Multivariate transdiagnostic neural biomarkers of Schizophrenia and Autism Spectrum Disorders during the empathic accuracy task

Poster Session C - Sunday, April 14, 2024, 5:00 – 7:00 pm EDT, Sheraton Hall ABC

Ju-Chi Yu¹ (ju-chi.yu@camh.ca), Colin Hawco^1,2, Lindsay D. Oliver¹, Maria T. Secara^1,2, Iska Moxon-Emre¹, Fariah A. Sandhu³, Zara Z. Khan⁴, Peter Szatmari^1,2, Meng-Chuan Lai^1,2, Miklos Argyelan⁵, James M. Gold⁶, Sunny X. Tang⁵, George Foussias^1,2, Robert W. Buchanan⁶, Anil K. Malhotra⁵, Aristotle N. Voineskos^1,2, Stephanie H. Ameis^1,2, Erin W. Dickie^1,2; ¹Centre for Addiction and Mental Health, ²University of Toronto, ³York University, ⁴McMaster University, ⁵Zucker Hillside Hospital, ⁶Maryland Psychiatric Research Center

Schizophrenia (SSD) and Autism Spectrum Disorders (autism) are both characterized by social cognitive deficits, which vary substantially within, but overlap across, diagnoses. In this study, we aimed to identify group-specific and shared brain functional network configurations present during a social processing functional magnetic resonance imaging (fMRI) task. Functional connectivity during the empathic accuracy (EA) task from 411 participants (autism: N=67; SSD: N=174; Controls (HC): N=170) was analyzed by DiSTATIS (i.e., multi-table multidimensional scaling). DiSTATIS extracts a combined latent dimension space characterizing prominent network configurations across participants and enabling comparison between groups in the same space. Bootstrap tests were used to examine network configuration and group differences. Two orthogonal dimensions were identified from DiSTATIS. The first dimension’s configuration (explaining 14.98% of the signal) was characterized by differentiation between language (LAN) and default mode versus visual (VIS) and dorsal attention networks (DAN). The second dimension’s configuration (explaining 11.97% of the signal) was characterized by differentiation between auditory, somatomotor versus frontoparietal networks. On group-wise comparison, HC showed specific differentiation between VIS and DAN (p < .05); autism showed specific differentiation between LAN and ventral-multimodal network (p < .05); and SSD showed specific non-differentiation between primary and secondary VIS (p > .05). The prominent latent dimensions of the overall EA-related network configuration were similar across groups, with specific configuration of autism in LAN and of SSD in VIS. Future examination of their relationships with cognition and clinical outcomes can reveal potential biomarkers for social cognitive deficits in autism and SSD.

Topic Area: METHODS: Neuroimaging