Altered temporal, but intact spatial, features of transient network dynamics in psychosis.
Danhong WangXiaolong PengAndrea Pelletier-BaldelliNatasza OrlovAmy FarabaughShahin NasrHamdi EryilmazMaurizio FavaAvram J HomesJoshua L RoffmanHesheng LiuDaphne J HoltPublished in: Molecular psychiatry (2021)
Contemporary models of psychosis suggest that a continuum of severity of psychotic symptoms exists, with subthreshold psychotic experiences (PEs) potentially reflecting some genetic and environmental risk factors shared with clinical psychosis. Thus, identifying abnormalities in brain activity that manifest across this continuum can shed new light on the pathophysiology of psychosis. Here, we investigated the moment-to-moment engagement of brain networks ("states") in individuals with schizophrenia (SCZ) and PEs and identified features of these states that are associated with psychosis-spectrum symptoms. Transient brain states were defined by clustering "single snapshots" of blood oxygen level-dependent images, based on spatial similarity of the images. We found that individuals with SCZ (n = 35) demonstrated reduced recruitment of three brain states compared to demographically matched healthy controls (n = 35). Of these three illness-related states, one specific state, involving primarily the visual and salience networks, also occurred at a lower rate in individuals with persistent PEs (n = 22), compared to demographically matched healthy youth (n = 22). Moreover, the occurrence rate of this marker brain state was negatively correlated with the severity of PEs (r = -0.26, p = 0.003, n = 130). In contrast, the spatial map of this state appeared to be unaffected in the SCZ or PE groups. Thus, reduced engagement of a brain state involving the visual and salience networks was demonstrated across the psychosis continuum, suggesting that early disruptions of perceptual and affective function may underlie some of the core symptoms of the illness.
Keyphrases
- resting state
- functional connectivity
- cerebral ischemia
- white matter
- bipolar disorder
- risk factors
- deep learning
- social media
- magnetic resonance
- sleep quality
- optical coherence tomography
- gene expression
- physical activity
- machine learning
- blood brain barrier
- convolutional neural network
- single cell
- dna methylation
- genome wide
- climate change
- subarachnoid hemorrhage
- high resolution
- copy number