The BrainWaves study of adolescent wellbeing and mental health: Methods development and pilot data

Adolescent mental health and wellbeing are of growing concern globally with increased incidence of mental health disorders in young people. BrainWaves provides a framework for relevant and diverse research programmes into adolescent mental health and wellbeing that can translate into practice and policy. The research programme is a partnership with schools centred on establishing a large (n > 50,000) cohort and trials platform. Reported here is the BrainWaves cohort pilot study. This was designed as proof-of-concept for our recruitment and data capture pipelines, and for cost-modelling. A network of research schools was recruited and a computer-driven questionnaire administered. The eligible population was 16 + year olds who were attending the research schools. Of 41 research schools, 36 (88%) participated over one three-week and one four-week data collection period. From an eligible population of 33,531 young people, 16,010 (48%) attended the study lesson and created an account. Of the 16,010 (100%) who created an account, 15,444 (96%) consented to participate, 9,321 (60%) consented to linkage of research data with educational records, and 6,069 (39%) consented to linkage of research with school/college attendance data. Participants were aged 16-19 years, 59% female, and 76% White. Higher levels of anxiety and depression were found in females than males. Higher levels of media-based social networking were found in females, whereas higher levels of media-based gaming were found in males. Females were more likely to report insufficient sleep whilst males were more likely to report high levels of exercise. This study confirmed an ability to recruit at pace and scale. Whilst the response-rate does not indicate a representative sample, the demographics describe an inclusive and diverse sample. Data collected confirmed findings from previous studies indicating that the electronic data collection methods did not materially bias the findings. Initial cost-modelling suggests these data were collected for around £20 per participant.