c o r t e x 5 9 ( 2 0 1 4 ) 1 0 3 e1 1 2

Available online at www.sciencedirect.com

ScienceDirect Journal homepage: www.elsevier.com/locate/cortex

Research report

Induction of a depression-like negativity bias by cathodal transcranial direct current stimulation Larissa Wolkenstein a,1, Monika Zeiller b,1, Philipp Kanske c and Christian Plewnia b,* a

Department of Psychology, Clinical Psychology and Psychotherapy, University of Tu¨bingen, Tu¨bingen, Germany Department of Psychiatry and Psychotherapy, Neurophysiology & Interventional Neuropsychiatry, and Werner Reichardt Centre of Integrative Neuroscience, University of Tu¨bingen, Tu¨bingen, Germany c Max Planck Institute for Human Cognitive and Brain Sciences, Department of Social Neuroscience, Leipzig, Germany b

article info

abstract

Article history:

Cognitive control (CC) over emotional distraction is of particular importance for adaptive

Received 11 March 2014

human behaviour and is associated with activity in the left dorsolateral prefrontal cortex

Reviewed 3 June 2014

(dlPFC). Deficient CC, e.g., presenting as negativity bias, has been suggested to underlie

Revised 6 June 2014

many of the core symptoms of major depression (MD) and is associated with impairments

Accepted 19 July 2014

of dlPFC function. Correspondingly, enhancement of dlPFC activity with anodal trans-

Action editor Andreas Meyer-

cranial direct current stimulation (tDCS) can ameliorate these impairments in patients

Lindenberg

with MD. Here, we tested the hypothesis that a reduction of dlPFC activity by cathodal

Published online 5 August 2014

tDCS induces CC deficits, thus triggering a depression-like negativity bias in healthy subjects. Twenty-eight individuals participated in a double-blinded, balanced randomized

Keywords:

crossover trial of cathodal (1 mA, 20 min) and sham tDCS applied to the left dlPFC. To

Cathodal tDCS

assess CC we conducted a delayed response working memory (DWM) task and an arith-

dlPFC

metic inhibition task (AIT) with pictures of varying valent content (negative, neutral,

Cognitive control

positive) during and immediately after stimulation. Cathodal tDCS led to impaired CC

Negativity bias

specifically over negative material as assessed by reduced response accuracy in the DWM

Extra-cephalic electrode placement

and prolonged response latency in the AIT. Hence, the current study supports the notion that left dlPFC is critically involved in CC over negative material. Together with previously reported beneficial anodal effects, it indicates that the hypoactivation of left dlPFC causes deficits in CC over negative material, which is a possible aetiological mechanism of depression. © 2014 Elsevier Ltd. All rights reserved.

* Corresponding author. Department of Psychiatry and Psychotherapy, University Clinic of Tu¨bingen, Calwerstraße 14, D-72076 Tu¨bingen, Germany. E-mail address: [email protected] (C. Plewnia). 1 LW and MZ contributed equally to this work. http://dx.doi.org/10.1016/j.cortex.2014.07.011 0010-9452/© 2014 Elsevier Ltd. All rights reserved.

104

1.

c o r t e x 5 9 ( 2 0 1 4 ) 1 0 3 e1 1 2

Introduction

Cognitive control (CC) is necessary to maintain goal-directed behaviour in the presence of competing, goal-irrelevant information and requires, for example, inhibition of the processing of previously relevant or goal-irrelevant information. A topographically distributed system with subcomponents in frontal and parietal cortices is involved in CC processes (D'Esposito, Postle, Jonides, & Smith, 1999; Nee, Wager, & John, 2007). An association between dorsolateral prefrontal cortex (dlPFC) activity and enhanced CC over emotional processing (Beer, Knight, & D'Espositio, 2006; Herrington et al., 2005) and reciprocal interconnection between dlPFC and the affective circuitry (Dolcos & McCarthy, 2006; Sheline, Price, Yan, & Mintun, 2010) suggest that reduced frontal activity causes amygdala hyperactivation and thus deficient CC over emotional material (Ochsner & Gross, 2005). Especially the left dlPFC participates in CC in the processing of emotional material (De Raedt et al., 2010; Herrington et al., 2005). An increased recruitment of dlPFC in cognitive tasks including emotional distraction is assumed to reflect its role in counteracting the distraction by increasing task-specific activity (Cromheeke & € nfelder, & Kanske, 2013). Mueller, 2014; Wessa, Heissler, Scho Individuals with major depression (MD) have difficulties in disengaging from processing negative material (Goeleven, de Raedt, Baert, & Koster, 2006; Gotlib, Krasnoperova, Yue, & € nfelder, & Wessa, Joormann, 2004; Kanske, Heissler, Scho 2012). In addition, depressed individuals do not activate dlPFC as efficiently as healthy controls when confronted with distracting negative material (Berman et al., 2011). Various studies using different methods (e.g., PET, fMRI, SPECT, rTMS) provide evidence for a hypoactivity of especially the left dlPFC (Fitzgerald, Laird, Maller, & Daskalakis, 2008; Grimm et al., 2008) and decreased connectivity between dlPFC and amygdala in patients with MD (Siegle, Thompson, Carter, Steinhauer, & Thase, 2007). Thus, it has been proposed that a hypoactivation of the dlPFC contributes to the onset and maintenance of MD (De Raedt & Koster, 2010; Mayberg, 1997). The hypothesis is that impaired CC over emotional interference, which is associated with decreased dlPFC activity, is one source of the negativity bias characteristic for MD (Fales et al., 2008). Research that has investigated the neural substrates of impaired CC processes and negative biases in depression has been mainly correlational. What remains unclear is as to whether a hypoactivation of the dlPFC causes deficits in CC and thereby negativity bias or whether both hypoactivation of the dlPFC as well as CC deficits are epiphenomena of depression. Direct experimental modulation of spontaneous brain activity in healthy individuals could provide crucial evidence regarding this question. Such modulation can be induced by transcranial direct current stimulation (tDCS), which e using specific stimulation configurations (most commonly applied at 1 mA for up to 20 min) e has been shown to facilitate (anodal) or inhibit (cathodal) cortical excitability (Nitsche & Paulus, 2000; Priori, 2003; Wolkenstein & Plewnia, 2013). While it has been shown that tDCS of the left dlPFC modulates functional connectivity in the CC network (Keeser et al., 2011), and that anodal tDCS applied to left dlPFC improves CC over emotional material in MD (Wolkenstein & Plewnia, 2013), no study thus far has

investigated whether cathodal tDCS of the left dlPFC impairs CC in healthy individuals. We therefore examined the effects of cathodal tDCS on CC in healthy subjects. We hypothesized that cathodal tDCS of the left dlPFC impairs CC and thereby evokes a negativity bias in healthy individuals.

2.

Material and methods

2.1.

Participants

Twenty-eight participants were recruited through advertisements posted on the internet. Potential participants first completed a phone screening, after which they were invited for an interview if deemed eligible. A trained interviewer administered the Structured Clinical Interview for DSM-IV Axis I and II (SCID; First, Spitzer, Gibbon, & Williams, 1996). Participants were excluded if they had a current or lifetime psychiatric disorder, were taking psychotropic medication, were left-handers or ambidexters (LQ < 70) as assessed by the Edinburgh Handedness Inventory (Oldfield, 1971), showed clinically relevant depressive symptoms (BDI-II > 13) as assessed by Becks Depression Inventory (Hautzinger, Keller, & Ku¨hner, 2009), or had a verbal IQ of less than 80 as assessed by the Multiple Choice Word Fluency Test (MWT-B; Lehrl, 1992). We further administered the Verbal Learning and Memory Test (VLMT; Helmstaedter, Lendt, & Lux, 2001) to account for verbal memory and the Trail Making Test (TMT; Reitan, 1992) to account for complex attention, motor speed, visual-motor conceptual screening, and executive functions. All participants provided written informed consent. The study was approved by the local ethics committee and was conducted in compliance with the Declaration of Helsinki.

2.2.

Delayed response working memory task (DWM)

The first task we used to assess CC over emotional material was a DWM task (see Fig. 1). The pictures included in this task were taken from the Emo-Pics (Wessa et al., 2010) and were of either negative, neutral, or positive valence. Picture selection was based on normative values (Wessa et al., 2010) and negative and positive pictures were counterbalanced regarding their valence and arousal scores (i.e., arousal: (positive ¼ negative) > neutral; valence: positive > neutral > negative). In each trial, participants first saw a 1000 msec fixation display presented on a computer monitor. This was followed by a set of six letters that was simultaneously presented for 2000 msec. During the following distraction phase, participants saw interfering pictures of either negative, neutral, or positive valence, or a blank slide (control condition), respectively. After the distraction phase, a probe letter was presented for 4000 msec and participants were instructed to indicate as quickly and as accurately as possible whether the probe letter was one of the original six letters presented before. The next trial began after 4000 msec had elapsed in which the probe letter was displayed independently of when the participants responded. Each condition (negative, neutral, positive, control) consisted of 15 trials resulting in 60 trials overall, which were presented in a randomized order. Responses and response latencies were

c o r t e x 5 9 ( 2 0 1 4 ) 1 0 3 e1 1 2

105

Fig. 1 e Study design: Subjects participated in two stimulation sessions separated by at least one day, up to as many as eight days. Within each test session, participants completed the PANAS immediately before and after both stimulation and the Delayed Working Memory Task (DWM) and Arithmetic Inhibition Task (AIT) were completed. The first session began with diagnostic assessments.

recorded. Participants were given 10 practice trials that were excluded from data analysis.

2.3.

Arithmetic inhibition task (AIT)

The second task we used to assess CC over emotional material was the AIT that was developed on the basis of a task used in € nfelder, Bongers, & former studies (Kanske, Heissler, Scho Wessa, 2011; Van Dillen, Heslenfeld, & Koole, 2009; Wessa et al., 2013). Again, the negative, neutral, and positive pictures included in the task were taken from the Emo-Pics (Wessa et al., 2010), were selected based on normative values, and were counterbalanced regarding their valence and arousal scores. Each trial began with the presentation of a 2000 msec fixation display. This was followed by two equations that were presented simultaneously for 6000 msec and were visually separated by a vertical line in the middle of the screen. The two equations were presented upon the background of a picture with either a negative, neutral or positive valence, or conversely upon a blank slide (control condition), respectively, and were formed with two operands including either a subtraction or an addition (e.g., 17 þ 2 vs 23  2). Participants were instructed to indicate as quickly and as accurately as possible, which one of the two equations yielded the higher result. The next trial began either after participants had responded or after 6000 msec had elapsed within which the two equations had been displayed. Again, each condition (negative, neutral, positive, control) consisted of 15 trials resulting in 60 trials overall, which were presented in a randomized order. Responses and response latencies were recorded. Again, participants were given 10 practice trials that were excluded from data analysis.

2.4.

tDCS

Transcranial direct current was delivered continuously by a battery-driven stimulator (NeuroConn GmbH, Ilmenau, Germany) using a pair of saline-soaked 5  7 cm sponge electrodes. To reduce activity of the left dlPFC the cathodal electrode was placed on the scalp over F3 according to the international 10e20 system of electrode placement (Jasper, 1958). The reference electrode was placed on the contralateral deltoid muscle to avoid unwanted effects in the brain which ensures that effects can be traced back exclusively to the cathodal stimulation of the left dlPFC (Priori et al., 2008; Wolkenstein & Plewnia, 2013). Active tDCS was administered for 20 min with a constant current of 1 mA and a linear fadein/fade-out phase of 5 sec. For sham stimulation, same electrode placement was used but the current was only applied for 40 sec at the onset of the sham session and then ramped down, thereby eliciting a transient tingling experience comparable to that elicited by verum stimulation without having any effects in the brain (Gandiga, Hummel, & Cohen, 2006). Predefined codes assigned to either sham or verum stimulation were used to start the stimulator and thus allowed for a double-blind study design. The order of verum and sham tDCS was counterbalanced across participants.

2.5.

Procedure

The current study was designed as a double-blind, balanced randomized, sham-controlled crossover trial. Accordingly, each participant performed two sessions (one cathodal and one sham tDCS session) within eight days separated by an

106

c o r t e x 5 9 ( 2 0 1 4 ) 1 0 3 e1 1 2

inter-session-interval of at least one day (Monte-Silva, Kuo, Liebetanz, Paulus, & Nitsche, 2010). Order of stimulation condition was counterbalanced across subjects. There were no significant correlations between the order of stimulation (cathodal-sham vs shamcathodal) and the dependent variables (all p > .06). On average the inter-session-interval was 4.18 (SD ¼ 2.55) days with over 78% of participants having an inter-sessioninterval of two days or longer. The mean inter-sessioninterval did not differ between the two subgroups built on the basis of the order of stimulation, t(26) ¼ .073, p ¼ .943. During the first session, sociodemographic data were assessed before the interviewer administered the SCID. Afterwards participants were asked to complete the Edinburgh Handedness Inventory, the BDI-II as well as neuropsychological tests (VLMT, MWT-B, TMT). To control for potential effects of the stimulations on the actual mood state of the participants, they completed the Positive and Negative Affect Schedule (PANAS; Krohne, Egloff, Kohlmann, & Tausch, 1996) immediately before and after stimulation and the DWM and AIT tasks were completed. All participants performed both tasks: The DWM started 5 min after the onset of stimulation to reach maximum effects (Nitsche & Paulus, 2000) and took approximately 15 min. The AIT started immediately after the completion of the DWM, specifically after 20 min of stimulation. During the second session, participants again filled out the PANAS immediately before and after the stimulation and completion of the DWM and AIT tasks. The procedure of the current study is illustrated in Fig. 1.

2.6.

3.1. DWM

Correct responses and decision latencies in the

Our main hypotheses predicted stimulation-dependent differences in the DWM as measured by the percentage of correct responses and mean latencies. Specifically, we predicted that participants would display lower accuracy scores and longer latency under cathodal stimulation compared to sham stimulation e especially when distracting pictures had a negative valence. The mean percentages of correct responses as well as the mean latencies in the different conditions are presented in Table 2. Overall error rates were low. The rmANOVA for response accuracy yielded no significant main effect of ‘stimulation’, F(1,27) ¼ .03, p ¼ .858, but a significant main effect of ‘valence’, F(2,54) ¼ 3.39, p ¼ .041, h2 ¼ .112. However, and as predicted, the analysis also yielded a significant interaction of ‘stimulation’ and ‘valence’, F(2,54) ¼ 5.11, p ¼ .009, h2 ¼ .159 (see Fig. 2A). Follow-up analyses revealed a reduced accuracy under cathodal tDCS for negative as compared to neutral, t(27) ¼ 3.42, p ¼ .002, d ¼ .71, and positive pictures, t(27) ¼ 2.686, p ¼ .012,

Statistical analyses

Data analysis was supported by SPSS 21.0. Mean values of answer accuracy and latencies were calculated for the participants over test trials (cathodal vs sham stimulation) and distractor valences (negative, neutral, positive) to determine indices of behavioural CC correlates. Hypotheses tests were executed in a two-tailed manner with a ¼ .05. We conducted four repeated-measures analyses of variance (rmANOVA) with ‘stimulation’ (cathodal vs sham) as the first withinsubject factor and ‘valence of the picture’ (negative, neutral, positive) as the second within-subject factor to examine stimulation effects on response accuracy and response latency in the DWM and AIT. Post-hoc paired t-tests were conducted to specify differences between conditions. We provide effect sizes (Cohen's d) for all significant post-hoc tests. To control for changes in the actual mood state (measured with the PANAS) due to the stimulation, we conducted a rmANOVA with the within-subject-factors ‘stimulation’ (cathodal vs sham), ‘measurement’ (pre vs post stimulation and task completion) and ‘affective state’ (positive vs negative subscale of the PANAS).

3.

measured by the number of correct responses and the response latency in the control condition of the CC tasks (blank slide, no distractor). Subjects did not report meaningful discomfort or adverse effects during or after stimulation. Only a minor tingling sensation beneath the stimulation electrode was described, predominantly at the beginning of stimulation. In addition, two participants exhibited a minor skin irritation at the anodal stimulation site.

Results

Demographic and neuropsychological characteristics of the participants are presented in Table 1. Table 1 further displays descriptive statistics of the participants' ability to calculate as

Table 1 e Demographic, neuropsychological and clinical characteristics of participants. Characteristic Gender (female) University entrance diploma (yes) Age TMT-A (s) TMT-B (s) TMT-B/TMT-A MWT-B IQ VLMT e immediate recall VLMT e delayed recall BDI-II DWM e accuracy in control condition DWM e response latency control condition AIT e accuracy in control condition AIT e response latency control condition

%

M

SD

Range

30.86 23.25 49.04 2.14 118.64 13.07 12.96 2.21 .94

10.18 6.67 15.46 .44 17.15 2.05 2.15 2.30 .06

19.00e54.00 13.00e40.00 23.00e100.00 1.41e3.13 94.00e145.00 9.00e15.00 8.00e15.00 .00e8.00 .80e1.00

1.15

.28

.60e1.77

.87

.11

.60e1.00

2.78

.64

1.81e4.15

71.40 75.00

Notes. TMT-A ¼ time to complete the Trail Making Test part A; TMT-B ¼ time to complete the Trail Making Test part B; MWT-B IQ ¼ Intelligence Quotient assessed with the MWT-B; VLMT ¼ Verbal Learning and Memory Test; BDI-II ¼ extent of current depressive symptoms as assessed with the BDI-II; DWM ¼ Delayed Response Working Memory Task; AIT ¼ Arithmetic Inhibition Task.

107

c o r t e x 5 9 ( 2 0 1 4 ) 1 0 3 e1 1 2

Table 2 e Descriptive statistics for answer accuracy (in %/100) and response latencies (in sec) in DWM and AIT. DWM sham

Accuracy negative neutral positive Response latency negative neutral positive

DWM cathodal

AIT sham

AIT cathodal

M

SD

M

SD

M

SD

M

SD

.919 .912 .933

.078 .098 .063

.886 .948 .938

.090 .084 .070

.900 .897 .896

.108 .112 .112

.881 .895 .919

.122 .095 .114

1.255 1.202 1.211

.278 .277 .291

1.327 1.265 1.288

.586 .514 .496

2.743 2.753 2.810

.619 .609 .632

2.928 2.817 2.802

.761 .557 .644

d ¼ .64. The accuracy rates of neutral and positive pictures, however, did not significantly differ from each other, t(27) ¼ .51, p ¼ .615. In contrast, under sham tDCS there were no differences between the answer accuracy rates in the three valence conditions of distracting pictures (all p > .10). Even though the accuracy in the negative condition relative to the neutral and positive condition was significantly impaired under cathodal stimulation we did not find significant differences between the accuracy rates under cathodal compared to sham stimulation (all p  .08). The rmANOVA for response latencies yielded neither a significant effect of ‘stimulation’, F(1,27) ¼ .51, p ¼ .482, nor a significant interaction of ‘stimulation’ and ‘valence’, F(2,54) ¼ .07, p ¼ .936 (see Fig. 2B). However, we found a small significant effect of ‘valence’, F(2,54) ¼ 4.38, p ¼ .017, h2 ¼ .139, which indicated that across both stimulation conditions participants responded significantly faster to positive, t(27) ¼ 2.06, p ¼ .044, d ¼ .12, and to neutral pictures, t(27) ¼ 2.48, p ¼ .020, d ¼ .17, compared to negative pictures. There was no difference between response latencies in the neutral and positive condition, t(27) ¼ .99, p ¼ .331.

The rmANOVA on accuracy revealed neither a significant effect of ‘stimulation’, F(1,27) ¼ .002, p ¼ .961, nor a significant effect of ‘valence’, F(2,54) ¼ 1.06, p ¼ .353. Furthermore, there was no interaction effect between these two factors, F(2,54) ¼ 1.04, p ¼ .361 (see Fig. 3A). The rmANOVA on response latencies yielded neither a significant effect of ‘stimulation’, F(1,27) ¼ 1.17, p ¼ .29, nor a significant effect of ‘valence’, F(2,54) ¼ .91, p ¼ .41. However, we found the expected interaction between ‘stimulation’ and ‘valence’, F(2,54) ¼ 3.42, p ¼ .04, h2 ¼ .112. Whereas response latencies did not differ from each other depending on the valence conditions under sham stimulation (all p > .19), response latencies were significantly longer in the negative condition compared to the neutral, t(27) ¼ 2.00, p ¼ .05, d ¼ .17, and the positive condition, t(27) ¼ 2.13, p ¼ .04, d ¼ .18, under cathodal stimulation. The latencies in the neutral and positive conditions, however, did not significantly differ from each other, t(27) ¼ .29, p ¼ .78. Whereas the difference between sham and cathodal stimulation was significant in the negative condition, t(27) ¼ 2.16, p ¼ .04, d ¼ .27, there were no stimulationdependent differences in the neutral, t(27) ¼ .92, p ¼ .37, or positive condition, t(27) ¼ .08, p ¼ .94. Results are shown in Fig. 3B.

3.2.

3.3.

Correct responses and decision latencies in the AIT

The mean percentages of correct responses as well as the mean latencies in the different conditions are presented in Table 2.

A

tDCS effect on mood state

We conducted a rmANOVA with the within-subject-factors ‘stimulation’ (cathodal vs sham), ‘measurement’ (pre vs post stimulation and task completion) and ‘affective state’

B

Fig. 2 e A, The figure shows the mean answer accuracy in negative, neutral, and positive conditions of the DWM under sham stimulation and cathodal transcranial direct current stimulation. The interaction of ‘stimulation’ and ‘valence’ is significant at p ≤ .01, indicating a significantly reduced accuracy in the negative compared to the neutral and the positive conditions only under cathodal stimulation. Error bars represent standard errors. *p ≤ .05; **p ≤ .01. B, The figure depicts mean response latencies in negative, neutral, and positive conditions of the DWM under sham stimulation and cathodal transcranial direct current stimulation. Error bars represent standard errors.

108

c o r t e x 5 9 ( 2 0 1 4 ) 1 0 3 e1 1 2

A

B

Fig. 3 e A, The figure displays the mean answer accuracy in negative, neutral, and positive conditions of the AIT under sham stimulation and cathodal transcranial direct current stimulation. Error bars represent standard errors. B, The figure shows the mean response latencies in negative, neutral, and positive conditions of the AIT under sham stimulation and cathodal transcranial direct current stimulation. The interaction of ‘stimulation’ and ‘valence’ is significant at p ≤ .05, indicating a significantly longer response latency in the negative compared to the neutral and the positive conditions only under cathodal stimulation. The response latency in the negative condition is significantly longer under cathodal compared to sham stimulation. Error bars represent standard errors. *p ≤ .05.

(positive vs negative subscale of the PANAS). This analysis yielded significant effects of ‘measurement’, F(1,27) ¼ 110.38, p < .001, h2 ¼ .803, and ‘affective state’, F(1,27) ¼ 14.47, p ¼ .001, h2 ¼ .349 and a significant interaction of ‘measurement’ and ‘affective state’, F(1,27) ¼ 35.45, p < .001, h2 ¼ .568. Follow-up tests indicated that independent of whether participants received cathodal or sham stimulation, their positive affect decreased significantly from pre- to post-session, t(27) ¼ 6.49, p > .001, whereas their negative affect increased from pre- to post-session, t(27) ¼ 2.44, p ¼ .022. Means and standard deviations are presented in Table 3.

4.

Discussion

This is the first study to examine whether cathodal tDCS of the left dlPFC impairs CC and thereby evokes a negativity bias in healthy individuals, which is typically described in MD. Our results show that applying cathodal tDCS impairs CC specifically in the processing of negative material, as assessed by DWM and AIT. In both tasks, participants showed a decreased

Table 3 e Positive and negative affect of participants depending on stimulation condition and time of measurement. Characteristic PANAS PANAS PANAS PANAS PANAS PANAS PANAS PANAS

PA, pre-session, sham stimulation PA, post-session, sham stimulation PA, pre-session, cathodal stimulation PA, post-session, cathodal stimulation NA, pre-session, sham stimulation NA, post-session, sham stimulation NA, pre-session, cathodal stimulation NA, post-session, cathodal stimulation

M

SD

28.93 25.04 28.79 27.04 10.82 12.07 10.54 11.07

8.09 8.47 8.71 8.63 1.16 3.69 .88 1.49

Notes. PANAS PA ¼ Positive affect as assessed by the PANAS; PANAS NA ¼ Negative affect as assessed by the PANAS.

performance in the negative compared to the neutral and positive conditions under cathodal but not under sham stimulation. Moreover, the response latency in the negative condition of the AIT was significantly longer under cathodal than under sham stimulation. This is in line with our hypotheses and suggests that hypoactivation of the left dlPFC causes deficits in CC, specifically in the processing of negative material and thereby induces a negativity bias. These findings are consistent with the assumption that hypoactivation of the left dlPFC might be the neurophysiological correlate of difficulties in disengaging from the processing of negative material (De Raedt & Koster, 2010; Fales et al., 2008; Roiser, Elliott, & Sahakian, 2012) and corroborate studies revealing an association between negativity bias and less efficient activation of the left dlPFC in depression (Berman et al., 2011). Our data also support recent cognitive neuropsychological models of depression, suggesting that impaired CC over emotional material plays a crucial role in promoting the development and maintenance of depression by causing negativity biases (Roiser et al., 2012). This assumption is further supported by studies showing that not only MD individuals but also their non-affected first-degree relatives exhibit negative biases (Joormann, Talbot, & Gotlib, 2007; Le Masurier, Cowen, & Harmer, 2007), as well as by studies showing that the risk of recurrence is reduced if patients with recurrent depression attend an attentional bias modification training (Browning, Holmes, Charles, Cowen, & Harmer, 2012). Further evidence for dlPFC hypoactivity and associated CC deficits as critical pathophysiological factors of MD has been provided by showing that antidepressant treatment normalizes deficient dlPFC activity during emotional distraction in MD (Fales et al., 2009) and that CC over emotional material in MD can be ameliorated by anodal tDCS of the left dlPFC (Wolkenstein & Plewnia, 2013). Our finding of decreased CC after cathodal tDCS cannot be generalized to other domains of executive functions that may not be affected (Boggio et al., 2010; Fregni et al., 2005) or even

c o r t e x 5 9 ( 2 0 1 4 ) 1 0 3 e1 1 2

improved by cathodal tDCS (Dockery, Hueckel-Weng, Birbaumer, & Plewnia, 2009; Monti et al., 2008). It has been suggested that inhibitory cathodal effects are domain-specific leading to heterogeneous results in non-motor areas most likely due to complex networks and very specific interactions of inhibitory and excitatory mechanisms (Jacobson, Koslowsky, & Lavidor, 2012). However, the current study clearly demonstrates valence-specificity of the inhibitory cathodal effect and thereby gives reason to believe that not only different classes of cognitive functions might exhibit a different vulnerability to inhibitory cathodal effects but also that the use of different stimulus material is of particular importance. While CC over negative material was significantly reduced under cathodal stimulation, we did not find cathodeinduced cognitiveeemotional interference in the positive or neutral condition. Considering studies demonstrating an asymmetry of information processing with greater propensity of attention allocation to negative- as compared to positivetainted entities (Smith, Cacioppo, Larsen, & Chartrand, 2003; Vaish, Grossmann, & Woodward, 2008), the herein demonstrated higher susceptibility to inhibitory cathodal effects of negative trials suggests that predisposed cognitive biases are more vulnerable to inhibitory cathodal effects in healthy individuals than non-predisposed biases. Whereas within the DWM cathodal tDCS impaired response accuracy but not response latency in the negative condition, within the AIT cathodal tDCS impaired response latency but not response accuracy in the negative condition. These differing patterns may be due to the fact that the DWM was conducted while cathodal tDCS was applied to the left dlPFC whereas the AIT was conducted after the cathodal tDCS had been applied. However, to our knowledge there are no studies examining differing effects of “online” versus “offline” tDCS on response accuracy versus response latency. Another possible reason for differing impact of cathodal tDCS on accuracy and latency lies in different task requirements. Within the DWM, subjects are forced to keep six letters in mind while being distracted with emotional stimuli in a delay period. Given that participants are only requested to check whether the probe letter has been presented just before, the time frame of 4000 msec to respond seems relatively long and time pressure appears not to play an important role. In contrast, within the AIT, participants are confronted with two relatively simple arithmetic tasks and are distracted by pictures presented in the background of the target task, leading to high perceptual load competing with a WM function. As proposed by Mu¨ller, Andersen, and Keil (2008) simultaneously presented, but goal-irrelevant input withdraws processing resources from the demanding primary task, leading to reduced processing speed. The AIT requires the calculation of two mathematical equations and the comparison of their results in a relatively short period of time (6000 msec), resulting in a considerably high cognitive load and time pressure. Even though the AIT as well as the DWM revealed a significant difference between the performance in the negative compared to the neutral and the positive conditions under cathodal stimulation, only the AIT further revealed a significant difference between cathodal and sham stimulation in the negative condition. Besides the just discussed differences in timing and requirements of the two tasks, this might also be

109

due to general performance-fluctuations between sessions. Furthermore, it is known that tDCS effects on cognitive performance can be enhanced by cognitive activity during tDCS (Andrews, Hoy, Enticott, Daskalakis, & Fitzgerald, 2011). While the DWM was performed during tDCS and prior to the AIT, the AIT was conducted immediately after the end of the stimulation and the completion of the DWM. Thus, in contrast to the DWM the AIT might have been influenced by the combination of tDCS and cognitive activity which might have enhanced the tDCS effects on the AIT. Further studies using independent samples would be desirable to reproduce the key findings of the current study. Interestingly, the standard errors for latencies in the DWM task are larger in the active compared to the sham condition. Having in mind that the standard error is a measure of how representative a sample is likely to be of the population (Field, 2005), this might be indicative for the high variability in responsiveness to tDCS across subjects which is assumed to be due to subject specific anatomy, genetic variation and different current flow patterns (Datta, Truong, Minhas, Parra, & Bikson, 2012; Plewnia et al., 2013; Wiethoft, Hamada, & Rothwell, 2014). For adequate interpretation of the results, the sensitivity of the tasks must be considered. Visual inspection of the raw data suggests a ceiling effect. However, even in the context of near optimal performance, deterioration of performance under cathodal tDCS manifests in both tests for the negative condition. Future research with systematic variations of workload (e.g., speeding up procedure) or emotional manipulations (e.g., more diverse and non-social material) in CC tasks is needed to further explore the influence of cathodal tDCS on CC over emotional material (Anticevic, Repovs, & Barch, 2010). Since the assessment of mood state immediately before and after both stimulation and task performance did not reveal any changes that were specific for the cathodal stimulation, indirect effects of a cathodal-induced increase of negative affect can be ruled out as an alternative explanation for the tDCS-induced CC deficit. We assume that the mood changes that were present in both experimental conditions point to an effect of the task performance. This is in line with a study by Nitsche et al. (2012), who also did not find moodaltering effects of cathodal tDCS in healthy subjects. Since we did not ask the participants whether they perceived any difference between verum and sham stimulation, one could put into question whether the blinding in the current study was effective. However, it has recently been shown that naı¨ve as well as experienced subjects are not able to distinguish 1 mA tDCS from sham stimulation (Ambrus et al., 2012). In contrast to the predominantly used supra-orbital reference electrode, the usage of an extra-cephalic position of the anode allowed us to avoid confounding effects to the contralateral frontal cortex. Together with previously demonstrated effects of anodal tDCS on CC with the same reference placement (Wolkenstein & Plewnia, 2013) these data further underline the feasibility of this approach. However, the limited spatial resolution of tDCS must also be considered. The stimulated area is relatively large and the stimulation probably affects multiple neural systems (e.g., more inferiorlateral areas of PFC) with different, possibly interacting behavioural effects. Furthermore, it cannot be ruled out that

110

c o r t e x 5 9 ( 2 0 1 4 ) 1 0 3 e1 1 2

an inhibition of the dlPFC may trans-synaptically modulate the activity of other regions involved in CC that are not located directly under the cathode (Polanı´a, Paulus, Antal, & Nitsche, 2011). As Miller (2000) pointed out, the PFC shows a pattern of connectivity with almost all sensory neocortical and motor systems and various subcortical structures which not only allows for synthesizing a wide range of information needed for complex behaviours but also for a top-down-influence on numerous brain processes. Thus, it might for example be that reducing the activity of the dlPFC results in a heightened activity of the amygdala or changes in other regions that have repeatedly been shown to be involved in the CC of emotion as for example the orbitofrontal cortex or the anterior cingulate cortex (Ochsner & Gross, 2005; Pessoa, McKenna, Gutlerrez, & Ungerleider, 2002). Moreover, tDCS to other parts of the frontoparietal CC network, e.g., the right dlPFC and inferior parietal cortex (Schweizer, Grahn, Hampshire, Mobbs, & Dalgleish, 2013), may also be suitable to modulate CC performance. Future studies particularly combining tDCS with concomitant functional neuroimaging methods are required to disentangle this network. To further elucidate the mechanisms underlying tDCS effects on CC over emotional distraction, it is also important to consider possible interactions of brain stimulation and further variables, as for example the individual base level of activity in the dlPFC related to CC, genetic polymorphisms (e.g., BDNF, COMT, 5-HT) (Plewnia et al., 2013), stress history (Williams et al., 2009), and personality charac~ a-Go  mez, Vidal-Pin ~ eiro, Clemente, Pascualteristics (Pen  s-Faz, 2011). The consideration of these interLeone, & Bartre active effects may not only help to predict tDCS effects but also allow for individualized therapeutic approaches.

5.

Conclusions

In summary, the current study provides new insights into possible causal mechanisms underlying deficits in CC over negative material and thus the development of a negativity bias, which in turn is of aetiological relevance for MD. Combined with previous findings of beneficial effects of anodal tDCS on CC in depressed individuals (Wolkenstein & Plewnia, 2013), our results establish the viability of tDCS for a targeted modulation of CC and point to the possibility that brain stimulation techniques may be helpful in the treatment of disorders related to dysfunctional CC.

Conflict of interest The authors declare no competing financial interests.

Acknowledgements Dr. Plewnia received a research grant from the German Research Council (DFG), Werner Reichardt Centre for Integrative Neuroscience (CIN, PP2011_11). We thank Marjorie Kinney, who kindly assisted with the proof-reading of the manuscript.

references

Ambrus, G. G., Al-Moyed, H., Chaieb, L., Sarp, L., Antal, A., & Paulus, W. (2012). The fade-ineshort stimulationefade out approach to sham tDCSereliable at 1 mA for naı¨ve and experienced subjects, but not investigators. Brain Stimulation, 5, 499e504. Andrews, S. C., Hoy, K. E., Enticott, P. G., Daskalakis, Z. J., & Fitzgerald, P. B. (2011). Improving working memory: the effect of combining cognitive activity and anodal transcranial direct current stimulation to the left dorsolateral prefrontal cortex. Brain Stimulation, 4, 84e89. Anticevic, A., Repovs, G., & Barch, D. M. (2010). Resisting emotional interference: brain regions facilitating working memory performance during negative distraction. Cognitive, Affective & Behavioral Neuroscience, 10, 159e173. Beer, J. S., Knight, R. T., & D'Espositio, M. (2006). Integrating emotion and cognition: the role of the frontal lobes in distinguishing between helpful and hurtful emotion. Psychological Science, 17, 448e453. Berman, M. G., Nee, D. E., Casement, M., Kim, H. S., Deldin, P., Kross, E., et al. (2011). Neural and behavioral effects of interference resolution in depression and rumination. Cognitive and Affective Behavioral Neuroscience, 2011(11), 85e96. Boggio, P. S., Campanha, C., Valasek, C. A., Fecteau, S., PascualLeone, A., & Fregni, F. (2010). Modulation of decision-making in a gambling task in older adults with transcranial direct current stimulation. European Journal of Neuroscience, 31, 593e597. Browning, M., Holmes, E. A., Charles, M., Cowen, P. J., & Harmer, C. J. (2012). Using attentional bias modification as a cognitive vaccine against depression. Biological Psychiatry, 72, 572e579. Cromheeke, S., & Mueller, S. C. (2014). Probing emotional influences on cognitive control: an ALE meta-analysis of cognition emotion interactions. Brain Structure and Function, 219, 995e1008. Datta, A., Truong, D., Minhas, P., Parra, L. C., & Bikson, M. (2012). Inter-individual variation during transcranial direct current stimulation and normalization of dose using MRI-derived computational models. Frontiers in Psychiatry, 3, 1e8. De Raedt, R., & Koster, E. H. W. (2010). Understanding vulnerability for depression from a cognitive neuroscience perspective: a reappraisal of attentional factors and a new conceptual framework. Cognitive, Affective, and Behavioral Neuroscience, 10, 50e70. De Raedt, R., Leyman, L., Baeken, C., Van Schuerbeek, P., Luypaert, R., Vanderhasselt, M.-A., et al. (2010). Neurocognitive effects of HF-rTMS over dorsolateral prefrontal cortex on the attentional processing of emotional information in healthy women: an event-related fMRI study. Biological Psychiatry, 85, 487e495. D'Esposito, M., Postle, B. R., Jonides, J., & Smith, E. (1999). The neural substrate and temporal dynamics of interference effects in working memory as revealed by event-related functional MRI. Proceedings of the National Acadamy of Sciences of the United States of America, 96, 7514e7519. Dockery, C. A., Hueckel-Weng, R., Birbaumer, N., & Plewnia, C. (2009). Enhancement of planning ability by transcranial direct current stimulation. The Journal of Neuroscience, 29, 7271e7277. Dolcos, F., & McCarthy, G. (2006). Brain systems mediating cognitive interference by emotional distraction. The Journal of Neuroscience, 26(7), 2072e2079. Fales, C. L., Barch, D. M., Rundle, M. M., Mintun, M. A., Mathews, J., Snyder, A. Z., et al. (2009). Antidepressant treatment normalizes hypoactivity in dorsolateral prefrontal cortex

c o r t e x 5 9 ( 2 0 1 4 ) 1 0 3 e1 1 2

during emotional interference processing in major depression. Journal of Affective Disorders, 112, 206e211. Fales, C. L., Barch, D. M., Rundle, M. M., Mintun, M. A., Snyder, A. Z., Cohen, J. D., et al. (2008). Altered emotional interference processing in affective and cognitive-control brain circuitry in major depression. Biological Psychiatry, 63, 377e384. Field, A. (2005). Discovering statistics using SPSS (2nd ed.). London: SAGE Publications Ltd. First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. (1996). Structured clinical interview for DSM-IV. Washington, DC: American Psychiatric Association. Fitzgerald, P. B., Laird, A. R., Maller, J., & Daskalakis, Z. J. (2008). A meta-analytic study of changes in brain activation in depression. Human Brain Mapping, 29(6), 683e695. Fregni, F., Boggio, P. S., Nitsche, M. A., Bermpohl, F., Antal, A., Feredoes, E., et al. (2005). Anodal transcranial direct current stimulation of prefrontal cortex enhances working memory. Experimental Brain Research, 166, 23e30. Gandiga, P. C., Hummel, F. C., & Cohen, L. G. (2006). Transcranial DC stimulation (tDCS): a tool for double-blind sham-controlled clinical studies in brain stimulation. Clinical Neurophysiology, 117, 845e850. Goeleven, E., de Raedt, R., Baert, S., & Koster, E. H. W. (2006). Deficient inhibition of emotional information in depression. Journal of Affective Disorders, 93, 149e157. Gotlib, I. H., Krasnoperova, E., Yue, D. L., & Joormann, J. (2004). Attentional biases for negative interpersonal stimuli in clinical depression. Journal of Abnormal Psychology, 113, 127e135. Grimm, S., Beck, J., Schuepbach, D., Hell, D., Boesiger, P., Bermpohl, F., et al. (2008). Imbalance between left and right dorsolateral prefrontal cortex in major depression is linked to negative emotional judgment: an fMRI study in severe major depressive disorder. Biological Psychiatry, 63, 369e376. Hautzinger, M., Keller, F., & Ku¨hner, C. (2009). BDI-II. BeckDepressions-Inventar. Revision (2nd ed.). Frankfurt: Pearson Assessment. Helmstaedter, C., Lendt, M., & Lux, S. (2001). Verbaler Lern- und €higkeitstest. Go € ttingen: Beltz Test GmbH. Merkfa Herrington, J. D., Mohanty, A., Koven, N. S., Fisher, J. E., Stewart, J. L., Banich, M. T., et al. (2005). Emotion-modulated performance and activity in left dorsolateral prefrontal cortex. Emotion, 5, 200e207. Jacobson, L., Koslowsky, M., & Lavidor, M. (2012). tDCS polarity effects in motor and cognitive domains: a meta-analytical review. Experimental Brain Research, 216, 1e10. Jasper, H. H. (1958). Report of the committee on methods of clinical examination in electroencephalography. Electroencephalography and Clinical Neurophysiology, 10, 370e371. Joormann, J., Talbot, L., & Gotlib, I. H. (2007). Biased processing of emotional information in girls at risk for depression. Journal of Abnormal Psychology, 116, 135e143. € nfelder, S., Bongers, A., & Wessa, M. Kanske, P., Heissler, J., Scho (2011). How to regulate emotion? Neural networks for reappraisal and distraction. Cerebral Cortex, 21, 1379e1388. € nfelder, S., & Wessa, M. (2012). Neural Kanske, P., Heissler, J., Scho correlates of emotion regulation deficits in remitted depression: the influence of regulation strategy, habitual regulation use, and emotional valence. NeuroImage, 61, 686e693. Keeser, D., Meindl, T., Bor, J., Palm, U., Pogarell, O., Mulert, C., et al. (2011). Prefrontal transcranial direct current stimulation changes connectivity of resting-state networks during fMRI. Journal of Neuroscience, 31, 15284e15293. Krohne, H. W., Egloff, B., Kohlmann, C. W., & Tausch, A. (1996). Untersuchungen mit einer deutschen Version der “Positive and Negative Affect Schedule” (PANAS). Diagnostica.

111

Lehrl, S. (1992). Mehrfachwahl-Wortschatz-intelligenztest: MWT-b (2nd ed.). Nu¨rnberg: Perimed-spitta. Le Masurier, M., Cowen, P. J., & Harmer, C. J. (2007). Emotional bias and waking salivary cortisol in relatives of patients with major depression. Psychological Medicine, 37, 403e410. Mayberg, H. S. (1997). Limbic-cortical dysregulation: a proposed model of depression. Journal of Neuropsychiatry and Clinical Neurosciences, 9, 471e481. Miller, E. K. (2000). The prefrontal cortex and cognitive control. Nature Reviews Neuroscience, 1, 59e65. Monte-Silva, K., Kuo, M. F., Liebetanz, D., Paulus, W., & Nitsche, M. A. (2010). Shaping the optimal repetition interval for cathodal transcranial direct current stimulation (tDCS). Journal of Neurophysiology, 103, 1735e1740. Monti, A., Cogiamanian, F., Marceglia, S., Ferrucci, R., MrakicSposta, S., Vergari, M., et al. (2008). Improved naming after transcranial direct current stimulation in aphasia. Journal of Neurology, Neurosurgery and Psychiatry, 79, 451e453. Mu¨ller, M. M., Andersen, S. K., & Keil, A. (2008). Time course of competition for visual processing resources between emotional pictures and foreground task. Cerebral Cortex, 18, 1892e1899. Nee, D. E., Wager, T. D., & John, J. (2007). Interference resolution: Insights from a meta-analysis of neuroimaging tasks. Cognitive, Affective & Behavioral Neuroscience, 7(1), 1e17. Nitsche, M. A., Koschack, J., Pohlers, H., Hullemann, S., Paulus, W., & Happe, S. (2012). Effects of frontal transcranial direct current stimulation on emotional state and processing in healthy humans. Frontiers in Psychiatry, 3, 58. Nitsche, M. A., & Paulus, W. (2000). Excitability changes induced in the human motor cortex by weak transcranial direct current stimulation. Journal of Physiology, 527, 633e639. Ochsner, K. N., & Gross, J. J. (2005). The cognitive control of emotion. Trends in Cognitive Sciences, 9. http://dx.doi.org/ 10.1016/j.tics.2005. Oldfield, R. C. (1971). The assessment and analysis of handedness: the Edinburgh Inventory. Neuropsychologia, 9, 97e113. ~ a-Go  mez, C., Vidal-Pin ~ eiro, D., Clemente, I. C., PascualPen  & Bartre s-Faz, D. (2011). Down-regulation of negative Leone, A., emotional processing by transcranial direct current stimulation: effects of personality characteristics. PLoS One, 6, e22812. Pessoa, L., McKenna, M., Gutlerrez, E., & Ungerleider, L. G. (2002). Neural processing of emotional faces requires attention. Proceedings of the National Academy of Sciences of the United States of America, 99, 11458e11463. € ngst, I., Maurer, B., Giel, K., & Kru¨ger, R. Plewnia, C., Zwissler, B., La (2013). Effects of transcranial direct current stimulation (tDCS) on executive functions: influence of COMT Val/Met polymorphism. Cortex, 49, 1801e1807. Polanı´a, R., Paulus, W., Antal, A., & Nitsche, M. A. (2011). Introducing graph theory to track for neuroplastic alterations in the resting human brain: a transcranial direct current stimulation study. NeuroImage, 54, 2287e2296. Priori, A. (2003). Brain polarization in humans: a reappraisal of an old tool for prolonged non-invasive modulation of brain excitability. Clinical Neurophysiology, 114, 589e595. Priori, A., Mameli, F., Cogiamanian, F., Marceglia, S., Tiriticco, M., Mrakic-Sposta, S., et al. (2008). Lie-specific involvement of dorsolateral prefrontal cortex in deception. Cerebral Cortex, 18, 451e455. Reitan, R. M. (1992). Trail Making Test. Manual for administration and scoring. Tucson, Arizona: Reitan Neuropsychological Laboratory. Roiser, J. P., Elliott, R., & Sahakian, B. J. (2012). Cognitive mechanisms of treatment in depression. Neuropsychopharmacology Reviews, 37, 117e136. Schweizer, S., Grahn, J., Hampshire, A., Mobbs, D., & Dalgleish, T. (2013). Training the emotional brain: Improving affective

112

c o r t e x 5 9 ( 2 0 1 4 ) 1 0 3 e1 1 2

control through emotional working memory training. The Journal of Neuroscience, 33, 5301e5311. Sheline, Y. I., Price, J. L., Yan, Z., & Mintun, M. A. (2010). Restingstate functional MRI in depression unmasks increased connectivity between networks via the dorsal nexus. Proceedings of the National Academy of Sciences of the United States of America, 107, 11020e11025. Siegle, G. J., Thompson, W., Carter, C. S., Steinhauer, S. R., & Thase, M. E. (2007). Increased amygdala and decreased dorsolateral prefrontal BOLD responses in unipolar depression: related and independent features. Biological Psychiatry, 61, 198e209. Smith, N. K., Cacioppo, J. T., Larsen, J. T., & Chartrand, T. L. (2003). May I have your attention, please: electrocortical responses to positive and negative stimuli. Neuropsychologia, 41, 171e183. Vaish, A., Grossmann, T., & Woodward, A. (2008). Not all emotions are created equal: the negativity bias in socialemotional development. Psychological Bulletin, 134, 383e403. Van Dillen, L. F., Heslenfeld, D. J., & Koole, S. L. (2009). Tuning down the emotional brain: an fMRI study of the effects of

cognitive load on the processing of affective images. NeuroImage, 45, 1212e1219. € nfelder, S., & Kanske, P. (2013). GoalWessa, M., Heissler, J., Scho directed behavior under emotional distraction is preserved by enhanced task-specific activation. Social Cognitive and Affective Neuroscience, 8, 305e312. Wessa, M., Kanske, P., Neumeister, P., Bode, K., Heissler, J., & € nfelder, S. (2010). EmoPics: Subjektive und Scho psychophysiologische Evaluation neuen Bildmaterials fu¨r die klinisch-biopsychologische Forschung. Zeitschrift fu¨r Klinische Psychologie und Psychotherapie, 39, 77. Wiethoft, S., Hamada, M., & Rothwell, J. C. (2014). Variability in response to transcranial direct current stimulation of the motor cortex. Brain Stimulation, 7, 468e475. Williams, L. M., Gatt, J. M., Schofield, P. R., Olivieri, G., Peduto, A., & Gordon, E. (2009). “Negativity bias” in risk for depression and anxiety: brain-body fear circuitry correlates, 5-HTT-LPR and early life stress. NeuroImage, 47, 804e814. Wolkenstein, L., & Plewnia, C. (2013). Amelioration of cognitive control in depression by transcranial direct current stimulation. Biological Psychiatry, 73, 646e651.