[Construction and analysis of the online social network topology among men who have sex with men in Guangzhou].

Objectives: To construct and analyze the social network of men who have sex with men (MSM) in Guangzhou on a geosocial networking platform, aiming to reveal its topological characteristics, connection patterns, and group stratification. The study aims to identify structural bottlenecks and key nodes in information dissemination, thereby supporting targeted public health interventions. Methods: User data were collected via web scraping from a major MSM-oriented social platform. A directed network was constructed with users as nodes and follow relationships as edges. Network topology was assessed using metrics such as degree distribution, clustering coefficient, reciprocity, assortativity, centrality, and community structure. Stratified comparisons and homophily analyses were conducted based on age, sexual role, and number of sexual partners to explore how user attributes relate to structural positioning and interaction patterns. Results: The network included 72 535 users and exhibited a highly sparse (density <0.001), low reciprocity (0.076), weakly clustered (transitivity=0.004), and strongly assortative (0.925) structure, reflecting decentralization and stratification. Younger users (<29 years) showed the highest age-based assortativity (0.369), while users with multiple sexual partners had the strongest behavioral clustering (0.220). The community structure was highly fragmented-70.9% of communities consisted of only two nodes-but overall connectivity was maintained by 12 089 boundary nodes, primarily composed of middle-aged, top-identified, and multi-partner users, who contributed 39.2% of cross-community links. Conclusions: The MSM online network is characterized by structural heterogeneity and hierarchy. High assortativity, low reciprocity, and fragmented communities can hinder the flow of health information and access to support. Future interventions should target high-centrality nodes, improve the visibility of peripheral users, and bridge fragmented clusters. Targeted dissemination based on network structure and strengthening cross-community bridging pathways may help optimize information flow, reduce structural visibility gaps, and build a fairer, more inclusive, and structurally friendly digital intervention environment.