New research has suggested patients with occasional high blood pressure are at increased risk of stroke. Professor Peter Rothwell and Dr Alastair Webb discuss the implications for GPs
Current hypertension guidelines emphasise the importance of estimating each patient's average underlying blood pressure, ignoring intermittent high measurements, and recommending treatment only if the average of repeated measurements in clinic or on home monitoring is above specific treatment thresholds.1 However, new research has shown that increased variability in systolic BP across multiple clinic visits, and the maximum systolic BP reached, are strong predictors of an increased risk of stroke, independent of mean systolic BP .1,2 Indeed, episodic hypertension is associated with a higher stroke risk than persistent hypertension, despite a lower mean BP.2
It has also been shown that calcium channel blockers, and to a lesser extent non-loop diuretics, reduce visit-to-visit BP variability, whereas ß-blockers increase variability.2,3,4 These drug-class effects on variability account for the previously unexplained differences between drugs in their effectiveness in prevention of stroke. Although mean systolic BP and diastolic BP on multiple readings is still the main indication for BP-lowering medication, variability in BP must also now be taken into account in diagnosis, treatment and monitoring of hypertension.
How common is episodic hypertension?
Episodic hypertension is very common. In the UK-TIA trial for example – with a cohort of patients with previous TIA – only 12% of patients had stable hypertension (systolic BP consistently over 140mmHg), whereas 69% had episodic hypertension (some readings of 140mmHg or below and some over 140mmHg).1 The correlation between systolic BP at one visit and the next is therefore poor in all such cohorts. Similar variability is also seen prior to a first stroke, but episodic hypertension is usually left untreated – systolic BP not being raised on the requisite number of repeat visits, and lower readings probably being interpreted as indicating the likely usual BP. For example, of 150 consecutive patients on no BP-lowering drugs prior to a first stroke in the Oxford Vascular Study, 87% had at least one systolic BP of 160mmHg or above during the previous 10 years, but 69% of these also had systolic BPs of 130mmHg or below on at least two other visits.
In which groups of patients is BP variability most important?
Systolic BP variability and maximum BP reached are most strongly associated with risk of stroke and, less so, with coronary artery disease and heart failure. So they are likely to be of most importance in patients at high risk of stroke, either in primary prevention because of family history or other risk factors, or in secondary prevention after a previous TIA or stroke.
There is currently significant undertreatment of hypertension after TIA or stroke, despite evidence of benefit in patients with apparently normal baseline BP.5 But increased variability should also be taken into account more generally. For example, residual variability in systolic BP is a strong predictor of stroke in otherwise well-controlled hypertensive patients.
More research is required to determine the importance of variable BP in otherwise healthy middle-aged individuals, but it is noteworthy that in the recent studies of treated hypertensive or TIA patients, increased variability was most powerfully predictive of stroke in younger patients.2
How should episodic hypertension be diagnosed?
Average BP above treatment threshold is still the main indication for BP-lowering treatment. But treatment should also be considered in patients who have intermittent systolic hypertension even if the average systolic BP is below threshold.
A fixed definition of what is excessive visit-to-visit variability in BP – or what level and frequency of episodic hypertension should necessitate treatment – is difficult to determine. But it is also not necessarily helpful because the risk relations reported so far have shown continuous increases in risk of stroke with increasing variability and maximum BP reached. However, a level of variability of systolic BP that results in clinic measurements with a range that is consistently 40mmHg or more should certainly raise concern.
Similarly, patients with mean systolic BP below treatment thresholds, but with some resting clinic systolic BP measurements above 160mmHg or with any measurement of 180mmHg or more should be considered for treatment, irrespective of their mean systolic BP over multiple visits.
Other issues that still require more research include:
• How should variability determined using 24-hour ambulatory monitoring (ABPM) be taken into account? Initial analyses suggest that variability measured on ABPM is only partially correlated with visit-to-visit variability in clinic BP.2
• Does variability determined using home measurements of BP have similar prognostic value to visit-to-visit variability? If so, how frequently and for how long should home measurements be performed?
• How is visit-to-visit variability in BP related to short-term instability in BP, such as postural hypo- or hypertension?
For the time being, the focus in primary care should probably be to make use of the often substantial number of clinic BP measurements that have been recorded over the years in individual patients. Assessments of variability based on seven to 10 clinic measurements at different visits over months or years provide a reasonably reproducible estimate of variability.2 For new patients, or patients who have recently started or changed treatment, identification of episodic hypertension will require repeated clinic visits, perhaps supplemented by self-measurement at home and by 24-hour ABPM.
How does episodic hypertension differ from white-coat hypertension?
There is now evidence that white-coat hypertension is associated with an increased cardiovascular risk,6 but this is probably independent of the predictive value of visit-to-visit variability in clinic
BP. Initial analyses have shown that visit-to-visit variability is not correlated with the ‘white-coat effect' (the initially high systolic BP on first measurement in clinic), and visit-to-visit variability in clinic BP is non-situational (that is, it occurs despite the same clinic setting). White-coat hypertension is purely situational, being defined as a high BP in clinic and a more normal BP at home or on ABPM. Episodic high BP in clinic should certainly not be dismissed as white-coat hypertension.
How should episodic hypertension be treated?
There has been controversy for some years as to whether calcium channel blockers and diuretics are particularly effective at prevention of stroke, whereas ß-blockers are less effective than would be expected on the basis of their effects on mean BP. The debate has recently been substantially advanced by two new findings.
First, that calcium channel blockers – and to a lesser extent diuretics – significantly reduce visit-to-visit variability in BP, whereas ß-blockers increase it.3,4 Secondly, that these drug-class effects are highly correlated both within trials3 and across trials4 with effects on risk of stroke.
These new findings strengthen the rationale for a move away from ß-blockers as first-line drugs in treatment of hypertension, and highlight the effectiveness of calcium channel blockers in patients with particularly variable BP or patients at particular risk of stroke, such as patients with previous TIA or stroke.
The reduction in BP variability by calcium channel blockers appears to be present for both dihydropyridine and non-dihydropyridine subclasses and to be independent of half-life.4 More research is required into the effectiveness of different combinations of drug-classes, but preliminary evidence suggests that the benefit of calcium channel blockers is maintained in patients receiving other classes of drugs, although it is uncertain whether the combination of a calcium channel blocker and a thiazide has additive benefits in terms of reducing variability. Calcium channel blockers are, of course, less effective than other agents in preventing heart failure.
Professor Peter Rothwell is professor of clinical neurology and Dr Alastair Webb is academic clinical fellow in neurology at the John Radcliffe Hospital, Oxford
Competing interests: none declared