Glycemic Risk Across Exercise Modalities in Adults with Type 1 Diabetes Using Continuous Glucose Monitoring and Wearable Sensors: A Prospective Cohort Study

Background: Exercise provides important health benefits for adults with type 1 diabetes; however, it remains associated with substantial glycemic instability that may vary according to exercise modality, intensity, duration, and clinical context. Continuous glucose monitoring (CGM) and wearable sensors offer an opportunity to characterize exercise-related glycemic responses under real-world conditions, yet prospective free-living data remain limited. Objective: This study aimed to evaluate glycemic risk across exercise modalities in adults with type 1 diabetes using CGM and wearable sensors in a real-world prospective cohort. Methods: This prospective cohort study was conducted under free-living conditions in 120 adults with type 1 diabetes. Participants were followed during habitual exercise using CGM, wearable sensor data, and session-level exercise classification. A total of 1568 valid exercise sessions were analyzed and categorized as aerobic, resistance, interval-based, or mixed exercise. The primary outcomes were immediate glucose change and time below range during exercise and within 6 h post-exercise. Secondary outcomes included severe biochemical hypoglycemia, time in range, time above range, glycemic variability, delayed hypoglycemia, nocturnal hypoglycemia, and rescue carbohydrate intake. Results: Glycemic risk differed across exercise modalities. Aerobic exercise was associated with the greatest immediate glucose decline, the highest time below range, the highest frequency of delayed post-exercise hypoglycemia, and the greatest need for rescue carbohydrate intake. Resistance exercise showed the most favorable acute glycemic profile, whereas interval-based and mixed exercise showed intermediate patterns. The associations between exercise modality and glycemic risk were modified by pre-exercise glucose level, time of day, and insulin delivery modality. Sensitivity analyses were consistent with the primary findings. Conclusions: In adults with type 1 diabetes monitored under real-world conditions, glycemic risk varies meaningfully across exercise modalities and is further shaped by clinically relevant contextual factors. These findings support a more individualized interpretation of exercise-related glycemic responses using CGM and wearable-derived data.