Ensemble stacked model for the treatment and prevention of Arsenic induced Diabetes Mellitus

Diabetes Mellitus is a long-standing metabolic condition. This happens when the body gradually loses its ability to convert the food into useable energy and the different chemical reactions that normally keep glucose under control stop working properly. Consequently, this disturbs the glucose regulation and overall balance of energy. In recent years, the scientists have raised the questions on the contribution of arsenic in the prevalence of diabetes. A major point of interest is the AS3MT gene, which supports the body methylate inorganic arsenic. During this process, a few reactive by-products, including dimethylarginine, may be produced and slowly accumulate. When these compounds build up with the passage of time, they can resist with routine metabolic activity and may increase the possibilities of diabetes.In our study, we focused on how arsenic methylation associated with AS3MT relates to the growth of Diabetes Mellitus. We included 2,000 participants in our dataset, divided evenly between diabetic patients and healthy controls. Along with arsenic measurements, we reviewed clinical indicators, physiological findings, lifestyle patterns, and anthropometric details and the goal was to understand how all variables and elements were fit together. Overall, the diabetic group consistently showed higher blood glucose readings and noticeable shifts in several biochemical parameters when compared with controls. Arsenic and its metabolites were also found in larger amounts among diabetic individuals, with concentrations reaching 56 µg/L, while the control group averaged around 2 µg/L. These observations support the idea that increased arsenic exposure may be tied to diabetic status.For deep analysis and beyond basic comparisons, we used several analytical procedures along with machine-learning methods. Interestingly noted that participants who had apparently stable HbA1c values but faced swear health issues like hypertension, abnormal lipid levels, or signs of kidney or liver issues. A steady pattern appeared in which higher arsenic-related measurements aligned with a greater likelihood of diabetes. Our customized Ensemble Stacked Model performed particularly well, reaching 99% predictive accuracy with Logistic Regression. This suggests that such models might help in early identification and support preventive healthcare efforts. All data processing and statistical work were completed in R Studio and SPSS (Version 23) to ensure transparency and reproducibility.