Across all stages of schizophrenia and related psychoses, impairments in episodic memory predict poor outcomes. Previous research has posited the hippocampus as a key brain region for understanding both negative symptoms and episodic memory. Our team proposed that hippocampal dysconnectivity impairs episodic memory, which in turn negatively affects social cognitive abilities and enhances the presence of negative symptoms (e.g., asociality, apathy), affecting functional outcome. While we were able to provide preliminary evidence for such a temporal progression, an investigation of this model considering a structural and functional assessment of hippocampal connectivity and controlling for important moderating factors including sex and illness chronicity is currently lacking. To validate this model, we will generate a multisite dataset consisting of 300 individuals with schizophrenia spectrum disorder or bipolar I disorder with psychosis and 150 non-clinical controls, to determine multimodal hippocampal connectivity, memory impairments, functional outcome, and potential mediating/moderating factors (e.g., social cognition, negative symptoms). The machine learning algorithm Subtype and Stage Inference will be applied to unravel data-driven disease phenotypes with temporal progression patterns and enable individual-level predictions. To promote open science principles, the resulting dataset and analytic code will be made publicly available. This study combines both data-driven and hypothesis-driven approaches to provide a novel mechanistic account of how memory impairments impact functioning in individuals with psychotic disorders. It also follows a translational framework aiming to reduce heterogeneity in biomarker discovery and promote precision psychiatry. The resulting open dataset and code will advance open science in psychiatric illness. Our open resources will pave the way for large-scale initiatives in psychosis and mental health research, supporting international collaborations and new analytical perspectives.
Totzek et al. (Thu,) studied this question.