Psychiatry Research: Neuroimaging 259 (2017) 42–53

Contents lists available at ScienceDirect

Psychiatry Research Neuroimaging

journal homepage: www.elsevier.com/locate/psychresns

Decreased response inhibition to sad faces during explicit and implicit tasks MARK
in females with depression: Evidence from an event-related potential study
Fengqiong Yua,b,c,1, Xiaoqing Zhoud,1, Wu Qinga,c, Dan Lib,c, Jing Lia, Xingui Chena,c,
⁎ ⁎
Gongjun Jia,c, Yi Dongc,d, Yuejia Luoe, Chunyan Zhua,c, , Kai Wanga,b,c,
a
Laboratory of Cognitive Neuropsychology, Department of Medical Psychology, Anhui Medical University, Hefei, China
b
Department of Neurology, The First Affiliated Hospital of Anhui Medical University, Hefei, China
c
Collaborative Innovation Centre of Neuropsychiatric Disorders and Mental Health, Anhui Province, China
d
Anhui Mental Health Center, Hefei, China
e
Institute of Social and affective Neuroscience, Shenzhen University, Shenzhen, China

A R T I C L E I N F O A BS T RAC T

Keywords: The present study aimed to investigate neural substrates of response inhibition to sad faces across explicit and
Depression implicit tasks in depressed female patients. Event-related potentials were obtained while participants performed
Response inhibition modified explicit and implicit emotional go/no-go tasks. Compared to controls, depressed patients showed
Implicit decreased discrimination accuracy and amplitudes of original and nogo-go difference waves at the P3 interval in
Explicit
response inhibition to sad faces during explicit and implicit tasks. P3 difference wave were positively correlated
Emotion
P3
with discrimination accuracy and were independent of clinical assessment. The activation of right dorsal
prefrontal cortex was larger for the implicit than for the explicit task in sad condition in health controls, but was
similar for the two tasks in depressed patients. The present study indicated that selectively impairment in
response inhibition to sad faces in depressed female patients occurred at the behavior inhibition stage across
implicit and explicit tasks and may be a trait-like marker of depression. Longitudinal studies are required to
determine whether decreased response inhibition to sad faces increases the risk for future depressive episodes
so that appropriate treatment can be administered to patients.

1. Introduction social interactions, inhibiting inappropriate negative emotions con-

Depression is one of the most common psychiatric disorders and a
leading cause of disability worldwide (Goodwin et al., 2006). Studies
suggest that at least 20% of adolescents experience a depressive
episode before adulthood, which is associated with high morbidity
and mortality (Brent and Birmaher, 2006). Women show a younger age
of onset and higher rate of recurrence of depression than men (Kessler
et al., 2005; Schuch et al., 2014). Repeated negative thoughts are
considered a hallmark of depression (Koster et al., 2011), and are
related to impaired cognitive control over negative information
(Ochsner and Gross, 2008; Wager et al., 2008). Response inhibition
is an important component of this cognitive control system; depressed
individuals are unable to inhibit their response to negative information
(De Raedt and Koster, 2010; Gotlib and Joormann, 2010), which is also
observed in unmedicated offspring of depressed mothers (Joormann
et al., 2007).
Sad facial expressions are one type of negative social stimulus. In
veyed by sad faces is an indispensible social skill. Depression patients
show attentional bias to sad faces and are unable to prevent being
troubled by sad expressions (Gollan et al., 2008; Gur et al., 1992;
Sylvester et al., 2015). In one study, depressed adolescents and those
who developed depression at follow-up made more commission errors
for sad than for happy faces (Kilford et al., 2015). Moreover,
unmedicated offspring of depressed mothers showed greater pupil
dilation compared to controls, which was associated with greater
depression risk (Burkhouse et al., 2015). Neuroimaging studies have
found that depressed individuals exhibit higher activity in the amyg-
dala in response to sad faces, whereas control subjects show higher
activity in response to happy faces (Suslow et al., 2010). Depression
patients also show less differential activation of the insular cortex in
response to sad vs. happy faces as compared to healthy controls; the
difference in activation is negatively correlated with depression severity
(Henje Blom et al., 2015). These results suggest that depressed
individuals show facilitated negative bias in the processing of sad

⁎
Corresponding authors at: Laboratory of Cognitive Neuropsychology, Department of Medical Psychology, Anhui Medical University, Hefei, China.
E-mail addresses: zhuchunyan@163.com (C. Zhu), wangkai1964@126.com (K. Wang).
1
These authors contribute equally to this work.

http://dx.doi.org/10.1016/j.pscychresns.2016.10.013
Received 25 February 2016; Received in revised form 22 October 2016; Accepted 27 October 2016
Available online 08 December 2016
0925-4927/ © 2016 Elsevier Ireland Ltd. All rights reserved.
F. Yu et al.
faces. Exerting cognitive control over sad emotions is critical for the
remission of depression, and understanding the neural correlates of
response inhibition to sad faces in depression can enable the initiation
of appropriate interventions.
In explicit processing, facial expressions are within the scope of
voluntary attention and are directly processed, whereas in implicit
processing they occur outside the scope of voluntary attention and are
incidentally processed. As such, attentional resources for stimulus
processing differ between the two conditions. Implicit and explicit
facial processing serve different functions (Taylor et al., 2003) and have
distinct neural substrates (Linden et al., 2010; Valdes-Conroy et al.,
2014; Winston et al., 2003), and subjects have reported different
emotional intensities associated with facial expressions processed
explicitly vs. implicitly. For instance, rating pictures was associated
with reduced intensity of sadness as compared to passive viewing,
likely because the rating task reduced the activation of brain regions
related to emotional experiences (Taylor et al., 2003). Psychiatric
patients show different responses to explicit and implicit emotional
stimuli. For example, schizophrenia and depression patients exhibit
greater automatic amygdala responses to implicit sad faces than
controls, in the former, these responses were positively correlated with
the negative subscale of the Positive and Negative Syndrome Scale
(Rauch et al., 2010). We previously found that response inhibition was
modulated by sad facial information at the action inhibition stage when
facial expressions are processed explicitly as opposed to implicitly (Yu
et al., 2014). Moreover, individuals with generalized anxiety disorder
showed impaired response inhibition to sad faces in an explicit but not
in an implicit task (Yu et al., 2015). However, it is not known whether
depression patients exhibit deficits in response inhibition across
implicit and explicit conditions.
To address this issue, we developed a modified emotional go/no-go
paradigm consisting of two sections. In the task-related section,
participants made their go/no-go decision according to recognition of
facial expressions; that is, the emotional expression was explicitly
processed. In the task-irrelevant section, participants responded or
inhibited their response based on identification of the gender of the
face; that is, the emotional processing was implicit. We used the same
set of stimuli in the two sections to exclude interference from
additional factors. We hypothesized that response inhibition deficits
in depression are reflected by lower discriminating accuracy and
shorter response time to sad stimuli across explicit and implicit tasks.
Event-related potentials (ERPs) were recorded during the evalua-
tion of behavioral measures. The main advantage of this non-invasive
method is the high time resolution that reflects the time course of brain
activity (Otten and Rugg, 2004); it has been frequently used to explore
the neural basis of response inhibition (Fabiani et al., 2000), which
comprises two different cognitive processes—namely, conflict monitor-
ing and action inhibition (Sehlmeyer et al., 2010). No-go N2 reflects the
former, while no-go P3 is associated with conflict resolution and
behavioral inhibition (Donkers and van Boxtel, 2004; Kropotov et al.,
2011). Neuroimaging studies have revealed that the right inferior
prefrontal cortex (rIFC) plays a crucial role in response inhibition to
emotional stimuli (Berkman et al., 2009; Goldstein et al., 2007;
Ochsner et al., 2004; Padmala and Pessoa, 2010; Shafritz et al.,
2006); ERP measurements have shown that response inhibition is
impaired in individuals with anxiety (Hum et al., 2013). Depression
patients showed decreased no-go N2 amplitudes in a modified auditory
go/no-go task (Kaiser et al., 2003; Katz et al., 2010). However, another
study in which participants performed a cued emotional conflict task
revealed differences between depression patients and controls only
during the late N450 time window (Vanderhasselt et al., 2012). These
inconsistent findings may be due to variations in experimental para-
digms and patient inclusion criteria in these studies. Although depres-
sion disrupts response inhibition, the neural basis of this effect is not
well understood. It has been suggested that depression is linked to a
reduced ability to monitor conflict or to apply active inhibition (Aarts
43
Psychiatry Research: Neuroimaging 259 (2017) 42–53
and Pourtois, 2010; Berggren and Derakshan, 2013; Botvinick et al.,
2001).
The present study investigated the neural substrates of response
inhibition to sad faces across implicit and explicit tasks in depression
patients based on ERP recordings. We hypothesized that depression
patients would show deficits in response inhibition to sad faces across
both tasks, and would therefore show decreased accuracy in discerning
sad faces as well as a shorter response time to sad faces relative to
control subjects. For ERP recordings, we expected decreased N2 and
P3 amplitudes in inhibition-related brain areas such as the rIFC in both
tasks. We also examined the correlation between behavioral measures,
ERP components, and clinical parameters to determine whether
decreases in N2 and P3 are markers of depression. We recruited only
women in this study since response inhibition and patterns of brain
activation show gender differences (Yuan et al., 2009) and the
incidence of depression is higher among females (Kessler et al.,
2005; Schuch et al., 2014).
2. Materials and methods
2.1. Participants
Female unipolar depression outpatients (n=20; age: 17–46 years)
were recruited from the depression disorder clinic in Anhui Mental
Health Center. Diagnostic assessments were independently completed
by two psychiatrists based on DSM-IV criteria. DSM-IV is the fourth
edition of the Diagnostic and Statistical Manual of Mental Disorders
(DSM) which has been praised for standardizing psychiatric diagnostic
categories and criteria (American Psychiatric Association, 1994). The
manual includes all currently recognized mental health disorder and
thus is usually used by mental health professionals to describe the
characteristics of a given mental disorder to distinguish the disorder
from other, similar problems. In addition, patients included in the
study scored > 14 on the Hamilton Depression Rating Scale (HAMD)-
17 which is used for adults and to rate the depression severity by
probing mood, feeling of guilt, suicide ideation, agitation, retardation,
anxiety, weight loss, and somatic symptom. The scale showed high
inter-rater reliability as 0.9 for 70 patients (Hamilton, 1960). Exclusion
criteria were as follows: any history of mania, psychosis or any other
Axis I disorders; a history of neurological conditions; substance abuse;
and current use of psychotropic medication. Patients had never had
behavioral or drug treatment for depression or had suspended treat-
ment for at least two months prior to the study. Age-matched female
adults (n=21) screened for current and past psychiatric and neurolo-
gical disorders were recruited through internet postings. All partici-
pants had normal or corrected-to-normal vision and HAMD-17 scores
< 7. Control tests and characteristics of the study population are shown
in Table 1. This study was carried out in accordance with the
Declaration of Helsinki and was approved by the Research Ethics
Committee of Anhui Medical University. All participants completed the
experiment on a voluntarily basis with receiving RMB50 yuan.
Table 1
Group characteristics of the DEP group and CON group.
DEP CON Between group
Group Group comparison
Mean (SD) Mean (SD) P-value
Age(years) 30.6(9.4) 25.9(6.2) t39=1.84, P=0.084
Education(years) 13.7(3.4) 13.9(0.9) t39=0.25, P=0.80
BDI 18.5(4.5) 4.9(2.1) t39=12.42, P < 0.001
SAI 50.2(10.3) 36.0(6.9) t39=5.37, P < 0.001
TAI 57.3(8.5) 34.4(6.7) t39=9.64, P < 0.001
Abbreviations: DEP, depression; CON, control; BDI, Beck Depression Inventory. SAI,
State Anxiety Inventory; TAI, Trait Anxiety Inventory.
F. Yu et al.
2.2. Materials
Stimuli used in the present study have been described in our
previous work (Yu et al., 2014). Sad and neutral faces (n=40 each) were
selected from the native Chinese Facial Affective Picture System and
have been assessed on a 9-point scale by 100 college students from two
colleges in Beijing and used in many studies (Fan et al., 2015; Luo
et al., 2010). In the present study, 20 female and 20 male faces
displaying each type of emotion were presented to subjects. The
stimulus were selected such that they differed significantly in valence
from one another [t=11.65, P < 0.001 (M ± SD, sad: 3.11 ± 0.63,
neutral: 4.49 ± 0.41)], but were similar in terms of arousal (P > 0.05).
Stimuli were also similar in size, spatial frequency, brightness, back-
ground, contrast grade, and other physical properties. Each picture was
cropped using Adobe Photoshop v.8.0 software into an elliptical shape
that incorporated facial characteristics. The screen resolution was72
pixels per inch, and the viewing angle was 5.7°×4.6°. All stimuli were
displayed in the center of a 17-in. screen. Subjects were seated in a
soundproof room with their eyes approximately 100 cm away from the
screen.
2.3. Experimental procedures
The experimental procedure was similar to that of previous studies
(Yu et al., 2014, 2015) and was developed using E-Prime software
(Psychology Software Tools, Pittsburgh, PA, USA). The experiment
included two parts: the implicit and explicit tasks. In the former,
participants were instructed to respond to the presentation of faces
depicting one gender (go trials) and to inhibit response after presenta-
tion of the other gender (no-go trial). In the explicit task, participants
were asked to respond or inhibit their response according to facial
expression category. There were 480 trials that included 144 no-go
stimuli and 336 go stimuli (30% vs. 70%) in implicit and explicit tasks,
which were presented in two separate parts, with the order of the parts
counterbalanced across participants. In addition, each part was sub-
divided into two blocks. Facial stimuli were counterbalanced in terms
of whether they were indicated as go or no-go trials in each part. Thus,
the experimental procedure consisted of four blocks: sad-go/neutral-
no-go and neutral-go/sad-no-go in the explicit task, and female-go/
male-no-go and male-go/female-no-go in the implicit task. In each
block, go and no-go stimuli were presented pseudo-randomly, and a
Fig. 1. Experimental design for (A) explicit and (B) implicit emotional go/no-go tasks. In the explicit task, subjects pressed a response button or inhibited their behavior according to the
facial expression (sad/neutral). In the implicit task, subjects responded to gender (male/female).
44
Psychiatry Research: Neuroimaging 259 (2017) 42–53
no-go trial was always preceded by a go trial. This was done in order to
induce pre-potent motor responses and conflict during response
inhibition. At the start of each block, an instruction screen that lasted
for 5 min was presented in order to prompt participants to press or
refrain from pressing the “J” key with their right hand according to the
facial expression or gender. Each trial was initiated by a small white
cross that was displayed for a variable duration (200–400 ms) on the
black background. An emotional face then appeared at the center of the
screen for 1000 ms, and participants were asked to respond as quickly
and accurately as possible. The inter-trial interval which means the
time interval from stimulus offset to stimulus onset was 1200–
1500 ms. The experimental procedure is outlined in Fig. 1. The
subjects in this experiment have given written informed consent to
publish these case details. A practice session consisting of 20 trials was
administered before the experiment using stimuli that differed from
those used in the actual experiment.
2.4. Event-related potential recording
A Geodesic Sensor Net System (Neuro Scan, Sterling, VA, USA) was
used to record 64-channel electroencephalography (EEG) according to
the international 10/20 system within an electrically and acoustically
shielded room. The system was grounded with a forehead electrode. All
EEG signals were referenced to the left mastoid and were re-referenced
off-line to the average of the left and right mastoids. Vertical electro-
oculogram (EOG) data were recorded supraorbitally and infraorbitally
in the left eye. Horizontal EOG data were recorded as the left versus
right orbital rim. EEG and EOG activity was amplified with a 0.01–
100 Hz band-pass filter and continuously sampled at 500 Hz/channel.
All electrode impedances were maintained below 5 kΩ. Ocular artifacts
were removed from the EEG signals using a regression procedure
implemented in Neuroscan software. Trials with remaining EOG
artifacts, amplifier clipping artifacts, or peak-to-peak deflections ex-
ceeding ± 100 μV were excluded from averaging. The EEG activities
during correct responses in each condition were aligned and averaged
separately. The ERP waveforms were time-locked to the onset of the
face stimuli, and the average epoch was 1200 ms, including a 200 ms
pre-stimulus base-line.
F. Yu et al.
2.5. Data analysis
The independent variables in the experiment were task (explicit/
implicit), valence (sad/neutral), trial type (go/no-go), and group
(patient/health control). Data were evaluated by a mixed measures
multivariate analysis of variance. The degrees of freedom of F ratios
were adjusted according to the Greenhouse–Geisser epsilon correction
in all analyses. The Bonferroni method was used for multiple compar-
isons.
2.5.1. Analysis of behavioral results
Signal detection theory was applied to the analysis of behavioral
results in order to distinguish signal from noise (Snodgrass and
Corwin, 1988); participants’ discrimination accuracy d' (distance
between signal and noise distributions) and decision bias β (tendency
to respond to signal and noise distributions) were assessed based on
this theory. d' was defined as d′=(μs−μn)/σ, where μs and μn are signal
and noise distributions, respectively, and σ is the common standard
deviation of both distributions. d' was calculated as d′=Zhit rate−Zfalse
alarm rate, which had four possible outcomes: hit (signal present and
subject's response is ‘yes’), false alarm (signal absent and subject's
response is ‘yes’), miss (signal present and subject's response is ‘no’),
and correct rejection (signal absent and subject's response is ‘no’). β
was defined as β=fs(λ)/fn(λ), where fs(λ) is the height of the signal
distribution at a given criterion λ, and fn(λ) is the height of the noise
distribution at the same λ; the value was calculated as β= exp (d′×C),
where C=–(Zhit rate+Zfalse alarm rate)/2. A three-way mixed design
analysis of variance (ANOVA) of the amplitudes of each component
was carried out with task (two levels: implicit and explicit), valence
(two levels: neutral and sad) as within-subject factors, and group
(depression and control) as a between-subject factor.
2.5.2. ERP measurement and analysis
According to the topographical distribution of grand-averaged ERP
activity and previous studies (Righart and de Gelder, 2007; Yu et al.,
2009; Yuan et al., 2008), nine electrode sites including F3, Fz, F4, C3,
Cz, C4, P3, Pz, and P4 were selected for statistical analysis of N2 (250–
350 ms) and P3 (500–600 ms) mean amplitudes. A five-way mixed
design analysis of variance (ANOVA) of the amplitudes of each
component was carried out with task (two levels: implicit and explicit),
valence (two levels: neutral and sad), trial type (two levels: go and no-
go), and electrode as within-subject factors, and group (depression and
control) as a between-subject factor.
2.5.3. Source localization analysis
Standardized low-resolution brain electromagnetic tomography
analysis (sLORETA) was used to determine the neural substrates of
response inhibition (Pascual-Marqui, 2002). Using this method, we
estimated current density distributions in the cortical gray matter and
hippocampus in the digitized Montreal Neurological Institute atlas with
6, 239 voxels at a spatial resolution of 5 mm. sLORETA assumes that
neighboring neuronal sources active similarity. The assumption was
implemented by computing the ‘smoothest’ of all possible activity
Table 2
Averaged reaction times for Go stimuli and the response accuracy rates for Go and Nogo trials in explicit and implicit tasks in each group (M ± SD).
Explicit task
Negative Neutral
DEP CON DEP
go ACC (%) 76.6 (23.9) 92.8 (6.8) 86.6 (14.7)
nogo ACC (%) 89.3 (17.8) 95.8 (3.1) 93.6 (4.8)
go RT (ms) 470.3 (154.9) 522.6 (91.6) 560.1 (99.9)
Abbreviations: DEP, depression; CON, control; ACC, accuracy rate; RT, response time.
Psychiatry Research: Neuroimaging 259 (2017) 42–53
distributions. Yet, the algorithm has been conformed to be cross-modal
validity (Pizzagalli et al., 2002). Studies have shown that sLORETA
results are consistent with those obtained by MRI (magnetic resonance
imaging) and functional MRI as well as electrocorticography from
subdural electrodes (Seeck et al., 1998; Worrell et al., 2000). However,
as the limited spacial resolution of sLORETA, the results should be
confirmed by other brain imaging methodologies with high spatial
resolution.
First, To detect the neural mechanism of response inhibition to
facial expression, voxel-based whole brain sLORETA images were
compared between go and no-go conditions; The sLORETA built-in
voxel-wise randomization tests (5000 permutations) based on statis-
tical non-parametric mapping methodology were used to yield a whole-
brain statistical correction (Nichols and Holmes, 2002). According to
previous sLORETA studies, to reduce the likelihood of false positives,
we set the nominal significance threshold to P < 0.005, with minimal
cluster at least 5 significant contiguous voxels (Mueller and Pizzagalli,
2016). Second, the current source density (CSD) of the clusters that
were significant in the first step would be subjected to ANOVA in order
to determine inter-group differences and assess the modulatory effects
of task and valence. Due to limited spatial resolution of sLORETA, we
extracted the mean CSD for all contiguous significant voxels of the
clusters for further statistical analysis to increase the credibility of the
results.
2. Results
2.1. Behavioral performance
Subjects’ response accuracy and time are presented in Table 2. An
analysis of accuracy by signal detection theory revealed that the main
effect of group was significant (F1,39=5.67, P < 0.05). Depression
patients showed smaller d' as compared to controls when inhibiting
inappropriate responses. The interaction effect of group and valence
was also significant (F1,39=4.80, P < 0.05) (see Fig. 2). Further analysis
revealed that the group effect was significant for sad faces (F1,39=6.86,
P < 0.05). Depression patients had significantly smaller d' values than
controls when asked to respond to neutral faces and inhibit response to
sad faces, indicating their diminished ability to suppress negative
information. However, the inter-group difference in the neutral condi-
tion was not significant (F1,39=0.11, P=0.74), and the difference in d'
between the two groups was not significant in the other conditions. d'
and β are independent variables, and therefore discrimination accuracy
does not correlate with decision bias. Discrimination accuracy in
response to neutral faces and inhibition of responses to sad faces were
impaired as indicated by d', but decision bias (as measured by β) was
not significant (F1,39=0.037, P=0.85). The mixed ANOVA of response
time showed a significant main effect for group (F1, 39=6.01, P < 0.05)
(see Fig. 2), with a shorter response time observed for the depression
than for the control group.
Implicit task
Negative Neutral
CON DEP CON DEP CON
89.1 (9.4) 92.5 (8.1) 96.2 (4.9) 91.8 (6.3) 97.5 (2.9)
95.5 (3.0) 93.1 (4.9) 92.6 (6.9) 94.6 (2.3) 94.3 (3.2)
526.1 (44.9) 608.1 (115.9) 515.9 (40.7) 589.5 (80.1) 519.3 (38.5)
45
F. Yu et al.
Fig. 2. Discrimination accuracy (left panel) compared between depression (gray bar) and control (black bar) groups during sad nogo/neutral go (left group) and neutral nogo/sad go
(right group) tasks and response time (right panel) compared between depression (left group) and control (right group) groups in sad (black bar) and neutral (gray bar) conditions. *, P <
0.05.
2.2. Raw ERP waveforms
ERP waveforms time-locked to go and no-go stimuli were averaged
(Fig. 3). As expected, no-go stimuli elicited larger N2 and P3
components than go stimuli. Task, valence, trial type, and group effects
are described below. The average N2 and P3 scores for the four factors
are shown in Table 3.
2.2.1. . N1
The five-way mixed ANOVA showed that valence, group and
electrodes main effects were significant on N1 amplitudes
(F1,39=5.49, P < 0.05; F1,39=8.96, P=0.005; F8,312=150.36, P <
0.0001). Neutral stimuli elicited larger N1 amplitudes than sad stimuli.
Depression group showed smaller amplitudes than control group. N1
amplitudes, which were largest at FCz electrode sites, were mainly
distributed in central prefrontal areas. No any other main effect or
interaction effect was detected during this stage.
2.2.2. . N2
The five-way mixed ANOVA showed that N2 amplitudes were
modulated by trial type (F1,39=20.88, P < 0.001), valence
(F1,39=24.36, P < 0.001), and electrode (F8,312=15.20, P < 0.001). No-
go stimuli elicited larger N2 amplitudes than go stimuli, which were
smaller in the sad than in the neutral condition. N2 amplitudes, which
were largest at FCz electrode sites, were mainly distributed in central
prefrontal areas. In addition, trial type and group interaction effects
were significant (F1,39=8.76, P=0.005) (Fig. 4); the former was
significant in the control (F1,20=18.96, P < 0.001) but not in the
depression group (F1,19=2.88, P=0.11). There was no any other main
effect or interaction effect that was detected between this time
windows.
2.2.3. . P3
The five-way mixed ANOVA of P3 amplitudes showed significant
main effects of trial type, valence, group, and electrode (F1,39=9.29, P <
0.01; F1,39=7.56, P < 0.01; F8,312=46.46, P < 0.001; F1,39=14.59, P <
0.001). P3 amplitudes were larger in the sad than in the neutral
condition. The depression group showed smaller P3 amplitudes than
controls. P3 amplitudes induced by no-go stimuli were larger than
those induced by go stimuli, with Pz eliciting the largest amplitudes. Cz,
C4, P3, and P4 elicited larger amplitudes than the other electrode sites.
The interaction effect between valence and trial type was significant
(F1,39=8.84, P=0.005). The valence effect was significant in nogo
condition (F1,39=15.12, P < 0.001). P3 amplitudes induced by sad
stimuli were significantly larger than that induced by neutral stimuli.
However, the valence effect was not significant in go condition
46
Psychiatry Research: Neuroimaging 259 (2017) 42–53
(F1,39=0.02, P=0.89). In addition, the interaction effect between
valence, trial type and group effect was significant (F=4.46, P < 0.05)
(Fig. 4). Further simple analysis showed that valence and trial type
interaction effect was significant in control group (F=8.16, P < 0.05).
The emotion main effect was significant in nogo condition (F=13.63,
P=0.001) but not in go condition (F=0.02, P=0.89). P3 amplitudes
induced by sad stimuli were significantly larger than that induced by
neutral stimuli in nogo condition in control group. However, the
interaction effect between valence and trial type was not significant
in depression group (F=1.06, P=0.32). Notably, the interaction effect
between task, valence, group, and trial type was also marginally
significant (F1,39=3.92, P=0.055). To clearly illustrate this effect, we
calculated go/no-go difference waves to determine the dissimilarity
between depression and control groups. No any other main effect or
interaction effect that was detected between this time windows.
2.3. Go/no-go differences ERP waveforms
As described above, to clearly illustrate the interaction effect
between task, valence, group, and trial type during P3 interval and
interaction effect between group and trial type during N2 interval,
amplitudes induced by go trials were directly subtracted from ampli-
tudes induced by nogo trials for each valence condition to more
precisely determine go/nogo difference waves.
2.3.1. N2 difference wave
The main effects of group (F1,39=6.64, P < 0.05) and electrode
(F8,312=5.56, P=0.001) were significant. N2 amplitudes were smaller
in the depression than in the control group and were larger at FC4 and
C4 than at other electrode sites.
2.3.2. P3 difference wave
The analysis of P3 difference waves showed significant main effects
of valence (F1,39=13.55, P=0.001), group (F1,39=7.69, P < 0.01), and
electrode (F8,312=13.81, P < 0.001). Sad faces elicited larger P3 ampli-
tudes than neutral faces, and the depression group showed smaller P3
amplitudes than the control group. These were larger at FCz and Cz
than at other electrode sites. We also detected an valence and group
interaction effect (F1,39=6.72, P < 0.05). The group effect was signifi-
cant in the sad condition (F1,39=5.27, P < 0.05). The depression group
showed lower P3 amplitudes than the control group, but the inter-
group difference was not significant in the neutral condition
(F1,39=0.007, P=0.935). Importantly, we found a significant task,
valence, and group interaction effect (F1,39=4.27, P < 0.05) (Fig. 5),
and the valence and group interaction effect was significant in the
explicit task (F1,39=8.10, P < 0.01). A group effect was also detected in
F. Yu et al. Psychiatry Research: Neuroimaging 259 (2017) 42–53

Fig. 3. Grand averages evoked by go (dashed line) and nogo (solid line) stimuli in sad and neutral conditions under implicit and explicit tasks in the control (red line) and depression
(blue line) groups at FCz and Cz sites. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

the sad condition during the explicit task (F1,39=5.68, P < 0.05). P3
amplitudes in the control group were larger than in the depression
group in the sad condition during the explicit task, while the group
effect was not significant in the neutral condition during this task.
However, in the implicit task, the valence and group interaction effect
was not significant (F1,39=0. 61, P=0.44).
2.3.3. Relationship between clinical characteristics and task-related
measures
Discrimination accuracy was positively correlated with amplitudes
of go/no-go difference waves in the P3 time window at FCz and Cz
electrode sites during the sad condition in the depression group
(r=0.45, P < 0.05). Discrimination accuracy and P3 amplitudes were

47
F. Yu et al. Psychiatry Research: Neuroimaging 259 (2017) 42–53

Table 3
Amplitudes in original ERP components of the study participants (M ± SD).

Explicit task Implicit task

Sad Neutral Sad Neutral

Go Nogo Go Nogo Go Nogo Go Nogo

N2(DEP) 3.7(2.6) 3.4(2.3) 2.9(1.9) 2.7(2.5) 2.7(2.1) 2.9(2.1) 2.9(2.5) 2.4(2.5)

N2(CON) 3.3(3.3) 3.5(2.9) 3.2(3.1) 1.7(3.1) 2.9(2.0) 2.8(2.5) 3.8(3.4) 2.2(2.3)
P3(DEP) 5.4(2.8) 7.0(3.4) 5.1(2.4) 6.2(3.1) 4.3(2.3) 6.4(3.5) 4.4(2.5) 6.3(3.5)
P3(CON) 7.0(5.2) 10.5(4.0) 7.3(4.0) 7.8(3.4) 6.9(3.7) 10.4(4.9) 6.7(3.8) 9.7(4.0)

Amplitudes in microvolt. All values expressed in mean (SD). Abbreviations: DEP, depression; CON, control.

not correlated in the other conditions. There were no correlations
between clinical measures (HAMD) and discrimination accuracy or
between clinical measures and P3 components in the sad condition
during the explicit task.
than in explicit task. The task main effect was not significant in neutral
condition in control group (F1,39=0.11, P=0.92).
3. Discussion

2.3.4. Source localization results
Based on ERP results that showed significant task, valence, group,
and trial type interaction effects during the P3 stage, voxel-based whole
brain sLORETA images were compared between go and no-go condi-
tions at this stage with non-parametric randomization tests. As
predicted, a cluster in the right dorsolateral prefrontal cortex
(rDLPFC; BA 8/9/10: maximum values obtained at MNI coordinates:
x=25, y=40, z=35) showed significantly enhanced activity (t39=0.87, P
< 0.005; k=54 voxels, volume: 6.75 cm3) in nogo trials compare to go
trials during this stage (Fig. 6). To evaluate experimental effects, the
CSD of the cluster was extracted using coordinates from voxels showing
the largest t value in the previous sLORETA comparison as centroid
point and 22.5 mm as radius which was computed from the volume of
the significant cluster (6.75 cm3). A significant main effect of trial type
was observed (F1,39=36.69, P < 0.001). The CSD was higher in the no-
go condition than in the go condition. We detected significant interac-
tion effect between task and group (F1,39=6.23, P < 0.05), and the task
effect was significant in the control (F1,20=4.76, P=0.041) but not in the
depression (F1,19=2.15, P=0.16) group. The CSD of rDLPFC was larger
in the implicit than in the explicit task in the control group. In addition,
the interaction effect between task, valence, and group was significant
(F1,39=8.85, P < 0.05) (Fig. 6). Further analysis showed that the task
and valence interaction effect was significant in control group
(F1,39=5.04, P < 0.05), but not in depression group (F1,39=3.812,
P=0.083). Task main effect was significant in sad condition in control
group (F1,39=13.59, P=0.001). The CSD was higher in implicit task
This study investigated the neural substrates of response inhibition
deficits to emotionally sad information in female depressive individuals
across explicit and implicit tasks using high-resolution ERP measure-
ment. The findings suggest that depression patients showed decreased
go/no-go difference waves during the P3 stage when inhibiting
responses to sad faces across implicit and explicit tasks. In addition,
the P3 difference wave was positively correlated with discrimination
accuracy in the depression group. When inhibiting neutral faces, these
individuals showed decreased P3 difference waves in implicit tasks in
which neutral faces were ambiguous and were likely to be recognized as
negative information. However, in explicit tasks, we did not detect this
difference between the two groups. The source localization analysis on
the P3 component showed that no-go trials activated the rDLPFC to a
greater degree than go trials. In addition, the CSD of the rDLPFC was
higher in the implicit than in the explicit task in the control as
compared to the depression group, confirming that this is a key area
for top-down cognitive control that plays a critical role in the
pathophysiology of depression.
As shown in Fig. 3, at the 40–140 ms interval, the early N1
components were induced by the facial stimuli. Numerous ERP studies
confirmed that N1 component over frontal areas were related to facial
expressions (Eimer and Holmes, 2002; Holmes et al., 2003). The
frontal N1 is related to rapid attentional alerting to facial expressions
(Eimer et al., 2008). Negative faces elicited larger N1 amplitudes
compared with neutral faces at frontal areas (Association, 1994; Luo
et al., 2010; Yang et al., 2015). However, in the present study, we

Fig. 4. Histograms described the original N2 (A) and P3 amplitudes (B). (A) The original N2 amplitudes compared between nogo (black bar) and go (gray bar) conditions in depression
(left group) and control (right group) groups. (B) The original P3 amplitudes compared between sad (black bar) and neutral (gray bar) conditions in depression and control groups
during nogo (left panel) and go (right panel) tasks. CON, control; DEP, depression; *, P < 0.05; **, P < 0.001.

48
F. Yu et al.
Fig. 5. P3 difference waves of no-go subtracted from go trials at the FCz site (upper panel) and corresponding scalp topography (middle panel) and amplitude histograms (bottom panel)
of the two groups in sad and neutral condition during implicit and explicit tasks. dep, depression; con, control; im, implicit; ex, explicit.
detected that N1 amplitudes were larger in neutral condition than in
sad condition. The inconsistent results may be due to the different task
sets in present study. In previous facial expression studies, participants
were asked to discern facial expressions or passive viewing (Eimer and
Holmes, 2002; Luo et al., 2010), while in the present study, we
required participants to response or inhibit their response according
to facial expression or facial gender. In the case, emotion information
was distracted stimuli. Participants pay more attention to sad emotion
processing, resulting in decreased frontal N1 amplitudes elicited by
response inhibition task. In addition, depression group showed smaller
N1 amplitudes than control group irrespective of task type and valence.
Studies reported that depression patients showed obvious attention
bias to negative expression and also exhibited difficulty in recognizing
six basic facial expressions, including neutral facial expression
(Feinberg et al., 1986; Leppanen et al., 2004). Therefore, more
attention resources were devoted to facial expression processing in
depression group, resulting in decreased frontal N1amplitudes in
response inhibition task irrespective of task and valence.
More important, as expected, larger N2 and P3 components were
elicited by no-go stimuli as compared to go stimuli. Previous electro-
physiological findings indicated that N2 and P3 represent distinct
functions in response inhibition: the former is related to conflict
monitoring, while the latter is an indicator of behavioral inhibition
(Albert et al., 2010; Bokura et al., 2001; Donkers and van Boxtel, 2004;
49
Psychiatry Research: Neuroimaging 259 (2017) 42–53
Kropotov et al., 2011). During the N2 stage, we detected a significant
interaction effect of trial type and group (Fig. 4). N2 amplitudes elicited
by no-go trials were decreased in the depression as compared to the
control group, indicating that the processing of conflict monitoring was
disrupted in depression irrespective of valence and task. These results
were consistent with previously examined non-emotional tasks such as
flank (Alderman et al., 2015) and go/no-go (Bailey et al., 2014) tasks,
which also showed decreased N2 amplitudes. It is worth noting that
there was no interaction effect of task, valence, and group within no-go-
N2. N2 sub-components are sensitive to involuntary cognitive proces-
sing but not to voluntary top-down cognitive regulation (Kahkonen
et al., 2002). In this study, we used a block design and participants
were required to respond to or inhibit their response according to the
same instructions presented at the beginning of each task. As such,
their attention was focused on facial expression or gender in order to
perform the task, and was not affected by task or valence.
It was worth noting that we observed a significant interaction effect
of group, task, valence, and trial type on the P3 component at 500–
600 ms, as demonstrated by difference waves. The subtraction of ERPs
elicited by go trials from those elicited by no-go trials in each valence
condition revealed a clear P3 difference wave during this stage. No-go-
related P3 represents behavioral inhibition to inappropriate stimuli.
We observed a significant interaction effect of valence and group in the
difference wave. The depression group showed decreased no-go-P3
F. Yu et al.
Fig. 6. Source localization results on P3 components. (A) sLORETA solutions to non-parametric randomization tests on P3 components showing voxels in which the go/no-go condition
contrast was significant (P < 0.05); (B) Histograms of the CSD of rDLPFC extracted from implicit (black bar) and explicit tasks (gray bar) under sad and neutral conditions in control (left
panel) and depression (right panel) groups. *, P < 0.05.
amplitudes relative to controls when inhibiting sad but not neutral
stimuli as well as lower discrimination accuracy during this task. The
original P3 amplitudes also showed interaction effect between valence,
trial type and group. When inhibiting sad stimuli, control group
showed increased P3 amplitudes compared to neutral stimuli, but
depression group showed no difference between the two valence stimuli
(Fig. 4). Both electrophysiological and behavior data indicated that
depression patients had impaired response inhibition to sad faces.
Behavioral studies that examined interference and inhibition of emo-
tional stimuli in depression subjects based on adapted paradigms have
shown that these subjects fail to inhibit negative information in a
behavioral context (Joormann and Gotlib, 2006, 2007, 2008; Koster
et al., 2005). Devoting more cognitive resources to the inhibition of
avoidance-based motivationally relevant stimuli is an evolutionary
adaptation in normal individuals (Cuthbert et al., 2000; Schupp
et al., 2000). Clinically depressed patients had decreased no-go P3 to
a sad expression, indicating maladapted cognition in this group
resulting in a diminished ability to inhibit response to sad faces. This
was underscored by the positive correlation between no-go P3 elicited
by sad faces and discrimination accuracy in the sad no-go/neutral go
condition. Depression patients display attention bias to mood-congru-
ent stimuli, which makes inhibition of negative stimuli difficult
(Krompinger and Simons, 2009). Negative attention bias and the
inability to inhibit negative information are closely related phenomena
that play key roles in depression pathogenesis, maintenance, and
relapse (Goeleven et al., 2006; Katz et al., 2010; Vanderhasselt et al.,
2012). The electrophysiological evidence showed that selective impair-
ment of response inhibition to sad faces in depression patients
occurred during the late inhibition no-go P3 stage.
Inconsistent with the present study, using a go/no-go task with
emotional images as stimuli, Krompinger and colleagues found that
individuals with high depression scores showed larger P3 amplitudes to
negative as compared to positive stimuli in both go and no-go trials,
50
Psychiatry Research: Neuroimaging 259 (2017) 42–53
while negative and positive stimuli elicited similar P3 in controls
(Krompinger and Simons, 2009). This may be explained by the
different P3 sub-components in the two studies. In the present study,
the P3 component was mainly distributed in the midline prefrontal
area (i.e., maximal at FCz, followed by Cz and Pz), which is typically
inhibited in association with the no-go P3 component; in the previous
work, P3 was not distributed at the midline and was actually a P3b
component related to attention/working memory processing. Another
possible reason is that research subjects in Krompinger and Simons's
work is sub-clinically depressed individuals which is different from
clinically depressed patients in current study. The different results
between the two studies demonstrate that neural mechanisms of
response inhibition differ between clinically and sub-clinically de-
pressed populations. A direct comparison between the two groups
using the same simple task paradigm can confirm or refute this
possibility.
The interaction effect between task, valence, and group was
significant during the P3 difference wave. In the explicit task, ampli-
tudes of P3 difference waves were reduced in depression patients
relative to controls in the sad condition, but in the neutral condition,
P3 difference waves were similar between the two groups. In the
implicit task, these waves decreased in patients relative to controls in
both the sad and neutral conditions. It is worth noting that depression
patients showed abnormal processing when inhibiting the response to
neutral faces in the implicit but not in the explicit task. In the former,
participants were asked to inhibit or execute behavior according to face
gender. Under this condition, neutral faces were not the focus of
attention and were therefore ambiguous stimuli. It has been reported
that depressed individuals are more inclined to interpret ambiguous
information in a negative manner (Mogg et al., 2006; Moser et al.,
2012). Indeed, in the implicit task, depression patients also showed
impaired response inhibition to neutral faces, which were recognized as
negative information. However, in the explicit task in which patients
F. Yu et al.
were asked to respond to facial expression, patients inhibited neutral
faces normally. These results suggest that interpretation bias in
depression can be modified by training individuals to interpret
ambiguous information as benign or positive (Mathews and
Mackintosh, 2000).
No-go N2 amplitudes were smaller in depression patients than in
controls irrespective of valence, while the amplitudes of no-go P3 were
decreased only in the sad condition. N2 sub-components are sensitive
to involuntary cognitive processing but not to voluntary top-down
cognitive regulation (Kahkonen et al., 2002), while components emer-
ging in late stages of the P3 family—including no-go P3—are sensitive
to motivationally relevant stimuli and processing stage during con-
sciousness (Bradley et al., 2007; Delplanque et al., 2006). In the N2
stage, depression patients showed decreased amplitudes as compared
to controls irrespective of task and valence; consequently, in the later
P3 stage, inappropriate neutral stimuli may be intentionally inhibited
by top-down cognitive control.
During the P3 time window, the rDLPFC was activated in the no-go
relative to the go condition (Fig. 4). Neuroimaging studies in humans
have reported that this brain region, which is part of the rIFC, plays a
critical role in response inhibition, task switching, updating, and other
cognitive control functions (Meyer et al., 2011; Verbruggen et al.,
2010). Our results confirm that the rDLPFC is a critical area for
response inhibition. Furthermore, the cluster analysis found that the
CSD of the rDLPFC was higher in the implicit than in the explicit task
in sad condition in the control rather than in depression group. Indeed,
in the explicit task, facial expressions were directly processed and were
within the scope of voluntary attention, while in the implicit task, they
were incidentally processed and were within the scope of involuntary
attention. Thus, attentional resources dedicated to stimulus processing
differ for the two tasks. Previous studies have reported that the DLPFC
plays a critical role in top-down modulation of selective attention,
working memory, response inhibition, and other cognitive control
functions (Harding et al., 2015; Shin and Kim, 2015). The differential
activation of the rDLPFC for the two tasks in the P3 phase that was
observed in normal subjects suggests that it is critical for top-down
cognitive control. However, there was no difference between the two
tasks in the depression group, indicating that functional abnormalities
exist in the rDLPFC of depression patients. Various reports have
highlighted the key role of the DLPFC in the pathophysiology of
depression (Koenigs and Grafman, 2009); exogenous stimulation of
this brain area has been shown to reduce depression symptoms
(Brunoni et al., 2014; Loo and Mitchell, 2005; Marangell et al.,
2007). The present results indicate that impairment in response
inhibition and deficits in the rDLPFC underlie depression.
There were several limitations to this study. Firstly, only sad
expressions were used as experimental stimuli; further study is
required to determine how other basic emotions are related to response
inhibition in female depression patients across implicit and explicit
tasks. Secondly, depression is often accompanied by anxiety. Anxiety
scores in our depression group were high, suggesting that differences
between depression patients and controls may be influenced by
anxiety. Future studies should distinguish between depression patients
with and without anxiety. Thirdly, the spatial resolution of the
electrophysiological method used in this study was low, and source
localization analysis was carried out according to a mathematical and
not a physiological approach. In addition, many sub-cortical areas are
involved in response inhibition to facial stimuli such as the amygdala,
ventral striatum, and fusiform gyrus (Kret et al., 2011; Shafritz et al.,
2006). Due to the limitations of electrophysiological source analysis, we
were unable to record from sub-cortical areas. Additional studies using
other brain imaging methodologies with high spatial resolution, such as
functional magnetic resonance imaging, are needed to substantiate
these findings. Last but not least, the participants recruited in the
present study were just depression and control women. Given the
evidence that sex hormones, such as testosterone and ovarian show
51
Psychiatry Research: Neuroimaging 259 (2017) 42–53
imbalance relationship with depression symptoms (Colangelo et al.,
2012), it remains unclear whether the results are generalizable to
depressed men. Moreover, Yuan and colleagues have reported that
women and men had different behavioral inhibition control ability in
human adults (Yuan et al., 2008). Future studies focus on response
inhibition ability of depression men to negative information are
required to clarify this question.
Funding
This research was supported by the National Natural Science
Foundation of China (31000503, 91232717, 81301176, 81300944,
and 81300944) and the Natural Science Foundation of the Colleges and
Universities in Anhui Province (KJ2016A355). The funders had no role
in study design, data collection and analysis, decision to publish, or
preparation of the manuscript.
References
Aarts, K., Pourtois, G., 2010. Anxiety not only increases, but also alters early error-
monitoring functions. Cogn. Affect. Behav. Neurosci. 10, 479–492.
Albert, J., Lopez-Martin, S., Carretie, L., 2010. Emotional context modulates response
inhibition: neural and behavioral data. NeuroImage 49, 914–921.
Alderman, B.L., Olson, R.L., Bates, M.E., Selby, E.A., Buckman, J.F., Brush, C.J., Panza,
E.A., Kranzler, A., Eddie, D., Shors, T.J., 2015. Rumination in major depressive
disorder is associated with impaired neural activation during conflict monitoring.
Front. Hum. Neurosci. 9, 269.
Association, A.P., 1994. Diagnostic and Statistical Manual of Mental Disorders 4 ed..
American Psychiatric Press, Washington, DC.
Bailey, N.W., Hoy, K.E., Maller, J.J., Segrave, R.A., Thomson, R., Williams, N.,
Daskalakis, Z.J., Fitzgerald, P.B., 2014. An exploratory analysis of Go/Nogo event-
related potentials in major depression and depression following traumatic brain
injury. Psychiatry Res. 224, 324–334.
Berggren, N., Derakshan, N., 2013. Attentional control deficits in trait anxiety: why you
see them and why you don't. Biol. Psychol. 92, 440–446.
Berkman, E.T., Burklund, L., Lieberman, M.D., 2009. Inhibitory spillover: intentional
motor inhibition produces incidental limbic inhibition via right inferior frontal
cortex. NeuroImage 47, 705–712.
Bokura, H., Yamaguchi, S., Kobayashi, S., 2001. Electrophysiological correlates for
response inhibition in a Go/NoGo task. Clin. Neurophysiol.: Off. J. Int. Fed. Clin.
Neurophysiol. 112, 2224–2232.
Botvinick, M.M., Braver, T.S., Barch, D.M., Carter, C.S., Cohen, J.D., 2001. Conflict
monitoring and cognitive control. Psychol. Rev. 108, 624–652.
Bradley, M.M., Hamby, S., Low, A., Lang, P.J., 2007. Brain potentials in perception:
picture complexity and emotional arousal. Psychophysiology 44, 364–373.
Brent, D.A., Birmaher, B., 2006. Treatment-resistant depression in adolescents:
recognition and management. Child Adolesc. Psychiatr. Clin. N. Am. 15, 1015–1034.
Brunoni, A.R., Boggio, P.S., De Raedt, R., Bensenor, I.M., Lotufo, P.A., Namur, V.,
Valiengo, L.C., Vanderhasselt, M.A., 2014. Cognitive control therapy and
transcranial direct current stimulation for depression: a randomized, double-
blinded, controlled trial. J. Affect. Disord. 162, 43–49.
Burkhouse, K.L., Siegle, G.J., Woody, M.L., Kudinova, A.Y., Gibb, B.E., 2015. Pupillary
reactivity to sad stimuli as a biomarker of depression risk: evidence from a
prospective study of children. J. Abnorm. Psychol. 124, 498–506.
Colangelo, L.A., Craft, L.L., Ouyang, P., Liu, K., Schreiner, P.J., Michos, E.D., Gapstur,
S.M., 2012. Association of sex hormones and sex hormone-binding globulin with
depressive symptoms in postmenopausal women: the Multiethnic Study of
Atherosclerosis. Menopause 19, 877–885.
Cuthbert, B.N., Schupp, H.T., Bradley, M.M., Birbaumer, N., Lang, P.J., 2000. Brain
potentials in affective picture processing: covariation with autonomic arousal and
affective report. Biol. Psychol. 52, 95–111.
De Raedt, R., Koster, E.H., 2010. Understanding vulnerability for depression from a
cognitive neuroscience perspective: a reappraisal of attentional factors and a new
conceptual framework. Cogn. Affect. Behav. Neurosci. 10, 50–70.
Delplanque, S., Silvert, L., Hot, P., Rigoulot, S., Sequeira, H., 2006. Arousal and valence
effects on event-related P3a and P3b during emotional categorization. Int. J.
Psychophysiol.: Off. J. Int. Organ. Psychophysiol. 60, 315–322.
Donkers, F.C., van Boxtel, G.J., 2004. The N2 in go/no-go tasks reflects conflict
monitoring not response inhibition. Brain Cognit. 56, 165–176.
Eimer, M., Holmes, A., 2002. An ERP study on the time course of emotional face
processing. Neuroreport 13, 427–431.
Eimer, M., Kiss, M., Holmes, A., 2008. Links between rapid ERP responses to fearful
faces and conscious awareness. J. Neuropsychol. 2, 165–181.
Fabiani, M., Gratton, G., Goles, M.G., 2000. Hand Book of Psychophysiology. Cambridge
University Press, London.
Fan, C., Chen, S., Zhang, L., Qi, Z., Jin, Y., Wang, Q., Luo, Y., Li, H., Luo, W., 2015. N170
changes reflect competition between faces and identifiable characters during early
visual processing. NeuroImage 110, 32–38.
Feinberg, T.E., Rifkin, A., Schaffer, C., Walker, E., 1986. Facial discrimination and
emotional recognition in schizophrenia and affective disorders. Arch. Gen.
F. Yu et al.
Psychiatry 43, 276–279.
Goeleven, E., De Raedt, R., Baert, S., Koster, E.H., 2006. Deficient inhibition of emotional
information in depression. J. Affect. Disord. 93, 149–157.
Goldstein, M., Brendel, G., Tuescher, O., Pan, H., Epstein, J., Beutel, M., Yang, Y.,
Thomas, K., Levy, K., Silverman, M., Clarkin, J., Posner, M., Kernberg, O., Stern, E.,
Silbersweig, D., 2007. Neural substrates of the interaction of emotional stimulus
processing and motor inhibitory control: an emotional linguistic go/no-go fMRI
study. NeuroImage 36, 1026–1040.
Gollan, J.K., Pane, H.T., McCloskey, M.S., Coccaro, E.F., 2008. Identifying differences in
biased affective information processing in major depression. Psychiatry Res. 159,
18–24.
Goodwin, R.D., Jacobi, F., Bittner, A., 2006. Text Book of Mood Disorders 2006.
American Psychiatric Publishing Washington, Washington, 33–54.
Gotlib, I.H., Joormann, J., 2010. Cognition and depression: current status and future
directions. Annu. Rev. Clin. Psychol. 6, 285–312.
Gur, R.C., Erwin, R.J., Gur, R.E., Zwil, A.S., Heimberg, C., Kraemer, H.C., 1992. Facial
emotion discrimination: II. Behavioral findings in depression. Psychiatry Res. 42,
241–251.
Hamilton, M., 1960. A rating scale for depression. J. Neurol. Neurosurg. Psychiatry 23,
56–62.
Harding, I.H., Yucel, M., Harrison, B.J., Pantelis, C., Breakspear, M., 2015. Effective
connectivity within the frontoparietal control network differentiates cognitive control
and working memory. NeuroImage 106, 144–153.
Henje Blom, E., Connolly, C.G., Ho, T.C., LeWinn, K.Z., Mobayed, N., Han, L., Paulus,
M.P., Wu, J., Simmons, A.N., Yang, T.T., 2015. Altered insular activation and
increased insular functional connectivity during sad and happy face processing in
adolescent major depressive disorder. J. Affect. Disord. 178, 215–223.
Holmes, A., Vuilleumier, P., Eimer, M., 2003. The processing of emotional facial
expression is gated by spatial attention: evidence from event-related brain potentials.
brain Res. Cogn. Brain Res. 16, 174–184.
Hum, K.M., Manassis, K., Lewis, M.D., 2013. Neural mechanisms of emotion regulation
in childhood anxiety. J. Child Psychol. Psychiatry Allied Discip. 54, 552–564.
Joormann, J., Gotlib, I.H., 2006. Is this happiness I see? Biases in the identification of
emotional facial expressions in depression and social phobia. J. Abnorm. Psychol.
115, 705–714.
Joormann, J., Gotlib, I.H., 2007. Selective attention to emotional faces following recovery
from depression. J. Abnorm. Psychol. 116, 80–85.
Joormann, J., Gotlib, I.H., 2008. Updating the contents of working memory in
depression: interference from irrelevant negative material. J. Abnorm. Psychol. 117,
182–192.
Joormann, J., Yoon, K.L., Zetsche, U., 2007. Cognitive inhibition in depression. Appl.
Prev. Psychol. 12, 128–139.
Kahkonen, S., Ahveninen, J., Pekkonen, E., Kaakkola, S., Huttunen, J., Ilmoniemi, R.J.,
Jaaskelainen, I.P., 2002. Dopamine modulates involuntary attention shifting and
reorienting: an electromagnetic study. Clin. Neurophysiol.: Off. J. Int. Fed. Clin.
Neurophysiol. 113, 1894–1902.
Kaiser, S., Unger, J., Kiefer, M., Markela, J., Mundt, C., Weisbrod, M., 2003. Executive
control deficit in depression: event-related potentials in a Go/Nogo task. Psychiatry
Res. 122, 169–184.
Katz, R., De Sanctis, P., Mahoney, J.R., Sehatpour, P., Murphy, C.F., Gomez-Ramirez, M.,
Alexopoulos, G.S., Foxe, J.J., 2010. Cognitive control in late-life depression:
response inhibition deficits and dysfunction of the anterior cingulate cortex. Am. J.
Geriatr. Psychiatry: Off. J. Am. Assoc. Geriatr. Psychiatry 18, 1017–1025.
Kessler, R.C., Chiu, W.T., Demler, O., Merikangas, K.R., Walters, E.E., 2005. Prevalence,
severity, and comorbidity of 12-month DSM-IV disorders in the National
Comorbidity Survey Replication. Arch. Gen. Psychiatry 62, 617–627.
Kilford, E.J., Foulkes, L., Potter, R., Collishaw, S., Thapar, A., Rice, F., 2015. Affective
bias and current, past and future adolescent depression: a familial high risk study. J.
Affect. Disord. 174, 265–271.
Koenigs, M., Grafman, J., 2009. The functional neuroanatomy of depression: distinct
roles for ventromedial and dorsolateral prefrontal cortex. Behav. Brain Res. 201,
239–243.
Koster, E.H., De Lissnyder, E., Derakshan, N., De Raedt, R., 2011. Understanding
depressive rumination from a cognitive science perspective: the impaired
disengagement hypothesis. Clin. Psychol. Rev. 31, 138–145.
Koster, E.H., De Raedt, R., Goeleven, E., Franck, E., Crombez, G., 2005. Mood-congruent
attentional bias in dysphoria: maintained attention to and impaired disengagement
from negative information. Emotion 5, 446–455.
Kret, M.E., Denollet, J., Grezes, J., de Gelder, B., 2011. The role of negative affectivity
and social inhibition in perceiving social threat: an fMRI study. Neuropsychologia
49, 1187–1193.
Krompinger, J.W., Simons, R.F., 2009. Electrophysiological indicators of emotion
processing biases in depressed undergraduates. Biol. Psychol. 81, 153–163.
Kropotov, J.D., Ponomarev, V.A., Hollup, S., Mueller, A., 2011. Dissociating action
inhibition, conflict monitoring and sensory mismatch into independent components
of event related potentials in GO/NOGO task. NeuroImage 57, 565–575.
Leppanen, J.M., Milders, M., Bell, J.S., Terriere, E., Hietanen, J.K., 2004. Depression
biases the recognition of emotionally neutral faces. Psychiatry Res. 128, 123–133.
Linden, S.C., Jackson, M.C., Subramanian, L., Wolf, C., Green, P., Healy, D., Linden,
D.E., 2010. Emotion-cognition interactions in schizophrenia: implicit and explicit
effects of facial expression. Neuropsychologia 48, 997–1002.
Loo, C.K., Mitchell, P.B., 2005. A review of the efficacy of transcranial magnetic
stimulation (TMS) treatment for depression, and current and future strategies to
optimize efficacy. J. Affect. Disord. 88, 255–267.
Luo, W., Feng, W., He, W., Wang, N.Y., Luo, Y.J., 2010. Three stages of facial expression
processing: ERP study with rapid serial visual presentation. NeuroImage 49,
52
Psychiatry Research: Neuroimaging 259 (2017) 42–53
1857–1867.
Marangell, L.B., Martinez, M., Jurdi, R.A., Zboyan, H., 2007. Neurostimulation therapies
in depression: a review of new modalities. Acta Psychiatr. Scand. 116, 174–181.
Mathews, A., Mackintosh, B., 2000. Induced emotional interpretation bias and anxiety. J.
Abnorm. Psychol. 109, 602–615.
Meyer, T., Qi, X.L., Stanford, T.R., Constantinidis, C., 2011. Stimulus selectivity in dorsal
and ventral prefrontal cortex after training in working memory tasks. J. Neurosci.:
Off. J. Soc. Neurosci. 31, 6266–6276.
Mogg, K., Bradbury, K.E., Bradley, B.P., 2006. Interpretation of ambiguous information
in clinical depression. Behav. Res. Ther. 44, 1411–1419.
Moser, J.S., Huppert, J.D., Foa, E.B., Simons, R.F., 2012. Interpretation of ambiguous
social scenarios in social phobia and depression: evidence from event-related brain
potentials. Biol. Psychol. 89, 387–397.
Mueller, E.M., Pizzagalli, D.A., 2016. One-year-old fear memories rapidly activate human
fusiform gyrus. Soc. Cogn. Affect. Neurosci. 11, 308–316.
Nichols, T.E., Holmes, A.P., 2002. Nonparametric permutation tests for functional
neuroimaging: a primer with examples. Hum. Brain Mapp. 15, 1–25.
Ochsner, K.N., Gross, J.J., 2008. cognitive emotion regulation: insights from social
cognitive and affective neuroscience. Curr. Dir. Psychol. Sci. 17, 153–158.
Ochsner, K.N., Ray, R.D., Cooper, J.C., Robertson, E.R., Chopra, S., Gabrieli, J.D., Gross,
J.J., 2004. For better or for worse: neural systems supporting the cognitive down-
and up-regulation of negative emotion. NeuroImage 23, 483–499.
Otten, L.J., Rugg, M.D., 2004. Event-Related Potentials: A Methods Handbook. The MIT
Press, United States of America.
Padmala, S., Pessoa, L., 2010. Interactions between cognition and motivation during
response inhibition. Neuropsychologia 48, 558–565.
Pascual-Marqui, R.D., 2002. Standardized low-resolution brain electromagnetic
tomography (sLORETA): technical details. Methods Find. Exp. Clin. Pharmacol. 24
(Suppl. D), S5–S12.
Pizzagalli, D.A., Nitschke, J.B., Oakes, T.R., Hendrick, A.M., Horras, K.A., Larson, C.L.,
Abercrombie, H.C., Schaefer, S.M., Koger, J.V., Benca, R.M., Pascual-Marqui, R.D.,
Davidson, R.J., 2002. Brain electrical tomography in depression: the importance of
symptom severity, anxiety, and melancholic features. Biol. Psychiatry 52, 73–85.
Rauch, A.V., Reker, M., Ohrmann, P., Pedersen, A., Bauer, J., Dannlowski, U., Harding,
L., Koelkebeck, K., Konrad, C., Kugel, H., Arolt, V., Heindel, W., Suslow, T., 2010.
Increased amygdala activation during automatic processing of facial emotion in
schizophrenia. Psychiatry Res. 182, 200–206.
Righart, R., de Gelder, B., 2007. Impaired face and body perception in developmental
prosopagnosia. Proc. Natl. Acad. Sci. USA 104, 17234–17238.
Schuch, J.J., Roest, A.M., Nolen, W.A., Penninx, B.W., de Jonge, P., 2014. Gender
differences in major depressive disorder: results from the Netherlands study of
depression and anxiety. J. Affect. Disord. 156, 156–163.
Schupp, H.T., Cuthbert, B.N., Bradley, M.M., Cacioppo, J.T., Ito, T., Lang, P.J., 2000.
Affective picture processing: the late positive potential is modulated by motivational
relevance. Psychophysiology 37, 257–261.
Seeck, M., Lazeyras, F., Michel, C.M., Blanke, O., Gericke, C.A., Ives, J., 1998. Non-
invasive epileptic focus localization using EEG-triggered functional MRI and
electromagnetic tomography. Clin. Phys. Physiol. Meas.: Off. J. Hosp. Phys. Assoc.
Dtsch. Ges. Med. Phys. Eur. Fed. Organ. Med. Phys. 106, 508–512.
Sehlmeyer, C., Konrad, C., Zwitserlood, P., Arolt, V., Falkenstein, M., Beste, C., 2010.
ERP indices for response inhibition are related to anxiety-related personality traits.
Neuropsychologia 48, 2488–2495.
Shafritz, K.M., Collins, S.H., Blumberg, H.P., 2006. The interaction of emotional and
cognitive neural systems in emotionally guided response inhibition. NeuroImage 31,
468–475.
Shin, G., Kim, C., 2015. Neural correlates of cognitive style and flexible cognitive control.
NeuroImage 113, 78–85.
Snodgrass, J.G., Corwin, J., 1988. Perceptual identification thresholds for 150
fragmented pictures from the Snodgrass and Vanderwart picture set. Percept. Mot.
Skills 67, 3–36.
Suslow, T., Konrad, C., Kugel, H., Rumstadt, D., Zwitserlood, P., Schoning, S., Ohrmann,
P., Bauer, J., Pyka, M., Kersting, A., Arolt, V., Heindel, W., Dannlowski, U., 2010.
Automatic mood-congruent amygdala responses to masked facial expressions in
major depression. Biol. Psychiatry 67, 155–160.
Sylvester, C.M., Hudziak, J.J., Gaffrey, M.S., Barch, D.M., Luby, J.L., 2015. Stimulus-
driven attention, threat bias, and sad bias in youth with a history of an anxiety
disorder or depression. J. Abnorm. Child Psychol.
Taylor, S.F., Phan, K.L., Decker, L.R., Liberzon, I., 2003. Subjective rating of emotionally
salient stimuli modulates neural activity. NeuroImage 18, 650–659.
Valdes-Conroy, B., Aguado, L., Fernandez-Cahill, M., Romero-Ferreiro, V., Dieguez-
Risco, T., 2014. Following the time course of face gender and expression processing:
a task-dependent ERP study. Int. J. Psychophysiol.: Off. J. Int. Organ. Psychophysiol.
92, 59–66.
Vanderhasselt, M.A., De Raedt, R., Dillon, D.G., Dutra, S.J., Brooks, N., Pizzagalli, D.A.,
2012. Decreased cognitive control in response to negative information in patients
with remitted depression: an event-related potential study. J. Psychiatry Neurosci.:
JPN 37, 250–258.
Verbruggen, F., Aron, A.R., Stevens, M.A., Chambers, C.D., 2010. Theta burst stimulation
dissociates attention and action updating in human inferior frontal cortex. Proc.
Natl. Acad. Sci. USA 107, 13966–13971.
Wager, T.D., Davidson, M.L., Hughes, B.L., Lindquist, M.A., Ochsner, K.N., 2008.
Prefrontal-subcortical pathways mediating successful emotion regulation. Neuron
59, 1037–1050.
Winston, J.S., O'Doherty, J., Dolan, R.J., 2003. Common and distinct neural responses
during direct and incidental processing of multiple facial emotions. NeuroImage 20,
84–97.
F. Yu et al.
Worrell, G.A., Lagerlund, T.D., Sharbrough, F.W., Brinkmann, B.H., Busacker, N.E.,
Cicora, K.M., O'Brien, T.J., 2000. Localization of the epileptic focus by low-resolution
electromagnetic tomography in patients with a lesion demonstrated by MRI. Brain
Topogr. 12, 273–282.
Yang, P., Wang, M., Jin, Z., Li, L., 2015. Visual short-term memory load modulates the
early attention and perception of task-irrelevant emotional faces. Front. Hum.
Neurosci. 9, 490.
Yu, F., Ye, R., Sun, S., Carretie, L., Zhang, L., Dong, Y., Zhu, C., Luo, Y., Wang, K., 2014.
Dissociation of neural substrates of response inhibition to negative information
between implicit and explicit facial Go/Nogo tasks: evidence from an
electrophysiological study. PloS One 9, e109839.
Yu, F., Yuan, J., Luo, Y.J., 2009. Auditory-induced emotion modulates processes of
53
Psychiatry Research: Neuroimaging 259 (2017) 42–53
response inhibition: an event-related potential study. Neuroreport 20, 25–30.
Yu, F., Zhu, C., Zhang, L., Chen, X., Li, D., Zhang, L., Ye, R., Dong, Y., Luo, Y., Hu, X.,
Wang, K., 2015. The neural substrates of response inhibition to negative information
across explicit and implicit tasks in GAD patients: electrophysiological evidence from
an ERP study. Front. Psychol. 6, 275.
Yuan, J., He, Y., Qinglin, Z., Chen, A., Li, H., 2008. Gender differences in behavioral
inhibitory control: ERP evidence from a two-choice oddball task. Psychophysiology
45, 986–993.
Yuan, J., Luo, Y., Yan, J.H., Meng, X., Yu, F., Li, H., 2009. Neural correlates of the
females' susceptibility to negative emotions: an insight into gender-related
prevalence of affective disturbances. Hum. Brain Mapp. 30, 3676–3686.