Neural Patterns for Empathic Versus Subjective Fear in Classifying Generalized Anxiety Disorder and Major Depressive...

Fear processing is multifaceted, involving dissociable neural and computational pathways that vary depending on its source and context. However, modality-specific fear processing, such as empathic versus subjective fear, remains poorly understood. Moreover, the role of these distinctions in clarifying the high comorbidity between generalized anxiety disorder (GAD) and major depressive disorder (MDD) has been understudied.

In the current study, we combined functional magnetic resonance imaging with multivariate pattern analysis to develop two neurofunctional models: one for empathic fear (face-induced empathic fear signature [FEFS], n = 81) and another for subjective fear (scene-induced subjective fear signature [SSFS], n = 81) evoked by visual stimuli. After validating these models in the independent cohorts (dynamic stimuli: n = 28), we generalized them to distinguish GAD and MDD during facial emotional-processing (n = 80) and pain empathy (n = 87) tasks.

Our findings revealed that both models engaged distributed brain systems, including cortical regions (e.g., prefrontal, cingulate, insula, and parietal cortices) and subcortical areas (e.g., thalamus and amygdala). Crucially, we identified distinct functional profiles: FEFS-predominant regions were linked to execution, whereas SSFS-predominant regions mapped onto emotion processing. Notably, compared with SSFS, FEFS consistently discriminated between GAD and MDD during both fearful expression and affective pain empathy processing.

These findings demonstrate modality-specific processing of subjective and empathic fear, revealing distinct empathy-related neural signatures that may serve as potential biomarkers for differentiating GAD and MDD, offering new insights into their neurobiological distinctions.