Distinct neurostructural signatures of anxiety-, fear-related and depressive disorders: a comparative voxel-based meta-analysis

Internalizing disorders encompass anxiety, fear and depressive disorders. While the DSM-5 nosology conceptualizes anxiety and fear-related disorders as an entity, dimensional psychopathology models suggest that generalized anxiety disorders (GAD) and major depression originate from an overarching "anxious-misery" factor whereas fear-related disorders originate from the "fear" factor. Given that a neurobiological evaluation is lacking, we conducted a comparative neuroimaging meta-analysis of gray matter volume alterations to determine common and disorder-specific brain structural signatures in these disorders. The PubMed, Web of Knowledge, and Scopus databases were searched for case-control voxel-based morphometric studies through December, 2020 in GAD, fear-related anxiety disorders (FAD, i.e., social anxiety disorders, SAD; specific phobias, SP; panic disorders, PD; and agoraphobia, AG) and major depressive disorder (MDD). Neurostructural abnormalities were assessed within each disorder group followed by quantitative comparison and conjunction analyses using Seed-based d-Mapping (SDM-PSI). GAD (9 studies, 226 patients) showed disorder-specific decreased volumes in left insula (z=-2.98, pFWE-corrected<0.05) and lateral/medial prefrontal cortex (z=-2.10, pFWE-corrected<0.05,) as well as increased right putamen volume (z=1.86, pFWE-corrected<0.05) relative to FAD (10 SAD, 11 PD, 2 SP studies, 918 patients). Both GAD and MDD (46 studies, 2,575 patients) exhibited decreased prefrontal volumes compared to controls and FAD. While FAD showed less robust alterations in lingual gyrus (p < 0.0025, uncorrected), this group presented intact frontal integrity. No shared structural abnormalities were found. Unique clinical features characterizing anxiety-, fear-related and depressive disorders are reflected by disorder-specific neuroanatomical abnormalities. Targeting the disorder-specific neurostructural signatures could improve therapeutic efficacy.