Objective

This study aimed to develop a risk engine that consists of multiple risk equations for predicting micro- and macrovascular complications for individuals with type 1 diabetes (T1D).

Methods

We used data from two US nationally representative study cohorts: the Diabetic Control and Complications Trial (DCCT) and Epidemiology of Diabetes Interventions and Complications (EDIC) study, to develop the risk engine. 1,441 participants were enrolled in DCCT study and 1,375 of them were followed by the EDIC study. The microvascular outcomes included retinopathy, neuropathy, and nephropathy. The macrovascular outcomes included myocardial infarction (MI), cerebrovascular events, death from cardiovascular diseases, silent MI, congestive heart failure (CHF). Cox proportional hazard models and logistic regression together with stepwise model selection method were used to select significant risk factors.

Results

The baseline HbA1c and duration of diabetes were the most common significant risk factors for microvascular outcomes (all p-values <0.05). Gender and drinking status were the other significant risk factors for retinopathy (both p-values <0.05). Albumin excretion rate (AER), age, body mass index (BMI), and high-density lipoprotein (HDL) were the other significant risk factors for nephropathy (all p-values< 0.05). Retinopathy at baseline and age were the other significant risk factors for neuropathy (both p-values < 0.05). For the macrovascular outcomes, significant risk factors included age, duration of diabetes, level of HbA1c, BMI, systolic blood pressure (SBP), high-density lipoprotein, low-density lipoprotein, and gender (each p-value < 0.05, all the data of potential risk factors used were collected at DCCT baseline) .

Conclusion

This risk engine for diabetes complications was developed for the US individuals with T1D. External validation process is needed for this new risk engine.