Human brain is a topologically complex network embedded in anatomical space, and anatomical distance may affect functional connectivity (FC) in schizophrenia. However, little is known if and how this effect occurs in adolescent-onset schizophrenia (AOS).Methods:
We explored long- and short-range FC through resting-state functional magnetic resonance imaging in 48 first-episode, drug-naive AOS patients and 31 healthy controls, and we examined if these abnormalities could be utilized to separate patients from controls using receiver operating characteristic curves and support vector machines (SVM).Results:
Patients had increased long-range positive FC (lpFC) and short-range positive FC (spFC) in the right middle frontal gyrus and right superior medial prefrontal cortex within the anterior default mode network (DMN), decreased lpFC and spFC in several regions of the posterior DMN, and decreased lpFC within the important hubs of salience network (SN). The decreased lpFC in the left superior temporal gyrus was positively correlated with cognitive impairment. We found that SVM has high accuracy (up to 92.4%) in classifying patients and control.Conclusion:
Disrupted anatomical distance would underlie network-level dysconnectivity, highlighting the importance of the DMN and SN in the neurodevelopment of schizophrenia. Abnormalities of long- and short-range FC in brain regions could discriminate patients from controls with high accuracy.