When spiders appear suddenly: Spider-phobic patients are distracted by task-irrelevant spiders

https://doi.org/10.1016/j.brat.2007.10.010Get rights and content

Abstract

Fear is thought to facilitate the detection of threatening stimuli. Few studies have examined the effects of task-irrelevant phobic cues in search tasks that do not involve semantic categorization. In a combined reaction time and eye-tracking experiment we investigated whether peripheral visual cues capture initial attention and distract from the execution of goal-directed eye movements. Twenty-one spider-phobic patients and 21 control participants were instructed to search for a color singleton while ignoring task-irrelevant abrupt-onset distractors which contained either a small picture of a spider (phobic), a flower (non-phobic, but similar to spiders in shape), a mushroom (non-phobic, and not similar to spiders in shape), or no picture. As expected, patients’ reaction times were longer on trials with spider distractors. However, eye movements revealed that this was not due to attentional capture by spider distractors; patients more often fixated on all distractors with pictures, but their reaction times were delayed by longer fixation durations on spider distractors. These data do not support automatic capture of attention by phobic cues but suggest that phobic patients fail to disengage attention from spiders.

Introduction

Rapid detection and preferential processing of biologically prepared fear cues is thought to be adaptive for all humans and should be specifically attuned to phobic cues in patients with specific phobias (Mineka & Öhman, 2002). It has been suggested that an underlying preattentive mechanism helps to detect threat cues and is followed by the enhanced attention toward the object (Öhman, Flykt, & Esteves, 2001).

The specialized neural circuitry needed for such preattentive analysis of fear-relevant stimuli is indeed available in all mammals: There are direct neural connections from sensory pathways of the thalamus to the amygdala, which is central to emotional processing (Davis & Whalen, 2001). This connection, often called the “low road” of visual processing (LeDoux, 1996), enables a fast and automatic processing of emotionally relevant information even without an intact visual cortex (Anders et al., 2004). Evidence for initial low-level processing of threat cues in healthy subjects comes from visual search paradigms. In these studies, the task is to search for a fear-relevant picture which is presented amongst multiple neutral pictures, or a neutral picture presented amongst fear-relevant ones. Pictures of fear-relevant animals or angry faces presented among neutral objects are detected more quickly (Öhman et al., 2001).

This “pop-out” effect is thought to take place at a preattentive perceptual stage because the speeded detection of threat is found to be unaffected by the number of distractors. Data from several dot-probe studies have also been interpreted as evidence for enhanced engagement of attention to threat cues (for reviews see Mathews & MacLeod, 2005; Mogg & Bradley, 1998). This attentional bias seems to be selectively enhanced in individuals who demonstrate fear of these specific cues. For example, spider-fearful participants responded faster than non-anxious control participants to dots which appeared at the location of a previously shown spider picture. This was interpreted as an attentional bias for spider stimuli (Mogg & Bradley, 2006). Similar findings by Rinck and Becker (2006) demonstrated that spider-fearful participants’ first fixation was more often on spider pictures in a free-viewing task.

In spite of supporting evidence and considerable theoretical persuasiveness, several other findings question the relevance of initial attentional capture by fear-relevant stimuli. First, the initial attentional bias for fear-relevant animals may be less specific than originally thought. The attentional bias is not limited to fear-relevant animals, but extends to other animals independent of how much they elicit fear (Lipp, 2006; Lipp, Derakshan, Waters, & Logies, 2004; Tipples, Young, Quinlan, Broks, & Ellis, 2002).

Second, a number of well-conducted studies did not find initial attentional capture by threat-relevant pictorial material. Instead, there is evidence that fear-relevant pictures may interfere with observers’ performance. For example, Koster, Crombez, Verschuere, and De Houwer (2004) suggest that effects found in previous dot-probe experiments may not so much be a result of attentional capture but that they may be caused by a difficulty to disengage attention from threat instead. In an adapted dot-probe experiment, Salemink, van den Hout, and Kindt (2007) have recently confirmed that trait anxiety is indeed related to a difficulty to disengage attention from threat rather than to speeded orienting towards threat.

Third, the observation that the perceptual threshold for the detection of threatening stimuli is not lowered is another challenge to the assumption of an effective low-level detection of fear-relevant cues. Phobic patients are not better at detecting phobic stimuli than healthy controls, but they have a lowered criterion for what they label as a phobic stimulus (Becker & Rinck, 2004).

In general, the effects of initial attentional capture or hampered attentional disengagement are difficult to separate—and this separation is especially difficult in search paradigms where intentional and automatic attention are confounded. The faster detection of fear-relevant stimuli among neutral distractors, as described in detail above (Öhman et al., 2001), could also emerge from slowed detection of neutral targets when they are presented among fear-relevant distractors. In two eye-tracking experiments, it has been found that fear-relevant targets interfere with the search for neutral targets (Miltner, Krieschel, Hecht, Trippe, & Weiss, 2004; Rinck, Reinecke, Ellwart, Heuer, & Becker, 2005). This effect could be interpreted as distractibility from fear. Direct evidence for distraction comes from search paradigms where the distractor and not the target was fear-relevant. Here, reaction times were slowed on trials with task-irrelevant but threat-related distractor stimuli in healthy and fearful participants (Lipp, 2006; Lipp & Waters, 2007). These findings may rather reflect delayed disengagement from fear-relevant distractors than enhanced attentional capture. However, measuring only reaction time does not allow to differentiate between these processes.

Recent eye-tracking studies with spider-phobic patients also do not show consistently that attention is captured when a phobic cue has to be detected, but that these cues distract. While failing to replicate the pop-out effect, eye-tracking revealed that phobic participants were slower when a neutral object was to be searched, and a task-irrelevant spider picture distracted them from this task (Miltner et al., 2004). Another eye-tracking study did not find attentional capture by threat cues during search tasks (Rinck et al., 2005). This shows more evidently how fearful participants are distracted by threatening background pictures and are slowed down in their search for neutral targets. However, one needs to keep in mind that in free-viewing tasks healthy controls were repeatedly found to initiate their initial saccades more often towards emotionally arousing pictures of scenes than to neutral scenes (Alpers, in press; Nummenmaa, Hyona, & Calvo, 2006).

Taken together, when fear-relevant stimuli are task-relevant an initial capture of attention cannot consistently be found. Instead, preferential processing may be more apparent when there is a competition for attentional resources (Mathews & Mackintosh, 1998). Eventually, more robust findings for an attentional bias are found when fear-relevant stimuli are presented as task-irrelevant distractors. Under this condition, it seems to be difficult for phobic patients to disengage their attention from these stimuli and this results in observable impairment in task performance. Recording eye movements provides the opportunity to distinguish these different processes. So far, the question remains, whether threatening stimuli distract from ongoing task execution due to enhanced engagement or slowed down disengagement of attention or whether both accompany the processing of threatening stimuli (for a further discussion see Rinck & Becker, 2006; Williams, Watts, MacLeod, & Mathews, 1997). Despite considerable evidence showing that phobic material distracts, the specific underlying processes behind slowed reaction times have not been examined in much details.

The purpose of our experiment was to investigate how phobia-related but task-irrelevant stimuli interfere with ongoing search tasks. In the majority of previous studies, the stimuli served as targets on some trials and distractors on others within the same experiment and a semantic stimulus processing was therefore always necessary. For our experiment, it was particularly important that the spider pictures as well as the neutral pictures never had to be responded throughout our experiment. While participants searched for a color singleton and identified a target letter inside of it, task-irrelevant phobic or neutral stimuli were presented in the periphery. Thus, our task did not require any semantic processing of the distractors. If attentional capture can be demonstrated under these conditions, this would be convincing evidence that threatening distractors capture attention independent of the necessity to differentiate between the pictures. In addition to measuring reaction times, we registered eye movements in order to dismantle the processes involved in the expected distractor effect. Our experimental procedure based on the finding, that abrupt-onset distractors can sometimes not be ignored during controlled saccades to a target: Distractors are fixated upon and reaction times to targets slowed down (Godjin & Theeuwes, 2003; Theeuwes, Kramer, Hahn, Irwin, & Zelinsky, 1999). Previous eye-tracking search tasks demonstrated a methodological vagueness in only reporting overall gaze durations on target regions but not locations and durations of discrete fixations (see Miltner et al., 2004; Rinck et al., 2005). To clearly interpret eye movement data in terms of attentional processes we focused on the characteristics of the initial fixation on distractors. In sum, we aimed to further examine the effects of task-irrelevant phobic distractors and to separate, whether differences in reaction times to neutral targets result from attentional capture or by slower disengagement when phobic cues are erroneously attended to.

Section snippets

Participants

Twenty-five spider-phobic patients and 25 non-anxious control participants were recruited by advertisements in a local newspaper. Potential candidates were invited when they passed a phone-screening using the German Spider Anxiety Screening (Rinck et al., 2002) with a score above 18 for the spider-phobic patient group (spider-fearful groups had a range of 18–24 in normative samples) or below 5 for the non-anxious control group (non-anxious participants ranged from 0 to 3 in normative control

Stimulus ratings

For the valence ratings of spider, flower, and mushroom pictures, there were main effects of picture category, F(2, 80)=32.36, p<.001, η2=.45, and group, F(1, 40)=33.52, p<.001, η2=.46, as well as a significant Picture Category×Group interaction, F(2, 80)=18.56, p<.001, η2=.32. Spider-phobic patients rated the spider pictures (M=−2.95, SD=1.36) significantly more negative than the non-anxious control group (M=−.10, SD=.625), t(28.09)=8.75, p<.001, d=2.69, but there were no reliable group

Discussion

The present study shows that phobic patients are distracted from ongoing tasks by abruptly appearing task-irrelevant phobic cues. We compared reaction times and eye movements of diagnosed spider-phobic patients with non-anxious control participants. The purpose was to highlight possible differences in the initial and subsequent allocation of attention in the presence of phobic distractor pictures presented in the visual periphery. Phobic patients were distracted from an easy search task when

Acknowledgments

The pictures used in this study were graciously provided by Wolfgang Miltner, Jena. We would like to thank Anne Springmann for her help with data collection and data analysis. We acknowledge the support by the Research Group Emotion and Behavior which is sponsored by the German Research Society (DFG).

References (51)

  • G.W. Alpers et al.

    Startle and automatic nervous system modulation while viewing emotional scenes or emotional facial expressions

    Psychophysiology

    (2006)
  • G.W. Alpers et al.

    Here is looking at you: Emotional faces predominate in binocular rivalry

    Emotion

    (2007)
  • Alpers, G. W., Gerdes, A. B. M., Lagarie, B., Tabbert, K., Vaitl, D., & Stark, R. (submitted). Attention modulates...
  • S. Anders et al.

    Parietal somatosensory association cortex mediates affective blindsight

    Nature Neuroscience

    (2004)
  • APA

    Diagnostic and statistical manual of mental disorders

    (1994)
  • E.S. Becker et al.

    Sensitivity and response bias in fear of spiders

    Cognition and Emotion

    (2004)
  • P.J. Bex et al.

    Spatial frequency, phase, and the contrast of natural images

    Journal of the Optical Society of America, A, Optics, Image Science, & Vision

    (2002)
  • K. Carlsson et al.

    Fear and the amygdala: Manipulation of awareness generates differential cerebral responses to phobic and fear-relevant (but nonfeared) stimuli

    Emotion

    (2004)
  • M. Davis et al.

    The amygdala: Vigilance and emotion

    Molecular Psychiatry

    (2001)
  • M.W. Eysenck

    Anxiety: The cognitive perspective

    (1992)
  • M.W. Eysenck et al.

    Anxiety and cognitive performance: Attentional control theory

    Emotion

    (2007)
  • M.S. Fanselow

    Neural organization of the defensive behavior system responsible for fear

    Psychonomic Bulletin and Review

    (1994)
  • M.S. Fanselow et al.

    A functional behavioristic approach to aversively motivated behavior: Predatory imminence as a determinant of the topography of defensive behavior

  • E. Fox et al.

    Do threatening stimuli draw or hold visual attention in subclinical anxiety?

    Journal of Experimental Psychology: General

    (2001)
  • E. Fox et al.

    Attentional bias for threat: Evidence for delayed disengagement from emotional faces

    Cognition and Emotion

    (2002)
  • Cited by (87)

    • Interpreting and responding to ambiguous natural images in spider phobia

      2019, Journal of Behavior Therapy and Experimental Psychiatry
    View all citing articles on Scopus
    View full text