Xiaochen Tang1, Tianhong Zhang1, Yingying Tang1, Lihua Xu1, Junjie Wang1, Margaret Niznikiewicz2, Hui Li3, Martha E. Shenton4,5,6, Susan Whitfield-Gabrieli7, Matcheri Keshavan8, William S. Stone8, JiJun Wang1; 1Shanghai Key Laboratory of Psychotic Disorders, Shanghai Mental Health Center, Shanghai Jiao Tong University School of Medicine, Shanghai, China;, 2Department of Psychiatry, Veterans Affairs Boston Healthcare System, Brockton Division, Brockton, MA, USA;, 3Department of Psychology, Florida A&M University, Tallahassee, FL, USA;, 4Psychiatry Neuroimaging Laboratory, Department of Psychiatry, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA;, 5Department of Radiology, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA;, 6Research and Development, Veterans Affairs Boston Healthcare System, Brockton Division, Brockton, MA, USA;, 7McGovern Institute for Brain Research and Poitras Center for Affective Disorders Research, Massachusetts Institute of Technology, Cambridge, MA, USA;, 8Massachusetts Mental Health Center, Public Psychiatry Division, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA;
Individuals at clinical high risk (CHR) for psychosis are defined primarily by the presence of clinical symptoms such as attenuated positive symptoms or brief intermittent psychotic symptoms. The heterogeneity in CHR subjects is likely as striking as is observed in schizophrenia, bipolar disorders, and/or major depressive disorders. To detect potential biotypes with distinct neurophysiological features, we performed a canonical correlation analysis to select linear combinations of MRI resting state functional connectivity indicators associated with clinical symptoms, as identified by Structured Interview of Psychosis-risk Syndromes (SIPS). This analysis yielded two sets of symptoms with negative and disintegration (loadings = 0.995, 0.427), and positive (loadings = 0.971). In terms of connection features, both sets of connectivity combinations include network-wide connections between fronto-parietal and salience networks. Furthermore, the first connectivity component defines functional connections between the default mode network and other networks, while the second component defines distinct connection combination between the default mode network and the visual network to other networks. Based on the two corresponding combinations of functional connectivity characteristics, cluster analysis was performed to classify 138 CHR subjects from the Shanghai At Risk Psychosis Project (SHARP) into 4 biological subtypes, with percentages of 10.8%, 16.7%, 14.5% and 60.1% respectively. These results demonstrate that individuals at CHR states can be classified into four biotypes defined by distinct patterns of connectivity primarily in default mode networks. Validation and further inquiry into clinical significance of these four biotypes in CHR is ongoing.
