Attention allocation to subliminally presented affective faces in high and low social anxiety

Abstract

Attention bias to affective stimuli, particularly threatening stimuli, has been well documented, with attention bias to affective faces often reported in populations with social anxiety (SA). However, inconsistencies exist in the literature regarding the direction of the bias, with some studies reporting a bias to attend toward social threat, and others reporting a bias to avoid social threat. This variability in the attention bias literature could be related to fluctuations in how attentional resources are allocated toward social stimuli over time. This study aimed to isolate early processing of affective faces using a backward masking paradigm in participants with high and low levels of subclinical SA. We used prepulse inhibition (PPI) of startle to index the amount of early attention allocated to masked affective faces. There was greater PPI to masked angry faces compared to neutral faces across all participants, suggesting that more attention was allocated to threatening faces. This effect was similar across face genders and participant SA levels. There was also a trend for more attention to be allocated to masked affective faces (angry and happy) compared to neutral faces across all participants. These findings demonstrate that attention bias to subliminal affect, and threat specifically, may not be modulated by subclinical levels of SA at very early stages of attention processing.

Keywords: Anxiety; Attention; Face processing; Startle blink.