The environment and eating disorders: developing novel measures and hypotheses through inter-disciplinary collaborations.

Eating disorders are severe mental health disorders that tend to start in childhood and adolescence. Around to 6% of women and 2% of men experience anorexia nervosa, bulimia nervosa, binge eating disorder, or other specified feeding and eating disorders. Up to one in five young people have some of their symptoms. It is not clear how many people have avoidant/restrictive food intake disorder (ARFID), Pica, and rumination disorder, which have been more recently included as feeding and eating disorder diagnoses. Overall, we still know little about who has eating disorders in the population and what causes these conditions.

Work package 1. Research on eating disorders is limited. Few large studies have collected eating disorders data. In some cases, this is because we do not have adequate measures. Our first goal is to improve availability of eating disorder data to facilitate future research. We will:
1. include a short eating disorder questionnaire in the Age of Wonder Cohort, a large general population study of 30,000 adolescents.
2. Create a short questionnaire for ARFID, Pica, and rumination disorder.
3. Link population studies to clinical data in order to better understand who is or is not diagnosed with an eating disorder and what are long-term health outcomes of people who are not diagnosed.

Work package 2. More young people experience eating disorders now than 20 years ago. Societal changes that have happened during this period could have led to this increase. There is evidence that aspects of urban environments - such as lack of parks, higher pollution, deprivation, fast-food availability - are associated with worse mental and physical health in young people. This might be because of increased problems with sleep, fewer opportunities for socialisation and physical activity, and weight gain. Sleep disturbances, loneliness, and weight gain are also seen in people with eating disorders, but there is no robust evidence on whether they are a cause of eating disorders which, if addressed, could help prevent them. Similarly, social media are thought to have led to an increase in eating disorders in young people, but there is no strong evidence of this. Understanding if and which content and type of social media engagement might cause eating disorder symptoms could help identify adolescents at risk and put in place measures to reduce this risk. Our second goal is to develop new approaches to study these societal risk factors. We will create a network of young people and researchers with different backgrounds to develop research hypotheses and pilot new ways of collecting data that will help researchers study these questions in the future. We will:
4. Collect information on location, sleep, physical activity, stress, mood, appetite on 250 adolescents in the Age of Wonder cohort using smartwatches.
5. Explore different ways to collect social media data from these adolescents
6. Understand how we can identify adolescents at risk of eating disorders using data from a large social media platform, TalkLife.

Work package 3. Young people spend much of their day in schools, making them important environments to study in relation to eating disorders. We know that some schools have higher rates of eating disorders, so a better understanding of which aspects of school culture might cause eating disorders could help developing preventative interventions. Our third goal is to create a network of researchers with expertise in different aspects of young people's mental health and school environments to work alongside young people, parents, and teachers to:
7. Generate a list of potential school-level risk factors for eating disorders and identify measures and datasets that allow investigating these questions.
This information will help future researchers testing hypotheses and conduct new research in this important area.

Eating disorders(ED) are severe mental health disorders; 6% of women and 2% of men experience anorexia nervosa, bulimia nervosa, binge eating disorder or other specified feeding and eating disorders. Less is known on the prevalence of Avoidant/Restrictive Food Intake Disorder (ARFID), Pica, and rumination disorder.

The incidence of ED has increased in the past twenty years, which is consistent with aetiological influences of environmental risk factors. Environments are modifiable and thus promising targets for preventative interventions but aetiological ED research in the context of ED is rare.

We will form inter-disciplinary networks of researchers, young people with lived experience, carers, and teachers to address these research gaps. We have designed three work packages (WP) to increase availability of ED data, pilot new ways of collecting and analysing environmental risk factors data, and develop a framework for understanding school-level risk factors for ED.

In WP1, we will:
1. include the EDE-QS in the Age of Wonder cohort (n=30,000) in adolescence.
2. develop and validate a short ARFID, Pica, and rumination disorder measure.
3. leverage population cohort-EHRs linkages to develop data analysis pipelines and open-access clinical codes identifying missed ED diagnoses.

In WP2, we will:
4. collect spatial, activity, and mood data on 250 Age of Wonder participants to pilot research on short- and long-term aetiological processes underlying ED risk.
5. investigate the feasibility of collecting and analysing social media engagement data on 250 Age of Wonder participants
6. explore the feasibility of developing algorithms to identify short- and long-term changes in eating disorder symptoms using large scale longitudinal social media language data.

In WP3, we will:
7. develop a consensus on putative modifiable school-level risk factors for eating disorders and identify existing measures and datasets to study these associations.