Salience network segregation and symptom profiles in psychosis risk subgroups among youth and early adults.

Understanding neurobiological similarities among individuals with psychosis risk symptoms can improve early identification and intervention strategies. We aimed to (i) identify neurobiologically similar psychosis risk subgroups by integrating resting-state functional connectivity and psychosis risk symptom data and (ii) discern discriminating symptom profiles and brain connectivity patterns in the identified sub-groups. Our sample (N = 922) was extracted from the Philadelphia Neurodevelopmental Cohort, a community group of individuals aged 12-21 years, with fMRI and self-reported psychopathology data. Analyses were conducted separately for youth and early adults. We constructed a two-layer network using pair-wise similarity distances between participants based on resting-state fMRI and psychosis risk symptoms measured with the PRIME screen. We then performed community detection via a multiplex stochastic block model to identify subject clusters. We identified 2 blocks or communities for both the youth (n = 458 and 179) and early adult (n = 173 and 112) groups. Connection parameter estimates of the neuroimaging layer were nearly identical between blocks for both age groups whereas there was significant variation for the symptom layer. Psychopathology symptom and brain system segregation profiles were consistent across age groups. The youth block (n = 458) with higher salience network segregation values had higher mean psychosis risk symptom scores while the early adult block (n = 173) with lower salience network segregation had higher mean psychosis risk symptom scores. By integrating global similarities in brain connectivity and psychosis risk symptoms, we identified distinct subgroups. These groups exhibit different symptom profiles and network segregation in youth and early adults, suggesting variations in developmental paths for psychosis spectrum.