Aim. - To assess whether reverse-dipping status is associated with cardiovascular (CV) events such as CV death, myocardial infarction (MI) or stroke in diabetic patients with hypertension. Methods. - A total of 97 diabetic patients underwent their first ambulatory blood pressure monitoring (ABPM 1). "Reverse dippers" were defined as patients with a nighttime systolic and/or diastolic blood pressure (BP) greater than daytime systolic and/or diastolic BP. Other patients were called "others". A second ABPM (ABPM 2) was clone after a median delay of 2.6 years. Patients were then followed for a further 2.9-year median period (total median follow-up: 5.5 years). Results. - AfterABPM 1,CV events Occurred in 53% of the reverse dippers (n = 15) and in 29% of the others (n = 82). Kaplan-Meier curves showed significant differences between the two groups (P = 0.003). Mean nighttime systolic BP on ABPM1 was 148 +/- 23 mmHg and 142 +/- 19 mmHg in patients who did and did not experience a CV event, respectively. With Cox analysis adjusted for confounders, a 10 mmHg increase in nighttime systolic BP was associated with a 35% increase in the risk of a CV event (hazard ratio [HR]: 1.35, P=0.003). The HR for a CV event in reverse-versus nonreverse-dipping status was 2.79 (P=0.023). After ABPM 2, the relationship between the reverse-clipping status and occurrence of CV events was no longer evident (P=0.678). Nighttime systolic BP remained predictive of CV events (P=0.001). Conclusion. - These findings suggest that nighttime systolic BP per se appeared to be a stronger predictor of an excess risk of CV events compared with reverse-clipping status. (C) 2008 Elsevier Masson SAS. All rights reserved.