Abstract

Background

Social Anxiety Disorder (SAD) is a prevalent and impairing condition that is often underdiagnosed and undertreated, leading to chronic symptoms and comorbidities. Identifying objective behavioral markers is critical for improving early detection. Aberrant visual attention such as gaze avoidance and hyperscanning have been proposed as key mechanisms underlying SAD, yet findings from traditional eye-tracking studies using static or limited social stimuli remain inconsistent. Virtual Reality (VR) combined with integrated eye-tracking offers an ecologically valid approach to studying visual attention within realistic, socially evaluative contexts.

Objective

This study examined visual attention patterns in individuals with SAD using an immersive VR paradigm with integrated eye-tracking designed to simulate naturalistic socio-evaluative situations.

Methods

Thirty individuals with SAD and thirty matched healthy controls (HCG) completed three 360° VR tasks: a job interview, an unstructured social interaction, and a presentation. Eye-tracking data were analyzed across predefined areas of interest (AOI) using temporal and spatial metrics.

Results

Individuals with SAD showed reduced fixation duration on faces and increased fixation duration on the environment compared with HCG. In the job interview task, significant group differences were found for the face AOI (p = .048) and environment AOI (p = .022). In the unstructured social interaction task, individuals with SAD again spent less time fixating on faces (p = .006) and more on the environment (p < .001). No group differences were observed in the presentation task. Individuals with SAD also demonstrated greater mean distance between fixations, indicative of hyperscanning across all tasks.

Conclusions

VR-based eye-tracking can capture distinct visual attention patterns in SAD, supporting its ecological validity and diagnostic potential as an objective tool for early detection and digital mental-health assessment method.