Identifying highly influential nodes in the complicated grief network.

Area of Science:

• Psychopathology
• Network Science
• Quantitative Psychology

Background:

• The network approach views mental disorders as interconnected symptom networks.
• Identifying influential symptoms is key to understanding and treating disorders.
• Existing centrality measures may not fully capture symptom influence due to unaddressed negative symptom relationships.

Purpose of the Study:

• To develop and evaluate novel expected influence (EI) indices for psychopathology networks.
• To assess if EI indices better reflect a node's influence compared to traditional centrality measures, especially in networks with negative edges.
• To apply EI indices to a complicated grief (CG) network and identify key symptoms for etiology and treatment.

Main Methods:

• Simulated single-node interventions on random networks with positive and negative edges.
• Developed two EI indices accounting for positive and negative edge weights.
• Analyzed longitudinal bereavement data to assess the association between centrality, EI, and network change in complicated grief.

Main Results:

• In simulations, both centrality and EI correlated with node influence in positive-edge networks.
• EI showed a stronger association with node influence than centrality in networks containing negative edges.
• In the CG network, both centrality and EI correlated with the strength of node-network change associations.

Conclusions:

• Expected influence (EI) indices offer a more nuanced assessment of symptom influence in psychopathology networks, particularly when negative symptom relationships exist.
• High-EI nodes, such as emotional pain and feelings of emptiness, are critical targets for understanding and intervening in complicated grief.
• The findings support the utility of EI in advancing network-based approaches to psychopathology research and clinical practice.