One of the most fundamental hypotheses concerning the pathophysiology of psychotic illnesses such as schizophrenia-spectrum disorders (SZ) is that they are disorders of dysconnectivity caused by abnormal interactions between brain regions. The hypothesis was first proposed by Wernicke (1906) as a loss of fiber tract integrity he termed a “sejunction” and Bleuler's theoretical “loosening of associations” (1911). The theory was then reconceptualized later by Friston and Frith (1995) as a “disconnection syndrome” as evidenced from neuroimaging studies suggesting that loss of connectivity may result in loss of “intrinsically generated action” and “aberrant perception resulting from misattribution of a self-induced sensory change to external agencies.” The hypothesis has since been frequently investigated using a wide range of imaging modalities, including structural and functional magnetic resonance imaging (fMRI) as well as diffusion tensor imaging (reviewed by Dong et al., 2018; Fitzsimmons et al., 2013; Kraguljac et al., 2021; Perry et al., 2019; Pettersson-Yeo et al., 2011; Wheeler and Voineskos, 2014). Overall, these studies suggests that these disorders are associated with widespread patterns of dysconnectivity between regions, in agreement with Wernicke's and Blueler's theories. Indeed, a 2018 meta-analysis of 52 studies by Dong et al. (2018) found reduced connectivity within the default mode, affective, ventral attention (VAN), thalamic, and somatosensory networks as well as between the VAN and thalamic, VAN and default mode, VAN and frontoparietal, frontoparietal and thalamic, and frontoparietal and default mode networks in SZ. Furthermore, evidence suggests the extent of functional dysconnectivity may also predict severity of positive symptoms in SZ (e.g., Damiani et al., 2022; O'Neill et al., 2019; Palaniyappan et al., 2013; Venkataraman et al., 2012). Importantly, identifying functional biomarkers that predict symptom severity is essential if psychiatric clinical research is to establish that investigational therapeutics are modifying their intended neuronal targets (Wylie et al., 2016).

Taken together, the results of previous studies suggest that abnormal functional connectivity may predict symptom severity in psychosis. One may additionally postulate that positive and disorganization symptoms (as well as general functioning) may be particularly affected by abnormal connectivity patterns in psychotic illness based on previous hypotheses. Here, we used resting state fMRI to examine the relationship between overall intrinsic functional connectivity and symptoms/functioning in early psychosis (EP). As we were broadly interested in relationships to psychosis symptoms, our EP group included individuals with schizophrenia-spectrum disorders as well as Type I bipolar disorder with psychotic features. To calculate global connectivity, we calculated a “similarity index,” in which the connectivity matrix of an EP individual was correlated to the average matrix of a sample of healthy control (HC) participants. This index was thus considered a measure of how closely each EP participant's global connectivity matrix resembled that of the average unaffected individual. We hypothesized EP individuals with matrices less similar to the HC average would show more severe positive and disorganization symptoms as well as lower functioning relative to those more similar to the HC average.