Prof. Denis Clement ,
The topic of blood pressure variability has gained new life in a number of recent publications. This paper will reviewits role in relation to the definition of blood pressure, the prognosis of hypertension and allude to the mechanisms involved in its variations.
All physicians will notice that blood pressure varies from one moment to another, over even very short periods of time (short term variations). This observation has forced investigators to record blood pressure repeatedly in well controlled situations in order to define blood pressure accurately- whether before or following interventions, whether by drugs or other means. The scale of these short-term variations will strike anyone as obvious if one realises that the difference there is between blood pressure values measured in one individual, over just a few minutes, using the same instrument, in a patient holding the same position, is within the same order of magnitude as the difference obtained as a result of the sought after effect of most antihypertensive drugs. Thus, when trying to determine the effect of a drug or a change in lifestyle on hypertensive patients, one cannot escape from the obligation of recording blood variations as well, to be sure that the observed changes are not just spontaneous variations in pressure.
As said above, the question of blood pressure variability is not new. In the early seventies already, James Conway wrote in an editorial on labile hypertension (1): “The definition of blood pressure has always been bedeviled by its variations”. In the early days of blood pressure monitoring, a symposium on this issue was organised and pioneers such as P. Sleight, G. Pickering, W. Birkehäger, E. Raftery participated (2). The problem was well understood by the participants of the symposium who discussed a series of techniques (invasive and non-invasive) to obtain objective documentation regarding the blood pressure variations in a given individual. Additionaly, a number of conclusions were formulated concerning the mechanism responsible for the variations (3). It was established that amplitude of variations is related to level of blood pressure; therefore blood pressure should not only be given as the standard deviation of the mean in absolute terms but also as variation coefficient (this is as a percentage of the level of blood pressure). Variability was shown to increase with age; there seemed - at least in these early studies, to be no relationship with heart rate variability. Lastly, no clear correlation could be found with the adrenergic nervous system as neither beta nor alpha adrenergic blockade decreased variability (in dogs, variations are linked to vagal tone, but this is less clear in humans). This symposium opened the way to the question of the value of longer term variations and their role in the definition of blood pressure which in turn, automatically lead to the introduction of longer term recordings. Nevertheless, although such longer (24 hours) recordings clearly showed superior correlation to prognosis than simple office pressure recordings (4), no attention was focused to the value of BPVar and prognosis.
Two early studies paved the way to incorporating BPVar as an entity that can impact prognosis. The first study (5) illustrated, in a group of 286 hypertensive patients followed over three years, that intima media thickness (recorded in the carotid arteries) is increased when systolic variability is higher compared to patients with similar mean blood pressure but lower variability. The other study (6) looked at a large group of 2012 subjects, randomly selected from the Monza population with a follow-up of over 148 months. Investigators reported that BPVar showed a positive correlation to cardiovascular mortality independently of 24 hour mean blood pressure values. A number of recent publications reinforced support for a possible impact of BPVar on prognosis. The best known study (7) was performed using a totally different group of patients, patients at risk for a cerebrovascular event and whose follow-up was carried out within the framework of a drug treatment program to prevent it. It came out that blood pressure variations defined as the difference between blood pressure measured at various visits, is a predictor of stroke and to a lesser extent, of coronary events and heart failure. Isolated systolic peaks seemed to play a more important role than had ever before been anticipated; results proved valid even after adjusting for absolute blood pressure levels. As shown in the early studies, no correlation to heart rate variations was shown. These results were confirmed shortly after with the data from the NHANES Study group (8). The obvious question that followed was whether the amplitude of variations could be influenced by therapy. The ASCOT trial (9) showed that variations could be decreased by calcium antagonists but not by beta blocking agents (which confirmed the early studies, see ref 3) nor by drugs acting on the RAA axis.
This new information is, undoubtedly, extremely interesting and stimulating. It is surprising how such a scientific message has remained unknown and needed so many years to unravel... Technical factors certainly have played a role in this, as the fact that all interest was focused on the question of the importance of office blood pressure compared to 24 hour recordings with respect to long term prognosis. It is important to understand the differences between recent studies and former approaches to BPVar. Recent studies are focused at the “inter visit” variations while all previous studies focused at the “intra visit” variations. It is not clear yet to what extent both approaches are similar, overlapping or largely different from one another. An important weakness certainly lies in the fact that recent study results (7-9) come from post-hoc analysis and not from analysis planned at the time of setting up the protocols for the studies. One should certainly be careful as well in the interpretation of the cause to effect correlation; for example, it may well be that the amplitude of variations is largely due to the degree of arterial stiffness, which would confirm the role of arterial stiffness in prognosis - as was shown by many investigators in the last years. Another interpretation would be that the occasional peaks are not part of what is regularly meant by “spontaneous variability” but rather, are the consequence of poor compliance with antihypertensive treatment; peaks could than just be the expression of the forgotten pill that day(s) and not a part of spontaneous variations. What to do with this new information? Should the 24 hour or home recordings and all techniques for recordings during daily life be forgotten? Certainly not! The value of both office and out-of-office blood pressure have been repeatedly demonstrated and is extremely well established. What the clinician should integrate in his or her daily activity is to record blood pressure repeatedly, at every visit; the clinician should not leave out the first nor the last figure of blood pressure or quickly make whatever sort of average value. The message here is that the clinician needs to record as many blood pressure values as possible, to allow for the determination of all “intra” and “inter” visit blood pressure values and their variations. Clinicians need to focus, by all means, on short and long term optimal blood pressure definitions and keep these values in the patients’ files. Blood pressure definitions should be made using well validated techniques at hand, thus avoiding altogether the problem that the so-called spontaneous variations would merely be due to technical failures. Lastly, as regards to the choice of antihypertensive therapy, one could get better help from calcium antagonists than from other antihypertensive classes. Obviously though, all of these recommendations need further confirmation.
Above all however, one should not forget that many of the recent data are original and quite stimulating but have been obtained from retrospective analysis. We are lacking well planned prospective studies, with set questions to be answered by the study contained within the protocol of the study, right from the beginning of the study. It is likely that such studies will get organised in the near future and will answer many questions remaining at the present time. The essential message does remain however, that all blood pressure values are important and that clinicians and investigators should record as many blood pressure values as possible using well validated techniques, and keeping them in their files hand, and ready for appropriate analysis.
1. Conway J. Labile Hypertension. Circulation research: 1970: 1 suppl 1: 43-47 2. Clement D. Blood Pressure Variability.1979. MTP Press Limited, Lancaster, England 3. Clement DL, Mussche MM, Vanhoutte G, Pannier R.Is blood pressure variability related to activity of the sympathetic system? Clin Sci (Lond). 1979 Dec;57 Suppl 5:217s-219s. 4. Clement D.L .et al. Prognostic value of ambulatory blood pressure. New England J.Medicine: 2003: 348:2407-15 5. Sander D. et al. Relationship between Circadian Blood Pressure Patterns and progression of early carotid atherosclerosis. A 3-year follow-up study. Circulation : 2000: 102: 1536-1541 6. Mancia G. et al. Long-term Prognostic Value of Blood Pressure Variability in the General Population. Results of the Pamela Study. Hypertension: 2007: 49: 1265-1270 7. Rothwell PM et al. Prognostic significance of visit-to-visit variability, maximum systolic pressure, and episodic hypertension. Lancet: 2010: March 13: 375: (9718): 895-905 8.Muntner P. et al. The relationship between visit-to-visit variability in systolic blood pressure and all-cause mortality in the General population. Hypertension: 2011: 57: 160-166 9. Rothwell PM. et al. Effects of beta blockers and calcium-channel blockers on within-individual variability in blood pressure and risk of stroke. Lancet Neurol. 2010: may 9 (5): 469-80 Epub 2010 Mar 11.
Our mission: To reduce the burden of cardiovascular disease
© 2018 European Society of Cardiology. All rights reserved