Electrophysiological evidence of a typical cognitive distortion in bipolar disorder

c o r t e x 6 6 ( 2 0 1 5 ) 1 0 3 e1 1 4

Available online at www.sciencedirect.com

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

Research report

Electrophysiological evidence of a typical cognitive distortion in bipolar disorder Juliane Kopf a,*,1, Julia Volkert b,1, Sarah Heidler b, Thomas Dresler c,d, Sarah Kittel-Schneider a, Alexandra Gessner b, Martin J. Herrmann b, Ann-Christine Ehlis c and Andreas Reif a a

Department of Psychiatry, Psychosomatics and Psychotherapy, University of Frankfurt, Frankfurt, Germany Department of Psychiatry, Psychosomatics and Psychotherapy, University of Wuerzburg, Wuerzburg, Germany c Department of Psychiatry and Psychotherapy, University of Tuebingen, Tuebingen, Germany d LEAD Graduate School, University of Tuebingen, Tuebingen, Germany b

article info

abstract

Article history:

Patients suffering from bipolar disorder often report negative thoughts and a bias to-

Received 3 June 2014

wards negative environmental stimuli. Previous studies show that this mood-congruent

Reviewed 4 September 2014

attentional bias could mediated by dysfunctions in anterior limbic regions. The Error-

Revised 11 September 2014

Related Negativity (ERN), which originates in the anterior cingulate cortex (ACC), has

Accepted 16 February 2015

been used to research this negativity bias in depressed patients, and could also help to

Action editor Andreas Meyer-

better understand the underlying mechanisms causing the negativity bias in bipolar

Lindberg

patients.

Published online 5 March 2015

In this study we investigated error processing in patients with bipolar disorder. Acute depressive bipolar patients (n ¼ 20) and age-matched healthy controls (n ¼ 20) underwent a

Keywords:

modified Eriksen Flanker Task to assess test performance and two error-related event-

Bipolar disorder

related potentials (ERPs), i.e., the ERN and Error Positivity (Pe) were measured by EEG. Half

Error-related negativity

of the patients were measured again in a euthymic state.

Error-related positivity

We found similar ERN amplitudes in bipolar patients as compared to healthy controls,

Flanker task

but significantly reduced Pe amplitudes. Moreover, acutely depressed bipolar patients

Cognitive distortion

displayed an ERN and Pe even if they responded accurately or too slow, which indicates that correct responses are processed in a way similar to wrong responses. This can be interpreted as a psychophysiological correlate of typical cognitive distortions in depression, i.e., an erroneous perception of personal failures. This biased error perception partially remained when patients were in a euthymic state.

Abbreviations: ERN, Error-related negativity; Pe, Error-related Positivity; ACC, anterior cingulate cortex; MADRS, Montgomery Asperger Depression Rating Scale; YMRS, Young Mania Rating Scale. * Corresponding author. Department of Psychiatry, Psychosomatics and Psychotherapy, University of Frankfurt, Heinrich-Hoffmann-Str. 10, D-60528 Frankfurt, Germany. E-mail address: [email protected] (J. Kopf). 1 These authors contributed equally. http://dx.doi.org/10.1016/j.cortex.2015.02.009 0010-9452/© 2015 Elsevier Ltd. All rights reserved.

104

c o r t e x 6 6 ( 2 0 1 5 ) 1 0 3 e1 1 4

Together, our data indicate that aberrant error processing of bipolar patients may be regarded a trait marker possibly reflecting a risk factor for depressive relapses in bipolar disorder. © 2015 Elsevier Ltd. All rights reserved.

1.

Introduction

Bipolar disorder is a severe mental disorder with a variety of affective, cognitive and behavioral symptoms. During acute depressive episodes almost every bipolar patient reports cognitive distortions which consist of irrational and dysfunctional overestimations of failure and blame (Beck, 1963). In contrast, patients have grandiose beliefs and unrealistic perceptions of their own abilities and interpersonal relationships during episodes of manic episodes (Morriss, van der Gucht, Lancaster, & Bentall, 2009; Van der Gucht, Morriss, Lancaster, Kinderman, & Bentall, 2009). In most cases, those symptoms resolve after remission, the bias towards negative experiences present during depressive episodes and the seeking of positive experience during manic episodes ease. However, recent research shows that, for example, errors in reasoning, especially the depressive pattern of thought, sometimes may persist even in euthymic bipolar patients and correspond to a high risk of relapse (Alloy, Abramson, Walshaw, Keyser, & Gerstein, 2006; Scott & Pope, 2003). To date, the underlying neuronal mechanisms of such cognitive distortions in bipolar disorder (and other affective disorders) are not fully understood. Psychological studies repeatedly show enhanced sensitivity towards negative environmental cues during depressive episodes and suggest a reduced responsiveness to pleasant stimuli and other positive reinforcements (Jongen, Smulders, Ranson, Arts, & Krabbendam, 2007). In contrast, patients show an increased responsiveness to positive and rewarding stimuli during an acute (hypo-)manic episode (Chen et al., 2006; Wessa & Linke, 2009). This mood-congruent attentional bias seems to be mediated by a dysbalanced activity in anterior limbic regions (Green, Cahill, & Malhi, 2007; Phillips, Ladouceur, & Drevets, 2008). This is underlined by functional magnetic resonance imaging (fMRI) findings which consistently reveal deviant activation in the anterior cingulate cortex (ACC), the amygdala and the ventral striatum (Adler, Levine, DelBello, & Strakowski, 2005; Phillips, et al., 2008; Strakowski, Delbello, & Adler, 2005) in bipolar disorder. Additionally, a meta-analysis of voxel-based morphometry (VBM) findings in bipolar disorder confirmed gray matter reductions in the left ACC (Bora, Fornito, Yucel, & Pantelis, 2010). The ACC is associated with the integration of cognitive and emotional processes in support of goal-directed behavior, and its dysfunction may represent a primary neurobiological basis for the cognitive and emotional abnormalities observed in bipolar and depressive disorder (Cerullo, Adler, Delbello, & Strakowski, 2009). Besides these findings, an increasing number of studies use event-related potentials (ERP) to neurophysiologically investigate anterior limbic alterations in affective disorders. The

error-related negativity (ERN) (Gehring, Goss, Coles, Meyer, Donchin, 1993) or error negativity (Ne) (Falkenstein, Hohnsbein, Hoormann, & Blanke, 1991) is a commonly used ERP for the assessment of error-processing and action monitoring. The ERN is a negative deflection in the electroencephalogram (EEG), which peaks 50e100 msec after the commitment of an error (Falkenstein, et al., 1991). The ERN is generated in the ACC (Dehaene, Posner, & Tucker, 1994; Gehring et al., 1993; Segalowitz & Dywan, 2009), which is active during response monitoring and the commission of errors (Carter et al., 1998; Mathalon, Whitfield, & Ford, 2003). The ERN is followed by a later positive deflection, i.e., error positivity (Pe), which peaks 200e500 msec after an erroneous response (Falkenstein, Hoormann, Christ, & Hohnsbein, 2000; Nieuwenhuis, Ridderinkhof, Blom, Band, & Kok, 2001). This component has a centro-parietal distribution and seems to reflect further action monitoring, especially conscious error perception (Falkenstein, et al., 2000). In order to understand the functional impact of the ERN and Pe, some theoretical models have been developed. Some authors suggest that the ERN reflects error-detection (Falkenstein, et al., 1991), whereas others assume general conflict-detection processes (Yeung, Botvinick, & Cohen, 2004) or reinforcement and reversal learning as crucial mechanisms (Holroyd & Coles, 2002). Irrespective of these considerations, there is increasing evidence that the ERN relates to motivational and affective variables and correlates with personality traits (Hoffmann, Wascher, & Falkenstein, 2012) and given incentives (Santesso & Segalowitz, 2008; Segalowitz & Dywan, 2009). Recent investigations suggest that task engagement as common underlying factor predicts the amplitude of the ERN, which in turn is influenced by personality traits such as concern about social evaluation and the outcome of events (Tops & Boksem, 2010). In line with these considerations are the findings of ERN and Pe alterations in mental disorders. Patients suffering from depression and anxiety disorders show a larger ERN and Pe deflection as compared to healthy controls (Chiu & Deldin, 2007; Holmes & Pizzagalli, 2008, 2010; Olvet & Hajcak, 2008), which was interpreted as an increased sensitivity to errors. Accordingly, an fMRI study revealed enhanced error-related rostral ACC activity in depressed patients (Steele, Meyer, & Ebmeier, 2004). Several studies demonstrated heightened error monitoring processes (increased ERN and Pe amplitudes) in patients with obsessivecompulsive disorder (Endrass, Klawohn, Schuster, & Kathmann, 2008; W. J. Gehring, Himle, & Nisenson, 2000), anxiety (Moser, Hajcak, & Simons, 2005) and generalized anxiety disorder (Weinberg, Olvet, & Hajcak, 2010). On the other hand, patients suffering from externalizing disorders such as substance abuse and impulsive personality were

105

c o r t e x 6 6 ( 2 0 1 5 ) 1 0 3 e1 1 4

found to display decreased error processing (Franken, van Strien, Franzek, & van de Wetering, 2007; Hall, Bernat, & Patrick, 2007; Potts, George, Martin, & Barratt, 2006). Patients with psychosis and schizophrenia showed a blunted error signal in the EEG (Bates, Kiehl, Laurens, & Liddle, 2002; Foti, Kotov, Bromet, & Hajcak, 2012). Based on these findings, some authors proposed the ERN as a useful endophenotype also for internalizing disorders (Olvet & Hajcak, 2008; Riesel, Endrass, Kaufmann, & Kathmann, 2011). To our knowledge, no study on error processing investigating ERN and Pe characteristics in bipolar disorder has been published. The aim of the present study was to investigate characteristics in error processing in patients with bipolar disorder and their temporal dynamics (acute depressive and remitted patients) to assess their suitability as a trait disease marker. For this purpose, we compared ERN and Pe amplitudes and the test performance of patients with bipolar depression to data obtained in the same patients during euthymic mood state as well as to age-matched healthy controls. The ERN and Pe were measured during a modified Eriksen Flanker Task, which is broadly used to investigate error processing (Ehlis, H, Bernhard, & Fallgatter, 2005; Herrmann et al., 2010; Olvet & Hajcak, 2008). Based on previous work in other mental disorders, we hypothesized that error processing would be enhanced in bipolar patients during acute depression, as compared to healthy controls, and would normalize when patients are euthymic.

2.

Methods

2.1.

Participants

Forty-seven participants were originally recruited, from which six subjects (3 from the patient and 3 from the control group) prematurely aborted the experiment. These e and three further subjects with an errorless performance e were excluded from analysis. The total sample thus comprised 20 patients and 18 controls. Patients were recruited through a specialized bipolar ward at the Department of Psychiatry, Psychosomatics and Psychotherapy at the University Hospital Wu¨rzburg. Their disorder status was diagnosed and confirmed by two trained psychiatrists (AR, SKS) and extensive testing with the OPCRIT diagnostic system (McGuffin, Farmer, & Harvey, 1991). Patients were measured in a depressive episode according to DSM-IV criteria [Montgomery Asperger Depression Rating Scale (MADRS) scores (Montgomery & Asberg, 1979): mean ¼ 19.6 ± 8.6; Young Mania Rating Scale (YMRS) scores (Young, Biggs, Ziegler, & Meyer, 1978): mean ¼ 4.9 ± 3.7]. Nine patients could be enrolled for a follow-up measurement at least 3 months later and after reaching remission (MADRS mean ¼ 3.4 ± 3.4; YMRS mean ¼ 1.8 ± 1.1). Medication was recorded to control for possible confounds of the EEG results with medication effects, for details please see Table 1. Nine of the 20 patients were diagnosed as bipolar I patients, eleven as bipolar II patients, no patient was expressing rapid cycling symptoms. Patients with schizoaffective disorder were not included in the study. Controls were matched to patients regarding age, sex and education level. All controls were free of past and current axis

Table 1 e Summary the medication taken by patients participating in the study. Medication

Number of patients

Lithium Acute Remitted Antipsychotics first gen Haloperidol Acute Antipsychotics sec gen Aripiprazol Acute Risperidon Acute Clozapine Acute Quetiapine Acute Remitted Antiepileptics Acute Remitted SSRIs Acute Remitted Tricyclica Acute Remitted Benzodiazepines Acute Melatonine Acute Opiods Acute

9 8

1

7 1 1 10 2 6 2 8 1 3 3 15 1 1

I (classified via the DSM-IV) disorders as assessed with the Mini International Neuropsychiatric Interview (MINI), German Version 5.0 (Sheehan et al., 1998). Details on sociodemographic features can be found in Table 2. The study was reviewed and approved by the Ethics Committee of the University of Wu¨rzburg, and all procedures were in accordance with the latest version of the Declaration of Helsinki. All participants gave written informed consent after comprehensive explanation of the experimental procedures.

2.2.

Eriksen Flanker task

A modified version of the Eriksen Flanker Task (Eriksen & Eriksen, 1979) was presented via a monitor in an electrically shielded, sound-attenuated, dimly-illuminated room. At the

Table 2 e Sociodemographics of the test subjects. Patients and controls did not significantly differ on any of the demographic variables. Education level ranges from 1 ¼ only high school diploma to 4 ¼ postdoctoral degree.

Sex _/\ Age (years) Education level

Bipolar patients

Healthy controls

Group comparison

11/9 44.3 ± 9.5 1.9 ± .8

6/12 44.2 ± 11.9 2.1 ± .8

Chi2 ¼ 1.8, p ¼ .2 T(36) ¼ .05, p ¼ .9 c2 ¼ 3.4, p ¼ .3

106

c o r t e x 6 6 ( 2 0 1 5 ) 1 0 3 e1 1 4

beginning of each trial, a combination of five arrows was presented for 125 msec (<<><<; ><><>; >><>>; <><><) and participants were instructed to press the right/left shift button on a standard keyboard when the arrow in the middle was pointing to the right/left. Directly after the button press, one of two graphic forms of feedback was presented for 500 msec: 1) neutral in form of a plus (for a correct response), a minus (for an erroneous response), an exclamation sign (for a correct, but slow response), or 2) emotional in form of a smiling face, a sad face, and a neutral face following a delayed answer. Type of feedback was fixed within each task block (see below), and order of blocks was counterbalanced across participants. Intertrial intervals were 1250 msec long. To correctly answer a trial, participants had to react within a certain time limit, which was different for each test subject and determined in a training block: Before the experiment, 42 trials were presented with a fixed response time limit of 500 msec. Not considering the first 10 trials the mean reaction time for correct responses was calculated and reduced for a further 5 percent representing the cut-off time for a correct response in the main experiment. This was done to ensure that test subjects generate enough errors. After analyzing the data, it became evident that some participants did not make enough errors in one feedback block, so data from the two blocks were merged and the analysis of the feedback ERN was abandoned. Experimental blocks were separated by a break, 200 trials were presented in each block.

electrodes (Cz, Pz) for the Pe. Absolute peak values were used to quantify the ERP amplitudes. Statistical analyses were performed using SPSS for Windows, version 21 (IBM Corp. Released 2012. IBM SPSS Statistics for Windows, Version 21.0. Armonk, NY: IBM Corp.). All data were checked for normal distribution with the KolmogoroveSmirnov test. This includes data comparing acute and remitted patients, since only 9 patients were included in this sub-analysis. Two-tailed t-tests were conducted to test within- and between-group differences in behavioral measures. Analyses of variance (ANOVAs) calculated for the latencies and amplitudes of the two components comprised the within-subject factors condition (errors, correct responses, slow answers) and electrode position (FCz, Cz and Cz, Pz, respectively), as well as the between-subject factor of group. The Greenhouse-Geisser procedure was used to correct degrees of freedom whenever necessary. One-way ANOVAs and two-tailed independent and paired t-tests were used for posthoc analyses and for the reaction time data. Equality of variances was tested by means of Levene's test and corrections for inequality were performed when necessary.

3.

Results

3.1.

Behavioral data

All behavioral data were normally distributed (p > .1).

2.3.

EEG recording 3.1.1.

The EEG was recorded from 28 scalp electrodes according to the international 10/20-System (Jasper, 1958). Three additional electrodes were placed at the outer canthi of both eyes and below the right eye for the registration of eye movements. The recording reference was placed between Fz and Cz. Electrode impedances were kept below 5 kU. Data were recorded with a 32-channel DC-amplifier and the software “vision recorder” (BrainProducts, Munich, Germany).

2.4.

Data analyses and statistics

ERP data were analyzed with the software “vision analyzer” (BrainProducts, Munich, Germany). First, a linear derivation was calculated for all data. Then the data were band pass filtered with a low cutoff of .1 Hz and a high cutoff of 50 Hz, and an eye movement artifact correction was applied (Gratton, Coles, & Donchin, 1983). Then, data were re-referenced to an average reference, and segmented into the three task conditions (correct responses, erroneous responses, delayed responses) in a response-locked manner (150e750 msec around the button press). Afterwards, an artifact rejection excluded all segments with amplitudes exceeding ±70 mV or voltage steps of more than 70 mV per sampling point. Corresponding segments were then averaged, baseline-corrected and an semiautomatic peak detection was completed for two distinct components of the ERP: an early negative deflection with a frontocentral distribution (ERN/Ne), and a subsequent centroparietal positive peak (Pe). Peaks were individually determined within defined time windows (ERN/Ne: 20e80 msec; Pe: 120e270 msec) at frontocentral sites (FCz, Cz) for the ERN and at two midline

Depressed patients versus controls

In the 400 trials, acutely depressed patients made an average of 94 ± 109 errors (mean ± SD), remitted patients an average of 164 ± 83 errors, and control subjects an average of 62 ± 50 errors (t36 ¼ 1.16, n.s.). Patients produced an average of 246 ± 135 correct responses, control subjects had 296 ± 117 correct responses (t36 ¼ 1.21, n.s.). Patients made an average of 233 ± 107 answers that were classified as delayed, and 224 ± 96 responses of control subjects were classified as delayed (t36 ¼ .26, n.s.). The mean reaction time threshold was 471 ± 51 msec for patients and 505 ± 127 msec for controls t36 ¼ 1.51, n.s.). None of these differences between groups were statistically significant. When comparing reaction times between depressed patients and controls, patients were marginally slower when making an error as compared to controls, and significantly slower when responding correctly but too late. This pattern changed when patients were measured again during their remitted state. Details on the data can be found in Table 3.

3.1.2.

Remitted patients versus controls

When comparing remitted patients with controls, both groups did not significantly differ in any of the reaction time measures. However, remitted patients made significantly fewer correct responses, more errors, and fewer slow responses. For details, please see Table 3.

3.1.3. phase

Patients in their acute phase versus in their remitted

When comparing acutely depressed patients with their remitted state, reaction times of the remitted patients were

c o r t e x 6 6 ( 2 0 1 5 ) 1 0 3 e1 1 4

107

Table 3 e Behavioral results for within- and betweengroup differences. Mean

SD

Acutely ill patients versus Controls Reaction time correct responses (msec) T(36) ¼ .87, n.s. Ill patients 474.1 91.6 Controls 448.1 92.7 Reaction time errors (msec) T(36) ¼ 1.82, p ¼ .08 Ill patients 557.1 102..4 Controls 488.6 129.4 Reaction time delayed responses (msec) T(36) ¼ 2.13, p ¼ .04 Ill patients 687.2 107.3 Controls 599.2 146.3 Remitted patients versus Controls Correct responses T(17.4) ¼ 9.11, p < .000 Remitted patients 41.8 9.2 Controls 296.7 117.9 Errors T(11.02) ¼ 3.4, p ¼ .006 Remitted patients 164.7 83.3 Controls 62.2 50.4 Delayed responses T(25) ¼ 3.85, p ¼ .001 Remitted patients 84.7 68.8 Controls 224.6 96.9 Acutely ill patients versus remitted patients Reaction time correct responses (msec) T(8) ¼ 3.6, p ¼ .007 Ill patients 511.5 69.1 Remitted patients 420.9 49.01 Reaction time errors (msec) T(8) ¼ 2.45, p ¼ .004 Ill patients 566.8 83.5 Remitted patients 495.8 112.6 Reaction time delayed responses (msec) T(8) ¼ 3.6, p ¼ .007 Ill patients 714.4 124.2 Remitted patients 575.8 63.1 Correct responses T(8) ¼ 4.94, p ¼ .001 Ill patients 304.3 156.5 Remitted patients 41.8 9.2 Errors T(8) ¼ 3.73, p ¼ .006 Ill patients 61.9 38.8 Remitted patients 164.7 83.3 Delayed responses T(8) ¼ 2.67, p ¼ .028 Ill patients 216.8 124.1 Remitted patients 84.7 68.8

Fig. 1 e Response-locked grand average ERP wave plots of ERN and Pe at Cz for incorrect, correct and delayed trials for depressed patients.

group [F(1,35) ¼ 3.7, p ¼ .063]. The interaction effect for response*group was significant [F(2,70) ¼ 7.4, p ¼ .001]. Since differences between electrodes were not of interest, this effect was not analyzed further. Regarding the significant interaction, independent post-hoc t-tests revealed differences between groups for correct and delayed responses for the mean value of both electrode positions, i.e., patients showed a more pronounced ERN compared with controls, even on responses where no ERN-like pattern should be elicited. Interestingly, there were no between-group differences for the erroneous response, which is supposed to elicit the greatest ERN. Details on the post hoc t-tests can be found in Table 4.

3.2.1.2. REMITTED PATIENTS VERSUS CONTROLS. A2  3  2 repeated measures ANOVA showed significant main effects for the

significantly faster for all conditions, i.e., the reaction time threshold was significantly faster, and answers to correct responses, errors, and responses that were classified as delayed became significantly faster as well. Details for all analyses are given in Table 3.

3.2.

ERP data

3.2.1.

ERN

Figs. 1e4 show the ERN amplitudes in the grand average curves of acutely depressed, remitted bipolar patients and healthy controls. Fig. 5 contains an overview of ERN amplitudes in the grand average curves of acutely depressed, remitted bipolar patients and healthy controls.

3.2.1.1. DEPRESSED PATIENTS VERSUS CONTROLS. A 2  3  2 repeated measures ANOVA revealed significant main effects for the factor electrode position (FCz versus Cz) [F(1,35) ¼ 27.006, p < .001], for the factor response (correct, erroneous, delayed) [F(1.6, 56.3) ¼ 30.4, p < .001] and a trend effect for the factor

Fig. 2 e Response-locked grand average ERP wave plots of ERN and Pe at Cz for incorrect, correct and delayed trials for euthymic patients.

108

c o r t e x 6 6 ( 2 0 1 5 ) 1 0 3 e1 1 4

3

2

1

ERN peak in μV

0

-1 depressed remitted control

-2

-3

-4

-5

-6 correct response

Fig. 3 e Response-locked grand average ERP wave plots of ERN and Pe at Cz for incorrect, correct and delayed trials for healthy controls.

factor electrode position [F(1,23) ¼ 15.9, p ¼ .001] and for the factor response [F(2,46) ¼ 26.2, p < .001], but no main effect for the factor group (remitted, controls) [F(1,23) ¼ .91, p ¼ .35]. An interaction effect for the factors response*group was also significant [F(2,46) ¼ 5.3, p ¼ .009]. Independent post hoc t-tests revealed a difference between groups only for the erroneous responses. The ERN was significantly less pronounced in remitted patients. Details can be found in Table 4.

3.2.1.3. PATIENTS IN THEIR ACUTE PHASE VERSUS IN THEIR REMITTED PHASE. For the comparisons of patients in their acute and remitted states, only depressed patients who were measured again in the remitted phase, were included in the calculations. A 2  3 repeated measures ANOVA with the within-group effect disorder status (acutely depressed versus remitted)

Fig. 4 e ERP waveforms showing response-locked plots of ERN and Pe for each of the three groups (depressed patients, euthymic patients and healthy controls).

erroneous response

delayed response

Fig. 5 e Comparison of ERN peaks for the mean value of both electrode positions for acutely depressed, remitted bipolar patients and healthy controls. Error bars indicate the standard error.

showed a significant main effect for the factor group [F(1,7) ¼ 13.5, p ¼ .008], a trend effect for the factor electrode [F(1,7) ¼ 4.9, p ¼ .062], and a significant main effect for the factor response [F(2,14) ¼ 8.5, p ¼ .004]. The interaction effect response*group was marginally significant [F(2,14) ¼ 3.4, p ¼ .062].

Table 4 e Post Hoc t-tests showing between-group differences in the ERN. ERN

Mean

SD

Acutely ill patients versus Controls Correct responses T(35.2) ¼ 2.5, p ¼ .016 Depressed patients 1.8 0.6 Controls 0.4 0.6 Errors T(35) ¼ .28, n.s. Depressed patients 3.8 2.7 Controls 4.1 2.9 Delayed responses T(36) ¼ ¡2.6, p ¼ .013 Depressed patients 3.3 3.8 Controls 0.6 2.4 Remitted patients versus Controls Correct responses T(25) ¼ .7, n.s. Remitted patients 1.2 4.02 Controls 0.4 2.6 Errors T(23) ¼ 2.1, p ¼ .052 Remitted patients 1.7 1.9 Controls 4.1 2.9 Delayed responses T(25) ¼ .1, n.s. Remitted patients 1.7 3.04 Controls 0.6 2.4 Acutely ill patients versus remitted patients Correct responses T(8) ¼ ¡3.4, p ¼ .01 Depressed patients 1.8 3.2 Rremitted patients 1.2 4.03 Errors T(7) ¼ 1.06, n.s. Depressed patients 3.2 3.1 Remitted patients 1.7 1.9 Delayed responses T(8) ¼ ¡1.9, p ¼ .08 Depressed patients 3.7 5.2 Remitted patients 1.7 3.04

109

c o r t e x 6 6 ( 2 0 1 5 ) 1 0 3 e1 1 4

3.2.1.4. WITHIN-GROUP DIFFERENCES. Dependent post hoc t-tests were calculated for the responses separately for all groups. For acutely depressed patients, significant differences were found between correct and erroneous answers as well as between correct and slow answers, but not between erroneous and slow answers, with more positive ERN values for correct compared to both erroneous and delayed answers. For the control group, as predicted, differences could be found between correct and erroneous responses, and between erroneous and slow answers, with more negative ERN values following incorrect responses. For the remitted group, the same differences persisted as in the acutely ill group: Significant differences could be found between correct and erroneous answers, as well as between correct and slow answers, but no differences could be found between erroneous and slow answers. All findings were true for both electrode positions and can be found in Table 5. 3.2.2.

Pe

Fig. 6 contains an overview of Pe amplitudes in the grand average curves of acutely depressed, remitted bipolar patients and healthy controls.

3.2.2.1. ACUTELY DEPRESSED PATIENTS VERSUS CONTROLS. A 2  3  2 repeated measures ANOVA revealed significant main effects for the factor electrode position (Cz, Pz) [F(1,35) ¼ 67.08, p < .000], for the factor response (correct, false, delayed) [F(1.5, 53.6) ¼ 14.8, p < .000] and for the factor group [F(1,35) ¼ 15.02, p < .000]. The interaction effect for electrode*response was significant [F(2,70) ¼ 5.6, p ¼ .006] as well. Since differences between electrodes were not of interest in this paper, the main effect was not further analyzed. Independent post-hoc t-tests revealed differences between groups for errors, correct and delayed responses for both electrode

Table 5 e Results of the within post hoc t-tests for each group separately. Means and standard deviations are already presented in Table 3. ERN within post hoc t-tests depressed patients Correct responses versus errors Correct responses versus delayed responses Delayed responses versus errors remitted patients Correct responses versus errors Correct responses versus delayed responses Delayed responses versus errors Controls Correct responses versus errors Correct responses versus delayed responses Delayed responses versus errors

T(19) ¼ 4.4, p < .001 T(19) ¼ 3.6, p ¼ .002 T(19) ¼ .7, n.s. T(7) ¼ 3.9, p ¼ .006 T(8) ¼ 3.3, p ¼ .01 T(7) ¼ .17, n.s. T(16) ¼ 6.8, p < .001 T(17) ¼ 1.9, p ¼ .076 T(16) ¼ 4.8, p < .001

8

7

6

5

mean Pe peak in μV

Post hoc dependent sample t-tests revealed significant differences for the average values of both electrode positions FCz and Cz for correct responses: the ERN was less pronounced in remitted patients, and a trend effect for the slow responses, with remitted patients showing almost no ERN and acutely ill patients showing a moderate ERN. Details on the post hoc testing can be found in Table 4.

4 depressed 3

remitted control

2

1

0 correct response

erroneous response

delayed response

-1

-2

Fig. 6 e Comparison of Pe peaks for the averaged values on electrode position Cz and Pz for acutely depressed, remitted bipolar patients and healthy controls. Error bars indicate the standard error.

positions; patients had an attenuated Pe compared with controls, even on responses where a Pe should be elicited. Details for means and standard deviations can be found in Table 6.

3.2.2.2. REMITTED PATIENTS VERSUS CONTROLS. A2  3  2 repeated measures ANOVA showed significant main effects for the factor electrode position [F(1,23) ¼ 39.4, p < .000] and for the factor response [F(1.3,30.7) ¼ 11.04, p ¼ .001], but no main effect for the factor group (acutely ill, remitted) [F(1,23) ¼ 1.005, p ¼ .33]. An interaction effect for the factors electrode*response was also significant [F(2,46) ¼ 4.9, p ¼ .012]. 3.2.2.3. PATIENTS IN THEIR ACUTE PHASE VERSUS IN THEIR REMITTED PHASE. A 2  3 repeated measures ANOVA with the within factor disorder status showed a significant main effect for the factor group [F(1,7) ¼ 10.4, p ¼ .015], for the factor electrode

Table 6 e Post hoc t-tests showing between-group differences in the Pe for both electrode positions. Pe

Mean

SD

Acutely ill patients versus Controls Correct responses Ill patients 1.4 3.2 Controls 4.7 4.3 Errors Ill patients 3.5 2.6 Controls 6.2 2.3 Delayed responses Ill patients .02 2.2 Controls 3.7 3.9 Acutely ill patients versus remitted patients Correct responses T(8) ¼ ¡4.1, p ¼ .004 Ill patients 0.6 2.8 Remitted patients 4.6 3.2 Errors T(7) ¼ 1.4, n.s. Ill patients 3.8 3.7 Remitted patients 5.1 3.8 Delayed responses T(8) ¼ .7, n.s. Ill patients 0.1 1.4 Remitted patients 0.5 2.4

110

c o r t e x 6 6 ( 2 0 1 5 ) 1 0 3 e1 1 4

[F(1,7) ¼ 17.2, p ¼ .004], and for the factor response [F(1.2,8.6) ¼ 4.1, p ¼ .04]. The interaction effect response*group was also significant [F(2,14) ¼ 3.8, p ¼ .049]. Post hoc dependent sample t-tests revealed significant differences for the averaged values of both electrode positions for correct responses, for which the Pe was significantly more pronounced in remitted patients, and a trend effect at Cz for errors, with acutely depressed patients showing less positive values than remitted patients. Interestingly, still no differences were observed for the Pe in slow responses. Details on the post hoc testing can be found in Table 6.

3.2.2.4. WITHIN-GROUP DIFFERENCES. Dependent post hoc t-tests were then calculated for the responses separately for all three groups. For acutely depressed patients, significant differences were found between correct and erroneous answers, between correct and slow answers, and between erroneous and slow answers. Patients showed the highest Pe after errors, and the lowest after delayed responses. For the remitted patients, significant differences were detected for comparisons of correct responses and delayed responses, as well as for delayed responses and errors. No differences could be found for the comparison between correct responses and errors. The highest Pe was again observed after errors, which was a little higher than the response to correct answers. The amplitude changes elicited by delayed responses were the smallest. For the controls, all comparisons were significantly different. Controls had the highest Pe amplitudes are errors, followed by responses to correct answers, and the lowest amplitude could be observed after delayed responses. Details on the t-tests can be found in Table 7.

4.

Discussion

To our knowledge, the present study is the first to investigate error processing using EEG components ERN and Pe in a sample of bipolar patients. We examined acute depressed and remitted bipolar patients with a modified Flanker task.

Table 7 e Post hoc t-tests for within group differences. Means and standard deviations used to compare amplitude changes for the different responses can be found in Table 5. Pe within post hoc t-tests Depressed patients Correct responses versus errors Correct responses versus delayed responses Delayed responses versus errors Remitted patients Correct responses versus errors Correct responses versus delayed responses Delayed responses versus errors Controls Correct responses versus errors Correct responses versus delayed responses Delayed responses versus errors

T(19) ¼ ¡2.2, p ¼ .039 T(19) ¼ 2.3, p ¼ .033 T(19) ¼ 4.1, p ¼ .001 T(7) ¼ .2, n.s. T(8) ¼ 5.9, p < .001 T(7) ¼ 2.6, p ¼ .035 T(16) ¼ ¡1.7, p ¼ .097 T(17) ¼ 2.3, p ¼ .038 T(16) ¼ 3.03, p ¼ .008

We did not find differences in ERN amplitudes between bipolar patients and healthy controls. That is, contrary to our hypothesis, an enhanced error processing in acutely depressed bipolar patients could not be discerned. While some studies did also fail in showing deviant error detection and processing in unipolar depressive patients (Georgiadi, Liotti, Nixon, & Liddle, 2011; Ruchsow et al., 2006, 2004; Schrijvers et al., 2008; Schrijvers et al., 2009), other studies reported an increased error processing in major depression (Chiu & Deldin, 2007; Holmes & Pizzagalli, 2008, 2010). Some authors argue that a differing symptom severity led to the inconsistent findings across studies (Olvet, Klein, & Hajcak, 2010; Schrijvers, et al., 2008), assuming a non-linear relationship between symptom severity and ERN amplitude: More severe depressive symptoms seem to be associated with smaller ERN amplitudes, moderate depressive symptoms with a potentiated ERN. Our sample, consisting of patients suffering from acute bipolar depression, was characterized by moderate symptom severity as measured by MADRS scores and we did not find a correlation between the severity of depression and the characteristics of the ERN/Pe signal. Thus, one of the main findings of our study is an unaltered ERN signal after erroneous responses in both acute and remitted bipolar patients in comparison to healthy controls. Additionally, we analysed the EEG signals after correct and delayed responses and found significant differences between bipolar patients and healthy controls on these trials. Patients with an acute depression showed an ERN signal after correct as well as delayed responses. As mentioned above, the ERN is associated with unconscious error monitoring and processing. Accordingly, the enhanced ERN signal of bipolar depressive patients in correct or delayed trials was unexpected as in these trials no error was made. However, being in accordance with the clinical appearance of depression it could be interpreted as a neurophysiological equivalent of the typical cognitive distortion in depression as patients often blame themselves for minor or even non-existing errors. This was also found by Morris et al. (Morris, Yee, & Nuechterlein, 2006) in schizophrenic patients, who displayed an enhanced ERN signal after correct responses. In the next step, this finding should be compared to data from patients suffering from major depression to ascertain the specificity regarding bipolar disorder. Another finding of our study is a reduced Pe signal in acute depressive patients compared to controls after incorrect as well as correct and delayed responses. This could reflect a blunted conscious error recognition in acute depression. Thus, depression may lead to a state of consistently searching for errors and consciously engaging in an error detection modus, respectively, which prevents adequate error processing. In line with this finding, other studies found either reduced or unaltered Pe amplitudes in unipolar depressive patients and attributed this result to the anhedonia and apathy of depressed patients (Chiu & Deldin, 2007; Holmes & Pizzagalli, 2010; Ruchsow, et al., 2004; Schrijvers, et al., 2009), arguing that this electrophysiological finding is not etiologically specific but rather inherent to a depressive syndrome irrespective of its origin. Recent data in healthy controls indicate that the Pe signal depends mainly on motivational salience (Overbeek, van Alfen, Bor, & Zwarts, 2005; Tops &

c o r t e x 6 6 ( 2 0 1 5 ) 1 0 3 e1 1 4

Boksem, 2010). As we did not use any incentives in our Flanker task to increase task engagement, the reduced Pe signal could be due to the absence of such motivational factors. Another interpretation could be that acute depressive patients are in some way accustomed to the subjective feeling of failure, which is why they may not have a strong Pe signal. After remission, the Pe amplitude of euthymic patients was comparable to healthy controls after incorrect and correct responses. However, after delayed responses, remitted patients still had a reduced Pe signal, indicating potential trait characteristics. But given that the Pe has not typically been the focus of studies of error monitoring in mental disorders, further investigations are needed. Of special interest is our result of the pre-post measurement within the bipolar patients. The biased error perception after correct and delayed responses mentioned above persisted in part when the patients were remitted. Euthymic patients showed a significantly decreased Pe and a trend for an increased ERN amplitude on delayed reactions in comparison to controls. This result indicates that the aberrant error processing of bipolar patients could be a trait rather than a state effect, possibly representing a marker for vulnerability to depressive episodes. It is known that the cognitive distortion of increased perception of personal failures constitutes a risk factor for depressive relapses (Beck, 1963; Leppanen, 2006; Teasdale, 1983). Similar to our results, recent studies showed increased ERN amplitudes in remitted unipolar depressive patients and suggest an overactive ACC and a hyper vigilant error-monitoring system (Botvinick, Cohen, & Carter, 2004; Georgiadi, et al., 2011). Regarding the behavioral performance in the Flanker task, we found that acutely depressed patients reacted slightly more slowly than euthymic patients and healthy controls, which is consistent with the characteristic psychomotor slowing in depression. However, they featured a comparable amount of correct and erroneous responses. Remitted patients had RTs similar to the healthy control group, but as a result of those faster reactions task accuracy was reduced. These performance results are in line with an emerging body of research showing poor neuropsychological functioning of bipolar patients compared to healthy controls (Mann-Wrobel, Carreno, & Dickinson, 2011).

4.1.

111

ERN (de Bruijn, et al., 2006). However, antidepressant medication is highly diverse, so that it cannot be directly compared between patients, and it is not clear if the antidepressant medication could play a role in the altered ERN signals in our sample of bipolar patients. The sample size of the remitted patients was rather small. This needs to be taken into consideration especially when interpreting the data from the remitted patient group. Small sample sizes in general decrease the power of the statistical analyses, and thereby reduce the likelihood that a statistically significant result reflects a true effect. Future research including more patients especially in their remitted state is necessary in order to make better educated guesses at the nature of the phenomenon of cognitive distortion as a trait factor in bipolar disorder. If a unipolar depressed patient group were included, maybe the contradicting results found in other studies with unipolar patients could have been interpreted better as well. Also, if a unipolar depressed group would have been included, one could have drawn better conclusions about the specificity of the studied effect, especially highlighting differences between these disorders or concluding on common mechanisms, especially in light of the fact that they phenomenologically overlap to a considerable extent.

4.2.

Conclusion and perspective

The acutely depressed patients showed an ERN signal after correct and delayed responses, which can be interpreted as cognitive bias in the perception of personal failures. The decreased Pe signal possibly reflects blunted conscious error recognition in acute depression. However, more studies in larger samples of euthymic bipolar patients are needed to answer the question if altered error processing in bipolar patients is a state or trait finding. Furthermore, it would be helpful to conduct longitudinal studies following the same patients over time across different phases. Particularly the examination of patients with an acute mania would be interesting. Considering the findings of Olvet et al. (2008), and keeping in mind the symptoms of impulsivity and increased self-esteem during mania, one could assume a decreased ERN and Pe signal in manic patients.

Limitations

It has been discussed that medication could influence ERP data and there is evidence for a general ERN attenuation under antipsychotic medication (de Bruijn, Sabbe, Hulstijn, Ruigt, & Verkes, 2006; Riba, Rodriguez-Fornells, Morte, Munte, & Barbanoj, 2005), although inconsistent findings exist (Zirnheld et al., 2004). Therefore, we calculated the chlorpromazine equivalents for the antipsychotic medication prescribed to our sample (Kroken, Johnsen, Ruud, Wentzel-Larsen, & Jorgensen, 2009; Woods, 2003). The results indicate that acutely depressive and remitted patients did not significantly differ regarding the mean dosage of antipsychotic medication. Hence, we can confidently assume that antipsychotic medication is not responsible for differences in the ERN and Pe in the pre post measurement. However, patients also received antidepressant medication. Some studies show that certain selective serotonin reuptake inhibitors do not influence the

Disclosures The study was supported by the DFG (Grant RE1632/5-1 to AR, He45411-1 to MJH, KFO 125 to AR; SFB TRR 58 Z02 to AR, C04 to MJH, RTG 1253 to AR, MJH and JK) and the IZKF (project Z-3/24, to SKS) JV was supported by a grant of the German Excellence Initiative to the Graduate School of Life Sciences, University of Wuerzburg. TD was partly supported by the LEAD graduate school [GSC1028], a project of the Excellence Initiative of the German federal and state governments.

Acknowledgments We are grateful to all individuals who participated in this study.

112

c o r t e x 6 6 ( 2 0 1 5 ) 1 0 3 e1 1 4

references

Adler, C. M., Levine, A. D., DelBello, M. P., & Strakowski, S. M. (2005). Changes in gray matter volume in patients with bipolar disorder [Controlled Clinical Trial Research Support, N.I.H., Extramural Research Support, Non-U.S. Gov't Research Support, U.S. Gov't, P.H.S.] Biological Psychiatry, 58(2), 151e157. http://dx.doi.org/10.1016/j.biopsych.2005.03.022. Alloy, L. B., Abramson, L. Y., Walshaw, P. D., Keyser, J., & Gerstein, R. K. (2006). A cognitive vulnerability-stress perspective on bipolar spectrum disorders in a normative adolescent brain, cognitive, and emotional development context [Research Support, N.I.H., Extramural Review] Development and Psychopathology, 18(4), 1055e1103. http:// dx.doi.org/10.1017/S0954579406060524. Bates, A. T., Kiehl, K. A., Laurens, K. R., & Liddle, P. F. (2002). Errorrelated negativity and correct response negativity in schizophrenia. Controlled Clinical Trial Research Support, Non-U.S. Gov't] Clinical Neurophysiology, 113(9), 1454e1463. Beck, A. T. (1963). Thinking and depression. I. Idiosyncratic content and cognitive distortions. Archives of General Psychiatry, 9, 324e333. Bora, E., Fornito, A., Yucel, M., & Pantelis, C. (2010). Voxelwise meta-analysis of gray matter abnormalities in bipolar disorder [Meta-Analysis Research Support, Non-U.S. Gov't] Biological Psychiatry, 67(11), 1097e1105. http://dx.doi.org/10.1016/ j.biopsych.2010.01.020. Botvinick, M. M., Cohen, J. D., & Carter, C. S. (2004). Conflict monitoring and anterior cingulate cortex: an update [Research Support, Non-U.S. Gov't Research Support, U.S. Gov't, P.H.S.] Trends in Cognitive Sciences, 8(12), 539e546. http://dx.doi.org/ 10.1016/j.tics.2004.10.003. de Bruijn, E. R., Sabbe, B. G., Hulstijn, W., Ruigt, G. S., & Verkes, R. J. (2006). Effects of antipsychotic and antidepressant drugs on action monitoring in healthy volunteers [Clinical Trial Comparative Study Randomized Controlled Trial] Brain Research, 1105(1), 122e129. http://dx.doi.org/10.1016/ j.brainres.2006.01.006. Carter, C. S., Braver, T. S., Barch, D. M., Botvinick, M. M., Noll, D., & Cohen, J. D. (1998). Anterior cingulate cortex, error detection, and the online monitoring of performance [Research Support, U.S. Gov't, P.H.S.] Science, 280(5364), 747e749. Cerullo, M. A., Adler, C. M., Delbello, M. P., & Strakowski, S. M. (2009). The functional neuroanatomy of bipolar disorder [Research Support, N.I.H., Extramural Review] International Review of Psychiatry, 21(4), 314e322. http://dx.doi.org/10.1080/ 09540260902962107. Chen, C. H., Lennox, B., Jacob, R., Calder, A., Lupson, V., BisbrownChippendale, R., et al. (2006). Explicit and implicit facial affect recognition in manic and depressed states of bipolar disorder: a functional magnetic resonance imaging study [Comparative Study Research Support, Non-U.S. Gov't] Biological Psychiatry, 59(1), 31e39. http://dx.doi.org/10.1016/j.biopsych.2005.06.008. Chiu, P. H., & Deldin, P. J. (2007). Neural evidence for enhanced error detection in major depressive disorder [Comparative Study Research Support, Non-U.S. Gov't] American Journal of Psychiatry, 164(4), 608e616. http://dx.doi.org/10.1176/ appi.ajp.164.4.608. Dehaene, S., Posner, M. I., & Tucker, D. M. (1994). Localization of a neural system for error detection and compensation. Psychological Science, 5, 303e305. Ehlis, A. C., H, M. J., Bernhard, A., & Fallgatter, A. J. (2005). Monitoring of internal and external error signals. Journal of Psychophysiology, 19(4), 263e269. Endrass, T., Klawohn, J., Schuster, F., & Kathmann, N. (2008). Overactive performance monitoring in obsessive-compulsive disorder: ERP evidence from correct and erroneous reactions

[Comparative Study] Neuropsychologia, 46(7), 1877e1887. http:// dx.doi.org/10.1016/j.neuropsychologia.2007.12.001. Eriksen, C., & Eriksen, B. (1979). Target redundancy in visual search: do repetitions of the target within the display impair processing? Perception & Psychophysics, 26(3), 195e205. http:// dx.doi.org/10.3758/bf03199869. Falkenstein, M., Hohnsbein, J., Hoormann, J., & Blanke, L. (1991). Effects of crossmodal divided attention on late ERP components. II. Error processing in choice reaction tasks. Electroencephalography and Clinical Neurophysiology, 78(6), 447e455. Falkenstein, M., Hoormann, J., Christ, S., & Hohnsbein, J. (2000). ERP components on reaction errors and their functional significance: a tutorial [Research Support, Non-U.S. Gov't Review] Biological Psychology, 51(2e3), 87e107. Foti, D., Kotov, R., Bromet, E., & Hajcak, G. (2012). Beyond the broken error-related negativity: functional and diagnostic correlates of error processing in psychosis. Biological Psychiatry, 71(10), 864e872. http://dx.doi.org/10.1016/ j.biopsych.2012.01.007. Franken, I. H., van Strien, J. W., Franzek, E. J., & van de Wetering, B. J. (2007). Error-processing deficits in patients with cocaine dependence [Research Support, Non-U.S. Gov't] Biological Psychology, 75(1), 45e51. http://dx.doi.org/10.1016/ j.biopsycho.2006.11.003. Gehring, W. J., Goss, B., Coles, M. G. H., Meyer, D. E., & Donchin, E. (1993). A neural system for error detection and compensation. Psychological Science, 4, 385e390. Gehring, W. J., Himle, J., & Nisenson, L. G. (2000). Actionmonitoring dysfunction in obsessive-compulsive disorder [Research Support, Non-U.S. Gov't] Psychological Science, 11(1), 1e6. Georgiadi, E., Liotti, M., Nixon, N. L., & Liddle, P. F. (2011). Electrophysiological evidence for abnormal error monitoring in recurrent major depressive disorder [Research Support, Non-U.S. Gov't] Psychophysiology, 48(9), 1192e1202. http:// dx.doi.org/10.1111/j.1469-8986.2011.01198.x. Gratton, G., Coles, M. G., & Donchin, E. (1983). A new method for off-line removal of ocular artifact. Electroencephalography and Clinical Neurophysiology, 55(4), 468e484. Green, M. J., Cahill, C. M., & Malhi, G. S. (2007). The cognitive and neurophysiological basis of emotion dysregulation in bipolar disorder [Research Support, Non-U.S. Gov't Review] Journal of Affective Disorders, 103(1e3), 29e42. http://dx.doi.org/10.1016/ j.jad.2007.01.024. Hall, J. R., Bernat, E. M., & Patrick, C. J. (2007). Externalizing psychopathology and the error-related negativity [Research Support, N.I.H., Extramural Research Support, Non-U.S. Gov't] Psychological Science, 18(4), 326e333. http://dx.doi.org/10.1111/ j.1467-9280.2007.01899.x. Herrmann, M. J., Mader, K., Schreppel, T., Jacob, C., Heine, M., Boreatti-Hummer, A., et al. (2010). Neural correlates of performance monitoring in adult patients with attention deficit hyperactivity disorder (ADHD). World Journal of Biological Psychiatry, 11(2 Pt 2), 457e464. http://dx.doi.org/10.1080/ 15622970902977552911417996. Hoffmann, S., Wascher, E., & Falkenstein, M. (2012). Personality and error monitoring: an update. Frontiers in Human Neuroscience, 6, 171. http://dx.doi.org/10.3389/ fnhum.2012.00171. Holmes, A. J., & Pizzagalli, D. A. (2008). Spatiotemporal dynamics of error processing dysfunctions in major depressive disorder [Research Support, N.I.H., Extramural Research Support, NonU.S. Gov't] Archives of General Psychiatry, 65(2), 179e188. http:// dx.doi.org/10.1001/archgenpsychiatry.2007.19. Holmes, A. J., & Pizzagalli, D. A. (2010). Effects of task-relevant incentives on the electrophysiological correlates of error processing in major depressive disorder [Research Support,

c o r t e x 6 6 ( 2 0 1 5 ) 1 0 3 e1 1 4

N.I.H., Extramural Research Support, Non-U.S. Gov't] Cognitive, Affective, & Behavioral Neuroscience, 10(1), 119e128. http:// dx.doi.org/10.3758/CABN.10.1.119. Holroyd, C. B., & Coles, M. G. (2002). The neural basis of human error processing: reinforcement learning, dopamine, and the error-related negativity [Clinical Trial Randomized Controlled Trial Research Support, U.S. Gov't, P.H.S.] Psychological Review, 109(4), 679e709. Jasper, H. H. (1958). The ten-twenty electrode system of the International Federation. Electroencephalography and Clinical Neurophysiology, 10, 371e375. Jongen, E. M., Smulders, F. T., Ranson, S. M., Arts, B. M., & Krabbendam, L. (2007). Attentional bias and general orienting processes in bipolar disorder [Research Support, Non-U.S. Gov't] Journal of Behavior Therapy and Experimental Psychiatry, 38(2), 168e183. http://dx.doi.org/10.1016/j.jbtep.2006.10.007. Kroken, R. A., Johnsen, E., Ruud, T., Wentzel-Larsen, T., & Jorgensen, H. A. (2009). Treatment of schizophrenia with antipsychotics in Norwegian emergency wards, a crosssectional national study [Comparative Study Research Support, Non-U.S. Gov't] BMC Psychiatry, 9, 24. http:// dx.doi.org/10.1186/1471-244X-9-24. Leppanen, J. M. (2006). Emotional information processing in mood disorders: a review of behavioral and neuroimaging findings [Research Support, Non-U.S. Gov't Review] Current Opinion in Psychiatry, 19(1), 34e39. http://dx.doi.org/10.1097/ 01.yco.0000191500.46411.00. Mann-Wrobel, M. C., Carreno, J. T., & Dickinson, D. (2011). Metaanalysis of neuropsychological functioning in euthymic bipolar disorder: an update and investigation of moderator variables [Meta-Analysis] Bipolar Disorder, 13(4), 334e342. http://dx.doi.org/10.1111/j.1399-5618.2011.00935.x. Mathalon, D. H., Whitfield, S. L., & Ford, J. M. (2003). Anatomy of an error: ERP and fMRI [Research Support, Non-U.S. Gov't Research Support, U.S. Gov't, Non-P.H.S. Research Support, U.S. Gov't, P.H.S.] Biological Psychology, 64(1e2), 119e141. McGuffin, P., Farmer, A., & Harvey, I. (1991). A polydiagnostic application of operational criteria in studies of psychotic illness: development and reliability of the opcrit system. Archives of General Psychiatry, 48(8), 764e770. http://dx.doi.org/ 10.1001/archpsyc.1991.01810320088015. Montgomery, S. A., & Asberg, M. (1979). A new depression scale designed to be sensitive to change. The British Journal of Psychiatry, 134(4), 382e389. http://dx.doi.org/10.1192/ bjp.134.4.382. Morriss, R. K., van der Gucht, E., Lancaster, G., & Bentall, R. P. (2009). Adult attachment in bipolar 1 disorder. Psychology and Psychotherapy, 82(Pt 3), 267e277. http://dx.doi.org/10.1348/ 147608309X415309. Morris, S. E., Yee, C. M., & Nuechterlein, K. H. (2006). Electrophysiological analysis of error monitoring in schizophrenia. Journal of Abnormal Psychology, 115(2), 239e250. http://dx.doi.org/10.1037/0021-843x.115.2.239. Moser, J. S., Hajcak, G., & Simons, R. F. (2005). The effects of fear on performance monitoring and attentional allocation [Clinical Trial] Psychophysiology, 42(3), 261e268. http:// dx.doi.org/10.1111/j.1469-8986.2005.00290.x. Nieuwenhuis, S., Ridderinkhof, K. R., Blom, J., Band, G. P., & Kok, A. (2001). Error-related brain potentials are differentially related to awareness of response errors: evidence from an antisaccade task [Clinical Trial Research Support, Non-U.S. Gov't] Psychophysiology, 38(5), 752e760. Olvet, D. M., & Hajcak, G. (2008). The error-related negativity (ERN) and psychopathology: toward an endophenotype. Clinical Psychology Review, 28(8), 1343e1354. http://dx.doi.org/10.1016/ j.cpr.2008.07.003. Olvet, D. M., Klein, D. N., & Hajcak, G. (2010). Depression symptom severity and error-related brain activity [Research Support,

113

Non-U.S. Gov't] Psychiatry Research, 179(1), 30e37. http:// dx.doi.org/10.1016/j.psychres.2010.06.008. Overbeek, B. U., van Alfen, N., Bor, J. A., & Zwarts, M. J. (2005). Sural/radial nerve amplitude ratio: reference values in healthy subjects. Muscle & Nerve, 32(5), 613e618. http://dx.doi.org/ 10.1002/mus.20421. Phillips, M. L., Ladouceur, C. D., & Drevets, W. C. (2008). A neural model of voluntary and automatic emotion regulation: implications for understanding the pathophysiology and neurodevelopment of bipolar disorder [Research Support, N.I.H., Extramural Research Support, Non-U.S. Gov't Review] Molecular Psychiatry, 13(9). http://dx.doi.org/10.1038/ mp.2008.65, 829, 833e857. Potts, G. F., George, M. R., Martin, L. E., & Barratt, E. S. (2006). Reduced punishment sensitivity in neural systems of behavior monitoring in impulsive individuals [Comparative Study Research Support, N.I.H., Extramural] Neuroscience Letters, 397(1e2), 130e134. http://dx.doi.org/10.1016/ j.neulet.2005.12.003. Riba, J., Rodriguez-Fornells, A., Morte, A., Munte, T. F., & Barbanoj, M. J. (2005). Noradrenergic stimulation enhances human action monitoring [Clinical Trial Comparative Study Randomized Controlled Trial Research Support, Non-U.S. Gov't] Journal of Neuroscience, 25(17), 4370e4374. http:// dx.doi.org/10.1523/JNEUROSCI.4437-04.2005. Riesel, A., Endrass, T., Kaufmann, C., & Kathmann, N. (2011). Overactive error-related brain activity as a candidate endophenotype for obsessive-compulsive disorder: evidence from unaffected first-degree relatives [Comparative Study] American Journal of Psychiatry, 168(3), 317e324. http:// dx.doi.org/10.1176/appi.ajp.2010.10030416. Ruchsow, M., Herrnberger, B., Beschoner, P., Gron, G., Spitzer, M., & Kiefer, M. (2006). Error processing in major depressive disorder: evidence from event-related potentials. Journal of Psychiatric Research, 40(1), 37e46. http://dx.doi.org/10.1016/ j.jpsychires.2005.02.002. Ruchsow, M., Herrnberger, B., Wiesend, C., Gron, G., Spitzer, M., & Kiefer, M. (2004). The effect of erroneous responses on response monitoring in patients with major depressive disorder: a study with event-related potentials [Clinical Trial] Psychophysiology, 41(6), 833e840. http://dx.doi.org/10.1111/ j.0048-5772.2004.00237.x. Santesso, D. L., & Segalowitz, S. J. (2008). Developmental differences in error-related ERPs in middle- to late-adolescent males [Comparative Study Research Support, Non-U.S. Gov't] Developmental Psychology, 44(1), 205e217. http://dx.doi.org/ 10.1037/0012-1649.44.1.205. Schrijvers, D., de Bruijn, E. R., Maas, Y., De Grave, C., Sabbe, B. G., & Hulstijn, W. (2008). Action monitoring in major depressive disorder with psychomotor retardation [Research Support, Non-U.S. Gov't] Cortex, 44(5), 569e579. http://dx.doi.org/ 10.1016/j.cortex.2007.08.014. Schrijvers, D., De Bruijn, E. R., Maas, Y. J., Vancoillie, P., Hulstijn, W., & Sabbe, B. G. (2009). Action monitoring and depressive symptom reduction in major depressive disorder [Research Support, Non-U.S. Gov't] International Journal of Psychophysiology, 71(3), 218e224. http://dx.doi.org/10.1016/ j.ijpsycho.2008.09.005. Scott, J., & Pope, M. (2003). Cognitive styles in individuals with bipolar disorders. Psychologie Medicale, 33(6), 1081e1088. Segalowitz, S. J., & Dywan, J. (2009). Individual differences and developmental change in the ERN response: implications for models of ACC function [Research Support, Non-U.S. Gov't Review] Psychological Research, 73(6), 857e870. http:// dx.doi.org/10.1007/s00426-008-0193-z. Sheehan, D. V., Lecrubier, Y., Sheehan, K. H., Amorim, P., Janavs, J., Weiller, E., et al. (1998). The Mini-International Neuropsychiatric Interview (M.I.N.I.): the development and

114

c o r t e x 6 6 ( 2 0 1 5 ) 1 0 3 e1 1 4

validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. Journal of Clinical Psychiatry, 59 (0160e6689 (Print)), 34e57. Steele, J. D., Meyer, M., & Ebmeier, K. P. (2004). Neural predictive error signal correlates with depressive illness severity in a game paradigm [Comparative Study Research Support, NonU.S. Gov't] NeuroImage, 23(1), 269e280. http://dx.doi.org/ 10.1016/j.neuroimage.2004.04.023. Strakowski, S. M., Delbello, M. P., & Adler, C. M. (2005). The functional neuroanatomy of bipolar disorder: a review of neuroimaging findings. Molecular Psychiatry, 10(1), 105e116. http://dx.doi.org/10.1038/sj.mp.40015854001585. Teasdale, J. D. (1983). Change in cognition during depressionepsychopathological implications: discussion paper [Research Support, Non-U.S. Gov't] Journal of the Royal Society of Medicine, 76(12), 1038e1044. Tops, M., & Boksem, M. A. (2010). Absorbed in the task: personality measures predict engagement during task performance as tracked by error negativity and asymmetrical frontal activity [Research Support, Non-U.S. Gov't] Cognitive, Affective, & Behavioral Neuroscience, 10(4), 441e453. http:// dx.doi.org/10.3758/CABN.10.4.441. Van der Gucht, E., Morriss, R., Lancaster, G., Kinderman, P., & Bentall, R. P. (2009). Psychological processes in bipolar affective disorder: negative cognitive style and reward processing [Multicenter Study Research Support, Non-U.S. Gov't] British Journal of Psychiatry, 194(2), 146e151. http:// dx.doi.org/10.1192/bjp.bp.107.047894.

Weinberg, A., Olvet, D. M., & Hajcak, G. (2010). Increased errorrelated brain activity in generalized anxiety disorder. Biological Psychology, 85(3), 472e480. http://dx.doi.org/10.1016/ j.biopsycho.2010.09.011. Wessa, M., & Linke, J. (2009). Emotional processing in bipolar disorder: behavioural and neuroimaging findings [Research Support, Non-U.S. Gov't Review] International Review of Psychiatry, 21(4), 357e367. http://dx.doi.org/10.1080/ 09540260902962156. Woods, S. W. (2003). Chlorpromazine equivalent doses for the newer atypical antipsychotics [Research Support, Non-U.S. Gov't Research Support, U.S. Gov't, P.H.S. Review] Journal of Clinical Psychiatry, 64(6), 663e667. Yeung, N., Botvinick, M. M., & Cohen, J. D. (2004). The neural basis of error detection: conflict monitoring and the error-related negativity [Clinical Trial Controlled Clinical Trial] Psychological Review, 111(4), 931e959. http://dx.doi.org/10.1037/0033295X.111.4.939. Young, R. C., Biggs, J. T., Ziegler, V. E., & Meyer, D. A. (1978). A rating scale for mania: reliability, validity and sensitivity. The British Journal of Psychiatry, 133(5), 429e435. http://dx.doi.org/ 10.1192/bjp.133.5.429. Zirnheld, P. J., Carroll, C. A., Kieffaber, P. D., O'Donnell, B. F., Shekhar, A., & Hetrick, W. P. (2004). Haloperidol impairs learning and error-related negativity in humans [Clinical Trial Comparative Study Randomized Controlled Trial] Journal of Cognitive Neuroscience, 16(6), 1098e1112. http://dx.doi.org/ 10.1162/0898929041502779.