Optimizing differential diagnostics and identifying transdiagnostic treatment targets using virtual reality

Abstract

Accurate diagnosis in psychiatry remains a significant challenge, often delaying appropriate treatment and resulting in poorer clinical outcomes. Identifying precise biomarkers for differential diagnosis is therefore crucial. This study aimed to identify distinct behavioral and psychophysiological markers of emotional reactivity in virtual reality (VR) settings among individuals with bipolar disorder (BD), borderline personality disorder (BPD), schizophrenia spectrum disorders (SSD), and healthy controls (HC). Participants (BD: n = 32, BPD: n = 21, SSD: n = 17, HC: n = 30) aged 19–60 were exposed to six immersive 360-degree social VR scenarios, ranging from neutral to highly emotional contexts (e.g., an elevator ride, a crying baby). Emotional responses were self-rated on a 1–5 scale, while galvanic skin response (GSR) was continuously recorded. Scenarios assessed feelings of unpleasantness, pleasantness, being observed, and the urge to comfort. Across diagnoses, individuals with mental health conditions reported more negative emotional responses (greater unpleasantness) across both neutral and negative scenarios (ps ≤ 0.02) despite similar GSR levels to HC. Specifically, in the elevator scenario, BPD and SSD experienced greater unpleasantness and feelings of being observed, coupled with stronger GSRs compared to BD (ps ≤ 0.03). SSD reported higher unpleasantness in the canteen scenario, less pleasantness in the happy baby scenario, and overall higher GSR than BD (ps ≤ 0.049). Negative emotional reactivity was consistent across BD, BPD, and SSD, with heightened emotional and physiological responses distinguishing SSD and BPD from BD in specific VR contexts. VR-based assessments of emotional and physiological markers show promise for improving differential diagnosis and identifying transdiagnostic treatment targets.