Suicide Transmission Dynamics on Social Media: Modeling Risk, Understanding Platform Differences, and Exploring Ethical Implications Among Young Adults - PROJECT SUMMARY Suicide is a leading cause of death among young adults, and rates have been steadily increasing in recent decades, highlighting the urgent need for effective prevention strategies. The rise of social media use among this population has coincided with these increasing suicide rates, raising concerns about its potential role in the spread of suicidal thoughts and behaviors. While evidence suggests a link between social media use and suicide risk, the underlying mechanisms through which social media influences suicide transmission, particularly over the rapid timescales in which suicidal crises unfold, remain poorly understood. This research project aims to investigate the dynamic transmission of suicidal thoughts and behaviors among young adults (18-25 years) on two prominent social media platforms: Twitter/X and Reddit. The study will employ a novel, multi-method approach integrating computational social science, natural language processing (NLP), dynamic social network analysis, and ecological momentary assessment (EMA) to model suicide transmission as a real-time process and identify modifiable risk factors. In Phase 1, we will recruit 800 young adults (400 from Twitter/X and 400 from Reddit) who are already engaging with suicide-related content on these platforms, minimizing the risk of exposing participants to potentially harmful content. Data will be collected longitudinally over 90 days using both active and passive methods. Participants will use a smartphone app to complete EMAs three times daily, capturing real-time fluctuations in their suicidal thoughts and behaviors. At the end of the study period, they will download their complete history of posts, likes, replies, and reposts from both platforms, enabling us to reconstruct their dynamic social networks. We will also conduct a psychiatric interview with each participant to assess suicide risk and gather detailed clinical information. Advanced NLP and dynamic social network analysis will be used to analyze social media content and interaction patterns, while deep learning models will predict changes in suicidal ideation based on both individual and network-level factors. We will compare transmission dynamics between Twitter/X and Reddit to identify platform-specific characteristics that influence the spread of suicide risk. Phase 2 will employ a policy Delphi study with a diverse panel of 60 stakeholders, including young adults with lived experience, digital health researchers, ethicists, mental health clinicians, and representatives from relevant organizations. This study will explore the ethical and practical feasibility of using the insights gained from Phase 1 to inform real-world suicide prevention interventions on these platforms. Through multiple rounds feedback, we aim to achieve consensus on the ethical considerations of leveraging social media data and technology for suicide prevention. This research has the potential to significantly advance our understanding of suicide transmission on social media, paving the way for the development of data-driven, ethically grounded, and scalable interventions that could ultimately contribute to a reduction in suicide risk among young adults.
