Subthalamic deep brain stimulation restores automatic response activation and increases susceptibility to impulsive behavior in patients with Parkinson’s disease

Brain and Cognition 87 (2014) 16–21

Contents lists available at ScienceDirect

Brain and Cognition journal homepage: www.elsevier.com/locate/b&c

Subthalamic deep brain stimulation restores automatic response activation and increases susceptibility to impulsive behavior in patients with Parkinson’s disease Franziska Plessow a,b,c,⇑, Rico Fischer a,b, Jens Volkmann d,1, Torsten Schubert a a

Department of Psychology, Humboldt University of Berlin, Germany Department of Psychology, Technische Universität Dresden, Germany Berenson-Allen Center for Noninvasive Brain Stimulation, Department of Neurology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA d Department of Neurology, Christian-Albrechts University, Kiel, Germany b c

a r t i c l e

i n f o

Article history: Accepted 20 February 2014 Available online 15 March 2014 Keywords: Parkinson’s disease Deep brain stimulation Subthalamic nucleus Action control Cognitive control Impulsivity Response inhibition Simon task Spatial compatibility

a b s t r a c t Repeatedly reported deficits of patients with Parkinson’s disease (PD) in selecting an appropriate action in the face of competing response alternatives has led to the conclusion of a basal ganglia (BG) involvement in response selection and impulse control. Despite capacious research, it remains elusive how BG dysfunction affects processes subserving goal-directed behavior. Even more problematically, since PD pathology transcends a BG dysfunction due to dopamine depletion in the nigrostriatal DA system (by also comprising alterations in extrastriatal dopamine availability and other neurotransmitter systems), it is not yet clear which aspects of these deficits are actually caused by BG dysfunction. To address this question, the present study investigated 13 off-medication PD patients with bilateral therapeutic subthalamic deep brain stimulation (DBS) both with and without stimulation (DBSON and DBSOFF, respectively) and 26 healthy controls. All participants performed a task that tests the relation between automatic response impulses and goal-directed action selection. Results show an improvement of automatic response activation under DBSON, increasing the susceptibility to impulsive responses, and a reduced impact of automatic response activation under DBSOFF. We argue that the BG determine the efficiency of the regulation and transmission of stimulus-driven bottom-up response activation required for efficient response selection. Ó 2014 Elsevier Inc. All rights reserved.

1. Introduction Impulse control refers to the ability to successfully refrain from responding to automatic stimulus-triggered response tendencies, when these tendencies interfere with intentional goal-directed actions. Therefore, the study of efficient action control requires an understanding of the interplay between automatic and habitual response impulses and inhibitory control to allow for goal-directed action selection (e.g., Goschke, 2000). Impulse control was found to primarily rely on a cortical network comprising the lateral prefrontal cortex (PFC), anterior cingulate cortex (ACC), and parietal areas (e.g., Bunge, Hazeltine, Scanlon, Rosen, & Gabrieli, 2002). In

⇑ Corresponding author. Present address: Berenson-Allen Center for Noninvasive Brain Stimulation, Department of Neurology, Beth Israel Deaconess Medical Center, Harvard Medical School, 330 Brookline Avenue KS-158, Boston, MA 02215, USA. E-mail address: [email protected] (F. Plessow). 1 Present address: Department of Neurology, Universitätsklinikum Würzburg, Germany. http://dx.doi.org/10.1016/j.bandc.2014.02.009 0278-2626/Ó 2014 Elsevier Inc. All rights reserved.

addition to that, recent approaches highlight the basal ganglia (BG; a complex of subcortical nuclei including striatum, globus pallidus, substantia nigra, and subthalamic nucleus [STN]) as significant contributors to goal-directed response selection processes by closely interacting with these cortical areas (Frank, 2011; Mink, 1996) based on the fronto-striato-thalamocortical circuits (Alexander, DeLong, & Strick, 1986; Jahfari et al., 2011). Evidence for this assumption mainly comes from studies on Parkinson’s disease (PD), using PD as a model of BG dysfunction. PD is primarily characterized by motor symptoms (e.g., tremor, rigidity, akinesia/bradykinesia, postural instability). Additionally, PD patients, compared to age-matched healthy controls, show impairments in action selection in situations with competing response tendencies, for example, in the Stroop task (e.g., Henik, Singh, Beckley, & Rafal, 1993) or Eriksen–Flanker task (Praamstra, Stegeman, Cools, & Horstink, 1998; Wylie, Stout, & Bashore, 2005; Wylie et al., 2009). These results seem to speak in favor of an altered response selection in PD patients and thus, of a BG involvement in response selection processes.

F. Plessow et al. / Brain and Cognition 87 (2014) 16–21

However, typical PD patient-control group designs cannot be conclusive with respect to a causal BG involvement in response selection processes, as pathological changes in PD are not limited to BG dysfunction, for example, due to nigrostriatal dopamine (DA) depletion (e.g., altered extrastriatal DA levels, changes in the cholinergic and serotoninergic systems; Thobois, Jahanshahi, Pinto, Frackowiak, & Limousin-Dowsey, 2004). Moreover, cognitive impairments in PD are assumed to rise from changes in extrastriatal DA (most notably in the PFC) or pathological alterations of other neurotransmitter systems in addition to the observed striatal DA shortage and BG dysfunction (Broussolle et al., 1999; Cools, Stefanova, Barker, Robbins, & Owen, 2002). For PD-related response selection deficits, DA depletion within the lateral PFC and ACC (Cools et al., 2002; Kaasinen et al., 2000) seems to be critical considering the significance of those brain structures for response selection. A method to test the BG involvement in response selection in a more straightforward way is deep brain stimulation (DBS) of the STN, which allows for a systematic and controlled within-subject variation of the BG functioning state. STN DBS represents a neurosurgical treatment method for advanced PD which leads to a normalization of the neural output activity of the BG on the neural level and an improvement of PD-related motor symptoms on the behavioral level (Benabid, Chabardes, Mitrofanis, & Pollak, 2009). By this means, STN DBS provides the unique possibility to switch between a pathological (DBSOFF) and a more physiologically intact (DBSON) state of BG function in a controlled way on a micro-time level (Jahanshahi et al., 2000; Witt et al., 2004). The efficacy of subthalamic DBS for treating motor dysfunctioning in PD patients has been well established and has led to its widespread use in clinical neurology. At the same time, however, the effects of this treatment on other functional domains are still to be fully understood, e.g., alterations of impulsive behavior have been reported and represent a major concern with rather contradictory results (Wylie et al., 2010). Based on this, in the present study we aimed to advance our understanding of how BG dysfunction affects the interplay between automatic and habitual response impulses on one hand, and inhibitory control to allow for goal-directed action selection on the other, by applying DBS to the STN. To this end, we tested medication-free (withdrawn) PD patients that received bilateral stimulation of the STN exclusively. Patients and participants of a healthy control sample performed an interference paradigm (i.e., a spatial-compatibility task) that specifically targets the interplay between two mechanisms of response selection, namely, the processes of fast automatic response activation (i.e., response impulses) and goal-directed action selection (i.e., inhibitory control), respectively (Kornblum, Hasbroucq, & Osman, 1990; Ridderinkhof, 2002b). In the present version of the task, participants responded with left and right key presses to the pointing direction of an arrow presented on the left or right of fixation. Although task-irrelevant, the stimulus location is assumed to automatically activate its spatially corresponding response, leading to automatic response tendencies that facilitate responding when the impulsive response tendency matches the required response (compatible condition), but interfere and need to be inhibited when it activates the competing response (incompatible condition). This typically results in longer response times and less response accuracy in the incompatible compared to the compatible condition denoting the spatial-compatibility effect (SCE). We specifically implemented a (Stroop-like) version of the spatial-compatibility task (Kornblum, Stevens, Whipple, & Requin, 1999). The current arrow stimuli involved by themselves a left-right dimension, which could or could not coincide with the left or right position of the stimulus on the display. The use of these stimuli should enhance the tendency for impulsivity in the first and the needs for inhibitory control in the latter case. In addition, Fischer, Plessow, and Kiesel

17

(2010) have shown that this particular task situation provides reliable effects across different response time (RT) levels which is important, since (1) we expect generally slowed RTs under DBSOFF compared to DBSON and (2) a recent study found DBS interactions to be limited to certain RT bins (Wylie et al., 2010). Taken together, the present design ensured a maximum of the effects of automatic response activation and inhibitory control processes on participants’ performance in the current task; in addition, it ensured that potential effects of DBS manipulations on the balance between automatic response impulses and goal-directed action selection would be observable irrespective of overall response speed. Regarding the effect of STN DBS-based manipulation of the BG functioning state on response selection processes, there are several possibilities. First, if especially inhibitory control processes are affected by pathological BG functioning (e.g., Praamstra et al., 1998; Wylie et al., 2005), impaired inhibitory control of competing response activation would be expected under DBSOFF. These less efficient inhibitory control processes under DBSOFF should result in increased susceptibility to impulsive and habitual response tendencies, which would be observable by increased response times/error rates when the automatically (by the stimulus location) activated response does not correspond to the required response (incompatible condition). This predicts a larger SCE under DBSOFF (Hypothesis #1). Alternatively, if especially automatic response activation processes are affected by pathological BG functioning, the efficiency of automatic response activation would be reduced under DBSOFF, resulting in a smaller SCE. The weakened influence of automatic stimulus location-driven response activation would less interfere in the incompatible condition and would reduce the performance advantage in the compatible compared to the incompatible condition compared to a condition with physiologically intact BG functioning (Hypothesis #2). We would therefore predict that the DBSON condition facilitates the efficiency of the transmission of automatic response activation; as a net effect this should result in an increased susceptibility to response capture by pre-potent action impulses and thus, to larger SCE effects for DBSON (Wylie et al., 2010). Finally, in line with studies that did not reveal SCE difference between PD patients and matched healthy controls (e.g., Praamstra & Plat, 2001; Wylie et al., 2009), it is also conceivable that DBS manipulation does not specifically affect the response selection mechanism or any specific part of it, yielding to a similar SCE in both DBSON and DBSOFF (Hypothesis #3). 2. Methods 2.1. Participants Thirteen patients with advanced idiopathic Parkinson’s disease and treatment with bilateral STN DBS (Medtronic 3389, Minneapolis, USA) and 26 age-matched healthy controls took part in this study (see Table 1 for details). All participants gave their written informed consent prior to their inclusion in the study in accordance with the Declaration of Helsinki and the guidelines of the local ethics committee. None of the patients reported any other neurological or psychiatric conditions. All participants had normal or corrected-to-normal vision and were studied exclusively without antiparkinsonian medication (in case of regular treatment with antiparkinsonian drugs after an overnight withdrawal). 2.2. Procedure and design Stimuli were presented white against black on a 17-inch monitor using Experimental Run Time System (www.berisoft.com, BeriSoft Corporation). Viewing distance was approximately 60 cm.

F. Plessow et al. / Brain and Cognition 87 (2014) 16–21

Patients

Table 1 Demographics for all subjects and clinical characteristics for patients with Parkinson’s disease.

Highest education Secondary school Vocational training University Clinical characteristics Time since Disease onset (years)a Surgery (months) UPDRS-3 motor scoreb DBSON DBSOFF

Healthy controls (n = 26)

64.31 (6.59) 11

65.08 (8.14) 13

2 7 4

3 15 8

800

Incompatible Compatible

700

RT (ms)

Demographics Age (years) Males (n)

Patients with Parkinson’s disease (n = 13)

Controls

Error (%)

18

600 500 400 300

17.00 (6.03) 20.31 (14.23)

– –

16.31 (8.38) 38.15 (17.88)

– –

Note: Means (standard deviation). Abbreviations: UPDRS = Unified Parkinson’s Disease Rating Scale (Fahn, Elton, & Committee, 1987), DBS = Deep brain stimulation. a n = 12. b Scores are provided for off medication.

Each trial started with a central fixation cross (0.81°  0.81°) for 1000 ms, followed by a blank screen for 600 ms. Subsequently, the target display consisting of two geometrical figures (a square of 1.91°  1.91° and a triangle-shaped arrow head of 1.43°  2.77°, pointing to either the left or right; line width 0.29°; total display size: 6.56°  2.77°) was displayed for 200 ms. One figure appeared 2.8 cm left, the other one 2.8 cm right of the screen centre. Participants responded with their index fingers by pressing external response keys (1.0 cm  1.0 cm) fixed to the table surface. They were instructed to press the left key if a left-pointing arrow appeared and vice versa. When the response location (e.g., left key for left-pointing arrow) matches the stimulus location (e.g., arrow presented on the left), this denotes a compatible condition. Conversely, when response location (e.g., left key for left-pointing arrow) mismatches the stimulus location (e.g., arrow presented on the right), this denotes an incompatible condition. The SCE was computed by incompatible conditions minus compatible condition. Participants were informed that both the other geometrical form and the arrow location were completely task-irrelevant. A correct response was followed by a blank screen for 300 ms. A trial was counted as erroneous, when, for example, a right key was pressed to a left-pointing arrow or vice versa. In this case, the word ‘‘falsch’’ (wrong) was provided as feedback instead of the blank screen for 300 ms. In case of no response within 2000 ms, the feedback ‘‘zu langsam’’ (too slow) appeared. After 500 ms the next trial started. The experiment consisted of four blocks with 48 trials each. Each participant conducted half of the trials under DBSON, the other half under DBSOFF. Order of the two experimental conditions was counterbalanced across patients. After a change in DBS state, a break of at least 30 min ensured stable clinical conditions. Additionally, a neurologist validated the clinical stability. Participants in the control group also performed two experimental conditions (i.e., a first and a second half of the experiment). For analysis reasons, for half of the control group the first part of the experiment was declared as DBSON and the second half as DBSOFF. Both parts were separated by a comparable break. For the other half of the control group this assignment was reversed.

0

ON

OFF DBS

ON

OFF

40 30 20 10 0

DBS

Fig. 1. Median response times (RTs; solid lines) and mean error rates (dashed lines) for incompatible and compatible trials as a function of group (patients with Parkinson’s disease in both deep brain stimulation [DBS] on and off vs. healthy controls). For half of the control group, the first part of the experiment was declared as DBSON and the second part as DBSOFF. For the other half this assignment was reversed. Error bars represent standard errors of the mean.

and DBS (DBSON vs. DBSOFF) as well as the between-subject factor group (PD patient vs. healthy control) was conducted on median RTs and percent error, respectively. 3. Results 3.1. RT

2.3. Data analysis

As expected, DBS interacted with group, F(1, 37) = 12.66, p = .001, g2 = .26. That is, DBS affected general performance for patients (DBSON: 577 ms, DBSOFF: 717 ms), F(1, 12) = 6.75, p = .023, g2 = .36, but not for controls, F(1, 25) = 1.00, p = .326, g2 = .04. Most importantly, we found a significant interaction between spatial compatibility, DBS, and group, F(1, 37) = 6.61, p = .014, g2 = .15.2 For patients only, DBS significantly affected the size of the SCE, F(1, 12) = 6.41, p = .026, g2 = .35. Consistent with Hypothesis #2, Post Hoc t-tests revealed a large SCE under DBSON (100 ms), t(12) = 5.04, p < .001, 95% CI [57, 143], which was reduced under DBSOFF (42 ms) where it did not reach statistical significance, t(12) = 1.49, p = .16, 95% CI [ 20, 105] (see Fig. 1 left panel). Controls showed a medium size SCE (68 ms), F(1, 25) = 69.92, p < .001, g2 = .74, that did not differ between DBS conditions, F < 1, and was numerically in between that of patients in the DBSON (100 ms) and DBSOFF (42 ms) conditions (see Fig. 1 right panel). Yet, the difference of the SCE between patients and control approached significance in the DBSON condition, F(1, 37) = 2.86, p = .099, g2 = .07, but not under DBSOFF, F(1, 37) = 0.99, p = .325, g2 = .03. Overall, patients responded slower than controls, F(1, 37) = 4.72, p = .036, g2 = .11, due to increased RTs for patients specifically in the DBSOFF condition (patients: 717 ms, controls: 548 ms), F(1, 37) = 11.23, p = .002, g2 = .23. In the DBSON condition, no RT difference was found between patients (577 ms) and controls (564 ms), F < 1. To further investigate whether the SCE difference between DBSON and DBSOFF in patients depends on the observed change in mean RT level between the two conditions (due to bradykenesia in DBSOFF), we analyzed the vincentized cumulative RTs (Ratcliff, 1979; Schubert, 1999). For that purpose, we computed the quartile values for each participant based on the whole RT distribution and calculated the mean values per condition and group (Fischer et al., 2010; Fischer, Plessow, & Kiesel, 2013).

A mixed-model analysis of variance (ANOVA) with the withinsubject factors spatial compatibility (incompatible vs. compatible)

2 An additional factor Order of Experimental Condition did not lead to significant interactions nor to a main effect, all F’s < 1.

19

F. Plessow et al. / Brain and Cognition 87 (2014) 16–21

Quartiles (%)

100

Patients

75

50

25

100

Quartiles (%)

Separate repeated-measures ANOVAs with the within-subject factors spatial compatibility, DBS, and quartile showed that the effects of DBS on the SCE were not modulated by differences in RT levels neither in the patient, F < 1, nor in the control group, F(3, 75) = 1.11, p = .306, g2 = .04. Furthermore, in a combined mixed-model ANOVA the factors spatial compatibility, DBS, quartile, and group did not interact, F(3, 111) = 1.28, p = .267, g2 = .03 (see Fig. 2). We further analyzed the effect of DBS on trial-to-trial sequential modulations of the SCE. Prior studies have shown that automatic information transmission from stimulus to motor response is strongest in an actual trial n if it follows a compatible trial in n 1 but is usually reduced when following incompatible trials in n 1 (e.g., Fischer, Dreisbach, & Goschke, 2008; Soutschek, Müller, & Schubert, 2012; Stürmer, Leuthold, Soetens, Schröter, & Sommer, 2002; Wühr & Ansorge, 2005). For this, we calculated the SCE in trial n depending on previous trial history (i.e., compatible versus incompatible in trial n-1). Post-hoc RT analyses confirmed that the sequential modulation of the SCE differed between groups according to DBS condition as reflected in the significant interaction between spatial compatibilityn, spatial compatibilityn 1, DBS, and group, F(1, 37) = 4.83, p = .034, g2 = .12. More specifically, whereas for controls, the sequential modulation of the SCE, F(1, 25) = 10.89, p = .003, g2 = .30, was evident irrespective of experimental condition, F(1, 25) = 1.30, p = .265, g2 = .05, for patients the sequential modulation of the SCE was found only for DBSON (following compatible: 123 ms, following incompatible: 32 ms), F(1, 12) = 9.64, p < .01, g2 = .45, but not for DBSOFF (following compatible: 19 ms, following incompatible: 33 ms), F < 1. The latter observation suggests an impaired transmission regulation of response activation under DBSOFF, which is consistent with findings of recent studies (e.g., Praamstra & Plat, 2001; Rustamov et al., 2013).

Controls

75

50 ON compatible ON incompatible OFF compatible OFF incompatible

25 0

400

600

800

1000

1200

1400

RT (ms) Fig. 2. Response time (RT) distribution analysis. Curves are averaged across individual median RTs for incompatible and compatible trials as a function of group (patients with Parkinson’s disease in both deep brain stimulation [DBS] on and off vs. healthy controls). For half of the control group, the first part of the experiment was declared as DBSON and the second part as DBSOFF. For the other half this assignment was reversed.

3.2. Error rates Patients committed significantly more errors (17.43%) than healthy controls (5.83%), F(1, 37) = 6.68, p = .014, g2 = .15, and displayed a larger SCE (incompatible condition: 23.08%, compatible condition: 11.78%) than healthy controls (incompatible condition: 7.81%, compatible condition: 3.85%) , F(1, 37) = 6.85, p = .013, g2 = .16. Mirroring the RT data, there was a significant interaction between DBS, spatial compatibility, and group, F(1, 37) = 5.23, p = .028, g2 = .12, suggesting that the SCE was differentially affected by the DBS condition between patients and controls. Indeed, for patients the SCE in the error data was numerically larger for DBSON (15.06%) compared to DBSOFF (7.53%). Although this result closely mirrors the RT data, in the patient group the interaction between spatial compatibility and DBS only approached statistical significance, F(1, 12) = 3.32, p = .09, g2 = .22 (see also Fig. 1 left panel). Importantly, no main effect of DBS was found, F(1, 12) = 1.53, p = .240, g2 = .11, ruling out assumptions of a speed accuracy trade-off. For controls, as expected, the SCE did not differ between DBS conditions, F < 1 (see Fig. 1 right panel). Finally, in the DBSON condition patients produced more errors by tendency, F(1, 37) = 4.03, p = .052, g2 = .10, and showed a larger SCE than controls, F(1, 37) = 12.04, p = .001, g2 = .25. In the DBSOFF condition, patients produced more errors than controls, F(1, 37) = 7.80, p = .008, g2 = .17. At the same time, no differences in the SCE in error rates were found under DBSOFF, F(1, 37) = 1.26, p = .269, g2 = .03. 4. Discussion The present study aimed at directly addressing whether and how the basal ganglia (BG) are involved in the interplay between automatic and habitual response impulses and inhibitory control.

Healthy controls and medication-free (withdrawn) patients with advanced idiopathic PD and treatment with bilateral DBS of the subthalamic nucleus conducted a spatial-compatibility task under DBSON and DBSOFF. This manipulation of the BG functioning revealed a main effect of DBS on response times (RTs) which reflects the therapeutic effect of DBS on the motor system and parallels earlier findings of our group with DBS of the globus pallidus internus (Schubert et al., 2002). More importantly, DBS modulated the size of the spatial-compatibility effect (SCE). For patients, a clear SCE was found only under DBSON that was reduced under DBSOFF which clearly rules out Hypothesis #3 (no effect of DBS on the SCE). This result is also not in accordance with Hypothesis #1 predicting stronger interference and therefore a larger SCE effect under DBSOFF due to impaired cognitive response inhibition. We refrain from interpreting the reduced SCE under DBSOFF as evidence for improved inhibitory control, as it seems highly unlikely that the more pathological BG functioning state is associated with more efficient response selection. Instead, our findings are in line with Hypothesis #2 which predicted a DBS-induced regulation of the efficiency of automatic information transmission with reduced versus increased efficiency in DBSOFF versus DBSON, respectively. Reduced efficiency of automatic information transmission in DBSOFF consequently decreases the impact of automatic location-based response activation on response selection. Less automatic response activation slows responses in compatible conditions but benefits the identification of the correct response in incompatible conditions reducing the SCE under DBSOFF. It is important to note that the smaller SCE under DBSOFF cannot be explained by the generally slower RT level or to specific blurring due to bradykinesia, since differences between the BG states were not tied to specific tails of the RT distribution.

20

F. Plessow et al. / Brain and Cognition 87 (2014) 16–21

Although we did not find evidence of a statistically reliable SCE under DBSOFF, it should also be noted that at least some numerical value of the SCE seemed to remain. One possibility might be that automatic information transmission under conditions of DBSOFF is only impaired to some extent but not completely blocked. Hypothesis #2 further predicted that DBSON might improve information transmission and may thus re-establish the impact of automatic response activation in the response selection mechanism. Intact automatic response activation facilitates compatible responses but results in increased interference in response incompatible trials. Consequently and in line with this prediction, we found a reliable SCE for patients in DBSON that was larger than in DBSOFF. Moreover, the SCE for patients in DBSON not only exceeded the one in DBSOFF but also appeared to be more pronounced (in error rates and at least by tendency in RTs) than the SCE revealed by healthy controls. This finding fits well with recent observations that PD patients with DBS reveal a higher susceptibility to impulsive response tendencies to habitual and automatized conflicting response alternatives (e.g., Wylie et al., 2010). It can therefore also perfectly account for findings of increased interference susceptibility in a Stroop task under DBSON compared to DBSOFF that have previously been discussed as evidence for impaired response inhibition under DBS (cf. Jahanshahi et al., 2000; Schroeder et al., 2002; Witt et al., 2004). The assumption of an increased efficiency of transmission of automatic response activation for patients in DBSON is also compatible with recent observations of increased impulsivity under DBSON compared to DBSOFF (e.g., Ballanger et al., 2009; Frank, Samanta, Moustafa, & Sherman, 2007; Wylie et al., 2010). At the same time, however, the error data closely mirrored RT data, ruling out the possibility of an unspecific shift in the speed accuracy trade-off. Our results extend previous findings, by providing further novel support of the argumentation of restored automatic information transmission under DBS. As a window into the adaptive regulation of pre-potent response tendencies, we analyzed trial-to-trial sequential modulations of SCE. It has been argued that automatic information transmission and thus, the influence of pre-potent response tendencies towards the stimulus location is typically strongest following compatible trials in n 1 (Stürmer et al., 2002). This pattern has clearly been demonstrated for the healthy controls. For patients, however, sequential modulations of the SCE were exclusively obtained in conditions of DBSON but not DBSOFF. For the latter, we did not find an effect of automatic information transmission following compatible trials in n-1, thus, providing further empirical support for the notion of impaired information transmission in PD patients that can be restored by means of subthalamic DBS. The present demonstration of restored automatic information transmission and increased impulse behavior by DBS is in line with and extends findings of previous studies (e.g., Wylie et al., 2010). Wylie and colleagues found increased susceptibility to pre-potent responses under DBS stimulation for the faster parts of the RT distribution, although no overall effect of DBS on the SCE was found. It should be noted though, that the authors implemented a standard spatial-compatibility task that is generally sensitive to response duration (i.e., smaller SCE for slower responses), which can be explained by either inhibitory control requiring time to develop (Ridderinkhof, 2002a) or alternatively by passive decay of automatic response activation, thus, diminishing the interfering impact over time (Hommel, 1994). In the present study, we particularly applied a spatial-compatibility paradigm that produces an SCE that is not affected by the response speed level (e.g., Fischer et al., 2010)3 3 Arrow stimuli in the present study contain an additional spatial left–right dimension which results in a combined stimulus–stimulus conflict and stimulusresponse conflict (Kornblum et al., 1999).

and that ensures strongest SCE effects because of combined stimulus–stimulus and stimulus–response conflicts. Thus, differences in the design of the spatial-compatibility tasks may explain why the DBS effects in the current study were not limited to the fastest RTs only and appeared across the whole RT distribution. Nevertheless, taken together, both studies suggest that the observation of DBSimproved automatic response activation processes is a replicable phenomenon across studies and that the issue of the extent of this improvement deserves additional specification. Further studies may clarify the possible impact of an interaction between decaydependent vs. decay-independent SCE and DBS manipulations on PD’s RTs in interference tasks (Wylie et al., 2010). An additional factor influencing the impulsivity control may be related to the medication state of the PD patients as the patients in the current study were investigated in medication off and in the Wylie et al. study under normal medication state conditions. Findings of Pattij and Vanderschuren (2008) point to a critical role of the dopaminergic medication for inhibitory control, which may affect the action control system of PD patients on different levels of the involved cortico–subcortical network. Because the current study investigated PD patients without levodopa medication, the current findings are suggestive for a specific involvement of the STN of the BG in the mechanism of response impulse activation, which may be neglected if medicated PD patients are investigated under different DBS conditions (Frank et al., 2007). Our findings have several further implications. First, the present within-subject manipulation of BG functioning provided causal proof of a BG involvement in response-selection processes, supporting previous suggestions of a strong basal ganglia involvement in high-level cognitive control (e.g., van Schouwenburg, den Ouden, & Cools, 2010; van Schouwenburg et al., 2013). Second, we were able to specify more clearly how BG dysfunction alters response selection by showing that particularly the transmission of automatic stimulus-driven response activation is regulated by the BG. Importantly, this assumption does not challenge the view that PD patients might also show deficits in inhibitory control (e.g., Praamstra et al., 1998; Wylie et al., 2005). It is conceivable, for example, that strongly reduced transmission of automatic response activation due to BG dysfunctioning also reduces the need for subsequent inhibitory control primarily revealed by frontal cortical areas. Finally, our results are also in line with current anatomical and functional BG models that postulate a close link between the BG and frontal cortical areas via multiple BG-thalamocortical circuits and, based on that, BG contribution to motor-program selection (Alexander et al., 1986; Mink, 1996). Acknowledgment The authors like to thank Stefanie Richter for assistance in control group data collection. References Alexander, G. E., DeLong, M. R., & Strick, P. L. (1986). Parallel organization of functionally segregated circuits linking basal ganglia and cortex. Annual Review of Neuroscience, 9, 357–381. http://dx.doi.org/10.1146/annurev.ne.09.030186. 002041. Ballanger, B., van Eimeren, T., Moro, E., Lozano, A. M., Hamani, C., Boulinguez, P., et al. (2009). Stimulation of the subthalamic nucleus and impulsivity: Release your horses. Annals of Neurology, 66(6), 817–824. http://dx.doi.org/10.1002/ ana.21795. Benabid, A. L., Chabardes, S., Mitrofanis, J., & Pollak, P. (2009). Deep brain stimulation of the subthalamic nucleus for the treatment of Parkinson’s disease. The Lancet Neurology, 8(1), 67–81. http://dx.doi.org/10.1016/S14744422(08)70291-6. Broussolle, E., Dentresangle, C., Landais, P., Garcia-Larrea, L., Pollak, P., Croisile, B., et al. (1999). The relation of putamen and caudate nucleus 18F-Dopa uptake to motor and cognitive performances in Parkinson’s disease. Journal of the Neurological Sciences, 166(2), 141–151.

F. Plessow et al. / Brain and Cognition 87 (2014) 16–21 Bunge, S. A., Hazeltine, E., Scanlon, M. D., Rosen, A. C., & Gabrieli, J. D. (2002). Dissociable contributions of prefrontal and parietal cortices to response selection. Neuroimage, 17(3), 1562–1571. Cools, R., Stefanova, E., Barker, R. A., Robbins, T. W., & Owen, A. M. (2002). Dopaminergic modulation of high-level cognition in Parkinson’s disease: the role of the prefrontal cortex revealed by PET. Brain, 125(Pt. 3), 584–594. Fahn, S., Elton, R. L., & Committee, M. o. t. U. D. (1987). The unified Parkinson’s disease rating scale. In S. Fahn, C. D. Marsden, D. B. Calne, & M. Goldstein (Eds.). Recent developments in Parkinson’s disease (Vol. 2). Macmillan Healthcare Information: Florham Park (pp. 153–163, 293–304). Fischer, R., Dreisbach, G., & Goschke, T. (2008). Context-sensitive adjustments of cognitive control: conflict-adaptation effects are modulated by processing demands of the ongoing task. Journal of Experimental Psychology: Learning, Memory and Cognition, 34(3), 712–718 (doi: 2008-04973-021, pii: 10.1037/ 0278-7393.34.3.712). Fischer, R., Plessow, F., & Kiesel, A. (2010). Auditory warning signals affect mechanisms of response selection: Evidence from a Simon task. Experimental Psychology, 57(2), 89–97 (doi: 626751V61L1XR227, pii: 10.1027/1618-3169/ a000012). Fischer, R., Plessow, F., & Kiesel, A. (2013). The effects of alerting signals in masked priming. Frontiers in Psychology, 4, 448. http://dx.doi.org/10.3389/ fpsyg.2013.00448. Frank, M. J. (2011). Computational models of motivated action selection in corticostriatal circuits. Current Opinion in Neurobiology, 21(3), 381–386. http:// dx.doi.org/10.1016/j.conb.2011.02.013. Frank, M. J., Samanta, J., Moustafa, A. A., & Sherman, S. J. (2007). Hold your horses: Impulsivity, deep brain stimulation, and medication in parkinsonism. Science, 318(5854), 1309–1312. http://dx.doi.org/10.1126/science.1146157. Goschke, T. (2000). Intentional reconfiguration and involuntary persistence in task set switching. In S. Monsell & J. Driver (Eds.), Attention and Performance XVIII: Control of cognitive processes (pp. 331–355). Cambridge, MA: MIT Press. Henik, A., Singh, J., Beckley, D. J., & Rafal, R. D. (1993). Disinhibition of automatic word reading in Parkinson’s disease. Cortex, 29(4), 589–599. Hommel, B. (1994). Spontaneous decay of response-code activation. Psychological Research, 56(4), 261–268. Jahanshahi, M., Ardouin, C. M., Brown, R. G., Rothwell, J. C., Obeso, J., Albanese, A., et al. (2000). The impact of deep brain stimulation on executive function in Parkinson’s disease. Brain, 123(Pt 6), 1142–1154. Jahfari, S., Waldorp, L., van den Wildenberg, W. P., Scholte, H. S., Ridderinkhof, K. R., & Forstmann, B. U. (2011). Effective connectivity reveals important roles for both the hyperdirect (fronto-subthalamic) and the indirect (fronto-striatalpallidal) fronto-basal ganglia pathways during response inhibition. Journal of Neuroscience, 31(18), 6891–6899. http://dx.doi.org/10.1523/JNEUROSCI.525310.2011. Kaasinen, V., Nagren, K., Hietala, J., Oikonen, V., Vilkman, H., Farde, L., et al. (2000). Extrastriatal dopamine D2 and D3 receptors in early and advanced Parkinson’s disease. Neurology, 54(7), 1482–1487. Kornblum, S., Hasbroucq, T., & Osman, A. (1990). Dimensional overlap: Cognitive basis for stimulus-response compatibility – A model and taxonomy. Psychological Review, 97(2), 253–270. Kornblum, S., Stevens, G. T., Whipple, A., & Requin, J. (1999). The effects of irrelevant stimuli: 1. The time course of stimulus-stimulus and stimulus-response consistency effects with stroop-like stimuli, simon-like tasks, and their factorial combinations. Journal of Experimental Psychology: Human Perception and Performance, 25(3), 688–714. Mink, J. W. (1996). The basal ganglia: Focused selection and inhibition of competing motor programs. Progress in Neurobiology, 50(4), 381–425. Pattij, T., & Vanderschuren, L. J. (2008). The neuropharmacology of impulse behaviour. Trends in Pharmacological Sciences, 29, 192–199. Praamstra, P., & Plat, F. M. (2001). Failed suppression of direct visuomotor activation in Parkinson’s disease. Journal of Cognitive Neuroscience, 13(1), 31–43. Praamstra, P., Stegeman, D. F., Cools, A. R., & Horstink, M. W. (1998). Reliance on external cues for movement initiation in Parkinson’s disease. Evidence from movement-related potentials. Brain, 121(Pt 1), 167–177.

21

Ratcliff, R. (1979). Group reaction time distributions and an analysis of distribution statistics. Psychological Bulletin, 86(3), 446–461. Ridderinkhof, K. R. (2002b). Micro- and macro-adjustments of task set: Activation and suppression in conflict tasks. Psychological Research, 66(4), 312–323. http:// dx.doi.org/10.1007/s00426-002-0104-7. Ridderinkhof, K. R. (2002a). Activation and suppression in conflict tasks: Empirical clarification through distributional analyses. In W. Prinz & B. Hommel (Eds.), Common Mechanisms in Perception and Action. Attention & Performance XIX. Oxford, UK: Oxford University Press (pp. 494–519). Rustamov, N., Rodriguez-Raecke, R., Timm, L., Agrawal, D., Dressler, D., Schrader, C., et al. (2013). Absence of congruency sequence effects reveals neurocognitive inflexibility in Parkinson’s disease. Neuropsychologia. http://dx.doi.org/10.1016/ j.neuropsychologia.2013.10.025. Schroeder, U., Kuehler, A., Haslinger, B., Erhard, P., Fogel, W., Tronnier, V., et al. (2002). Subthalamic nucleus stimulation affects striato-anterior cingulate cortex circuit in a response conflict task: A PET study. Brain, 125(Pt 9), 1995–2004. Schubert, T. (1999). Processing differences between simple and choice reactions affect bottleneck localization in overlapping tasks. Journal of Experimental Psychology: Human Perception and Performance, 25(2), 408–425. Schubert, T., Volkmann, J., Muller, U., Sturm, V., Voges, J., Freund, H. J., et al. (2002). Effects of pallidal deep brain stimulation and levodopa treatment on reactiontime performance in Parkinson’s disease. Experimental Brain and Research, 144(1), 8–16. http://dx.doi.org/10.1007/s00221-002-1020-1. Soutschek, A., Müller, H. J., & Schubert, T. (2012). Conflict-specific effects of accessory stimuli on cognitive control in the stroop task and the simon task. Experimental Psychology, 1–8 (doi: 3036070P23161K4M, pii: 10.1027/16183169/a000181). Stürmer, B., Leuthold, H., Soetens, E., Schröter, H., & Sommer, W. (2002). Control over location-based response activation in the Simon task: Behavioral and electrophysiological evidence. Journal of Experimental Psychology: Human Perception and Performance, 28(6), 1345–1363. Thobois, S., Jahanshahi, M., Pinto, S., Frackowiak, R., & Limousin-Dowsey, P. (2004). PET and SPECT functional imaging studies in Parkinsonian syndromes: From the lesion to its consequences. Neuroimage, 23(1), 1–16. http://dx.doi.org/10.1016/ j.neuroimage.2004.04.039. van Schouwenburg, M. R., den Ouden, H. E., & Cools, R. (2010). The human basal ganglia modulate frontal–posterior connectivity during attention shifting. Journal of Neuroscience, 30(29), 9910–9918. http://dx.doi.org/10.1523/ JNEUROSCI.1111-10.2010. van Schouwenburg, M. R., Onnink, A. M., Ter Huurne, N., Kan, C. C., Zwiers, M. P., Hoogman, M., et al. (2013). Cognitive flexibility depends on white matter microstructure of the basal ganglia. Neuropsychologia, 53C, 171–177. http:// dx.doi.org/10.1016/j.neuropsychologia.2013.11.015. Witt, K., Pulkowski, U., Herzog, J., Lorenz, D., Hamel, W., Deuschl, G., et al. (2004). Deep brain stimulation of the subthalamic nucleus improves cognitive flexibility but impairs response inhibition in Parkinson disease. Archives of Neurology, 61(5), 697–700. http://dx.doi.org/10.1001/archneur.61.5.697. Wühr, P., & Ansorge, U. (2005). Exploring trial-by-trial modulations of the Simon effect. The Quarterly Journal of Experimental Psychology Section A, 58(4), 705–731. Wylie, S. A., Ridderinkhof, K. R., Elias, W. J., Frysinger, R. C., Bashore, T. R., Downs, K., et al. (2010). Subthalamic nucleus stimulation influences expression and suppression of impulsive behaviour in Parkinson’s disease. Brain, 133(Pt 12), 3611–3624 (doi: awq239 pii: 10.1093/brain/awq239). Wylie, S. A., Stout, J. C., & Bashore, T. R. (2005). Activation of conflicting responses in Parkinson’s disease: Evidence for degrading and facilitating effects on response time. Neuropsychologia, 43(7), 1033–1043. http://dx.doi.org/10.1016/ j.neuropsychologia.2004.10.008. Wylie, S. A., van den Wildenberg, W. P., Ridderinkhof, K. R., Bashore, T. R., Powell, V. D., Manning, C. A., et al. (2009). The effect of Parkinson’s disease on interference control during action selection. Neuropsychologia, 47(1), 145–157. http:// dx.doi.org/10.1016/j.neuropsychologia.2008.08.001.