دوره زمانی توجه برای چهره عاطفی در هراس اجتماعی

This study investigated the time-course of attentional bias in socially phobic (SP) and non-phobic (NP) adults. Participants viewed angry and happy faces paired with neutral faces (i.e., face-face pairs) and angry, happy and neutral faces paired with household objects (i.e., face-object pairs) for 5000 ms. Eye movement (EM) was measured throughout to assess biases in early and sustained attention. Attentional bias occurred only for face-face pairs. SP adults were vigilant for angry faces relative to neutral faces in the first 500 ms of the 5000 ms exposure, relative to NP adults. SP adults were also vigilant for happy faces over 500 ms, although there were no group-based differences in attention to happy-neutral face pairs. There were no group differences in attention to faces throughout the remainder of the exposure. Results suggest that social phobia is characterised by early vigilance for social cues with no bias in subsequent processing.

Research has increasingly emphasized the role of emotion in regulating attention to fear-relevant stimuli (Öhman, Flykt, & Lundqvist, 2000). In the context of social anxiety, it is generally assumed that socially anxious individuals are biased toward selective processing of social threat cues, known as attentional bias ( Bögels & Mansell, 2004). However studies have also shown that socially anxious individuals may avoid processing social threat ( Chen et al., 2002, Mansell et al., 1999 and Rinck and Becker, 2005). Findings of both vigilance and avoidance are prominent in social anxiety (for review see Heinrichs & Hofmann, 2001) and have prompted researchers to develop more sophisticated models of attentional bias that allow for variations in the direction of attentional allocation. Several authors have proposed that anxiety is characterised by a vigilant-avoidant pattern of attentional allocation, in which anxious individuals initially orientate toward threat but then shift attention away in order to alleviate anxious mood ( Mogg et al., 1997 and Mogg et al., 1987). This two-stage model assumes that the direction of attentional bias might vary throughout the information processing stream, and as a result, has prompted research into the time-course of attentional bias. The main goal of this research has been to determine whether attentional bias occurs in both the early and late stages of attentional processing, and whether the direction of attention bias varies throughout these stages. To date, however little is known about the time-course of attentional bias in social phobia. Traditional models of social phobia propose that attentional bias is important in maintaining the disorder, however they make competing predictions about the time-course of attentional bias. For example, Rapee and Heimberg's (1997) model of social phobia proposes that socially phobic individuals detect social threat cues rapidly and have difficulty disengaging attention from them. Once external threat is detected, the individual also begins monitoring an internal representation of themselves as they appear to others. Overall, the model predicts early and sustained vigilance for threat; however it remains possible that splitting attention toward internal cues may cause the outward-directed attentional bias to diminish over time. In contrast, Clark and Wells' (1995) model of social phobia proposes that socially phobic individuals avoid social threat, however it is unclear how threat is first detected. If avoidance follows initial overt orientation toward threat, the model would predict a vigilant-avoidant pattern of attention. Few studies have examined the time-course of attentional bias in social anxiety, in part due to limitations in existing technologies. Attentional bias is generally measured using dot probe tasks in which pairs of social threat and non-threat stimuli are presented for 500 ms and replaced by a probe. Vigilance for threat is indicated by faster detection of probes replacing threat stimuli compared with neutral stimuli. This task provides only a static snap-shot of attention and is ill equipped to study the time-course of attentional bias. One approach to assessing the time-course of attentional bias has been to study the eye movements (EMs) of anxious individuals while viewing threatening pictures. EM provides a dynamic, ‘on-line’ registration of attention, enabling researchers to examine the time-course of selective attention over much longer stimulus durations. However, EM studies with socially anxious individuals have also produced equivocal results. Horley, Williams, Gonsalvez, and Gordon (2003) monitored the EMs of socially phobic and non-phobic individuals to single photographs of sad, happy and neutral faces for 10 s. Social phobics spent less time viewing informative facial regions, particularly the eyes, but showed extensive scanning of non-feature areas of the face. Horley et al. replicated these results with the addition of angry faces (Horley Williams, Gonsalvez, & Gordon, 2004) and found that hyperscanning with eye avoidance was particularly pronounced in angry faces, suggesting elements of both vigilance and avoidance of social threat. However, it is difficult to extrapolate these results to draw conclusions about the time-course of attentional bias as EM data were averaged across the entire 10 s viewing exposure. Consequently, although socially phobic individuals avoided viewing facial features overall, it is possible that they initially looked toward facial features but then averted their gaze to non-feature regions of the face. In addition, given that individuals viewed only one facial expression at a time, results do not reflect selective processing of affective vs. neutral social cues. More recently, Garner, Mogg, and Bradley (2006) assessed the first EM of high and low socially anxious individuals to face pairs. When participants were threatened with having to give a speech, high socially anxious individuals were faster to initiate a first EM to emotional (happy and angry) faces relative to neutral faces. However, this fixation was brief, with socially anxious individuals tending to fixate on emotional faces for less time than low anxious individuals. The authors concluded that under conditions of social-threat, high socially anxious adults rapidly orientate toward emotional faces, but then quickly disengage their attention. The authors further suggested that a threat-specific bias might be more apparent in those with clinical levels of social anxiety. To that end, it is important to replicate this work in those diagnosed with social phobia. The current study aimed to extend previous research in several ways. We examined the time-course of attentional bias in individuals diagnosed with social phobia and healthy controls using EM as a continuous index of selective attention. Previous dot probe and EM studies have produced reliable evidence for a bias in early attentional processing (Heinrichs & Hofmann, 2001) therefore we predicted that socially phobic adults would show an attentional bias during the first 500 ms of stimulus viewing. Studies have also shown, however, that the precise direction of early attentional bias might vary depending upon the context in which social threat occurs. Clark and McManus (2002) argued that if socially anxious individuals have the opportunity to avoid social cues (e.g., faces), they will do so and will preferentially attend instead to external non-social cues, such as objects (e.g., Chen et al., 2002 and Mansell et al., 1999). However, if presented with two faces, socially anxious individuals will preferentially attend to the more negative face (e.g., Mogg & Bradley, 2002). To control for this, EM was monitored to stimuli pairs containing: (a) happy and angry faces matched with neutral faces (i.e., face-face pairs); and (b) happy, angry and neutral faces matched with household objects (i.e., face-object pairs). Following Clark and McManus's theory, we predicted that socially anxious adults would show an early bias toward negative faces on face-face trials and would avoid faces in favour of objects on face-object trials. This is the first study to examine EMs to face pairs in socially phobic adults over several seconds. Studies to date have produced mixed support for the existence of a bias in sustained processing. Studies examining scanning patterns over long periods (Horley et al., 2003 and Horley et al., 2004) have used singular faces (not stimulus pairs) and have found elements of both vigilance and avoidance. Two studies using face pairs (Garner et al., 2006 and Mogg et al., 2004) failed to demonstrate attentional bias in sustained processing, however they used much shorter time frames. Based on the two dominant models of social phobia we predicted that social phobia might be characterised by either sustained vigilance for social cues (Rapee & Heimberg, 1997) or late-stage avoidance (Clark & Wells, 1995).