None listed
Conditions
Brief summary
Type 2 diabetes requires ongoing day-to-day self-management, including adherence to medication and sustained lifestyle changes. Many individuals find it challenging to maintain these behaviours between clinic visits, where structured support is limited. This study will evaluate the feasibility, acceptability, and preliminary efficacy of DiNa, an AI-powered digital health coaching program delivered via a mobile application. Participants allocated to the intervention group will complete structured coaching sessions over 12 weeks addressing key self-management domains, including healthy eating, physical activity, taking medication, glucose monitoring, problem solving, reducing risks, and healthy coping. The system integrates data from a smartwatch and continuous glucose monitor (CGM) to deliver data-informed feedback and goal review during sessions. This is a 12-week pilot randomised controlled trial comparing the DiNa program with an active control group using a commercially available diabetes self-management application (mySugr). Both groups will use a smartwatch and CGM for monitoring, while AI-powered coaching is provided exclusively to the intervention group.
Interventions
The Digital Health Navigator (DiNa) intervention is a digital health coaching program delivered via a mobile application, supported by wearable devices (smartwatch and continuous glucose monitoring [CGM] sensors), for adults with type 2 diabetes.
Participants allocated to the intervention arm receive an individually tailored coaching program delivered by an AI Coach. The intervention targets the Association of Diabetes Care and Education Specialists 7 (ADCES7) Self-Care Behaviors: healthy eatin
The Digital Health Navigator (DiNa) intervention is a digital health coaching program delivered via a mobile application, supported by wearable devices (smartwatch and continuous glucose monitoring [CGM] sensors), for adults with type 2 diabetes. Participants allocated to the intervention arm receive an individually tailored coaching program delivered by an AI Coach. The intervention targets the Association of Diabetes Care and Education Specialists 7 (ADCES7) Self-Care Behaviors: healthy eating, being active, monitoring, taking medication, problem solving, healthy coping, and reducing risks. Participants receive 24 remote coaching sessions over 12 weeks (minimum two sessions per week). Each coaching session is approximately 10–15 minutes in duration. Sessions are accessed on a study-provided smartphone in participants' everyday settings, such as at home. The AI Coach delivers coaching through an interactive conversational voice agent built into the mobile application, guiding participants through real-time spoken conversations. During each session, the AI Coach answers questions, provides feedback, and supports participants with self-management behaviours, drawing on participant-reported data and physiological data from wearable devices (e.g., activity, heart rate, sleep) and glucose data from CGM to support goal review, reflection, and progress tracking. The intervention is individually tailored, with coaching content, behavioural goals, and session focus areas shaped by each participant's self-management priorities. This tailoring is intended to make the program more personally relevant and to help participants overcome their own barriers to behaviour change. Tailoring starts at baseline using each participant's current self-management behaviours and personal goals, and is refined over the 12-week program as the AI Coach learns from past conversations, tracks each participant's progress, and draws on real-time data from wearable devices and CGM. In this way, the coaching becomes increasingly personalised as the AI Coach builds an understanding of each participant's needs, preferences, and patterns over time. The intervention uses a range of established behaviour change techniques (BCT), such as goal setting, action planning, problem solving, self-monitoring, prompts and cues. Motivational interviewing and cognitive behavioural therapy (CBT) principles are used to support participant reflection and engagement. Intervention delivery is automated within the platform. Engagement with the intervention is tracked through platform analytics, including the number of sessions completed, how often sessions are accessed, and how long participants spend in each session.
Sponsors
The University of Auckland
Study design
Allocation
Randomised controlled trial
Intervention model
Parallel
Primary purpose
Educational / counselling / training
Masking
Blinded (masking used) (Outcomes Assessor)
Eligibility
Sex/Gender
All
Age
18 Years to No maximum
Healthy volunteers
No
Inclusion criteria
1. Aged 18 years and older with a confirmed diagnosis of T2D. 2. Willingness and ability to use a smartphone for study participation (a study-provided smartphone will be supplied for participants in the DiNa intervention arm). 3. Be willing to use the DiNa system and wearable devices (e.g., smartwatch, continuous glucose monitor) throughout the study duration. 4. Be proficient in English to ensure they can understand and interact with the DiNa system and complete study questionnaires. 5. Provide informed consent to participate in the study. 6. Access to a reliable internet connection to enable data synchronisation and system functionality.
Exclusion criteria
1. Participants with severe comorbid conditions that could interfere with the study (e.g., advanced heart disease, renal failure, severe mental health disorders). 2. Pregnant or lactating women due to potential variability in diabetes management needs during pregnancy. 3. Participants who have participated in another clinical trial within the past 30 days to avoid confounding effects. 4. Individuals with significant cognitive impairment or dementia that would prevent them from effectively using the DiNa system and completing study tasks. 5. Participants with severe diabetes-related complications (e.g., recent diabetic ketoacidosis, severe hypoglycemia or hyperglycemia requiring medical intervention) within the past six months. 6. Incompatibility with the continuous glucose monitoring (CGM) system or other study devices (e.g. device or software incompatibility). 7. Participants with a history of all other types of diabetes, including gestational diabetes and LADA (latent onset adult T1D)
Outcome results
None listed