Vascular adhesion protein-1 (VAP-1) participates in inflammation and can catalyze oxidative deamination to produce hydrogen peroxide and aldehydes, leading to generation of AGEs and ALEs. We have reported that plasma VAP-1 is higher in subjects with diabetes and predicts the incidence of diabetic complications. VAP-1 expression was increased in plaques in humans and in apolipoprotein E (ApoE)-deficient mice. Patients with CAD had higher plasma VAP-1 than those without CAD. Plasma VAP-1 was positively associated with the extent of CAD. In Apo E-deficient mice,VAP-1 inhibition reduced atheroma and decreased oxidative stress. VAP-1 inhibition attenuated the expression of adhesion molecules, chemoattractant proteins and proinflammatory cytokines in the aorta, suppressed monocyte adhesion and transmigration across human umbilical vein endothelial cells, and suppressed proliferation and migration of A7r5 smooth muscle cells . Our data suggest that VAP-1 inhibition is a potential treatment for atherosclerosis. In addition, we also explored the physiologic role of VAP-1. In the literature, VAP-1 can enhance tissue glucose uptake in cell studies and normalize hyperglycemia in animal studies. However, serum VAP-1 concentration (sVAP-1) is higher in subjects with diabetes in cross-sectional studies. In a cohort study, we tested our hypothesis that sVAP-1 is increased in prediabetes to counteract hyperglycemia and is associated with incident diabetes negatively. We found that sVAP-1 was higher in subjects with prediabetes and increased during an OGTT. Fasting sVAP-1 was associated with the response of sVAP-1 during an OGTT. Besides, sVAP-1 was associated negatively with body mass index, waist circumference, abdominal visceral and subcutaneous fat areas, and serum high-sensitivity C-reactive protein concentration, and positively with plasma adiponectin concentration. After 4.7 years, 12.2% of subjects developed incident diabetes. High sVAP-1 predicted a lower incidence of diabetes, adjusted for age, gender, BMI, family history of diabetes, HbA1c, HOMA2-%B and HOMA2-IR. These findings provide evidence to support our hypothesis.