2004-P: Real-World Evaluation of a Personalized, Multilingual Digital Diabetes Care Model across Metropolitan and Nonmetropolitan India

Introduction and Objective: In India, digital diabetes care outside metropolitan regions (MR) is limited by access to care, native dialects and cuisine differences. This study evaluates Sugar.Fit app use and clinical outcomes of a personalized, nationwide digital diabetes program with language-specific coaching across MR and non-MR (NMR). Methods: A retrospective real-world analysis of individuals with diabetes using the Sugar.Fit program across India was conducted, stratified by location (MR vs NMR) and age. The program integrates continuous digital tracking with proactive health coach support in regional languages. App utilization metrics and clinical outcomes were assessed over 3 months. Results: A total of 13,379 users were analyzed (MR: n = 7,067; NMR: n = 6,312). MR users had higher early engagement, while NMR users showed greater mid to late engagement with more text interactions; audio/video call use was similar across locations. In age groups 21-40, 41-60, and >60 years, NMR users had similar reductions in HbA1c (1.03, 1.55, 1.35% vs. 1.13, 1.61, 1.32%), FBS (20.58, 36.10, 30.22 vs. 25.58, 36.38, 30.41 mg/dL), and weight (1.79, 1.89, 2.50 vs. 2.42, 1.89, 2.02 kg) compared with MR users. Overall reductions were similar between locations for HbA1c (1.35% vs. 1.36%), FBS (30.19 vs. 31.20 mg/dL), and weight (2.24 vs. 2.04 kg), indicating comparable effectiveness, with all changes statistically significant (all p<0.001). Conclusion: A personalized, multilingual digital diabetes care model demonstrated high engagement and clinically meaningful outcomes across age groups and geographically diverse Indian population. Comparable improvements among users from MR and NMR highlight the adaptability, value, and scalability of culturally tailored, coaching-driven digital care in resource-constrained settings. Disclosure S. Kumar: None. A.M. Raymond: None. A. Sequeira: None. J. Joseline: None. C.G. Mehra: None.