# Examination of self-harm clustering in adolescent peer networks: a nationwide registry cohort study in Finland

## TL;DR

This study finds that adolescents are more likely to self-harm if their peers have, suggesting a social influence effect, especially around age 16.

## Contribution

The study provides population-level evidence of self-harm clustering in adolescent peer networks using nationwide Finnish registry data.

## Key findings

- Having schoolmates who self-harmed is associated with a small increased risk of subsequent self-harm.
- The highest risk occurs around age 16, suggesting peer influence is strongest during life transitions.
- Exposure to peer self-harm in eighth or ninth grade shows similar elevated risk around age 16.

## Abstract

Clusters of self-harming behaviour among adolescents have been observed, yet population-based epidemiological evidence is lacking. This study aims to address this lack by examining the clustering of self-harming behaviour within adolescent peer networks at the population level.

We used nationwide registry data on Finnish people born between January 1, 1985, and December 31, 2000, to examine whether having same-grade schoolmates who had self-harmed was associated with greater subsequent self-harm risk. Cohort members were followed up until first recorded self-harm episode, emigration, death, or December 31, 2020, whichever came first. Hazard ratios (HRs) were estimated using mixed-effects Cox proportional hazards models adjusted for a comprehensive set of individual-, parental-, school-, and area-level covariates.

The cohort comprised 913,149 Finnish residents. Having same-grade schoolmates who had self-harmed between school-starting age and finishing ninth grade was associated with a higher, albeit small in magnitude, HR of subsequent self-harm over a median of 11.6 years of follow-up (HR 1.05, [95% CI 1.01–1.09]). HR was not consistently higher over follow-up time but was highest in the beginning of follow-up when the cohort members were around age 16 (1.45 [1.25–1.69]). Limiting exposure to schoolmates’ self-harm episodes to 1 year consistently showed the highest risk around age 16, regardless of whether the exposure occurred in ninth grade (1.49 [1.21–1.82]) or eighth grade (1.36 [1.07–1.74]), with follow-up commencing after the respective grade.

While we cannot rule out residual confounding, our findings suggest that self-harm may socially transmit within adolescent peer networks. The observed highest risk around age 16 suggests that external stressors associated with transitioning to new life stages at this age may moderate the impact of peer self-harm exposure. Prevention and intervention measures that consider possible peer influences on adolescents’ self-harming behaviour may help reduce the public health burden of self-harm.

European Research Council and 10.13039/501100002341Research Council of Finland.