The smoothness index: a new, reproducible and clinically relevant measure of the homogeneity of the blood pressure reduction with treatment for hypertension

Objective To introduce a new method, the smoothness index, for assessing the homogeneity of 24 h blood pressure reduction by antihypertensive treatment and to compare it with the trough :peak ratio; and to assess the ability of both indices to predict a reduction in the left ventricular mass index induced by treatment. Patients and methods In 174 patients with essential hypertension and left ventricular hypertrophy, enrolled in the Study on Ambulatory Monitoring of Pressure and Lisinopril Evaluation (SAMPLE), aged 20-65 years, we measured clinic blood pressure, 24 h ambulatory blood pressure and the left ventricular mass index (echocardiography) before and after treatment with lisinopril at 20 mg with the addition of 12.5 or 25 mg hydrochlorothiazide as needed to reach a sufficient brood pressure reduction. The following parameters were computed for systolic and diastolic ambulatory blood pressure: (1) hourly and 24 h blood pressure averages (+/- SD) at baseline and after 3 and 12 months of treatment; (2) hourly blood pressure changes from baseline after 3 and 12 months of treatment, and their average (+/- SD) over 24 h; (3) the trough:peak ratio after 3 and 12 months of treatment; and (4) the smoothness index after 3 and 12 months of treatment Similar calculations were also performed at the end of a final study month during which active treatment was withdrawn and placebo was substituted (n =164). Results The smoothness index for systolic and diastolic ambulatory blood pressure computed after 3 months of treatment was more closely related to its corresponding values after 12 months of treatment than the trough:peak ratio values computed after the same time periods were (r=0.68 versus 0.38 for systolic and 0.68 versus 0.42 for diastolic blood pressure, respectively). The smoothness index showed an inverse correlation with the 24 h standard deviation of systolic and diastolic blood pressure (r=-0.25 and -0.16, P<0.01 and < 0.05, respectively, for 12 months of treatment), while the trough:peak ratio did not (r=-0.01 to -0.12, NS). A treatment-induced reduction in the left ventricular mass index was related to the smoothness index for systolic and diastolic blood pressure (r=-0.35 and -0.32, P< 0.001 with 12 months of treatment), but not to the corresponding trough:peak ratios. Conclusions The smoothness index identifies the occurrence of a balanced 24 h blood pressure reduction with treatment and correlates with the favourable effects of treatment on left ventricular hypertrophy better than the commonly used trough : peak ratio. J Hypertens 1998, 16:1685-1691 (C) Lippincott Williams & Wilkins.