Adult users of a digital binge-eating intervention wanted adaptations for type 2 diabetes: Qualitative study of post-intervention preferences

Technical Summary: Adult Users of a Digital Binge-Eating Intervention Wanted Adaptations for Type 2 Diabetes

Problem Statement
Binge eating, characterized by consuming large quantities of food with a sense of loss of control, affects up to 25% of adults with Type 2 Diabetes Mellitus (T2DM). This comorbidity creates a bidirectional relationship where binge eating worsens glycemic control, insulin sensitivity, and weight management, while T2DM management strategies (such as restrictive diets or insulin therapy) can exacerbate binge eating behaviors. Despite the high prevalence and mutual impairment of these conditions, existing binge-eating interventions have rarely been adapted specifically for adults with T2DM. Prior efforts have focused on screening or clinical conversations rather than tailoring digital interventions to the specific physiological, cognitive, and structural barriers faced by this population. Furthermore, previous digital interventions for this group, such as POSE-D, relied heavily on specialist clinical support, limiting scalability, and were often designed with limited end-user input.

Methodology
This qualitative study employed human-centered design methods to analyze post-intervention preferences of adults with T2DM and recurrent binge eating who had completed a digital binge-eating intervention called "FoodSteps."

• Participants: The sample consisted of 42 adults (74% female) drawn from two concurrent trials (a pilot trial focusing on T2DM and food insecurity, and a randomized trial focusing on obesity) conducted in 2024–2025. Eligibility required a self-reported current T2DM diagnosis and recurrent binge eating (≥12 episodes in the past 3 months).
• Intervention Context: Participants engaged in FoodSteps, a 16-week guided self-help digital intervention supported by non-specialist coaches via messaging. The intervention was not originally adapted for T2DM.
• Data Collection: Following the intervention, participants completed semi-structured interviews (10–20 minutes) regarding their experiences, challenges in managing T2DM alongside binge eating, and suggestions for intervention improvements.
• Analysis: Interview data were transcribed and analyzed using thematic analysis. A coding team, including a clinical psychology doctoral student, developed an inductive codebook through iterative meetings to identify themes and subthemes. The analysis focused on identifying barriers to T2DM management and potential adaptation opportunities.

Key Results
The analysis identified three primary themes regarding the intersection of T2DM and binge eating:

1. Disease Management Challenges: Participants reported four categories of barriers:

   • Knowledge Barriers: Uncertainty regarding optimal nutrition and the specific impact of foods on blood glucose.
   • Cognitive/Psychological Barriers: Difficulties with memory (forgetting medications or glucose checks), resisting sugar cravings, and managing restrictive eating patterns that co-occur with binge eating.
   • Physiological Barriers: Challenges with physical activity due to pain or limitations, managing weight, and the specific phenomenon where low blood glucose readings triggered binge-eating episodes.
   • External/Structural Barriers: Financial constraints limiting access to healthy food and medication, insufficient medical guidance, and a lack of social support.

2. Intervention Adaptation Opportunities: Participants proposed specific features to address the identified barriers:

   • Psychoeducation and Nutrition: Content explaining the relationship between T2DM and binge eating, diabetes-friendly meal plans, and tools like continuous glucose monitor integration.
   • Cognitive/Psychological Supports: Automated nudges, personalized reminders for medication and self-care tasks, and incentives for completing health goals.
   • Physiological Supports: Adaptive physical activity suggestions for those with limitations and in-app blood glucose tracking to visualize the impact of eating behaviors.
   • External/Structural Supports: Social forums for peer connection, in-person meet-up options, mechanisms to share data with clinicians, and links to community resources (e.g., food delivery, medication subsidies).

3. Dissemination and Service Delivery Considerations:

   • Target Population: Participants suggested that individuals recently diagnosed with T2DM would benefit most from an adapted intervention, as they are in a critical learning phase.
   • Motivators: Engagement could be driven by positive motivators (e.g., family well-being) or negative motivators (e.g., fear of acute health events), though positive motivators may offer more sustainable engagement.

Key Contributions
This study provides qualitative evidence supporting the necessity of adapting binge-eating interventions for adults with T2DM. By engaging users immediately post-intervention, the research identifies specific, user-driven design features that address the unique intersection of eating disorders and diabetes management. The findings highlight that current "one-size-fits-all" digital interventions may not adequately address the structural and physiological complexities faced by this dual-diagnosis population. The study also advances the literature by documenting the specific phenomenon of hypoglycemia-induced binge eating in T2DM within the context of digital intervention design.

Significance and Claims
The authors claim that their findings offer a foundation for future work to design and validate adapted intervention features. They assert that while the proposed design ideas (such as glucose tracking and psychoeducation) show promise, they must be validated with new users before deployment. The paper concludes that addressing eating behaviors is a critical, yet under-addressed, component of routine diabetes care. It suggests that clinicians could benefit from increased awareness of how eating behaviors undermine T2DM self-management and that future interventions should aim to be self-contained service models to reduce the burden on clinical staff. The research emphasizes the importance of human-centered design in creating scalable, relevant digital health tools for comorbid conditions.