MON-553 The Long-Term Impact of Sedentary Lifestyles and Occupational Risks on Diabetes in High- and Low-Income Countries: Forecasting Mortality and Disability Trends Using Machine Learning

Abstract Disclosure: A. Dixit: None. N. Koduri: None. B. Padamati: None. G. Palaniswamy: None. P.R. Patel: None. A.J. Augustine: None. S. Sahu: None. T. Jain: None. V.A. Mendpara: None. Background: Type 2 Diabetes Mellitus (T2DM) is a growing public health concern, contributing significantly to global mortality and disability-adjusted life years (DALYs). The impact of low physical activity and occupational risk factors on diabetes burden differs across high-income and low-income countries due to disparities in healthcare access, urbanization, and economic transitions. This study examines trends in diabetes deaths, DALYs, Age-Standardized Mortality Rates (ASMR), and physical inactivity from 1990 to 2021 and projects diabetes mortality up to 2035 using machine learning-based forecasting. Methods: Data from the Global Burden of Disease (GBD) study (1990-2021) was analyzed for World Bank-classified high-income and low-income countries. Metrics included diabetes-related deaths, DALYs, ASMR, and physical inactivity prevalence. A correlation analysis (r = 0.34, p < 0.05) examined the relationship between physical inactivity and diabetes mortality. A Random Forest regression model (R² = 0.79, RMSE = 7.2) was applied to forecast diabetes mortality trend from 2022 to 2035, incorporating historical patterns of low physical activity, DALYs, and ASMR. Results: Diabetes-related deaths increased by 108% globally over three decades, rising from 1.2 million in 1990 to 2.5 million in 2021. Low-income countries experienced a steeper rise in mortality, while high-income nations saw a more gradual increase. ASMR increased from 12.5 to 19.4 per 100,000 in high-income countries and from 17.8 to 29.7 per 100,000 in low-income countries, reflecting widening disparities. DALYs surged from 25.3 million in 1990 to 50.7 million in 2021, with low-income regions now contributing nearly 60% of the global burden. Physical inactivity prevalence increased from 23% to 38% in high-income countries and from 14% to 26% in low-income regions, with moderate correlation to diabetes mortality. Machine Learning-Based Forecasting (2022-2035): The Random Forest model predicts diabetes mortality in high-income countries stabilizing at approximately 91 deaths per 100,000 per year. In low-income countries, deaths are expected to reach 53 per 100,000, though underreporting and urbanization may cause higher actual mortality. DALYs and ASMR are projected to continue rising in low-income regions, reaching 35 per 100,000 by 2035, emphasizing the need for policy intervention. Conclusions: Diabetes mortality and disability will continue to rise, disproportionately affecting low-income countries. Governments should prioritize workplace health programs, diabetes screening, and urban planning interventions to counteract rising inactivity and occupational risks. Strengthening healthcare access and public awareness initiatives is crucial. Future research should integrate obesity, diet, and socioeconomic stressors to refine predictive models and guide global diabetes prevention efforts. Presentation: Monday, July 14, 2025