Abstract

Gilles de la Tourette syndrome is a neuropsychiatric disorder characterized by an impaired ability to inhibit unwanted behaviour. Although the presence of chronic motor and vocal tics defines Tourette’s syndrome, other distinctive behavioural features like echo- and coprophenomena, and non-obscene socially inappropriate behaviour are also core features. We investigated neuronal activation during stimulus-driven execution and inhibition of prepared movements in Tourette’s syndrome. To this end, we performed event-related functional magnetic resonance imaging and structural diffusion tensor imaging in 15 moderately affected uncomplicated patients with ‘pure’ Tourette’s syndrome and 15 healthy control participants matched for age and gender. Subjects underwent functional magnetic resonance imaging during a Go/NoGo reaction time task. They had to withhold a prepared finger movement for a variable time until a stimulus instructed them to either execute (Go) or inhibit it (NoGo). Tics were monitored throughout the experiments, combining surface electromyogram, video recording, and clinical assessment in the scanner. Patients with Tourette’s syndrome had longer reaction times than healthy controls in Go trials and made more errors in total. Their functional brain activation was decreased in left primary motor cortex and secondary motor areas during movement execution (Go trials) but not during response inhibition (NoGo trials) compared with healthy control subjects. Volume of interest analysis demonstrated less task-related activation in patients with Tourette’s syndrome in primary and secondary motor cortex bilaterally, but not in the basal ganglia and cortical non-motor areas. They showed reduced co-activation between the left primary sensory-motor hand area and a network of contralateral sensory-motor areas and ipsilateral cerebellar regions. There were no between-group differences in structural connectivity of the left primary sensory-motor cortex as measured by diffusion tensor imaging-based probabilistic tractography. Our results link reduced sensory-motor cortical activation during movement execution to a decreased co-activation between the sensory-motor cortex and other brain areas involved in motor processing. These functional changes in patients with Tourette’s syndrome might result from adaptive reorganization in fronto-parietal brain networks engaged in motor and behavioural control, possibly triggered by abnormal processing and presumably overactivity in cortico-striato-cortical circuits. This might enable patients with Tourette’s syndrome to better suppress unwanted movements but comes at a price of behavioural deficits in other domains.

Introduction

The ability to inhibit unwanted behaviour depending on the context is crucial for human social life. This includes both the inhibition of behaviour driven by internal motivation, responses to external stimuli, and, depending on the context, the inhibition of planned and already prepared actions. An impairment of this ability can be found in a number of neuropsychiatric disorders including Gilles de la Tourette syndrome. Tourette’s syndrome is a neuropsychiatric disorder with childhood-onset that is characterized by chronic motor and vocal tics (American Psychiatric Association, 2000; Leckman et al., 2001). Tics are the clinical hallmark of Tourette’s syndrome, but echo- and coprophenomena, and non-obscene socially inappropriate behaviour are also often present (Robertson, 2000). These phenomena represent a failure to inhibit unwanted or inappropriate behaviours, and can be driven by inner urges or triggered by external stimuli. For instance, echophenomena are immediate and unwanted responses to observed movements (echopraxia) or speech (e.g. echolalia) (Finis et al., 2012; Ganos et al., 2012). Tics can also be set off by external stimuli, e.g. sounds or eye-catching observations (Leckman et al., 2001). Most patients with Tourette’s syndrome can suppress tics and other unwanted behaviour to some extent and for certain periods. Such inhibition though is incomplete. Sooner or later suppressed behaviour has to be released or diverted to related actions considered less inappropriate.

Complementing clinical observation there is experimental evidence of impaired control of behavioural responses in different tasks such as grip force control (Nowak et al., 2005), choice reaction time experiments (Sheppard et al., 2000), Go/NoGo tasks (Eichele et al., 2010), or complex stimulus response tasks in patients with Tourette’s syndrome (Georgiou et al., 1995). On the other hand, other studies have failed to demonstrate distinct behavioural differences between patients with Tourette’s syndrome and control participants in simple tasks of behavioural inhibition or motor control (Channon et al., 2006), manual response switching (Jung et al., 2013), or simple response inhibition (Roessner et al., 2008). Even enhanced inhibitory performance, e.g. in an oculomotor switching task (Mueller et al., 2006) or enhanced control of motor output (Jackson et al., 2011) has been reported in Tourette’s syndrome and interpreted as being compensatory. Heterogeneous results in behavioural studies have also been attributed to the clinical heterogeneity of Tourette’s syndrome study populations in terms of clinical characteristics and associated co-morbidities (Ozonoff et al., 1998; Robertson, 2000; Jung et al., 2013).

Although the aetiology of Tourette’s syndrome is still unknown there is consensus about a number of pathophysiological assumptions based on clinical observations, experimental data, and post-mortem studies, which offer a pathophysiological framework to explain the observed behavioural characteristics. Abnormal neuronal processing in cortico-striato-thalamo-cortical circuits and altered dopaminergic neurotransmission is assumed to play a role in the pathophysiology of Tourette’s syndrome (Stern et al., 2000; Peterson, 2001; Singer and Minzer, 2003; Singer, 2005; Mink, 2006; Buse et al., 2013). Inhibition of actions is assumed to rely on fronto-parietal control networks (Swick et al., 2011), thus the impairment of inhibitory capabilities in Tourette’s syndrome may be explained by alteration in these networks (Werner et al., 2011; Jung et al., 2013). Structural brain imaging results demonstrating alterations in basal ganglia and frontal cortices in patients with Tourette’s syndrome seem to support this assumption (Peterson, 2001; Peterson et al., 2001; Sowell et al., 2008). Functional brain imaging studies provide further support for altered processing in fronto-parietal control networks in these patients (Church et al., 2009a; Werner et al., 2011; Worbe et al., 2012; Jung et al., 2013).

In this study, we aimed at exploring brain activation in Tourette’s syndrome during the execution and inhibition of prepared movements and relating possible alterations of neuronal activation both to behavioural findings and brain structure. To this end, we performed a functional MRI experiment along with structural brain imaging using diffusion tensor imaging in a comprehensively assessed group of patients with Tourette’s syndrome. Previous studies have identified co-morbid conditions such as attention-deficit hyperactivity disorder or obsessive compulsive disorder as potential determinants of executive dysfunction in patients with Tourette’s syndrome (Georgiou et al., 1995), an interpretation corroborated by the fact that several studies of patients with uncomplicated Tourette’s syndrome, excluding individuals with co-morbid attention deficit hyperactivity disorder and obsessive compulsive disorder, have failed to demonstrate marked behavioural differences between groups (Channon et al., 2006; Roessner et al., 2008; Werner et al., 2011; Jung et al., 2013). To rule out an influence of these co-morbid conditions we thus focused on patients with uncomplicated Tourette’s syndrome.

In Tourette’s syndrome behavioural deficits in tasks of motor and executive control were identified with increasing task difficulty (Werner et al., 2011). Although no behavioural abnormalities were observed in simple motor or stimulus-response tasks (Roessner et al., 2008; Jung et al., 2013), patients with Tourette’s syndrome demonstrated behavioural deficits during complex motor control (Nowak et al., 2005), or demanding choice reaction time or Go/NoGo experiments (Sheppard et al., 2000; Eichele et al., 2010). We opted for a complex Go/NoGo reaction time paradigm likely to involve activation in widespread cortical networks.

We hypothesized that patients with Tourette’s syndrome would show behavioural costs reflected by longer reaction times and higher error rates, particularly errors of commission, i.e. failures to withhold prepared movements in the NoGo condition. We expected increased task-related blood oxygen level-dependent (BOLD) signal in fronto-parietal control networks during the NoGo condition (Baym et al., 2008). We also hypothesized that more severely affected patients with Tourette’s syndrome would demonstrate more pronounced alterations in brain activation patterns.

Materials and methods

Participants and clinical assessment

We studied 15 patients with Tourette’s syndrome without psychiatric co-morbidities [13 males, mean age = 34 years, standard deviation (SD) ± 9]. Patients were recruited from the specialized Tourette’s syndrome outpatient clinics in the Department of Neurology, University Medical Centre Eppendorf, Hamburg and the Clinic of Psychiatry, Social Psychiatry and Psychotherapy, Hannover Medical School, Hannover. Each patient was clinically assessed by a neurologist or psychiatrist experienced in diagnosing and treating Tourette’s syndrome. Lifetime clinical information was systematically collected using standardized clinical assessment and a structured interview. The diagnosis of Tourette’s syndrome was made according to Diagnostic and Statistical Manual of Mental Disorders (4th edition, text revision, DSM-IV-TR) criteria (American Psychiatric Association, 2000). Obsessive compulsive disorder and mood disorders were tested for using the appropriate modules of the German version of the structured clinical interview for DSM-IV Axis I disorders (SCID-I) (Wittchen et al., 1997). Attention deficit hyperactivity disorder was assessed according to DSM-IV-TR. Patients fulfilling criteria for obsessive compulsive disorder, attention deficit hyperactivity disorder or other co-morbidities, particularly depression, were excluded from the study. Patients were either medication-naïve or had stopped neuroleptic medication before inclusion into the study. Patients still on medication were instructed to stop medication at least 3 weeks before participating in the study. The Diagnostic Confidence Index was used to assess the lifetime likelihood of a diagnosis of Tourette’s syndrome (Robertson et al., 1999). Tic severity was measured using the Yale Global Tic Severity Scale (YGTSS) (Leckman et al., 1989). Standardized video recording was performed and data were scored using the Modified Rush Videotape Rating Scale (Goetz et al., 1999), which provides a total tic impairment score ranging from 0 to 20. Two patients refused to be filmed.

As a control group, 15 healthy participants, matched for age and gender, were studied (13 males, mean age 35 years, SD ± 9). All control participants had a normal neurological and psychiatric examination and no history of neurological or psychiatric disorders. Except for one patient and one control subject who were ambidextrous, all participants were right-handed according to the Edinburgh Handedness Inventory (Oldfield, 1971). Participants gave informed written consent to participate in the study in accordance with the Declaration of Helsinki (1964). The local Ethics Committee had approved the study protocol (No. 2514). Patients and controls represent the same sample as in a previous study (Thomalla et al., 2009).

Experimental design

Subjects were studied by functional MRI during a Go-NoGo reaction time paradigm. A visual pre-cue ‘S1’ (meaningless geometric shape: square or cross) instructed participants to prepare an extension-flexion movement of either the index or middle finger of their right hand. A subsequent cue ‘S2’ (coloured circle: yellow or blue) indicated whether they should perform (Go) or withhold (NoGo) the prepared response. S1 and S2 were presented for 500 ms each. Stimulus onset asynchronies of S1 and S2 in the same trial, and of S2 and S1 in consecutive trials, were both jittered between 2000 and 6000 ms (Fig. 1).

Figure 1

Go/NoGo reaction time paradigm used for functional MRI. Top left: the shape of stimulus (S1) instructs preparing a lifting and lowering of either the index or middle finger of the right hand. Top middle/right: the colour of stimulus S2 prompts movement execution (Go) or inhibition (NoGo). Bottom: schematic display of trial timing.

Figure 1

Go/NoGo reaction time paradigm used for functional MRI. Top left: the shape of stimulus (S1) instructs preparing a lifting and lowering of either the index or middle finger of the right hand. Top middle/right: the colour of stimulus S2 prompts movement execution (Go) or inhibition (NoGo). Bottom: schematic display of trial timing.

We used an event-related functional MRI design with stimuli presented in a pseudo-randomized order. Sixty trials were presented in each of two runs. To create a high expectancy to act we presented Go and NoGo conditions with a ratio of 2:1 (40:20 trials). Participants were instructed by S1 to prepare a movement of the index finger or the middle finger in 50% of the Go trials each. This resulted in a 2 × 2 × 2 design including between-subjects factor ‘group’ (patients with Tourette’s syndrome, controls), and within-subject factors ‘S1’ (index, middle) and ‘S2’ (Go, NoGo). Assignments of geometric shape of S1 (square/cross) to finger movements (index/middle finger), and colour of S2 (yellow/blue) to Go/NoGo conditions, respectively, were both counterbalanced between participants. Before the experiment, participants performed 20 practice trials outside the scanner.

Presentation software (Neurobehavioral Systems, Inc) was used for stimulus presentation and recording of motor responses. Visual stimuli were back-projected by an LCD beamer onto a screen in the bore of the scanner and viewed through a mirror fixed to the head coil. Motor responses were recorded with an MRI-compatible custom-made opto-electronic device. The device consisted of two light barriers mounted on a panel. The endings of two optical fibres were positioned laterally to the tip of the index and middle finger with the fingers resting on a board in a slightly flexed position. The light barrier was opened as soon as participants lifted the finger. Before the experimental sessions, participants performed 20 practice trials outside the scanner.

Tic monitoring

Patients were instructed to focus on the experimental task and to let go, i.e. not to suppress their tics. To identify tics as potential confounders during the experiment, we performed a multimodal tic monitoring time-locked to the experiment. First, during functional MRI measurements, a Tourette’s syndrome-experienced neurologist (M.O. or A.M.) familiar with the individual patient’s tic repertoire stood next to the magnetic resonance scanner and indicated every observed motor tic (except for head and facial tics, which were hidden by the coil) by pressing a mouse button. Anatomical distribution of tics (face, arm, leg, multiple body areas) was coded on the basis of the number of mouse button presses. Second, the patients’ faces were continuously videotaped by an MRI-compatible miniature camera fixed to the head coil. Third, surface EMG was recorded using customized magnetic resonance-compatible electrodes and a magnetic resonance-compatible amplifier (BrainAmp MR, Brain Products GmbH) from the following muscles bilaterally: M. frontalis, M. orbicularis oculi, M. zygomaticus, M. splenius, M. trapezius, M. deltoideus, M. extensor digitorum and M. flexor digitorum. EMG signals were sampled at 5 kHz. Offline, EMG data were post-processed using Brain Vision Analyzer software (Brain Products GmbH). Scanner artefacts were corrected and EMG signal was filtered between 10–120 Hz to further remove artefacts. To identify tic onsets in each patient, EMG data were analysed and assessed by G.T. along with the video and expert tic ratings both recorded during scanning. Tic-related EMG activity was thus identified, and onset of EMG activity used to define tic onset for all tics captured by EMG monitoring (Fig. 2). Only the onset of tics involving body parts not captured by surface EMG (i.e. trunk and legs) was defined on the basis of expert tic ratings. These were delayed by a mean of 1.1 s ( ± 0.5) with respect to EMG onset. For tics not covered by EMG monitoring we corrected for this delay by subtracting 1.1 s from the recorded onset in order to minimize the potential error resulting from delayed judgement of tic onset by expert ratings in those tics.

Figure 2

Example of multimodal tic monitoring as applied in our study. (A) EMG recordings from selected muscles (given here as bipolar recording between muscles of the left and right side) during functional MRI scanning after post-processing. Arrows indicate tic onsets (F = facial tic, M = tic involving multiple body areas) that were also identified by the expert in the scanner and/or on videotape. (B) Example of the view of the patients face as recorded by the miniature camera fixed to the head coil (snapshots during tics recorded by EMG in A as indicated by numbers 1, 2 and 4).

Figure 2

Example of multimodal tic monitoring as applied in our study. (A) EMG recordings from selected muscles (given here as bipolar recording between muscles of the left and right side) during functional MRI scanning after post-processing. Arrows indicate tic onsets (F = facial tic, M = tic involving multiple body areas) that were also identified by the expert in the scanner and/or on videotape. (B) Example of the view of the patients face as recorded by the miniature camera fixed to the head coil (snapshots during tics recorded by EMG in A as indicated by numbers 1, 2 and 4).

Magnetic resonance imaging protocol

MRI was conducted at 3 T (Magnetom Trio, Siemens) with a standard head coil (Bruker). A T1-weighted FLASH 3D sequence was used for structural MRI of the whole brain (repetition time = 15 ms, echo time = 4.92 ms, flip angle 25°, 192 slices, slice thickness = 1 mm, gap: 20%, matrix: 256 × 256 mm, field of view 256 × 256 mm). Single-shot gradient-echo echo-planar imaging covering the whole brain (repetition time 2190 ms, echo time 24 ms, flip angle 70°, 35 slices, slice thickness = 3 mm, gap: 33%, matrix 64 × 64 mm, field of view 210 × 210 mm) was used to measure task-related changes in BOLD signal indicative of regional neuronal activity. The functional MRI experiment was divided into two runs of ∼8 min 20 s each where 228 brain volumes were acquired in each.

Diffusion tensor imaging was performed using an echo planar imaging sequence (echo time/repetition time = 105/18.500 ms, 128 × 128 matrix, field of view 256 × 192 mm, 60 axial slices, 2 mm slice thickness without interslice gap, resulting in an isotropic voxel size of 2 × 2 × 2 mm3) with diffusion sensitization along 24 different directions with a b-value of 1000 s/mm2. The diffusion tensor imaging sequence included a fluid attenuated inversion recovery technique (inversion time = 2400 ms) for CSF suppression. Scanning was repeated twice, resulting in a total scanning time of 16 min 59 s. During the entire MRI acquisition the position of the head was stabilized with foam pads to minimize head movements. Functional MRI was conducted first in each participant, followed by diffusion tensor imaging and structural MRI.

Data analysis

Behavioural data

Behavioural data were analysed using R (R Core Team, 2012). In each subject, we calculated mean reaction times excluding trials with errors and outliers defined by reaction time <100 ms and exceeding the individual mean reaction time by more than 2.5 SD. Average error rates were also determined, in total and separately for different types of errors in Go trials (i.e. responses with the wrong finger, omissions, too early responses before S2), and NoGo trials (errors of commission). For both reaction times and error rates, responses made with the index or middle finger were pooled. Total error rate was based on a weighted sum of errors from Go and NoGo trials according to their proportion (2:1). Because part of the behavioural data showed a high dispersion, we performed non-parametrical Wilcoxon rank sum-tests for independent samples to detect group differences in data distribution. Significance level was set to P < 0.05.

Functional magnetic resonance imaging data

Functional MRI data were analysed using SPM5 (Wellcome Department of Cognitive Neurology, London, UK) in Matlab (Mathworks). The first four scans of each session were discarded from data analysis because of the non-equilibrium state of magnetization. Preprocessing of image data included slice timing and spatial realignment of individual scan volumes by rigid body transformation to the first image of the time series. Additionally, the time series was undistorted using the unwarping procedure implemented in SPM5 to minimize unwanted variance introduced by movement × susceptibility interactions. Unwarped images were spatially normalized to a standard echoplanar imaging template and spatially smoothed using an isotropic Gaussian kernel of 9 mm full-width at half-maximum. To remove low-frequency drifts the time series data at each voxel was processed using a high-pass filter with a cut-off of 128 s.

For within-subject event-related analysis (first level), a general linear model was specified using multiple regression analysis (Friston et al., 1995). Both sessions were entered separately into one model each. Event-related changes in BOLD signal were estimated at each voxel by modelling the onsets of each stimulus as delta functions convolved with a haemodynamic response function. Onsets of S1, S2-Go and S2-NoGo entered the design matrices as separate regressors (events of interest). Within each condition (Go/NoGo), S1s instructing a movement of the index or middle finger were pooled. Tic onset identified in patients by the monitoring procedure was modelled additionally as a confounding event. Error events were also entered into the model as events of no interest. The statistical significance of the model was tested in each subject voxel-wise using F- and t-statistics.

For each participant, we generated statistical parametric maps (SPMs) of t-statistics for each voxel using linear contrasts of the parameter estimates for the regressors of interest. We tested for BOLD-signal changes related to events S1, S2-Go and S2-NoGo. These individual contrast images were then entered into second level analyses using one- and two-sample t-tests to assess within- and between-group effects, respectively. To test for a relation between behavioural parameters and activation patterns we correlated BOLD-signal changes during S2-Go with reaction time as well as with core clinical markers of disease severity (Modified Rush Video Scale, YGTSS, tic count per minute). In our analysis of task-related brain activation across groups we corrected for multiple comparisons within the whole brain. In our group comparisons of brain activation related to NoGo and Go, we took into account multiple comparisons by using small volume correction within predefined anatomical volumes of interest based on our hypotheses. For the group comparison of brain activation related to NoGo we identified a network of brain areas that were reported to be active during NoGo experiments in a recent quantitative meta-analysis of functional MRI studies of NoGo experiments (Swick et al., 2011). We used the significant clusters identified to be active in NoGo experiments by the Activation Likelihood Estimate method in this paper to correct for multiple comparisons. This volume of interest included the following brain regions: right insula, bilateral middle frontal gyrus, right medial frontal gryrus, bilateral inferior frontal gyrus, bilateral inferior parietal lobule, left claustrum, left putamen, left superior frontal gyrus, right precuneus, right superior temporal gyrus, left supramarginal gyrus, right inferior occipital gyrus, left fusiform gyrus, right cingulate gyrus, and comprised a total of 3.146 voxels. For the group comparison of brain activation during the Go condition we hypothesized that patients with Tourette’s syndrome would show altered brain activation in motor brain areas based on findings from previous functional MRI studies in patients with Tourette’s syndrome (Stern et al., 2000; Werner et al., 2011; Worbe et al., 2012). Following the description of the human motor area template in a meta-analysis reporting maxima of the main cortical motor regions from numerous functional MRI studies (Mayka et al., 2006), we created a volume of interest mask defined by bilateral primary motor cortex (M1), dorsal premotor cortex and supplementary motor area (SMA) proper. This volume of interest comprised a total of 2.539 voxels. All corrections for multiple comparisons were made using the family-wise error (FWE) method, and P-values <0.05 were considered significant. For parameter maps with significant differences between groups we additionally report voxels with between-group differences of P < 0.001 (uncorrected) as trends. For exploratory purposes we also tested for tic-related BOLD signal changes. Based on the exploratory character of this analysis of tic-related brain activation as a potential confounder of task related BOLD-signal changes SPMs for this analysis were thresholded at P < 0.001 uncorrected.

Based on our hypothesis of specific activation changes in areas involved in motor control in patients with Tourette’s syndrome (Stern et al., 2000; Werner et al., 2011; Worbe et al., 2012), we further explored brain activation during experimental conditions with significant brain activation differences between groups by the analysis of event-related BOLD signal changes in predefined anatomical volumes of interest. To test the hypothesis of altered brain activation in cortical sensory-motor areas we compared event-related signal changes in volumes of interest from brain regions involved in sensory-motor processing comprising cortical (M1, dorsal premotor cortex, SMA-proper, primary somatosensory cortex) and subcortical areas (caudate nucleus, putamen, globus pallidus internus, globus pallidus externus). In addition, we included a large prefrontal non-motor volume of interest for comparison. For each volume of interest event-related per cent signal change during S2-Go was extracted using the rfxplot toolbox in SPM5 (Glascher, 2009). Volumes of interest were grouped into three areas: cortical sensory-motor areas (M1, primary somatosensory cortex, dorsal premotor cortex, SMA-proper), cortical non-motor areas (prefrontal cortex), and basal ganglia (caudate nucleus, putamen, globus pallidus externus, globus pallidus internus) and values were entered into a 3 × 2 × 3 ANOVA in SPSS (SPSS Inc.) with within-subject factors ‘Side’ (right, left) and ‘Area’ (cortical sensory-motor areas, cortical non-motor areas, basal ganglia) and between subjects factor ‘Group’ (Tourette’s syndrome, Control). Greenhouse-Geisser correction was applied to correct for non-sphericity of data. Post hoc group comparison of event-related per cent signal change was performed using two-sample t-test. To control for the FWE rate in the context of multiple testing, Bonferroni correction was applied. Given a total of 18 tests this results in a threshold of P < 0.0028 for each individual test to comply with a type I error rate of α0.05.

We performed an additional analysis of context dependent co-activation between motor brain areas activated during the Go/NoGo experiment. Therefore, we correlated the whole BOLD time series in a seed volume of interest with all other areas, as suggested previously (Heinz et al., 2005). The seed volume of interest was defined by a spherical binary mask of 10 mm diameter around the global maximum in the one-sample t-test on the main effect of S2-Go across both groups from the previous second level analysis (left primary sensory-motor cortex, MNI coordinates −42 −3 54). For each subject, the BOLD time series were extracted from all voxels within the seed volume of interest using the rfxplot toolbox (Glascher, 2009) implemented in SPM5. Time series were filtered as in the previous SPM analysis, adjusted for block and nuisance variables, and their first eigenvariate was calculated. Within-subject linear regression analyses were performed, using the time series in the left primary sensory-motor cortex as a regressor that was not convolved with the haemodynamic response function and whole brain FWE correction was applied.

Processing of diffusion tensor imaging data was performed using the Diffusion Toolbox from the FMRIB software library (Smith et al., 2004). Images were corrected for eddy currents and motion by using affine registration to the non-diffusion volumes. Probability distributions of fibre orientations in each voxel were calculated (Behrens et al., 2003) and probabilistic tractography was performed from a seed region defined matching the seed region in left primary sensory-motor cortex that was used to study co-activation. A total of 5000 individual streamlines were drawn for each seed voxel with a step length of 0.5 mm and a maximum step number of 2000, curvature threshold 0.2. Probabilistic tracking was performed on individual data sets in original diffusion space. Individual seed regions were created by back-transformation of the left primary sensory-motor cortex volume of interest from MNI space into individual anatomical space using the parameters derived from spatial normalization. Individual probabilistic map were thresholded at 10% of streamlines to exclude noise and transferred to MNI space. For display, spatially normalized probability map images were binarized and added to create group maps of structural connectivity to the left primary sensory-motor cortex. Statistical comparison of the individual probabilistic maps between groups was performed using permutation-based non-parametric two-tailed t-test with exhaustive testing and FWE correction for multiple comparisons (http://www.fmrib.ox.ac.uk/fsl/randomise/).

Results

Clinical characteristics of patients with Tourette’s syndrome

Patient characteristics are given in Table 1. In summary, the population studied comprised moderately affected patients with Tourette’s syndrome reflected by a mean YGTSS score of 42 ± 16 (SD) and a mean YGTSS tic score of 21 ± 7. Mean disease duration was 26.5 ± 8.5 years, mean diagnostic confidence index was 61 ± 15. Five patients were medication-naïve, all others had stopped medical treatment at least 3 weeks up to years before participation in the study. Only four patients had been on neuroleptic medication during the last year. During the functional MRI experiment we observed an average number of 6.1 tics per minute ( ± 4.1, range: 0.3–13.6). There was a slight increase in the mean number of tics per minute from 5.5 ± 4.0 in the first session to 6.7 ± 5.4 in the second session (P < 0.01). A significantly higher percentage of tics occurred between the trials (i.e. between the offset of S2 and the onset of the next S1) than during the trials (i.e. between the onset of S1 and the offset of S2) [Wilcoxon signed rank test: V = 88, P < 0.05, median (range): between trials 54.17% (40–71.88%), within trials 45.83% (28.13–60%)]. Of note, the mean number of tics per minute observed during the experiment in the scanner was markedly lower than the mean number of tics per minute during video recording outside the scanner (41.8 ± 20.5).

Table 1

Patients characteristics

Subject Age Sex Age at onset DCI (0–100) YGTSS (0–100) YGTSS tics (0–50) Motor tics
 
Vocal tics
 
Tics / min during fMRI Tics / min outside scannera MRVS (total) Medication history Time since last medical treatment 
       Simple Complex Simple Complex      
P01 33 12 63 46 26 3.9 34 11 Tiapride, sulpiride >1 year 
P02 23 61 57 27 − 2.8 44 Tiapride, sulpiride, risperidone, aripiprazole, olanzapine, pimozide 3 weeks 
P03 29 68 30 20 0.4  Missingb − Naïve 
P04 27 68 31 21 − 3.5  Missingb Tiapride, risperidone, quetiapine 3 months 
P05 22 10 47 26 16 − − 4.3 27 Tiapride 6 weeks 
P06 39 12 100 77 37 12.9 55 14 Tiapride, sulpiride 6 months 
P07 31 12 37 18 − − − 0.6 − Naïve 
P08 54 13 57 49 19 − 10.1 68 15 Haloperidole >10 years 
P09 28 50 44 24 − − 7.1 36 Tiapride >1 year 
P10 42 76 42 22 − 13.6 63 13 − Naïve 
P11 45 11 64 35 25 − 4.7 63 10 − Naïve 
P12 29 54 36 16 − 2.6 17 − Naïve 
P13 34 52 56 16 − − 9.2 48 Not known >10 years 
P14 38 11 45 22 12 − − 7.0 15 Tiapride, pimozide >1 year 
P15 43 67 60 30 8.8 65 11 Tiapride, sulpiride >2 years 
Mean 34  61 42 21     6.1 41.8   
Subject Age Sex Age at onset DCI (0–100) YGTSS (0–100) YGTSS tics (0–50) Motor tics
 
Vocal tics
 
Tics / min during fMRI Tics / min outside scannera MRVS (total) Medication history Time since last medical treatment 
       Simple Complex Simple Complex      
P01 33 12 63 46 26 3.9 34 11 Tiapride, sulpiride >1 year 
P02 23 61 57 27 − 2.8 44 Tiapride, sulpiride, risperidone, aripiprazole, olanzapine, pimozide 3 weeks 
P03 29 68 30 20 0.4  Missingb − Naïve 
P04 27 68 31 21 − 3.5  Missingb Tiapride, risperidone, quetiapine 3 months 
P05 22 10 47 26 16 − − 4.3 27 Tiapride 6 weeks 
P06 39 12 100 77 37 12.9 55 14 Tiapride, sulpiride 6 months 
P07 31 12 37 18 − − − 0.6 − Naïve 
P08 54 13 57 49 19 − 10.1 68 15 Haloperidole >10 years 
P09 28 50 44 24 − − 7.1 36 Tiapride >1 year 
P10 42 76 42 22 − 13.6 63 13 − Naïve 
P11 45 11 64 35 25 − 4.7 63 10 − Naïve 
P12 29 54 36 16 − 2.6 17 − Naïve 
P13 34 52 56 16 − − 9.2 48 Not known >10 years 
P14 38 11 45 22 12 − − 7.0 15 Tiapride, pimozide >1 year 
P15 43 67 60 30 8.8 65 11 Tiapride, sulpiride >2 years 
Mean 34  61 42 21     6.1 41.8   

DCI = Diagnostic Confidence Index; MRVS = modified Rush Video Scale; M = male; F = female; Tic count / min during fMRI = average number of tics counted per minute during the functional MRI experiment based on comprehensive judgement of video, EMG, and observer inside the scanning room.

aTotal tic count per minute during 2 min video recording while unobserved.

bTwo patients refused to be videotaped. + = present; − = not present; medication history lists only neuroleptic medication.

Behavioural data

Table 2 shows the behavioural data. Patients with Tourette’s syndrome responded significantly slower to S2-Go stimuli than healthy control subjects [median reaction time (range) in ms: patients with Tourette’s syndrome 479 (341–571), controls 406 (333–550); W = 41, P < 0.01]. Moreover, patients showed a slight tendency to make more errors than control subjects in total [median % error (range): patients with Tourette’s syndrome 7.5 (0–25), controls 3.75 (0–16.25); W = 72, P = 0.095]. This was because of patients omitting more responses in Go-trials than controls [patients with Tourette’s syndrome 1.25 (0–7.5), controls 0 (0–3.75); W = 57, P < 0.05], whereas there were no differences in false and too early responses. Groups were also comparable as to the number of errors of commission in NoGo-trials. After a correction of P-values for multiple comparisons on behavioural data using the Holm method, the reaction time difference was still significant (P < 0.05), while the difference between omission errors was present as a trend (P = 0.06).

Table 2

Behavioural results

 Tourette’s patients Control participants Group comparison 
 Median (range) Median (range) P-value (W
Reaction time (ms) 479 (341−571) 406 (333−550) 0.002 (41)** 
Error rate (%) 
    Go: wrong finger 1.25 (0−12.5) 1.25 (0−2.5) 0.118 (76.5) 
    Go: omissions 1.25 (0−7.5) 0 (0−3.75) 0.011 (57)* 
    Go: too early responses 0 (0−1.25) 0 (0−1.25) 0.307 (97.5) 
    NoGo: commission errors 2.5 (0−22.5) 2.5 (0−15) 0.966 (111) 
    Total errors 7.5 (0−25) 3.75 (0−16.25) 0.095 (72) 
 Tourette’s patients Control participants Group comparison 
 Median (range) Median (range) P-value (W
Reaction time (ms) 479 (341−571) 406 (333−550) 0.002 (41)** 
Error rate (%) 
    Go: wrong finger 1.25 (0−12.5) 1.25 (0−2.5) 0.118 (76.5) 
    Go: omissions 1.25 (0−7.5) 0 (0−3.75) 0.011 (57)* 
    Go: too early responses 0 (0−1.25) 0 (0−1.25) 0.307 (97.5) 
    NoGo: commission errors 2.5 (0−22.5) 2.5 (0−15) 0.966 (111) 
    Total errors 7.5 (0−25) 3.75 (0−16.25) 0.095 (72) 

Descriptive statistics are indicated for reaction times in Go-trials and different error types in Go- and NoGo-trials, respectively.

Results of Wilcoxon rank sum-tests between independent samples (patients with Tourette’s syndrome, control participants) are indicated as P- and U-values.

Asterisks show (uncorrected) significance levels: *P < 0.05, **P < 0.01.

Functional magnetic resonance imaging data

Whole brain analysis: main effects in task-related BOLD responses

Events of interest led to expected and plausible regional increases in BOLD signal across groups (data not shown). During S1 we found increased signal in visual cortex, intraparietal sulcus, and superior parietal lobe bilaterally, i.e. cortical areas known to be involved in visual perception, control of visual attention, and integration of sensory information. During S2-Go there was widespread bilateral activation of visual cortex, secondary motor areas including premotor cortex and SMA-proper, primary sensory-motor areas sparing right M1, parietal cortex, and dorsolateral prefrontal cortex. During S2-NoGo there was a similar widespread pattern of visual, sensory-motor, parietal and prefrontal activation bilaterally including the main activation clusters observed in previous Go/NoGo studies (Swick et al., 2011), i.e. anterior insula, inferior and middle frontal gyrus, SMA-proper, inferior parietal lobule, with some lateralization to the right in the anterior insula and middle frontal gyrus.

Volume of interest analysis: group comparison of task-related BOLD response

There were no areas with higher BOLD signal increases in patients with Tourette’s syndrome compared to healthy controls in any of the experimental conditions. There were also no brain areas with significant differences in activation between groups during NoGo. However, there was a cluster with significant higher signal increase during Go in controls as compared to patients with Tourette’s syndrome in the left primary and secondary motor cortex (SMA-proper, dorsal premotor cortex, M1; Table 3). There were further areas with higher signal increase during Go in controls that did not remain significant after correction for multiple comparisons in left parietal areas (Fig. 3).

Figure 3

Go condition: group comparison controls > patients with Tourette’s syndrome. Statistical parametric maps showing increases in regional BOLD signal in control participants compared with patients with Tourette’s syndrome during Go trials. Voxels significantly more activated in controls than in patients during Go trials are overlaid on a single-subject surface rendering (A) and on selected transversal, sagittal and coronal sections, respectively (B) (thresholded at P < 0.001 uncorrected, cluster extent threshold ≥10 voxels). Co-ordinates refer to MNI space. L = left; R = right.

Figure 3

Go condition: group comparison controls > patients with Tourette’s syndrome. Statistical parametric maps showing increases in regional BOLD signal in control participants compared with patients with Tourette’s syndrome during Go trials. Voxels significantly more activated in controls than in patients during Go trials are overlaid on a single-subject surface rendering (A) and on selected transversal, sagittal and coronal sections, respectively (B) (thresholded at P < 0.001 uncorrected, cluster extent threshold ≥10 voxels). Co-ordinates refer to MNI space. L = left; R = right.

Table 3

S2-Go: group comparison control participants > Tourette’s patients

MNI peak coordinates (mm)
 
T P-value FWE-corrected (voxel level)* Anatomical allocation 
x y z  
−27 −15 66 4.90 0.024 Left middle frontal gyrus, medial frontal gyrus, precentral gyrus, superior frontal gyrus (dPM, SMA-proper) 
MNI peak coordinates (mm)
 
T P-value FWE-corrected (voxel level)* Anatomical allocation 
x y z  
−27 −15 66 4.90 0.024 Left middle frontal gyrus, medial frontal gyrus, precentral gyrus, superior frontal gyrus (dPM, SMA-proper) 

Local maxima showing significant increases in BOLD signal for control participants compared with patients with Tourette’s syndrome.

*A threshold of P < 0.05 and correction for FWE was applied to correct for multiple comparisons within predefined volumes of interest (bilateral primary and secondary motor areas).

dPM = dorsal premotor cortex.

Volume of interest analysis of per cent BOLD signal change

ANOVA of event-related signal changes during S2-Go in grouped volume of interest identified significant main effects of Area [F(2,56) = 47.8; P < 0.001] and Side [F(1,28) = 8.9; P = 0.006], whereas there was no effect of Group. In addition, there was significant two-way interaction between Area and Side [F(2,56) = 16.1; P < 0.001]. Finally, an interaction between Area and Group [F(2,56) = 10.9, P < 0.001] demonstrated a region-specific group difference (Fig. 4). Post hoc t-tests identified significantly lower per cent signal changes in patients with Tourette’s syndrome compared to the control group in the cortical sensory-motor areas [0.151 ± 0.096 versus 0.235 ± 076; t(28) = 2.6, P = 0.014] whereas there was no group difference in the basal ganglia (P = 0.752) and cortical non-motor areas (P = 0.414). Event-related per cent signal change during S2-Go in the individual volume of interest is plotted in Fig. 5. Post hoc t-tests identified significantly lower per cent signal in patients with Tourette’s syndrome in the following volume of interest: left M1 (P = 0.040), left primary somatosensory cortex (P = 0.029), right primary somatosensory cortex (P = 0.050), left SMA-proper (P = 0.002), right SMA-proper (P = 0.049), left dorsal premotor cortex (P = 0.001), and right dorsal premotor cortex (P = 0.028). After correction for multiple testing using the Bonferroni method, group differences in left dorsal premotor cortex (P = 0.001) and left SMA-proper (P = 0.002) remained statistically significant.

Figure 4

Volume of interest analysis of per cent BOLD signal change during S2-Go: grouped volumes of interest. Mean per cent signal change during S2-Go as a function of volume of interest (cortical sensory-motor areas, basal ganglia, cortical non-motor areasa) and group [patients with Tourette’s syndrome (GTS), control subjects (CON)]. Error bars indicate standard error of mean (SEM).

aCortical sensory-motor areas = primary motor cortex, primary somatosensory cortex, dorsal premotor cortex, supplementary motor area proper; cortical non-motor areas = prefrontal cortex (PFC); basal ganglia = caudate nucleus, putamen, globus pallidus externus/internus.

Figure 4

Volume of interest analysis of per cent BOLD signal change during S2-Go: grouped volumes of interest. Mean per cent signal change during S2-Go as a function of volume of interest (cortical sensory-motor areas, basal ganglia, cortical non-motor areasa) and group [patients with Tourette’s syndrome (GTS), control subjects (CON)]. Error bars indicate standard error of mean (SEM).

aCortical sensory-motor areas = primary motor cortex, primary somatosensory cortex, dorsal premotor cortex, supplementary motor area proper; cortical non-motor areas = prefrontal cortex (PFC); basal ganglia = caudate nucleus, putamen, globus pallidus externus/internus.

Figure 5

Volume of interest analysis of per cent BOLD signal change during S2-Go. The figure shows plots of signal changes during S2-Go in predefined anatomical regions of interest: primary motor cortex (M1), primary somatosensory cortex (S1), dorsal premotor cortex (dPM), SMA-proper, prefrontal cortex (PFC), caudate nucleus (Caud), putamen (Put), globus pallidus externus (GPe), globus pallidus internus (GPi). CON = control participants; GTS = Gilles de la Tourette syndrome patients. Error bars indicate standard error of mean (SEM).

*Significant group difference with P < 0.05 uncorrected.

**Significant group difference with P < 0.05 Bonferroni corrected.

Figure 5

Volume of interest analysis of per cent BOLD signal change during S2-Go. The figure shows plots of signal changes during S2-Go in predefined anatomical regions of interest: primary motor cortex (M1), primary somatosensory cortex (S1), dorsal premotor cortex (dPM), SMA-proper, prefrontal cortex (PFC), caudate nucleus (Caud), putamen (Put), globus pallidus externus (GPe), globus pallidus internus (GPi). CON = control participants; GTS = Gilles de la Tourette syndrome patients. Error bars indicate standard error of mean (SEM).

*Significant group difference with P < 0.05 uncorrected.

**Significant group difference with P < 0.05 Bonferroni corrected.

Co-activation with left primary sensory-motor cortex

In patients with Tourette’s syndrome a significantly decreased co-activation with the left sensory-motor cortex was found in the contralateral sensory-motor cortex (maximum x = 23 mm, y = −12 mm, z = 72 mm, T = 6.14, PFWE-corrected = 0.024). Examining group differences at P < 0.001 uncorrected, revealed further brain areas with reduced co-activation with the left primary sensory-motor cortex (Fig. 5), representing a widespread network of cortical, subcortical and cerebellar brain areas with preponderance of primary and secondary sensory-motor areas (right M1, bilateral primary somatosensory cortex, right dorsal premotor cortex, bilateral SMA-proper, left putamen, left insula and frontal operculum, right thalamus) and areas involved in visual processing (bilateral fusiform gyrus and lingual gyrus, right middle and inferior occipital gyrus, right middle temporal gyrus).

Structural connectivity of left primary sensory-motor cortex

Group maps of probability of structural connectivity to left sensory-motor cortex revealed no obvious group differences (Fig. 6). This impression was corroborated by the results of permutation testing which did not reveal any significant difference in the probabilistic maps between groups.

Figure 6

Co-activation with left primary sensory-motor cortex: group comparison controls > patients with Tourette’s syndrome. Statistical parametric maps showing group differences in correlation with the maximum BOLD signal increase in left primary sensory-motor cortex during S2-Go with the whole BOLD time course. Voxels showing a significantly higher correlation in control participants than in patients with Tourette’s syndrome are overlaid on a single-subject surface rendering (A) and selected sections (B) (thresholded at P < 0.001 uncorrected, ≥10 voxels). The seed region for analysis of co-activation was defined by a spherical binary mask of 10 mm diameter around the global maximum in the one-sample t-test on the main effect of S2-Go across both groups: left M1/somatosensory cortex, MNI coordinates −42 −3 54. Coordinates refer to MNI space. L = left, R = right.

Figure 6

Co-activation with left primary sensory-motor cortex: group comparison controls > patients with Tourette’s syndrome. Statistical parametric maps showing group differences in correlation with the maximum BOLD signal increase in left primary sensory-motor cortex during S2-Go with the whole BOLD time course. Voxels showing a significantly higher correlation in control participants than in patients with Tourette’s syndrome are overlaid on a single-subject surface rendering (A) and selected sections (B) (thresholded at P < 0.001 uncorrected, ≥10 voxels). The seed region for analysis of co-activation was defined by a spherical binary mask of 10 mm diameter around the global maximum in the one-sample t-test on the main effect of S2-Go across both groups: left M1/somatosensory cortex, MNI coordinates −42 −3 54. Coordinates refer to MNI space. L = left, R = right.

Correlation of imaging findings with clinical and behavioural parameters

Linear regression analysis revealed a negative correlation of Go-reaction time and BOLD-signal changes during S2-Go in both groups, located bilaterally in two clusters in the anterior cingulum (first cluster: 32 voxels; maximum at x = 9, y = 9, z = 42; T = 5.16; second cluster: 10 voxels; maximum at x = −12, y = 15, z = 33; T = 4.03) but no group interaction. BOLD response during S2-Go did not correlate with Modified Rush Video Scale, YGTSS, or tic count per minute.

Tic related BOLD signal change

Tic-related brain activation throughout the entire experiment in patients with Tourette’s syndrome was found bilaterally in the precentral gyrus (45, −12, 39; T = 3.79; P < 0.001 uncorrected; and −42, −12, 42; T = 3.79; P < 0.001 uncorrected; Fig. 7).

Figure 7

Group maps of structural connectivity of left sensory-motor cortex. Group maps of probability of structural connectivity of left sensory-motor cortex for control participants (CON; upper row) and patients with Tourette’s syndrome (GTS; lower row). The seed region for probabilistic tractography was defined identical to that for functional connectivity analysis, i.e. a spherical binary mask of 10 mm diameter around the global maximum in the one-sample t-test on the main effect of S2-Go across both groups (MNI coordinates −42 −3 54). Results of probabilistic tractography are presented as sum images giving for each voxel the number of participants, in whom this voxel was reached by a pathway running through the seed region in >10% of individual samples. Results are superimposed in colour on the mean fractional anisotropy image from all participants. Coordinates refer to MNI space. L = left, R = right.

Figure 7

Group maps of structural connectivity of left sensory-motor cortex. Group maps of probability of structural connectivity of left sensory-motor cortex for control participants (CON; upper row) and patients with Tourette’s syndrome (GTS; lower row). The seed region for probabilistic tractography was defined identical to that for functional connectivity analysis, i.e. a spherical binary mask of 10 mm diameter around the global maximum in the one-sample t-test on the main effect of S2-Go across both groups (MNI coordinates −42 −3 54). Results of probabilistic tractography are presented as sum images giving for each voxel the number of participants, in whom this voxel was reached by a pathway running through the seed region in >10% of individual samples. Results are superimposed in colour on the mean fractional anisotropy image from all participants. Coordinates refer to MNI space. L = left, R = right.

Discussion

Using functional MRI during a Go/NoGo reaction time experiment in patients with ‘pure’ Tourette’s syndrome and healthy control participants, we showed that task-related BOLD signal change was decreased in cortical motor areas in patients with Tourette’s syndrome whereas activation was not different from controls in other areas. Moreover, co-activation between primary sensory-motor cortex and contralateral motor cortex, and to a lesser extent also with distributed brain areas involved in motor control, was decreased in patients. This finding of decreased co-activation was not explained by structural alterations, as there was no group difference in structural connectivity of primary sensory-motor cortex. Altered activation patterns of cortical motor areas in patients with Tourette’s syndrome were accompanied by a significantly poorer behavioural task performance with patients with Tourette’s syndrome both responding slower and committing more errors, but were not associated with clinical measures.

Behavioural shortcomings of patients with Tourette’s syndrome in our experiment are in line with previous studies. Although several previous studies of patients with uncomplicated Tourette’s syndrome excluding individuals with co-morbid attention deficit hyperactivity disorder and obsessive compulsive disorder have failed to demonstrate marked behavioural differences between groups in rather simple tasks of behavioural inhibition or motor control (Channon et al., 2006; Roessner et al., 2008; Werner et al., 2011; Jung et al., 2013), behavioural deficits in tasks of motor and executive control have become prominent with increasing task difficulty in these patients (Werner et al., 2011). They were also slower in a continuous performance task with increasing inhibitory demands (Channon et al., 2009) and exhibited deficits in a previous study of grip force control (Nowak et al., 2005) as well as in a serial choice reaction time experiment (Sheppard et al., 2000). Their reaction times were also slower in a Go/NoGo task (Eichele et al., 2010). The fact that patients with Tourette’s syndrome did not show higher numbers of errors of commission in NoGo-trials but instead more often failed to respond to Go-stimuli points towards a strategy focusing on the avoidance of false and preponderant responses, which is likewise reflected by the markedly slower reaction time. Thus, the performance of patients with Tourette’s syndrome in our study may be considered to reflect inhibitory overachievement at the cost of both speed and accuracy. These findings may also reflect the fact that the paradigm used in our study was not a classical Go/NoGo or stop paradigm but involved the preparation of an externally determined movement for a relatively long time period (2000–6000 ms) until a second signal instructed participants to either execute (Go) or inhibit (NoGo) this movement. We cannot rule out that behavioural shortcomings of patients with Tourette’s syndrome result, at least in part, from interference by tics that were frequently observed during the experiment. However, as tics constitute a defining symptom and integral part of Tourette’s syndrome, considerations on how patients with Tourette’s syndrome might perform without tics are speculative.

Activation during S2-Go and S2-NoGo was decreased in sensory-motor areas but not in the basal ganglia and non-motor cortical areas in patients with Tourette’s syndrome. This result was contrary to our initial hypothesis of increased brain activation in fronto-parietal control networks and to earlier functional MRI studies reporting patterns of over-activation of brain areas during cognitive control tasks (Peterson et al., 1998; Baym et al., 2008). However, these findings are in keeping with recent functional MRI studies that rather consistently described decreased activation in primary and secondary motor cortex during different tasks of motor or behavioural control, or simple movements (Jackson et al., 2011; Roessner et al., 2012; Jung et al., 2013). These imaging findings are paralleled by reduced motor excitability prior to finger movements (Jackson et al., 2013) or increases of intracortical inhibition during the preparation of finger movement (Heise et al., 2010) as determined by transcranial magnetic stimulation.

We interpret these findings as an effect of functional down-regulation of cortical motor areas in patients with Tourette’s syndrome reflecting adaptive compensatory mechanisms counteracting over-activity within cortico-striato-cortical loops that are a characteristic feature of Tourette’s syndrome (Ganos et al., 2013). Volume of interest analysis demonstrated that this supposed down-regulation is specific for primary and secondary sensory motor areas including M1, primary somatosensory cortex, dorsal premotor cortex, and SMA-proper. We were unable to identify a specific brain region driving this assumed downregulation. However, we found decreased co-activation between the left primary sensory-motor cortex and a widespread network of brain areas involved in sensory-motor integration and processing, which may constitute a potential mechanism of reducing the gain of the cortical motor system. This interpretation is in line with recent studies pointing towards compensatory mechanisms that alter cortical excitability and functional connectivity of distributed brain regions in Tourette’s syndrome (Jackson et al., 2011; Werner et al., 2011; Jung et al., 2013). In this context it is interesting to note that brain areas showing reduced activation during Go-trials in our study largely match the sensory-motor brain network activated at the beginning of tics (Stern et al., 2000; Bohlhalter et al., 2006). We thus speculate that a compensatory downregulation specifically concerns those brain areas involved in the behavioural hallmarks of Tourette’s syndrome. In our analysis, a reduction of co-activation between left and right sensory-motor cortex was the most robust observation. This may reflect alterations of interhemispheric connectivity, which is in line with behavioural, electrophysiological, and imaging results from previous studies. Reports of altered size and microstructure of the corpus callosum in patients with Tourette’s syndrome were interpreted as a result of neural plasticity within the context of the disease (Plessen et al., 2004, 2006). In another study, adults with Tourette’s syndrome showed impairment in bimanual tasks (Margolis et al., 2006). In a combined diffusion tensor imaging and transcranial magnetic stimulation analysis of a subgroup of the sample reported here we were also able to demonstrate abnormal functional interhemispheric connectivity that was accompanied by an altered structure-function relationship in motor parts of the corpus callosum (Baumer et al., 2010).

As to potentially underlying mechanisms, we did not find any signs of altered structural connectivity mirroring the observed decrease in co-activation between primary sensory-motor cortex and other brain areas. Thus, the effect observed may not directly be ascribed to structural brain changes. However, there is previous evidence of altered structure of the sensory-motor brain areas with findings of somatosensory or motor cortex thinning (Sowell et al., 2008; Fahim et al., 2010) and altered structural connectivity in sensory-motor circuits (Baumer et al., 2010; Jackson et al., 2011; Neuner et al., 2011), which at least in part are considered to reflect compensatory structural plasticity. We also demonstrated structural alterations in somatosensory white-matter pathways in our sample as reported previously (Thomalla et al., 2009), whereas in the current analysis we did not find a group difference in structural connectivity of primary sensory motor cortex. These observations may point towards a more complex interaction between distributed structural alterations and changes of functional connectivity within the sensory-motor system that act together in compensatory downregulation of the cortical motor brain.

Figure 8

Tic-related brain activation in patients with Tourette’s syndrome. Statistical parametric maps showing tic-related increase in regional BOLD signal in patients with Tourette’s syndrome overlaid on selected transversal, sagittal and coronal sections (thresholded at P < 0.001 uncorrected). Coordinates refer to MNI space. L = left, R = right.

Figure 8

Tic-related brain activation in patients with Tourette’s syndrome. Statistical parametric maps showing tic-related increase in regional BOLD signal in patients with Tourette’s syndrome overlaid on selected transversal, sagittal and coronal sections (thresholded at P < 0.001 uncorrected). Coordinates refer to MNI space. L = left, R = right.

There are alternative explanations. One could argue that the observed decreased task-related activation simply reflects artefacts resulting from either tic-related activity throughout the experiment or increased movement-by-susceptibility interactions induced by tic-related head movements. To account for tic-related activity we used a comprehensive set-up to monitor tics during the functional MRI experiment. Although we certainly may have missed some tics it is unlikely that a great number of tics escaped our monitoring. In an exploratory analysis we identified tic-related brain activation bilaterally in the precentral gyrus (Fig. 8). These two clusters are within the larger regions of brain activation related to tics in previous studies (Stern et al., 2000; Bohlhalter et al., 2006). Although our study was not specifically designed to identify tic-related brain activation, this observation further supports our hypothesis of specific downregulation of brain areas involved in the behavioural hallmarks of Tourette’s syndrome, including tics. On the other hand, the fact that tic-related brain activation does not overlap with the areas of decreased brain activation during Go in patients with Tourette’s syndrome argues against the concern that decreased regional brain activation in patients with Tourette’s syndrome might simply result from tic-related brain activation during the experiment.

To further ensure that the additional tic-regressor (i.e. tics) used in the regression model for functional MRI data in patients with Tourette’s syndrome was not assigned variance to the disadvantage of our regressors of interest we performed an additional analysis that completely disregarded tics. This resulted in virtually the same activation maps during S2-Go and -NoGo in patients (not illustrated). Finally, none of these alternatives would convincingly explain the regional specificity of decreased activations in cortical motor areas. Visual cortex and parietal association areas were not differently activated in groups although main effects demonstrated that both were largely involved in the task. If simply technical artefacts or some sort of continuous over-activation of the Tourette brain was considered as an explanation, we would rather expect increased co-activation between the brain areas demonstrating continuous task-unrelated hyperactivity, but the opposite was the case in our study.

Compensatory mechanisms resulting in decreased cortical motor activation come at the price of behavioural cost: patients with Tourette’s syndrome were markedly slower and missed more responses while maintaining inhibitory control and avoiding both preponderant responses and errors of commission. On the other hand, brain activation decreases did not correlate with clinical measures. How can this be reconciled with current pathophysiology models of Tourette’s syndrome?

Dysfunction in cortico-striato-thalamo-cortical circuits leading to hyper-excitability of sensory-motor areas has been proposed as a core problem in Tourette’s syndrome (Mink, 2006). Recent neuroimaging data support this idea providing evidence of altered brain networks demonstrating immature patterns in children with Tourette’s syndrome (Channon et al., 2009; Church et al., 2009a,b; Worbe et al., 2012). In older children and adults with Tourette’s syndrome, however, any observed alterations both in brain structure and function are likely to be a composite of primary pathophysiology and compensatory mechanisms (Serrien et al., 2005; Jung et al., 2013). This combination of primary pathology and compensatory changes may explain the partly contradictory association between brain activation changes and behaviour or clinical hallmarks in Tourette’s syndrome in previous studies and the lack of an association between clinical characteristics and brain activation in our sample.

In conclusion, our study demonstrated regionally distributed decreases in brain activation and decreased co-activation of cortical sensory-motor brain regions in patients with pure Tourette’s syndrome. We interpret these findings as a result of adaptive reorganization in fronto-parietal networks representing a response to primary abnormal processing and over-activation in cortico-striato-cortical circuits in these patients. We speculate that this compensatory functional plasticity enables patients with Tourette’s syndrome to maintain inhibitory control over motor output even in demanding behavioural tasks and presumably daily life. Control, however, comes at the price of behavioural costs.

Acknowledgements

We thank Prof. Diane Swick for providing activation likelihood estimation maps from quantitative meta-analysis of functional MRI studies during Nogo experiments.

Funding

This work was supported by the Deutsche Forschungsgemeinschaft (DFG) [grant number MU1692/2-1 and /2-2 to A.M., H.R.S. and A.S.], and by the University of Hamburg [grant number F168-1 to G.T.].

Abbreviations

    Abbreviations
  • BOLD

    blood oxygen level-dependent

  • M1

    primary motor cortex

  • SMA

    supplementary motor area

  • YGTSS

    Yale Global Tic Severity Scale

References

American Psychiatric Association (APA)
Diagnostic and statistical manual of mental disorders
 
4th edn
 
text revision (DSM-IV-TR). Arlington, VA: American Psychiatric Publishing; 2000
Baumer
T
Thomalla
G
Kroeger
J
Jonas
M
Gerloff
C
Hummel
FC
, et al.  . 
Interhemispheric motor networks are abnormal in patients with Gilles de la Tourette syndrome
Mov Disord
 , 
2010
, vol. 
25
 (pg. 
2828
-
37
)
Baym
CL
Corbett
BA
Wright
SB
Bunge
SA
Neural correlates of tic severity and cognitive control in children with Tourette syndrome
Brain
 , 
2008
, vol. 
131
 (pg. 
165
-
79
)
Behrens
TE
Woolrich
MW
Jenkinson
M
Johansen-Berg
H
Nunes
RG
Clare
S
, et al.  . 
Characterization and propagation of uncertainty in diffusion-weighted MR imaging
Magn Reson Med
 , 
2003
, vol. 
50
 (pg. 
1077
-
88
)
Bohlhalter
S
Goldfine
A
Matteson
S
Garraux
G
Hanakawa
T
Kansaku
K
, et al.  . 
Neural correlates of tic generation in Tourette syndrome: an event-related functional MRI study
Brain
 , 
2006
, vol. 
129
 (pg. 
2029
-
37
)
Buse
J
Schoenefeld
K
Munchau
A
Roessner
V
Neuromodulation in Tourette syndrome: dopamine and beyond
Neurosci Biobehav Rev
 , 
2013
, vol. 
37
 (pg. 
1069
-
84
)
Channon
S
Drury
H
Martinos
M
Robertson
MM
Orth
M
Crawford
S
Tourette's syndrome (TS): inhibitory performance in adults with uncomplicated TS
Neuropsychology
 , 
2009
, vol. 
23
 (pg. 
359
-
66
)
Channon
S
Gunning
A
Frankl
J
Robertson
MM
Tourette's syndrome (TS): cognitive performance in adults with uncomplicated TS
Neuropsychology
 , 
2006
, vol. 
20
 (pg. 
58
-
65
)
Church
JA
Fair
DA
Dosenbach
NU
Cohen
AL
Miezin
FM
Petersen
SE
, et al.  . 
Control networks in paediatric Tourette syndrome show immature and anomalous patterns of functional connectivity
Brain
 , 
2009a
, vol. 
132
 (pg. 
225
-
38
)
Church
JA
Wenger
KK
Dosenbach
NU
Miezin
FM
Petersen
SE
Schlaggar
BL
Task control signals in pediatric tourette syndrome show evidence of immature and anomalous functional activity
Front Hum Neurosci
 , 
2009b
, vol. 
3
 pg. 
38
 
Eichele
H
Eichele
T
Hammar
A
Freyberger
HJ
Hugdahl
K
Plessen
KJ
Go/NoGo performance in boys with Tourette syndrome
Child Neuropsychol
 , 
2010
, vol. 
16
 (pg. 
162
-
8
)
Fahim
C
Yoon
U
Das
S
Lyttelton
O
Chen
J
Arnaoutelis
R
, et al.  . 
Somatosensory-motor bodily representation cortical thinning in Tourette: effects of tic severity, age and gender
Cortex
 , 
2010
, vol. 
46
 (pg. 
750
-
60
)
Finis
J
Moczydlowski
A
Pollok
B
Biermann-Ruben
K
Thomalla
G
Heil
M
, et al.  . 
Echoes from childhood–imitation in Gilles de la Tourette syndrome
Mov Disord
 , 
2012
, vol. 
27
 (pg. 
562
-
5
)
Friston
KJ
Worsley
KJ
Frackowiak
RS
Frith
CD
Holmes
A
statistical parametric maps in functional imaging: a general linear approach human brain mapping
Hum Brain Mapp
 , 
1995
, vol. 
2
 (pg. 
189
-
210
)
Ganos
C
Ogrzal
T
Schnitzler
A
Munchau
A
The pathophysiology of echopraxia/echolalia: relevance to Gilles de la Tourette syndrome
Mov Disord
 , 
2012
, vol. 
27
 (pg. 
1222
-
9
)
Ganos
C
Roessner
V
Munchau
A
The functional anatomy of Gilles de la Tourette syndrome
Neurosci Biobehav Rev
 , 
2013
, vol. 
37
 (pg. 
1050
-
62
)
Georgiou
N
Bradshaw
JL
Phillips
JG
Bradshaw
JA
Chiu
E
The Simon effect and attention deficits in Gilles de la Tourette's syndrome and Huntington's disease
Brain
 , 
1995
, vol. 
118
 
Pt 5
(pg. 
1305
-
18
)
Glascher
J
Visualization of group inference data in functional neuroimaging
Neuroinformatics
 , 
2009
, vol. 
7
 (pg. 
73
-
82
)
Goetz
CG
Pappert
EJ
Louis
ED
Raman
R
Leurgans
S
Advantages of a modified scoring method for the Rush Video-Based Tic Rating Scale
Mov Disord
 , 
1999
, vol. 
14
 (pg. 
502
-
6
)
Heinz
A
Braus
DF
Smolka
MN
Wrase
J
Puls
I
Hermann
D
, et al.  . 
Amygdala-prefrontal coupling depends on a genetic variation of the serotonin transporter
Nat Neurosci
 , 
2005
, vol. 
8
 (pg. 
20
-
1
)
Heise
KF
Steven
B
Liuzzi
G
Thomalla
G
Jonas
M
Muller-Vahl
K
, et al.  . 
Altered modulation of intracortical excitability during movement preparation in Gilles de la Tourette syndrome
Brain
 , 
2010
, vol. 
133
 (pg. 
580
-
90
)
Jackson
SR
Parkinson
A
Jung
J
Ryan
SE
Morgan
PS
Hollis
C
, et al.  . 
Compensatory neural reorganization in Tourette syndrome
Curr Biol
 , 
2011
, vol. 
21
 (pg. 
580
-
5
)
Jackson
SR
Parkinson
A
Manfredi
V
Millon
G
Hollis
C
Jackson
GM
Motor excitability is reduced prior to voluntary movements in children and adolescents with Tourette syndrome
J Neuropsychol
 , 
2013
, vol. 
7
 (pg. 
29
-
44
)
Jung
J
Jackson
SR
Parkinson
A
Jackson
GM
Cognitive control over motor output in Tourette syndrome
Neurosci Biobehav Rev
 , 
2013
, vol. 
37
 (pg. 
1016
-
25
)
Leckman
JF
Peterson
BS
King
RA
Scahill
L
Cohen
DJ
Phenomenology of tics and natural history of tic disorders
Adv Neurol
 , 
2001
, vol. 
85
 (pg. 
1
-
14
)
Leckman
JF
Riddle
MA
Hardin
MT
Ort
SI
Swartz
KL
Stevenson
J
, et al.  . 
The Yale Global Tic Severity Scale: initial testing of a clinician-rated scale of tic severity
J Am Acad Child Adolesc Psychiatry
 , 
1989
, vol. 
28
 (pg. 
566
-
73
)
Margolis
A
Donkervoort
M
Kinsbourne
M
Peterson
BS
Interhemispheric connectivity and executive functioning in adults with Tourette syndrome
Neuropsychology
 , 
2006
, vol. 
20
 (pg. 
66
-
76
)
Mink
JW
Neurobiology of basal ganglia and Tourette syndrome: basal ganglia circuits and thalamocortical outputs
Adv Neurol
 , 
2006
, vol. 
99
 (pg. 
89
-
98
)
Mayka
MA
Corcos
DM
Leurgans
SE
Vaillancourt
DE
Three-dimensional locations and boundaries of motor and premotor cortices as defined by functional brain imaging: a meta-analysis
Neuroimage
 , 
2006
, vol. 
31
 (pg. 
1453
-
74
)
Mueller
SC
Jackson
GM
Dhalla
R
Datsopoulos
S
Hollis
CP
Enhanced cognitive control in young people with Tourette's syndrome
Curr Biol
 , 
2006
, vol. 
16
 (pg. 
570
-
3
)
Neuner
I
Kupriyanova
Y
Stocker
T
Huang
R
Posnansky
O
Schneider
F
, et al.  . 
Microstructure assessment of grey matter nuclei in adult tourette patients by diffusion tensor imaging
Neurosci Lett
 , 
2011
, vol. 
487
 (pg. 
22
-
6
)
Nowak
DA
Rothwell
J
Topka
H
Robertson
MM
Orth
M
Grip force behavior in Gilles de la Tourette syndrome
Mov Disord
 , 
2005
, vol. 
20
 (pg. 
217
-
23
)
Oldfield
RC
The assessment and analysis of handedness: the Edinburgh handedness inventory
Neuropsychologia
 , 
1971
, vol. 
9
 pg. 
97
 
Ozonoff
S
Strayer
DL
McMahon
WM
Filloux
F
Inhibitory deficits in Tourette syndrome: a function of comorbidity and symptom severity
J Child Psychol Psychiatry
 , 
1998
, vol. 
39
 (pg. 
1109
-
18
)
Peterson
BS
Neuroimaging studies of Tourette syndrome: a decade of progress
Adv Neurol
 , 
2001
, vol. 
85
 (pg. 
179
-
96
)
Peterson
BS
Skudlarski
P
Anderson
AW
Zhang
H
Gatenby
JC
Lacadie
CM
, et al.  . 
A functional magnetic resonance imaging study of tic suppression in Tourette syndrome
Arch Gen Psychiatry
 , 
1998
, vol. 
55
 (pg. 
326
-
33
)
Peterson
BS
Staib
L
Scahill
L
Zhang
H
Anderson
C
Leckman
JF
, et al.  . 
Regional brain and ventricular volumes in Tourette syndrome
Arch Gen Psychiatry
 , 
2001
, vol. 
58
 (pg. 
427
-
40
)
Plessen
KJ
Gruner
R
Lundervold
A
Hirsch
JG
Xu
D
Bansal
R
, et al.  . 
Reduced white matter connectivity in the corpus callosum of children with Tourette syndrome
J Child Psychol Psychiatry
 , 
2006
, vol. 
47
 (pg. 
1013
-
22
)
Plessen
KJ
Wentzel-Larsen
T
Hugdahl
K
Feineigle
P
Klein
J
Staib
LH
, et al.  . 
Altered interhemispheric connectivity in individuals with Tourette's disorder
Am J Psychiatry
 , 
2004
, vol. 
161
 (pg. 
2028
-
37
)
R Core Team
R: a language and environment for statistical computing
 , 
2012
Vienna, Austria
R Foundation for Statistical Computing
Robertson
MM
Tourette syndrome, associated conditions and the complexities of treatment
Brain
 , 
2000
, vol. 
123
 
Pt 3
(pg. 
425
-
62
)
Robertson
MM
Banerjee
S
Kurlan
R
Cohen
DJ
Leckman
JF
McMahon
W
, et al.  . 
The Tourette syndrome diagnostic confidence index: development and clinical associations
Neurology
 , 
1999
, vol. 
53
 (pg. 
2108
-
12
)
Roessner
V
Albrecht
B
Dechent
P
Baudewig
J
Rothenberger
A
Normal response inhibition in boys with Tourette syndrome
Behav Brain Funct
 , 
2008
, vol. 
4
 pg. 
29
 
Roessner
V
Wittfoth
M
Schmidt-Samoa
C
Rothenberger
A
Dechent
P
Baudewig
J
Altered motor network recruitment during finger tapping in boys with Tourette syndrome
Hum Brain Mapp
 , 
2012
, vol. 
33
 (pg. 
666
-
75
)
Serrien
DJ
Orth
M
Evans
AH
Lees
AJ
Brown
P
Motor inhibition in patients with Gilles de la Tourette syndrome: functional activation patterns as revealed by EEG coherence
Brain
 , 
2005
, vol. 
128
 (pg. 
116
-
25
)
Sheppard
DM
Bradshaw
JL
Georgiou
N
Bradshaw
JA
Lee
P
Movement sequencing in children with Tourette's syndrome and attention deficit hyperactivity disorder
Mov Disord
 , 
2000
, vol. 
15
 (pg. 
1184
-
93
)
Singer
HS
Tourette's syndrome: from behaviour to biology
Lancet Neurol
 , 
2005
, vol. 
4
 (pg. 
149
-
59
)
Singer
HS
Minzer
K
Neurobiology of Tourette's syndrome: concepts of neuroanatomic localization and neurochemical abnormalities
Brain Dev
 , 
2003
, vol. 
25
 
Suppl 1
(pg. 
S70
-
84
)
Smith
SM
Jenkinson
M
Woolrich
MW
Beckmann
CF
Behrens
TE
Johansen-Berg
H
, et al.  . 
Advances in functional and structural MR image analysis and implementation as FSL
Neuroimage
 , 
2004
, vol. 
23
 
Suppl 1
(pg. 
S208
-
19
)
Sowell
ER
Kan
E
Yoshii
J
Thompson
PM
Bansal
R
Xu
D
, et al.  . 
Thinning of sensorimotor cortices in children with Tourette syndrome
Nat Neurosci
 , 
2008
, vol. 
11
 (pg. 
637
-
9
)
Stern
E
Silbersweig
DA
Chee
KY
Holmes
A
Robertson
MM
Trimble
M
, et al.  . 
A functional neuroanatomy of tics in Tourette syndrome
Arch Gen Psychiatry
 , 
2000
, vol. 
57
 (pg. 
741
-
8
)
Swick
D
Ashley
V
Turken
U
Are the neural correlates of stopping and not going identical? Quantitative meta-analysis of two response inhibition tasks
Neuroimage
 , 
2011
, vol. 
56
 (pg. 
1655
-
65
)
Thomalla
G
Siebner
HR
Jonas
M
Baumer
T
Biermann-Ruben
K
Hummel
F
, et al.  . 
Structural changes in the somatosensory system correlate with tic severity in Gilles de la Tourette syndrome
Brain
 , 
2009
, vol. 
132
 (pg. 
765
-
77
)
Werner
CJ
Stocker
T
Kellermann
T
Bath
J
Beldoch
M
Schneider
F
, et al.  . 
Altered motor network activation and functional connectivity in adult Tourette's syndrome
Hum Brain Mapp
 , 
2011
, vol. 
32
 (pg. 
2014
-
26
)
Wittchen
H-U
Wunderlich
U
Gruschwitz
S
Zaudig
M
SKID-I. Strukturiertes Klinisches Interview für DSM-IV
 
Achse I: Psychische Störungen. Göttingen Bern Toronto Seattle: Hogrefe; 1997
Worbe
Y
Malherbe
C
Hartmann
A
Pelegrini-Issac
M
Messe
A
Vidailhet
M
, et al.  . 
Functional immaturity of cortico-basal ganglia networks in Gilles de la Tourette syndrome
Brain
 , 
2012
, vol. 
135
 (pg. 
1937
-
46
)