Abstract
Diabetes is a chronic condition affecting millions worldwide and severely impacts health and quality of life. According to the International Diabetes Federation (IDF), over 463 million adults, which is 9.3% of the global population, live with diabetes. Diabetes ranks among the most prevalent chronic diseases and was the ninth-leading cause of mortality in 2019, with 4.2 million deaths reported. This article introduces DiaBD, a novel dataset of 5,288 patient records from Bangladesh, designed to address critical gaps in diabetes research and aid in healthcare planning, risk analysis, and predictive modelling. The dataset comprises 14 attributes including age, gender, clinical vitals (pulse rate, systolic and diastolic blood pressure, glucose levels), anthropometrics (height, weight, body mass index (BMI)), family history of diabetes and hypertension, cardiovascular disease (CVD), and stroke, with a dependent attribute, Diabetic, indicates whether an individual has diabetes or not. The dataset ensures demographic diversity and precise measurements, supporting the study of diabetes and its related health issues. Features like CVD and stroke enable broader research on comorbidities. This dataset facilitates machine learning applications, risk assessment, and personalized healthcare strategies. Researchers can explore the links between diabetes, hypertension, CVD, and stroke, while healthcare providers and policymakers can leverage DiaBD to identify trends, allocate resources efficiently, and enhance public health strategies.