MOVE-CGM: Personalized Exercise Coaching for Adults with Type 1 Diabetes

The Big Picture

Making Exercise Work for People with Type 1 Diabetes

What is the specific goal of this study?

MOVE-CGM is a telehealth and mobile platform delivering personalized exercise coaching for adults with T1D or closely related conditions, specifically middle-aged adults managing a complex daily diabetes regimen alongside work, childcare, and other demands. Coaching tips are timed and tailored to each participant's individual patterns: what their biometric data reveals about their best windows for exercise, and what motivational and logistical considerations their coach identifies through ongoing client-centered conversations.

The program generates integrated visual summary reports through the Labfront Companion mobile app, covering each participant's glucose, activity, heart rate, and sleep. These are reviewed in semi-monthly telehealth coaching sessions led by exercise physiologists with expertise in diabetes management. This human-directed personalization is also a step toward a future automated just-in-time adaptive intervention (JITAI), where an algorithm triggers support at the precise moments individuals are most likely to benefit.

What real-world problem is this research trying to solve?

Exercise for people with T1D can trigger hypoglycemia or hyperglycemia, requiring complex adjustments to insulin and diet. This physiological unpredictability alone is a powerful deterrent. Existing diabetes technology, including CGM and automated insulin delivery systems, can support and even partially automate these adjustments. But most platforms still fall short of addressing the logistical and psychological burdens of being a middle-aged adult: competing demands on time, fluctuating motivation, and the memory of a past hypoglycemic episode that can make someone avoid exercise for days or weeks.

MOVE-CGM bridges this gap by aligning biometrically favorable windows for exercise with logistically and psychologically favorable ones, which can be equally hard to find. By treating motivation as a dynamic state that shifts throughout the day rather than a fixed personality trait, the program delivers support that meets people where they are to create exercise habits that feel sustainable and effective.

‍

The Study Blueprint

A Third-Generation Intervention, Built on What Worked

How many participants, and what's the study timeline?

The current study will enroll 60 participants with T1D or related conditions, ages 30–65, through July 2027. This builds on two prior iterations of MOVE-CGM: a 20-person pilot (2019–2020) and a 24-person feasibility study, which together established that the platform is feasible, acceptable, and showed promising early effects on exercise behavior. The intervention design was itself informed by formative qualitative research, including a study of adults with T1D that examined how they perceived the relationship between diabetes management and exercise, and what they actually wanted from a support program (Swaminath, Yale University, 2023).

Each version has directly shaped the next:

• Iteration 1 — Pilot Study (2019–2020)
  20 participants, 10 weeks. CGM-integrated biometric reports and monthly motivational coaching. Results published in Clinical Journal of Sport Medicine, 2023.
• Iteration 2 — Feasibility Study
  24 participants. Added a biometric dashboard and weekly written coaching tips, responding to pilot data showing that monthly coaching produced transient motivation.
• Iteration 3 — Current Clinical Trial (through July 2027)
  60 participants. Coaching tips delivered between sessions and timed to each participant's individual biometric and motivational patterns. This iteration is building the evidence base for a future automated JITAI.

Why a wearable, and why the Garmin Vivosmart 5 specifically?

Traditional clinical tools are episodic and environment-bound, capturing snapshots rather than the continuous, contextually rich data streams that personalized coaching requires. The Garmin Vivosmart 5 provides continuous heart rate, step count, sleep actigraphy, and movement data throughout daily life.

Participants keep the device after completing the study. This isn't just a retention incentive; it reflects the program's goal of building sustainable habits that extend beyond the trial period.

The Garmin data is compiled and organized in Labfront, enabling the team to generate integrated visual reports combining Garmin biometrics with CGM values, insulin logs, and ecological momentary assessment survey data. Examples include comparing glucose on exercise and non-exercise days or sleep quality after workouts. Labfront's ability to bring these data streams together in a participant-facing visualization is central to this intervention's effectiveness.

The Methodology

Intersectional Analysis Across Biometric and Behavioral Data

Which physiological metrics are most critical?

The most critical metrics are continuous glucose values, heart rate, activity and sleep actigraphy. But what makes this program distinctive is not any single metric. It's the analysis of these metrics in combination.

Key intersections the team examines:

• Blood glucose on exercise versus non-exercise days, to quantify the glycemic benefit of specific workouts for each individual
• Sleep quality the evening before and after exercise sessions
• Heart rate intensity alongside glucose trends during and after exercise
• Daily step counts as a proxy for overall activity beyond structured exercise

‍

This intersectional analysis is what allows coaches to move beyond generic recommendations and point participants to patterns in their own data. For example, showing that a specific type of workout at a particular time of day is associated with better glucose control and next-morning sleep for that individual.

‍

What other data streams does the study collect?

Labfront is the integration hub for the study's primary data streams.

Through Labfront, the team collects:

• Garmin Vivosmart 5: Steps, heart rate, and sleep actigraphy
• EMA surveys: Exercise barriers, mood, sleep quality, and fear of hypoglycemia
• Exercise video usage logs. Early iterations of this research used GlucoseZone's mobile application to deliver exercise video content; in the current version, GlucoseZone's video library is accessed via YouTube and linked through Labfront, allowing the research team to retain the same diabetes-specific content while leveraging Labfront's data infrastructure.

‍

Collected outside Labfront

CGM values, blood pressure, validated psychosocial instruments (motivation, self-efficacy, fear of hypoglycemia scales), body weight and waist circumference, research-grade waking hip actigraphy for objective physical activity verification, and self-reported exercise logs for cross-validation.

‍

The Impact

From Zero to 64 Minutes a Week

What have prior iterations of MOVE-CGM shown?

Clinical outcome analyses for the current study are ongoing, but prior iterations have already produced compelling findings. In the published pilot (Ash et al., Clin J Sport Med 2023), participants increased weekly exercise from a median of zero minutes to 64 minutes per week. Many credited not the coaching conversations, but the moment they saw their own data visualized: glucose placed side-by-side with exercise, sleep, and mood. Several described viewing their integrated report as the point at which things "clicked."

A key limitation the prior work identified: motivation from monthly sessions was often transient. That finding directly shaped the current study, which introduces coaching tips delivered between sessions and timed to each participant's individual biometric and motivational patterns. Early feedback is encouraging. Participants report the timing and content feel personally relevant rather than generic.

"I did notice after doing exercise sessions the glucose went down, so the exercise does work." - _{MOVE-CGM Pilot Participant}

How will this research improve diabetes prevention or treatment?

Regular exercise reduces cardiovascular risk, improves glycemic control, supports weight management, and benefits mental health. Most people with T1D do not exercise regularly, largely because the physical, logistical, and psychological barriers are not fully addressed by existing tools. MOVE-CGM makes coaching personal, timely, and grounded in each participant's own data rather than generic population-level guidelines.

The current study will provide the first controlled evidence base for whether motivationally-informed, biometric-guided coaching can produce meaningful and sustained exercise increases in this population. Beyond T1D, the approach is scalable. The framework of combining wearable biometric data with real-time psychological assessment to deliver tailored behavioral support applies across chronic conditions where lifestyle behavior change is both essential and difficult to sustain.

‍

Key Reference

Ash GI, Nally LM, Stults-Kolehmainen M, et al. Personalized Digital Health Information to Substantiate Human-Delivered Exercise Support for Adults With Type 1 Diabetes. Clin J Sport Med. 2023;33(5):512–520. doi:10.1097/JSM.0000000000001078.

Swaminath, Meera. Perceptions of Adults with Type 1 Diabetes About Diabetes Management and Exercise. Yale University, 2023. https://elischolar.library.yale.edu

Support

An NIH-NIDDK mentored research scientist development award (K01DK129441) supports this study. The NIH is not involved with the views expressed in this article or the decision to disseminate them.

‍