Background Eating disorders are psychiatric disorders defined by extreme and life-threatening behaviours surrounding weight and food intake. Every year, eating disorders affect 2.4m people in the UK, costing £12bn, and 28.8m in the US, costing $65b (Beat, 2015; Deloitte Access Economics, 2020; The Hearts Minds and Genes Coalition for Eating Disorders, 2021). The problem is growing: hospitalisations for eating disorders are up 84% and over half of female university students report some symptoms of an eating disorder. Symptoms are very disruptive and eating disorders are associated with significant and persistent impairments in quality of life. Medical costs for those with eating disorders are up to 7.2x higher than those without behavioural health conditions (Beat, 2015).
Current treatments for eating disorders are moderately effective at best, and, on average, only 1 in 4 individuals with eating disorders access specialist help for their condition (Ali et al., 2017; Hart et al., 2011). While women represent the majority of sufferers, eating disorders affect people of any weight, age, gender, socioeconomic background and ethnicity (Beat, 2015). Yet, less than a third of published papers reflect that diversity, fuelling the stereotype of the typical sufferer being an underweight, White young woman.
The sector is ripe for innovation, especially today when technological tools can serve clinical needs, and the vast majority of the UK \& US population has a smartphone in their pocket. Juniver Limited collaborated with leading clinicians and hundreds of sufferers including underserved groups to establish what could be done better. We found that all eating disorders are characterised by urges, and that the moment of experiencing an urge is when help is most needed and impactful. Overcoming an urge paves the way to recovery by decreasing the intensity and frequency of subsequent urges - this is how most smokers quit smoking, for example. It's particularly difficult for eating disorder sufferers to overcome urges unsupported, as they have to eat for sustenance and are triggered multiple times every day. We developed a solution rooted in the science of addiction that delivers micro-interventions for on-demand help in compulsive moments (Baumel, Fleming \& Schueller, 2020; Johnson et al., 2022; Juarascio et al., 2018).
A pilot with 97 users validated demand and engagement, with strong signals that our solution works (92% of those who tried our main tool successfully overcame an urge). Additionally, we conducted direct research with 500 participants. We uncovered the following:
* Sufferers are desperate to get better: 97% have sought treatment before, in most (92%) cases, unsuccessfully.
* When asked why they haven't yet recovered, 73% 'haven't found something that's worked' or 'don't know how to recover.'
* The top 3 barriers to access are cost, stigma, and the complexity of navigating specialist care systems.
The Juniver program was developed to directly address these findings. This clinical trial will help advance understanding of how Juniver works and for whom, allow for research with a more diverse population of individuals than those presenting to clinical services, and inform the ways self-guided, on-demand help can contribute to better, more equitable outcomes.
Clinical trial The study will include a sample of 300 participants aged 16 years and over, with participants randomly allocated at a 1:1 ratio to the immediate Juniver group (n=150) or wait-list control group (n=150). The total sample size of 300 participants was guided by statistical power calculations performed using G\*Power 3 software. This factors in a drop-off rate of 30%, estimated from similar studies involving digital interventions for eating disorders and other mental health disorders. Participants assigned to the Juniver group will gain access to the app starting on their respective Week 1. In contrast, participants in the control group will receive access to the app at the conclusion of their respective Week 12. The Juniver group will complete outcome measures at Week 4, 8, 12, 16, 20 and 24 while the control group will complete measures at Week 4, 8 and 12. Statistical analysis will involve linear mixed models to compare differences in outcome measures in the Juniver group compared to the control group, including any overall between-group differences and any differences in how variables change over time.
