INTERconNEcT-Eds: a Guided Self-help Mobile App to Improve Outcome in Eating Disorders

Despite the disabling nature of eating disorders, many individuals with EDs do not receive appropriate mental health care. Recently, the use of eHealth technology has been proposed as a potentially effective alternative to traditional, inperson treatment delivery for those with EDs. Specifically, self-help programs include the use of materials, such as books, workbooks and videoclips, to empower patients through information and skills to cope with the illness. Guided self-help adds the element of guidance to the use of these materials.

The project aims to exploit the use of highly scalable and inclusive digital technologies (i.e., mobile app and online GSH) to improve the transdiagnostic understanding of EDs and their psychological treatment. Online GSH will be developed based on pre-existing UK-based model of intervention (RecoveryMANTRA and ECHOMANTRA) and will be adapted to Italian young adults through the introduction of a smartphone app aimed to monitor and support individuals to treatment. This project aims to establish at what stage of illness the evidence-based interventions used are most valued and beneficial to individuals. Therefore, online GSH will be tested in two different settings and clinical context: (1) the outpatient setting for those reporting a diagnosis of ED; and (2) the community setting for people reporting symptoms of eating disorders who are not receiving psychological treatment. GSH translated for delivery through smartphone app may be one solution towards reducing treatment gap for people with ED who do not still receive a proper support, fostering awareness and motivation to start treatment as well as an augmenting strategy to improve clinical outcomes for patients undergoing routine treatment delivered by clinical services.

Primary aims: Regarding the outpatient setting, the first aim is to assess if GSH intervention is effective to improve the following primary outcome for patients with ED diagnosis: 1) overall psychological distress. Secondary outcomes are: 1) eating disorder psychopathology; 2) interpersonal distress; 3) quality of life; 4) motivation for treatment; 5) emotion dysregulation; 6) Body Mass Index, 7) satisfaction with treatment, at the end of the 8-week intervention and 3-month follow-up compared to the control condition.

In the community setting, the study aims to examine if GSH is effective to improve the following primary outcome for individuals with symptoms of disordered eating: 1) overall psychological distress; Secondary outcomes are: 1) eating disorder psychopathology; 2) interpersonal distress; 3) quality of life; 4) motivation for treatment; 5) emotion dysregulation, 6) Body Mass Index, 7) satisfaction with treatment, 8) first contact with clinical services to manage their dysfunctional eating behaviors, at the end of the 8-week GSH and at 3-month follow-up compared to the control condition. Secondary aims: A secondary aim is to determine the extent to which the working alliance with clinicians delivering weekly online group guidance is associated with clients' clinical changes on a session-by-session basis during the intervention. Given the difficult nature of building a strong therapeutic alliance with ED patients (Warren, Crowley, Olivardia \& Schoen, 2008), an increased willingness to trust in clinician during the online GSH may be a basis for increased treatment involvement and symptomatic improvement.

Study 1. The first study is a multicentric RCT where participants with a proper diagnosis of eating disorders will be recruited at the Eating Disorders Unit of the University Hospital of Catanzaro and Padova. The aim of this RCT is to answer to the question: "Does online guided self-help through mobile app for eating disorders provided in addition to TAU improve overall psychological distress, eating disorder psychopathology, BMI, interpersonal distress, quality of life, motivation for treatment, emotion dysregulation, satisfaction with treatment compared to treatment as usual alone (outpatient treatment) among young people with a diagnosis of eating disorders?".

Participants will be randomized to one of two groups: (1) online guided self-help + treatment as usual (EDs standard treatment by participating centres) or (2) treatment as usual only). After completing informed consent and baseline questionnaires, participants will be randomised to the intervention or control group, using a web-based randomisation tool. Randomization will take place in blocks, in size of 16 participants, with a 1:1 allocation ratio to ensure a random allocation of 8 participants in each group per time. The participants allocated in the intervention arm will get access to self-help materials (workbook and videoclips), forum groups (interactional materials) and online interpersonal group sessions for 8 weeks.

Study 2. In this RCT, participants will be recruited from the community through advertisements posted in the psychoeducation Instagram profile "Dicci Come Aiutarti" (Albano et al. 2023). The aim of this RCT is to answer to the question: "Does online guided self-help through mobile app for eating disorders improve overall psychological distress as well as eating disorder psychopathology, BMI, interpersonal distress, quality of life, motivation for treatment, emotion dysregulation, satisfaction with treatment and service use, compared to a no treatment condition?". A secondary aim of this trial is to evaluate if during the ongoing trial or at the end of the intervention (8 weeks) participants start a clinical intervention in a specialized service to manage their eating difficulties.

Participants will be randomized to one of two groups: (1) online guided self-help (without online interpersonal sessions) or (2) TAU (only GP consultation). After completing informed consent and baseline questionnaires, participants will be randomised to the intervention or control group using a web-based randomisation tool, with a 1:1 allocation ratio.

Guided self-help intervention. The proposed 8-week online GSH will be adapted from two online interventions developed and tested with people with eating disorders by the Eating Disorders Research Unit team at King's College London, with a cooperation of the University of Palermo research team (Cardi et al., 2020). The GSH will include self-help materials, i.e. a written workbook and a library of 74 brief video-clips with the goal of providing psychoeducation about eating disorders and interpersonal difficulties and describing and modelling the use of helpful strategies to reverse abnormal eating behaviours. Participants will use self-help materials through a mobile app and online platform, which will be developed and implemented by the Tech Digit Easy s.r.l., a company leader in the design and development of E-health services and devices which has been cooperating with the Research Unit of the University of Palermo for 2 years. Regarding anorexic symptoms and behaviours, the self-help material will focus on the topics of the New Maudsley model. Therefore, participants will get the chance to work on six main areas related to EDs maintenance factors. On the other hand, further modules will cover the ED overeating behaviors through the Italian translation of the manual Getting Better Bite by Bite: A Survival Kit for Sufferers of Bulimia Nervosa and Binge Eating Disorders (Schmidt, Treasure \& Alexander, 2012). Finally, a module on interpersonal difficulties related to EDs will be provided to all participants as a transdiagnostic maintenance illness factor to be addressed (Rieger, Van Buren, Bishop, Tanofsky-Kraff, Welch, \& Wilfley, 2010; Wilfley et al., 2002).

A further service offered through the mobile app will be the chance to get access to a weekly online forum group (interactional material) aimed to foster the adherence to the self-help program and the use of online materials. A focus on motivation to change, instilling hope on the recovery, will be promoted through motivational interviewing framework and the lived experience of influencers who recovered from an EDs (collaborators of #DCA). This additional interactional material will be delivered weekly through a synchronous 1-hour mobile chat with a peer mentor (a recovered influencer from an EDs) and a low experienced professional (graduated students in psychology). Each session lasts 60 minutes and is themed following the structure of the patient video-clips materials.

In this project, interpersonal group sessions (IPT) involve 90-minute weekly online sessions (Murphy et al., 2012). Group sessions will be led by clinicians with a good expertise in the treatment of eating disorders. The interpersonal group sessions could be attended by those participants enrolled in the study and recruited from clinical centres with a proper diagnosis of eating disorder. The interpersonal group sessions will be delivered through a zoom link available in the online platform. A group of 6-8 participants, diagnosed with various eating disorders including AN, BN, and BED, will be included with the aim of addressing the interpersonal difficulties related to nutrition and eating management.

Statistical Analysis Power analysis: An a-priori sample size was estimated using G*Power, based on previous research findings (e.g., Cardi et al., 2020 Traviss-Turner et al., 2017). Considering a mixed design factorial ANOVA, a statistical power of (1 - β) = .90, alpha level of .05, a small effect size (Cohen's f = .10) in mean improvement for outcome variables, and medium to high stability of the outcomes across time points (r= .60), the power analysis suggested a sample to include 86 subjects per group, with a total of 172 participants in the study 1. An a-priori sample size was estimated using G*Power, based on previous research findings (e.g., Linardon, Shatte, Rosato \& Fuller-Tyszkiewicz, 2022). Considering a mixed design factorial ANOVA, a statistical power of (1 - β) = .80, a two-tailed alpha level and a small effect size (Cohen's f = .10) in mean improvement for outcome variables, and medium to high stability of the outcomes across time points (r = .60), the power analysis suggested a sample to include 35 subjects per group, with a total of 70 participants in the study 2.

For first aim: Univariate analyses of variance will be used to compare the intervention and control arms on the clinical outcomes. All analyses will be based on the intention-to-treat approach. Missing data will be imputed using multiple imputation based upon the Markov chain Monte Carlo method and maximum likelihood imputation based upon the Expectation-Maximization (EM) algorithm. Results will be compared across the two methods. Linear Mixed-effects Models applying an intent-to-treat approach will be used to compare the intervention arm to the control group in terms of both primary and secondary outcomes. Effect sizes will be calculated using Cohen's d.

For the secondary aim: The random intercepts cross-lagged panel model will assess the bidirectional association between working alliance and post-session weekly outcomes (EDE-Q, CORE-10 etc). The random intercept in the model removes between-person variance (i.e., Level 2) such that the lagged relationships in the cross-lagged model characterize within-person change over time. Three fit indices will be used to evaluate the fit of the model: the comparative fit index (CFI), root mean square error of approximation (RMSEA), and standardized root mean square residual (SRMR). Two models will be estimated, one with the auto-correlation and the cross-lagged paths freely estimated, and a second model with the auto-correlation and the cross-lagged paths constrained to be equal across time periods.