BACKGROUND

Existing functional connectivity studies of psychosis use population-averaged functional network maps, despite highly variable topographies of these networks across the brain surface. We aimed to define the functional network areas and topographies in the general population and the changes associated with psychotic experiences (PEs) and disorders.

METHODS

Maps of 8 functional networks were generated using an individual-specific template-matching procedure for each participant from the Human Connectome Project Young Adult cohort ( = 1003) and from a matched case cohort (schizophrenia [SCZ],  = 27; bipolar disorder,  = 35) scanned identically with the same Connectom scanner. In the Human Connectome Project Young Adult cohort, PEs were estimated based on scores from the Achenbach Self-Report Scale. The relationship of symptoms to the probability of network representation at each cortical vertex was assessed using logistic regression.

RESULTS

In Human Connectome Project Young Adult participants, PE severity on the Achenbach thought problems scale was predicted by increased language network (LAN) and dorsal attention network (DAN) areas and decreased cingulo-opercular network area ( < 0.12). Significant effects were found in SCZ, with a larger DAN and LAN and a smaller frontoparietal network. Network pattern analysis in SCZ showed an increased probability of LAN in the posterior region of the left superior temporal gyrus and of the visual network in the left insula. Regression analyses in SCZ found that mood dysregulation was related to increased DAN surface area.

CONCLUSIONS

Those with PEs and SCZ showed abnormal functional network cortical topographies, particularly involving DAN and LAN. Network findings may predict psychosis progression and guide earlier intervention.