Identifying shared and distinct alterations in brain structure across childhood psychiatric disorders using the ENIGMA consortium

Epidemiological studies estimate that around 16% of children have a diagnosable mental health problem. Childhood psychiatric disorders are associated with adverse outcomes such as poorer quality of life and functioning, criminality, and continued mental health difficulties. Consequently, they incur significant individual and societal costs. Studying mental health difficulties in childhood is key to promoting early prevention and intervention strategies.

While modern psychiatry has largely focused on categorical diagnoses, there is increasing evidence of comorbidity transcending these distinct diagnoses, as well as shared genetic influences across disorders. Still, to date most neuroimaging studies focus on single, specific disorders, rather than comparing multiple diagnoses with each other and control groups. Those that do attempt cross-disorder comparisons often focus exclusively on adult samples, and yield inconsistent findings that prevent a robust understanding of transdiagnostic neurobiological substrates. Small sample sizes and heterogenous participant groups (e.g., in terms of age and sex) reduce our ability to compare across studies and negatively impact statistical power. Additionally, methodological differences (e.g., data acquisition and analysis methods) further limit the comparability of findings.

The application of meta- and mega-analyses are therefore vital to study shared and distinct neurobiological alterations across disorders, using harmonised protocols for data acquisition, processing and analysis. The feasibility of this large-scale research has been established in studies of brain structure by Goodkind et al. (2015) and Opel et al. (2020), and in white-matter connectivity by Koshiyama et al. (2020), all of whom found evidence of both disorder-specific and cross-disorder alterations. However, there is a dearth of evidence from child and adolescent samples.

This project will therefore address a gap in the literature, with the aim of investigating shared versus distinct alterations in brain structure or connectivity across different childhood psychiatric disorders. Furthermore, we will explore different patterns of brain maturation across disorders, and study the impact of factors such as comorbidity and sex. This study will utilise data from the Enhancing Neuro-Imaging Genetics through Meta-Analysis (ENIGMA) consortium, an international team-science collaboration aiming to pool data and resources, and overcome common limitations by avoiding methodological heterogeneity and publication bias (Thompson et al. 2020). While ENIGMA is comprised of over 50 working groups, this project will draw on data from the working groups focusing on emotional and behavioural disorders. The final sample size of those with disorders is estimated to be 6,000-10,000, with an even larger number of typically-developing controls.

This project will be the largest and most comprehensive neuroimaging study of transdiagnostic psychopathology in children and adolescents to date, while mega-analysis of white-matter connectivity data is a particularly novel methodology in the field. Whether disorder-general or specific, the findings will inform key debates regarding the nature of psychopathology, potentially building on behavioural genetic research to provide an organising model of childhood mental health, with important implications for our diagnostic systems (e.g., DSM, ICD). Stratifying the samples by age group will allow the investigation of different patterns of brain maturation across disorders, a phenomenon which may have been a confound in previous research. The findings of this project could contribute to early diagnosis and treatment of childhood disorders, with significant benefits to society at large.

I hope to go on an institutional visit to the University of Southern California, where ENIGMA Central is based.