ObjectiveTo explore the potential association between the serum levels of 25-hydroxyvitamin D [25(OH)D] and carotid atherosclerosis in patients with type 2 diabetes. Material and methodsThree hundred and fifty patients with type 2 diabetes were enrolled in this study in Shanghai, China. B-mode ultrasound was used to detect carotid plaques as indicators of atherosclerosis and measure carotid artery intima-media wall thickness (C-IMT) at two sites of carotid artery. Subjects were divided into group A (patients with carotid plaques) and group B (patients without carotid plaques) and be assessed clinically. Serum levels of 25(OH)D and other clinical parameters were measured. Multivariate logistic regression was performed to find predictors of carotid atherosclerosis in the entire group. ResultsThe levels of serum 25(OH)D were lower in group A than in group B[19.60 (13.30–25.73) vs 23.19 (18.10–30.06)ng/ml, P<0.001]. The C-IMT levels [(1.00±0.17 vs 0.88±0.20)mm, Ptrend<0.001] and proportion of people with carotid plaques(44/88 vs 20/87, Ptrend<0.001) in the lowest quartile of 25(OH)D were higher than in the highest quartile. Vitamin D concentrations were inversely associated with HbA1c in women(r=−0.194, P=0.006), and C-IMT in men(r=−0.409, P<0.001). Logistic regression analysis showed age, male sex, current smoke, history of hypertension, SBP, LDL-C and lg[25(OH)D] (OR: 0.924, 95%CI: 0.893–0.955, P<0.001) were independently associated with the presence of carotid plaques in T2DM. ConclusionsSerum vitamin D level is significantly and independently associated with carotid atherosclerosis in patients with T2DM in Shanghai, China.