Subcortical structures, which include the basal ganglia and parts of the limbic system, have key roles in learning, motor control and emotion, but also contribute to higher-order executive functions. Prior studies have reported volumetric alterations in subcortical regions in schizophrenia. Reported results have sometimes been heterogeneous, and few large-scale investigations have been conducted. Moreover, few large-scale studies have assessed asymmetries of subcortical volumes in schizophrenia. Here, as a work completely independent of a study performed by the ENIGMA consortium, we conducted a large-scale multisite study of subcortical volumetric differences between patients with schizophrenia and controls. We also explored the laterality of subcortical regions to identify characteristic similarities and differences between them. T1-weighted images from 1680 healthy individuals and 884 patients with schizophrenia, obtained with 15 imaging protocols at 11 sites, were processed with FreeSurfer. Group differences were calculated for each protocol and meta-analyzed. Compared with controls, patients with schizophrenia demonstrated smaller bilateral hippocampus, amygdala, thalamus and accumbens volumes as well as intracranial volume, but larger bilateral caudate, putamen, pallidum and lateral ventricle volumes. We replicated the rank order of effect sizes for subcortical volumetric changes in schizophrenia reported by the ENIGMA consortium. Further, we revealed leftward asymmetry for thalamus, lateral ventricle, caudate and putamen volumes, and rightward asymmetry for amygdala and hippocampal volumes in both controls and patients with schizophrenia. Also, we demonstrated a schizophrenia-specific leftward asymmetry for pallidum volume. These findings suggest the possibility of aberrant laterality in neural pathways and connectivity patterns related to the pallidum in schizophrenia.

When collecting large amounts of neuroimaging data associated with psychiatric disorders, images must be acquired from multiple sites because of the limited capacity of a single site. However, site differences represent a barrier when acquiring multisite neuroimaging data. We utilized a traveling-subject dataset in conjunction with a multisite, multidisorder dataset to demonstrate that site differences are composed of biological sampling bias and engineering measurement bias. The effects on resting-state functional MRI connectivity based on pairwise correlations because of both bias types were greater than or equal to psychiatric disorder differences. Furthermore, our findings indicated that each site can sample only from a subpopulation of participants. This result suggests that it is essential to collect large amounts of neuroimaging data from as many sites as possible to appropriately estimate the distribution of the grand population. Finally, we developed a novel harmonization method that removed only the measurement bias by using a traveling-subject dataset and achieved the reduction of the measurement bias by 29% and improvement of the signal-to-noise ratios by 40%. Our results provide fundamental knowledge regarding site effects, which is important for future research using multisite, multidisorder resting-state functional MRI data.

Identifying both the commonalities and differences in brain structures among psychiatric disorders is important for understanding the pathophysiology. Recently, the ENIGMA-Schizophrenia DTI Working Group performed a large-scale meta-analysis and reported widespread white matter microstructural alterations in schizophrenia; however, no similar crossdisorder study has been carried out to date. Here, we conducted mega-analyses comparing white matter microstructural differences between healthy comparison subjects (HCS; N = 1506) and patients with schizophrenia (N = 696), bipolar disorder (N = 211), autism spectrum disorder (N = 126), or major depressive disorder (N = 398; total N = 2937 from 12 sites). In comparison with HCS, we found that schizophrenia, bipolar disorder, and autism spectrum disorder share similar white matter microstructural differences in the body of the corpus callosum; schizophrenia and bipolar disorder featured comparable changes in the limbic system, such as the fornix and cingulum. By comparison, alterations in tracts connecting neocortical areas, such as the uncinate fasciculus, were observed only in schizophrenia. No significant difference was found in major depressive disorder. In a direct comparison between schizophrenia and bipolar disorder, there were no significant differences. Significant differences between schizophrenia/bipolar disorder and major depressive disorder were found in the limbic system, which were similar to the differences in schizophrenia and bipolar disorder relative to HCS. While schizophrenia and bipolar disorder may have similar pathological characteristics, the biological characteristics of major depressive disorder may be close to those of HCS. Our findings provide insights into nosology and encourage further investigations of shared and unique pathophysiology of psychiatric disorders.

Compelling evidence in Caucasian populations suggests a role for copy-number variations (CNVs) in autism spectrum disorder (ASD) and schizophrenia (SCZ). We analyzed 1,108 ASD cases, 2,458 SCZ cases, and 2,095 controls in a Japanese population and confirmed an increased burden of rare exonic CNVs in both disorders. Clinically significant (or pathogenic) CNVs, including those at 29 loci common to both disorders, were found in about 8% of ASD and SCZ cases, which was significantly higher than in controls. Phenotypic analysis revealed an association between clinically significant CNVs and intellectual disability. Gene set analysis showed significant overlap of biological pathways in both disorders including oxidative stress response, lipid metabolism/modification, and genomic integrity. Finally, based on bioinformatics analysis, we identified multiple disease-relevant genes in eight well-known ASD/SCZ-associated CNV loci (e.g., 22q11.2, 3q29). Our findings suggest an etiological overlap of ASD and SCZ and provide biological insights into these disorders.

Many studies have highlighted the difficulty inherent to the clinical application of fundamental neuroscience knowledge based on machine learning techniques. It is difficult to generalize machine learning brain markers to the data acquired from independent imaging sites, mainly due to large site differences in functional magnetic resonance imaging. We address the difficulty of finding a generalizable marker of major depressive disorder (MDD) that would distinguish patients from healthy controls based on resting-state functional connectivity patterns. For the discovery dataset with 713 participants from 4 imaging sites, we removed site differences using our recently developed harmonization method and developed a machine learning MDD classifier. The classifier achieved an approximately 70% generalization accuracy for an independent validation dataset with 521 participants from 5 different imaging sites. The successful generalization to a perfectly independent dataset acquired from multiple imaging sites is novel and ensures scientific reproducibility and clinical applicability.