiety and depression (MASC, Children’s Depression Rating Scale—Revised) and GABA levels were examined in both the TD group and the combined group, as control subjects exhibited some range in these symptoms despite not meeting clinical significance. Pearson correlations were used for analyses that included normally distributed data, and Spearman rank-order correlations were used for non-normal data. We reran the correlation analyses controlling for age, as some studies²⁸ have suggested that, over time, individuals develop compensatory mechanisms in the brain to regulate or suppress tics. Given our relatively small sample size, and therefore limited statistical power, we report both statistically significant findings (P ≤ 0.05) and those with a trend toward significance (P ≤ 0.10).

Results

Participant characteristics

Table 1 provides participant demographic and clinical characteristics. Two-tailed t tests and chi-square tests revealed no significant group differences in age (t = 1.27, P = 0.21), but TD participants were more likely to be male (at a trend level; χ² = 3.27, P = 0.07) and Caucasian (χ² = 10.83, P = 0.001) compared with control subjects; therefore, sex and ethnicity (Caucasian versus non-Caucasian) were included as covariates in group-comparison analyses. Within the TD group, no demographic variables were correlated with tic severity, tic-related impairment, anxiety severity, or depression severity.

TABLE 1.

Demographic and Clinical Data for Youth With Tourette’s Disorder (TD) and Healthy Control Subjects (HC)

Demographic or Clinical Variable	TD (n = 15)	HC (n = 36)
Age in years (mean ± S.D. [range])	15.0 ±2.7 (12.0–21.8)	15.9 ±2.1 (12.6–21.3)
Sex [n (%)]
Male	10 (67*)	14 (39%)
Female	5 (33%)	22 (61%)
Ethnicity (n [%])
Caucasian	13 (87%)	13 (36%)
African American	0 (0%)	15 (42%)
Other	2 (13%)	8 (22%)
Clinical measures (mean ± S.D. [range])
MASC*	37.0 ± 12.5 (15–66)	28.1 ± 13.5 (4–57)
CDRS-R&*,†	23.4 ± 6.5 (17–38)	18.7 ± 1.7 (17–25)
YGTSS motor tics	11.8 ± 2.8 (6–17)
YGTSS vocal tics	6.0 ± 4.9 (0–13)
YGTSS all tics	17.8 ± 6.2 (6–27)
YGTSS impairment	16.67 ± 8.2 (10–30)
Comorbid diagnoses (n [%])
Anxiety disorder	2 (15%)
OCD‡	4 (27%)
Attention deficit hyperactivity disorder	8 (53%)
Dysthymic disorder	1 (7%)

Abbreviations:

CDRS-R = Children’s Depression Rating Scaled—Revised

MASC = Multidimensional Anxiety Scale for Children

OCD = Obsessive compulsive disorder

YGTSS = Yale Global Tic Severity Scale

^*Data missing for one TD participant.

^†Data missing for one control participant.

^‡Includes clinical and subclinical presentations.

¹H MRS voxel tissue heterogeneity

According to two-tailed t tests, there were no significant or trending group differences in tissue proportions of gray matter or white matter, nor were there differences in ACC mean unsuppressed voxel tissue water signal between participants with TD and control subjects (Table 2). Group differences in ACC CSF approached significance (t = −1.89, P = 0.07). CSF was therefore entered as a covariate in subsequent analyses comparing groups on ACC GABA/W.

TABLE 2.

GABA/W, Water, Gray Matter, White Matter, and CSF in the ACC for Youth With Tourette’s Disorder (TD) and Healthy Control Subjects (HC)

MR Outcome Variable	HC	ID	t (P Value)
GABA/W (mean ± S.D.)	2.85 × 10−3 ±0.44 × 10−3	2.54 × 10−3 ±0.39 × 10−3	2.38 (0.02)
Water (mean ± S.D.)	1.75 × 1012 ± 4.31 × 1011	1.79 × 1012 ± 5.38 × 1011	−0.29 (0.77)
Gray matter % (mean ± S.D.)	57.0 ± 3.9	57.9 ± 3.2	−0.73 (0.47)
White matter % (mean ± S.D.)	30.4 ± 4.1	28.4 ± 4.1	1.38 (0.17)
CSF % (mean ± S.D.)	11.8 ± 2.5	13.3 ± 2.0	−1.89 (0.07)

Abbreviations:

ACC = Anterior cingulate cortex

CSF = Cerebrospinal fluid

GABA/W = γ-aminobutyric acid level relative to unsuppressed voxel tissue water

Group GABA level comparisons

As shown in Table 2 and Fig 2, TD participants had lower ACC GABA/Wrelative to control subjects ([2.54 ± 0.39] × 10⁻³ vs [2.85 ± 0.44] × 10⁻³; t = 2.38, P = 0.02). Results persisted after controlling for sex, ethnicity, and CSF (F = 4.03, P = 0.05). There were no group differences in other neurometabolite levels that were simultaneously detected in the ACC: Glx (F = 1.85, P = 0.18), choline (F = 0.04, P = 0.84), creatine (F = 0.00, P = 0.98), and N-acetylaspartate (F = 1.23, P = 0.26).

FIGURE 2.

Mean GABA concentrations in the ACC for youth with Tourette’s disorder (TD) and healthy control subjects (HC). ACC, anterior cingulate cortex; GABA/W, γ-aminobutyric acid level relative to unsuppressed voxel tissue water.

Associations with clinical measures

Table 3 shows the correlations between ACC GABA/W and clinical measures. For the YGTSSs, no correlations were statistically significant; however, the correlation between YGTSS total tics and GABA/W was moderate (r = 0.44) and approached significance at P = 0.10. When we reran analyses controlling for age, correlations were similar. Measures of anxiety and depression were also not associated with GABA/W in the ACC at either a significant level or a trend level.

TABLE 3.

Correlations^* and Partial Correlations (Controlling for Age) in Youth With Tourette’s Disorder (TD) and Full Sample

Clinical Measure	ACC GABA/W
	TD Group	Full Sample
YGTSS motor tics	0.17 (0.19)
YGTSS vocal tics	0.36 (0.36)
YGTSS total tics	0.44† (0.43)
YGTSS impairment	−0.07 (−0.08)
MASC‡	0.12 (0.13)	−0.08 (−0.07)
CDRS-R ‡,§	0.15 (0.16)	−0.05 (−0.02)

Abbreviations:

ACC = Anterior cingulate cortex

CDRS-R = Children’s Depression Rating Scaled—Revised

GABA/W = γ-aminobutyric acid level relative to unsuppressed voxel tissue water

MASC = Multidimensional Anxiety Scale for Children

YGTSS = Yale Global Tic Severity Scale

Numbers in parentheses indicate partial correlations when controlling for age.

^*Pearson correlations used for normally distributed data, and Spearman Rank correlations used for non-normal data.

^†P = 0.10.

^‡Data missing for one TD participant.

^§Data missing for one TD participant and one control.

Discussion

Although frontostriatal alterations have been documented in TD, the mechanisms involved and the role of specific neurochemical and neurotransmitter dysfunction in TD are still unknown, yet new research directions suggest that decreased GABA activity may be contributory. The present study used ¹H MRS to examine GABA/W levels in the ACC in psychotropic medication-free youth with TD compared with healthy control subjects. As hypothesized, ACC GABA levels were significantly lower in youth with TD compared with healthy youth. Contrary to expectation, GABA levels were not inversely correlated with TD severity. Indeed, within the TD group, GABA levels neither correlated significantly with clinical measures of tic severity or tic-related impairment, nor with measures of anxiety or depression severity.

Literature on ¹H MRS of GABA in TD is still emerging, and numerous questions remain regarding the specific influence of GABA on frontostriatal circuitry in the developing brain. Our finding of lower ACC GABA in TD is consistent with studies using different assessment modalities, including psychopharmacologic,⁶ neuroimaging,⁸ postmortem,^9,10 and genetic,^29,30 which have suggested abnormalities in the GABAergic system in TD. It is important to note that our sample evidenced low tic symptom severity relative to prior TD studies, which was expected given that use of psychotropic medication was exclusionary. This suggests that GABA abnormalities may be present even in more mild presentations of the disorder.

As noted previously, ¹H MRS studies in TD have documented conflicting findings with decreased, increased, and no differences in GABA between individuals with TD and control subjects.^11–13 These conflicting findings may be because of differences in brain regions examined, methodologic inconsistencies, or the inclusion of subjects treated with psychotropic medications in prior studies. In addition, the one study that failed to document any group differences was carried out with adults aged 18 to 52 years, which may play an important confounding role. Research suggests that tics tend to lessen over time,³¹ potentially because of the development of compensatory mechanisms in the brain to regulate or suppress tics. Indeed, data suggest that individuals with TD undergo changes in brain structure and function, allowing the development of greater top-down control over volitional movements, including enhanced tic suppression, over time.²⁸

Although we anticipated that decreased GABA levels would be inversely correlated with tic severity in our TD group, our hypothesis was not supported. In fact, there was a positive trend for association between overall tic severity and GABA/W within the ACC, but this finding was not statistically significant. Draper et al.¹² reported that motor tic severity strongly predicted greater GABA within the sensory motor cortex among their sample of adolescents with TD. However, Puts et al.¹¹ found the opposite pattern of results in their younger TD cohort (age eight to twelve years), with GABA levels inversely correlated with motor tic severity. The discrepancies between studies may reflect age differences in TD samples, and the development of compensatory mechanisms in older adolescents, as mentioned previously. Even so, in our sample, the relationship between GABA and tic severity and impairment showed little change when we controlled for age. A more plausible explanation for divergent findings may be differences in symptom presentation between the samples, as there is great individual symptom variability in TD,³² and studies have suggested region-specific symptom expression.³³ Furthermore, as mentioned previously, tic severity and impairment in our sample was relatively low, and therefore analyses were limited by restricted variability.

Although previous ¹H MRS studies by our group and others reported decreased GABA in individuals with depression and anxiety in the medial prefrontal cortex, including the ACC,^19,34,35 the present study did not document relationships between GABA levels and symptoms of anxiety and depression in either the TD sample or the group as a whole. Again, this might be because our TD sample had a restricted range of depression and anxiety symptom severity, thus limiting our ability to detect significant correlations. Scores on measures of depression and anxiety symptom severity in our sample were generally lower than other TD samples,^36,37 which may be explained by the mild nature of their TD presentation.

Our study has a number of additional limitations. First, our sample size was not sufficiently large to permit meaningful analyses of the effects of comorbid conditions on GABA in TD. Second, we were unable to match the TD and control groups for age, gender, or ethnicity, which might affect GABA levels, given that control subjects were recruited as part of other ongoing studies. However, we did examine group differences in these variables and, when evident, covaried them in our analyses to statistically control for any impact on findings. In addition, it is important to note that ¹H MRS measures total GABA concentration—both extracellular and intracellular—and therefore may not provide an adequate measure of GABA neurotransmission. However, ¹H MRS provides the only technology to date that allows for the assessment of in vivo GABA levels non-invasively and is therefore the only method appropriate for studies of youth.

Despite these limitations, our study adds new data to the growing literature examining the role of the GABAergic system in neurodevelopmental disorders. One important advantage of our sample was that participants were youth who were psychotropic medication-free at the time of the scan, therefore eliminating the possibility that medication use may have influenced levels of GABA and other neurometabolites. If replicated, the present findings may lay the groundwork for translational approaches aimed at enhancing our understanding of the neurobiology of TD, eventually leading to novel and more effective treatments.