Abstract: Modification of risky health behaviors to improve brain health and early detection of the disease are pivotal components to alleviate the societal burden of Alzheimer's Disease and Related Dementias (ADRD). Both these components are heavily influenced by individuals' in-person and online social ties. Social media has become predominant as a source of information for many individuals as they seek health-related information and care. However, our understanding of the ways through which we can effectively harness these social mechanisms to support individuals' care-seeking behaviors and behavior change efforts is quite limited. In the proposed research, we will seek to develop a methodological framework that enables us to (a) examine individual- and group-level dynamics associated with information-seeking and care-seeking behaviors in in- person and digital contexts for risk reduction and early detection of ADRD, and (b) model population-level network diffusion characteristics and their relationship with individuals' behavioral state/stage, resistance to misinformation, manifestation of sociobehavioral processes. Grounded in theories of human behavior, cognitive psychology, and network sciences, we will employ mixed-methods approach to analyze social dynamics using data collected through surveys and interviews in in-person settings and ADRD-related peer interactions in online social media platforms. Specifically, we will employ theories (e.g., Integrated Behaviora Model, Transtheoretical Model of Change, Case seeking delay model) and methods that facilitate granular (e.g., grounded theory analysis), high-throughput (e.g., natural language processing, deep learning), and collective (e.g., ego-centered networks, affiliation networks) modeling of information seeking and care seeking behaviors as self-reported in surveys and as manifested in online social media. While generally applicable, the methods will be developed based on communications among individuals engaging in social intercourse related to risk reduction and early detection of ADRD. This research proposal will result in 1) novel methods to incorporate behavioral/care seeking stages and processes into network models of social influence and 2) new proposals for the development of network interventions that harness social ties and behavioral constructs to support individuals engaging in modification of risky health behaviors and care seeking behaviors within and beyond ADRD domain.