Awm278 1.15

Brain Advance Access published December 3, 2007
Brain (2007) Page 1 of 15 Neural correlates of tic severity and cognitivecontrol in children with Tourette syndrome C. L. Baym,1,4, B. A. Corbett,1,2,3, S. B. Wright1,5 and S. A. Bunge1,5,6 1Center for Mind and Brain, 2Department of Psychiatry and Behavioral Sciences, 3MIND Institute, University of Californiaat Davis, 4Department of Psychology, University of Illinois at Urbana-Champaign, 5Helen Wills Neuroscience Institute and6Department of Psychology, University of California at Berkeley, USA These authors contributed equally to this work.
Correspondence to: Silvia A. Bunge, PhD, Department of Psychology &, Helen Wills Neuroscience Institute,University of California, Berkeley, 132 Barker Hall, Berkeley, CA 94720 -3190, USAE-mail: [email protected] Tourette syndrome (TS) is a neurodevelopmental disorder characterized by involuntary motor and phonic tics.
It is hypothesized that excess dopamine leads to an imbalance in the pathways through the basal ganglia, result-ing in unchecked movements via thalamic disinhibition. It has been unclear whether TS is associated with cogni-tive control deficits as well as pure motor control deficits, or whether cognitive deficits are associated with thepresence of comorbid conditions. Furthermore, little is known about the neural underpinnings of TS in child-hood, prior to the long-term effects of medication on brain function. Here, children withTS and typically devel-oping children performed a cognitive control task during event-related fMRI data acquisition.The study included18 native English-speaking 7 13 -year-old children with TS (M = 10.42; 15 males), and 19 healthy, age-matchednative English-speaking volunteers (M = 10.33; 11 males). The task involved three separate manipulations of cog-nitive control. Behaviourally, higher tic severity was correlated with slower task performance on the mostdemanding task conditions. Neurally, higher tic severity was associated with enhanced activation of dopaminer-gic nuclei (substantia nigra/ventral tegmental area) and cortical, striatal and thalamic regions in the direct path-way. Heightened tic severity was also associated with greater engagement of the subthalamic nucleus area,suggestive of a compensatory mechanism. Overall, patients engaged left prefrontal cortex more strongly thantypicals during task performance. These data suggest that children aged 7 13 unmedicated for TS exhibitincreased activation in the direct pathway through the basal ganglia, as well as increased compensatory activa-tion in prefrontal cortex and the subthalamic nucleus.
Keywords: frontal; striatal; paediatric; brain imaging; disorder Abbreviations: ADHD = attention deficit disorder; fMRI = functional magnetic resonance imaging;OCD = obsessive-compulsive disorder; PFC = prefrontal cortex; TS = Tourette syndrome Received June 29, 2007. Revised September 20, 2007. Accepted October 5, 2007 disorder (OCD) and attention deficit disorder (ADHD), Tourette syndrome (TS) is characterized by the expression as well as mood disorders, other anxiety disorders and of tics, or involuntary movements, such as sudden gestures conduct disorders (Comings et al., 1990; Leckman, 1993).
or facial movements, and vocalizations, such as throat- The heredity and genetic underpinnings of TS and clearing or utterances. The diagnostic criteria require the associated disorders are a subject of active investigation expression of involuntary tics defined as sudden, rapid, (e.g. Pauls and Leckman, 1986; Pauls, 2003; Curtis et al., recurrent, non-rhythmic, stereotyped motor movements 2004; Abelson et al., 2005).
and vocalizations (APA, 1994). Both the expression of tics, TS has been associated with a deficit in executive and the effort expended in suppressing them, result in function or cognitive control—i.e. the set of cognitive marked disturbance in the individual's daily functioning.
processes that allow an individual to produce meaningful, Pure cases of TS are the exception rather than the rule; goal-directed behaviour by selecting relevant thoughts and actions. In particular, deficits have been observed in ß 2007 The Author(s)This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License whichpermits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
C. L. Baym et al.
inhibitory control and cognitive flexibility, two components PFC, resulting in an amelioration of symptoms later in life of cognitive control (Bornstein and Baker, 1991; Johannes (Gerard and Peterson, 2003).
et al., 2001; Watkins et al., 2005). However, these studies do Consistent with the anatomical findings, functional not conclusively demonstrate the presence of a cognitive magnetic resonance imaging (fMRI) studies of TS have control deficit, for several reasons. First, studies have not always considered the comorbidity profiles of TS patients.
(e.g. Peterson et al., 2003; Frey and Albin, 2006; Marsh Some researchers have found that TS patients who do not et al., 2007). Disturbances in projections from PFC to the additionally have OCD or another comorbid neurological caudate nucleus or from caudate to other basal ganglia disorder like ADHD are not impaired on tests of executive structures are thought to be responsible for the inability to function (Ozonoff et al., 1998; Como, 2001; Verte et al., suppress tics (Gerard and Peterson, 2003). To date, only a 2005). Second, it is a challenge to find medication-naı¨ve few studies have examined brain activation of TS patients— either adults or children—during performance of a adults, perhaps) are likely to engage in compensatory cognitive task (Johannes et al., 2001; Hershey et al., 2004; strategies that diminish their deficits. Indeed, one study Serrien et al., 2005; Marsh et al., 2007) that is unrelated to found evidence for enhanced performance on a cognitive tic generation or suppression (Stern et al., 2000; Bohlhalter control task in patients with TS (Mueller et al., 2006).
et al., 2006).
Fourth, it may be that some cognitive control processes are The first fMRI study involving children with TS was impaired, whereas others are intact. As such, it is important recently published by Marsh and colleagues (2007). In this to determine which cognitive control processes, if any, are study, children and adults with TS and healthy volunteers deficient in TS.
performed a Stroop task during fMRI data acquisition. The Widely characterized as a frontostriatal disorder, TS is patients and typicals did not differ in terms of performance believed to involve abnormalities in the neural brain on this cognitive control task, but did differ in terms of brain activation patterns. Among patients, stronger behav- prefrontal cortex (PFC) and the striatum (Albin and ioural Stroop interference effects were associated with Mink, 2006). Clues regarding frontostriatal abnormalities increased activation of dorsolateral PFC (DLPFC) and in TS—or, more specifically, a disruption in cortico- ventrolateral PFC (VLPFC), lenticular nucleus, thalamus striatal-thalamic-cortical circuitry (CSTC; Butler et al., and mesial PFC. In contrast, healthy individuals exhibited 2006)—come primarily from anatomical studies in adults increased activation in right VLPFC with better perfor- and children (Gerard and Peterson, 2003; Pringsheim et al., mance on their task. Interestingly, within the patient group, 2003; Frey and Albin, 2006; Ludolph et al., 2006), positron greater tic severity was associated with increased activation emission tomography studies, primarily in adults (Chase in DLPFC, possibly reflecting an ineffectual compensatory et al., 1984; Baxter and Guze, 1992; Ernst et al., 1999; mechanism (Marsh et al., 2007). While this prior study Gilbert et al., 2006), magnetic resonance spectroscopy does offer insight into the mechanisms of cognitive control (DeVito et al., 2005), and neuroscientific research on across children and adults with TS, it did not examine the non-human animals (Saka and Graybiel, 2003). Recent whole brain, leaving open the possibility that there may be studies have shown promise for the use of deep brain stimulation of several components of the basal ganglia in Furthermore, as many of the patients (in particular the the treatment of patients with severe TS (Mink et al., 2006).
children) were medicated, it remains unclear whether the Peterson and colleagues (2003) conducted volumetric differential patterns of activation observed between groups analyses in a large sample, and showed reduced caudate were attributable to Tourette syndrome itself or to changes nucleus volumes in both children and adults with TS.
in brain structure/function as a result of medication.
Notably, smaller bilateral caudate volumes in childhood A number of lines of evidence point to dysfunctional have been shown to be predictive of tic severity and dopamine neurotransmission in TS (Pringsheim et al., obsessive compulsive symptoms in adulthood (Bloch et al., 2003; Albin and Mink, 2006). In particular, medications 2005). Additionally, other anatomical studies have shown that block dopamine receptors are used to control tics, and a variety of anatomical measures point to abnormal (Ludolph et al., 2006), midbrain, potentially including dopaminergic function in TS (Albin and Mink, 2006). It dopaminergic nuclei (Garraux et al., 2006), and thalamus is hypothesized that excess dopamine leads to an imbalance (Lee et al., 2006). In contrast, smaller volumes in patients of activity in the direct and indirect pathways through the with TS have been found in the hippocampus (Lee et al., basal ganglia, resulting in involuntary movements via 2006). With respect to PFC volumes, the data have been disinhibition of the motor thalamus (Albin and Mink, more mixed (Peterson et al., 2001; Fredericksen et al., 2006). However, the precise nature of the dopamine 2002). Based on the pattern of findings for PFC volumes in dysfunction in TS is not yet understood. Some of the children and adults with TS relative to controls, it has been brain differences between adults with and without TS, such hypothesized that children who learn to effectively suppress as decreased availability of D2 receptors in those with TS their tics during development show increased growth of (Gilbert et al., 2006), may be indicative of compensatory Cognitive control in Tourette syndrome mechanisms counteracting the underlying disorder. Indeed, were intentionally minimized on this task by the fact that some of the differences observed between adults with and participants had multiple seconds to prepare for an without TS may be a result of long-term use of upcoming trial, and that they were explicitly given the medications, and/or the long-term consequences of volun- rule at the start of each new trial. Nonetheless, we expected tary and/or involuntary recruitment of compensatory brain that children would have greater difficulty when switching from one rule to the other than when employing the same Several studies have found evidence of compensatory rule as on the preceding trial. Based on prior behavioural mechanisms in adults with TS: i.e. enhanced prefrontal research (Watkins et al., 2005), we predicted that children activation in TS patients relative to healthy adults, despite with TS would exhibit magnified behavioural costs for similar performance levels (Johannes et al., 2001; Hershey Switch versus Repeat trials relative to typical children.
et al., 2004). Intriguingly, as noted previously, one study Finally, the response selection manipulation compared showed enhanced cognitive control in adults with TS, Incongruent and Congruent trials, in which the two possibly as a result of years of tic suppression (Mueller potentially relevant features of a stimulus—its colour and et al., 2006). If the compensatory developmental hypothesis orientation—were associated with different responses or the of Gerard and Peterson (2003) summarized above is same response. Based on the prior behavioural research correct, it would emphasize the point that it is imperative described previously, we predicted that children with TS to diagnose TS as early as possible, so as to provide would exhibit magnified behavioural costs for incongruent children with coping strategies that could lead to beneficial versus congruent trials compared with typical children.
long-term brain plasticity and a better long-term prognosis.
The rule manipulation contrasted arbitrary S-R mappings These studies provide a strong rationale for characterizing brain anomalies—both anatomical and functional—in Direction rule). Demands on rule representation were paediatric patients as well as in adults.
minimal on this task, in that the currently relevant rule In our event-related fMRI study, we sought to test for was explicitly indicated on each trial. However, we still differences between unmedicated children with TS and age- expected that the Color rule would tax rule representation matched typically developing children (referred to here as more strongly than the Direction rule, because it relied on typicals) in terms of performance and brain activation newly learned S-R associations. We did not have any associated with a cognitive control task. We devised a task predictions regarding group differences in terms of rule (Fig. 1A) that included manipulations of three separate representation, but sought to test whether children with TS aspects of cognitive control: (i) task-switching, or the ability would exhibit magnified behavioural costs for Color versus to properly update the currently relevant task information; Direction trials relative to typical children.
(ii) response selection, or the ability to ignore competing In terms of brain imaging analyses, we sought to test for information to select between response options and (iii) signs of hyperdopaminergic transmission in patients with rule representation, or the ability to retrieve and use TS, as indexed by increased activity in the substantia nigra currently relevant response contingencies. By including (SN) and/or ventral tegmental area (VTA), dopamingergic three separate manipulations, we could assess whether nuclei with neuromodulatory effects on widespread regions children with TS were impaired on one or more of these of the brain. These nuclei are located sufficiently close to facets of cognitive control, and examine the neural one another that they are difficult to distinguish from one correlates of task performance. Based on the few behav- another (Aron et al., 2004). Thus, rather than attempting to ioural studies of cognitive control that have been conducted examine the SN and VTA independently from one another, in children with TS, we predicted that this group would we sought to test whether the SN/VTA area would be more have difficulty switching between tasks and selecting strongly engaged by patients with TS.
between competing response alternatives.
With respect to the relationship between brain activation The task was intended to be challenging, but not overly and performance in TS, we entertained three possible difficult, for children aged 7–13 (Fig. 1A). Prior to MRI outcomes: namely, that unmedicated children with TS scanning, participants were taught stimulus-response (S-R) would (i) exhibit a deficit in cognitive control, associated mappings based on ‘Direction' and ‘Color' rules. On with specific differences in brain activation relative to Direction trials, participants were to press a left button in typicals; (ii) exhibit no cognitive control deficits and no response to a leftward pointing stimulus, and to press a differences in task-related brain activation, after taking into right button for a rightward pointing stimulus. On Color account other comorbidities, in particular ADHD or (iii) trials, participants were to press a left button in response to exhibit no deficits in performance on a cognitive control a red stimulus, and a right button in response to a blue task, but exhibit enhanced brain activation relative to stimulus, or vice versa.
typicals, suggestive of the engagement of compensatory The task-switching manipulation compared trials in which neural mechanisms.
the rule was different from that of the previous trial A pathway between right inferior frontal cortex (IFC) (Switch) to those in which the rule was the same as that of and the right subthalamic nucleus (STN) has recently been the previous trial (Repeat). Demands on task-switching characterized in humans (Aron et al., 2007). Through this C. L. Baym et al.
Fig. 1 Task design and behavioural data. (A) Participants were instructed to respond with a left or right button press to each stimulus,based on the relevant rule for that trial (Color or Direction). Variable ITIs (2 8 s) were interleaved between each trial. (B) Box-and-whisker plots of accuracy and RT showing min/max/first quartile/third quartile/median divided by activation for ConRep and IncSw, and bygroup. (C) Behavioural correlations between estimated IQ and accuracy on IncSw trials, separately for patients and typicals and betweentic severity and RT on IncSw trials for the patients only. RTs are reported for correct trials only. Medication naive patients are indicatedwith circle markers. Patients previously medicated for ADHD symptoms are indicated with triangle markers.
Cognitive control in Tourette syndrome and/or other hyperdirect pathways to the basal ganglia, Excluded participants motor inhibition could proceed rapidly, bypassing the Apart from the 37 participants included in the study, eight inhibitory indirect pathway through the basal ganglia additional participants were excluded from the final analysis. Five (Mink, 2001; Aron and Poldrack, 2006). We hypothesized patients and one typical were excluded on the basis of excessive that patients with TS might compensate for weakened basal head motion during scanning (greater than 5 mm within a run), ganglia control via increased engagement of a hyperdirect one patient on the basis of poor task performance, and one typical pathway. Thus, we sought to test the prediction that on the basis of low estimated IQ.
children with TS would engage the STN more strongly thantypicals. Such a pattern would be consistent with a Recruitment and screening compensatory mechanism; in contrast, stronger STN Children with TS were recruited via the Tourette Syndrome activation in typicals than in patients with TS would be Association based in the United States, clinical referrals, local consistent with diminished activity in the indirect pathway advertisements, physician referrals and through the University of California at Davis. The diagnosis of TS was based on DSM-IV dopamine transmission and stimulation of D2 receptors criteria (APA, 1994) and established by: (i) a previous diagnosis of (Mink, 2001). Like the SN/VTA, the STN is a small nucleus; TS by a psychologist, psychiatrist, neurologist or behavioural however, relying on coordinates from prior fMRI studies paediatrician; (ii) clinical judgment (B.A.C.) and (iii) confirma-tion of the presence of motor and phonic tics by the Yale Global (Aron et al., 2004, 2007; Aron and Poldrack, 2006), we Tic Severity Scale (YGTSS). The parents of all the participants also sought to test our theoretically motivated prediction that completed a semi-structured interview, the Diagnostic Interview there would be group differences in the STN area.
Schedule for Children (DISC; Shaffer et al., 2000) to determine theextent of symptomology across other related disorders (e.g. OCD,ADHD, anxiety).
Materials and Methods Typically developing children were recruited through area Inclusion and exclusion criteria schools, fliers and recreational centres. Following initial contact,potential participants were screened via parent interview for the Children aged 7–13 with TS and healthy age-matched controls absence of neurodevelopmental disorders. Individuals with a were recruited for participation in this study, and were history of serious physical illness (e.g. endocrine, cardiovascular compensated financially for their involvement. Children with or neurological disorders) were excluded from enrollment in the serious neurological, psychiatric or medical conditions, other than study. The groups were not matched on the basis of intellectual TS with or without comorbid ADHD or OCD, or with an functioning (Table 1). After thorough explanation of the study estimated IQ below 75 were not included in the study. Rather was completed, informed written consent was obtained from than considering OCD and ADHD as exclusion criteria for the parents and verbal assent was obtained from all research subjects patients, we included scores on standard measures of OCD and prior to inclusion in the study. This study was approved by ADHD as continuous variables in multiple regression analyses of the Institutional Review Board of the University of California performance and brain activation. Children who had taken medication to treat tics were excluded from the study.
Each child underwent an extensive clinical evaluation prior toscanning by, or under the supervision of, paediatric neuropsychol- The study included 18 native English-speaking 7–13-year-old ogist (B.A.C.). Diagnostic testing was completed following children with TS (M = 10.42; 15 males, 3 females), and 19 healthy, informed consent procedures on the first of two visits. The age-matched native English-speaking volunteers (M = 10.33; 11 Wechsler Abbreviated Intelligence Scale (WASI; Wechsler, 1999) was males, 8 females). Children in the two groups did not differ in age used as an estimate of IQ. This measure was administered to all (P = 0.996). Although the two groups had different proportions of participants, unless an IQ score from a more comprehensive males and females, two-sample t-tests revealed no significant measure had been recently completed (52 years). Children with a performance differences between males and females (all P40.05).
score of 575 on the WASI were excluded from the study.
Three patients and one typical child included in the study were On the first visit, each child completed a psychological left-handed. One patient had a history of epilepsy, but was not an assessment that included 15 subsections of the DISC (Shaffer outlier relative to the other patients on any of our other measures.
et al., 2000), the YGTSS (Leckman et al., 1998), and the ChildYale-Brown Obsession and Compulsion Scale (CY-BOCS; Scahillet al., 1997). Additionally, parents were given the following Medication status measures to complete at home, prior to the scanning visit: the Sixteen out of 18 patients were medication-naı¨ve, as per parental Behavior Rating Inventory of Executive Function (BRIEF; Gioia reports. The other two children had taken atomoxetine (brand et al., 2000), Conners' Parent Rating Scale-Revised (Short) (CPRS- name StratteraTM) to treat ADHD symptoms. One child had R:S; Conners, 2001), Child Behavior Checklist (CBCL; Achenbach ended treatment approximately 1 month prior to testing; the other and Rescorla, 2001), Social Communication Questionnaire (SCQ; had been off atomoxetine for 40 h prior to testing. Behavioural Rutter et al., 2003), Multidimensional Anxiety Scale for Children and fMRI analyses performed with and without these two children (MASC; March, 1999) and Parenting Stress Index (PSI; Abidin, indicated that they were not outliers.
1995). More details regarding each of the assessments are provided C. L. Baym et al.
Table 1 Results of psychological assessments Executive dysfunction Behavioural problems Social communication problems Table of mean scores for assessment results, ordered from most to least significant difference between groups. P-values indicate level ofsignificance of unpaired comparisons between groups. Standard deviations are reported in parentheses. Test results are ordered from mostto least significant effect.
in the Supplementary Material. Questionnaires filled out at home (ITIs) interspersed between them. The order of trial type were properly and entirely completed for 29/37 participants (14 presentation and jittered fixation was determined using Optseq2, patients and 15 typicals); for the remaining participants, 1–4 of an optimizing program designed to allow for maximal efficiency the questionnaires could not be scored.
in deconvolving trials from each condition and baseline activation(Dale, 1999). Jittered fixation lasted from 2 to 8 s and was interspersed throughout the functional runs.
Participants were instructed on a trial-by-trial basis to use one oftwo visually presented rule cues (‘Color' or ‘Direction') to Experimental procedure determine the appropriate response for a given target stimulus Prior to MRI scanning, children were introduced to the scanner (Fig. 1A). The target stimuli were cartoons from the Disney movie environment through the use of a mock scanner. They were asked ‘Finding Nemo'; the colour and orientation of these stimuli were to lie down in the mock scanner, listen to prerecorded sounds of manipulated in Adobe Photoshop CS2. Participants were taught the pulse sequences, and practice lying still and making button the relevant stimulus-response (S-R) mappings for the two rules presses. After participants were familiarized with the scanning prior to scanning (Fig. 1A). On Direction trials, participants were procedure, they learned S-R mappings for the Color and Direction to press a left button in response to a leftward pointing stimulus, rules, and practiced applying them both on paper and in a 3 min and a right button for a rightward pointing stimulus. On Color computerized session. During this training session, participants trials, half of the participants were taught that a red stimulus provided 1 ml saliva samples at 20 min intervals following initial indicated a left-button press and a blue stimulus indicated a right- exposure to the mock scanner as part of a study on the biological button press, and half were taught the reverse S-R mapping.
response to stress in patients with TS.
Participants used the index and middle fingers of their right handto respond.
MRI data acquisition Experimental design Imaging was performed using an 8-channel phased-array coil on a3-Tesla Siemens Trio MRI scanner (Siemens Medical Solutions, The experimental task followed a 3  2 factorial design, enabling Erlangen, Germany) at the University of California at Davis the assessment of independent contributions of rule difficulty, Imaging Research Center (Sacramento, CA). Participants viewed task-switching and interference suppression. The rule difficulty visual stimuli on a projection screen using an angled mirror and manipulation contrasted arbitrary S-R mappings learned on theday of testing (Color) with previously learned S-R mappings responded using a button box in the right hand.
After acquisition of a T2 localizer scan, four functional runs (Direction). The task-switching manipulation compared trialsinvolving a rule switch relative to the prior trial (Switch) with were collected, each 4.5 min long (TR = 2000 ms, TE = 25 ms, 34 those in which the rule was the same as on the previous trial axial slices, no interslice-gap, 3.4  3.4  4 mm3 voxels, flip (Repeat). The interference suppression manipulation compared angle = 90, field of view = 220 mm, 135 volumes per run).
trials in which two features of a stimulus were associated during training with different responses (Incongruent) or the same Correction (3D-PACE) sequence was used to minimize motion artifacts during fMRI data acquisition. The 3D-PACE sequenceprospectively adjusts scan parameters throughout a run on thebasis of real-time assessment of head motion (Siemens Medical Timing parameters Solutions; Thesen et al., 2000). Four volumes from the start of On each trial of the event-related fMRI paradigm, a fixation cross each functional scan were removed from analysis to account for appeared for 200 ms, followed by an instructional cue (‘Color' or magnetic field equilibration. Following the functional scans, high- ‘Direction') for 2300 ms, and a target stimulus for 1500 ms resolution T1 MPRAGE anatomical images were acquired.
(Fig. 1A). Color and Direction trials were pseudorandomly Across the four functional scans, participants completed 48 ordered throughout a scan, with variable intertrial intervals trials of each of the following conditions: Congruent Repetition Cognitive control in Tourette syndrome (ConRep), Congruent Switch (ConSw), Incongruent Repetition basal ganglia and thalamus. Voxel-based analyses were restricted to (IncRep) and Incongruent Switch (IncSw) trials. For each of these voxels within two anatomical masks created in the MarsBar conditions, an equal number of trials involved the Color rule and toolbox for SPM2 (Brett et al., 2002): a lateral prefrontal mask the Direction rule. Each scan included six trials of each of the including 4816 voxels (bilateral B.A. 6, 8, 9, 10, 44, 45, 46, 47), eight resulting trial types. The order of the functional scans was and a subcortical mask including 1475 voxels (bilateral caudate, counterbalanced across subjects.
putamen, globus pallidus and thalamus). We report results forcontiguous clusters of 5 or more voxels within one of these masks Image preprocessing meeting an uncorrected threshold of P50.005. At this threshold,41 cluster would be expected by chance alone within either the Data were analysed using SPM2 (Wellcome Department of prefrontal or subcortical mask.
Cognitive Neurology, London, UK). Functional volumes from Region-of-interest (ROI) analyses were performed to further each participant were corrected for interleaved slice acquisition, characterize the activation profile across all four task conditions and were translated using a rigid-body motion correction.
for regions of a priori interest. ROIs in lateral PFC were defined Functional volumes were normalized to an EPI template using functionally from a voxel-based t-test (patients4typicals for all correct trials4baseline). ROIs in SN/VTA and nucleus accumbens 3  3  4 mm3 voxels. The SPM EPI template has been validated were defined functionally from a voxel-based multiple regression for use in normalization of brain volumes for children aged 56 analysis used to identify activations positively correlated with (Burgund et al., 2002; Kang et al., 2003). Finally, functional YGTSS. For basal ganglia structures and the thalamus, we used volumes were smoothed using an 8 mm full-width at half anatomical ROIs from the MarsBar toolbox. Mean parameter maximum isotropic Gaussian kernel.
estimates for a given ROI were computed for each participant and To examine individual movement, we computed root mean condition, and were submitted to repeated measures ANOVAs and square (RMS) values for translation for each participant across post-hoc comparisons. For anatomical ROIs, both left- and right- sided regions (e.g. left and right putamen) were submitted to (M = 0.398; range = 0.567). As expected, patients exhibited more analysis, and hemisphere was included as a within-subject variable.
head movement than typicals (t = 2.82, P50.009). However, In total, eight different regions from lateral PFC, basal ganglia multiple steps were taken to ensure that the fMRI data were not and the thalamus were submitted to ROI analyses (Table 2).
compromised by excessive head motion, as noted above. First, we Taking into account this number of ROIs, the corrected statistical used a 3D-PACE prospective acquisition sequence during data threshold would be P50.00625 with an alpha of 0.05. None of acquisition, so as to minimize the impact of head motion on our our ROI-based results met correction for multiple comparisons.
results. Secondly, we excluded participants for whom head motion However, the pattern of results observed in these ROIs were highly was greater than 5 mm within a run. Finally, we included motion consistent with our predictions, and are featured here as initial regressors as covariates of no interest in the analysis of individual findings regarding the neural mechanisms of cognitive control in subjects' data.
unmedicated children with TS.
Single-subject fMRI analysesStatistical analyses were performed using the general linear model in SPM2. FMRI time-series data were modeled as a series of events, time-locked to the onset of each trial, and were convolvedwith a canonical hemodynamic response function. The resulting The TS group differed from the typical group in terms of functions were used as covariates in a general linear model, along tic severity, as measured by YGTSS. Indeed, this was the with a basis set of cosine functions that high-pass filtered the data, factor that most prominently distinguished the two groups as well as a covariate for session effects. Six motion regressors (Table 1). However, the groups differed in other ways as (indicating amount of translation and rotation in the x, y and z well. Listed in order from strongest to weakest group dimensions for each 2 s TR) were included in the analysis as differences, the TS group also scored higher on a parental covariates of no interest. The least-squares parameter estimates of assessment of executive dysfunction (BRIEF), Attention height of the best-fitting synthetic hemodynamic response Deficit Hyperactivity symptoms (Conners'), Obsessive- function for each condition were used in pair-wise contrasts, Compulsive symptoms (CY-BOCS), behavioural problems and the resulting contrast images, computed on a subject-by- (CBCL), social communication problems (SCQ) and subject basis, were submitted to group analyses. Incorrect trialswere modeled separately, and were not included in the fMRI anxiety levels (MASC). Although the patients with TS in our sample were high-functioning (mean estimated IQscore  SD: 106  14), they had lower estimated IQ scores Group fMRI analyses than our sample of typicals. Indicative of the fact that TStakes a toll on the family rather than merely the individual, Several analytic approaches were used to characterize differences parents of children with TS scored higher on a parental in brain activation between children with and without TS. First, stress index than parents of typical children (mean  SD for voxel-based two-sample t-tests were used to directly compare thegroups. Second, voxel-based multiple regression analyses were parents of patients versus typicals: 76  21 versus 57  16; used to assess the effects of various factors both within and across groups. For these two types of analyses, we limited the search Because of the strong prevalence of obsessive-compulsive space to regions for which we had specific predictions: lateral PFC, and attention-deficit symptoms in patients with TS, and C. L. Baym et al.
Table 2 Results of multiple regression analyses conducted on ROIs Bilateral caudate Bilateral globus pallidus Bilateral putamen YGTSS R2 = 0.27, P50.04 Bilateral thalamus YGTSS R2 = 0.32, P50.02 YGTSS R2 = 0.39, P50.009 EstIQ R2 = 0.34, P50.017 BRIEF R2 change = 0.177, P50.033 YGTSS R2 = 0.32, P50.02 EstIQ R2 = 0.36, P50.011 EstIQ R2 = 0.46, P50.003 Nucleus Accumbens YGTSS R2 = 0.39, P50.007 Estimated IQ, YGTSS, CY-BOCS, Conners', MASC, BRIEF and age were included as independent variables. All correlations were in thepositive direction. No hemispheric differences were observed in the regions noted as bilateral; average R2- and P-values are reported forvariables that account for a significant portion of the between-subject variance of activation for an ROI (P50.05).
because these symptoms are not always severe enough to placed increased demands on both task-switching and rule meet full diagnostic criteria for OCD or ADHD, we purposefully used continuous measures of ADHD and Taken as a whole, the two groups did not differ OCD rather than categorical ones. However, using con- significantly from one another in performance (accuracy: servative cut-offs (t-score 570 for the Conners' and raw F1,35 = 1.6, P = 0.21, RTs: F1,3551, P = 0.99), and there were score 5 8 for the CY-BOCS), seven patients scored in the no significant interactions involving Group. However, clinical range for TS + OCD, one for TS + ADHD and two because our TS group exhibited a range of tic severity, for TS + OCD + ADHD. The remaining eight patients and from mild to marked (total YGTSS scores ranged from all typical participants scored in the normal range on both 8 to 31), we sought to examine more closely the relation- the Conners' and CY-BOCS.
ship between tic severity and task performance.
Multiple regression analyses of behavioural Repeated-measures ANOVAs were performed on accuracy and response time (RT) data from the Nemo task. We We performed stepwise multiple regression analyses to examined the between-subjects effect of Group (Patients, examine the effects on performance of age, estimated IQ Typicals), and the within-subjects effects of Congruency and total scores on the YGTSS, BRIEF, Conners', MASC (Congruent, Incongruent), Switching (Switch, Repeat) and and CY-BOCS. These regression analyses were performed Rule (Color, Direction).
separately for accuracy and RTs on the more difficultcondition in each of the three task manipulations: Effects of task manipulations on Incongruent, Color and Switch trials. Taking into account performance across all participants the number of whole-brain regressions computed (four- Across groups, accuracy and RTs were affected by each of the teen), the corrected statistical significance was P = 0.0036 three task manipulations (Fig. 1B). Children performed worse with an alpha of 0.05. Correlations that met correction for when they had to switch from one rule to another (Switch multiple comparisons are noted with an asterisk (). Given versus Repetition trials; accuracy: F the dearth of behavioural data on cognitive control 1,35 = 71, P50.001; RTs: paradigms for unmedicated children with TS, correlations 1,35 = 48, P50.001). Additionally, performance was worse when the irrelevant dimension of the stimulus was associated that did not meet correction for multiple comparisons are with the opposite response from the relevant dimension also listed below.
Across all children, accuracy was positively correlated with P50.001; RTs: F1,35 = 30, P50.001). Finally, performance IQ for Incongruent trials (R = 0.50; P50.003), Switch trials was worse for the arbitrary rule than for the non-arbitrary (R = 0.46; P50.006), and Color trials (R: 0.37; P50.03) one (Color4Direction trials; accuracy: F1,35 = 43, P50.001; (Fig. 1C). Additionally, Switch accuracy was correlated with age (R2 change = 0.13; P50.02). A stronger correlation Congruency  Switching (F1,35 = 18, P50.001) and between accuracy and estimated IQ across the entire group Congruency  Rule (F1,35 = 24, P50.001) were observed in was observed for IncSw trials (R = 0.54; P50.0001) than for terms of accuracy. In terms of RTs, there was a significant ConRep trials (R = 0.282; P50.05). RTs were positively interaction of Congruency  Rule (F1,35 = 13, P = 0.001).
correlated with Age for Incongruent trials (R = 0.40; These results indicate that the congruency manipulation P50.02), and for Switch trials (R = 0.35; P50.05), but not Cognitive control in Tourette syndrome Fig. 2 Voxel-based activation across all correct trials relative to fixation, shown across all participants (P50.005, 5 -voxel extent, withinlateral prefrontal and subcortical masks). Regions within the lateral PFC mask are shown in magenta and regions within the subcorticalmask are shown in cyan. Top row: axial slices for z-coordinates of +48, +32 and 0, and coronal slice for y-coordinate of sagittal slices for x-coordinates of 17 and +42. Numbers correspond to approximate Brodmann areas. C = Caudate nucleus.
for Color trials (P40.05). Thus, performance on Incongruent and Switch trials, but not Color trials, was correlated with IQand age across the whole group.
Task-related activation across groups Focusing on the patient group only, we again found that As a first pass, we conducted voxel-based contrasts within accuracy was positively correlated with IQ for Incongruent our anatomical masks to identify regions in lateral PFC, trials (R = 0.55; P50.023), and Switch trials (R = 0.61; basal ganglia and/or thalamus that were generally engaged P50.009), but not for Color trials (all P40.05). RTs during task performance across the two groups, collapsing were positively correlated with tic severity (as measured by across conditions (Fig. 2). For correctly performed trials YGTSS) across the TS group for Incongruent trials relative to fixation, clusters of activation included bilateral (R = 0.51; P50.037) and Switch trials (R = 0.52; P50.04), middle and inferior frontal gyri (BA 6, 8, 9, 10 and 46), and but not for Color trials (P40.05) (Fig. 1C). The strongest bilateral caudate nucleus.
correlation between task performance and YGTSS wasobserved for the IncSw trials (R = 0.53, P50.02); a slightlyweaker Comparisons between patients and controls (R = 0.42, P50.04).
Voxel-based two-sample t-tests were performed to identify regions in PFC, basal ganglia and/or thalamus that were Conners' and BRIEF) were not significantly correlated engaged differentially by the two groups. No regions were with RTs for the TS group (all P40.05). This latter finding engaged more strongly by typicals than patients. As does not necessarily mean that performance on this task is predicted, greater activation for patients than typicals was correlated with TS but not with other disorders, such as observed in lateral PFC. Two small clusters were located in ADHD or OCD; rather, it is likely to be related to the fact left middle frontal gyrus (MFG; BA 8, 46) and left superior that we recruited children with a primary diagnosis of TS.
frontal gyrus (SFG; BA 9).
The exclusion of the two participants who had been Follow-up ROI analyses revealed that the left SFG cluster medicated for ADHD symptoms only slightly weakened was engaged relative to baseline by patients but not typicals, the behavioural correlations with YGTSS for Incongruent in a manner that was insensitive to our task manipulations trials (R = 0.44) and Switch trials (R = 0.45). In summary, (all P40.13). In contrast, the left MFG cluster exhibited we found that (i) overall, better performance on the significant effects of both switching (F1,35 = 5.4, P = 0.027) Incongruent and Switch trials was associated with higher IQ and rule (F1,35 = 4.3, P = 0.046) (Fig. 3). Thus, patients and older ages and (ii) children with TS who responded exhibited overall greater activation in two focal regions of most slowly on Incongruent and Switch trials tended to PFC, one of which was specifically associated with rule have the highest tic severity.
representation and task-switching.
C. L. Baym et al.
accumbens (Fig. 4; Supplementary Table 1). The SN/VTA,an a priori predicted region that was not included in themask of subcortical regions created with Marsbar, alsoexhibited a positive correlation with YGTSS. These findingsindicate that patients with more severe TS symptomsexhibited heightened task-related activation in a number ofregions in the CSTC loop.
To test whether the regions identified as showing a positive correlation with YGTSS could have been correlatedinstead with ADHD or OCD symptoms, an additionalmultiple regression analysis was performed on the fMRIdata, including YGTSS, Conners', CY-BOCS and IncSwaccuracy as regressors. In effect, none of the activationsidentified as being associated with tic severity could beattributed to either OCD or ADHD severity. Further, theexclusion of the two children who had been medicated forADHD symptoms did not change the set of regions foundto be correlated with YGTSS.
ROIs in SN/VTA and nucleus accumbensAs noted previously, we sought evidence suggestive ofelevated dopaminergic activity in patients relative tocontrols. To this end, ROI analyses were performed tofurther characterize the activation profile and groupdifferences in the nucleus accumbens and SN/VTA area,as identified from the positive correlation with YGTSS(Fig. 5). To isolate the SN/VTA area from a larger cluster ofactivation, we masked the cluster with 5 mm spherescentered on MNI coordinates of 9, these regions were identified from a positive correlation ofIncSw4fixation brain activation with YGTSS among Fig. 3 Lateral prefrontal ROI (BA 8/46; 48, 23, 24; 16 contiguous patients, multiple regression analyses were performed to voxels) derived from all correct trials 4 fixation for patients 4typicals. Box-and-whisker plot showing min/max/first quartile/third determine whether additional behavioural variables modu- quartile/median divided by activation for ConRep and IncSw, and by lated activation in either the patient or typical group (Table 2). The following variables were included in theregression analyses: YGTSS, CY-BOCS, Conners' andBRIEF, as well as three additional covariates of no interest: Regions exhibiting correlations with tic MASC, estimated IQ and age.
severity within the patient group In addition to being correlated with YGTSS, activation in Because the patient group varied considerably in terms of bilateral SN/VTA on IncSw trials was also correlated across tic severity, we sought to identify brain regions in PFC, patients with scores on the BRIEF, a parental-report basal ganglia and/or thalamus for which level of activation measure of executive dysfunction, such that patients with on the challenging Incongruent Switch trials was either higher scores—indicative of worse executive function— positively or negatively correlated with tic severity. To this engaged this region more strongly (right: R2 change = 0.186, end, a multiple regression analysis was performed in SPM, P50.02; left: R2 change = 0.17, P50.05; Fig. 5). Although including YGTSS scores as the covariate of interest, and Age SN/VTA activation was positively correlated with both and Incongruent Switch accuracy as covariates of no YGTSS and BRIEF, scores on these two questionnaires were interest. No regions were observed for the negative not correlated with one another (R2 = 0.009). For typicals, correlation with YGTSS. In contrast, several regions by contrast, greater SN/VTA activation was associated with within our anatomical masks were engaged more strongly as a function of higher tic severity, including premotor P50.03). Thus, greater SN/VTA activation was associated cortex (bilateral BA 6), superior frontal gyrus (left BA 9, with greater tic severity and executive dysfunction in the right BA 8), inferior frontal gyrus pars triangularis (right patient group, but higher intellectual functioning in BA 45), middle frontal gyrus (left BA 9, right BA 8, 46), the typical group. In contrast to SN/VTA, nucleus accumbens

Cognitive control in Tourette syndrome Fig. 4 Regions correlated with tic severity (YGTSS) for patients during performance of IncSw trials (covariates of no interest: IncSwaccuracy and age). Clusters of activation within the lateral prefrontal mask are shown in magenta, and clusters within the subcortical maskare shown in cyan. Within the masks, clusters are shown if they met a statistical threshold of P50.005 with a 5 -voxel extent. Top row:sagittal slices for x-coordinates of 52, 14 and +18. Lower row: axial slices for z-coordinates of +64, +32, 0 and 8. Numbers correspond to approximate Brodmann areas. Th = thalamus; P = putamen; NA = nucleus accumbens.
Fig. 5 Bilateral ROIs in the SN/VTA are shown (spheres with 5 mm radius centered at 9, 8). Left graph: correlation plot for tic severity versus right SN/VTA parameter estimates for patients demonstrating that increased tic severity is positively correlated withactivity in this region. Right graph: correlation plot for executive dysfunction (BRIEF) versus right SN/VTA parameter estimates in thepatient group.
C. L. Baym et al.
activation was not correlated with other variables; higher conditions (ADHD, OCD or other anxiety disorders or levels of nucleus accumbens activation was associated only conduct disorders). The finding that children with more with increased tic severity.
severe TS have greater difficulty than their peers whenforced to switch between tasks or select between competingresponses underscores the point that TS is not simply a ROIs in basal ganglia and thalamus movement disorder, but a cognitive one as well.
Anatomical ROI analyses were conducted on regions of the Children with TS may compensate for underlying basal ganglia (caudate nucleus, putamen and globus difficulties with cognitive control. Consistent with this idea, we observed that children with TS, particularly those Additionally, because of our predictions relating to the with greatest tic severity, engaged lateral PFC more strongly STN, we conducted an ROI analysis for a region in the STN than typically developing children. This finding is consistent area (Fig. 6B), defined anatomically based on the studies of that of Marsh and colleagues (2007), who reported Aron and Poldrack (2006, 2007).
increased activity in bilateral PFC (left BA 9/46, 10 and For each ROI, we compared activation values between right BA 45, 10) with increasing tic severity across a large groups, separately for IncSw and ConRep trials (Fig. 6 for sample of children and adults. Several studies have provided plots of putamen, thalamus and STN area). These plots evidence of compensatory prefrontal activity in adults with revealed greater individual variability in the TS group than TS (Johannes et al., 2001; Hershey et al., 2004); the present the typicals, particularly on IncSw trials. We then study is suggestive of the early stages of a compensatory conducted multiple regression analyses for these regions, mechanism present already in childhood and in the absence to identify factors that could account for the large of tic medication. The possibility that this enhanced PFC individual differences in activation observed for IncSw recruitment is part of a compensatory mechanism, rather trials in the patients. These analyses included YGTSS, than being a consequence of heightened dopaminergic CY-BOCS, Conners', MASC, BRIEF, estimated IQ and age activity in TS, is bolstered by our finding of increased as independent variables, and average parameter estimates activation in the STN (an inhibitory nucleus linked to the for ConRep and IncSw as dependent variables. These ability to suppress inappropriate responses) as a function of tic severity, as described below.
because of the number of regressors included. However, While the current data suggest that children with TS only we sought to test the specific prediction that YGTSS was weakly engaged compensatory mechanisms, more dramatic correlated with activation in the ROIs, specifically or most compensation may be observed over development or with prominently on IncSw trials; the other regressors were included to ensure that our findings were not misattributed Our fMRI data revealed that tic severity scores were to tic severity.
strongly associated with elevated activation in a number of These ROI analyses revealed that tic severity among brain regions aside from lateral PFC. Notably, these patients was positively correlated with IncSw activation in correlations were observed primarily for the most cogni- bilateral putamen and thalamus (Table 2), as well as in the tively challenging trials (Incongruent Switch), strongly right STN area. This finding for the right STN ROI is suggesting that they are not the result of a confound, consistent with the results of the regression with YGTSS such as increased movement artifacts in the TS group.
reported above. Additionally, bilateral STN activation in the Rather, these results indicate that children with TS, in typical group was positively correlated with IQ (averaged particular those with higher tic severity, engage a number of across hemispheres; R2 = 0.43, P50.007). No significant regions more strongly than do typical children when asked correlations were observed for ConRep activation in these to perform a cognitively challenging task. Below we discuss regions, or for the caudate or globus pallidus.
four sets of brain regions for which a positive correlationwas observed with tic severity.
First, greater tic severity was correlated with stronger activation of the SN/VTA, consistent with the dopamine Overall, the children with TS in our study performed the hypothesis of TS (Nomura and Segawa, 2003; Yoon et al., cognitive control task as well as age-matched children.
2007; Fig. 7). To our knowledge, this is the first functional However, our sample included children with a range of tic brain imaging study to provide evidence of elevated severity. Our behavioural data revealed that greater tic activation in the SN/VTA, although this finding fits well severity among patients was associated with dispropor- with other results, such as the PET finding of higher [18F] tionate slowing of responses in both the switching and fluorodopa accumulation in the caudate nucleus of children response selection manipulations, but not the rule manip- with TS than typicals (Ernst et al., 1999).
ulation. Unlike prior studies examining performance of Second, tic severity was positively correlated with patients with TS on cognitive tasks, we have shown that this activation of regions involved in the direct pathway through cognitive deficit can be attributed to TS, rather than to the the basal ganglia (striatum, globus pallidus interna, symptoms of any of the most common comorbid thalamus and motor cortex). These findings provide Cognitive control in Tourette syndrome Fig. 6 (A) Subcortical ROIs in the right basal ganglia and thalamus. C = caudate; P = putamen; GP = globus pallidus; Th = thalamus. Box-and-whisker plots for right putamen and thalamus, showing min/max/first quartile/third quartile/median divided by activation for ConRep andIncSw, and by group. Plots for maximally active voxel in this ROI from IncSw-fix correlated with YGTSS (additional regressors: IncSwAccuracy and age). (B) ROI of STN, including box-and-whisker plots and correlation plot.
support for the idea that TS involves an imbalance between mesoaccumbens pathway (Saka and Graybiel, 2003; Albin the direct and indirect pathways, whereby excess dopamine and Mink, 2006). It should be noted that our results leads to elevated activity in the direct pathway through provide evidence for elevated mesolimbic activity as a stimulation of D1 receptors, resulting in unchecked move- function of tic severity specifically during performance of ments via disinhibition of the motor thalamus (Fig. 7). The the most cognitively challenging trials, and are therefore disinhibition of motor pathways in TS is nicely illustrated indicative of transient rather than sustained elevation of by the observation that tic severity was positively correlated mesolimbic activity in TS (this same observation holds for with regions in the corticospinal tract, including primary the other regions whose activation levels were correlated motor cortex.
with tic severity). Even though our study does not focus on Third, we showed that tic severity was positively tic production per se, our findings provide insight into the correlated with activity in the nucleus accumbens. Given mechanism by which spontaneous motor acts may be the role of the nucleus accumbens in the reinforcement of reinforced in TS.
behaviour, this finding suggests that tics, as habitual motor Finally, tic severity was positively correlated with activation of the STN, consistent with our prediction that C. L. Baym et al.
Supplementary materialSupplementary material is available at Brain online.
AcknowledgementsThis study was supported by a Merck Scholarship in theBiology of Developmental Disorders. Additional support forBlythe A. Corbett was provided by a mentored clinicalresearch award NIMH K08 MH072958. We thank J. Pinter,MD for clinical referrals, M. Souza, S. Donohue andR. Libove for assistance with data collection and analysis,D. Galik for participant recruitment, C. Wendelken forcontribution to the experimental design and analysis code, Fig. 7 Model of CSTC pathways affected in TS. Neural circuitry F. Sharp for collaboration on a larger Tourette syndrome showing dopaminergic, GABAergic and glutamatergic pathways project, and M. D'Esposito, B. Voytek, R.Cools and R. Ivry through the basal ganglia. The current study supports the for helpful feedback. Funding to pay the Open Access hypothesis that patients exhibit increased activity in the directpathway as well as increased compensatory activation of the publication charges for this article was provided by a John hyperdirect pathway (both pathways bolded). The indirect pathway Merck Scholarship in the Biology of Developmental is shown with dotted lines. Arrows indicate excitatory connections Disabilities in Children.
and circles indicate inhibitory connections.
children with TS would engage the hyperdirect pathway Abelson JF, Kwan KY, O'Roak BJ, Baek DY, Stillman AA, Morgan TM, from cortex to the STN, which would downregulate activity et al. Sequence variants in SLITRK1 are associated with Tourette'ssyndrome. Science 2005; 310: 317–2.
in the thalamus via the inhibition of the globus pallidus Abidin R. The parenting stress index. Lutz, FL: Psychological Assessment interna. Activation of this hyperdirect pathway would help Resources: Inc.; 1995.
to compensate for the disinhibition of the thalamus caused Achenbach TM, Rescorla, LA. Manual For the ASEBA school-age forms by an imbalance of activity between the direct and indirect and profiles: an integrated system of multi-informant assessment.
pathways—i.e. excessive direct pathway activation (Fig. 7).
Burlington, VT: University of Vermont, Research Center for Children,Youth, & Families; 2001.
It is known that dopamine stimulates D2 receptors, which Albin RL, Mink JW. Recent advances in Tourette syndrome research.
leads to inhibition of the STN via the indirect pathway, and Trends Neurosci 2006; 29: 175–82.
therefore to the facilitation of movement. Given that we APA. Diagnostic and statistical manual of mental disorder. 4th edn., observed elevated activity in dopamine-producing nuclei in Washington, DC: American Psychiatric Association; 1994.
TS, we might have predicted lower STN activation in Aron AR, Behrens TE, Smith S, Frank MJ, Poldrack RA. Triangulating a cognitive control network using diffusion-weighted magnetic resonance patients than typicals, as a result of reduced activity in the imaging (MRI) and functional MRI. J Neurosci 2007; 27: 3743–52.
indirect pathway. The fact that we observed greater STN Aron AR, Poldrack RA. Cortical and subcortical contributions to Stop activation in patients with more severe tics suggests that a signal response inhibition: role of the subthalamic nucleus. J Neurosci compensatory mechanism is at work.
2006; 26: 2424–33.
Aron AR, Shohamy D, Clark J, Myers C, Gluck MA, Poldrack RA. Human midbrain sensitivity to cognitive feedback and uncertainty during classification learning. J Neurophys 2004; 92: 1144–1152.
Baxter LR, Guze BA. eds. Neuroimaging. New York, NY: Marcel Dekker; Children with TS must exert a conscious effort to suppress their involuntary thoughts and actions; the need to monitor Bloch MH, Leckman JF, Zhu H, Peterson BS. Caudate volumes in and control themselves throughout the day can lead to poor childhood predict symptom severity in adults with Tourette syndrome.
Neurology 2005; 65: 1253–8.
performance in school. In future research, it will be critical Bohlhalter S, Goldfine A, Matteson S, Garraux G, Hanakawa T, to examine the relationships between TS, OCD and ADHD, Kansaku K, et al. Neural correlates of tic generation in Tourette all of which are frontostriatal disorders that affect cognitive syndrome: an event-related functional MRI study. Brain 2006; 129: control (e.g. Vaidya et al., 2005; Penades et al., 2006), and that can occur together or separately. Additionally, it is Bornstein RA, Baker GB. Neuropsychological performance and urinary phenylethylamine in Tourette's syndrome. J Neuropsychiatry Clin important to broaden the scope of research to include a Neurosci 1991; 3: 417–21.
larger population of children with tic disorders who do not Brett M, Anton JL, Valabregue R, Poline JB. Region of interest analysis meet criteria for a formal diagnosis of TS. It would be most using an SPM toolbox. In 8th International Conference on Functional helpful to be able to determine whether a child who is Mapping of the Human Brain, 2002, Sendai, Japan.
exhibiting tics is likely to develop TS. In this way, Burgund ED, Kang HC, Kelly JE, Buckner RL, Snyder AZ, Petersen SE, et al. The feasibility of a common stereotactic space for children and behavioural interventions targeting younger children could adults in fMRI studies of development. NeuroImage 2002; 17: 184–200.
help to improve the long-term prognosis of those who are Butler T, Stern E, Silbersweig D. Functional neuroimaging of Tourette likely to experience persistent tics.
syndrome: advances and future directions. Adv Neurol 2006; 99: 115–29.
Cognitive control in Tourette syndrome Chase TN, Foster NL, Fedio P, Brooks R, Mansi L, Kessler R, et al. Gilles Mink JW, Walkup J, Frey KA, Como P, Cath D, DeLong MR, et al. Patient de la tourette syndrome: studies with the fluorine-18-labeled fluor- selection and assessment recommendations for deep brain stimulation in odeoxyglucose positron emission tomographic method. Ann Neurol Tourette syndrome. Mov Disord 2006; 21: 1831–8.
1984; 15 (Suppl): S175.
Mueller SC, Jackson GM, Dhalla R, Datsopoulos S, Hollis CP. Enhanced Comings DE, Himes JA, Comings BG. An epidemiologic study of Tourette's cognitive control in young people with Tourette's syndrome. Curr Biol syndrome in a single school district. J Clin Psychiatry 1990; 51: 463–9.
2006; 16: 570–3.
Como PG. Neuropsychological function in Tourette syndrome. Adv Nomura Y, Segawa M. Neurology of Tourette's syndrome (TS) TS as a Neurol 2001; 85: 103–11.
developmental dopamine disorder: a hypothesis. Brain Dev 2003; 25 Conners KC. Conners' rating scales revised manual. Tonawanda, NY: (Suppl 1): S37–42.
Multi-Health Systems, Inc.; 2001.
Ozonoff S, Strayer DL, McMahon WM, Filloux F. Inhibitory deficits in Curtis D, Brett P, Dearlove AM, McQuillin A, Kalsi G, Robertson MM, Tourette syndrome: a function of comorbidity and symptom severity.
et al. Genome scan of Tourette syndrome in a single large pedigree J Child Psychol Psychiatry 1998; 39: 1109–18.
shows some support for linkage to regions of chromosomes 5, 10 and Pauls DL. An update on the genetics of Gilles de la Tourette syndrome.
13. Psychiatric Genetics 2004; 14: 83–7.
J Psychosom Res 2003; 55: 7–12.
Dale AM. Optimal experimental design for event-related fMRI. Hum Brain Pauls DL, Leckman JF. The inheritance of Gilles de la Tourette's syndrome Mapp 1999; 8: 109–14.
and associated behaviors. Evidence for autosomal dominant transmis- DeVito TJ, Drost DJ, Pavlosky W, Neufeld RW, Rajakumar N, McKinlay BD, sion. N Engl J Med 1986; 315: 993–7.
et al. Brain magnetic resonance spectroscopy in Tourette's disorder. J Am Penades R, Catalan R, Rubia K, Andres S, Salamero M, Gasto C. Impaired Acad Child Adolesc Psychiatry 2005; 44: 1301–8.
response inhibition in obsessive compulsive disorder. Eur Psychiatry Ernst M, Zametkin AJ, Jons PH, Matochik JA, Pascualvaca D, Cohen RM.
2007; 22: 404–10.
High presynaptic dopaminergic activity in children with Tourette's Peterson BS, Pine DS, Cohen P, Brook JS. Prospective, longitudinal study disorder. J Am Acad Child Adolesc Psychiatry 1999; 38: 86–94.
of tic, obsessive-compulsive, and attention-deficit/hyperactivity disorders Fredericksen KA, Cutting LE, Kates WR, Mostofsky SH, Singer HS, in an epidemiological sample. J Am Acad Child Adolesc Psychiatry 2001; Cooper KL, et al. Disproportionate increases of white matter in right 40: 685–95.
frontal lobe in Tourette syndrome. Neurology 2002; 58: 85–9.
Peterson BS, Thomas P, Kane MJ, Scahill L, Zhang H, Bronen R, et al.
Frey KA, Albin RL. Neuroimaging of Tourette syndrome. J Child Neurol Basal Ganglia volumes in patients with Gilles de la Tourette syndrome.
2006; 21: 672–7.
Arch Gen Psychiatry 2003; 60: 415–24.
Garraux G, Goldfine A, Bohlhalter S, Lerner A, Hanakawa T, Hallett M.
Pringsheim T, Davenport WJ, Lang A. Tics. Curr Opp Neurol 2003; 16: Increased midbrain gray matter in Tourette's syndrome. Ann Neurol 2006; 59: 381–5.
Rutter M, Bailey A, Berument SK, Lord C, Pickles A. Social communica- Gerard E, Peterson BS. Developmental processes and brain imaging studies tion questionnaire. Los Angeles, CA: Western Psychological Services; in Tourette syndrome. J Psychosom Res 2003; 55: 13–22.
Gilbert DL, Christian BT, Gelfand MJ, Shi B, Mantil J, Sallee FR. Altered Saka E, Graybiel AM. Pathophysiology of Tourette's syndrome: striatal mesolimbocortical and thalamic dopamine in Tourette syndrome.
pathways revisited. Brain Dev 2003; 25 (Suppl 1): S15–9.
Neurology 2006; 67: 1695–7.
Scahill L, Riddle MA, McSwiggin-Hardin M, Ort SI, King RA, Gioia GA, Isquith PK, Guy SC, Kenworthy L. The behavior rating Goodman WK, et al. Children's Yale-Brown Obsessive Compulsive inventory of executive function. Lutz, FL: Psychological Assessments Scale: reliability and validity. J Am Acad Child Adolesc Psychiatry 1997; Resources, Inc.; 2000.
36: 844–52.
Hershey T, Black KJ, Hartlein J, Braver TS, Barch DM, Carl JL, et al.
Serrien DJ, Orth M, Evans AH, Lees AJ, Brown P. Motor inhibition in Dopaminergic modulation of response inhibition: an fMRI study. Brain patients with Gilles de la Tourette syndrome: functional activation Res Cogn Brain Res 2004; 20: 438–48.
patterns as revealed by EEG coherence. Brain 2005; 128: 116–25.
Johannes S, Wieringa BM, Mantey M, Nager W, Rada D, Muller-Vahl KR, Shaffer D, Fisher P, Lucas CP, Dulcan MK, Schwab-Stone ME. NIMH et al. Altered inhibition of motor responses in Tourette Syndrome and Diagnostic Interview Schedule for Children Version IV (NIMH DISC- Obsessive-Compulsive Disorder. Acta Neurol Scand 2001; 104: 36–43.
IV): description, differences from previous versions, and reliability of Kang HC, Burgund ED, Lugar HM, Petersen SE, Schlaggar BL.
some common diagnoses. J Am Acad Child Adolesc Psychiatry 2000; 39: Comparison of functional activation foci in children and adults using a common stereotactic space. NeuroImage 2003; 19: 16–28.
Stern E, Silbersweig DA, Chee KY, Holmes A, Robertson MM, Trimble M, Leckman JF. ed. Tourette syndrome. Washington, DC: American et al. A functional neuroanatomy of tics in Tourette syndrome. Arch Psychiatric Press; 1993.
Gen Psychiatry 2000; 57: 741–8.
Leckman JF, Riddle MA, Hardin MT, Ort SI, Swartz KL, Stevenson J, et al.
Thesen S, Heid O, Mueller E, Schad LR. Prospective acquisition correction The Yale Global Tic Severity Scale: initial testing of a clinician-rated for head motion with image-based tracking for real-time fMRI. Magn scale of tic severity. J Am Acad Child Adolesc Psychiatry 1998; 28: Reson Med 2000; 44: 457–65.
Vaidya CJ, Bunge SA, Dudukovic NM, Zalecki CA, Elliott GR, Gabrieli JD.
Lee JS, Yoo SS, Cho SY, Ock SM, Lim MK, Panych LP. Abnormal thalamic Altered neural substrates of cognitive control in childhood ADHD: volume in treatment-naive boys with Tourette syndrome. Acta Psychiatr evidence from functional magnetic resonance imaging. Am J Psychiatry Scand 2006; 113: 64–7.
2005; 162: 1605–13.
Ludolph AG, Juengling FD, Libal G, Ludolph AC, Fegert JM, Kassubek J.
Verte S, Geurts HM, Roeyers H, Oosterlaan J, Sergeant JA. Executive Grey-matter abnormalities in boys with Tourette syndrome: Magnetic functioning in children with autism and Tourette syndrome. Dev resonance imaging study using optimised voxel-based morphometry.
Psychopathol 2005; 17: 415–45.
Br J Psychiatry 2006; 188: 484–5.
Watkins LH, Sahakian BJ, Robertson MM, Veale DM, Rogers RD, March J. The multidimensional anxiety scale for children. North Pickard KM, et al. Executive function in Tourette's syndrome and Tonawanda, NY: Multi-Health Systems, Inc.; 1999.
obsessive-compulsive disorder. Psychol Med 2005; 35: 571–82.
Marsh R, Zhu H, Wang Z, Skudlarski P, Peterson BS. A Developmental Wechsler D. Wechsler Abbreviated Scale of Intelligence. San Antonio, TX: fMRI Study of Self-Regulatory Control in Tourette's Syndrome. Am J Psychological Corporation; 1999.
Psychiatry 2007; 164: 955–66.
Yoon DY, Gause CD, Leckman JF, Singer HS. Frontal dopaminergic Mink JW. Basal ganglia dysfunction in Tourette's syndrome: a new abnormality in Tourette syndrome: a postmortem analysis. J Neurol Sci hypothesis. Pediatr Neurol 2001; 25: 190–8.
2007; 255: 50–6.

Source: http://bungelab.berkeley.edu/wp-content/uploads/2012/01/Baym_Brain20071.pdf