Longitudinal Illness- and Medication-Related Brain Volume Changes in Psychosis are Shaped by Connectome Architecture

Distributed grey matter brain regions are connected by a complex structural network of white matter fibres, which are responsible for the propagation of action potentials and the transport of neurochemicals. In neurodegenerative disease, these connections constrain the way in which grey matter volume (GMV) loss progresses. Here, we investigated whether connectome architecture also shapes the spatial patenting of longitudinal GMV changes attributable to illness and/or antipsychotic medication in first-episode psychosis (FEP).We conducted a triple-blind randomised placebo-control trial where 62 people with FEP received either an antipsychotic or placebo over 6 months. A healthy control group was also recruited. Anatomical MRI scans were acquired at baseline, 3-months and 12-months. Deformation-based morphometry was used to estimate GMV changes over time. Structural brain connectivity patterns were derived from healthy controls using diffusion-weighted imaging. We tested the hypothesis that GMV changes in any given brain region could be predicted by changes in areas to which it is structurally connected. We found a strong correlation between longitudinal regional illness-related change (r=.613; p< 0.001) and medication-related volume change (r=.591; p<.001) with volumetric changes in structurally connected regions. No such associations were found for functionally connected regions (all r< .391). Psychosis- and antipsychotic-related GMV changes are strongly shaped by structural brain connections. Psychosis- and antipsychotic-related GMV changes are strongly shaped by structural brain connections. This result is consistent with findings in other neurological disorders and implies that structural brain connections may act as a conduit for the spread of pathological processes causing brain dysfunction in FEP.