Development of clinical risk models for diabetic cardiovascular autonomic neuropathy in a Chinese population using logistic regression analysis

Abstract

Background: We developed clinical risk models for predicting diabetic cardiovascular autonomic neuropathy (DCAN) in Chinese diabetic patients. Methods: A Chinese cohort of 455 diabetic participants underwent a short heart rate variability (HRV) test which was recruited between 2011 and 2013. Clinical risk models were developed that included independent and significant risk factors by using multiple variable stepwise regressions. These clinical risk models were tested in another independent cohort of Chinese individuals. Results: The clinical risk models included age, fasting plasma glucose, 2-h plasma blood glucose, triglycerides, resting HRs, and duration of diabetes mellitus. The area under the receiver-operating characteristic (ROC) curve of the study group was 0.794. In the model with the continuous variables, the area under the ROC curve was 0.810. A cutoff score of 12.54 which produced the optimal sensitivity (68.20%) and specificity (76.80%) and identified the percentage (35.77%) of the population that required subsequent testing. Conclusions: The clinical risk models showed high sensitivity and specificity for the prediction of DCAN in Chinese diabetic patients.