Using Logistic Regression to Identify The Most Important Factors That Lead To Diabetes

Abstract

This study aimed to shed light on one of the important statistical methods, which is the use of the binary response logistic regression method, which is a special case of the general linear model and is the most common in descriptive data analysis. The reason behind addressing this type of statistical methods is the importance of logistic regression in predicting the most important variables affecting the incidence of diabetes in the city of Al-Khums 162 people, divided into 81 diabetic patients and 81 non-diabetic patients, for the year 2024. The independent variables were total cholesterol, triglycerides, LDL cholesterol, high-density lipoprotein cholesterol, free thyroxine, and thyroid-stimulating hormone The probability of developing diabetes as a dependent variable was highly significant in both the level of total cholesterol, triglycerides, the level of harmful cholesterol in the blood, and the level of high-density lipoprotein cholesterol. Therefore, these factors are considered the most important causes of developing diabetes, while it was not significant in the level of free thyroxine hormone  and the level of thyroid-stimulating hormone.