Associations among Continuous Glucose Monitoring, Engagement, and Weight Loss in a Digital Diabetes Self-Management Education and Support Program: A Non-randomized Retrospective Cohort Study (Preprint)

BACKGROUND Modest weight loss is recommended for individuals with type 2 diabetes (T2D) to help improve glycemic control and reduce cardiovascular disease risk. Continuous glucose monitoring (CGM) is recommended as a standard of care for individuals with T2D who use insulin, but more research is needed to inform recommendations for CGM use in all T2D populations. Digital diabetes self-management programs can help individuals with T2D lose weight and improve glycemic control. Less is known about the impact of CGM on program engagement within digital programs for diabetes self-management and the relationship between CGM use and weight loss for those with T2D. OBJECTIVE The primary objective of this study was to examine the difference in percent weight loss from baseline to 6 months between CGM users and non-users among commercially insured members with T2D in a digital program for diabetes self-management (Omada for Diabetes). METHODS A non-randomized retrospective observational cohort study was performed using data from 2,612 Omada for Diabetes program members (mean age=52.3 years, 56.4% female, 63.5% white) who started the program between January 1, 2021 and September 30, 2021. Data analyses examined the difference in percent weight loss over 6 months between CGM users and non-users overall as well as stratified by body mass index, program engagement, and CGM-adherence, and associations between program engagement and CGM group. RESULTS A higher percentage of CGM users (57.6%) than non-users (48.4%) were classified as ‘highly engaged’ with the Omada for Diabetes program (p<.001). Both groups showed significant within-group mean percent weight loss from baseline to 6 months (-2.0% CGM users, -1.8% non-users, p<.001), but no differences were detected between groups. When stratified by program engagement, highly engaged CGM users and non-users had significantly greater percent weight loss compared to those with normal/low engagement (CGM users: -2.50% vs. -1.33%, p=.004; non-users: -2.43% vs. -1.30%, p<.001). In fully adjusted models, CGM users and non-users had significant reductions in percent weight loss (β=-2.0%, 95% CI (-2.42, -1.57), β=-1.87%, 95% CI (-2.11, -1.63), respectively). CONCLUSIONS Members participating in the Omada for Diabetes program had a significant change in mean percent body weight from program start to 6 months regardless of CGM status. A higher percentage of CGM users were more engaged with the program than non-users, and higher engagement across both groups was associated with greater percent weight loss. Further understanding the impact that CGM has in digital diabetes self-management programs for T2D could enhance program effectiveness, encourage sustained engagement, and inform standard of care recommendations. CLINICALTRIAL N/A