Common and Distinct Neuroanatomical Abnormalities in Adult ASD and Schizophrenia

Background: Autism spectrum disorder (ASD) and schizophrenia are characterized by significant impairments in social and non-social information processing, which may reflect similar underlying brain abnormalities, yet few studies have directly examined convergence in the neuroanatomical basis of these conditions.

Objectives: The purpose of this study was to use high-resolution structural magnetic resonance imaging to identify and compare regional gray matter morphological abnormalities between adults ASD and schizophrenia relative to healthy volunteers.

Methods: A cross-sectional, cross-diagnostic case-control structural neuroimaging study was conducted with 14 verbal adults with ASD, 42 stabilized adult schizophrenia outpatients, and 20 healthy adult volunteers.  Regional measurements of gray matter volume were collected using high-resolution structural magnetic resonance imaging using a 3T Siemens Trio whole-body scanner and head coil.  Broad fronto-temporal regions of interest were defined using Wake Forest University Pickatlas including relevant areas of the prefrontal cortex, medial-temporal lobe, and superior and middle temporal gyri based on previous studies of ASD and schizophrenia.  All data were preprocessed and analyzed using Statistical Parametric Mapping, version 12.

Results: Adults with ASD demonstrated significant and broadly increased gray matter in the bilateral middle temporal gyrus (all k > 100, all p < .001) and insular cortex (all k > 158, all p < .001) relative to healthy volunteers.  In addition, a significant cluster of increased gray matter in the left medial prefrontal cortex was also present in adults with ASD (k = 162, p < .001) compared to healthy individuals.  Individuals with schizophrenia demonstrated similarly increased gray matter volume in the left middle temporal gyrus (k = 66, p = .001), although this finding was less pronounced than in ASD.  In contrast to adults with ASD, patients with schizophrenia also showed significantly reduced amygdala volume bilaterally (all k > 20, all p < .003).  When comparing ASD and schizophrenia directly, individuals with ASD had significantly greater gray matter volume in a bilaterial medial-temporal cluster encompassing the amygdala and parahippocampal gyrus (all k > 130, all p < .001).  Increased right putamen volume was also observed in ASD relative to patients with schizophrenia (k = 49, p = .001).

Conclusions: Autism and schizophrenia are associated with neuroanatomical abnormalities in medial-temporal regions that are only partially overlapping along the middle temporal gyrus implicated in social information processing.  Overall, ASD was characterized by widespread increases in cortical gray matter not observed in schizophrenia, which was primarily associated with reduced amygdalar volume.  These findings support models of distinct neurodevelopmental processes converging on shared social disability in people with ASD and schizophrenia.