Digital Health
AI-mediated health information seeking in a university-centered Chinese sample: Information overload, health anxiety, and conditional acceptance of AI-Assisted health communication support
By Huafeng Qu, Xiacheng Song, Min Dai and 3 others.
Abstract
Background Consumer-facing AI tools are becoming part of digital health information seeking, but their potential benefits and information-safety risks remain insufficiently understood. Objective This exploratory cross-sectional study examined AI-mediated health information seeking, information overload, uncertainty-related searching, health anxiety, digital health self-management, and conditional acceptance of AI-assisted health communication support in a university-centered Chinese sample. Methods An online survey of 364 participants assessed AI health information use, queried health topics, preferred AI functions, boundary expectations, privacy control, and related psychological and self-management outcomes. Analyses included descriptive statistics, reliability and measurement-validity checks, correlations, HC3 robust regression models, clean-sample sensitivity analyses, and exploratory network analysis. Results Health anxiety was associated primarily with information overload and uncertainty-related searching, rather than AI use itself in adjusted models. Digital health self-management was positively associated with AI use and boundary expectations. Conditional acceptance of AI-assisted health communication support was associated with AI use, boundary expectations, and perceived privacy control. Sensitivity analyses showed substantively similar patterns after excluding potentially low-quality responses. Conclusions In this university-centered, self-report sample, AI-mediated health information seeking appeared to involve both support-oriented and information-safety-related patterns. The findings should be interpreted as preliminary cross-sectional associations. Safe and acceptable AI health information support may depend on reducing overload, managing uncertainty, protecting privacy, and maintaining clear boundaries between AI assistance and professional care.



