1042-P: Intradermal Continuous Glucose Monitor with Embedded Activity and Sleep Tracking

Introduction and Objective: Lifestyle behaviors, including physical activity and sleep, significantly impact glycemia in individuals with diabetes. While activity and sleep trackers, as well as glucose sensors, have been available for many years, these technologies have not yet been integrated into a single wearable device. Here we evaluate the performance of activity and sleep tracking using sensors embedded in a wearable intradermal glucose sensor. Methods: Biolinq enrolled 26 people (wearing a total of 32 sensors) with non-intensively managed type 2 diabetes in a clinical trial for its first-generation wearable. Twenty-two wearables were used for training of a novel sleep detection algorithm and the remaining ten wearables were used for validation. In addition to Biolinq devices, subjects wore a Fitbit Inspire 3 and kept sleep diaries. Fitbit-detected sleep was used to correct any diary entry errors (e.g., AM and PM mix-ups); otherwise, sleep diaries served as reference for subjects’ sleep. Results: In the validation set (n=10 sensors), the sleep algorithm achieved an AUC-ROC of 0.97 for binary sleep/wake classification. Sleep periods were correctly detected with 91.3% accuracy; awake period detection was 94.2% accurate. Activity correlated well with the Fitbit (R^2 = 0.92). Conclusion: This study demonstrates the feasibility of collecting concurrent and continuous glucose, sleep, and activity data using sensors integrated within a single wearable device. Such data may enhance the accuracy and utility of CGM measurements, and adds valuable contextual information on how activity and sleep can impact glycemia in people with diabetes. Disclosure A. Cai: Employee; Biolinq. K. Nogueira: Employee; Biolinq. J.J. Bishop: Employee; Biolinq. A.L. VandenBerg: None. R. Yang: None. J.R. Tangney: None.