Using Structural Equation Modeling to Untangle Pathways of 5 Risk Factors Associated with Incident Type 2 Diabetes: The 6 Lifelines Cohort Study
Risk factors for type 2 diabetes are multifaceted and interrelated. Unraveling the complex pathways of modifiable risk factors related to incident type 2 diabetes will help prioritize prevention targets. The current analysis extended a previously proposed conceptual model by Bardenheier et al. (Diabetes Care, 36(9), 2655-2662, 2013) on prediabetes with a cross-sectional design. The model described the pathways of four aspects of modifiable risk factors in relation to incident type 2 diabetes, including socioeconomic status (income and education); lifestyle behaviors (diet quality, physical activity, TV watching, smoking, risk drinking, and unhealthy sleep duration); clinical markers (HDL-cholesterol, triglycerides, BMI, and waist circumference); and blood pressure. We performed structural equation modeling to test this conceptual model using a prospective population-based sample of 68,649 participants (35-80 years) from the Lifelines cohort study. During a median follow-up of 41 months, 1124 new cases of type 2 diabetes were identified (incidence 1.6%). The best-fitting model indicated that among all modifiable risk factors included, waist circumference had the biggest direct effect on type 2 diabetes (standardized β-coefficient 0.214), followed by HDL-cholesterol (standardized β-coefficient - 0.134). Less TV watching and more physical activity were found to play an important role in improving clinical markers that were directly associated with type 2 diabetes. Education had the biggest positive effects on all lifestyle behaviors except for unhealthy sleep duration. Our analysis provides evidence to support that structural equation modeling enables a holistic assessment of the interplay of type 2 diabetes risk factors, which not only allows the estimation of their total effects but also prioritization of prevention targets. Regarding the current guideline for diabetes prevention, waist management in addition to BMI control (clinical level), as well as less TV watching in addition to more physical activity (behavioral level), may provide additional public health benefits. Better education would be the main societal goal for the prevention of type 2 diabetes.
Keywords: Conceptual model; Path analysis; Risk factors; Structural equation modeling; Type 2 diabetes.