Components of attentional bias to threat in high trait anxiety: Facilitated engagement, impaired disengagement, and attentional avoidance
Introduction
Cognitive theories of anxiety disorders have proposed that biased attentive processing of threatening information is a prominent cognitive factor in the causation and maintenance of anxiety (Beck, Emery, & Greenberg, 1985; Eysenck, 1992; Mathews & Mackintosh, 1998; Mathews & MacLeod, 1994; Mogg & Bradley, 1998; Wells & Matthews, 1994; Williams, Watts, MacLeod, & Mathews, 1988). In these models on anxiety-vulnerability it is proposed that biased attention to threat is a function of trait anxiety. Indeed, there is a considerable body of empirical evidence demonstrating enhanced attention to threat in high trait anxious (HTA) compared to low trait anxious (LTA) individuals (for reviews, see Mogg & Bradley, 1998; Williams, Watts, MacLeod, & Mathews, 1997). This attentional bias in favor of threat has been found to exaggerate the effect of negative experiences on anxiety (MacLeod & Hagan, 1992; MacLeod, Rutherford, Campbell, Ebsworthy, & Holker, 2002).
Contemporary models of attention to threat have tried to account for the differential attentive processing of threat in HTA and LTA individuals (Mathews & Mackintosh, 1998; Mogg & Bradley, 1998). A basic premise of these models is that attention to threat essentially is a normal and adaptive process, with attention to threat being determined by both task demands and stimulus input. Inhibitory links are proposed between attention to current tasks and attention to novel, potentially threatening, stimulus input. Attention to current tasks and ongoing behavior is interrupted when stimulus input exceeds a certain threshold and is appraised as highly threatening. In the case of high threat (e.g., the sound of a firing shotgun nearby), everyone will allocate attention to the incoming threatening information. According to the aforementioned models, HTA individuals are characterized by an oversensitive threat appraisal system that appraises incoming stimuli more readily as highly threatening. Consequently, HTA individuals will also direct their attention to moderately threatening information, whereas LTA individuals will ignore such information. This predicted pattern of attentional bias has been observed in several studies (Mogg et al., 2000; Wilson & MacLeod, 2003; but not in Yiend & Mathews, 2001).
A core assumption in theoretical models of attention to threat is that threatening stimuli are more apt to attract visual attention, especially in HTA individuals. However, there is an ongoing debate about the constituents of attentional bias to threat. Posner and colleagues (e.g., Posner, 1980; Posner & Peterson, 1990) have demonstrated that attending to a new stimulus consists of three operations: (1) an initial transient shift of attention to the stimulus; (2) engaging attention with the stimulus; and (3) disengaging attention from the stimulus. Applying a similar distinction with attentional bias seems useful as biased attention in HTA individuals may be related to one or more of these processes. The first two operations have been related to attentional capture by threat, whereas the third operation has been related to increased difficulties in shifting attention away from threat (Fox, Russo, Bowles, & Dutton, 2001; Koster, Crombez, Van Damme, Verschuere, & De Houwer, 2004a; Koster, Crombez, Verschuere, & De Houwer, 2004b). It has been argued that the nature of attentional bias is important in understanding the effects of attentional bias in the development of anxiety disorders. For instance, facilitated attentional capture by threat could be related to improved awareness of threat in the environment. Impaired attentional disengagement may be related to prolonged anxiety states and difficulties in task-performance in the presence of threat. The components involved in attentional bias to threat have mainly been studied in tasks using (1) visual search methodology and (2) the exogenous cueing task.
Various versions of the visual search task have been used to examine attentional engagement and disengagement (for a review, see Öhman, Flykt, & Lundqvist, 2000). In this task, participants are presented with a range of stimuli (i.e., 3×3 matrix) and need to indicate whether or not all stimuli are the same. On some trials, a threatening target is presented amongst neutral distracters. By comparing search slopes for threatening vs. neutral targets one can examine faster threat detection. Attentional disengagement can be investigated by presenting neutral target stimuli in an array of threatening stimuli: slowed responding to the target indicates distraction by threat. Studies in “normal” populations have indicated that threatening target stimuli, for instance angry faces, are detected more rapidly than neutral or positive stimuli (Fox et al., 2000; Hansen & Hansen, 1988). Furthermore, when threatening stimuli are used as distracters they interfere more strongly with target search of, for instance, happy faces (e.g., Fox et al., 2000). Anxiety-related impaired disengagement from threat has been well-established using this task (Byrne & Eysenck, 1995; Rinck, Reinecke, Ellwart, Heuer, & Becker, 2005). Less robust effects have been observed on anxiety-related speeded detection of threat. That is, anxiety-related speeded target detection of threat has been observed in some studies (Byrne & Eysenck, 1995; Öhman et al., 2000; Rinck et al., 2005), whereas other studies repeatedly failed to replicate this effect (e.g., Rinck, Becker, Kellermann, & Roth, 2003). However, an important problem in this task concerns the simultaneous presentation of emotional and neutral stimuli. This causes difficulties in isolating attentional effects for the targets from the distracters. For instance, for high anxious individuals the presentation of neutral distracters can be more threatening than for low anxious individuals, perhaps leading to a stronger attentional capture by the distracters in these individuals.
Another methodology is offered by the emotional modification of the exogenous cueing task. In the exogenous cueing task (Posner, 1980), participants are asked to detect a visual target presented at the left or right side of a fixation cross. On most of the trials, a peripheral cue precedes the target at the same spatial location (“valid” trials). On the remaining trials, the target is presented at the opposite spatial location of the cue (“invalid” trials). Exogenous cues that are presented for a short duration (<300 ms) facilitate responding to target stimuli on valid trials, whereas on invalid trials a reaction time cost is observed. This pattern is referred to as the “cue validity effect”. In the emotional modification of this paradigm, the emotional value of the cue is varied (e.g., threat/neutral). This allows investigating attentional engagement by the threat cue, which is reflected in reaction time benefits in responding to valid threatening trials compared with valid neutral trials. Attentional disengagement is reflected by reaction time costs in responding to invalid threatening trials compared with invalid neutral trials.
Research investigating the components of attention in “normal” individuals has established that threat can modulate attentional engagement and disengagement in the modified cueing paradigm (e.g., Koster et al., 2004a; Koster, Crombez, Van Damme, Verschuere, & De Houwer, 2005). Studies including trait anxiety, however, found that HTA individuals show impaired disengagement from threat but no facilitated attentional capture by threat (Fox, Russo, & Dutton, 2002; Yiend & Mathews, 2001). This could imply that there are no anxiety-related differences in attentional engagement by threat. However, the absence of anxiety-related differences in attentional engagement may also be attributed to methodological issues in the previous studies using the modified cueing task. First, in the study by Fox et al. (2001) schematic faces were used, which may have lacked a sufficient threat value to elicit attentional capture effects. Second, in other studies (Yiend & Mathews, 2001) relatively long picture presentations (500 ms) were used, whereas research in “normal” individuals indicate that attentional engagement effects by threat are only found at shorter presentation durations (200 ms, Koster et al., 2004a).
In the present study, we examined whether attentional bias in HTA individuals is related to facilitated attentional engagement with and/or delayed disengagement from threat. In two experiments, HTA and LTA individuals performed the modified cueing task. The characteristics of this task were optimized to examine both components of attention by (1) presenting neutral, moderately, and highly threatening pictorial cues, and (2) varying the presentation duration of the cues (Experiment 1: 100 and 500 ms; Experiment 2: 200 and 500 ms).
Section snippets
Participants
A large group of first-year psychology students (N=477) were administered the trait version of the State and Trait Anxiety Inventory (STAI-T; Spielberger, Gorsuch, Lushene, Vagg, & Jacobs, 1983). When students scored within the extremes (upper 10%, lower 10%) of the larger group, we invited them to participate in the present study. A total of 47 students were allowed to participate. The LTA group consisted of 24 individuals (male/female ratio was 7:17; mean age=18.9 years). The HTA group was
Experiment 2
This study was designed to further examine some of the unexpected findings from experiment. A first objective was to examine the attentional disengagement component by exploring the time-course of attentional disengagement from threat. At 100 ms picture presentation clear difficulties in disengaging attention were found in Experiment 1. However, at 500 ms no such differences were found. Therefore, in Experiment 2, the 100 ms condition was replaced by a 200 ms condition to assess whether impaired
Acknowledgements
The authors would like to acknowledge Michelle Craske and two anonymous reviewers for their helpful suggestions.
References (40)
- et al.
Selective attention to threat in the dot probe paradigm: Differentiating vigilance and difficulty to disengage
Behaviour Research and Therapy
(2004) - et al.
Individual differences in the selective processing of threatening information, and emotional responses to a stressful life event
Behaviour Research & Therapy
(1992) - et al.
A cognitive-motivational analysis of anxiety
Behaviour Research and Therapy
(1998) - et al.
Anxiety disorders and phobias: A cognitive perspective
(1985) - et al.
Trait anxiety, anxious mood, and threat detection
Cognition and Emotion
(1995) Statistical power analysis for the behavioural sciences
(1988)Origins of phobias and anxiety disorders: Why more women than men?
(2003)- et al.
A simple and sensitive method to measure timing accuracy
Behavior Research Methods, Instruments, and Computers
(2003) - et al.
Anxiety-related attentional biases and their regulation by attentional control
Journal of Abnormal Psychology
(2002) Anxiety: The cognitive perspective
(1992)