اضطراب خصلتی و پویایی کنترل توجه

According to recent theoretical approaches dispositional anxiety is fundamentally linked to neural mechanisms of cognitive control (Braver et al., 2007 and Eysenck et al., 2007). The present study was conducted to further investigate this topic by focusing on the relation between trait anxiety, conflict-processing and dynamic adjustments in attentional allocation. Participants completed a modified version of the face–word Stroop task while an electroencephalogram was recorded. We analyzed behavioral and electrophysiological correlates of conflict processing and conflict-driven modulations in target and distractor processing. Anxiety was not related to general conflict-sensitivity but to individual differences in conflict-driven adjustments in attentional allocation: following a high level of stimulus–response conflict, highly anxious participants allocated more attentional resources to the processing of predominantly task-relevant information and withdrew attention from the processing of predominantly task-irrelevant information. Thus, trait anxiety appears to be closely related to individual differences in dynamic adjustments of attentional control.

In their influential attentional control theory (ACT), Eysenck and colleagues ( Derakshan and Eysenck, 2009 and Eysenck et al., 2007) have recently claimed that anxiety as personality trait is closely related to individual differences in higher-order functions of cognitive control. Specifically, high trait anxiety is thought to bias the balance between a goal-directed attention system and a stimulus-driven attention system in favor of the latter (also see, Corbetta and Shulman, 2002). Consequently, task-irrelevant information should be more intrusive in highly anxious compared to low anxious individuals. Moreover, Eysenck et al. hypothesize that the deficit in attentional control should affect processing efficiency (as typically indexed by reaction times [RTs]) rather than effectiveness (as typically indexed by error-rates), especially in tasks requiring inhibitory control (i.e., inhibiting the distracting influence of task-irrelevant information) and/or attentional set shifting (i.e., shifting attention between multiple task rules). A further postulate of the ACT is that anxious subjects can compensate for the deficiencies in attentional control by recruiting additional cognitive resources. On the neuronal level this should lead to an enhanced activity of brain circuits involved in cognitive control. By now these theoretical assumptions have received considerable empirical support by studies using behavioral indices of attentional processing. For instance, there is a large body of research, demonstrating that emotionally arousing but task-irrelevant stimuli are more intrusive in highly compared to low anxious subjects (Bar-Haim et al., 2007). Other findings indicate that high anxiety is indeed associated with a general deficit in inhibitory control and attentional set-shifting (e.g., Ansari et al., 2008, Derakshan et al., 2009b, Fox, 1993, Fox, 1994, Wieser et al., 2009 and Wood et al., 2001) and that these deficits affect processing efficiency rather than processing effectiveness (Derakshan et al., 2009a). In contrast to this rather homogeneous picture of behavioral findings, prior studies investigating the link between anxiety and neural correlates of attention and cognitive control have yielded more inconsistent results. While some authors report an increased recruitment of neural control mechanisms in highly trait anxious subjects (e.g., Ansari and Derakshan, 2011, Basten et al., 2011, Gray and Braver, 2002 and Telzer et al., 2008) others report the opposite (e.g., Bishop, 2009, Bishop et al., 2004 and Klumpp et al., 2011). Recent theoretical and empirical works suggest that these inconsistencies are partly caused by disregarding the temporal dynamics of cognitive control. For instance, Braver et al. (2007) postulated that low and highly anxious individuals generally differ in the way they exert top–down control. Specifically, highly anxious subjects are proposed to recruit neurocognitive control resources in a transient and reactive manner (i.e., only when control is needed) whereas low anxious subjects are thought to engage control in a rather sustained and proactive way (also see, Fales et al., 2008). The assumption of a reactive recruitment of cognitive-control in highly anxious subjects is also supported by data from our group (Osinsky et al., 2010). In this study, we investigated dynamic adjustments in conflict-processing, measuring event-related potentials (ERP) of the electroencephalogram (EEG) while subjects performed a face–word version of the Stroop task. Results of this study indicate that highly anxious subjects only more strongly engage neural mechanisms of conflict-monitoring when previously exposed to a high level of stimulus–response conflict. Similarly, findings from two other recent studies suggest that highly anxious subjects more strongly recruit mechanisms of conflict-monitoring (Dennis and Chen, 2009) and inhibitory control (Hardin et al., 2009) after seeing fearful faces. In sum these studies indicate that highly anxious subjects especially show a reactive and compensatory recruitment of control resources and goal-directed attention when previously exposed to a highly cognitive demanding or distracting event. The present study was conducted to further investigate the link between anxiety and cognitive control by focusing on behavioral and neurophysiological indices of conflict-processing and attentional allocation. Our main purposes were (1) to investigate the potential relation between anxiety and sensitivity for distracting and task-irrelevant information and (2) to examine how a potential reactive recruitment of cognitive control in highly anxious individuals affects dynamic adjustments in attentional processing. We recorded EEG while subjects performed a modified version of the face–word Stroop task (see Fig. 1). In most trials of this task, a single female or male face was combined with the word man or woman written across the face, resulting in a congruent or incongruent face–word pairing (hereafter ‘face–word stimuli’). In these trials, subjects were asked to discriminate the gender of the face by button-press. In the remaining infrequent trials only a single face (hereafter ‘face-only stimuli’) or word (hereafter ‘word-only stimuli’) was presented, requiring the same discrimination as in the frequent face–word trials. This task allows for testing several predictions of the ACT