Send to

Choose Destination
J Child Psychol Psychiatry. 2019 Oct;60(10):1065-1075. doi: 10.1111/jcpp.13102. Epub 2019 Aug 8.

Peer-adult network structure and suicide attempts in 38 high schools: implications for network-informed suicide prevention.

Author information

Department of Psychiatry, School of Medicine and Dentistry, University of Rochester, Rochester, NY, USA.
Department of Preventive Medicine, Keck School of Medicine, University of Southern California, Los Angeles, CA, USA.
Department of Public Health Education, University of North Carolina at Greensboro, Greensboro, NC, USA.
Departments of Psychiatry and Epidemiology, Columbia University Medical Center, New York, NY, USA.
Sources of Strength, Inc., Bismarck, ND, USA.
Department of Psychiatry and Behavioral Sciences, Feinberg School of Medicine, Northwestern University, Chicago, IL, USA.



Strengthening social integration could prevent suicidal behavior. However, minimal research has examined social integration through relationship network structure. To address this important gap, we tested whether structural characteristics of school networks predict school rates of ideation and attempts.


In 38 US high schools, 10,291 students nominated close friends and trusted adults to construct social networks. We used mixed-effects logistic regression models to test individual student networks and likelihood of suicidal ideation (SI) and suicide attempts (SA); and linear regression models to estimate associations between school network characteristics and school rates of SI, SA, and SA among all with ideation.


Lower peer network integration and cohesion increased likelihood of SI and SA across individual and school-level models. Two factors increased SA: student isolation from adults and suicidal students' popularity and clustering. A multivariable model identified higher SA in schools where youth-adult relationships were concentrated in fewer students (B = 4.95 [1.46, 8.44]) and suicidal students had higher relative popularity versus nonsuicidal peers (B = 0.93 [0.10, 1.77]). Schools had lower SA rates when more students named the same trusted adults named by friends and many students named the same trusted adults. When adjusting for depression, violence victimization and bullying, estimates for adult network characteristics were substantially unchanged whereas some peer effects decreased.


Schoolwide peer and youth-adult relationship patterns influence SA rates beyond individual student connections. Network characteristics associated with suicide attempts map onto three theory-informed domains: social integration versus thwarted relational needs, group cohesion, and suicidal students' social influence. Network interventions addressing these processes, such as maximizing youth-adult connections schoolwide and heightening influence of youth with healthy coping, could create more protective schools. Longitudinal and intervention studies are needed to determine how schools differentiate in network structure and clarify reciprocal dynamics between network characteristics and suicidal behavior.


Suicide prevention; network interventions; social integration; social networks

[Available on 2020-10-01]

Supplemental Content

Full text links

Icon for Wiley Icon for PubMed Central
Loading ...
Support Center