Volume of subcortical brain regions in social anxiety disorder: mega-analytic results from 37 samples in the ENIGMA-Anxiety Working Group.

There is limited convergence in neuroimaging investigations into volumes of subcortical brain regions in social anxiety disorder (SAD). The inconsistent findings may arise from variations in methodological approaches across studies, including sample selection based on age and clinical characteristics. The ENIGMA-Anxiety Working Group initiated a global mega-analysis to determine whether differences in subcortical volumes can be detected in adults and adolescents with SAD relative to healthy controls. Volumetric data from 37 international samples with 1115 SAD patients and 2775 controls were obtained from ENIGMA-standardized protocols for image segmentation and quality assurance. Linear mixed-effects analyses were adjusted for comparisons across seven subcortical regions in each hemisphere using family-wise error (FWE)-correction. Mixed-effects d effect sizes were calculated. In the full sample, SAD patients showed smaller bilateral putamen volume than controls (left: d = -0.077, pFWE = 0.037; right: d = -0.104, pFWE = 0.001), and a significant interaction between SAD and age was found for the left putamen (r = -0.034, pFWE = 0.045). Smaller bilateral putamen volumes (left: d = -0.141, pFWE < 0.001; right: d = -0.158, pFWE < 0.001) and larger bilateral pallidum volumes (left: d = 0.129, pFWE = 0.006; right: d = 0.099, pFWE = 0.046) were detected in adult SAD patients relative to controls, but no volumetric differences were apparent in adolescent SAD patients relative to controls. Comorbid anxiety disorders and age of SAD onset were additional determinants of SAD-related volumetric differences in subcortical regions. To conclude, subtle volumetric alterations in subcortical regions in SAD were detected. Heterogeneity in age and clinical characteristics may partly explain inconsistencies in previous findings. The association between alterations in subcortical volumes and SAD illness progression deserves further investigation, especially from adolescence into adulthood.

Relationships Between Telomere Length, Plasma Glucagon-like Peptide 1, and Insulin in Early-Life Stress-Exposed Nonhuman Primates.

Background: Early-life stress is associated with alterations in telomere length, a marker of accumulated stress and aging, and a risk factor for psychiatric disorders. Nonhuman primate maternal variable foraging demand (VFD) is a validated early-life stress model, resulting in anxiety- and depressive-like symptoms in offspring. Previous studies reported increased plasma glucagon-like peptide 1 (pGLP-1) along with insulin resistance in this model. We investigated whether VFD rearing related to adult telomere length and to these neuroendocrine markers. Methods: Adult leukocyte telomere length was measured in VFD-reared (12 males, 13 females) and non-VFD-reared (9 males, 26 females) bonnet macaques. Associations between adult telomere length and adolescent fasting pGLP-1 or insulin resistance in VFD-reared versus non-VFD-reared groups were examined using regression modeling, controlling for sex, weight, and age. Results: VFD subjects had relatively longer telomeres than non-VFD subjects (p = .017), and females relatively longer than males (p = .0004). Telomere length was positively associated with pGLP-1 (p = .0009) and with reduced insulin sensitivity (p < .0001) in both sexes, but not as a function of rearing group. Conclusions: Unexpectedly, VFD was associated with longer adult telomere length. Insulin resistance may lead to higher pGLP-1 levels in adolescence, which could protect telomere length in VFD offspring as adults. Associations between adult telomere length and adolescent insulin resistance and high pGLP-1 may reflect an adaptive, compensatory response after early-life stress exposure.

Structural Brain Correlates of Childhood Inhibited Temperament: An ENIGMA-Anxiety Mega-analysis.

Temperament involves stable behavioral and emotional tendencies that differ between individuals, which can be first observed in infancy or early childhood and relate to behavior in many contexts and over many years.1 One of the most rigorously characterized temperament classifications relates to the tendency of individuals to avoid the unfamiliar and to withdraw from unfamiliar people, objects, and unexpected events. This temperament is referred to as behavioral inhibition or inhibited temperament (IT).2 IT is a moderately heritable trait1 that can be measured in multiple species.3 In humans, levels of IT can be quantified from the first year of life through direct behavioral observations or reports by caregivers or teachers. Similar approaches as well as self-report questionnaires on current and/or retrospective levels of IT1 can be used later in life.

Mega-analysis methods in ENIGMA: The experience of the generalized anxiety disorder working group.

The ENIGMA group on Generalized Anxiety Disorder (ENIGMA-Anxiety/GAD) is part of a broader effort to investigate anxiety disorders using imaging and genetic data across multiple sites worldwide. The group is actively conducting a mega-analysis of a large number of brain structural scans. In this process, the group was confronted with many methodological challenges related to study planning and implementation, between-country transfer of subject-level data, quality control of a considerable amount of imaging data, and choices related to statistical methods and efficient use of resources. This report summarizes the background information and rationale for the various methodological decisions, as well as the approach taken to implement them. The goal is to document the approach and help guide other research groups working with large brain imaging data sets as they develop their own analytic pipelines for mega-analyses.

Cortical and subcortical brain structure in generalized anxiety disorder: findings from 28 research sites in the ENIGMA-Anxiety Working Group.

The goal of this study was to compare brain structure between individuals with generalized anxiety disorder (GAD) and healthy controls. Previous studies have generated inconsistent findings, possibly due to small sample sizes, or clinical/analytic heterogeneity. To address these concerns, we combined data from 28 research sites worldwide through the ENIGMA-Anxiety Working Group, using a single, pre-registered mega-analysis. Structural magnetic resonance imaging data from children and adults (5-90 years) were processed using FreeSurfer. The main analysis included the regional and vertex-wise cortical thickness, cortical surface area, and subcortical volume as dependent variables, and GAD, age, age-squared, sex, and their interactions as independent variables. Nuisance variables included IQ, years of education, medication use, comorbidities, and global brain measures. The main analysis (1020 individuals with GAD and 2999 healthy controls) included random slopes per site and random intercepts per scanner. A secondary analysis (1112 individuals with GAD and 3282 healthy controls) included fixed slopes and random intercepts per scanner with the same variables. The main analysis showed no effect of GAD on brain structure, nor interactions involving GAD, age, or sex. The secondary analysis showed increased volume in the right ventral diencephalon in male individuals with GAD compared to male healthy controls, whereas female individuals with GAD did not differ from female healthy controls. This mega-analysis combining worldwide data showed that differences in brain structure related to GAD are small, possibly reflecting heterogeneity or those structural alterations are not a major component of its pathophysiology.

Developmental Antecedents of Adult Macaque Neurogenesis: Early-Life Adversity, 5-HTTLPR Polymorphisms, and Adolescent Hippocampal Volume.

INTRODUCTION: Attenuated adult hippocampal neurogenesis may manifest in affective symptomatology and/or resistance to antidepressant treatment. While early-life adversity and the short variant ('s') of the serotonin transporter gene's long polymorphic region (5-HTTLPR) are suggested as interacting risk factors for affective disorders, no studies have examined whether their superposed risk effectuates neurogenic changes into adulthood. Similarly, it is not established whether reduced hippocampal volume in adolescence, variously identified as a marker and antecedent of affective disorders, anticipates diminished adult neurogenesis. We investigate these potential developmental precursors of neurogenic alterations using a bonnet macaque model. METHODS: Twenty-five male infant bonnet macaques were randomized to stressed [variable foraging demand (VFD)] or normative [low foraging demand (LFD)] rearing protocols and genotyped for 5-HTTLPR polymorphisms. Adolescent MRI brain scans (mean age 4.2y) were available for 14 subjects. Adult-born neurons were detected post-mortem (mean age 8.6y) via immunohistochemistry targeting the microtubule protein doublecortin (DCX). Models were adjusted for age and weight. RESULTS: A putative vulnerability group (VG) of VFD-reared 's'-carriers (all 's/l') exhibited reduced neurogenesis compared to non-VG subjects. Neurogenesis levels were positively predicted by ipsilateral hippocampal volume normalized for total brain volume, but not by contralateral or raw hippocampal volume. LIMITATIONS: No 's'-carriers were identified in LFD-reared subjects, precluding a 2×2 factorial analysis. CONCLUSION: The 's' allele (with adverse rearing) and low adolescent hippocampal volume portend a neurogenic deficit in adult macaques, suggesting persistent alterations in hippocampal plasticity may contribute to these developmental factors' affective risk in humans.

Is treatment-resistant schizophrenia associated with distinct neurobiological callosal connectivity abnormalities?

BACKGROUND: Resistance to antipsychotic treatment affects up to 30% of patients with schizophrenia. Although the time course of development of treatment-resistant schizophrenia (TRS) varies from patient to patient, the reasons for these variations remain unknown. Growing evidence suggests brain dysconnectivity as a significant feature of schizophrenia. In this study, we compared fractional anisotropy (FA) of brain white matter between TRS and non-treatment-resistant schizophrenia (non-TRS) patients. Our central hypothesis was that TRS is associated with reduced FA values. METHODS: TRS was defined as the persistence of moderate to severe symptoms after adequate treatment with at least two antipsychotics from different classes. Diffusion-tensor brain MRI obtained images from 34 TRS participants and 51 non-TRS. Whole-brain analysis of FA and axial, radial, and mean diffusivity were performed using Tract-Based Spatial Statistics (TBSS) and FMRIB's Software Library (FSL), yielding a contrast between TRS and non-TRS patients, corrected for multiple comparisons using family-wise error (FWE) < 0.05. RESULTS: We found a significant reduction in FA in the splenium of corpus callosum (CC) in TRS when compared to non-TRS. The antipsychotic dose did not relate to the splenium CC. CONCLUSION: Our results suggest that the focal abnormality of CC may be a potential biomarker of TRS.

Cellular correlates of cortical thinning throughout the lifespan.

Cortical thinning occurs throughout the entire life and extends to late-life neurodegeneration, yet the neurobiological substrates are poorly understood. Here, we used a virtual-histology technique and gene expression data from the Allen Human Brain Atlas to compare the regional profiles of longitudinal cortical thinning through life (4004 magnetic resonance images [MRIs]) with those of gene expression for several neuronal and non-neuronal cell types. The results were replicated in three independent datasets. We found that inter-regional profiles of cortical thinning related to expression profiles for marker genes of CA1 pyramidal cells, astrocytes and, microglia during development and in aging. During the two stages of life, the relationships went in opposite directions: greater gene expression related to less thinning in development and vice versa in aging. The association between cortical thinning and cell-specific gene expression was also present in mild cognitive impairment and Alzheimer's Disease. These findings suggest a role of astrocytes and microglia in promoting and supporting neuronal growth and dendritic structures through life that affects cortical thickness during development, aging, and neurodegeneration. Overall, the findings contribute to our understanding of the neurobiology underlying variations in MRI-derived estimates of cortical thinning through life and late-life disease.

Assessment of Neurobiological Mechanisms of Cortical Thinning During Childhood and Adolescence and Their Implications for Psychiatric Disorders.

Importance: Many psychiatric disorders can be conceptualized as disorders of brain maturation during childhood and adolescence. Discovering the neurobiological underpinnings of brain maturation may elucidate molecular pathways of vulnerability and resilience to such disorders. Objective: To investigate the underlying neurobiological mechanisms of age-associated cortical thinning during maturation and their implications for psychiatric disorders. Design, Setting, and Participants: This multicohort analysis used data from 3 community-based studies. The Saguenay Youth Study provided data from 1024 adolescents who were recruited at a single site in Quebec, Canada. The IMAGEN cohort provided data from 1823 participants who were recruited in 8 European cities. The Brazil High Risk Cohort Study for the Development of Childhood Psychiatric Disorders provided data from 815 participants who were recruited in 2 Brazilian cities. Cortical thickness was estimated from the results of magnetic resonance imaging (MRI) scans, and age-associated cortical thinning was estimated in 34 cortical regions. Gene expression from the Allen Human Brain Atlas was aligned with the same regions. Similarities in the interregional profiles of gene expression and the profiles of age-associated cortical thinning were measured. The involvement of dendrites, dendritic spines, and myelin was tested using 3 gene panels. Enrichment for genes associated with psychiatric disorders was tested among the genes associated with thinning and their coexpression networks. Data analysis was conducted between March and October 2019. Main Outcomes and Measures: MRI-derived estimates of age-associated cortical thinning and gene expression in 34 cortical regions. Results: A total of 3596 individuals aged 9 to 21 years were included in this study. Of those, 1803 participants (50.1%) were female, and the mean (SD) age was 15.2 (2.6) years. Interregional profiles of age-associated cortical thinning were associated with interregional gradients in the expression of genes associated with dendrites, dendritic spines, and myelin; the variance in thinning explained by the gene panels across different points ranged from 0.45% to 10.55% for the dendrite panel, 0.00% to 9.98% for the spine panel, and 0.19% to 26.39% for the myelin panel. These genes and their coexpression networks were enriched for genes associated with several psychiatric disorders. Conclusions and Relevance: In this study, genetic similarity between interregional variation in cortical thinning during maturation and multiple psychiatric disorders suggests overlapping molecular underpinnings. This finding adds to the understanding of the neurodevelopmental mechanisms of psychiatric disorders.

Are serum brain-derived neurotrophic factor concentrations related to brain structure and psychopathology in late childhood and early adolescence?

OBJECTIVE: Mental disorders can have a major impact on brain development. Peripheral blood concentrations of brain-derived neurotrophic factor (BDNF) are lower in adult psychiatric disorders. Serum BDNF concentrations and BDNF genotype have been associated with cortical maturation in children and adolescents. In 2 large independent samples, this study tests associations between serum BDNF concentrations, brain structure, and psychopathology, and the effects of BDNF genotype on BDNF serum concentrations in late childhood and early adolescence. METHODS: Children and adolescents (7-14 years old) from 2 cities (n = 267 in Porto Alegre; n = 273 in São Paulo) were evaluated as part of the Brazilian high-risk cohort (HRC) study. Serum BDNF concentrations were quantified by sandwich ELISA. Genotyping was conducted from blood or saliva samples using the SNParray Infinium HumanCore Array BeadChip. Subcortical volumes and cortical thickness were quantified using FreeSurfer. The Development and Well-Being Behavior Assessment was used to identify the presence of a psychiatric disorder. RESULTS: Serum BDNF concentrations were not associated with subcortical volumes or with cortical thickness. Serum BDNF concentration did not differ between participants with and without mental disorders, or between Val homozygotes and Met carriers. CONCLUSIONS: No evidence was found to support serum BDNF concentrations as a useful marker of developmental differences in brain and behavior in early life. Negative findings were replicated in 2 of the largest independent samples investigated to date.

The impact of epilepsy duration in a series of patients with mesial temporal lobe epilepsy due to unilateral hippocampal sclerosis.

PURPOSE: To evaluate if the duration of epilepsy influences MRI volumes of the hippocampus, amygdala, parahippocampal gyrus, entorhinal cortex and temporal pole of both hemispheres and epileptogenic hippocampus neuronal cell density and dentate gyrus granular cells distribution in patients with refractory mesial temporal lobe epilepsy due to hippocampal sclerosis (MTLE/HS). METHODS: Seventy-seven patients with refractory MTLE/HS submitted to surgery were included. Histopathological analysis included: (1) quantitative: hippocampal subfields and total estimated hippocampal cell density (HCD), thickness of the dentate gyrus - normal, thinning or dispersion; (2) qualitative: type of HS and granule cells pathology in the dentate gyrus (normal, neuronal cell loss, dispersion and bilamination). Automated MRI-derived measurements from bilateral temporal structures (hippocampus, amygdala, parahippocampal gyrus, temporal pole, entorhinal cortex) were obtained for 58 subjects. Histopathological and imaging findings were compared with data from specimens obtained in autopsies of age-matched individuals and living controls, respectively, and the data were adjusted for the age at epilepsy onset and the frequency of focal impaired awareness seizures/month. RESULTS: Forty-two (54.5%) patients presented right HS. The greater the duration of epilepsy, the smaller the total estimated HCD (p = 0.025; r = -0.259). Patients with a normal distribution of the granular cells had a shorter epilepsy duration than those with dispersion (p = 0.018) or thinning (p = 0.031). A reduced ipsilateral hippocampal volume (r = -0.551, p = 0.017) and a smaller hippocampal asymmetry index (r = -0.414, p = 0.002) were correlated to a longer epilepsy duration. The estimated HCD was correlated to the volume of the ipsilateral hippocampus (r = 0.420, p = 0.001). CONCLUSION: Our study showed an increasing atrophy of the ipsilateral hippocampus in patients with a longer epilepsy duration. Our data suggest that this reduction in hippocampal volume is related to neuronal loss. Besides that, we also showed an increased probability of exhibiting an abnormal distribution of the granular cells in the dentate gyrus in patients with longer epilepsy duration.

Early Life Stress Associated With Increased Striatal N-Acetyl-Aspartate: Cerebrospinal Fluid Corticotropin-Releasing Factor Concentrations, Hippocampal Volume, Body Mass and Behavioral Correlates.

INTRODUCTION: Using proton magnetic resonance spectroscopy imaging (1H-MRSI), the effects of early life stress (ELS) on nonhuman primate striatal neuronal integrity were examined as reflected by N-acetyl aspartate (NAA) concentrations. NAA measures were interrogated through examining their relationship to previously documented ELS markers -- cerebrospinal fluid (CSF) corticotropin-releasing factor (CRF) concentrations, hippocampal volume, body mass and behavioral timidity. Rodent models of depression exhibit increases in neurotrophic effects in the nucleus accumbens (NAcc). We hypothesized that rearing under conditions of ELS [Variable Foraging Demand: (VFD)] would produce persistent elevations of NAA concentrations (in absolute or ratio form) in ventral striatum/caudate nucleus (VS/CN) with altered correlation to ELS markers. METHODS: Eleven bonnet macaque males reared under VFD conditions and seven age-matched control subjects underwent 1H-MRSI during young adulthood. Voxels were placed over ventral striatum/caudate nucleus (VS/CN) to capture NAcc. Cisternal CSF CRF concentrations, hippocampal volume, body mass and response to a human intruder had been previously determined. RESULTS: VFD-reared monkeys exhibited significantly increased NAA/Cr concentrations in right VS/CN in comparison to normally-reared controls, controlling for multiple comparisons. In comparison to controls, VFD CSF CRF concentrations were directly associated with right VS/CN absolute NAA. Left hippocampal volume was inversely associated with left VS/CN N-acetyl aspartate/creatine (NAA/Cr) in VFD-reared but not in controls. Disruption of a normative inverse correlation between left VS/CN NAA and body mass was noted in VFD. Only non-VFD subjects exhibited a direct relationship between timidity response to an intruder and right VS/CN NAA. CONCLUSION: ELS produced persistent increases in VS/CN NAA, which demonstrated specific patterns of association (or lack thereof) to ELS markers in comparison to non-VFD subjects. The data are broadly consistent with a stable nonhuman primate phenotype of anxiety and mood disorder vulnerability whereby in vivo indicators of neuronal integrity, although reduced in hippocampus, are increased in striatum. The findings may provide a catalyst for further studies in humans and other species regarding a reciprocal hippocampal/NAcc relationship in affective disorders.

Investigating brain structural patterns in first episode psychosis and schizophrenia using MRI and a machine learning approach.

In this study, we employed the Maximum Uncertainty Linear Discriminant Analysis (MLDA) to investigate whether the structural brain patterns in first episode psychosis (FEP) patients would be more similar to patients with chronic schizophrenia (SCZ) or healthy controls (HC), from a schizophrenia model perspective. Brain regions volumetric data were estimated by using MRI images of SCZ and FEP patients and HC. First, we evaluated the MLDA performance in discriminating SCZ from controls, which provided a score based on a model for changes in brain structure in SCZ. In the following, we compared the volumetric patterns of FEP patients with patterns of SCZ and healthy controls using these scores. The FEP group had a score distribution more similar to patients with schizophrenia (p-value = .461; Cohen's d=-.15) in comparison with healthy subjects (p-value=.003; Cohen's d = .62). Structures related to the limbic system and the circuitry involved in goal-directed behaviours were the most discriminant regions. There is a distinct pattern of volumetric changes in patients with schizophrenia in contrast to healthy controls, and this pattern seem to be detectable already in FEP.

Low frequency fluctuation of brain spontaneous activity and obsessive-compulsive symptoms in a large school-age sample.

BACKGROUND: The present study was designed to explore alterations in brain dynamics at rest that are associated with Obsessive Compulsive Symptoms (OCS) in childhood by measuring low frequency fluctuation of spontaneous brain activity in a large school community sample from a developing country. METHOD: Resting state functional magnetic resonance imaging data were collected in a sample of 655 children and adolescents (6-15 years old) from the brazilian 'High Risk Cohort Study for Psychiatric Disorders (HRC)'. OCS were assessed using items from the Compulsion and Obsessions section of the Development and Well-Being Assessment (DAWBA). The correlation between the fractional amplitude of low frequency fluctuations (fALFF) and the number of OCS were explored by using a general linear model, considering fALFF as response variable, OCS score as regressor and age, gender and site as nuisance variables. RESULTS: The number of OCS was positively correlated with the fALFF coefficients at the right sensorimotor cortex (pre-motor, primary motor cortex and post-central gyrus) and negatively correlated with the fALFF coefficients at the insula/superior temporal gyrus of both hemispheres. Our results were specific to OCS and not due to associations with overall psychopathology. CONCLUSIONS: Our results suggest that brain spontaneous activity at rest in the sensorimotor and insular/superior-temporal cortices may be involved in OCS in children. These findings need independent replication and future studies should determine whether brain spontaneous activity changes within these regions might be predictors of risk for obsessive-compulsive disorder latter in life.

Assessing Music Perception in Young Children: Evidence for and Psychometric Features of the M-Factor.

Given the relationship between language acquisition and music processing, musical perception (MP) skills have been proposed as a tool for early diagnosis of speech and language difficulties; therefore, a psychometric instrument is needed to assess music perception in children under 10 years of age, a crucial period in neurodevelopment. We created a set of 80 musical stimuli encompassing seven domains of music perception to inform perception of tonal, atonal, and modal stimuli, in a random sample of 1006 children, 6-13 years of age, equally distributed from first to fifth grades, from 14 schools (38% private schools) in So Paulo State. The underlying model was tested using confirmatory factor analysis. A model encompassing seven orthogonal specific domains (contour, loudness, scale, timbre, duration, pitch, and meter) and one general music perception factor, the "m-factor," showed excellent fit indices. The m-factor, previously hypothesized in the literature but never formally tested, explains 93% of the reliable variance in measurement, while only 3.9% of the reliable variance could be attributed to the multidimensionality caused by the specific domains. The 80 items showed no differential item functioning based on sex, age, or enrolment in public vs. private school, demonstrating the important psychometric feature of invariance. Like Charles Spearman's g-factor of intelligence, the m-factor is robust and reliable. It provides a convenient measure of auditory stimulus apprehension that does not rely on verbal information, offering a new opportunity to probe biological and psychological relationships with music perception phenomena and the etiologies of speech and language disorders.

Does pediatric post-traumatic stress disorder alter the brain? Systematic review and meta-analysis of structural and functional magnetic resonance imaging studies.

Several studies have recently demonstrated that the volumes of specific brain regions are reduced in children and adolescents with post-traumatic stress disorder (PTSD) compared with those of healthy controls. Our study investigated the potential association between early traumatic experiences and altered brain regions and functions. We conducted a systematic review of the scientific literature regarding functional magnetic resonance imaging and a meta-analysis of structural magnetic resonance imaging studies that investigated cerebral region volumes in pediatric patients with PTSD. We searched for articles from 2000 to 2014 in the PsycINFO, PubMed, Medline, Lilacs, and ISI (Web of Knowledge) databases. All data regarding the amygdala, hippocampus, corpus callosum, brain, and intracranial volumes that fit the inclusion criteria were extracted and combined in a meta-analysis that assessed differences between groups. The meta-analysis found reduced total corpus callosum areas and reduced total cerebral and intracranial volumes in the patients with PTSD. The total hippocampus (left and right hippocampus) and gray matter volumes of the amygdala and frontal lobe were also reduced, but these differences were not significant. The functional studies revealed differences in brain region activation in response to stimuli in the post-traumatic stress symptoms/PTSD group. Our results confirmed that the pediatric patients with PTSD exhibited structural and functional brain abnormalities and that some of the abnormalities occurred in different brain regions than those observed in adults.

Using deep belief network modelling to characterize differences in brain morphometry in schizophrenia.

Neuroimaging-based models contribute to increasing our understanding of schizophrenia pathophysiology and can reveal the underlying characteristics of this and other clinical conditions. However, the considerable variability in reported neuroimaging results mirrors the heterogeneity of the disorder. Machine learning methods capable of representing invariant features could circumvent this problem. In this structural MRI study, we trained a deep learning model known as deep belief network (DBN) to extract features from brain morphometry data and investigated its performance in discriminating between healthy controls (N = 83) and patients with schizophrenia (N = 143). We further analysed performance in classifying patients with a first-episode psychosis (N = 32). The DBN highlighted differences between classes, especially in the frontal, temporal, parietal, and insular cortices, and in some subcortical regions, including the corpus callosum, putamen, and cerebellum. The DBN was slightly more accurate as a classifier (accuracy = 73.6%) than the support vector machine (accuracy = 68.1%). Finally, the error rate of the DBN in classifying first-episode patients was 56.3%, indicating that the representations learned from patients with schizophrenia and healthy controls were not suitable to define these patients. Our data suggest that deep learning could improve our understanding of psychiatric disorders such as schizophrenia by improving neuromorphometric analyses.