[Purpose/significance] Multi-subject interaction has emerged as a prominent characteristic and develop⁃ment model in online chronic disease communities. Exploring the relational network structure characteristics and be⁃havioral patterns of multi-subject interaction in these communities can provide valuable insights for formulating target⁃ed strategies to promote multi-subject interaction and sustainable development of online chronic disease communities.[Method/process] Based on the analysis of the conceptual characteristics, interaction behavior models and interaction relationship models of multi-subject interaction in online chronic disease communities, a multi-subject interaction re⁃lationship network was constructed using sample data from the“Lymphoma House”. Subsequently, multi-dimensional network structure characteristic indicators were introduced to analyze the resilience, dynamic evolution features, and behavioral patterns of the multi-subject interaction. [Result/conclusion] The multi-subject of online chronic disease community participates in community interaction activities with cross-agent interaction mode, key subject-led interac⁃tion mode, multi-core subject interaction mode and mediator-coordinated interaction mode. Under these interaction models, a sparse relational network with relatively weak communication interactivity, low resilience, and dynamically evolving network structures has been established.