ABSTRACT
Introduction It has been approved that information sources would affect public behaviors. However, due to the outbreak of COVID-19, this influence was enhanced and showed a distinctive pattern among different populations, which has been less noticed before. We aimed to investigate the potential roles of different information sources in COVID-19 preventive behaviors of different publics. Methods A cross-sectional online survey with 11,190 participants from 33 province-level regions in China was conducted during the COVID-19 pandemic. Sociodemographic characteristics, COVID-19 preventive behaviors, and information sources for COVID-19-related information were assessed. A mixed linear model was used to analyze risk factors of COVID-19 preventive behaviors. The effects of different information sources on COVID-19 prevention behaviors of different publics were analyzed. Results Generally, the Chinese public had good COVID-19 preventive behaviors, and the top three COVID-19 preventive behaviors with the higher action rate were avoiding eat bushmeat (76.1%), a healthy diet (74.8%), and avoiding contact with people with symptoms of respiratory diseases (73.0%). About information sources, 12320 telephone (National Public Health Hotline) (−0.62, 95% CI: −0.94 to −0.31) and acquaintances consulting (−1.00, 95% CI: −1.31 to −0.69) were negatively associated with COVID-19 preventive behaviors, while internet resources, family doctors, hospitals, and community health centers were positively associated with COVID-19 preventive behaviors (1.00 vs. 0.47 vs. 0.46 vs. 0.33, P < 0.05). For older adults, accessing to COVID-19-related information through family doctors and community health centers were positively associated with COVID-19 preventive behaviors. For the non-educated, family doctors and community health centers had positive effects on their COVID-19 preventive behaviors. Family doctors and internet resources were positively associated with COVID-19 preventive behaviors among those earning 5,000 yuans and above. The effects of family doctors, hospitals, and internet resources were higher for COVID-19 preventive behaviors of urban publics than for rural publics. Finally, the effect of internet resources on COVID-19 preventive behaviors of females was lower than males. Conclusions Obtaining COVID-19-related information through internet resources had the most significant effect on COVID-19 preventive behaviors, but was not significant among publics with old age, low education, low income, and living in rural area.