Risk factor assessment of prediabetes and diabetes based on epidemic characteristics in new urban areas: a retrospective and a machine learning study.

Abstract

To explore in depth the characteristics of the risk factors for diabetes and prediabetes pathogenesis and progression in special regions. We investigated medical data from 160 thousand cases in the newly developing urban area of a large modern city from 2015 to 2021. After excluding the population with incomplete data, a total of 47,608 people who underwent physical examinations and blood tests were included in this study. A total of 5.0 ± 0.6% of the population aged 41.3 ± 12.6 years had diabetes, and 5.3 ± 2.0% had prediabetes. Risk factor assessment in different states suggested that early risk factors for diabetes pathogenesis were associated with aging, metabolic disorders and obesity, and the consequent risk factors for disease progression were liver, cardiovascular and kidney dysfunction. Our machine learning model was used for disease risk estimation. After the model was trained, the precision and recall rate of the prediction reached 0.76 and 0.86, respectively, with an F1 score of 0. 81. Moreover, there was a greater incidence of diabetes in men than in women (6.68% vs. 2.61%, χ2 = 1415.68, p < 0.001). They all live in the same urban area and have similar age. Diabetes and prediabetes can improve and even reverse to a normal state through a healthy lifestyle. Taken together, the risk factors were independent, but they had synergistic effects on different factors responsible for the pathogenesis and progression of diabetes. Early intervention in health management, especially individual strategies associated with obesity and metabolism, is very helpful for diabetes prevention with increasing age.