Abstract:

Objective To construct predictive models for lower limb artery stenosis in patients with type 2 diabetes. Methods A total of 4 368 patients with type 2 diabetes underwent ultrasonography examination of lower limb arteries and serum biochemical assay and were assigned to the stenosis group (the stenosis rate ≥50%) and non-stenosis group according to the results of ultrasonography examination of lower limb arteries. Risk factors for lower limb artery stenosis were rated according to the results of binary regression analysis. The model of general clinical characteristics and the model of general clinical characteristics combined with biochemical indices were constructed. Receiver operating characteristic (ROC) curves were used to analyze the predictive potency of two models. Results Multiple Logistic regression analysis indicated that gender, age, course of diabetes, hypertension, poor blood glucose control, and decreased HDL-C level were associated with lower limb artery stenosis. The areas under ROC curves of the two models were 0.77 (0.74-0.80) and 0.78 (0.75-0.81)(P=0.000), respectively. The difference in the area under ROC curve between two models was not statistically significant. Models with scores ≥7 and ≥10 had the best sensitivity (86% and 74%) and specificity (59% and 68%). Conclusion General clinical characteristics can effectively predict the lower limb artery stenosis. For scores ≥7, examinations such as lower limb artery ultrasonography are recommended to exclude the lower limb artery stenosis.

Key words: type 2 diabetes, lower limb artery stenosis, predictive model, ultrasonography