Differential brain responses to affective sounds in misophonia and hyperacusis: A task-based fMRI approach

Abstract Humans are naturally attuned to emotionally salient sounds, such as screams signaling danger, which trigger survival-related responses. In sound-sensitivity disorders, such as misophonia and hyperacusis, everyday sounds provoke intense emotional and behavioral reactions. Misophonia is typically triggered by specific sounds, such as chewing, whereas hyperacusis involves hypersensitivity to sounds above certain intensity thresholds. Because these disorders share overlapping symptoms and frequently co-occur, disentangling their neural bases is essential for improving diagnosis and treatments. We recruited 91 young adults categorized into four groups: misophonia, hyperacusis, comorbid misophonia and hyperacusis, and controls. We conducted task-based fMRI, where participants listened to 90 emotionally valenced sounds from the International Affective Digitized Sounds-2 database and rated their valence during scanning. Whole-brain functional activation and seed-to-voxel functional connectivity analyses revealed both distinct and overlapping pattern of neural correlates associated with these disorders. The misophonia group, regardless of comorbid hyperacusis (relative to controls), showed hyperactivation in visual association areas and reduced connectivity between salience and visual networks during unpleasant versus neutral sound processing. This suggests atypical cross-modal sensory involvement. The hyperacusis group exhibited reduced connectivity between salience hubs and frontal control regions compared to misophonia and controls, indicating impaired top-down regulation. In contrast, this connectivity was preserved in misophonia for generally unpleasant sounds, suggesting intact regulation. The comorbid group showed neural patterns associated with both disorders. Overall, these findings reveal overlapping and disorder-specific neural patterns across sound tolerance profiles. Future research should combine neural and behavioral data to refine mechanistic models and guide targeted interventions.