Background Adolescence involves increased peer interactions, where peers become a significant source of influence. This study explores mental well-being homogeneity within adolescent peer groups. While prior research found similarities in health outcomes among friends, our focus extends to the broader network meso-level. Building on a previous study's moderate link between mental well-being and peer group membership, we utilize recent data to investigate similarities in various mental health outcomes among adolescents within the same peer group.

Methods In this cross-sectional study, data were gathered in 2020, 2022, and 2023 from five self-selected secondary schools in Scotland's central belt. All Secondary 2 (12–13 years) and 4 (14–15 years) students were invited, resulting in the participation of N = 1407 individuals (49% male, 46% female, and 4% other). The survey instrument included questions about health and school-based friendships. Outcomes of interest were general mental health self-esteem, and body image, assessed using validated scales. Ethical approval for the study was obtained from the University of Glasgow (200190035), and participants provided consent through an online form. A series of multilevel models were employed to assess the association between variation in health outcomes and peer group membership, while controlling for other potentially relevant variables such as school year, gender, ethnicity, family affluence, etc. Peer groups were identified using the Walktrap community detection algorithm, with sensitivity analyses employing alternative community detection algorithms.

Results In the model including only peer group membership, intraclass correlations (ICCs) for general mental health, self-esteem, and body image were 0.05 (CI95%: 0.02, 0.08), 0.06 (CI95%: 0.03, 0.10), and 0.04 (CI95%: 0.02, 0.07), respectively. More complex models revealed that being in a larger peer group correlated with a more positive body image (p = 0.054), while other individual level variables (e.g., gender, school year and parental behaviours) were also found to be significant predictors of the outcomes and their inclusion further reduced ICCs.

Conclusion Our findings indicate that approximately 5% of the total variance in mental health-related outcomes is attributable to differences between peer groups. While the clustering effect is modest, individuals within the same peer group show greater similarity in mental health compared to those in different peer groups. Additionally, our results suggest that certain characteristics of peer groups, such as size, may be associated with individual outcomes. This nuanced understanding of the interplay between peer group membership and structure can inform targeted school-based interventions.