Hidden pressures: How teen friendships shape self-harm risk

Most of us remember the feeling of being a teenager at school – fitting in, or not, within its social hierarchy. There are always the popular kids, the loners, and the in-betweeners whose friendships cross boundaries. But these hierarchies aren’t just social structures – they can shape mental health too.

In a recent study, my colleagues and I found that school friendship networks relate to one of the most serious health issues among young people today: self-harm.

Self-harm – hurting oneself on purpose – is worryingly common in adolescence. Around 16% to 22% of teenagers report having done it at least once, and rates appear to be rising, particularly among girls. Adolescents who self-harm face greater risks of worsening mental health and even suicide.

There are many reasons for self-harm. Some use it as a coping mechanism to relieve emotional pain (intra-personal), while others may be influenced by peers or use it to communicate distress (inter-personal). Since peer relationships are central during adolescence, understanding how social networks influence this behaviour is crucial.

Little research, however, has explored how school-based friendship networks connect to self-harm. To address this, we analysed data from the Resilience, Ethnicity and Adolescent Mental Health (Reach) study – the UK’s largest and most recent adolescent mental health cohort, following about 4,000 young people from 12 secondary schools in inner-city south London.

We focused on the first year of data, when participants were aged 11 to 14. Fourteen per cent reported having ever tried to harm or hurt themselves. Each student was also asked to name their school-year friends. This is the first UK cohort to collect data on both friendship networks and self-harm.

Using social network analysis, we mapped students’ friendships and calculated their “social positions”. This included popularity (how many peers named them as a friend), bridging (connecting otherwise separate friendship groups), and social isolation (having one or no friends). We also examined how many of their friends had reported self-harm, to see how these factors related.

We found that who adolescents are friends with – and how they’re connected – mattered. Nearly half had at least one friend who self-harmed, and those with such friends were more likely to report self-harm themselves, suggesting peer influence.

Certain social positions were linked to higher or lower risk. Social isolation clearly increased risk, but so did popularity and bridging to some extent. Popularity may bring social pressure, while bridging might reflect the stress of being between groups.

Protective positions included being sociable (nominating many friends) and belonging to a close-knit friendship group. However, the strongest associations were for isolation and having friends who self-harm.

We expected stronger effects among girls but found little gender difference, suggesting that these dynamics operate similarly for boys and girls in diverse, inner-city schools.

Our findings are correlational, not causal. We can’t yet tell whether social networks predict self-harm or the other way around. But they do highlight that self-harm cannot be understood in isolation. Both disconnection (isolation) and connection (friendship with those who self-harm) may play a role – meaning schools need multi-layered prevention approaches that address wider peer dynamics, not just individuals.

Peer relationships shape who we are – and understanding these networks may be key to improving young people’s mental health for years to come.

The Conversation