Hierarchical Drift Diffusion Modeling of Self-Referential Processing in Social Anxiety and Body Dysmorphic Disorder

Abstract

Body dysmorphic disorder (BDD) and social anxiety disorder (SAD) share hallmark features of heightened self-consciousness and maladaptive self-referential processing (SRP), yet the cognitive mechanisms underlying SRP in these disorders remain poorly understood. This study employed hierarchical drift diffusion modeling (HDDM) to examine SRP during a self-referential encoding task (SRET) among individuals with BDD, SAD, and healthy controls (HC). Drift rate, a model parameter indexing the efficiency of evidence accumulation, served as a proxy for SRP, with lower drift rates indicating greater decisional inefficiency in processing self-relevant information. A total of 58 participants (15 BDD, 15 SAD, 28 HC) judged positive and negative trait adjectives in self- and other-referential contexts. Both clinical groups endorsed fewer positive and more negative self-descriptors than HC, with no group differences in other-referential trials. Critically, both BDD and SAD showed reduced drift rates relative to HC in both self- and other-referential trials regardless of valence, although deficits were substantially larger for self-referential decisions. Within-group analyses showed marked self–other drift rate reductions in clinical groups. Across patients, disorder‑specific symptom severity predicted lower drift rates in self-referential trials even after controlling for depression. No drift rate differences emerged between BDD and SAD groups, suggesting a shared cognitive process. Together, these findings integrate computational modeling with experimental task data to reveal that BDD and SAD involve a generalized inefficiency in evaluating social information, most pronounced for self-relevant content, highlighting drift rate as a potential transdiagnostic marker of cognitive impairment.