The importance of instability and visit-to-visit variability of blood pressure

Abstract

Evaluation of: Rothwell PM, Howard SC, Dolan E et al. Prognostic significance of visit-to-visit variability, maximum systolic blood pressure, and episodic hypertension. Lancet 375, 895–905 (2010).

Spontaneous fluctuations of blood pressure from visit to visit have usually been disregarded as a trivial factor that confounds the ‘true’ associations of an individual’s long-term average blood pressure with disease. The paper under evaluation shows that visit-to-visit blood pressure variability is an independent predictor of future cardiovascular events in hypertensive patients and in subjects surviving a transient cerebral ischemia. Episodic elevations of blood pressure in nonhypertensive subjects seem to carry similar adverse prognostic significance. There is some evidence that different blood pressure-lowering drug classes may differ in their effects on visit-to-visit blood pressure variability, but these findings need to be confirmed in further studies.

Keywords::

• ambulatory blood pressure
• blood pressure
• blood pressure variability
• prognosis

The adverse cardiovascular consequences of hypertension are believed to depend primarily on absolute blood pressure (BP) values, and it is currently accepted that average BP is the main target for preventing cardiovascular morbidity and mortality in hypertensive subjects. The clinical value of BP fluctuations around the mean, and especially of the unsystematic BP variations from visit to visit, which are usually recorded during a prolonged clinical observation of months or years, has received little attention thus far.

This article evaulates a recent paper by Rothwell et al.[1] and a number of related papers [2–4], which questioned the axiom that average BP is the most important (or even the only) factor to be considered in the clinical management of hypertensive patients, and raised the issue of the importance of ‘random’ fluctuations of BP over long periods of observation.

Method & results

Visit-to-visit BP variability

The authors conducted a post hoc analysis of the data of the UK Transient Ischemic Attack (TIA) aspirin trial, in which 1324 subjects with previous transient cerebral ischemia received at least seven office BP measurements, once every 4 months. A large variability of BP was found within a single subject from one visit to the next (r² = 0.25–0.35), even without changes in antihypertensive treatment regimen. Such within-subject variability was a major component of the group BP variability at each follow-up visit, accounting for 41.5% of the variance in group BP at each visit. In other words, in a given population, the within-visit between-subject BP variability (expressed as standard deviation) is accounted for not only by different average BP values, but – to a substantial extent – also by within-subject BP variability over time.

In the UK-TIA study, a total of 104 patients had a subsequent stroke. Visit-to-visit systolic BP variability predicted future stroke, and its predictive value:

• • Was stronger than that of average systolic BP (multivariate-adjusted hazard ratio [HR], top vs bottom decile HR: 4.37; 95% CI: 2.73–76.99 vs HR: 2.44; 95% CI: 1.56–53.82);

• • Was independent of the effect of average systolic BP (top vs bottom decile adjusted HR: 3.27; 95% CI: 2.06–5.21);

• • Was confirmed both in patients receiving and not receiving antihypertensive treatment;

• • Increased with the number of BP readings included in the analysis as a measure of estimate precision;

• • Was not confirmed for diastolic BP variability.

The aforementioned findings were confirmed in an analysis of the data from three further randomized controlled trials of patients with previous transient cerebral ischemia or ischemic stroke (European Stroke Prevention Study [ESPS]-1, n = 2500; Dutch TIA, n = 3150; and the subgroup from the Anglo–Scandinavian Cardiac Outcomes Trial-Blood Pressure Lowering Arm [ASCOT-BPLA] with previous TIA or stroke, n = 2011), in which BP was measured during clinical visits every 3–6 months over a period of a few years. In an analysis of 18,530 hypertensive subjects from the ASCOT-BPLA trial who had two or more scheduled follow-up visits from 6 months onwards, visit-to-visit BP variability was a strong predictor of both stroke and coronary events, while average systolic BP was a weaker predictor of both.

Effects of treatment on visit-to-visit BP variability

The ASCOT-BPLA study randomized high-risk hypertensive patients to receiving an atenolol- or amlodipine-based antihypertensive regimen. In a post hoc analysis of the study results, the authors found that visit-to-visit systolic BP variability was greater in the atenolol group than in the amlodipine group.

Maximum systolic BP & episodic severe hypertension

In the UK-TIA group, maximum systolic BP predicted stroke independently of average systolic BP (multivariate-adjusted HR, top vs bottom decile HR: 15.01; 95% CI: 6.56–34.38). In the same cohort, patients who were not persistently hypertensive but had episodic severe hypertension (systolic BP >180 mmHg) at one or more of the BP visits had a higher risk of stroke than did those with stable hypertension (age- and sex-adjusted HR: 3.58; 95% CI: 1.58–8.10), despite having lower average systolic BP.

Other findings

24-h ambulatory BP monitoring was performed yearly in a subgroup of 1905 participants enrolled in the ASCOT-BPLA study. In these subjects, daytime BP variability (expressed as standard deviation of daytime systolic BP) correlated with visit-to-visit office systolic BP variability (r = 0.29–0.38; both p < 0.001), and both measures predicted vascular events independently of daytime average systolic BP, although visit-to-visit BP variability was somewhat superior in this regard.

Expert commentary

Blood pressure fluctuates around average values, both over the short term and over longer periods, its fluctuations being the result of a complex interaction between external environmental stimuli and the response of cardiovascular control mechanisms. Day–night and minute-to-minute BP variability, assessed mainly through 24-h ambulatory BP monitoring, have both been recognized as important cardiovascular risk markers in hypertension [5,6]. Aside from circadian and infradian BP variability, a substantial variation of BP exists from visit to visit when a subject is observed over days, weeks or months with home measurements or repeated clinical visits, even in the absence of treatment or in the presence of a steady therapeutic regimen.

Up to now, visit-to-visit BP variability had been mostly dismissed as a ‘background noise’ that dilutes the prognostic effects of average BP (the so-called ‘regression dilution bias’), and which must be neutralized by appropriate statistic techniques in order to appreciate the ‘true’ associations of the usual BP (i.e., an individual’s long-term average BP) with disease [7]. The demonstration that visit-to-visit BP variability, maximum systolic BP and episodic hypertension carry independent prognostic information has the potential of modifying our current understanding of the importance of BP as a cardiovascular risk factor, although confirmation in further studies is clearly needed.

The study results may also have implications for antihypertensive treatment. In the ASCOT-BPLA study, an amlodipine-based regimen was associated with a lower in-treatment visit-to-visit systolic BP variability than an atenolol-based regimen. In another analysis of the ASCOT-BPLA study [2], in which amlodipine was significantly better than atenolol at preventing stroke, the superiority of amlodipine was not attributable to the small between-group difference in average systolic BP, but could be explained by the reduction in visit-to-visit BP variability. These data are in agreement with the findings of the Medical Research Council trial, in which atenolol-treated subjects had a higher stroke risk in the first 2 years compared with subjects receiving placebo and a diuretic, and this excess risk was in parallel with an increased visit-to-visit variability in systolic BP. Subsequent temporal trends in BP variability during follow-up in the atenolol group correlated with trends in stroke risk [2].

In an analysis of 398 intervention BP-lowering trials [3], calcium antagonists reduced between-subject BP variability, expressed as the ratio between in-treatment BP variance and baseline BP variance (‘variance ratio’), while b-blockers, ACE inhibitors and ARBs increased the variance ratio. In a meta-analysis of 21 prospective event-based studies, treatment-induced reduction of variance ratio predicted a lower risk of stroke, but not of myocardial infarction, and such benefit was additional to those conferred by average BP reduction [3]. The authors take this as indirect evidence of BP variability as an important target for treatment, and a further mechanism through which antihypertensive drugs reduce stroke risk independently of their effect on average BP. An obvious limitation in interpreting these data, however, is that between-subject BP variability is only an approximate surrogate of visit-to-visit BP variability. The relationship between the two measures is rather weak [1], and the existence of drug-specific effects on visit-to-visit BP variability remains to be demonstrated.

Other limitations of the study also need to be addressed. While the mechanistic connections linking average BP and cardiovascular disease are well known, there are few data regarding the factors that determine visit-to-visit BP variability, and the mechanisms that could underlie its relationship with cardiovascular outcomes remain hypothetical, although potential mechanisms include baroceptor dysfunction, increased arterial stiffness, and the risk of cerebral ischemia in the presence of elevated BP variability and instability.

The study by Rothwell et al. is based on a post hoc analysis of data from studies in which BP was measured in a nonstandardized way, and often only once at every visit [1]. The hypothesis cannot be excluded that a higher visit-to-visit BP variability simply means inadequate compliance to BP-lowering treatment, which has been shown to be a major risk factor for future cardiovascular events in hypertensive subjects [8]. However, compliance was very high in the ASCOT-BPLA cohort, and similar results were found in untreated as well as in treated subjects.

Overall, the study by Rothwell et al. advances the hypothesis that spontaneous oscillations of systolic BP from visit to visit, as well as BP instability and isolated episodes of severe hypertension in otherwise normotensive subjects, should not be dismissed as trivial findings, but may carry important prognostic information, especially for the risk of stroke [1]. This hypothesis is substantiated by evidence from post hoc analyses and meta-analyses of clinical trials [1–4], but confirmation in other large databases is needed before implementing it in daily clinical practice.

Five-year view

Does the study by Rothwell et al.[1] and the related papers [2–4] represent the beginning of a new era in our understanding of the cardiovascular effects of BP and its fluctuations? Should indications for defining hypertension and for starting antihypertensive treatment be modified according to these new findings? Not yet, given the above limitations in the study design and the fact that visit-to-visit BP variability is difficult to measure in daily clinical practice, in which patients are usually treated with different drug regimens that change frequently according to clinical needs. However, the new findings are provocative enough to stimulate, in the next few years, a re-analysis of the many available databases from large-scale observational studies and randomized clinical trials. If the hypothesis of Rothwell et al.[1] is confirmed in such studies, this will probably lead to major changes in our understanding of hypertension as a cardiovascular risk factor.

Key issues

• • Variability of systolic blood pressure (BP) from visit to visit is an important predictor of future cardiovascular disease in people with previous cerebrovascular disease and in hypertensive subjects.

• • The prognostic value of visit-to-visit BP variability is independent from, and additional to, that of average BP.

• • An amlodipine-based regimen might reduce visit-to-visit BP variability more effectively than an atenonol-based regimen, and this might partly explain its stronger protective effect against stroke.

Financial & competing interests disclosure

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.