Treatment of Hypertension in the Elderly

Abstract

The aging of the United States population means that treatment of hypertension in the elderly is becoming an increasingly important issue, and the pathophysiology, consequences, and benefits of treatment are quantitatively different in older and younger people. There are several issues to be addressed, which include the significance of pulse pressure as a risk factor, the reliability of measurement of blood pressure in the elderly, choice of drugs and lifestyles, and management of the very old.

When plotted on a logarithmic scale, the relationship between risk and cardiovascular events is linear, but the slope is steeper for strokes than for heart attacks. ¹ One consequence of this is that the benefits of treating hypertension are greater for preventing stroke. Also, there is no flattening of the curve at low pressures, so that there is no obvious lower limit for blood pressure (BP) reduction. Systolic BP tends to increase progressively with age, but diastolic pressure does not, and after the age of 50 actually starts to fall. The explanation for this is increased stiffness of the arteries that usually accompanies aging, such that when the stroke volume is pumped into the less compliant arterial system of older people, there is a higher peak of systolic pressure. ² This is compounded by wave reflection due to increased pulse‐wave velocity, which adds a late systolic peak to the pressure wave.

There have been a large number of studies indicating that the best predictor of risk in the elderly patient is actually the pulse pressure, and that patients with low diastolic pressure may be at higher risk than patients with high diastolic pressure for the same level of systolic pressure. ³ This raises the issue of how we should treat a patient who has a BP of 170/50 mm Hg, because if we lower the systolic pressure, we will probably lower the diastolic pressure further, which perhaps would put the patient at increased risk, particularly for coronary events. Unfortunately, we do not at the present time have any drugs that will selectively lower systolic pressure, and what is needed is a drug that would primarily affect arterial stiffness. So, for practical purposes in making treatment decisions, we still have to go by systolic pressure and try to get that reasonably well controlled, and not pay too much attention to diastolic pressure in this elderly population.

THE BENEFITS OF TREATMENT

It is well established that treating hypertension in elderly patients reduces cardiovascular morbid events, often to an even greater extent than in younger patients, presumably because the older patients have a higher risk before treatment. The earlier studies recruited elderly patients with both systolic and diastolic hypertension, and their results have recently been reviewed in The Journal of Clinical Hypertension. ⁴ Three more recent placebocontrolled trials (Systolic Hypertension in the Elderly Program [SHEP], ⁵ Systolic Hypertension in Europe [Syst‐Eur], ⁶ and Systolic Hypertension in China [Syst‐China] ⁷ ) have recruited patients with isolated systolic hypertension, (an entry systolic pressure of at least 160 mm Hg, and a diastolic pressure <90–95 mm Hg), and although the treatment regimens were different, the results were almost superimposable, with 33%–42% reductions in stroke events, and 27%–30% reductions in coronary heart disease events in all three studies. The minimum age for entry in all three studies was 60, and the vast majority of patients were <80 years old.

THE IMPORTANCE OF OUT‐OF‐OFFICE BP MEASUREMENTS

The way in which BP is measured in elderly patients can have a huge impact on the treatment decision, because white coat hypertension is a significant problem in older patients. This point was well illustrated by the Syst‐Eur study, which recruited older patients who had a clinic systolic pressure >160 mm Hg. A subgroup of the patients underwent 24‐hour ambulatory monitoring at baseline ⁸ and, as shown in the Figure, the ambulatory BPs gave a much better prediction of risk than the clinic pressure. For the clinic pressure, which by definition starts at 160 mm Hg, there is a very small spread of risk at different levels of pressure. But the ambulatory pressure measurements (expressed as night, day, and average 24‐hour values) show a much steeper slope relating BP and risk. In addition, it is possible to identify a very high‐risk group, and a very low‐risk group with low ambulatory pressures who are the patients with white coat hypertension.

Figure.

Prediction of risk in elderly patients using clinic and ambulatory monitoring from the Systolic Hypertension in Europe study. SBP=systolic blood pressure. Adapted with permission from JAMA 1999;282:539–546. ⁸

Another important analysis of the Syst‐Eur data looked at the consequences of treating those patients who were recruited because they had high clinic BPs, but on ambulatory monitoring turned out to have white coat hypertension. ⁹ This was possible because the study was placebo‐controlled, and half the patients received active treatment (with a long‐acting dihydropyridine). In the subgroup of patients with sustained hypertension, the active treatment lowered both the clinic and the ambulatory BPs, whereas in the white coat hypertensives, only the clinic pressure was lowered. This lack of effect on the ambulatory pressure may be explained by the fact that it was normal to begin with. In the patients with sustained hypertension, treatment resulted in a significant reduction in the number of strokes, whereas in the white coat hypertensives, there was no significant treatment effect, although there were fewer strokes in the treated group. This lack of significance was because the number of strokes was so small in both the treated and untreated white coat hypertensives.

There are now a number of other prospective studies, showing that of all the measures of BP that we can obtain, it is the ambulatory pressure that predicts the risk best in the individual patient. ¹⁰ Medicare is now reimbursing elderly patients with suspected white coat hypertension for 24‐hour BP monitoring on the grounds that they are at lower risk and need to be treated less aggressively. To qualify for reimbursement, patients must have a diagnosis of suspected white coat hypertension, that is an office pressure above 140/90 mm Hg on several occasions, and measurements taken outside the office which are within the normal range, and the absence of target organ damage. Thus, if a patient has left ventricular hypertrophy, it is already clear that he or she is at risk, and ambulatory monitoring is not needed to make a treatment decision. The actual reimbursement levels are quite low, so that the technique is still not widely used, but we are using it in our clinic for clinical evaluation.

The unreliability of the conventional clinical measurements was nicely illustrated by a study conducted by general practitioners in England, where they took consecutive hypertensive patients and measured their BP by three different techniques. ¹¹ , ¹² These were the traditional clinic readings, home (or self) measurement, and ambulatory monitoring. On the grounds that the daytime ambulatory readings give the best prediction of risk, the other measurements were compared against this value and expressed as the white coat effect (WCE), which is defined as the difference between the clinic reading and the daytime average. The physician's readings gave a huge scatter of values for the WCE, but the average was 19/11 mm Hg. In contrast, readings taken by two nurses, also in the clinic, resulted in a WCE of only 5/6 mm Hg and 5/8 mm Hg (Table I). The home readings were of particular interest. When the patients took their readings at home the WCE was also 5/6 mm Hg, but when they brought their monitors into the clinic and took readings in the presence of the doctor, the WCE doubled, to 10/13 mm Hg. The message here is that it is the physician readings that are the outliers, and all the other measurements were much closer to the true daytime average measured with the ambulatory monitor.

Table I.

Comparison of Blood Pressure Measurements Made in Different Ways

BP Measurement Method	White Coat Effect* (mm Hg)
Physician	19/11
Nurse 1	08‐May
Nurse 2	06‐May
Patient at home	06‐May
Patient in clinic	13‐Oct
*Defined as the difference between the listed measurements and the daytime average measured by ambulatory monitoring. Data derived from BMJ. 2002;325(7358):254, 258–259. 11 , 12

The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC 7) ¹³ has acknowledged the use of self‐measurement, and it may be particularly useful for monitoring the response to antihypertensive therapy. Home readings have been shown to correlate more closely with changes of target organ damage during antihypertensive drug treatment than clinic readings. ¹⁴ There is some evidence that it may improve adherence to therapy, but the jury is still out on that.

Another potential use of self‐monitoring is to be able to record the early morning pressures. This has recently gained attention because of a study which looked at the predictive value of different components of the diurnal BP pattern for stroke. ¹⁵ The main finding was that the morning surge, which was the difference between the lowest pressure at night and the pressure during the first 2 hours after waking, had additional predictive value even after controlling for the average 24‐hour level. So patients who have very high BPs when they first wake up may be at increased risk. For patients who are monitoring their pressure at home, it may be important to pay particular attention to readings taken first thing in the morning.

THE GOALS OF TREATMENT IN THE ELDERLY

Getting older patients to the target pressure of 140/90 mm Hg is particularly difficult. Most of them have isolated systolic hypertension, and diastolic pressure is much easier to control than systolic. An analysis from the National Health and Nutrition Examination Survey (NHANES) study ¹⁶ looked at predictors of uncontrolled hypertension. One of the most important of these was being over the age of 65 ¹⁶ ; this was more significant than the patients' gender or race, and also how frequently they were seeing physicians. The good news is that several of the large intervention trials have shown that much better control rates (in the region of 60%–70%) ¹⁷ can be achieved in general practice by just following a definite protocol.

How far should BP be lowered in the elderly? The answer to this is not known, and as shown in Table II, some of the major trials that included older patients required a systolic pressure of at least 160 mm Hg to get into the trial and only achieved a final systolic pressure in the range of 140–150 mm Hg. ¹⁸ So while we say that the target systolic pressure is below 140 mm Hg, it seems reasonable to allow ourselves a little leeway and not be too concerned if a patient's systolic pressure is between 140–150 mm Hg when they come back to the clinic. While it may be true that if we lowered the pressure more we would prevent more strokes, and there are some intervention studies that hint that this could occur, ¹⁹ it remains unproven.

Table II.

Entry and Final Systolic Pressures in Four Trials of the Treatment of Hypertension in the Elderly

Trial	Entry BP (mm Hg)	Final BP (mm Hg)
HOT	170	140–144
EWPHE	183	149
SHEP	170	144
Syst‐Eur	174	151
BP=blood pressure; HOT=Hypertension Optimal Treatment randomized trial; EWPHE=European Working Party on High Blood Pressure in the Elderly; SHEP=Systolic Hypertension in the Elderly Program; Syst‐Eur=Systolic Hypertension in Europe Study; Data derived from QTM. 2000;93(4):203–205. 18

THE CHOICE OF ANTIHYPERTENSIVE DRUGS IN THE ELDERLY

Over the past 15 years or so, the sales of all classes of antihypertensive drugs have increased (with the possible exception of α blockers), so that it might be expected that control rates of hypertension would have improved. This has not happened to any great extent, which may mean that many patients are still not taking their medications regularly, that the medications are not working well, or that the doses are inappropriate. We should remember to ask our patients how regularly they are taking their medications.

Another problem is the limited efficacy of monotherapy. Some years ago, a study was carried out comparing different types of monotherapy, and it was found that most of the drugs tried (which included all the major classes except angiotensin receptor blockers [ARBs]) brought the pressure under control in only 50%–70% of patients, which is even less impressive when one considers that the placebo succeeded in 31%. ²⁰ The message here is that single‐drug therapy is rarely going to control a patient's BP. Unfortunately, we do not have any good way of predicting which patient is going to respond best to which type of medication. There is some suggestion that African Americans may respond better to diuretics and calcium channel blockers than to renin inhibiting drugs, because they tend to have lower renin levels, but there are so many exceptions to that rule that it is is not terribly helpful.

One of the interesting things about the recently released British guidelines for the management of hypertension is that they are recommending what is called the ABCD rule. ²¹ The letters correspond to the four major categories of antihypertensive drugs: A is for angiotensin blocking drugs (ARBs and angiotensin‐converting enzyme [ACE] inhibitors), B is for β blockers, C is for calcium channel blockers, and D is for diuretics. The AB drugs tend to work principally by blocking the renin‐angiotensin system, while the CD drugs tend to antagonize sodium retention. Thus, it is better to combine an AB drug with a CD drug, than an A with a B, or a C with a D. The proposal is that you start off with an A/B drug or a C/D drug, and if adequate control is not achieved, to add one from to the other pair. This rule was based on a study in which patients were systematically rotated between all four of the categories. ²²

The initial choice of antihypertensive drugs has received a great deal of attention as a result of the Antihypertensive and Lipid‐Lowering Treatment to Prevent Heart Attack Trial (ALLHAT) study results. ²³ This was the largest antihypertensive drug study ever performed, comparing an ACE inhibitor and a calcium channel blocker against chlorthalidone. Although the primary end point of the trial (heart attacks) was the same in all three groups, the official conclusion was that diuretics are to be preferred, partly because their use resulted in fewer strokes and heart failure admissions, and also because they are less expensive. However, the interpretation of the results is not straightforward, since the diuretic group had consistently lower BPs than the other groups throughout the trial. This may be because by design ALLHAT had a relatively high proportion of African Americans who, as mentioned above, tend to respond better to diuretics than to ACE inhibitors, and also because this was a relatively elderly population.

The ALLHAT publication was followed by an apparently contradictory finding from the Second Australian National Blood Pressure Study (ANBP2) trial ²⁴ that compared diuretics to ACE inhibitors in elderly Australian patients, and concluded that there was a slight (11%) and marginally significant reduction of events in the ACE inhibitor group. In this trial, BP changes were identical in the two groups. One of the unexplained findings was that the superiority of ACE inhibitors was seen only in men. This is anomalous, because no other study has shown such a gender difference with ACE inhibitors.

Can these two studies be reconciled? In ALLHAT, the diuretics were better than the ACE inhibitors for preventing heart failure, but this included hospitalizations for heart failure, where the main reason may have been sodium retention, which would be more likely to occur in the ACE inhibitor group since diuretics were withheld from them. There are other studies that show that ACE inhibitors prevent mortality from heart failure. ²⁵ The greater reduction of strokes than of heart attacks by diuretics in ALLHAT may have been partly due to the lower BP in the diuretic group. We know from results from other studies that, for the same reduction of pressure, there is more protection from strokes than heart attacks.

A recent analysis promoted the idea that diuretics may be more protective against strokes than ACE inhibitors. ²⁶ The two studies that have looked at the prevention of recurrent strokes are compatible with this idea. In the first, the Perindopril Protection Against Recurrent Stroke Study (PROGRESS), ¹⁹ patients were either treated with an ACE inhibitor (perindopril) or in combination with a diuretic (indapamide), and in the second Post‐Stroke Anti‐hypertensive Study (PATS) ²⁷ the treatment was indapamide vs. placebo. What this analysis appears to show is that for a similar reduction of systolic BP, there was a bigger reduction of stroke in the diuretic‐treated group (PATS) than in the ACE inhibitor group (PROGRESS), while in PROGRESS, the combination group, but not the ACE inhibitor monotherapy group, showed a very good effect on stroke. In addition, a meta‐analysis of the major trials found that ACE inhibitor‐based treatment was inferior to diuretic and β blocker‐based treatment at preventing strokes.

Another major landmark in the treatment of elderly patients was the Losartan Intervention for Endpoint Reduction in Hypertension Study (LIFE), ²⁸ which found that treatment based on an ARB (losartan) was superior at preventing stroke than β blocker‐based treatment. There has been some debate as to whether this difference occurred because ARBs are good at preventing stroke, or whether β blockers are ineffective. In fact, as shown in JNC VI, ²⁹ there is evidence from the older trials that β blockers are inferior to diuretics at preventing morbid events, and it has been argued that β blockers should no longer be drugs of first choice for the treatment of elderly hypertensive patients, unless there is some specific indication known, such as coronary heart disease.

One other finding is worth mentioning, and that is from Syst‐Eur, where it was found that treatment with a calcium channel blocker resulted in a dramatic (55%) reduction in new‐onset dementia. ³⁰

HYPERTENSION IN THE VERY OLD

A major unresolved issue is how to manage hypertension in the very old (patients ≥80 years old). One reason for not simply extrapolating results of intervention trials carried out in younger patients is that the natural history may be different in the very old. Such people are by definition survivors, and an observational study from Finland reported that in this population, individuals with lower BPs actually had a shorter survival than those with high pressures. ³¹ The major intervention trials conducted in the elderly did not include enough patients in this “very old” category to be able to make any statements as to how they should be treated. Fortunately, there is one study that is examining this question—the Hypertension in the Very Elderly Trial (HYVET). ³² The minimum age for inclusion is 80, and since it cannot be assumed that the benefits of lowering the BP in this group outweigh the harm, the study includes a no treatment group. There are two treatment arms, one using an ACE inhibitor, and the other a diuretic. The primary end point is a reduction in strokes, since this is the biggest problem in this population. Although the study is still ongoing, the results of the pilot study have been published ³² and indicate no effect of treatment on cardiovascular or total mortality, but a significant reduction of stroke events in the diuretic group. A similar trend was seen with the ACE inhibitor group, which is not yet significant. This is another argument in the very elderly for favoring the use of diuretics over ACE inhibitors, but we have to proceed cautiously here.

NON‐DRUG TREATMENT IN THE ELDERLY

We should not forget the benefits of non‐drug treatment in older people. Physical activity is critically important, not just for keeping the BP down, but also because of all the other bad things that happen to old people who are inactive. Walking is one of the best therapies for older people, ³³ whatever their medical condition. Older people are more likely to be salt‐sensitive than the young, but dietary advice in the elderly should focus on making sure that they are eating a balanced diet rather than simply a low‐salt one. Weight reduction is probably less of a problem in the very elderly, but the Dietary Approaches to Stop Hypertension (DASH) ³⁴ diet, which is high in minerals such as calcium, potassium, and magnesium may be particularly useful because many elderly patients are probably not getting enough calcium in their diet anyway. The effect of these interventions on BP is not necessarily additive, however. The recently published results of the study, ³⁵ which looked at the effects of different combinations of non‐drug treatments found that combining the DASH diet with other lifestyle changes had a very small incremental effect on BP. One reason for this may have been that, in this study, the DASH food was not provided already prepared, as had been done in the earlier DASH studies. ³⁶ Patients had to try to change their existing diet to the DASH program, which requires some degree of sophistication.

CONCLUSION

The number of elderly hypertensive patients is going to continue to increase, so we need to be better prepared to treat them appropriately. The first requirement is an accurate assessment of the BP, which means that we should no longer rely exclusively on traditional clinic measurements, which may be particularly misleading in older patients. Once the decision to start treatment is made, diuretics are definitely the drugs of first choice, but the majority of patients will need a combination of several different classes of medication. For patients >80 years of age, the need for BP reduction is not yet clearly established, although cautious diuretic‐based treatment would seem appropriate. Lifestyle changes are important at all ages, and are key changes in older people who measure their own BPs at home.