Obesity and physical inactivity during adolescence are associated with an increased risk of mental and physical health outcomes including heart disease, and depression. There are widely accepted benefits of intervening to address excess adiposity and insufficient physical activity. It is important to address the settings that are most pertinent in shaping these. Given adolescents’ social reorienting towards friends and away from parents, the social environment, and in particular, the school, and digital social environment are key focusses of this thesis. This thesis is organised into two linked parts.

In the first part of this thesis, I focus on the school social environment. I used data from GoActive, a large-scale physical activity intervention of adolescents living in the East of England (n=1,765). I explored associations between the school policy, social and physical environment and change in adolescent accelerometer-assessed physical activity. I tested many potential predictors of change as a hypotheses-generating exercise (Chapter 3). I found that friendship support for physical activity was predictive of change in physical activity, and that sex and socioeconomic status modified these relationships.

In the second part of this thesis, I explored the association between social media use and adiposity using data from the Millennium Cohort Study, a large population-based cohort of young people in the UK. I first investigated these associations cross-sectionally, showing that social media use and measured BMI z-score at age 14 years was associated in girls at the ≥5 hours/day level (vs. 0 to <1 hours/day, no associations for boys) (n=10,798). This association was partially explained by sleep duration, depressive symptoms, body weight satisfaction, and wellbeing (Chapter 5). I then explored the prospective association between social media use at age 14 years and change in BMI z-score from age 14 to 17 years in boys and girls (n=8,024). I showed that greater social media use was associated with a lower change in BMI z-score at the 3 to <5 hours/day range (vs. 0 to <1) for boys, and at the 1 to <3 hours/day range for girls (Chapter 6).

Overall, the results from my thesis suggests two key findings. First, friendship support for physical activity may be particularly important for increasing physical activity during adolescence. Second, although a negative association was found between social media use and change in BMI z-score at the population level, this finding must be interpreted cautiously given that associations were small, and are not consistent with the few studies completed to date which used self-reported BMI. However, associations differed by sex and there may be specific sub-populations at greater risk (e.g. extreme users, adolescents living with mental health challenges). The overall findings of this thesis highlight a need for further research to understand the potential influence of friendship support for physical activity, and to better understand and harness adolescents’ high engagement with social media use for combatting adolescent obesity.