Recovery is Relational: Digital Support Needs for Patients and Supporters in Eating Disorder Recovery

Paper Title

Recovery is Relational: Digital Support Needs for Patients and Supporters in Eating Disorder Recovery

Publication Info

• Topic area: Digital interventions for eating disorder recovery and relational support dynamics.
• Keywords: Eating disorders, digital health, social support, recovery, just-in-time interventions, relational dynamics, cognitive-behavioral therapy, patient-supporter interaction, context-sensitive tools, mental health technology.

Background and Problem

• Problem / challenge: Eating disorder (ED) recovery often occurs outside clinical settings, leaving patients vulnerable and supporters uncertain about how to provide effective help. Existing digital tools lack sufficient contextual and relational sensitivity to address these gaps.
• Significance: Addressing these challenges can improve recovery outcomes by providing timely, situationally aware support, reducing caregiving burdens, and fostering collaborative recovery efforts.
• Motivation and related work: Prior work has explored clinical interventions (e.g., CBT, family-based therapy) and digital tools (e.g., self-tracking, chatbots), but these approaches often focus on individuals and overlook relational dynamics. There is limited understanding of when and how support is needed in everyday life, particularly in the context of patient-supporter relationships.

Solution

• Proposed approach: A mixed-methods study combining a design session and a two-week in-situ diary study to explore everyday support needs of ED patients and their social supporters, with a focus on opportunities for digital interventions.
• Novelty:
  1. Empirical insights into the everyday contexts of support needs for both ED patients and their supporters.
  2. Identification of misalignments in support due to timing and relational dynamics, and their impact on engagement patterns.
  3. Design implications for context-sensitive digital tools that mediate communication and complement human care.
• Procedure and key techniques:
  • Conducted a design session to help participants reflect on support needs and familiarize them with structured journaling.
  • Implemented a two-week diary study using a custom mobile app, capturing in-situ support needs and reflections.
  • Conducted mid- and post-study interviews to clarify diary entries and discuss potential digital interventions.
  • Analyzed data through thematic coding to identify recurring contexts, support needs, and relational dynamics.

Results

• Concrete findings:
  • Identified 19 everyday contexts where support was needed, such as meal-related anxieties, symptom urges, and post-symptom struggles.
  • Documented 17 types of desired support for patients (e.g., emotional comfort, cognitive restructuring, sleep support) and 7 types for supporters (e.g., ED education, emotional support, communication mediation).
  • Highlighted timing as critical, with support most effective when provided just before or after vulnerable moments.
• Advantage over baselines: Unlike existing tools, the proposed approach emphasizes relational dynamics and situational awareness, addressing gaps in timing, communication, and coordination.
• Experiments / evaluation:
  • Participants: 27 ED patients and 12 social supporters, recruited via online communities and snowball sampling.
  • Data collection: 605 diary entries and interviews analyzed for themes related to support needs and relational challenges.
  • Metrics: Contexts of vulnerability, types of support desired, and relational dynamics.
• Limitations and future work:
  • Limited sample diversity, particularly among supporters.
  • Findings reflect perceived needs rather than validated intervention outcomes.
  • Future work should explore broader populations and evaluate proposed digital interventions in real-world settings.

Summary

This study investigates the everyday support needs of ED patients and their social supporters, revealing critical gaps in timing, communication, and relational dynamics. It identifies 19 contexts of vulnerability and 17 types of desired support, emphasizing the importance of situationally aware, relationally sensitive digital tools. The findings suggest opportunities for designing interventions that complement human care, mediate communication, and extend support beyond clinical settings. Future work should focus on broader populations and in-situ evaluations to validate these insights.