Murmuring of Alzheimer Caring: Analyzing Social Dynamics in Online Alzheimer’s Disease and Related Dementias Communities

Abstract

This dissertation investigates the social dynamics in online Alzheimer’s disease and related dementias (ADRD) communities. As more caregivers turn to online forums to seek support and share experiences, understanding the patterns of interaction and the roles of different contributors becomes increasingly important. Across five aims, this work examines how topic initiators catalyze engagement, how replies affect community participation, how the emotional and linguistic characteristics of posts influence response behaviors, and how AI-augmented models compare with human caregivers and clinicians in providing support. Methodologically, this study integrates social computing, natural language processing, and causal inference techniques. It introduces novel metrics to assess community dynamics, evaluates the linguistic and emotional dimensions of caregiving discourse, and tests retrieval-augmented generative models within this context. Results highlight the importance of reciprocal interaction, emotional expression, and readability in sustaining community engagement. Moreover, the findings inform the design of hybrid human-AI systems that may supplement caregiving resources in an emotionally intelligent and context-aware manner. Together, these findings contribute to our understanding of digital health communities and offer insights for future interventions that bridge data-driven models and compassionate caregiving.