Ripples of Emotion: Unraveling the Dynamics of Emotion Flows and User Engagement in Online Healthcare Communities

The sharing of emotions within online healthcare communities (OHCs) is central to peer support, yet little is known about whether emotional encounters continue to shape users’ behaviors beyond the original interaction. While prior research has primarily examined immediate emotional responses within posts or discussion threads, this study investigates how emotional influence unfolds over time across a community. Drawing on a resource-based perspective, we examine two complementary mechanisms of community-level emotion flow: Emotional Resonance Cascades (ERC), whereby accumulated emotions directed at focal users in prior interactions shape their subsequent engagement and emotional expressions toward different users in later discussions, and Emotional Feedback Exchange (EFE), whereby focal users’ own emotional expressions shape the responses they subsequently receive from other community members across the network. Using relational event modeling of 65,740 reply interactions from an online HIV community, we find support for both mechanisms. Receiving positive (negative) sentiment increases (decreases) users’ reply activity and, conditional on replying, the likelihood of expressing positive sentiment in later interactions. At the same time, expressing positive (negative) sentiment increases (decreases) the likelihood of receiving replies and of those replies expressing positive sentiment. These dynamics extend beyond topics, threads, and local social clusters, and are amplified for feeling-oriented messages. A complementary experiment provides evidence consistent with a resource-based explanation by showing that receiving emotional messages alters individuals' mental resources, while observing emotional exchanges shapes perceptions of others' resourcefulness, which in turn influence individuals' behavioral tendencies toward them. Together, these findings suggest that emotional influence in OHCs extends beyond immediate emotional mirroring and provide insights into sustaining supportive digital health communities.