Objective: To assess whether modifications in the nighttime blood pressure fall caused by antihypertensive treatment predict the regression of end-organ damage of hypertension.
Methods: The analysis was performed in patients with essential hypertension and echocardiographically detected left ventricular hypertrophy involved in the SAMPLE study. For each patient, ambulatory blood pressure monitoring and echocardiographic determination of left ventricular mass index were performed at the end of a 4-week wash-out pretreatment period, after 3 and 12 months of treatment with lisinopril or with lisinopril plus hydrochlorothiazide and after a final 4-week placebo period. For each ambulatory blood pressure monitoring the 24 h average, daytime average (0600-2400 h), night-time average (2400-0600 h) and day-night difference was computed. The percentages of dipper and non-dipper patients (i.e. the patients with night blood pressure falls greater and less than 10% of the daytime average, respectively) were also computed.
Results: The reproducibility of the day-night difference was low, both for comparison of the pretreatment and final placebo periods (n = 170) and for comparison of the third and the 12th month of treatment (n = 180). The reproducibility of the dipper-non-dipper dichotomy was also low, 35-40% of patients becoming non-dippers if they were dippers and vice versa, both with and without treatment The changes in left ventricular mass index after 12 months of treatment were significantly (P<0.01) related to the changes in 24 h, daytime and night-time blood pressure (r always > 0.33), but this was not the case for the treatment-induced modification of the day-night difference (r= -0.03 and -0.008 for systolic and diastolic blood pressures, respectively).
Conclusions: Our results show that day-night blood pressure changes and the classification of patients into dippers and non-dippers are poorly reproducible over time. It also provides the first prospective evidence that treatment-induced changes in day-night blood pressure difference are not related to treatment-induced regression of left ventricular mass index, thus having a limited clinical significance.