Prognostic relevance of blood pressure variability.

The possible prognostic relevance of different patterns characterizing blood pressure (BP) fluctuations during the day and night was first suggested in the late 1980s,1 based on observations made possible by 24-hour continuous intraarterial ambulatory BP monitoring in normotensive and hypertensive subjects. Since then, this issue has been repeatedly addressed in studies carried out either in animals and in humans and is still matter of debate. Evidence is available that in normal subjects 24-hour ambulatory BP recordings are characterized by a fairly smooth BP profile during daytime activities and by a significant BP decline during night sleep. Most, although not all, studies exploring the prognostic relevance of BP variations have shown that alterations in these 24-hour BP variability patterns are associated with hypertensive target organ damage and with a higher rate of cardiovascular events.2 This is the case for an increased amplitude and frequency of BP variations occurring during the daytime,3,4 the night-time,5 or the entire 24-hour period.6 It is also the case for alterations in the components of 24-hour BP variability related to the shift between wakefulness and sleep, namely the nocturnal BP decline and the morning BP surge. A reduced degree of nocturnal BP fall, and even more so the reverse pattern characterized by a BP increase at night, has been found to be associated with cardiac, cerebral, renal, and vascular damage. Several studies have also shown that lack of nocturnal dipping carries a higher risk of stroke,7–10 whereas conflicting findings were reported on the prognostic relevance of an excessive BP fall at night.11,12 Finally, data are available that a steeper morning BP surge is associated with a higher risk of cardiac and cerebrovascular events,11,13,14 although such an association was not found by other investigators.7 The article by Metoki et al published in this issue of Hypertension15 offers additional information in this field by exploring the risk either of ischemic and hemorrhagic cerebrovascular events, respectively, in association with different patterns of morning BP surge and nocturnal BP fall. This was done by analyzing the ambulatory BP recordings obtained in the general population of Ohasama, Japan, and by considering an average follow-up period of 10.4 years. Although the degree of nocturnal BP dipping and the magnitude of morning BP surge bore no relationship with total stroke incidence, a large morning BP surge, and extreme dipping (ie, a nocturnal BP fall larger than 20% of daytime levels) were both associated with a higher frequency of hemorrhagic stroke, while no dipping or inverse dipping were related to a higher frequency of cerebral infarctions. The study by Metoki et al thus provides additional support to the concept that, in the management of patients with arterial hypertension, care has to be paid not only to the reduction of mean BP levels by treatment, but also to the normalization of deranged BP variability patterns, such as those between day and night. In particular, the data provided by this and by other similar articles appear to suggest that inhibition of a steep morning BP surge by antihypertensive treatment might reduce the risk of events and, specifically, the risk of cerebrovascular hemorrhage. They also suggest that in nondipper patients or, even more so, in patients with an inversion of the physiological nocturnal BP dipping, improving the nocturnal BP decline through proper titration and scheduling of antihypertensive treatment might improve prognosis by specifically reducing the risk of ischemic stroke. These perspectives are of great potential interest. Their generalized extrapolation to patients’ management in clinical practice, however, must be considered with some caution. This is because of a number of problems still affecting the assessment of the clinical relevance of BP variability through the day and night (Table). First, there are differences between studies in the methods for the quantification of overall blood pressure variability. This is exemplified by the variable use of standard deviations of mean BP values or of coefficients of variations. It is also exemplified by some studies focusing on the whole 24-hour time, whereas others considered daytime or nighttime subperiods only. Second, different definitions of daytime and nighttime were often used. Some studies based their identification of day and night on information derived from patients’ logbooks; other studies adopted a fixed definition of daytime and nighttime. This was done either by considering all 24 recording hours (wide fixed criteria) or by skipping transitional periods between wakefulness and sleep (ie, by excluding a few hours in the evening and in the early morning during which different subjects were engaged in different behaviors [narrow fixed criteria]). Third, the changes in BP between day and night as well as those between night and day may not be reproducible in all subjects, a variability of at least 40% having been reported in the definition of night-time BP dipping pattern when comparing data from 2 consecutive ambulatory BP recordings.16 Fourth, as shown also by Metoki et al in their article, the degree of nocturnal BP fall is significantly related to the rate of morning BP rise, and such a correlation between 2 hemodynamic phenomena that are reported to have opposite prognostic value makes their clinical interpretation more The opinions expressed in this editorial are not necessarily those of the editors or of the American Heart Association. From the Department of Clinical Medicine, Prevention, and Applied Biotechnology, University of Milano-Bicocca, and II Cardiology Unit, San Luca Hospital, IRCCS, Istituto Auxologico Italiano, Milan, Italy. Correspondence to Gianfranco Parati, MD, Dept of Clinical Medicine, San Luca Hospital, Istituto Auxologico Italiano, Via Spagnoletto, 3, 20149 Milan; Italy. E-mail [email protected] (Hypertension. 2006;47:137-138.) © 2006 American Heart Association, Inc.