Suicide ideation of individuals in online social networks

Suicide explains the largest number of death tolls among Japanese adolescents in their twenties and thirties. Suicide is also a major cause of death for adolescents in many other countries. Although social isolation has been implicated to influence the tendency to suicidal behavior, the impact of social isolation on suicide in the context of explicit social networks of individuals is scarcely explored. To address this question, we examined a large data set obtained from a social networking service dominant in Japan. The social network is composed of a set of friendship ties between pairs of users created by mutual endorsement. We carried out the logistic regression to identify users’ characteristics, both related and unrelated to social networks, which contribute to suicide ideation. We defined suicide ideation of a user as the membership to at least one active user-defined community related to suicide. We found that the number of communities to which a user belongs to, the intransitivity (i.e., paucity of triangles including the user), and the fraction of suicidal neighbors in the social network, contributed the most to suicide ideation in this order. Other characteristics including the age and gender contributed little to suicide ideation. We also found qualitatively the same results for depressive symptoms.

💡 Research Summary

The paper investigates how an individual’s position and activity within an online social network relate to suicidal ideation, using a massive dataset from Japan’s most popular social networking service (SNS). The authors define “suicidal ideation” operationally as membership in at least one active, user‑created community whose title contains suicide‑related keywords. This behavioral definition bypasses self‑report bias and allows the analysis of millions of users simultaneously.

Data comprise several million users, their bidirectional “friend” links, and the full list of communities each user has joined. In addition to demographic variables (age, gender), the study extracts a rich set of network‑structural metrics: degree (number of friends), clustering coefficient, transitivity (proportion of closed triangles), and its complement, intransitivity (the scarcity of triangles around a node). The authors also compute the fraction of a user’s neighbors who belong to suicide‑related communities (“suicidal neighbor ratio”). Community‑related variables include the total number of communities a user belongs to and the proportion of those communities that are topic‑specific (e.g., gaming, romance).

A logistic regression model is fitted with suicidal‑community membership as the binary outcome. Variable selection is performed stepwise with cross‑validation to avoid over‑fitting, and multicollinearity diagnostics are reported. The results reveal three dominant predictors, in descending order of effect size: (1) the sheer number of communities a user joins, (2) intransitivity (i.e., a low count of triangles involving the user), and (3) the suicidal neighbor ratio. Users who belong to many communities are more likely to encounter negative or reinforcing content, suggesting that breadth of online affiliation can be a risk factor when not accompanied by supportive ties. High intransitivity indicates that a user’s friends are not themselves friends, reflecting a fragmented local network and a lack of mutual reinforcement; this structural isolation is strongly associated with suicidal ideation. Finally, a higher proportion of suicidal neighbors points to a contagion‑like effect, consistent with prior findings that exposure to suicidal content or peers increases personal risk.

Demographic variables—age and gender—show little explanatory power in this context, implying that within the SNS environment, structural and behavioral factors outweigh traditional demographic predictors. The authors repeat the analysis using “depressive symptoms” defined analogously (membership in depression‑related communities) and obtain qualitatively identical results, reinforcing the notion that suicide and depression share common network‑based risk mechanisms.

The study acknowledges several limitations. First, equating community membership with suicidal ideation does not capture the full spectrum of suicidal thoughts, plans, or attempts. Second, the observational design precludes causal inference; the identified associations could be bidirectional. Third, the findings are specific to a single Japanese SNS and may not generalize to other platforms or offline social structures. Despite these constraints, the research demonstrates the power of large‑scale, behavior‑driven data to uncover risk patterns that are invisible in traditional survey research.

Future directions suggested include longitudinal tracking of network evolution to detect temporal precursors, text mining of community posts to assess sentiment and content exposure, and the development of machine‑learning classifiers that integrate network metrics for early warning systems. The authors argue that public health officials and mental‑health professionals should monitor structural isolation (e.g., low clustering) and the diffusion of suicide‑related content within online networks, and design targeted interventions—such as algorithmic prompts, community moderation, or outreach to users with high intransitivity and suicidal neighbor ratios—to mitigate the risk of suicide among young adults.