Network Models of Psychopathology: Exogenous and Endogenous Vulnerability and the Integration of Individual Symptom Profiles (NetPsy)
Contact: PD Dr. Markus Moessner (markus.moessner(at)med.uni-heidelberg.de)
Project period: 2022-2024
In network models of psychopathology psychological functioning is modelled as a complex system. The patterns of mutual relations between symptoms and functional domains define the system’s vulnerability for psychological disorders. Despite increasing popularity since their introduction to psychopathology research approx. 10 years ago, current applications suffer from some severs shortcomings: - A formal model of psychopathology is still missing – Individual symptom profiles are not taken into account although the majority of applications aim at identifying targets for interventions. – No methods are available to integrate additional levels of data like resilience or biological markers. This project will address these shortcomings of network analysis in psychopathology research by adjusting and introducing network analytic methods, and applying them to large clinical datasets: Based on the robustness a formal model of psychopathology will be proposed. Stress tests and reverse stress tests will be conducted to assess exogenous (a system’s vulnerability to external shocks) and endogenous (a system destabilizes itself due to its internal structures) vulnerability in symptom networks, and quantify the network’s nodes and edges contributions. Individual symptom profiles will be included in symptom networks as knot weights to improve the identification of personalized targets for interventions. Finally, in order to include additional levels of data in psychopathology networks, methods to model and analyse bipartite graphs will be adjusted for applications in psychopathology networks. The project will increase our understanding of the structures and processes that constitute robustness in psychopathology networks, and introduce network analytic methods that will advance psychopathology research significantly.