This project will use longitudinal neuroimaging to study the network features of structural and functional connectivity that characterize subclinical psychotic experiences (PEs) and their development over time. Previous work with the ALSPAC cohort has shown that compared to healthy controls, subjects with a history of PEs show schizophrenia-like topological changes in structural connectivity (Drakesmith et al. 2015). Moreover, the presence of PEs has been associated with mild changes in frontoparietal network configuration during a working memory task (Fonville et al. 2015).

Extending this prior work, this study will examine how changes in connectivity over time correspond with the persistence of psychotic experiences. We hypothesize that subjects with PEs that persist will show increasingly divergent connectivity phenotypes from healthy controls, while people with transient PEs will show evidence of ‘recovery’ towards the phenotype of people with no history of PEs. We will then use machine learning to integrate multimodal connectivity data and predict PE trajectories. Finally, we will compare our results to those identified as showing case/control differences in clinical populations to bridge the gap in our understanding of the connectivity signatures of PEs and fully-fledged psychotic disorders.