Abstract

The precise localizations of the neural substrates of voluntary inhibition are still debated. It has been hypothesized that, in humans, this executive function relies upon a right-lateralized pathway comprising the inferior frontal gyrus and the presupplementary motor area, which would control the neural processes for movement inhibition acting through the right subthalamic nucleus (STN). We assessed the role of the right STN, via a countermanding reaching task, in 10 Parkinson’s patients receiving high-frequency electrical stimulation of the STN of both hemispheres (deep brain stimulation, DBS) and in 13 healthy subjects. We compared the performance of Parkinson’s patients in 4 experimental conditions: DBS-ON, DBS-OFF, DBS-OFF right, and DBS-OFF left. We found that 1) inhibitory control is improved only when both DBS are active, that is, the reaction time to the stop signal is significantly shorter in the DBS-ON condition than in all the others, 2) bilateral stimulation of STN restores the inhibitory control to a near-normal level, and 3) DBS does not cause a general improvement in task-related motor function as it does not affect the length of the reaction times of arm movements, that is, in our experimental context, STN seems to play a selective role in response inhibition.

Introduction

The ability to stop a pending action is fundamental for survival in a natural environment where events cannot be fully predicted. Sudden events, such as the appearance of a physical obstacle, often require a quick change of the planned motor strategy and the first step toward this goal is to suppress the preprogrammed actions. Thus voluntary inhibition plays a crucial role in cognitive control and behavioral flexibility. The neural underpinnings of this executive function are still debated and controversial. In particular, little is known about the neural processes underlying the inhibitory control of arm reaching movements despite their ecological relevance in primates. In fact, differently from saccades or key presses, these movements are those that, outside neurophysiology laboratories, allow physical interactions with the environment, thus leading to material outcomes such as acquisition of food or tools. As a consequence, reaches might require different neural processing from other relatively simpler movements.

It has been hypothesized that, in humans, voluntary inhibition of manual movements (key presses) relies upon a right-lateralized frontal–basal ganglia–thalamic pathway (Aron et al. 2007b). This network comprises the inferior frontal gyrus (Aron et al. 2003; Chambers et al. 2006), the presupplementary motor area (Aron et al. 2007a), and the subthalamic nucleus (STN; Aron and Poldrack 2006; van den Wildenberg et al. 2006) of the right hemisphere. Aron et al. (2007b) suggested that prefrontal areas countermand planned movements through the right STN. However, other evidence suggests that a larger network supports the executive control of hand motor suppression (e.g., Li et al. 2008; Zandbelt and Vink 2010).

The aim of the present work was to assess the role of one of the components of the network, the right STN, in the inhibitory performance. The STN plays a critical role in the control of movements by integrating cortical inputs from several motor areas (Mink 1996; Romanelli et al. 2005). Not surprisingly, alteration of STN functioning leads to loss of the ability to control movements, as in the case of Parkinson’s disease (PD; Obeso et al. 2008), and this control can be partially restored by deep brain stimulation (DBS) (Breit et al. 2004; Perlmutter and Mink 2006). Even though DBS of the STN is frequently adopted for treating patients with advanced PD (e.g., Benabid et al. 2009), the mechanisms underlying its effects remain to be clarified. DBS probably switches off pathological disruption of activity in the STN and imposes a new type of discharge with beneficial effects (Garcia et al. 2005). As a consequence, DBS patients are an ideal model in which to study arm suppression because by activating or deactivating the stimulators, we can control the status of the STN. To this end, we administered a reaching version of the countermanding task (Mirabella et al. 2006, 2008, 2009) to Parkinson’s patients after bilateral STN DBS placement. In this paradigm, a subject’s ability to control the production of voluntary movements is probed by randomly intermixing trials in which a speeded responses is required (no-stop trials) with trials where movements has to be withheld (stop trials). The paradigm yields an estimate of the duration of the suppression process (stop-signal reaction time, SSRT), a phenomenon which is not directly measurable but which can be estimated using the race model (Logan and Cowan 1984, see Fig. 1A).

Figure 1.

(A) Logic of the race model. The race model assumes that the performance in the countermanding reaching task is the outcome of a race between 2 independent processes: a GO process (black dashed line) and a STOP process (gray dashed line). The GO and STOP processes are initiated by the presentation of the go and the stop signals, respectively. On stop trials, the GO process begins before the STOP process because the stop signal is always presented at variable delays (SSDs) after the go signal. When the GO process finishes before the STOP process, the reaching movement is not cancelled (panel A, right) and vice versa (panel A, left). The time it takes to respond to the stop signal is the SSRT. (B) Temporal sequence of the visual displays for no-stop and stop trials in the countermanding reaching task. All trials began with the presentation of a central stimulus. After a variable holding delay (500–800 ms), it disappeared and simultaneously a target appeared to the right, acting as a go-signal. In the no-stop trials, subjects had to start a speeded reaching movement toward the peripheral target. Randomly, on a fraction of interleaved trials (33%), the central stimulus reappeared after variable delays (SSDs), informing subjects to inhibit movement initiation. In these stop trials, if subjects countermanded the planned movement keeping the arm on the central stimulus the trial was scored as a stop-success trial. Otherwise, if subjects executed the reaching movement, the trial was scored as a stop-failure trial (not shown). (C) Estimate of the SSRT. The independence assumption of the race model implies that the distribution of RTs on stop trials (whether a response is cancelled or not) is the same as the distribution of RTs of no-stop trials. If the SSD is varied so that the probability of making an error equals 0.5, then SSRT can be estimated by subtracting the SSD from the mean value of the RT distribution of no-stop trials (Logan and Cowan 1984).

Figure 1.

(A) Logic of the race model. The race model assumes that the performance in the countermanding reaching task is the outcome of a race between 2 independent processes: a GO process (black dashed line) and a STOP process (gray dashed line). The GO and STOP processes are initiated by the presentation of the go and the stop signals, respectively. On stop trials, the GO process begins before the STOP process because the stop signal is always presented at variable delays (SSDs) after the go signal. When the GO process finishes before the STOP process, the reaching movement is not cancelled (panel A, right) and vice versa (panel A, left). The time it takes to respond to the stop signal is the SSRT. (B) Temporal sequence of the visual displays for no-stop and stop trials in the countermanding reaching task. All trials began with the presentation of a central stimulus. After a variable holding delay (500–800 ms), it disappeared and simultaneously a target appeared to the right, acting as a go-signal. In the no-stop trials, subjects had to start a speeded reaching movement toward the peripheral target. Randomly, on a fraction of interleaved trials (33%), the central stimulus reappeared after variable delays (SSDs), informing subjects to inhibit movement initiation. In these stop trials, if subjects countermanded the planned movement keeping the arm on the central stimulus the trial was scored as a stop-success trial. Otherwise, if subjects executed the reaching movement, the trial was scored as a stop-failure trial (not shown). (C) Estimate of the SSRT. The independence assumption of the race model implies that the distribution of RTs on stop trials (whether a response is cancelled or not) is the same as the distribution of RTs of no-stop trials. If the SSD is varied so that the probability of making an error equals 0.5, then SSRT can be estimated by subtracting the SSD from the mean value of the RT distribution of no-stop trials (Logan and Cowan 1984).

The SSRT is a key behavioral parameter for evaluating neural substrates of volitional inhibition. In fact, those brain regions whose changes in activity produce a systematic modification of the SSRT can be taken as neural substrates of the suppression processes. To our knowledge, there are 3 studies that have tried to assess the role of the STN in response inhibition during a countermanding task exploiting the features of the DBS implant, and they provided opposing results. On the one hand, Ray et al. (2009) compared the effects of unilateral left- and right-sided stimulation with the condition in which both DBS were off. They found that irrespective of which DBS was on the SSRT was longer than when they were both off. Furthermore, this effect was larger when the left DBS was on. On the other hand, van den Wildenberg et al. (2006) and Swann et al. (2011) found that bilateral DBS decreased the length of the SSRT. Thus, it is unclear whether the DBS of the STN does or does not improve the speed of inhibitory responses. In addition, the findings of Ray et al. (2009) do not support the idea of the existence of a right-lateralized inhibitory network, while van den Wildenberg et al. (2006) leave the question open. An important feature of both studies was that patients could take their medication. As it is known that l-dopa and dopamine agonists affect cognition (Cools 2006), the specific effects of STN DBS might be difficult to isolate.

Therefore, we surveyed the performance of Parkinson’s patients with a bilateral DBS implant during the countermanding task, after washing out of medications, in 4 experimental conditions: 1) DBS-OFF on both sides, 2) DBS-ON on both sides, 3) DBS-OFF only on the left side, 4) DBS-OFF only on the right side. The advantage of this experimental design was 2-fold: 1) each patient served as his own control and 2) all possible combinations of the DBS states were studied.

This approach also allowed assessment of changes in the response strategies of PD patients in the context of the countermanding task. In fact, the rules of this task create a conflict on all no-stop trials because subjects are instructed to move as fast as possible, but at the same time, they tend to delay movement initiation to wait for the occurrence of a possible stop signal. As a consequence healthy subjects had longer reaction times (RTs) when executing go-trials intermixed with stop trials than when executing go-trials alone (“delay strategy”; Mirabella et al. 2006; Verbruggen and Logan 2009; Zandbelt and Vink, 2010). In addition, the occurrence of stop trials affects the RTs of responses produced in the immediately subsequent no-stop trials (“procrastination strategy”; Verbruggen and Logan 2009; Zandbelt and Vink 2010; Jahfari et al. 2009; Mirabella et al. 2006). These are both examples of proactive control, that is, control over response execution in anticipation of known task demands, driven by endogenous signals which, in the case of the countermanding task, are produced by the awareness of the presentation of stop trials. Hence, we tested whether and eventually how DBS stimulation affects these control strategies.

Finally, to provide an indication of the degree of recovery of PD patients following the DBS treatment, we compared both the SSRT and the RT in no-stop trials in patients with those measured in age-matched healthy subjects.

Materials and Methods

Subjects and Clinical Assessment

From the outpatients of the Parkinson’s unit of the Istituto di ricovero e cura a carattere scientifico Neuromed Hospital, we selected 13 patients who had undergone bilateral STN DBS implantation to address poor control of cardinal PD symptoms, the onset of freezing and dystonia, and/or disabling side effects induced by chronic administration of dopaminergic drugs. The time elapsing from implant to this study ranged from 1 to 5 years (mean ± standard error [SE], 3.4 ± 0.4). All patients were affected by idiopathic PD, which had been diagnosed on average 17.5 ± 1.6 years (range: 8–24 years) before the present study. All patients were in stable treatment with chronic bilateral subthalamic stimulation complemented by administration of l-dopa and dopamine agonists, and they did not present severe sensory deficits. We excluded patients with overt signs of dementia and those with severe tremor. As we were interested in analyzing the effects of DBS, patients did not take medications overnight prior to the study and thus, at the time of testing, were in the “practical defined off-state” (Moro et al. 1999). Furthermore, all tests were performed 60 min after any stimulator was switched off or on, so that they were tested in near-steady motor status (Lopiano et al. 2003; Temperli et al. 2003; Sturman et al. 2008). Before starting the task, the patient’s motor symptoms were rated by a neurologist (N.M.) using the Unified Parkinson’s Disease Rating Scale part 3 (UPDRS3). Three patients could not complete the study because the bilateral cessation of DBS caused severe impairment of their motor abilities. The clinical features of the remaining patients are summarized in Table 1. We also tested 13 age-matched healthy subjects (3 females, 10 males; age range: 55–68, mean ± SE: 60.7 ± 1.3 years; t-test t21 = 0.27, P = 0.79) with normal or corrected-to-normal vision.

Table 1

Clinical data of PD patients with bilateral implantation of DBS participating in the experiment

 Age Sex Years since diagnosis Years since surgery Hoehn & Yahr (ON) l-Dopa eq/Kg UPDRS 3
 
 ON OFF right OFF left OFF 
CR 54 24 400 15 19 21 24 
CG 67 17 890 30 35 36 43 
CL 53 16 920 21 33 31 50 
DA 64 720 34 40 38 35.5 
FS 53 15 304 12 15 17 22.5 
ME 64 22 920 28 36 34 46 
SA 61 17 1050 21 36 35 33 
ScA 64 23 150 12 15.5 19 22 
SG 55 22 640 14 12 18 
TN 66 16 640 14 30 23 34 
Mean (SEM) 60.1 (1.8)  17.5 (1.6) 3.5 (0.4) 1.4 (0.2) 699.2 (101.3) 19.3 (2.9) 27.3 (3.2) 26.6 (2.9) 32.8 (3.5) 
 Age Sex Years since diagnosis Years since surgery Hoehn & Yahr (ON) l-Dopa eq/Kg UPDRS 3
 
 ON OFF right OFF left OFF 
CR 54 24 400 15 19 21 24 
CG 67 17 890 30 35 36 43 
CL 53 16 920 21 33 31 50 
DA 64 720 34 40 38 35.5 
FS 53 15 304 12 15 17 22.5 
ME 64 22 920 28 36 34 46 
SA 61 17 1050 21 36 35 33 
ScA 64 23 150 12 15.5 19 22 
SG 55 22 640 14 12 18 
TN 66 16 640 14 30 23 34 
Mean (SEM) 60.1 (1.8)  17.5 (1.6) 3.5 (0.4) 1.4 (0.2) 699.2 (101.3) 19.3 (2.9) 27.3 (3.2) 26.6 (2.9) 32.8 (3.5) 

Note: For each patient sex, age, years since diagnosis, years since the implantation of the second DBS electrode, Hoehn & Yahr scores (indicating the relative level of disability due to PD disease) in the “ON” phase, L-dopa equivalents per kilogram and the UPDRS-3 in each DBS condition are given. SEM, standard error of the mean.

All participants were right-handed, gave their informed consent and were free to withdraw from the study at any time. The general procedures were approved by the local Institutional Ethics Committee and were performed in accordance with the ethical standards laid down in the Declaration of Helsinki of 1964.

Surgical Procedure

DBS macroelectrodes (model 3389; Medtronic Ltd, Minneapolis, MN) were inserted one at a time during 2 separate surgical interventions. During the first intervention, the STN contralateral to the most affected body side was chosen. The second intervention was performed after subsidence of the postsurgical microlesional effect (usually within 6 months after the first). Preoperative 1.5 T magnetic resonance (MR; Signa, General Electrics, Milwaukee, WI) with a volumetric T1 postcontrast sequence and with axial and coronal T2 Fast Spin Echo sequences (2-mm slices, no gaps) was acquired. Dopaminergic medications were withdrawn in order to perform the surgery in the off-state. After placement of the stereotactic frame (Radionics, Burlington, MA) under local anesthesia, a head computerized tomography (CT; 0.625-mm contiguous axial cuts without gantry tilting) with contrast was acquired. CT and MR were then fused on a stereotactic workstation (Stealth Station, Medtronic) to provide a direct visualization of the STN while correcting MR geometric distortion. DBS electrodes were then implanted via a penetration trajectory computed to reach the dorsolateral STN. Macrostimulation was performed after DBS placement to assess therapeutic and side effects. DBS electrodes had 4 platinum–iridium cylindrical surfaces (1.27 mm diameter) with an edge-to-edge spacing of 0.5 mm. Postsurgical reconstruction of the position of the quadripolar lead with respect to STN was performed according to a previously described localization method (Stancanello et al. 2006, 2008). The locations of the active electrode contacts are reported in Table 2.

Table 2

For each patient, the position of active electrodes, amplitude, pulse, and frequency of stimulation are shown separately for the right (R) and the left (L) hemisphere

Patient Active electrodes
 
Amplitude (V)
 
Pulse (μs)
 
Frequency (Hz)
 
 
CR STNd STNv(+), STNd(−)3.2 3.5 60 60 185 185 
CG Zi Zi 2.1 3.8 90 90 185 185 
CL STNd STNv 3.5 3.3 60 60 185 185 
DA Zi STNd(+), Zi(−)3.2 3.5 60 60 185 185 
FS STNd STNv 3.7 3.3 60 60 185 185 
ME STNv Zi 3.6 3.3 60 60 185 185 
SA STNv, STNd STNv, STNv 3.5 3.5 60 60 185 185 
ScA STNd, STNv STNv, SNr 3.6 2.6 60 60 185 185 
SG STNv Zi 3.8 4.3 90 90 185 185 
TN Zi STNv 3.2 60 60 185 185 
Patient Active electrodes
 
Amplitude (V)
 
Pulse (μs)
 
Frequency (Hz)
 
 
CR STNd STNv(+), STNd(−)3.2 3.5 60 60 185 185 
CG Zi Zi 2.1 3.8 90 90 185 185 
CL STNd STNv 3.5 3.3 60 60 185 185 
DA Zi STNd(+), Zi(−)3.2 3.5 60 60 185 185 
FS STNd STNv 3.7 3.3 60 60 185 185 
ME STNv Zi 3.6 3.3 60 60 185 185 
SA STNv, STNd STNv, STNv 3.5 3.5 60 60 185 185 
ScA STNd, STNv STNv, SNr 3.6 2.6 60 60 185 185 
SG STNv Zi 3.8 4.3 90 90 185 185 
TN Zi STNv 3.2 60 60 185 185 

Note: Active electrodes indicated by a star have a bipolar configuration, with cathode (+) and anode (−); all other electrodes have a monopolar configuration. SNr, Substantia Nigra pars reticulata; STNd, dorsal subthalamic nucleus; STNv, ventral subthalamic nucleus; Zi, zona incerta.

Postsurgical DBS programming was individually tailored to the patient and dopaminergic medications were assessed on the basis of the motor response to stimulation.

Experimental Apparatus

Subjects were seated in a darkened and silent room, in front of a 17-inch PC monitor (CRT noninterlaced, refresh rate 75 Hz, 640 × 480 resolution, 32-bit color depth) on which visual stimuli, consisting of red circles (2.434 cd/m2) with a diameter of 2.8 cm against a dark background of uniform luminance (<0.01cd/m2), were presented. The PC monitor was equipped with a touch screen (MicroTouch; sampling rate 200 Hz) for touch-position monitoring. A non-commercial software package, CORTEX (www.cortex.salk.edu), was used to control stimulus presentation and to collect behavioral responses. The temporal arrangements of stimulus presentation were synchronized with the monitor update rate.

Task

Parkinson’s patients performed a reaching version of the countermanding task (Mirabella et al. 2006, 2008, 2009) in 4 experimental conditions: 1) both DBS-OFF, 2) both DBS-ON, 3) right DBS-OFF, and 4) left DBS-OFF. Stimulation conditions were counterbalanced across patients and administered in 2 different experimental sessions occurring on different days. In each condition, patients were required to complete 4 blocks of 60 trials (240 trials); overall, therefore, each subject performed 960 trials. Resting periods were allowed between blocks whenever requested. Before starting the task, about 50 practice trials were given for familiarizing subjects with the apparatus. Age-matched healthy subjects performed a single countermanding block consisting of 240 trials.

The countermanding task consisted of a random mix of 67% of “no-stop trials” and 33% of “stop trials” (Fig. 1B). All trials began with the appearance of a stimulus at the center of the display and subjects had to maintain their right index finger on this position for a variable holding period (500–800 ms). Then, in no-stop trials, the central stimulus disappeared and, simultaneously, a target appeared (go-signal) on the horizontal plane at 8 cm to the right of the central stimulus. Subjects had to perform a speeded reaching movement toward the peripheral target. Stop trials differed from the no-stop trials because at a variable delay (stop-signal delay, SSD) after the presentation of the go-signal the central stimulus reappeared (stop-signal). In this instance, subjects were instructed to inhibit their movements, holding the central stimulus for a period of 400–600 ms. Trials in which subjects successfully withheld the movement were defined as stop-success trials and those in which they moved were defined as stop-failure trials. Auditory feedback was given for correct responses.

The SSD represents the critical dependent variable in this paradigm because stopping becomes increasingly difficult with the lengthening of the SSD. We dynamically changed the length of the SSDs using a staircase procedure (Levitt 1971; Osman et al. 1986, 1990) in order to allow the participants to succeed in canceling the response in about 50% of the stop trials. The SSD duration varied from one stop trial to the next according to the behavioral performance of the subjects: if they succeeded in withholding the response, then the SSD increased by 3 refresh rates (or 39.9 ms); if they failed, SSD decreased by the same amount of time. The staircase started from a SSD of 119.7 ms (9 refresh rates), which pilot experiments showed to be an appropriate delay for quickly obtaining the desired amount of inhibition in stop trials.

To discourage as much as possible participants from slowing down responses, a common strategy adopted in order to make the inhibition on stop trials easier, we verbally informed them about the tracking procedure and that the probability of stopping would approximate to 50%, irrespective of whether they were postponing their response or not. In addition, we set an upper RT limit for no-stop trials: whenever the RTs were longer than 800 ms, no-stop trials were signaled as errors to the subjects and aborted.

In addition, in order to accurately measure simple RTs of reaching arm movements, participants completed one block of 75 “go-only” trials. These trials were similar to no-stop trials, but subjects were aware of the fact that in this task, stop signals could never be presented. In half of the cases, the go-only task was administered before the countermanding task and in the other half, the order was reversed.

Behavioral Data Analysis

For each experimental condition, the corresponding SSRT was estimated, according to the procedure described in detail in Mirabella et al. (2009). Briefly, we exploited the so-called integration method, which allows calculation of the SSRT duration by subtracting the finishing time of the stop process from its starting time (i.e., the SSD value; Logan and Cowan 1984; Logan 1994; Band et al. 2003). Using the mid-run estimate method (Wetherill and Levitt 1965; Wetherill 1966; Levitt 1971), we computed the delay that better corresponds to the time needed for the subject to withhold a response half of the time, the “representative SSD.” This value was calculated as follows. In each session, the sequence of SSDs displayed either runs of increasing values or runs of decreasing values according to the performance of the subject; the representative SSD was estimated by averaging the values corresponding to the midpoint of every second run. The finishing time of the stop process was calculated by integrating the RT distribution of no-stop trials from the onset of the go-signal until the integral equaled the corresponding observed proportion of stop-failure trials (Logan 1994, see Fig. 1C). Finally, the SSRT estimate was obtained by subtracting the finishing time of the stop process from the representative SSD value.

The SSRT obtained using staircase algorithms is very reliable as it is derived from the central part of the RT distribution of no-stop trials and hence is relatively insensitive to violations of the assumptions of the race model (Logan et al. 1997; Band et al. 2003).

Statistics

Repeated measures analyses of variance (ANOVAs) were employed for assessing changes in RT and in SSRT across the 4 DBS stimulation conditions (both DBS-OFF, both DBS-ON, right DBS-OFF, and left DBS-OFF). Mauchley’s test evaluated the sphericity assumption and, where appropriate, correction of the degrees of freedom was made according to the Greenhouse–Geisser procedure. When needed post hoc tests (pairwise comparisons) with Bonferroni correction were employed. Additionally, t-tests were employed to contrast the RT and SSRT of age-matched healthy subjects with the corresponding values obtained in each single DBS stimulation condition.

Results

UPDRS3

We compared the motor symptoms of patients, measured on the UPDRS3 scale (see Table 1), across DBS conditions, in order to find out whether and when a significant improvement took place. To this aim, we ran a one-way repeated measures ANOVA which showed that the UPDRS3 score was significantly lowered only when both DBS were on (F1,73 = 20.99, P < 0.001; post hoc tests pairwise comparisons with Bonferroni correction, Ps<0.005).

Estimation of SSRT in DBS Patients and Healthy Controls

To check whether the staircase algorithm worked equally well in all DBS experimental conditions, we compared the average proportion of stop-failure trials [P(failure)] using a repeated measures ANOVA. No significant differences in the P(failure) values were found (F3,27 = 0.85, P = 0.47), so the staircase algorithm produced similar behavioral outcomes. Across all conditions the average P(failure) was 0.51 ±0.02 (SE; see also Table 3).

Table 3

Summary of behavioral values of arm movements for each DBS condition and for age-matched controls during the countermanding and the go-only sessions

 DBS ON DBS OFF right DBS OFF left DBS OFF Age-matched CTRL 
Representative SSD 238.4 ± 41.9 167.1 ± 30.6 204.3 ± 31.2 157.8 ± 33.8 348.9 ± 36.3 
P(failure) 0.51 ± 0.04 0.53 ± 0.03 0.48 ± 0.02 0.5 ± 0.02 0.48 ± 0.03 
SSRT 230.8 ± 12.4 286.5 ± 17.9 283.1 ± 20.8 293.4 ± 16.1 240.7 ± 6.2 
RT no-stop trials 497.5 ± 36.6 469.9 ± 28 520.3 ± 23.5 471.9 ± 39.2 611 ± 38.6 
RT stop-failure trials 389.6 ± 27.6 370.4 ± 20.9 399.6 ± 16.9 372.5 ± 26.8 498.6 ± 36.4 
RT go-only trials 318.5 ± 20.6 322.3 ± 19.8 349.9 ± 16.5 314.6 ± 19.9 324.1 ± 15.7 
 DBS ON DBS OFF right DBS OFF left DBS OFF Age-matched CTRL 
Representative SSD 238.4 ± 41.9 167.1 ± 30.6 204.3 ± 31.2 157.8 ± 33.8 348.9 ± 36.3 
P(failure) 0.51 ± 0.04 0.53 ± 0.03 0.48 ± 0.02 0.5 ± 0.02 0.48 ± 0.03 
SSRT 230.8 ± 12.4 286.5 ± 17.9 283.1 ± 20.8 293.4 ± 16.1 240.7 ± 6.2 
RT no-stop trials 497.5 ± 36.6 469.9 ± 28 520.3 ± 23.5 471.9 ± 39.2 611 ± 38.6 
RT stop-failure trials 389.6 ± 27.6 370.4 ± 20.9 399.6 ± 16.9 372.5 ± 26.8 498.6 ± 36.4 
RT go-only trials 318.5 ± 20.6 322.3 ± 19.8 349.9 ± 16.5 314.6 ± 19.9 324.1 ± 15.7 

Note: In all cases, the average value across the samples (±SE) is reported (see “Materials and Methods” for further details).

To obtain a reliable estimate of the SSRT, the basic assumption of the race model, that is, the stochastic independence between the go process (the process initiated by the go-signal leading to the execution of the movement) and the stop process (the process initiated by the stop-signal leading to the inhibition of the movement), has to be satisfied (Logan and Cowan 1984; Logan 1994; see also Boucher et al. 2007). Reaching movements were produced in both the no-stop trials and the stop-failure trials, but the latter were initiated because the go process finished before the stop process. Therefore, considering the distribution of the RTs of the no-stop trials, the responses that escape inhibition should be those corresponding to reaching movements that had RTs shorter than the SSD plus the estimated SSRT. Given the above reasoning, the mean RT of stop-failure trials should be shorter than the mean RT of no-stop trials (Logan and Cowan 1984). A 2-way repeated measures ANOVA (factors: DBS condition and RT no-stop trials/stop-failure trials) showed that stop-failure trials were faster than no-stop trials (F1,9 = 81.9, P < 0.0001). No other significant effects were found (DBS condition: F3,27 = 0.86, P = 0.47; interaction: F3,27 = 0.68, P = 0.57). Furthermore, for each subject (10) and for each task condition (4), we found that in all occurrences but 4 (36/40; 90%), the distributions of the RTs of stop-failure trials were different from those of no-stop trials (Kolmogorov–Smirnov test, Ps < 0.05). The average RTs in stop-failure trials were shorter than those in no-stop trials. Overall, these results indicate that our data gave a good estimate of the SSRT.

Figure 2A and Table 3 illustrate one of the core results of the present paper, namely that the SSRT increased whenever any of the DBS was turned off. In fact the SSRT was significantly shorter when both DBS were on than in all other conditions (one-way repeated measures ANOVA, F3,27 = 7.7, P < 0.001; post hoc tests pairwise comparisons with Bonferroni correction, Ps < 0.05).

Figure 2.

SSRT and RT in DBS Parkinson’s patients and age-matched controls. Panel (A) shows the average SSRT (±SE) in age-matched controls and in PD patient in each DBS condition. Panel (B) illustrates the average RT (±SE) of no-stop trials (dark grey) and the average RT (±SE) of go-only trials (light gray) in age-matched controls and in Parkinson’s patients in each DBS condition.

Figure 2.

SSRT and RT in DBS Parkinson’s patients and age-matched controls. Panel (A) shows the average SSRT (±SE) in age-matched controls and in PD patient in each DBS condition. Panel (B) illustrates the average RT (±SE) of no-stop trials (dark grey) and the average RT (±SE) of go-only trials (light gray) in age-matched controls and in Parkinson’s patients in each DBS condition.

Furthermore, the average SSRT estimated in age-matched healthy subjects was not significantly different from the SSRT of Parkinson’s patients when both DBS were on (t-tests, t21 = 0.7, P = 0.45) but it was significantly faster than in all other DBS conditions (t-tests DBS-OFF, t21 = 3.4, P < 0.005; DBS-OFF-right, t21 = 2.9, P < 0.01; and DBS-OFF-left, t21 = 2.2, P < 0.05).

Recently, Hershey et al. (2010) claimed that the individual variability in the effects of STN DBS stimulation on inhibitory control might be due to the location of active electrode contacts. In fact, they found that the stimulation of the ventral but not of the dorsal portion of STN affected response inhibition. Importantly, Hershey et al. (2010) compared the performance of patients when either the dorsal or the ventral contact of one DBS was active. In order to be as faithful as possible to this paradigm, we checked whether the unilateral stimulation of the dorsal versus the ventral portion of STN elicited a different speed of inhibition. In this analysis, we considered one at a time both DBS electrodes across all our patients, and we compared the SSRT recorded when one DBS electrode was on and the active contact was either in the dorsal (n = 10) or in the ventral (n = 7) portion of STN. We excluded those cases in which there was a bipolar configuration and the 2 active contacts were one in the ventral and one in the dorsal STN (n = 3). In addition, we considered contacts in the zona incerta as they were in the dorsal portion of STN and those contacts in the substantia nigra pars reticulata as they were in the ventral portion of STN. We did not find any differences in the lengths of SSRT for the stimulation of contacts placed in the dorsal and in the ventral portion of STN (mean SSRT 287.7 ± 19.3 vs. 289.9 ± 19.66 ms, respectively; t-test, t15 = −0.07, P = 0.9).

Effect of Stop Trials on the RT of No-Stop Trials: The “Delay Strategy”

As it is known that subjects delay their responses when go-trials are intermixed with stop trials (e.g., Mirabella et al. 2006), we compared the RTs of Parkinson’s patients during go-only trials versus those of no-stop trials using a 2-way repeated measures ANOVA (factors: DBS condition and RT no-stop trials/go-only trials). As expected, we found that patients had shorter RTs when executing go-only trials than when executing no-stop trials (see Fig. 2B and Table 3; F1,9 = 69.4, P < 0.0001), but they were equally fast in all DBS conditions (F3,27 = 1.37, P = 0.27). The interaction was not significant (F2.05,18.4 = 0.38, P = 0.69).

As a next step, we compared the RTs of no-stop and of go-only trials of control subjects with the corresponding values recorded at each DBS condition using a 2-way ANOVA mixed model (within-subject factor: RT no-stop trials/go-only trials; between-subjects factor: group [patients in a given DBS condition and controls]). Again the RTs of no-stop trials were always longer than those of go-only trials (see Table 4). Importantly, the lengths of the RTs were not significantly different in the 2 groups; however, the interaction was always significant, revealing that RTs in no-stop trials of controls were significantly longer than those of Parkinson’s patients in any DBS condition (pairwise comparisons, P < 0.05) except in the DBS-OFF left condition (P = 0.07). Conversely, the RTs of go-only trials never differed between the 2 groups.

Table 4

Results of the statistical comparison between the RTs of no-stop and go-only trials of controls with that of PD patients in each DBS conditions

 RTs of no-stop versus go-only trials
 
Group
 
RTs of no-stop/go-only trials × group
 
 Fdf P value η2 Fdf P value η2 Fdf P value η2 
DBS ON versus CTRL F1,21 = 99.6 <0.0001 0.83 F1,21 = 2.8 0.11 0.12 F1,21 = 5.3 <0.05 0.20 
DBS OFF RIGHT versus CTRL F1,21 = 116.9 <0.0001 0.85 F1,21 = 4.3 0.06 0.17 F1,21 = 12.03 <0.01 0.36 
DBS OFF LEFT versus CTRL F1,21 = 123.8 <0.0001 0.85 F1,21 = 1.0 0.32 0.05 F1,21 = 8.06 <0.01 0.28 
DBS OFF versus CTRL F1,21 = 115.9 <0.0001 0.85 F1,21 = 3.8 0.06 0.15 F1,21 = 9.87 <0.01 0.32 
 RTs of no-stop versus go-only trials
 
Group
 
RTs of no-stop/go-only trials × group
 
 Fdf P value η2 Fdf P value η2 Fdf P value η2 
DBS ON versus CTRL F1,21 = 99.6 <0.0001 0.83 F1,21 = 2.8 0.11 0.12 F1,21 = 5.3 <0.05 0.20 
DBS OFF RIGHT versus CTRL F1,21 = 116.9 <0.0001 0.85 F1,21 = 4.3 0.06 0.17 F1,21 = 12.03 <0.01 0.36 
DBS OFF LEFT versus CTRL F1,21 = 123.8 <0.0001 0.85 F1,21 = 1.0 0.32 0.05 F1,21 = 8.06 <0.01 0.28 
DBS OFF versus CTRL F1,21 = 115.9 <0.0001 0.85 F1,21 = 3.8 0.06 0.15 F1,21 = 9.87 <0.01 0.32 

Note: Two-way ANOVA mixed model; within-subjects factor: RTs of no-stop/go-only trials; between-subjects factor: group (patients in a given DBS condition, controls). For each ANOVA, the F value (and the corresponding degrees of freedom, df), the significance value (P value), and the η2 were shown for each factor and for interaction between factors. η2 is the partial eta-squared, and values >0.14 indicate strong effect sizes, namely that the F values obtained are unlikely to depend on the sample size.

Effect of Stop Trials on the RT of No-Stop Trials: The “Procrastination Strategy”

In healthy subjects, the occurrence of stop trials induces a “procrastination strategy,” that is, they affect the RTs of responses produced in the immediately subsequent no-stop trials (Verbruggen and Logan 2009; Zandbelt and Vink 2010; Jahfari et al. 2009; Mirabella et al. 2006). We tested whether 1) controls and DBS Parkinson’s patients exhibit a procrastination strategy, 2) DBS stimulation affects it, and 3) there is any difference between control subjects and patients with this respect.

The results of these analyses are illustrated in Figure 3. Figure 3A shows the average RT of reaching movements of the 3 no-stop trials immediately after a stop-success trial for both age-matched controls and Parkinson’s patients in each DBS conditions. In controls, the RTs of the first no-stop trial was significantly longer than the following 2 no-stop trials (one-way repeated measures ANOVA with factor RT of the 3 no-stop trials after a stop-success trial: F2,24 = 12.05, P < 0.0001, post hoc test pairwise comparisons with Bonferroni correction). A 2-way repeated measures ANOVA (factors: DBS condition and RT of the 3 no-stop trials after a stop-success trial) showed that in DBS Parkinson’s patients the first no-stop trial was significantly slower than the following 2 no-stop trials (F1.2,11.3 = 16.2, P < 0.001; post hoc test pairwise comparisons with Bonferroni correction). However, there was no difference among the DBS conditions (F3,27 = 2.03, P = 0.13) and no interaction (F6,54 = 1.6, P = 0.16).

Figure 3.

Procrastination strategy in DBS Parkinson’s patients and age-matched controls Panel (A) illustrates the effect of a stop signal on the RTs of 3 consecutive no-stop trials in age-matched controls and in each DBS condition. Dots represent the mean RT (±standard error of the mean) of no-stop trials. The dotted line represents the mean RT of no-stop trials for age-matched controls. Panel (B) shows the percentage of age-matched controls and PD patients in each DBS condition exhibiting the procrastination strategy.

Figure 3.

Procrastination strategy in DBS Parkinson’s patients and age-matched controls Panel (A) illustrates the effect of a stop signal on the RTs of 3 consecutive no-stop trials in age-matched controls and in each DBS condition. Dots represent the mean RT (±standard error of the mean) of no-stop trials. The dotted line represents the mean RT of no-stop trials for age-matched controls. Panel (B) shows the percentage of age-matched controls and PD patients in each DBS condition exhibiting the procrastination strategy.

Finally, we compared the procrastination strategy of controls with that of patients in each DBS condition using a 2-way repeated measures ANOVA (between-subjects factor: group [patients in a given DBS condition, controls]; within-subject factors: RT of the 3 no-stop trials after a stop-success trial). In all the 4 ANOVAs, the within-subject factor was always significant (see Table 5). Post hoc analysis indicated that the RT of the first trial after a stop-success was the slowest (pairwise comparisons with Bonferroni correction, Ps < 0.05). More importantly, patients in the DBS-OFF right condition and in the DBS-OFF condition were significantly faster than controls.

Table 5

Results of the statistical comparison of the procrastination strategy of controls with that of PD patients in each DBS conditions

 RTs of the 3 no-stop trials after a stop-success trial
 
Group
 
RTs of the 3 no-stop trials after a stop-success trial × group
 
 Fdf P value η2 Fdf P value η2 Fdf P value η2 
DBS ON versus CTRL F2,42 = 29.9 <0.0001 0.59 F1,21 = 2.3 0.14 0.10 F2,42 =1.5 0.22 0.07 
DBS OFF RIGHT versus CTRL F1.5,32.2 = 31.1 <0.0001 0.59 F1,21 = 5.7 <0.05 0.21 F1.5,32.2 = 4.3 <0.05 0.17 
DBS OFF LEFT versus CTRL F1.3,28.1 = 14.7 <0.0001 0.41 F1,21 = 1.4 0.25 0.06 F1.3,28.1 = 0.5 0.54 0.02 
DBS OFF vs CTRL F2,42 = 22.7 <0.0001 0.52 F1,21 = 5.8 <0.05 0.22 F2,42 = 0.4 0.69 0.02 
 RTs of the 3 no-stop trials after a stop-success trial
 
Group
 
RTs of the 3 no-stop trials after a stop-success trial × group
 
 Fdf P value η2 Fdf P value η2 Fdf P value η2 
DBS ON versus CTRL F2,42 = 29.9 <0.0001 0.59 F1,21 = 2.3 0.14 0.10 F2,42 =1.5 0.22 0.07 
DBS OFF RIGHT versus CTRL F1.5,32.2 = 31.1 <0.0001 0.59 F1,21 = 5.7 <0.05 0.21 F1.5,32.2 = 4.3 <0.05 0.17 
DBS OFF LEFT versus CTRL F1.3,28.1 = 14.7 <0.0001 0.41 F1,21 = 1.4 0.25 0.06 F1.3,28.1 = 0.5 0.54 0.02 
DBS OFF vs CTRL F2,42 = 22.7 <0.0001 0.52 F1,21 = 5.8 <0.05 0.22 F2,42 = 0.4 0.69 0.02 

Note: Two-way ANOVA mixed model: within-subjects factor, RTs of the 3 no-stop trials after a stop-success trial; between-subjects factor, group (patients in a given DBS condition, controls). For each ANOVA, the F value (and the corresponding degrees of freedom, df), the significance value (P value), and the η2 were shown for each factor and for interaction between factors. η2 is the partial eta-squared and values >0.14 indicate strong effect sizes, namely that the F values obtained are unlikely to depend on the sample size.

In order to see how many subjects individually showed the procrastination strategy, we compared the RTs of the 3 no-stop trials after a stop-success trial with a one-way ANOVA and we counted the number of times in which a main effect occurred and the first no-stop trial was slower than the others. As can be seen in Figure 3B, the majority of controls (∼69%) showed a consistent procrastination strategy. On the other hand, patients exhibited this behavior less frequently, especially when one or both DBS were off.

Discussion

Changes in Inhibitory Control According to STN Stimulation Conditions

It is known that inhibitory control is impaired in Parkinson’s patients (Gauggel et al. 2004; Seiss and Praamstra 2004; van den Wildenberg et al. 2006). Gauggel et al. (2004) showed that patients who were receiving dopaminergic medications suffer from a specific deficit in inhibitory functions. In fact patients had longer SSRTs than controls but RTs of no-stop trials of similar length.

Studies on healthy subjects (Aron and Poldrack 2006; Li et al. 2008) and on Parkinson’s patients (van den Wildenberg et al. 2006; Ray et al. 2009) indicate that STN plays a critical role in this executive function. Our results support this conclusion: we found that, when both DBS were turned on, the speed of inhibitory responses markedly improved, that is, the SSRT shortened. These results fit with those of van den Wildenberg et al. (2006) and of Swann et al. (2011) but not with those of Ray et al. (2009). Several methodological differences can account for this divergence. The most important one is that Ray et al. (2009) allowed patients to rest for only 10–15 min after any stimulator was switched off or on. This procedure could lead to an unsteady motor status (Lopiano et al. 2003; Temperli et al. 2003; Sturman et al. 2008). Another possibility for explaining inconsistencies comes from the location of active contacts in the STN. Recently, Hershey et al. (2010) found that the stimulation of electrode contacts placed in the ventral STN, but not of those in the dorsal STN, impairs inhibitory control. However, by comparing the SSRT obtained when the active contact was in the dorsal portion of STN with respect to when it was in the ventral portion of STN, we did not find any difference. Many differences in the experimental protocols could explain this discrepancy. First of all, Hershey et al. (2010) compared the behavioral performance of the same individual after the stimulation of the ventral and the dorsal portion of STN, while we could only make these comparisons across participants. Second, Hershey et al. (2010) used a Go/No-Go paradigm, and in this procedure, it is a potential movement and not an ongoing response that has to be halted. Further studies are needed to clarify this point.

Once it is established that bilateral STN DBS improves the inhibitory control one could speculate that if the right STN plays a key role in inhibition, a shortening of the SSRT should occur even when only the right stimulator is active. However, we found that when either the right or the left DBS was switched off, the SSRT became as long as when both stimulators were off. This is in contrast with the idea that the right STN plays a key role in inhibition (Aron and Poldrack 2006), rather indicating that the 2 DBS have a synergistic effect. This difference could be explained in several ways. First of all, we used a visual stop signal while Aron and Poldrack (2006) employed an auditory stop signal. Second, subjects in the latter study had to inhibit key press movements, while in the present work, participants had to inhibit reaching movements. Third, our results should be evaluated taking into account the possibility that either PD and/or the STN stimulation might have induced subtle alterations with respect to the circuitry of healthy subjects. For instance, the DBS placement into the STN and the subsequent chronic stimulation might have altered the circuitry by inducing some form of brain plasticity. Also, the placement of a DBS electrode might be responsible for severing fiber bundles such as the pallidothalamic tract traveling around the STN as well as other structures encountered along the DBS penetration trajectory. It also has to be stated that in the literature many other investigators did not find signs of lateralization of inhibitory functions (e.g., Li et al. 2008). The issue of lateralization needs further studies.

It has to be remarked that, if confirmed, our findings have clinical relevance indicating that, as far as response inhibition is concerned, it is better to implant DBS bilaterally than unilaterally. However, a caveat should be borne in mind: on average, our patients had had a bilateral implant for 3.5 years. It is possible that after such a long time they had become accustomed to this stimulation pattern so that, in this condition, the unilateral DBS was not longer effective.

A key finding is that the improved inhibitory control could not be explained by a general improvement of task-related motor function. In fact, we found that the DBS stimulation affected the length of the SSRTs but it did not change the length of the RTs of either go-only trials, no-stop trials, or stop-failure trials. This finding is in line with the results of Gauggel et al. (2004) and of Aron and Poldrack (2006), suggesting that the go process activates different brain circuits with respect to those activated by the stop process. van den Wildenberg et al. (2006) showed that STN stimulation improves the RTs of no-stop trials in the countermanding context but does not affect the RTs of go trials during a go/no-go task. The critical difference between the 2 tasks was that, in the countermanding task, subjects could move either to the right or to the left side (2-choice task) whereas in the go/no-go task, they could move just in one direction (simple choice task). van den Wildenberg et al. (2006) concluded that the stimulation probably affects the efficiency of the response selection processes rather than the generation of a movement per se. As we used a simple choice task a response selection process was not involved, so the stimulation would not be expected to decrease the RTs. Other studies have reported different results. For instance, Frank et al. (2007) found that bilateral DBS reduces the RTs during a decision-making task. However, in this case, patients could receive their medications and in the same paper it is shown that drugs deeply affect the RTs (Frank et al. 2007: Fig. 3A). The same reasoning might be applied to the results of Ray et al. (2009). In fact, in conditions more similar to ours, Hershey et al. (2010), using a Go/No-Go task, did not find any difference across DBS conditions for RTs of either correct Go trials or incorrect No-Go trials.

All in all our results indicate that STN DBS does not impair inhibitory functions but it selectively improves them. Therefore, in agreement with the hypothesis advanced by Garcia et al. (2005), it is likely that DBS silences the pathological activity of STN and simultaneously generates a new firing pattern with beneficial effects.

The precise role of the STN in suppression is still a matter of debate. On the one hand, Aron and Poldrack (2006) suggested that the right STN plays a direct role in response inhibition but on the other hand, there is some evidence that the activation of both STN is related to attentional monitoring, that is, to orient selective attention onto the stop signal. Our study could not distinguish between the 2 hypothesized roles of STN and further studies are needed to clarify the issue.

Proactive Control and DBS Treatment on STN

We found 2 aspects of the behavioral strategy that are affected by PD but that do not change with STN stimulation.

Healthy subjects tend to balance the speed on no-stop trials with successful stopping and thus they slow down their responses as much as possible (Mirabella et al. 2006; Verbruggen and Logan 2009; Zandbelt and Vink 2010). We found that this form of proactive control is not affected by DBS but the fact that Parkinson’s patients tend to react faster than age-matched controls indicates that this behavioral strategy is affected by PD.

Similarly, the procrastination strategy (Mirabella et al. 2006; Verbruggen and Logan 2009; Jahfari et al. 2009; Zandbelt and Vink 2010) was also not influenced by STN stimulation but it was less frequently adopted by Parkinson’s patients than by age-matched controls.

It has been shown that medial frontal cortices are implicated in proactive control mechanisms (Brown and Braver 2005; Li et al. 2006; Forstmann et al. 2008; Chen et al. 2010; Domenech and Dreher 2010) and it is known that these regions project to the ventral portion of STN (Krack et al. 2010). According to us, there are 2 possible and not mutually exclusive explanations for the absence of DBS effects on these aspects of the countermanding performance. DBS contacts in our patients might be located too dorsally with respect to the STN regions where this information is elaborated and/or STN might not play a key role in proactive control. It is noteworthy that the RTs of go-only trials were not affected by DBS and moreover they were not different from those of controls. In other words, the RTs, in a context in which proactive control is not required, are not affected by PD.

As there is some evidence that the slowing down of no-stop trials could involve a mechanism similar to that used to halt responses (Jahfari et al. 2009) and that the striatum activation is critical for this function (Zandbelt and Vink 2010), we can conclude that the reduced slowing in patients is due to damage to the striatum and/or to the connected circuitry and this affects the patients’ ability to keep in mind the context in which they operate.

Conclusions

To summarize, we have demonstrated that, in our experimental context, STN DBS selectively improves inhibitory functions as its electrical stimulation significantly shortened the SSRT but did not influence the RTs of no-stop trials. This finding is consistent with the idea that DBS not only shuts off the pathological activity of STN but also imposes a new pattern of activity with beneficial effects. In addition, we have shown that STN seems to be more involved in reactive than in proactive control even though both are affected by PD.

Funding

Italian Ministry of Work, Health and Social Policies (Bando Giovani Ricercatori 2007 to G.M.) and Italian Ministry of University and Research (PRIN n.2008_RBFNLH_005 to G.M.).

We thank all patients for their participation in the study. We also thank one anonymous reviewer for very useful suggestions. Conflict of Interest : None declared

References

Aron
AR
Behrens
TE
Smith
S
Frank
MJ
Poldrack
RA
Triangulating a cognitive control network using diffusion-weighted magnetic resonance imaging (MRI) and functional MRI
J Neurosci
 , 
2007
, vol. 
27
 (pg. 
3743
-
3752
)
Aron
AR
Durston
S
Eagle
DM
Logan
GD
Stinear
CM
Stuphorn
V
Converging evidence for a fronto-basal-ganglia network for inhibitory control of action and cognition
J Neurosci
 , 
2007
, vol. 
27
 (pg. 
11860
-
11864
)
Aron
AR
Fletcher
PC
Bullmore
ET
Sahakian
BJ
Robbins
TW
Stop-signal inhibition disrupted by damage to right inferior frontal gyrus in humans
Nat Neurosci
 , 
2003
, vol. 
6
 (pg. 
115
-
116
)
Aron
AR
Poldrack
RA
Cortical and subcortical contributions to Stop signal response inhibition: role of the subthalamic nucleus
J Neurosci
 , 
2006
, vol. 
26
 (pg. 
2424
-
2433
)
Band
GP
van der Molen
MW
Logan
GD
Horse-race model simulations of the stop-signal procedure
Acta Psychol (Amst)
 , 
2003
, vol. 
112
 (pg. 
105
-
142
)
Benabid
AL
Chabardes
S
Mitrofanis
J
Pollak
P
Deep brain stimulation of the subthalamic nucleus for the treatment of Parkinson's disease
Lancet Neurol
 , 
2009
, vol. 
8
 (pg. 
67
-
81
)
Boucher
L
Stuphorn
V
Logan
GD
Schall
JD
Palmeri
TJ
Stopping eye and hand movements: are the processes independent?
Percept Psychophys
 , 
2007
, vol. 
69
 (pg. 
785
-
801
)
Breit
S
Schulz
JB
Benabid
AL
Deep brain stimulation
Cell Tissue Res
 , 
2004
, vol. 
318
 (pg. 
275
-
288
)
Brown
JW
Braver
TS
Learned predictions of error likelihood in the anterior cingulate cortex
Science
 , 
2005
, vol. 
307
 (pg. 
1118
-
1121
)
Chambers
CD
Bellgrove
MA
Stokes
MG
Henderson
TR
Garavan
H
Robertson
IH
Morris
AP
Mattingley
JB
Executive “brake failure” following deactivation of human frontal lobe
J Cogn Neurosci
 , 
2006
, vol. 
18
 (pg. 
444
-
455
)
Chen
X
Scangos
KW
Stuphorn
V
Supplementary motor area exerts proactive and reactive control of arm movements
J Neurosci
 , 
2010
, vol. 
30
 (pg. 
14657
-
14675
)
Cools
R
Dopaminergic modulation of cognitive function-implications for L-DOPA treatment in Parkinson's disease
Neurosci Biobehav Rev
 , 
2006
, vol. 
30
 (pg. 
1
-
23
)
Domenech
P
Dreher
JC
Decision threshold modulation in the human brain
J Neurosci
 , 
2010
, vol. 
30
 (pg. 
14305
-
14317
)
Forstmann
BU
Dutilh
G
Brown
S
Neumann
J
von Cramon
DY
Ridderinkhof
KR
Wagenmakers
EJ
Striatum and pre-SMA facilitate decision-making under time pressure
Proc Natl Acad Sci U S A
 , 
2008
, vol. 
105
 (pg. 
17538
-
17542
)
Frank
MJ
Samanta
J
Moustafa
AA
Sherman
SJ
Hold your horses: impulsivity, deep brain stimulation, and medication in parkinsonism
Science
 , 
2007
, vol. 
318
 (pg. 
1309
-
1312
)
Garcia
L
D'Alessandro
G
Bioulac
B
Hammond
C
High-frequency stimulation in Parkinson's disease: more or less?
Trends Neurosci
 , 
2005
, vol. 
28
 (pg. 
209
-
216
)
Gauggel
S
Rieger
M
Feghoff
TA
Inhibition of ongoing responses in patients with Parkinson's disease
J Neurol Neurosurg Psychiatry
 , 
2004
, vol. 
75
 (pg. 
539
-
544
)
Hershey
T
Campbell
MC
Videen
TO
Lugar
HM
Weaver
PM
Hartlein
J
Karimi
M
Tabbal
SD
Perlmutter
JS
Mapping Go-No-Go performance within the subthalamic nucleus region
Brain
 , 
2010
, vol. 
133
 (pg. 
3625
-
3634
)
Jahfari
S
Stinear
CM
Claffey
M
Verbruggen
F
Aron
AR
Responding with restraint: what are the neurocognitive mechanisms?
J Cogn Neurosci
 , 
2010
, vol. 
22
 (pg. 
1479
-
1492
)
Krack
P
Hariz
MI
Baunez
C
Guridi
J
Obeso
JA
Deep brain stimulation: from neurology to psychiatry?
Trends Neurosci
 , 
2010
, vol. 
33
 (pg. 
474
-
484
)
Levitt
H
Transformed up-down methods in psychoacoustics
J Acoust Soc Am
 , 
1971
 
49 (suppl2):467–477
Li
CS
Huang
C
Constable
RT
Sinha
R
Imaging response inhibition in a stop-signal task: neural correlates independent of signal monitoring and post-response processing
J Neurosci
 , 
2006
, vol. 
26
 (pg. 
186
-
192
)
Li
CS
Yan
P
Sinha
R
Lee
TW
Subcortical processes of motor response inhibition during a stop signal task
Neuroimage
 , 
2008
, vol. 
41
 (pg. 
1352
-
1363
)
Logan
GD
Dagenbach
D
Carr
TH
On the ability to inhibit thought and action: a users’ guide to the stop signal paradigm
Inhibitory processes in attention, memory and language
 , 
1994
San Diego
Academic Press
(pg. 
189
-
239
)
Logan
GD
Cowan
WB
On the ability to inhibit thought and action: a theory of an act of control
Psychol Rev
 , 
1984
, vol. 
91
 (pg. 
295
-
327
)
Logan
GD
Schachar
RJ
Tannock
R
Impulsivity and inhibitory control
Psychol Sci
 , 
1997
, vol. 
8
 (pg. 
60
-
64
)
Lopiano
L
Torre
E
Benedetti
F
Bergamasco
B
Perozzo
P
Pollo
A
Rizzone
M
Tavella
A
Lanotte
M
Temporal changes in movement time during the switch of the stimulators in Parkinson's disease patients treated by subthalamic nucleus stimulation
Eur Neurol
 , 
2003
, vol. 
50
 (pg. 
94
-
99
)
Mink
JW
The basal ganglia: focused selection and inhibition of competing motor programs
Prog Neurobiol
 , 
1996
, vol. 
50
 (pg. 
381
-
425
)
Mirabella
G
Pani
P
Ferraina
S
Context influences on the preparation and execution of reaching movements
Cogn Neuropsychol
 , 
2008
, vol. 
25
 (pg. 
996
-
1010
)
Mirabella
G
Pani
P
Pare
M
Ferraina
S
Inhibitory control of reaching movements in humans
Exp Brain Res
 , 
2006
, vol. 
174
 (pg. 
240
-
255
)
Mirabella
G
Pani
P
Ferraina
S
The presence of visual gap affects the duration of stopping process
Exp Brain Res
 , 
2009
, vol. 
192
 (pg. 
199
-
209
)
Moro
E
Scerrati
M
Romito
LM
Roselli
R
Tonali
P
Albanese
A
Chronic subthalamic nucleus stimulation reduces medication requirements in Parkinson's disease
Neurology
 , 
1999
, vol. 
53
 (pg. 
85
-
90
)
Obeso
JA
Rodriguez-Oroz
MC
Bintez-Temino
B
Blesa
FJ
Guridi
J
Marin
C
Rodriguez
M
Functional organization of the basal ganglia: therapeutic implications for Parkinson's disease
Mov Disord
 , 
2008
, vol. 
23
 
Suppl 3
(pg. 
S548
-
S559
)
Osman
A
Kornblum
S
Meyer
DE
The point of no return in choice reaction time: controlled and ballistic stages of response preparation
J Exp Psychol Hum Percept Perform
 , 
1986
, vol. 
12
 (pg. 
243
-
258
)
Osman
A
Kornblum
S
Meyer
DE
Does motor programming necessitate response execution?
J Exp Psychol Hum Percept Perform
 , 
1990
, vol. 
16
 (pg. 
183
-
198
)
Perlmutter
JS
Mink
JW
Deep brain stimulation
Annu Rev Neurosci
 , 
2006
, vol. 
29
 (pg. 
229
-
257
)
Ray
NJ
Jenkinson
N
Brittain
J
Holland
P
Joint
C
Nandi
D
Bain
PG
Yousif
N
Green
A
Stein
JS
, et al.  . 
The role of the subthalamic nucleus in response inhibition: evidence from deep brain stimulation for Parkinson's disease
Neuropsychologia
 , 
2009
, vol. 
47
 (pg. 
2828
-
2834
)
Romanelli
P
Esposito
V
Schaal
DW
Heit
G
Somatotopy in the basal ganglia: experimental and clinical evidence for segregated sensorimotor channels
Brain Res Brain Res Rev
 , 
2005
, vol. 
48
 (pg. 
112
-
128
)
Seiss
E
Praamstra
P
The basal ganglia and inhibitory mechanisms in response selection: evidence from subliminal priming of motor responses in Parkinson's disease
Brain
 , 
2004
, vol. 
127
 (pg. 
330
-
339
)
Stancanello
J
Muacevic
A
Sebastiano
F
Modugno
N
Cerveri
P
Ferrigno
G
Uggeri
F
Romanelli
P
3T MRI evaluation of the accuracy of atlas-based subthalamic nucleus identification
Med Phys
 , 
2008
, vol. 
35
 (pg. 
3069
-
3077
)
Stancanello
J
Romanelli
P
Modugno
N
Cerveri
P
Ferrigno
G
Uggeri
F
Cantore
G
Atlas-based identification of targets for functional radiosurgery
Med Phys
 , 
2006
, vol. 
33
 (pg. 
1603
-
1611
)
Sturman
MM
Vaillancourt
DE
Shapiro
MB
Metman
LV
Bakay
RA
Corcos
DM
Effect of short and long term STN stimulation periods on parkinsonian signs
Mov Disord
 , 
2008
, vol. 
23
 (pg. 
866
-
874
)
Swann
N
Poizner
H
Houser
M
Gould
S
Greenhouse
I
Cai
W
Strunk
J
George
J
Aron
AR
Deep brain stimulation of the subthalamic nucleus alters the cortical profile of response inhibition in the beta frequency band: a scalp EEG study in Parkinson's disease
J Neurosci
 , 
2011
, vol. 
31
 (pg. 
5721
-
5729
)
Temperli
P
Ghika
J
Villemure
JG
Burkhard
PR
Bogousslavsky
J
Vingerhoets
FJ
How do parkinsonian signs return after discontinuation of subthalamic DBS?
Neurology
 , 
2003
, vol. 
60
 (pg. 
78
-
81
)
van den Wildenberg
WP
van Boxtel
GJ
van der Molen
MW
Bosch
DA
Speelman
JD
Brunia
CH
Stimulation of the subthalamic region facilitates the selection and inhibition of motor responses in Parkinson's disease
J Cogn Neurosci
 , 
2006
, vol. 
18
 (pg. 
626
-
636
)
Verbruggen
F
Logan
GD
Proactive adjustments of response strategies in the stop-signal paradigm
J Exp Psychol Hum Percept Perform
 , 
2009
 
35: 835–854
Wetherill
GB
Levitt
H
Sequential estimation of points on a psychometric function
Br J Math Stat Psychol
 , 
1965
, vol. 
18
 (pg. 
1
-
10
)
Zandbelt
BB
Vink
M
On the role of the striatum in response inhibition
PLoS One
 , 
2010
 
5:e13848