Are schizophrenia, autistic, and obsessive spectrum disorders dissociable on the basis of neuroimaging morphological findings?: A voxel‐based meta‐analysis

Abstract

Schizophrenia spectrum disorder (SCZD), autism spectrum disorder (ASD), and obsessive-compulsive spectrum disorder (OCSD) are considered as three separate psychiatric conditions with, supposedly, different brain alterations patterns. From a neuroimaging perspective, this meta-analytic study aimed to address whether this nosographical differentiation is actually supported by different brain patterns of gray matter (GM) or white matter (WM) morphological alterations. We explored two possibilities: (a) to find out whether GM alterations are specific for SCZD, ASD, and OCSD; and (b) to associate the identified brain alteration patterns with cognitive dysfunctions by means of an analysis of lesion decoding. Our analysis reveals that these psychiatric spectra do not present clear distinctive patterns of alterations; rather, they all tend to be distributed in two alteration clusters. Cluster 1, which is more specific for SCZD, includes the anterior insular, anterior cingulate cortex, ventromedial prefrontal cortex, and frontopolar areas, which are parts of the cognitive control system. Cluster 2, which is more specific for OCSD, presents occipital, temporal, and parietal alteration patterns with the involvement of sensorimotor, premotor, visual, and lingual areas, thus forming a network that is more associated with the auditory-visual, auditory, premotor visual somatic functions. In turn, ASD appears to be uniformly distributed in the two clusters. The three spectra share a significant set of alterations. Our new approach promises to provide insight into the understanding of psychiatric conditions under the aspect of a common neurobiological substrate, possibly related to neuroinflammation during brain development. Autism Res 2017. © 2017 International Society for Autism Research, Wiley Periodicals, Inc. Autism Res 2017, 10: 1079-1095. © 2017 International Society for Autism Research, Wiley Periodicals, Inc.

The morphometric co‐atrophy networking of schizophrenia, autistic and obsessive spectrum disorders

“…This result suggests that the normal anatomical covariance of insular and cingulate areas tends to be progressively altered toward the development of a pathological anatomical covariance (co‐atrophy). Our finding is in line with the frequent observation that the patterns of brain co‐alterations match in part the patterns of brain connectivity (Cauda, et al, ; Cauda, et al, ; Crossley, et al, ; Crossley, et al, ; Evans, ; Fornito, et al, ; Menon, ; Raj, et al, ; Saxena & Caroni, ; Seeley, et al, ; Yates, ; Zhou, et al, ). In contrast, when different edges are involved, as it is the case of the other altered or co‐atrophic areas of the MCN, especially those with a lower network degree, the normal pattern of anatomical covariance does not overlap with the alteration pattern.…”

“…Of note, the first diagnostic criterion of DSM‐5 for ASD is strictly related to the negative symptoms of SCZD, while the second diagnostic criterion is similar to the OCSD symptomatology. The relative symptomatic similarity between ASD and SCZD is consistent with a neurobiological model that suggests a common basis for SCZD and ASD, with a number of genetic alterations (SHANK 3 variations, DISC 1, dysregulation of CYFIP1, SCN2A, NRXN1 neurexin gene or RELN), cytoarchitectural abnormalities (about proliferation, migration and lamination defects), neuropsychological deficit, neuroimaging investigations (about GM/WM abnormalities and structural/functional connectivity alterations), and clinical observations (Baribeau & Anagnostou, ; Cauda, et al, ; Cauda, et al, ; Cauda, et al, ; Cheung, et al, ; Chisholm, Lin, Abu‐Akel, & Wood, ; de Lacy & King, ; King & Lord, ; Pathania, et al, ).…”

“…As Figure illustrates, when we leave SCZD out, the number of edges substantially diminishes; this, along with the fact that ASD and OCSD have more inhomogeneous sample data, as shown in Cauda et al (), explains why the correlation value between the matrix constructed with the SCZD data only and the matrix constructed with the whole dataset is low. Moreover, it is worth noting that, even though results are significantly reduced when SCZD is left out, the remaining edges are those that form the most connected part of the MCN, which is also the core of the pattern constructed with the SCZD data only (please see both the upper right panel of Figure and the upper panel of Supporting Information, Figure S5).…”

“…Often brain areas in which alterations co‐occur also exhibit patterns of anatomical covariance (Evans, ). Furthermore, converging evidence suggests that pathological alterations caused by brain disorders are likely to follow patterns of distribution that strongly have a network‐like architecture, which depends on the organization of both the structural and functional connectivity (Cauda, et al, ; Cauda, et al, ; Crossley, et al, ; Crossley, et al, ; Fornito, Zalesky, & Breakspear, ; Menon, ; Raj, Kuceyeski, & Weiner, ; Saxena & Caroni, ; Seeley, Crawford, Zhou, Miller, & Greicius, ; Yates, ; Zhou, Gennatas, Kramer, Miller, & Seeley, ).…”

Autism Spectrum Disorder Diagnosis Using Sparse Graph Embedding of Morphological Brain Networks

“…Despite this growing body of research on such networks, however, there is still a gap in the literature where morphological networks have not been explored to the same degree. This gap needs to be filled, considering there are studies that indicate morphological features of the brain, such as cortical thickness, can be affected in neurological disorders, including ASD [6,7]. As such, the use of networks based on morphological data in neurological disorder diagnosis, using machine learning, could prove fruitful.…”