Online health communities (OHCs) have already become essential medium for people to obtain medical knowledge, share experiences and emotions. OHC users are able to post user-generated content (UGC) to interact with each other. However, the large amount of UGC may lead to personal information even privacy disclosed online. Although such disclosure may help users to trade some social support, which is the basis of sustaining a successful OHC, the users should be aware of the risks of leaving such traces online. This study selects a popular online Q & A community “Zhihu” in China as the research target. By collecting all questions and corresponding answers from 4 diabetes sub-communities, we would like to identify online privacy-traces of users from UGC. According to the theory of Communication Privacy Management, we build an explanatory model to understand user behaviors of concealing or revealing private information from the aspects of user characteristics, peer attention, and social support effects.