Objective  To analyze the dissemination patterns and textual features of health science popularization rumors, establishing a foundation for the development of rumor mitigation and control mechanisms.

Methods A content analysis was conducted on a sample of 348 health-related rumors published in August 2024 by People's Daily Online, China Science Communication, and the China Internet Joint Rumor Debunking Platform. Rumor themes, emotional valence, involved stakeholders, and linguistic tone were systematically coded and statistically evaluated by chi-square tests to assess distributional differences across platforms.

Results Predominant themes included healthy lifestyles (35.34%), disease prevention and control (34.49%), and food safety issues (30.17%), with no significant variation in theme distribution among platforms (P > 0.05). Negative emotional contents accounted for 55.75%, exhibiting significant inter-platform differences (P < 0.05). The primary stakeholders involved were the general public (62.64%) and patients (22.99%), with no significant inter-platform differences (P > 0.05). The dominant linguistic structures were declarative sentences (41.95%) and interrogative sentences (33.33%), with notable differences in distribution across platforms (P < 0.05).

Conclusions Health science popularization rumors demonstrate strong links to public health concerns, notable emotional valence, targeted dissemination strategies, and diverse linguistic approaches. Strengthening surveillance, early warning systems, and cross-platform collaboration is essential for enhancing public health literacy and preventing rumor proliferation at the source.