# Spillover effect of mental disorders in adolescent peer networks on likelihood of dropping out of secondary school

## TL;DR

Having classmates with mental disorders increases the chance of dropping out of secondary school, even after accounting for one's own mental health.

## Contribution

This study identifies direct spillover effects of classmates' mental disorders on dropout likelihood, separate from the individual's own diagnosis.

## Key findings

- For academic students, having three or more classmates with mental disorders increases dropout likelihood by 16%.
- Vocational students with three or more such classmates face a 24% higher dropout risk.
- The direct spillover effect was larger than the indirect effect mediated by the individual's own mental disorder diagnosis.

## Abstract

Spillover effects of mental disorders on family members are well-documented, but their impact on adolescent peers remains unclear. We investigated whether having classmates with a diagnosed mental disorder in the ninth grade of lower secondary school (ages 15–16) was associated with greater likelihood of dropping out of upper secondary education (ages 16–19). We used registry data on Finnish people born between Jan 1, 1985, and Dec 31, 1997. After lower secondary school, 378,453 cohort members started general (academic) upper secondary school and 284,713 started vocational upper secondary school. Using causal mediation analysis, we disentangled the total effect of having ninth-grade classmates with a diagnosed mental disorder on upper secondary dropout into direct spillover effect and indirect effect mediated by the individual’s own mental disorder diagnosis received during upper secondary education. For academic students, having one ninth-grade classmate diagnosed with a mental disorder was associated with 6% greater likelihood of dropping out of upper secondary education (95% CI 3–9%); two diagnosed classmates, with 10% greater likelihood (6–14%); and three or more diagnosed classmates, with 16% greater likelihood (10–22%). For vocational students, the respective increases in likelihood were 5% (2–8%), 11% (6–15%), and 24% (18–32%). For both academic and vocational students, the indirect effect of dropping out mediated by one’s own mental disorder diagnosis was notably smaller than the direct effect attributable to having diagnosed ninth-grade classmates. Our results suggest that mental disorders may have spillover effects on the educational attainment of socially connected adolescent peers.

The online version contains supplementary material available at 10.1007/s00787-025-02723-8.