This longitudinal secondary analysis examined whether internalising and externalising symptoms converged or diverged between adolescents (mean age 14.89) and their best friend(s) over one year, and whether gender or victimisation moderated these patterns, using two waves of CILS4EU data from Sweden and Germany (n=5,705).
On average, adolescents **diverged** from friends in internalising symptoms over time. For externalising symptoms, only girls (not boys) showed significant **convergence** with their friendship group. Victimisation predicted higher symptoms and greater peer dissimilarity but did not amplify longitudinal convergence. Friendship stability was the key driver of convergence in exploratory LASSO analyses.
- Two-wave design limits ability to separate social selection from social influence and to capture dynamic change. - Non-clinical composite symptom measures (3–11 items) may lack sensitivity to detect subtle symptom changes. - Complete-case analysis required for network outcomes may introduce selection bias and reduce generalisability.
Peer friendship networks cluster adolescent mental health symptoms, so school-based screening and interventions should consider the peer group, not just the individual. Girls may be a specific target for externalising symptom peer-influence programs, and stable friendships appear to amplify this transmission.
Explore related topics
