Abstract
OBJECTIVE: Diabetes mellitus (DM) is an important global health risk factor in the twenty-first century and one of the 10 major causes of mortality world wide. The generalized structural equation model (GSEM) is a family of statistical techniques in the analysis of multivariate data, classified and sequential, which measures the hidden variables and the relationships between them. Finding risk factors for type 2 diabetes and providing a model for lifestyle changes is the aim of the study.. METHODS: This exploratory, cross-sectional study was performed to evaluate the risk factors in a cohort of Iranian diabetic patients aged over 35 years (N = 9975). Among 9975 people over 35 years old participating in the first phase of the Yazd cohort study, 1736 people (17.95%) with diabetes and people who were unaware of their diabetes status and pregnant women were excluded and finally we selected 7431 non-diabetics who had FBG test. By presenting the model of initial generalized structural equations using stata software (version 15), we investigated the risk factors affecting type 2 diabetes. RESULTS: The risk factors of BMI (Impact coefficients0.010), triglyceride (0.005), hypertension (0.086), and high cholesterol level (0.005) directly affected the DM status (P < 0.001). Meanwhile, BMI and triglyceride played a mediating role in this regard, and the factors of high-density lipoprotein cholesterol, physical activity (-0.23), and diet (0.001) indirectly affected the DM status. CONCLUSION: Using a large sample, this study provides a clear and direct model of the risk factors for diabetes. The main finding is that the risk factors for diabetes 2 that directly affect Iranians in this study are high blood pressure, BMI, triglycerides and cholesterol.