I can’t take my eyes off of your face : the impact of delayed attentional disengagement from threat in the maintenance of social anxiety

Heeren, Alexandre [UCL]

Attention bias modification is a recent procedure that allows examining the causal involvement of attentional bias for threat on the maintenance of anxiety. We examined which attentional bias’s features are involved in this phenomenon and how top-down processing modulates it. We have obtained converging evidence that the critical feature of attentional bias for threat in the maintenance of anxiety is the phenomenon of delayed disengagement from threat. Consistent with Cisler and Koster (2010), we have evidenced that difficulty in disengaging from threat may occur independently of facilitated attention because the modification of the former does not impact on the latter one. At a theoretical level, our findings provide support to the hypothesis that attentional bias for threat may be causally involved in the maintenance of the disorder (i.e., Eysenck et al., 2007; Mogg & Bradley, 1998; Williams et al., 1997), and decipher the nature of the attentional bias’s feature involved in this phenomenon. We also provided evidence that the maintenance of anxiety is not the result of a vigilance-avoidance pattern of attention (Pflugshaupt et al., 2005; Mogg & Bradley, 2006). According to that view, early vigilance may serve to heighten anxiety whereas the later avoidance maintains anxiety. Our data are not consistent with this view and indicate that the delayed attentional disengagement from threat appears as a relevant candidate for the maintenance of anxiety. Similarly, our data does not converge with the hypothesis of Foa and Kozak (1986) that the inhibition of detailed processing of threatening stimuli is the core deficit of anxiety, which is reflected in avoidance of threatening stimuli and that fear reduction will be hampered when attention is directed away from the feared stimulus (e.g., Foa, Huppert, & Cahill, 2006). We proposed that the phenomenological difference between fear and anxiety can partly explain the paradox. According to some models, attentional bias toward threat could be overridden by strategic effort (e.g., Eysenck, 2007; Matthews & Mackintosh, 1998). More predominantly, some theoretical models suggest that the difficulty in disengagement may occur because of generally poor regulatory top-down attention control (i.e., Cisler & Koster, 2010; Eysenck et al., 2007). One may thus expect that change in attention training may be mediated by change in top-down attention control. We have also drawn attention on the hypothesis that some potential changes in bottom-up processing might mediate the effects of top-down processing on delayed disengagement from threat. Indeed, as predicted by Cisler and Koster’s model (2010), difficulty in disengaging from threat may be the result of a too strong influence of bottom-up processing (i.e., threat detection mechanism, amygdala) on attention control (i.e., prefrontal regulatory process and functionally-related structures). Finally, we provide evidence to the prediction of Bar-Haim et al., (2007) that anxiety disorders may stem from attentional bias for threat. We also pointed out that the development of attentional bias for threat leads to a specific difficulty in disengaging attention from threat, which in turn, lead to increase emotional reactivity.