Neuroimaging Dimensions at the Extremes of the Schizophrenia Spectrum

PROJECT SUMMARY Schizophrenia (SZ) is a severe mental health disorder currently treated with antipsychotic drugs which often have serious side effects, are ineffective in ~30% of patients, and are not useful as a preventive treatment. With its burden estimated at $343.2 billion in the US alone, there is a pressing need to improve early identification strategies for SZ and treatments that improve functioning. A better understanding of the underlying neurobiology is key to achieving these goals. Advances in global, large-scale, collaborative neuroimaging efforts are able to generate replicable findings on brain abnormalities in SZ and assess contributions of clinical and confounding variables. Our prior work within the ENIGMA (Enhancing Neuroimaging Genetics through Meta-Analysis) SZ working group identified predominantly gray matter deficiencies such as smaller subcortical volumes and thinner cortex in SZ. However, our findings also highlighted the importance of confounding variables on such clinical neuroimaging data, such as antipsychotic medication or disease chronicity. In contrast, the ENIGMA Schizotypy (SZT) working group, studying well-functioning healthy individuals who self-report subclinical psychotic traits, recently found that higher SZT was associated with thicker cortex. Moreover, the cortical thickness profile in SZT was inversely related with the profile of cortical thinning in SZ. Based on these findings, we posit that a comprehensive characterization of neural abnormalities at the extremes of the SZ spectrum continuum may not only further our understanding of the low liability (SZT) and high liability (SZ) ends of the spectrum but may also have wider implications for the prediction of risk and resilience in SZ (functioning) and in individuals at clinical high-risk for psychosis. This innovative global initiative will be first to integrate data across ENIGMA SZT and SZ cohorts with structural MRI (sMRI), detailed diffusion tensor imaging (DTI), and resting-state functional MRI (rsfMRI) to address the following key questions in SZ research: what are the functional and structural connectivity signatures of SZT and SZ? How are these signatures related to symptom severity, and global functioning? Can label-noise reduced dimensional measures of SZT and SZ liability based on these multimodal neuroimaging measures predict poor functioning? Leveraging global data and expert teams from minimally 24 to possibly well over 100 cohorts, we will tackle imaging, clinical, and predictive questions about the SZ spectrum with unprecedented power. This project will employ standardized image analyses, quality assurance, and statistical analysis procedures across cohorts with multimodal neuroimaging and clinical data from the same individuals. This will yield replicated FC and SC signatures of SZT and SZ, determine relationships between neuroimaging markers of SZT and SZ, and relationships with symptom dimensions and functioning. It will deliver machine- learning based models that can be used to generate imaging-based, dimensional, biomarkers that may serve as treatment targets, predictors of clinical outcome, or predictors of conversion to psychosis in at-risk populations.