From Health Data to Exercise Decisions: A Multistakeholder Qualitative Study of Technology-Mediated Feedback for Older Adults With Type 2 Diabetes (Preprint)

BACKGROUND Wearable and extended reality (XR) technologies can provide exercise and health feedback, but how older adults with type 2 diabetes (T2DM) interpret this information and use it to inform exercise decisions within community services remains insufficiently understood. OBJECTIVE To explore how older adults with T2DM and other stakeholders understand technology-mediated feedback and envisage its delivery, use, and support in community exercise self-management. METHODS A five-stage qualitative study in Guangzhou, China, involved 24 participants (10 older adults with T2DM, 7 community workers, 5 healthcare professionals, and 2 extended reality (XR) designers). Data were generated through exploratory interviews, 2-week design probes incorporating activity tracking and contextual records, clarification interviews, a co-design workshop, and evaluation interviews. Materials from these stages were analyzed using reflexive thematic analysis, comparing interpretations across participants, stakeholder groups, and stages. RESULTS Five themes and 13 subthemes described conditions shaping feedback value, interpretation, delivery, exercise decisions, and support. Perceived feedback usefulness did not necessarily imply greater motivation to exercise, and preferences for information content and detail varied. Measured, self-reported, and inferred information raised different interpretive questions. Understandable descriptions did not necessarily clarify assessment criteria, while individualized exercise recommendations required further consideration of exercise goals, personal circumstances, and professional input. Preferences for feedback timing, presentation, in-exercise assistance, and automation differed according to use context, task demands, and desired personal control. Feedback was expected to inform a range of exercise responses, including continuing, adjusting, resting, stopping, or postponing exercise. Previous exercise experiences and longer-term information were also considered relevant to subsequent planning. Preferences for information sharing differed, while access to feedback did not necessarily establish responsibility for its interpretation or subsequent action. Proposed community support arrangements raised further questions about consent, staffing, clinical expertise, and response responsibility. CONCLUSIONS Technology-mediated feedback should help users interpret information and make contextually appropriate exercise decisions rather than merely deliver information. Further evaluation should assess the usability, safety, and feasibility of proposed feedback and community support arrangements during sustained use. CLINICALTRIAL N/A