Predictive language processing in first episode psychosis – the link between positive and negative symptoms?

• Applicant: Dr. Franziska Knolle
• Subject Area: Biological Psychology and Cognitive Neuroscience ; Human Cognitive and Systems Neuroscience
• Term: since 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 529207886

Schizophrenia is a highly debilitating disorder. Despite many research-advances the link between the pathophysiology and the occurrence of the complex set of positive (e.g. hallucinations, delusions), negative (e.g., social withdrawal) and cognitive symptoms (e.g. memory or language difficulties) is still poorly understood. The recent hierarchical Bayesian predictive coding account of psychosis provides a convincing theory for the explanation of positive symptoms, and to some extend also negative symptoms. In predictive coding, general brain processing is described as an inferential process in which prior beliefs are used to infer states of the world. In psychosis, however, it has been suggested that there is an imbalance between the precision of prior beliefs, which guide predictions about future events, and the precision of the incoming sensory data, leading to an abnormal weighting of prediction errors. The overweighting of prior information has been associated with positive symptoms (e.g. hallucinations), while the abnormally weighted prediction error has been linked to negative symptoms (e.g. lack of motivation). Using a novel language paradigm, this project will investigate for the first time whether an increased ratio of prior precision and the precision of the sensory likelihood during predictive language processing is present in acute (i.e. mainly positive symptoms) and remitted (i.e. mainly negative symptoms) psychosis, and may be linked to positive and negative symptoms. This project is focussing on predictive language processing, because, first, alterations in language processing are common cognitive symptoms in psychosis; second, language processing itself can be described using the predictive coding framework; and third, language processing may be linked to positive and negative symptoms. In the positive symptoms, language is often the medium of auditory hallucinations, whereas in negative symptoms disrupted language processing may lead to unsuccessful social interaction, and, ultimately, symptoms like social withdrawal. In a longitudinal design this project combines behavioural computational modelling analyses, which allow a Bayesian approximation of prior precision relative to the precision of the sensory information, and electroencephalography, to explore the temporal and neural signature of these processes. Thus, the overall aim of the project is to understand (1) the relationship of prior precision and the sensory likelihood in an acute and remitted state of early psychosis, and the underlying electrophysiology; (2) how a potential imbalance changes from an acute disease state to the state of remission; and (3) how altered predictive language processing links to positive and negative symptoms. This project will therefore increase the understanding of the pathophysiology of the complex set of symptoms in psychosis, and ultimately contribute to the development of better diagnostic tools and treatment options.

• DFG Programme: Research Grants