Mediation Effect of Suicide-Related Social Media Use Behaviors on the Association Between Suicidal Ideation and Suicide Attempt: Cross-Sectional Questionnaire Study

Xingyun Liu, Jiasheng Huang, Nancy Xiaonan Yu, Qing Li, Tingshao ZhuJournal of Medical Internet Research 2020 · DOI

Abstract: Our findings thus support the social cognitive theory, and there are implications for population-based suicide prevention that can be achieved by identifying behavioral signals.

Introduction

The "ideation-to-action" gap remains a significant hurdle in suicide prevention. While many individuals experience suicidal thoughts (ideation), only a fraction transition to suicide attempts. Identifying those at highest risk is notoriously difficult in traditional clinical settings. This study investigates how specific social media behaviors—consuming, reposting, or discussing suicide-related content—act as a bridge in this transition. By viewing social media as an interactive environment through the lens of Social Cognitive Theory (SCT), researchers aimed to identify behavioral markers that signal an increased risk of moving from thought to action.

Methodology

The study framework is based on SCT, which posits that behavior is acquired through four processes: attention, retention, production, and motivation. These were mapped to three measurable social media behaviors:

  1. Attending to: Consuming suicide-related news or posts.
  2. Commenting-reposting: Engaging with content by sharing or commenting.
  3. Talking about: Publicly discussing one's own suicide plans or joining online suicide communities.

The researchers analyzed a dataset of 569 Chinese social media users from Sina Weibo, recruited from the comment section of a deceased blogger's suicide note—a hub for individuals struggling with depression. Participants completed the Adult Suicidal Ideation Questionnaire (ASIQ) and a 6-item scale for suicide-related social media behaviors.

Statistical methodology for testing the mediation hypothesis (H2)

To test the mediation effects, the team employed a multi-step statistical approach. As illustrated in the figure above, bivariate correlations were used to test the premise of mediation, followed by Structural Equation Modeling (SEM) to estimate the effect. Finally, the Monte Carlo method was used to assess the indirect effects, ensuring the robustness of the identified "mediational chain."

Results

The study found that the relationship between suicidal ideation and suicide attempts was fully mediated by a chain of social media behaviors: Ideation → Attending → Commenting/Reposting → Talking About → Attempt.

Key findings include:

  • Attempters vs. Ideators: Suicide attempters showed significantly higher levels of suicidal ideation and engaged in more suicide-related social media behaviors across all categories compared to those who only had thoughts.
  • The "Upward Spiral": Social media acts as an accelerant; "attending to" suicide information can lead to deeper processing, normalizing the behavior and potentially increasing "suicide capability" (habituation to pain and fear).
  • Critical Signals: "Talking about" behavior—representing the "production" phase of SCT—was the strongest predictor of an actual attempt. When a user moves from consuming content to discussing their own plans, the risk level spikes significantly.

Discussion and Conclusion

These findings suggest that social media behaviors are not merely reflections of distress but are active components in the progression toward suicide. Identifying these behavioral signals offers a pathway for population-based suicide prevention.

While the study is limited by its cross-sectional design and reliance on self-reporting, it paves the way for future "Big Data" approaches. Using automated logs and machine learning to identify these markers in real-time could enable proactive, automated intervention systems.

Summary updated

Abstract

Our findings thus support the social cognitive theory, and there are implications for population-based suicide prevention that can be achieved by identifying behavioral signals.
BibTeX
@article{liu2020mediation,
  author = {Xingyun Liu and Jiasheng Huang and Nancy Xiaonan Yu and Qing Li and Tingshao Zhu},
  title = {{Mediation Effect of Suicide-Related Social Media Use Behaviors on the Association Between Suicidal Ideation and Suicide Attempt: Cross-Sectional Questionnaire Study}},
  journal = {Journal of Medical Internet Research},
  year = {2020},
  doi = {10.2196/14940},
}
Figure