Cardiovascular disease is the leading cause of death and disability in diabetes. Previous studies have shown that diabetes can increase the risk of stroke by 2 to 5 times. Reduced glucose tolerance is an intermediate state between normoglycemia and diabetes mellitus, as evidenced by non-fasted blood glucose levels between 7.8 and 11.0 mmol/L. Reduced glucose tolerance is present in approximately one quarter of patients with a previous TIA or mini-ischemic stroke. It has been shown that reduced glucose tolerance increases the risk of stroke in patients with coronary artery disease, and it is inconclusive whether reduced glucose tolerance increases the risk of stroke in patients with a previous TIA or mini-ischemic stroke. For this reason, Sarah E et al. conducted a prospective study of 3127 patients with previous TIA or minor ischemic stroke who participated in the Dutch TIA trial over a period of 2.6 years (this article was published in the journal Stroke, 2006, no. 6). Three groups were divided: two groups were given different doses of aspirin (30 or 283 mg) and atenolol (50 mg), and the control group was given placebo. All patients were followed up to obtain symptoms, medical history, smoking history, medication use, perform CT examinations, and measure venous plasma glucose levels. Patients were divided into five groups according to their glucose, 11.1 mmol/L. Diabetic patients were classified as >11.1 mmol/L group with or without glucose-lowering drugs and insulin. RESULTS: The follow-up period was from the beginning of the study until the occurrence of stroke, myocardial infarction, death, or the end of follow-up. During 2.6 years of follow-up, 272 patients (9%) had a stroke and 200 patients (6%) had a myocardial infarction or cardiac death. The incidence of cardiovascular events in each group was examined using the Kaplan-Meyer curve, and the risk of stroke and the risk of myocardial infarction, cardiac death, and blood glucose levels (mean 6.0, standard deviation 2.2 mmol/L) in relation to the non-fasted state were analyzed using Cox relative risk regression, with adjustment for cardiovascular risk factors. No difference was observed between the different doses of aspirin (30 or 283 mg) and atenolol (50 mg) in the trial. A J-shaped relationship was found between blood glucose levels and stroke in the non-fasted state.