Search

Schizophrenia (SCZ) is a chronic and serious mental disorder with a high mortality rate. At present, there is a lack of objective, cost-effective and widely disseminated diagnosis tools to address this mental health crisis globally. Clinical electroencephalogram (EEG) is a noninvasive technique to measure brain activity with high temporal resolution, and accumulating evidence demonstrates that clinical EEG is capable of capturing abnormal SCZ neuropathology. Although EEG-based automated diagnostic tools have obtained impressive performance on individual datasets, the transportability of potential EEG biomarkers in cross-site real-world application is still an open question. To address the challenges of small sample sizes and population heterogeneity, we develop an advanced interpretable deep learning model using multimodal clinical EEG features and demographic information as inputs to graph neural networks, and further propose different transfer learning strategies to adapt to different clinical scenarios. Taking the disease discrimination of health control (HC) and SCZ with 1030 participants as a use case, our model is trained on a small clinical dataset (N= 188, Chinese) and enhanced using a large-scale public dataset (N= 508, American) of adult participants. Cross-site validation from an independent dataset of adult participants (N= 157, Chinese) produced stable performance, with AUCs of 0.793–0.852 and accuracies of 0.786–0.858 for different SCZ prevalence, respectively. In addition, cross-site validation from another dataset of adolescent boys (N= 84, Russian) yielded an AUC of 0.702 and an accuracy of 0.690. Moreover, feature visualization further revealed that the ranking of feature importance varied significantly among different datasets, and that EEG theta and alpha band power appeared to be the most significant and translational biomarkers of SCZ pathology. Overall, our promising results demonstrate the feasibility of SCZ discrimination using EEG biomarkers in multiple clinical settings.

Across the major psychiatric disorders (MPDs), a shared disruption in brain physiology is suspected. Here we investigate the neural variability at rest, a well-established behavior-relevant marker of brain function, and probe its basis in gene expression and neurotransmitter receptor profiles across the MPDs. We recruited 219 healthy controls and 279 patients with schizophrenia, major depressive disorder, or bipolar disorders (manic or depressive state). The standard deviation of blood oxygenation level-dependent signal (SDBOLD) obtained from resting-state fMRI was used to characterize neural variability. Transdiagnostic disruptions in SDBOLDpatterns and their relationships with clinical symptoms and cognitive functions were tested by partial least-squares correlation. Moving beyond the clinical sample, spatial correlations between the observed patterns of SDBOLDdisruption and postmortem gene expressions, Neurosynth meta-analytic cognitive functions, and neurotransmitter receptor profiles were estimated. Two transdiagnostic patterns of disrupted SDBOLDwere discovered. Pattern 1 is exhibited in all diagnostic groups and is most pronounced in schizophrenia, characterized by higher SDBOLDin the language/auditory networks but lower SDBOLDin the default mode/sensorimotor networks. In comparison, pattern 2 is only exhibited in unipolar and bipolar depression, characterized by higher SDBOLDin the default mode/salience networks but lower SDBOLDin the sensorimotor network. The expression of pattern 1 related to the severity of clinical symptoms and cognitive deficits across MPDs. The two disrupted patterns had distinct spatial correlations with gene expressions (e.g., neuronal projections/cellular processes), meta-analytic cognitive functions (e.g., language/memory), and neurotransmitter receptor expression profiles (e.g., D2/serotonin/opioid receptors). In conclusion, neural variability is a potential transdiagnostic biomarker of MPDs with a substantial amount of its spatial distribution explained by gene expressions and neurotransmitter receptor profiles. The pathophysiology of MPDs can be traced through the measures of neural variability at rest, with varying clinical-cognitive profiles arising from differential spatial patterns of aberrant variability.

Schizophrenia is a complex disorder associated with aberrant brain functional connectivity. This study aims to demonstrate the relation of heterogeneous symptomatology in this disorder to distinct brain connectivity patterns within the triple-network model. The study sample comprised 300 first-episode antipsychotic-naive patients with schizophrenia (FES) and 301 healthy controls (HCs). At baseline, resting-state functional magnetic resonance imaging data were captured for each participant, and concomitant neurocognitive functions were evaluated outside the scanner. Clinical information of 49 FES in the discovery dataset were reevaluated at a 6-week follow-up. Differential features between FES and HCs were selected from triple-network connectivity profiles. Cutting-edge unsupervised machine learning algorithms were used to define patient subtypes. Clinical and cognitive variables were compared between patient subgroups. Two FES subgroups with differing triple-network connectivity profiles were identified in the discovery dataset and confirmed in an independent hold-out cohort. One patient subgroup appearing to have more severe clinical symptoms was distinguished by salience network (SN)-centered hypoconnectivity, which was associated with greater impairments in sustained attention. The other subgroup exhibited hyperconnectivity and manifested greater deficits in cognitive flexibility. The SN-centered hypoconnectivity subgroup had more persistent negative symptoms at the 6-week follow-up than the hyperconnectivity subgroup. The present study illustrates that clinically relevant cognitive subtypes of schizophrenia may be associated with distinct differences in connectivity in the triple-network model. This categorization may foster further analysis of the effects of therapy on these network connectivity patterns, which may help to guide therapeutic choices to effectively reach personalized treatment goals.

Antipsychotic-induced metabolic disturbance (AIMD) is a common adverse effect of antipsychotics with genetics partly underpinning variation in susceptibility among schizophrenia patients. Melanocortin4 receptor (MC4R) gene, one of the candidate genes for AIMD, has been under-studied in the Chinese patients. We conducted a pharmacogenetic study in a large cohort of Chinese patients with schizophrenia. In this study, we investigated the genetic variation of MC4Rin Chinese population by genotyping two SNPs (rs489693 and rs17782313) in 1,991 Chinese patients and examined association of these variants with the metabolic effects that were often observed to be related to AIMD. Metabolic measures, including body mass index (BMI), waist circumference (WC), glucose, triglyceride, high-density lipoprotein (HDL), and low-density lipoprotein (LDL) levels were assessed at baseline and after 6-week antipsychotic treatment. We found that interaction of SNP×medication status (drug-naïve/medicated) was significantly associated with BMI, WC, and HDL change %, respectively. Both SNPs were significantly associated with baseline BMI and WC in the medicated group. Moderate association of rs489693 with WC, Triglyceride, and HDL change % were observed in the whole sample. In the drug-naïve group, we found recessive effects of rs489693 on BMI gain more than 7%, WC and Triglyceride change %, with AA incurring more metabolic adverse effects. In conclusion, the association between rs489693 and the metabolic measures is ubiquitous but moderate. Rs17782313 is less involved in AIMD. Two SNPs confer risk of AIMD to patients treated with different antipsychotics in a similar way.

In this study, we investigated the association between bipolar I disorder (BDI) and between cognitive deficits therein and SNPs in GABAergic receptor genes. The sample comprised 477 patients with BDI and 438 healthy controls, with three neurocognitive tests being administered in 123 patients and 164 controls. For three SNPs, rs505474, rs1398175, and rs4868029 in the GABRA2, GABRA4, and GABRPgenes, respectively, their allele frequencies were significantly different between patients and controls (Bonferroni‐adjusted p= values 3.84 × 10−4, 9.92 × 10−3, and 1.22 × 10−2, respectively). Four haplotypes were significantly associated with BDI (TA and AG for rs3815762 and rs4868029 in GABRP, GG for rs11636988 and rs8024256 in GABRB3and GAGG for rs2197414, rs4921195, rs13188991, and rs11956731 in GABRA6, with pvalues of 0.0038, 0.044, 0.0176, and 0.0267, respectively, on 10,000 permutations). Furthermore, the SNP (rs2912585) within 250 kb upstream of the GABRB3gene displayed a strong association with the Tower of Hanoi (TOH) executive time in the patient group (p= 2.844 × 10−6). One other SNP (rs754661), which is located at the intronic region of the same gene, was associated with the global trait of the executive function and post hoc analysis showed significant SNP by group effect (p= 0.0094). Our study supports previous findings that GABAAreceptor genes are associated with bipolar disorder; it also suggests that the GABAAgenes, especially the GABRB3gene, might play a role in the executive function deficit in bipolar disorder, although future replication with a larger sample size is needed.

Objective Major depressive disorder is associated with abnormal functioning of the hypothalamic–pituitary–adrenal (HPA) axis. Studies using hair cortisol to measure the effect of antidepressants on the HPA axis are lacking. The aim of this study was to explore the long-term effects of antidepressants on hair cortisol concentration (HCC).Methods Participants were 42 patients and 36 healthy individuals. The patients took antidepressants for 4 weeks. Patient HCC was measured pre-treatment and post-treatment. The HCC of healthy controls was also measured.Results Patient post-treatment HCC (mean ± standard deviation: 34.40 ± 32.57 pmol/mg) was significantly higher than patient pre-treatment HCC (17.42 ± 12.40 pmol/mg) and healthy control HCC (10.22 ± 7.99 pmol/mg). No significant correlation was found between Hamilton Depression Rating Scale scores and HCC at pre-treatment or post-treatment.Conclusions Hair cortisol concentration analysis could be used to monitor the dynamics of the effects of antidepressants on the HPA axis.