3V\FKRORJLFD Coussement, C., et al. (2019). Does Change in Attention Control Mediate the Impact of tDCS on Attentional Bias for Threat? Limited Evidence ψ %HOJLFD from a Double-blind Sham-controlled Experiment in an Unselected Sample. Psychologica Belgica, 59(1), pp. 16–32. DOI: https://doi.org/10.5334/pb.449

RESEARCH ARTICLE Does Change in Attention Control Mediate the Impact of tDCS on Attentional Bias for Threat? Limited Evidence from a Double-blind Sham-controlled Experiment in an Unselected Sample Charlotte Coussement*,†, Pierre Maurage†,‡, Joël Billieux§ and Alexandre Heeren†,‡

Neurocognitive models of attentional bias for threat posit that attentional bias may result from a decreased activation of the left prefrontal cortex, and espe- cially of its dorsolateral part (dlPFC), resulting in an impaired attention control. Consequently, a transient increase of neural activity within the left dlPFC via non-invasive brain stimulation reduces attentional bias among both anxious and nonanxious participants. Yet, it is still unclear whether the impact of dlPFC acti- vation on attentional bias is mediated by improvement in attention control. In this experiment, we sought to test this hypothesis in an unselected sample (n = 20). Accordingly, we adopted a double-blind within-subject protocol in which we deliv- ered a single-session of anodal versus sham transcranial Direct Current Stimulation (tDCS) over the left dlPFC during the completion of a task assessing attention control. We also assessed its subsequent impact on attentional bias. Neither atten- tion control nor attentional bias did significantly improve following anodal tDCS. Although our results do not support our main hypothesis, we believe the present null results to be particularly useful for future meta-research in the field. We also formulated a series of methodological recommendations for future research aiming at testing the tDCS-induced modification of attentional bias.

Keywords: neuromodulation; transcranial direct current stimulation; attentional bias for threat; attention control; prefrontal cortex

* Cellule de Recherche et Publications Neuroscience, Université catholique de Louvain, ­Scientifiques (CRPS), Hôpital Psychiatrique du Brussels, BE Beau Vallon, Namur, BE § Integrative Research Unit on Social and † Psychological Science Research Institute, ­Individual Development (INSIDE), University of ­Université catholique de Louvain, Louvain-la- Luxembourg, Esch-sur-Alzette, LU Neuve, BE Corresponding author: Alexandre Heeren, Ph.D., ‡ Clinical Neuroscience Division, Institute of ([email protected]) Coussement et al: Does Change in Attention Control Mediate the Impact of 17 tDCS on Attentional Bias for Threat? Limited Evidence from a Double-blind Sham-controlled Experiment in an Unselected Sample Prominent cognitive theorists of anxiety Eysenck & Derakshan, 2011; Heeren, De disorders have argued that attentional bias Raedt, Koster, & Philippot, 2013; Peers et al., for threat—that is, a differential attentional 2013). This hypothesis relies on the seminal allocation for threat-related stimuli rela- study of Derryberry and Reed (2002) dem- tive to neutral ones—figures prominently onstrating that attention control (assessed in the maintenance, and perhaps the etiol- using a self-report measure) moderated the ogy, of anxiety disorders (Beck & Clark, 1997; magnitude of attentional bias among indi- Heimberg, Brozovich, & Rapee, 2010; Mogg viduals with elevated trait anxiety—that & Bradley, 2002; Van Bockstaele et al., 2014; is, individuals with lower attention con- Wong & Rapee, 2016). Accordingly, reducing trol exhibited stronger attentional bias for attentional bias via attention bias modifi- threat whereas those with higher attention cation (ABM) procedures— a computerized control showed reduced attentional bias. training procedure targeting attentional Several replications of this initial study have bias—may have clinical benefits (for meta- been reported across distinct paradigms analyses, see Heeren, Mogoaşe, Philippot, and clinical and nonclinical samples (e.g., & McNally 2015; Linetzky, Pergamin-Hight, Bardeen & Orcutt, 2011; Reinholdt-Dunne, Pine, & Bar-Haim, 2015; Mogoaşe, David, & Mogg, & Bradley, 2009; Taylor, Cross, & Amir, Koster, 2014). Likewise, transiently increas- 2016). Moreover, changes in attentional bias ing attentional bias promotes anxiety via ABM procedures depended on the initial proneness among nonanxious individuals level of attention control (Paulewicz, Blaut, & (e.g., Heeren, Peschard, & Philippot, 2012; Kłosowska, 2012). MacLeod, Rutherford, Campbell, Ebsworthy, Altogether, these findings dovetail with & Holker, 2002). longstanding neurocognitive models of The aforementioned studies suggested attentional bias (e.g., Bishop, 2008, 2009; that attentional bias for threat can be modi- Vuilleumier, 2005) suggesting that the fied and, in turn, reduce anxiety symptoms. deployment of attention vis-à-vis threaten- Yet, despite these promising initial results, ing material is regulated by two primary neu- recent meta-analyses indicated that modify- ral systems: (1) a bottom–up amygdala-based ing attentional bias via ABM yielded only a system that produces a signal reflecting the very limited impact—albeit significant—on perceived salience of stimuli and directs reducing attentional bias and anxiety symp- attention toward salient stimuli (Adolphs, toms (e.g., Cristea, Kok, & Cuijpers, 2015; Tranel, Damasio, & Damasio, 1995; Davis Heeren, Mogoaşe, et al., 2015; Mogoaşe & Whalen, 2001), and (2) a top–down sys- et al., 2014). However, as pointed out by tem mainly relying on the prefrontal cortex Grafton & MacLeod (2016), most ABM stud- (PFC) that produce a signal when conflicting ies that failed to reduce anxiety also failed demands are made on attention and down- to modify attentional bias for threat as regulate amygdala activation in the presence intend. Moreover, the mechanisms respon- of threat (Bishop, 2004). From this perspec- sible for ABM remains poorly understood tive, neuroscientists have hypothesized that (e.g., Heeren, Coussement, & McNally, 2016; attentional bias may result from a failure to Heeren, De Raedt, Koster, & Philippot, 2013; recruit regulatory PFC regions that are man- Mogg & Bradley, 2016). Therefore, the critical datory to down-regulate amygdala activa- next step is thus to improve our understand- tion in the presence of threat (Bishop, 2009; ing of the mechanisms involved in the modi- Britton, Lissek, Grillon, Norcross, & Pine, fication of attentional bias for threat. 2011). Accordingly, reduced left PFC activa- One common theoretical explanation for tions, especially of its dorsolateral (dlPFC; the maintenance of attentional bias focuses Bishop, 2009) and ventrolateral (vlPFC; Monk on general attention control — i.e. the abil- et al., 2006, 2008) sections, have been found ity to voluntarily regulate the allocation of among anxious individuals when complet- attentional resources (Cisler & Koster, 2010; ing tasks requiring such a top-down control 18 Coussement et al: Does Change in Attention Control Mediate the Impact of tDCS on Attentional Bias for Threat? Limited Evidence from a Double-blind Sham-controlled Experiment in an Unselected Sample in the presence of threat. Likewise, PFC areas orienting (i.e., selective engagement and dis- directly regulate amygdala activity during engagement with certain stimuli rather than threat processing (e.g., Monk et al., 2008). others), and an executive component (i.e., Moreover, the reduction of attentional bias top-down control of attention exemplified by via ABM is associated with increased activa- maintenance of attention on certain stimuli tion of the left dlPFC among healthy volun- and resisting distraction by other stimuli). teers (Browning, Holmes, Murphy, Goodwin, However, although some research have sug- & Harmer, 2010). Likewise, ABM increased gested that the three attentional networks vlPFC (Taylor et al., 2014) and attenuated might be distinctively associated with pro- bilateral amygdala activations (Britton et cesses assumedly involved in the maintenance al., 2013; Månsson et al., 2013; Taylor et al., of anxiety and related psychopathology (e.g., 2014) in patients with anxiety disorders. Heeren, Maurage, & Philippot, 2015; Heeren Convincing evidence regarding the causal & McNally, 2016), prior research in the field influence of PFC-areas in the maintenance of attentional bias for threat have almost of attentional bias also arise from experi- exclusively treated attention control as a uni- mental studies whose manipulation directly tary construct and did not differentiate the targeted the activity of these brain regions attentional networks (for a discussion, see during threat processing. Accordingly, a Heeren, Billieux, Philippot, & Maurage, 2015). single session of high-frequency repetitive Second, although previous research has sug- Transcranial Magnetic Stimulation (HF-rTMS) gested that PFC-related areas might be con- applied over the left dlPFC subsequently sidered as proxy of attention control, there is decreased attentional bias in a sample of no study directly testing whether attention nonanxious healthy participants (De Raedt control improvement does indeed mediate et al., 2010). Likewise, boosting the activ- the impact of PFC-modulation on attentional ity of the left dlPFC via transcranial Direct bias for threat. This is especially unfortunate Current Stimulation (tDCS)—another nonin- given the current trends in the development vasive method of brain stimulation allowing of neurocognitive therapeutic procedures the modulation of the cortical activities dur- aiming at directly targeting attentional bias ing the completion of a task—did mitigate to mitigate anxiety (e.g., Clarke et al., 2014; attentional bias for threat among patients Heeren et al., 2017). with a DSM-5 diagnosis of social anxiety dis- Accordingly, in the present study, we sought order (Heeren et al., 2017). Moreover, results to clarify the exact attention control mecha- indicated that, among healthy undergradu- nisms whereby the modulation of PFC-areas ate volunteers, combining tDCS over the via tDCS mitigates attentional for threat. left dlPFC with ABM yielded larger reduc- This study therefore represents a critical step tion in attentional bias than a sham stimu- towards elucidating the mechanism whereby lation combined with ABM (Clarke, Browing, left dlPFC activation via anodal tDCS may mit- Hammond, Notebaert, & MacLeod, 2014; igate attentional bias. To do so, we adopted Heeren, Baeken, Vanderhasselt, Philippot, & a double-blind within-subject protocol in De Raedt, 2015). which we delivered single-session of anodal However, despite increasing research link- versus sham tDCS over the left dlPFC during ing attentional bias for threat, attentional the completion of a task assessing attention control, and PFC activations, there are sev- control and, subsequently to the stimula- eral limitations to the previous studies. tion, on a second task indexing attentional First, prominent models of attentional sys- bias for threat. Following the aforemen- tems postulate that attention control is a mul- tioned HF-rTMS (e.g., De Raedt et al., 2010) tifaceted construct (e.g., Petersen & Posner, and tDCS studies (Heeren et al., 2017), we 2012; Posner & Rothbart, 2007), including decided to stimulate the left dlPFC, and not at least three distinct attentional networks: the left vlPFC. Of primary interest was to test alerting (i.e., maintenance of alertness), whether the reduction of attentional bias for Coussement et al: Does Change in Attention Control Mediate the Impact of 19 tDCS on Attentional Bias for Threat? Limited Evidence from a Double-blind Sham-controlled Experiment in an Unselected Sample threat following anodal tDCS is mediated by of French language. These criteria were ver- improvement in attention control, as com- bally assessed through a medical ­interview. pared to sham tDCS. Moreover, given extant Participants’ characteristics appear in publications of cognitive models bridging Table 1. executive control to attentional bias for threat The research was approved by the (for a review, see Heeren et al., 2013) and the Ethical Committee of the Medical School previous observation that the neuromodula- of the Université catholique de Louvain tion of dlPFC can foster the executive com- (Belgium) and carried out according to the ponent of attention control (Miler, Meron, Declaration of Helsinki. Participants pro- Baldwin, & Garner, 2017), we reasoned that vided informed consent, were debriefed if anodal tDCS mitigates attentional bias for upon completion,­ and were compensated threat via improvement of attention control, 30€ for their participation. then this mediational effect should be par- ticularly substantiated for the executive com- Measures ponent of attention. Descriptive measures. To best character- ize our sample, participants completed the Method Beck Depression Inventory (BDI-II; Beck, Participants Steer, & Brown, 1998), the self-report ver- The sample consisted of 20 Caucasian sion of the Liebowitz Social Anxiety Scale ­participants (65% females). All participants (LSAS; Liebowitz, 1987), and the Trait were right handed. They were recruited via Anxiety Inventory (STAIT-T; Spielberger, flyers posted in the Université catholique Gorsuch, Lushene, Vagg, & Jacobs, 1983). de Louvain community. Exclusion crite- Those scales were administered prior to ria included metal or electronic implants, start the experiment. The BDI-II is a 21-item ­epilepsy, pregnancy, cardiovascular disease, self-reported questionnaire measure of lifetime history of psychiatric/alcohol/drug symptoms of depression. The LSAS is a dependence, current pharmacological or psy- 24-item scale that measures fear and avoid- chological treatments, corrective eyewear for ance experienced in a range of social and altered vision, and insufficient knowledge performance situations over the last two

Table 1: Demographic and clinical characteristics of the participants.

Mean (SD) Cronbach’s alpha Demographic measures Age 24.45 (2.70) Educational level (in years) 16.95 (2.37) Clinical measures BDI-II 6.20 (7.24) .92 STAI-T 40.1 (7.32) .82 LSAS 42.30 (17.20) .91 Note: Education level was assessed according to the numbers of years of education completed after starting primary school. Ages ranged from 19 to 29 years old. Cronbach’s alphas were computed over the data of the current sample. BDI-II, Beck Depression Inventory; STAI-T, Spielberger State-Trait Anxiety Inventory-Trait version; LSAS, Liebowitz Social Anxiety Scale. 20 Coussement et al: Does Change in Attention Control Mediate the Impact of tDCS on Attentional Bias for Threat? Limited Evidence from a Double-blind Sham-controlled Experiment in an Unselected Sample weeks prior to completion. The STAI-T is a trials, based on the combination of four cues 20-item self-report questionnaire assessing (no cue, center cue, double cue, spatial cue), anxiety proneness. We used the validated three flankers (congruent, incongruent, neu- French version of these scales (BDI, Beck tral), two directions of the target arrow (left, et al., 1998; STAI-T, Bruchon-Schweitzer & right) and two localizations (upper or lower Paulhan, 1993; LSAS, Heeren, Maurage, et part of the screen). Trials were presented in al., 2012). Cronbach’s alphas among the cur- a random order and each possible trial was rent sample are presented in Table 1. presented twice within a block. The task was Attention networks task (ANT). The effi- programmed and presented via E-Prime 2.0 ciency of three independent attentional net- Professional® (Psychology Software Tools, works (i.e. alerting, orienting, and executive Pittsburgh, PA, USA). control) was assessed using the ANT (Fan, Probe discrimination task. To assess McCandliss, Sommer, Raz, & Posner, 2002). attentional bias, we used a probe discrimina- Participants had to determine as quickly and tion task modeled on the dot-probe detec- accurately as possible the direction of a cen- tion task (MacLeod, Mathews, & Tata, 1986). tral arrow (the target) located in the middle of The task consisted of 320 trials delivered in a horizontal line projected either at the top or one block. Each trial began with a central at the bottom of the screen. They responded fixation cross which appeared on the screen by pressing the corresponding button (left or for 500 ms. Immediately following the disap- right) on the keyboard. Each target was pre- pearance of the cross, a pair of faces appeared ceded by either no cue, a center cue (an aster- on the screen for 500 ms. One face appeared isk replacing the fixation cross), a double cue on the top of center screen, whereas the (two asterisks, one appearing above and one other face appeared on the bottom of center below the fixation cross), or a spatial cue (an screen. Each pair of faces displayed neutral- asterisk appearing above or below the fixation disgust facial expressions. cross and indicating the location of the upcom- Immediately after their disappearance, a ing target). Moreover, flankers appeared hori- probe appeared in the location previously zontally on each side of the target. There were occupied by one of the two faces. Participants three possible flanker types: either two arrows were asked to indicate whether the probe pointing in the same direction as the target was a dot (i.e.“.”) or a colon (i.e. “:”) by press- (congruent condition), two arrows pointing in ing a corresponding button using the right the opposite direction of the target (incongru- hand as quickly and accurately as possible. ent condition), or two dashes (neutral condi- They were also instructed to look at the fixa- tion). Each trial had the following structure: tion cross at the start of each trial. The probe (1) a central fixation cross (random duration remained on screen until a response was between 400 and 1600 ms); (2) a cue (100 ms); given. The inter-trial interval was 1500 ms. (3) a central fixation cross (400 ms); (4) a tar- There were an equal number of trials for get and its flankers, appearing above or below each type of stimuli location (top or down), the fixation cross (the target remained on the probe location (top or down), and probe type screen until the participant responded or for (“.” or “:”). We used an equal number of tri- 1700 ms if no response occurred); (5) a central als in each condition as a function of these fixation cross [lasting for 3500 ms minus the parameters (i.e., 320 trials = 40 face-pairs × sum of the first fixation period’s duration and 2 face positions × 2 cue types × 2 cue posi- the reaction time (RT)]. RT (in ms) and accu- tions). Each of the 320 trials appeared in a racy (percentage of correct responses) were different random order for each participant recorded for each trial. and each type of stimulation (anodal tDCS The ANT task comprised 288 trials, divided versus sham). Stimuli consisted of 40 differ- in three blocks of 96 trials each (with a short ent face pairs (20 male, 20 female), each pair break between blocks). There were 48 possible displaying neutral-disgust facial expressions, Coussement et al: Does Change in Attention Control Mediate the Impact of 21 tDCS on Attentional Bias for Threat? Limited Evidence from a Double-blind Sham-controlled Experiment in an Unselected Sample randomly selected from a validated version after the stimulation. To be consistent with (Goeleven, Raedt, Leyman, & Verschuere, previous tDCS studies in the field (e.g., Fregni 2008) of the Karolinska Directed Emotional et al., 2005; Heeren et al., 2017), the second Faces (Lundqvist, Flykt, & Öhman, 1998), stimulation was carried out after an exact which is a standardized set of Caucasian 48h-interval to avoid carry-over effect. The emotional faces. Faces were standardized order of the anodal and sham stimulation was for size (326 × 329 pixels). The task was pro- counterbalanced across participants. grammed and presented using E-Prime 2.0 Professional® (Psychology Software Tools, Procedure Pittsburgh, PA, USA). Participants first filled in the question- naires. Then, the two stimulation-sessions Transcranial direct current stimulation were conducted. At the beginning of each Direct electrical current was delivered by a ­stimulation, electrodes were soaked in saline battery-driven stimulator (Neuroconn, GmbH, solution and placed on the participant’s scalp Ilmenau, Germany) and applied via a saline- using the electrode montage depicted above. soaked pair of surface sponge rubber elec- Following 5 min of stimulation (anodal or trodes (35 cm2). We used a sham-controlled sham tDCS), participants started with the ANT. within-subject design in which all partici- The ANT lasted approximately 20 minutes. pants serve as their own control, a design that Immediately after the stimulation, partici- substantially increases statistical power. To pants started with the probe discrimination stimulate the left dlPFC, the anode electrode task. The probe discrimination task lasted was vertically positioned centered over the F3 approximately 15 minutes. Participants were according to the 10–20 international system asked to perform both tasks as quickly and for electroencephalogram electrode place- accurately as possible. The order of the two ment. The reference electrode (i.e., the cath- tDCS-stimulation conditions was randomly ode) was placed vertically at the ipsilateral arm counterbalanced across participants (i.e. 10 (Cogiamanian, Marceglia, Ardolino, Barbieri, participants received the anodal stimula- & Priori, 2007; Priori et al., 2008). During the tion first, 10 participants received the sham first 30 seconds of stimulation, the current stimulation first), and the second stimulation was ramped up to 2 mA and then delivered was carried out after an exact 48h-interval. constantly for 25 minutes. At the end of the Each session was administrated individually stimulation, the current was ramped down to in a dimly lit and quiet room. 0 mA over 30 seconds. For the sham stimula- tion, the position of the electrodes was iden- Preprocessing and data analytic plan tical to the anodal stimulation; however, the Power analysis current was ramped down after 30 seconds. An a priori power analysis was conducted This procedure is commonly used in tDCS to determine the appropriate total sample research and is known to be an optimal way size for testing hypotheses with the primary to provide the initial sensation of stimulation outcome variables. Based upon previous without the subsequent effects on cortical tDCS studies on attentional networks (e.g., excitability (Nitsche et al., 2008; Ohn et al., Miler et al., 2017) and attentional bias fort 2008). Predefined codes assigned to either threat (e.g., Heeren et al., 2017), we expected sham or real stimulation were used to start a medium-to-large effect size of Cohens’ the stimulator and thus allowed for a double- d = 0.7. Setting the level of α at 0.05, power blind study design. Anodal stimulation, or (1 − β) at 0.80 and expecting a conservative sham stimulation, respectively, started 5 min- correlation of ρ = 0.50 between repeated utes before the beginning of the ANT and was measurements, the power analysis (G*Power delivered for a further 20 minutes. Thus, the 3.1.3, Faul, Erdfelder, Lang, & Buchner, 2007) ANT was performed parallel to the stimulation indicated that a total sample size of 18 par- and the probe discrimination was performed ticipants would yield an adequate power. 22 Coussement et al: Does Change in Attention Control Mediate the Impact of tDCS on Attentional Bias for Threat? Limited Evidence from a Double-blind Sham-controlled Experiment in an Unselected Sample

Data reduction Data analytic plan ANT. Following previous studies (e.g., To investigate the impact of tDCS on the Heeren et al., 2014; Lannoy et al., 2017; three ANT networks, we first computed a Maurage et al., 2014, 2017), we excluded 2 (Stimulation) × 3 (attention networks) data from trials with incorrect responses repeated-measures ANOVAs for RT with and RTs lower than 200 ms or greater than Stimulation (anodal tDCS, Sham) and 2000 ms for each participant at each ses- Attention Network (Alerting, Orienting, sion. Following Fan et al. (2002), we com- Executive Conflict) with repeated measure- puted the alerting effect by subtracting the ment on the two factors and latencies as mean (i.e., RT or accuracy score) for double dependent variable. To investigate the impact cue trials from the mean for no cue trials of tDCS on attentional bias, we then com- (No cue – Double cue); the orienting effect puted a paired t-test to compare bias scores by subtracting the mean for spatial cue tri- following sham and anodal stimulation. als from the mean result for center cue trials Following previous studies in the field (Center cue – Spatial cue); and the execu- (e.g., Fregni et al., 2005), we also examined tive conflict effect by subtracting the mean potential stimulation-order effect. To do for congruent trials (summed across cue so, we computed a 2 (Stimulation: anodal types) from the mean for incongruent trials versus sham) × 3 (Attentional Networks: (Incongruent – Congruent). For both alert- Alerting versus Orienting versus Executive ing and orienting effects, greater subtrac- control) × 2 (Order: anodal first versus sham tion scores for RT indicate greater efficiency. first) ANOVA with repeated measurement In contrast, greater subtraction scores for on the first two factors and ANT latencies as RT on executive conflict indicate increased dependent variable. Likewise, we computed difficulty with executive control of atten- a 2 (Stimulation: anodal versus sham) × 2 tion (Fan, McCandliss, Fossella, Flombaum, (Order: anodal first versus sham first) ANOVA & Posner, 2005). with repeated measurement on the first fac- Probe discrimination task. First, trials tor and probe discrimination’s d scores as with incorrect responses were excluded dependent variable. (2.59% of all the trials following sham; All statistical analyses were performed 2.83% of all the trials following anodal using SPSS software package (version 20.0). tDCS). Second, RTs lower than 200ms or The significance level was set at an alpha level greater than 2000ms were removed from of .05 (bilateral). Effect sizes are reported 2 analyses (0.72% of all the trials follow- in the form of partial eta-squared (η p) for ing sham; 0.70% of all the trials following ANOVA and Cohen’s d using the formula for anodal tDCS). Then, to assess attentional paired comparison (i.e. mean pairs difference bias, we calculated a bias score for each par- divided by the pooled SD). ticipant at each session by subtracting the mean latencies when the probe appeared in Results the same location as the threatening stimuli Change in attentional networks from the mean latency when the probe and The Stimulation × Network interaction the threatening stimuli appeared at differ- was not significant, F(2,38) = 0.07, p = .93, 2 ent locations. Hence, positive bias scores η p < .01, implying that the stimulation did represent attention bias toward social not modulate the attentional networks. threat, and negative bias scores represent Likewise, the main effect of Stimulation, 2 bias away from threat, or equivalently, bias F(1,19) = .24, p = .63, η p = .01, was not sig- toward neutral faces. This bias score is the nificant. Yet, consistent with earlier studies, most frequently used index to determine the main effect of Network, F(2,19) = 130.04, 2 attentional bias from a probe discrimina- p < .0001, η p = .87, was significant, implying tion task procedure (e.g., MacLeod et al., that the three networks did differ. Results are 1986; Mogg, Philippot, & Bradley, 2004). shown in Table 2. Coussement et al: Does Change in Attention Control Mediate the Impact of 23 tDCS on Attentional Bias for Threat? Limited Evidence from a Double-blind Sham-controlled Experiment in an Unselected Sample Table 2: Differential latencies (in milli- following anodal tDCS. As such, our findings seconds) for bias scores and the three are at odds with previous observation among ­attentional networks as a function of healthy volunteers of dlPFC-based beneficial the stimulation. impact of non-invasive brain stimulation pro- cedures on attentional bias for threat (e.g., Sham Anodal De Raedt et al., 2010). Likewise, our results stimulation tDCS are also at odds with the previous observa- Mean (SD) Mean (SD) tion that a 20-minute anodal tDCS over the Alerting 28.42 (18.44) 31.27 (12.43) left dlPFC was associated with greater exec- Orienting 33.83 (21.79) 34.53 (16.29) utive network of the attention in healthy participants (Miler et al., 2017). Altogether, Executive 97.16 (25.67) 97.61 (23.09) our failures to replicate these findings ren- Bias score –.87 (9.32) .46 (13.04) dered unstable our main hypothesis that the improvement in attention control mediates Change in attentional bias the impact of anodal tDCS on attentional As shown in Table 2, there was no signifi- bias for threat. There are various potential cant difference regarding attentional bias explanations for our failure to replicate. between anodal and sham stimulations, First, as pointed out in the results section, t(19)= .34, p = .74, d = .33. A comparison our participants did not exhibit an atten- between mean latencies when the probe tional bias in the absence of stimulation—that appeared in the same location as the threat- is, following the sham stimulation. This is ening stimuli and the mean latency when the unfortunate as several studies indicated that probe and the threatening stimuli appeared the presence of attentional bias can be con- at different locations indicated that there sidered as a critical factor for the plasticity was no significant difference between of attentional bias (e.g., Heeren, Philippot, & latencies of the former relative to the latter Koster, 2015; Kuckertz et al., 2014; Mogoaşe ­following the sham condition, t(19) = 0.46, et al., 2014). On the other hand, the observa- p = .88, d = .10 (hence, no attentional bias tion of a significant change in the magnitude in the absence of stimulation, i.e. baseline). of attentional bias for threat following the Similarly, there was no attentional bias for combination of anodal tDCS and ABM proce- threat following the anodal stimulation, dure among participants who were explicitly t(19) = 0.12, p = .90, d = .03. selected to not possess an attentional bias for threat tends to rule out the hypothesis that Stimulation-order effect the modification of attention bias for threat The ANOVA revealed a non-significant Order × does mandatorily requires the presence of Stimulation × Network interaction for the ANT, an attentional bias at baseline (Clarke et 2 F(2,38) = 2.06, p = .14, η p = .10. Likewise, the al., 2014). Likewise, a recent meta-analysis Order × Stimulation interaction was not sig- revealed that, in average, clinical anxious nificant for the probe discrimination task, individuals enrolled in randomized con- 2 F(1,19) = .20, p = .66, η p = .01. These results trolled trials for ABM are not characterized confirmed that the present findings did not by attentional bias for threat (Kruijt, Parsons, mirror a stimulation-order effect. & Fox, 2018). Although this observation sug- gests that the beneficial impact of ABM on Discussion anxiety symptoms may operate via path- The main aim of the present study was to ways other than through attentional bias for examine the mediational role of attention threat (e.g., Kraft, Jonassen, Heeren, Harmer, control in the impact of anodal tDCS over Stiles, & Landrø, in press), it also seemingly the left dlPFC on attentional bias for threat, challenge the claim that anxiety is associ- as compared to sham tDCS. Neither atten- ated with attentional bias for threat. In our tion control nor attentional bias did improve sample, the lack of significant correlation 24 Coussement et al: Does Change in Attention Control Mediate the Impact of tDCS on Attentional Bias for Threat? Limited Evidence from a Double-blind Sham-controlled Experiment in an Unselected Sample between trait anxiety scores and attentional Reed, 2002), a self-report measure assessing bias for threat in the absence of stimulation, attention control as a trait-like construct. r(20) = .15, p = .52, dovetails with this claim. As such, one cannot exclude that our find- Yet, most of those studies were conducted ings would be been different using the among individuals with clinical anxiety dis- Attention Control Scale. On the other hand, orders and trait anxiety score might be less the examination of the impact of a single than ideal to depict anxiety and related session of tDCS on a self-report measure psychopathology (e.g., Heeren, Bernstein, has no relevance, especially given the trait- & McNally, 2018). As such, it remains par- like nature of attention control as assessed ticularly difficult to interpret the absence of using the Attention Control Scale. Moreover, attentional bias in the present sample with- the ANT has been repeatedly used in anxiety out an anxious comparison group. research (e.g., Heeren, Maurage, & Philippot, Second, we used a dot-probe task to assess 2015; Moriya & Tanno, 2009; Pacheco- attentional bias for threat. Yet, like most Unguetti et al., 2011) and has been already extant procedures for assessing attentional used in tDCS research (e.g., Miler et al., 2017). bias, the dot-probe task exhibits poor psy- Likewise, prior research has shown that the chometric properties (for a recent review, distinct ANT’s attentional networks, and par- see McNally, 2018). Recently developed ticularly the executive conflict index, were experimental paradigms enabling optimal strongly associated with attentional bias assessment of attentional bias might be for threat (e.g., Enock, Hofmann, & McNally, more appropriate in future research agen- 2014; Heeren & McNally, 2016; Heeren, das (e.g., Price et al., 2015; Sanchez-Lopez, Mogoaşe, McNally, Schmitz, & Philippot, Vanderhasselt, Allaert, Baeken, & De Raedt, 2015). The presence of a significant correla- 2018; Zvielli, Bernstein, & Koster, 2014). tion between the executive conflict index Third, we chose disgust faces as threat and attentional bias for threat in the absence cues. The rationale behind our decision of stimulation, r(20) = .46, p < .05, corrobo- was twofold. First, disgust conveys a mes- rated this observation in our sample. sage of aversion or rejection in both clinical Fifth, while previous tDCS studies admin- and nonclinical samples (Rozin, Lowery, & istered the probe discrimination task during Ebert, 1994). Second, previous studies sup- the stimulation, ours did after. Although we porting the effectiveness of either anodal decided to set up our experiment in accord- tDCS or ABM in reducing attentional bias ance to our mediational hypothesis vis-à-vis for threat relied on faces expressing disgust the potential impact of dlPFC-induced atten- as threatening stimuli (e.g., Heeren et al., tion control improvement on attentional bias 2017; Pieters et al., 2016; Sanchez-Lopez et mitigation, to the best of our knowledge, no al., 2018). However, other studies have relied previous studies did collect post-stimulation on fearful or angry faces as threat cues for data. Accordingly, our failure to replicate the assessing attentional bias for threat via the tDCS-induced mitigation of attentional bias dot-probe task (for a review, see van Rooijen, may merely mirror the non-persistence of the Ploeger, & Kret, 2017). As such, one cannot dlPFC-induced benefits at post-stimulation, exclude that the findings would have been that is, during the completion of the probe radically different using fearful or angry faces discrimination task. Consequently, an impor- as threat cues. Accordingly, future research tant next step would thus be to compare the could examine whether the impact of anodal online versus offline tDCS-induced mitigation tDCS on attentional bias for threat does vary of attentional bias for threat in anxious and across different types of threat cues. nonanxious samples. Fourth, we used the ANT to assess atten- Sixth, although our decision to target the tion control. Yet, most of the attentional bias left dlPFC relied on previous rTMS and tDCS research has assessed attention control using studies (Clarke et al., 2014; De Raedt et al., the Attention Control Scale (Derryberry & 2010; Heeren, Baeken, et al., 2015; Heeren Coussement et al: Does Change in Attention Control Mediate the Impact of 25 tDCS on Attentional Bias for Threat? Limited Evidence from a Double-blind Sham-controlled Experiment in an Unselected Sample et al., 2017), vlPFC has been also associated a complementary power analyses indicated with attentional bias for threat and anxiety that a total sample size of at least 787 par- (Fox & Pine, 2012; Hartley & Phelps, 2010; ticipants would be required to yield enough Heeren, Dricot, et al., 2017). Consequently, power to detect a small effect size (i.e., Cohen’s one cannot exclude that vlPFC would have d = .10) in the present study. However, such been a better target. On the other hand, sev- small effect sizes have extremely limited rel- eral reviews and meta-analysis suggested that evance for translational research. left dlPFC does constitute an optimal tar- In conclusion, these limitations notwith- get to improve top-down control processes standing, this study constitutes, to the best among healthy volunteers (e.g., Brunoni & of our knowledge, the first attempt to test Vanderhasselt, 2014). Future studies should whether attention control mediates the thus further delineate the respective contri- impact of dlPFC-based anodal tDCS on atten- bution of both ventral and dorsal compart- tional bias for threat. Although our findings ments of PFC vis-à-vis the modification of do not dovetail with prior research, we pro- attentional bias in clinical and nonclinical posed several potential explanations for our samples. In the same vein, although our deci- failure to replicate. Altogether, we believe sion to not rely on bipolar-balanced mon- the present null findings study to be particu- tage—that is, the anode centered over the larly useful for future empirical investigation left dlPFC and the cathode centered over the and meta-research in the field. right dlPFC—was based upon prior research in the field of attentional bias (e.g., Clarke Acknowledgements et al., 2014; Heeren, Baeken, et al., 2015; This research was supported by a PhD stu- Heeren et al., 2017) and ANT (e.g., Miler et al., dentship Grant from “L’Hôpital Psychiatrique 2017), different montage may have yielded du Beau Vallon asbl” awarded to Charlotte various outputs. Especially, the position of Coussement. This work also received the the reference electrode—in the present case, support from the Belgian Foundation for the cathode— may have impacted on the Vocation («Vocatio»), The Helaers Foundation overall current flow pattern as the wider the for Medical Research in Neuroscience, and the distance between the two electrodes, the «World Excellence postdoctoral fellowship smaller the current density under the elec- grant from the Competitive Cluster in Health trodes (e.g., DaSilva, Volz, Bikson, & Fregni, and Life Sciences of Wallonia—BioWin» 2011; Moliadze, Antal, & Paulus, 2010). (sub/2015/228106243177), all awarded In this way, although we relied on an extra- to Alexandre Heeren. Alexandre Heeren cephalic placement for the reference elec- (Research Associate) and Pierre Maurage trode to avoid any cortical influence of the (Research Associate) are funded by the cathode, this montage may have reduced the Belgian National Fund for Scientific Research overall current density. Future experiments (F.R.S.-FNRS, Belgium). These foundations are thus also needed to clarify this issue. did not exert any editorial influence over Finally, our a priori power analysis was this article. based upon previous tDCS studies on atten- tional networks (e.g., Miler et al., 2017) and Competing Interests attentional bias (e.g., Heeren et al., 2017). The authors have no competing interests to However, although our sample size had ade- declare. quate power to detect medium effect sizes, our analysis would have benefited from a References larger sample size. On the other hand, nei- Adolphs, R., Tranel, D., Damasio, H., ther the p-values nor the effect sizes associ- & Damasio, A. R. (1995). Fear and ated with our nonsignificant effects even the human amygdala. The Journal of approaches statistical significance. Moreover, ­Neuroscience: The Official Journal of the 26 Coussement et al: Does Change in Attention Control Mediate the Impact of tDCS on Attentional Bias for Threat? Limited Evidence from a Double-blind Sham-controlled Experiment in an Unselected Sample Society for Neuroscience, 15(9), 5879–5891. threat appraisal and fear learning. DOI: https://doi.org/10.1523/JNEURO- Depression and Anxiety, 28(1), 5–17. DOI: SCI.15-09-05879.1995 https://doi.org/10.1002/da.20733 Bardeen, J. R., & Orcutt, H. K. (2011). Browning, M., Holmes, E. A., Murphy, ­Attentional control as a moderator of S. E., Goodwin, G. M., & Harmer, C. J. the relationship between posttrau- (2010). Lateral prefrontal cortex ­mediates matic stress symptoms and attentional the cognitive modification of atten- threat bias. Journal of Anxiety Disorders, tional bias. Biological Psychiatry, 67(10), 25(8), 1008–1018. DOI: https://doi. 919–925. DOI: https://doi.org/10.1016/j. org/10.1016/j.janxdis.2011.06.009 biopsych.2009.10.031 Beck, A. T., & Clark, D. A. (1997). An infor- Bruchon-Schweitzer, M., & Paulhan, mation processing model of anxiety:­ I. (1993). Adaptation francophone de Automatic and strategic processes. l’inventaire d’anxiete Trait-Etat (Forme Behaviour­ Research and Therapy, 35(1), Y) de Spielberger. (Editions du Centre 49–58. DOI: https://doi.org/10.1016/ ­Psychologie Appliquée). Paris, France. S0005-7967(96)00069-1 Brunoni, A. R., & Vanderhasselt, M.-A. Beck, A. T., Steer, R. A., & Brown, G. K. (2014). Working memory improvement (1998). Inventaire de Dépression de with non-invasive brain stimulation of Beck – 2ème édition (IDB-II) [Beck Depres- the dorsolateral prefrontal cortex: A sion Inventory – 2nd edition (BDI-II)] ­systematic review and meta-analysis. Brain (­Editions du Centre de Psychologie Appli- and Cognition, 86, 1–9. DOI: https://doi. quée). Paris, France. org/10.1016/j.bandc.2014.01.008 Bishop, S. J. (2004). State Anxiety Modula- Cisler, J. M., & Koster, E. H. W. (2010). tion of the Amygdala Response to Unat- Mechanisms of attentional biases tended Threat-Related Stimuli. Journal towards threat in anxiety disorders: An of Neuroscience, 24(46), 10364–10368. integrative review. Clinical Psychology DOI: https://doi.org/10.1523/JNEURO- Review, 30(2), 203–216. DOI: https://doi. SCI.2550-04.2004 org/10.1016/j.cpr.2009.11.003 Bishop, S. J. (2008). Neural Mechanisms Clarke, P. J. F., Browning, M., Hammond, G., Underlying Selective Attention to Threat. Notebaert, L., & MacLeod, C. (2014). The Annals of the New York Academy of Causal Role of the Dorsolateral Prefrontal ­Sciences, 1129(1), 141–152. DOI: https:// Cortex in the Modification of Attentional doi.org/10.1196/annals.1417.016 Bias: Evidence from Transcranial Direct Bishop, S. J. (2009). Trait anxiety and Current Stimulation. Biological Psychia- ­impoverished prefrontal control of atten- try, 76(12), 946–952. DOI: https://doi. tion. Nature Neuroscience, 12(1), 92–98. org/10.1016/j.biopsych.2014.03.003 DOI: https://doi.org/10.1038/nn.2242 Cogiamanian, F., Marceglia, S., Ardolino, Britton, J. C., Bar-Haim, Y., Clementi, M. G., Barbieri, S., & Priori, A. (2007). A., Sankin, L. S., Chen, G., Shechner, T., Improved isometric force endurance Pine, D. S., et al. (2013). Training-associ- after transcranial direct current stimu- ated changes and stability of attention lation over the human motor cortical bias in youth: Implications for Attention areas. The European Journal of Neurosci- Bias Modification Treatment for pedi- ence, 26(1), 242–249. DOI: https://doi. atric anxiety. Developmental Cognitive org/10.1111/j.1460-9568.2007.05633.x Neuroscience, 4, 52–64. DOI: https://doi. Cristea, I. A., Kok, R. N., & Cuijpers, P. org/10.1016/j.dcn.2012.11.001 (2015). Efficacy of cognitive bias modifica- Britton, J. C., Lissek, S., Grillon, C., tion interventions in anxiety and depres- ­Norcross, M. A., & Pine, D. S. (2011). sion: Meta-analysis. The British Journal Development of anxiety: The role of of Psychiatry: The Journal of ­Mental Coussement et al: Does Change in Attention Control Mediate the Impact of 27 tDCS on Attentional Bias for Threat? Limited Evidence from a Double-blind Sham-controlled Experiment in an Unselected Sample ­Science, 206(1), 7–16. DOI: https://doi. pendence of attentional networks. org/10.1192/bjp.bp.114.146761 Journal of Cognitive Neuroscience, DaSilva, A. F., Volz, M. S., Bikson, M., & 14(3), 340–347. DOI: https://doi. Fregni, F. (2011). Electrode positioning org/10.1162/089892902317361886 and montage in transcranial direct cur- Faul, F., Erdfelder, E., Lang, A.-G., & rent stimulation. Journal of Visualized Buchner,­ A. (2007). G*Power 3: A flex- Experiments: JoVE, 51. DOI: https://doi. ible statistical power analysis program for org/10.3791/2744 the social, behavioral, and biomedical sci- Davis, M., & Whalen, P. J. (2001). The amyg- ences. Behavior Research Methods, 39(2), dala: Vigilance and emotion. Molecular 175–191. DOI: https://doi.org/10.3758/ Psychiatry, 6(1), 13–34. DOI: https://doi. BF03193146 org/10.1038/sj.mp.4000812 Fox, N. A., & Pine, D. S. (2012). Tempera- De Raedt, R., Leyman, L., Baeken, C., ment and the emergence of anxiety dis- Van Schuerbeek, P., Luypaert, R., orders. Journal of the American Academy ­Vanderhasselt, M.-A., & Dannlowski, of Child and Adolescent Psychiatry, 51(2), U. (2010). Neurocognitive effects of HF- 125–128. DOI: https://doi.org/10.1016/j. rTMS over the dorsolateral prefrontal jaac.2011.10.006 cortex on the attentional processing Fregni, F., Boggio, P. S., Nitsche, M., of emotional information in healthy ­Bermpohl, F., Antal, A., Feredoes, E., women: An event-related fMRI study. Pascual-Leone, A., et al. (2005). Anodal Biological Psychology, 85(3), 487–495. transcranial direct current stimulation DOI: https://doi.org/10.1016/j.biopsy- of prefrontal cortex enhances working cho.2010.09.015 ­memory. Experimental Brain Research, Derryberry, D., & Reed, M. A. (2002). 166(1), 23–30. DOI: https://doi. ­Anxiety-related attentional biases and org/10.1007/s00221-005-2334-6 their regulation by attentional con- Goeleven, E., Raedt, R. D., Leyman, L., & trol. Journal of Abnormal Psychology, Verschuere, B. (2008). The ­Karolinska 111(2), 225–236. DOI: https://doi. Directed Emotional Faces: A valida- org/10.1037/0021-843X.111.2.225 tion study. Cognition and Emotion, Enock, P. M., Hofmann, S. G., & McNally, 22(6), 1094–1118. DOI: https://doi. R. J. (2014). Attention bias modification org/10.1080/02699930701626582 training via smartphone to reduce social Hartley, C. A., & Phelps, E. A. (2010). anxiety: A randomized, controlled, multi- ­Changing fear: The neurocircuitry of session experiement. Cognitive Therapy ­emotion regulation. Neuropsychophar- and Research, 38, 200–216. DOI: https:// macology, 35(1), 136–146. DOI: https:// doi.org/10.1007/s10608-014-9606-z doi.org/10.1038/npp.2009.121 Eysenck, M. W., & Derakshan, N. (2011). Heeren, A., Baeken, C., Vanderhasselt, New perspectives in attentional control M.-A., Philippot, P., & De Raedt, R. theory. Personality and Individual Differ- (2015). Impact of anodal and cathodal ences, 50(7), 955–960. DOI: https://doi. transcranial direct current stimulation org/10.1016/j.paid.2010.08.019 over the left dorsolateral prefrontal cor- Fan, J., McCandliss, B. D., Fossella, tex during attention bias modification: J., ­Flombaum, J. I., & Posner, M. I. An eye-tracking study. PloS One, 10(4), (2005). The activation of attentional e0124182. DOI: https://doi.org/10.1371/ networks. NeuroImage, 26(2), 471–479. journal.pone.0124182 DOI: https://doi.org/10.1016/j.neuroim- Heeren, A., Bernstein, E. E., & McNally, R. age.2005.02.004 J. (2018). Deconstructing trait anxiety: Fan, J., McCandliss, B. D., Sommer, A network perspective. Anxiety, Stress, & T., Raz, A., & Posner, M. I. (2002). Coping, 31, 262–276. DOI: https://doi. Testing the efficiency and inde- org/10.1080/10615806.2018.1439263 28 Coussement et al: Does Change in Attention Control Mediate the Impact of tDCS on Attentional Bias for Threat? Limited Evidence from a Double-blind Sham-controlled Experiment in an Unselected Sample Heeren, A., Billieux, J., Philippot, P., De Research, 269, 147–154. DOI: https:// Raedt, R., Baeken, C., de Timary, P., doi.org/10.1016/j.bbr.2014.04.043 Vanderhasselt, M.-A., et al. (2017). Heeren, A., Maurage, P., & Philippot, P. Impact of transcranial direct current (2015). Revisiting attentional process- stimulation on attentional bias for threat: ing of non-emotional cues in social A proof-of-concept study among individ- anxiety: A specific impairment for the uals with social anxiety disorder. Social, orienting network of attention. Psy- Cognitive, and Affective Neuroscience, 21, chiatry Research, 228(1), 136–142. 251–260. DOI: https://doi.org/10.1093/ DOI: https://doi.org/10.1016/j.psy- scan/nsw119 chres.2015.04.030 Heeren, A., Billieux, J., Philippot, P., & Heeren, A., Maurage, P., Rossignol, Maurage, P. (2015). Looking under the M., Vanhaelen, M., Peschard, V., & hood of executive function impairments Eeckhout,­ C. (2012). The self-report ver- in psychopathology: A commentary on sion of the Liebowitz social anxiety scale: “Advancing understanding of executive Psychometric properties of the French function impairments and psychopa- version. Canadian Journal of Behavioural thology: Bridging the gap between clini- Science, 44(2), 99–107. DOI: https://doi. cal and cognitive approaches”. Frontiers org/10.1037/a0026249 in Psychology, 6, 1170. DOI: https://doi. Heeren, A., & McNally, R. J. (2016). An org/10.3389/fpsyg.2015.01170 integrative network approach to social Heeren, A., Coussement, C., & McNally, R. anxiety disorder: The complex dynamic J. (2016). Untangling attention bias mod- interplay among attentional bias for ification from emotion: A double-blind threat, attentional control, and symp- randomized experiment with individu- toms. Journal of Anxiety Disorders, 42, als with social anxiety disorder. Journal 95–104. DOI: https://doi.org/10.1016/j. of Behavior Therapy and Experimental janxdis.2016.06.009 Psychiatry, 50, 61–67. DOI: https://doi. Heeren, A., Mogoaşe, C., McNally, R. J., org/10.1016/j.jbtep.2015.05.005 Schmitz, A., & Philippot, P. (2015). Does Heeren, A., De Raedt, R., Koster, E. H. W., attention bias modification improve & Philippot, P. (2013). The (neuro)cogni- attentional control? A double-blind rand- tive mechanisms behind attention bias omized experiment with individuals with modification in anxiety: Proposals based social anxiety disorder. Journal of Anxiety on theoretical accounts of attentional Disorders, 29, 35–42. DOI: https://doi. bias. Frontiers in Human Neuroscience, org/10.1016/j.janxdis.2014.10.007 7, 119. DOI: https://doi.org/10.3389/ Heeren, A., Mogoaşe, C., Philippot, P., fnhum.2013.00119 & McNally, R. J. (2015). Attention Heeren, A., Dricot, L., Billieux, J., bias modification­ for social anxiety: ­Philippot, P., Grynberg, D., de Timary, A systematic­ review and meta-anal- P., & Maurage, P. (2017). Correlates of ysis. Clinical Psychology Review, 40, social exclusion in social anxiety disor- 76–90. DOI: https://doi.org/10.1016/j. der: An fMRI study. Scientific Reports, cpr.2015.06.001 7, 260. DOI: https://doi.org/10.1038/ Heeren, A., Peschard, V., & Philippot, P. s41598-017-00310-9 (2012). The Causal Role of Attentional Heeren, A., Maurage, P., Perrot, H., De Bias for Threat Cues in Social Anxiety: A Volder, A. G., Renier, L., Araneda, R., Test on a Cyber-Ostracism Task. Cognitive Lacroix, E., Decat, M., Deggouj, N., & Therapy and Research, 36(5), 512–521. Philippot, P. (2014). Tinnitus ­specifically DOI: https://doi.org/10.1007/s10608- alters the top-down ­executive sub-com- 011-9394-7 ponent of attention: Evidence from the Heeren, A., Philippot, P., & Koster, E. H. Attention Network Task. Behavioural Brain W. (2015). Impact of the temporal sta- Coussement et al: Does Change in Attention Control Mediate the Impact of 29 tDCS on Attentional Bias for Threat? Limited Evidence from a Double-blind Sham-controlled Experiment in an Unselected Sample bility of preexistent attentional bias for attention bias modification treatment threat on its alteration through attention for anxiety disorders. Depression and bias modification. Journal of Behavior ­Anxiety, 32(6), 383–391. DOI: https:// Therapy and Experimental Psychiatry, 49, doi.org/10.1002/da.22344 69–75. DOI: https://doi.org/10.1016/j. Lundqvist, D., Flykt, A., & Öhman, jbtep.2014.10.012 A. (1998). The Karolinska Directed Heimberg, R. G., Brozovich, F. A., & Rapee, ­Emotional Faces – KDEF, CD ROM from R. M. (2010). Chapter 15 – A Cognitive­ Department of Clinical Neuroscience, Behavioral Model of Social Anxiety Psychology section, Karolinska Institutet. ­Disorder: Update and Extension. In: MacLeod, C., Mathews, A., & Tata, P. Social Anxiety (Second Edition), 395–422. (1986). Attentional bias in ­emotional dis- San Diego: Academic Press. DOI: https:// orders. Journal of Abnormal ­Psychology, doi.org/10.1016/B978-0-12-375096- 95(1), 15–20. DOI: https://doi. 9.00015-8 org/10.1037/0021-843X.95.1.15 Kraft, B., Jonassen, R., Heeren, A., MacLeod, C., Rutherford, E., ­Campbell, Harmer, C., Stiles, T., & Landrø, N. I. L., Ebsworthy, G., & Holker, L. (2002). (2018). Attention bias modification in Selective attention and emotional remitted depression is associated with ­vulnerability: Assessing the causal increased interest and leads to reduced basis of their association through adverse impact of anxiety symptoms and the experimental manipulation of negative cognition. Clinical Psychological attentional bias. Journal of Abnormal­ Science, In press. Psychology, 111(1), 107–123. Kruijt, A., Parsons, S., & Fox, E. (2018). A DOI: https://doi.org/10.1037/0021- meta-analysis of bias at baseline in RCTs 843X.111.1.107 of attention bias modification: No evi- Månsson, K. N. T., Carlbring, P., Frick, A., dence for bias towards threat in clinical Engman, J., Olsson, C.-J., Bodlund, anxiety and PTSD. PsyArXiv. DOI: https:// O., Andersson, G., et al. (2013). Altered doi.org/10.31234/osf.io/rfjup neural correlates of affective processing Kuckertz, J. M., Gildebrant, E., ­Liliequist, after internet-delivered cognitive behav- B., Karlström, P., Väppling, C., ior therapy for social anxiety disorder. ­Bodlund, O., Carlbring, P., et al. (2014). Psychiatry Research, 214(3), 229–237. ­Moderation and mediation of the effect of DOI: https://doi.org/10.1016/j.pscy- attention training in social anxiety disor- chresns.2013.08.012 der. Behaviour Research and Therapy, 53, Maurage, P., de Timary, P., Billieux, J., 30–40. DOI: https://doi.org/10.1016/j. ­Collignon, M., & Heeren, A. (2014). brat.2013.12.003 Attentional alterations in alcohol- Lannoy, S., Heeren, A., Moyaerts, N., dependence are underpinned by specific ­Bruneau, N., Evrard, S., Billieux, J., & executive control deficits. Alcoholism: Maurage, P. (2017). Differential impair- Clinical and Experimental Research, 38(7), ments across attentional networks in 20105–12. DOI: https://doi.org/10.1111/ binge drinking. Psychopharmacology, acer.12444 234, 1054–1068. DOI: https://doi. Maurage, P., Heeren, A., Lahaye, org/10.1007/s00213-017-4538-4 M., ­Jeanjean, A., Guettat, L., Liebowitz, M. R. (1987). Social phobia. ­Verellen-Dumoulin, C., Halkin, S., Modern Problems of Pharmacopsychia- Billieux,­ J., & Constant, E. (2017). try, 22, 141–173. DOI: https://doi. ­Attentional impairments in Hungtin- org/10.1159/000414022 ton’s disease: A specific deficit for Linetzky, M., Pergamin-Hight, L., Pine, D. the executive conflict. Neuropsychol- S., & Bar-Haim, Y. (2015). ­Quantitative ogy, 31, 424–436. DOI: https://doi. evaluation of the clinical efficacy of org/10.1037/neu0000321 30 Coussement et al: Does Change in Attention Control Mediate the Impact of tDCS on Attentional Bias for Threat? Limited Evidence from a Double-blind Sham-controlled Experiment in an Unselected Sample McNally, R. J. (2018). Attentional bias for tional bias in response to angry faces threat: Crisis or opportunity? Clinical in adolescents with generalized anxiety psychology review. DOI: https://doi. disorder. The American Journal of Psychia- org/10.1016/j.cpr.2018.05.005 try, 163(6), 1091–1097. DOI: https://doi. Miler, J. A., Meron, D., Baldwin, D. S., & org/10.1176/ajp.2006.163.6.1091 Garner, M. (2017). The Effect of Prefron- Monk, C. S., Telzer, E. H., Mogg, K., tal Transcranial Direct Current Stimula- ­Bradley, B. P., Mai, X., Louro, H. M. C., tion on Attention Network Function in Pine, D. S., et al. (2008). Amygdala and Healthy Volunteers: TDCS AND ATTEN- ventrolateral prefrontal cortex activa- TION NETWORK FUNCTION. Neuromodu- tion to masked angry faces in children lation: Technology at the Neural Interface. and ­adolescents with generalized anxi- DOI: https://doi.org/10.1111/ner.12629 ety disorder. Archives of General Psychia- Mogg, K., & Bradley, B. P. (2002). Selective try, 65(5), 568–576. DOI: https://doi. orienting of attention to masked threat org/10.1001/archpsyc.65.5.568 faces in social anxiety. Behaviour Research Moriya, J., & Tanno, Y. (2009). and Therapy, 40(12), 1403–1414. ­Dysfunction of attentional networks DOI: https://doi.org/10.1016/S0005- for non-emotional processing in 7967(02)00017-7 negative affect. Cognition & Emotion­ Mogg, K., & Bradley, B. P. (2016). Anxiety 23, 1090–1105. DOI: https://doi. and attention to threat: Cognitive mecha- org/10.1080/02699930802335018 nisms and treatment with attention bias Nitsche, M. A., Cohen, L. G., ­Wassermann, modification. Behaviour Research and E. M., Priori, A., Lang, N., Antal, Therapy, 87, 76–108. DOI: https://doi. A., ­Pascual-Leone, A., et al. (2008). org/10.1016/j.brat.2016.08.001 ­Transcranial direct current stimulation: Mogg, K., Philippot, P., & Bradley, B. P. State of the art 2008. Brain Stimula- (2004). Selective attention to angry tion, 1(3), 206–223. DOI: https://doi. faces in clinical social phobia. Journal of org/10.1016/j.brs.2008.06.004 Abnormal Psychology, 113(1), 160–165. Ohn, S. H., Park, C.-I., Yoo, W.-K., Ko, M.-H., DOI: https://doi.org/10.1037/0021- Choi, K. P., Kim, G.-M., Kim, Y.-H., et al. 843X.113.1.160 (2008). Time-dependent effect of tran- Mogoaşe, C., David, D., & Koster, E. H. scranial direct current stimulation on the W. (2014). Clinical Efficacy of Atten- enhancement of working memory. Neu- tional Bias Modification Procedures: An roreport, 19(1), 43–47. DOI: https://doi. Updated Meta-Analysis: Clinical Effi- org/10.1097/WNR.0b013e3282f2adfd cacy of Attention Retraining. Journal of Pacheco-Unguetti, A. P., Acosta, A., Clinical Psychology, 70(12), 1133–1157. ­Marqués, E., & Lupiáñez, J. (2011). DOI: https://doi.org/10.1002/jclp.22081 Alterations of the attentional net- Moliadze, V., Antal, A., & Paulus, W. (2010). works in patients with anxiety Electrode-distance dependent after- ­disorders. ­Journal of Anxiety Disor- effects of transcranial direct and random ders 25, 888–895. DOI: https://doi. noise stimulation with extracephalic ref- org/10.1016/j.janxdis.2011.04.010 erence electrodes. Clinical Neurophysiol- Paulewicz, B., Blaut, A., & Kłosowska, J. ogy: Official Journal of the International (2012). Cognitive effects of attentional Federation of Clinical Neurophysiology, training depend on attentional con- 121(12), 2165–2171. DOI: https://doi. trol. Journal of Anxiety Disorders, 43(4), org/10.1016/j.clinph.2010.04.033 272–277. DOI: https://doi.org/10.2478/ Monk, C. S., Nelson, E. E., McClure, E. B., v10059-012-0030-5 Mogg, K., Bradley, B. P., Leibenluft, E., Peers, P. V., Simons, J. S., & Lawrence, A. D. Pine, D. S., et al. (2006). Ventrolateral (2013). Prefrontal control of attention to prefrontal cortex activation and atten- threat. Frontiers in Human Neuroscience, Coussement et al: Does Change in Attention Control Mediate the Impact of 31 tDCS on Attentional Bias for Threat? Limited Evidence from a Double-blind Sham-controlled Experiment in an Unselected Sample 7, 24. DOI: https://doi.org/10.3389/ DOI: https://doi.org/10.1037/0022- fnhum.2013.00024 3514.66.5.870 Petersen, S. E., & Posner, M. I. (2012). The Sanchez-Lopez, A., Vanderhasselt, M.-A., Attention System of the Human Brain: 20 Allaert, J., Baeken, C., & De Raedt, Years After. Annual Review of Neuroscience, R. (2018). Neurocognitive mechanisms 35, 73–89. DOI: https://doi.org/10.1146/ behind emotional attention: Inverse annurev-neuro-062111-150525 effects of anodal tDCS over the left and Pieters, E. K., Raedt, R. D., Enock, P. right DLPFC on gaze disengagement M., Putter, L. M. S. D., Braham, H., from emotional faces. Cognitive, Affec- McNally, R. J., & Koster, E. H. W. (2016). tive, & Behavioral Neuroscience, 18(3), Examining a Novel Gamified Approach 485–494. DOI: https://doi.org/10.3758/ to Attentional Retraining: Effects of s13415-018-0582-8 Single and Multiple Session Training. Spielberger, C. D., Gorsuch, R. L., ­Lushene, Cognitive Therapy and Research, 1–17. R., Vagg, P. R., & Jacobs, G. (1983). Man- DOI: https://doi.org/10.1007/s10608- ual for the state-trait anxiety inventory. 016-9803-z (Consulting Psychologists Press). Palo Posner, M. I., & Rothbart, M. K. (2007). Alto, CA. Research on Attention Networks­ Taylor, C. T., Aupperle, R. L., Flagan, T., as a Model for the Integration Simmons, A. N., Amir, N., Stein, M. B., of ­Psychological ­Science. Annual & Paulus, M. P. (2014). Neural ­correlates Review of Psychology, 58(1), 1–23. of a computerized attention ­modification DOI: https://doi.org/10.1146/annurev. program in anxious subjects. Social psych.58.110405.085516 Cognitive and Affective Neuroscience­ , Price, R. B., Kuckertz, J. M., Siegle, G. J., 9(9), 1379–1387. DOI: https://doi. Ladouceur, C. D., Silk, J. S., Ryan, N. org/10.1093/scan/nst128 D., Amir, N., et al. (2015). Empirical Taylor, C. T., Cross, K., & Amir, N. (2016). recommendations for improving the Attentional control moderates the rela- stability of the dot-probe task in clinical tionship between social anxiety symp- research. Psychological Assessment, 27(2), toms and attentional disengagement 365–376. DOI: https://doi.org/10.1037/ from threatening information. Journal pas0000036 of Behavior Therapy and Experimental Priori, A., Mameli, F., Cogiamanian, Psychiatry, 50, 68–76. DOI: https://doi. F., Marceglia, S., Tiriticco, M., org/10.1016/j.jbtep.2015.05.008 ­Mrakic-Sposta, S., Sartori, G., et al. Van Bockstaele, B., Verschuere, B., (2008). Lie-Specific Involvement of Dor- ­Tibboel, H., De Houwer, J., Crombez, solateral Prefrontal Cortex in Decep- G., & Koster, E. (2014). A review of cur- tion. Cerebral Cortex, 18(2), 451–455. rent evidence for the causal impact of DOI: https://doi.org/10.1093/cercor/ attentional bias on fear and anxiety. bhm088 Psychological­ Bulletin, 140(3), 682– Reinholdt-Dunne, M. L., Mogg, K., & 721. DOI: https://doi.org/10.1037/ Bradley,­ B. P. (2009). Effects of anxi- a0034834 ety and attention control on process- van Rooijen, R., Ploeger, A., & Kret, M. E. ing pictorial and linguistic emotional (2017). The dot-probe task to measure ­information. Behaviour Research and emotional attention: A suitable measure­ Therapy, 47(5), 410–417. DOI: https:// in comparative studies? Psychonomic Bulle- doi.org/10.1016/j.brat.2009.01.012 tin & Review, 24, 1686–1717. DOI: https:// Rozin, P., Lowery, L., & Ebert, R. (1994). doi.org/10.3758/s13423-016-1224-1 Varieties of disgust faces and the struc- Vuilleumier, P. (2005). How brains beware: ture of disgust. Journal of Personality Neural mechanisms of emotional and Social Psychology, 66, 870–881. ­attention. Trends in Cognitive Sciences­ , 32 Coussement et al: Does Change in Attention Control Mediate the Impact of tDCS on Attentional Bias for Threat? Limited Evidence from a Double-blind Sham-controlled Experiment in an Unselected Sample 9(12), 585–594. DOI: https://doi. tive Disorders, 203, 84–100. DOI: https:// org/10.1016/j.tics.2005.10.011 doi.org/10.1016/j.jad.2016.05.069 Wong, Q. J. J., & Rapee, R. M. (2016). The aeti- Zvielli, A., Bernstein, A., & Koster, E. ology and maintenance of social anxiety H. W. (2014). Temporal Dynamics of disorder: A synthesis of complimentary Attentional Bias. Clinical Psychological theoretical models and formulation of a Science, 3, 772–788. DOI: https://doi. new integrated model. Journal of Affec- org/10.1177/2167702614551572

How to cite this article: Coussement, C., Maurage, P., Billieux, J., & Heeren, A. (2019). Does Change in Attention Control Mediate the Impact of tDCS on Attentional Bias for Threat? Limited Evidence from a Double-blind Sham-controlled Experiment in an Unselected Sample. Psychologica Belgica, 59(1), 16–32. DOI: https://doi.org/10.5334/pb.449

Submitted: 03 April 2018 Accepted: 06 December 2018 Published: 15 January 2019

Copyright: © 2019 The Author(s). This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International License (CC-BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. See http://creativecommons.org/licenses/by/4.0/.

Psychologica Belgica is a peer-reviewed open access journal published by Ubiquity Press. OPEN ACCESS