results of antihypertensive treatment trials in the elderly

ANTIHYPERTENSIVE TREATMENT TRIALS THE AMERICAN JOURNAL OF GERIATRIC CARDIOLOGY 2002 VOL. 11 NO. 1 41

When a physician encounters an elderly patientwith arterial hypertension, the medical deci-

sions depend on the answers to the following threequestions:

1. Is arterial hypertension a risk factor for car-diovascular disease in elderly patients?

2. Do pharmacologic and nonpharmacologic in-terventions in elderly patients with elevatedblood pressure decrease the incidence of car-diovascular events?

3. Is it possible to lower elevated blood pressurewithout negatively influencing the quality oflife of elderly hypertensives?

As reported by Fagard in this issue (pages24–27),1 a positive answer to the first questionhas been given. In this paper, we review thestudies providing evidence that blood pressurereduction is indeed associated with a decreasedincidence of cardiovascular events in elderlyhypertensives.

CLINICAL TRIALS OF ANTIHYPERTEN-SIVE TREATMENT IN THE ELDERLYSince the late 1950s, many studies have demonstrat-ed the benefit of blood pressure lowering in reducing

cardiovascular morbidity and mortality.1 When anti-hypertensive therapy was first introduced, emphasiswas placed on treatment of the most severe forms ofhypertension, but by the late 1960s, the positive re-sults of the first prospective, placebo-controlled, dou-ble-blind trial in patients with nonmalignanthypertension were published.2 Thereafter, numerousprospective, controlled intervention trials in hyper-tension have been conducted. A meta-analysis of 14major trials in patients with nonmalignant hyperten-sion indicated that antihypertensive treatment signifi-cantly reduced the incidence of both fatal andnonfatal strokes (by 42%) and fatal and nonfatal coro-nary heart disease events (by 14%), when comparedto placebo treatment.3

In trials completed before 1985, elderly hyper-tensive patients were either not included or wereonly a minor component of the population investi-gated. However, the results of a number of inter-vention trials specifically devoted to elderlyhypertensives have become available since 1985.These studies can be subdivided into: 1) three trialsdevoted to patients with isolated systolic hyperten-sion (SHEP, Syst-Eur, and Syst-China); 2) threestudies in patients with systolic-diastolic hyperten-

Results of Antihypertensive TreatmentTrials in the Elderly

Gastone Leonetti, MD; Alberto Zanchetti, MDFrom Istituto Auxologico Italiano, Istituto Scientifico San Luca, and Centro di Fisiologia Clinica e Ipertensione,Università di Milano, Ospedale Maggiore, Milan, ItalyAddress for correspondence/reprint requests: Prof. Gastone Leonetti, MD, Ospedale S. Luca Via Spagnoletto 3, 20149Milan, ItalyE-mail: [email protected]

The prevalence of arterial hypertension is age-dependent, and with the prolongation of life expectancythe number of elderly subjects with arterial hypertension is very relevant. Epidemiologic studies haveshown that arterial hypertension is a risk factor in elderly patients and therefore the physician mustknow if the pharmacologic and nonpharmacologic reduction of blood pressure values is associated witha corresponding decrease in systolic-diastolic or isolated systolic hypertension. Clinical trials haveshown that the lowering of blood pressure values is commonly associated with a decrease in cardiovas-cular events. So far, the reduction of blood pressure per se appears more relevant to the cardiovascularbenefit than does a particular class of antihypertensive agents. The benefit of antihypertensive treat-ment has been shown up to the age of 80 years, while there are no clear indications of a benefit in per-sons older than 80 years. While sufficient data suggest that a diastolic blood pressure between 80 and90 mm Hg is associated with a clear benefit in elderly patients, the data in support of a systolic reduc-tion below 140 mm Hg require further direct confirmation. (AJGC. 2002;11:41–47, 57) ©2002 CVRR, Inc.

sion (EWPHE, STOP-Hypertension, and STONE);3) two studies in patients with both forms of arterialhypertension (Coope and Warrender and MRC tri-als); and 4) two studies in which the benefits of dif-ferent antihypertensive regimens were compared(MRC and STOP-Hypertension 2). On the basis ofthe evidence provided by these trials, we shall dis-cuss in sequence: 1) the benefit of antihypertensivetreatment in the elderly; 2) the antihypertensiveagents to be prescribed to elderly hypertensives; 3)the upper age limit for pharmacologic interven-tion; and 4) the blood pressure goal of antihyper-tensive treatment in the elderly.

BENEFIT OF ANTIHYPERTENSIVETREATMENT IN THE ELDERLYEight studies have assessed the benefit of antihy-pertensive treatment in the elderly by comparingactive drug treatment with placebo administrationand monitoring cardiovascular events for a suffi-cient length of time. The characteristics of thesestudy populations, and the treatment regimenscompared, are listed in Table I. The effects of ther-apy on blood pressure, cardiovascular events, andmortality are presented in Table II.

ISOLATED SYSTOLIC HYPERTENSIONIn elderly patients with isolated systolic hyperten-sion, three studies have been published since 1990.

Systolic Hypertension in the Elderly Program(SHEP). The report of this study,4 which is a land-mark in the history of antihypertensive therapy research, appeared in 1991, after the first demon-stration, in 1967, by Veterans Administration re-search on the benefit of treating patients with severehypertension2 and, in 1970, of treating patients withmild to moderate hypertension.5 Indeed, the major-

ity of research on antihypertensive efficacy in pre-venting cardiovascular events had been centered onmild (90–105 mm Hg) or moderate (105–114 mmHg) diastolic hypertension in the aged adult. TheSHEP results concerning cerebrovascular and car-diac event prevention with antihypertensive therapyin subjects aged 60 or older with isolated systolic hy-pertension are important for at least two reasons:the relevance of the high percentage of the elderlypopulation with this type of hypertension and theemphasis placed, for the first time, on antihyperten-sive trials addressing reduction of systolic ratherthan diastolic blood pressure.

This study enrolled 4736 patients ≥60 years ofage, with systolic blood pressure of >160 mm Hgand diastolic blood pressure of <90 mm Hg. Thepatients randomly allocated to chlorthalidone re-ceived a low dose (12.5 mg/day) as first-step therapy;in nonresponders (systolic blood pressure remainedat >160 mm Hg and/or systolic blood pressure re-duction was <20 mm Hg), the chlorthalidone dosewas doubled and eventually combined with atenololor reserpine. The primary end point of the studywas stroke, while secondary end points were coro-nary and all cardiovascular events. As expected,blood pressure was reduced to a greater extent inactively treated patients and the differences betweenthe active and control groups were 11 mm Hg forsystolic and 3 mm Hg for diastolic blood pressure,both differences being statistically significant. Thegreater blood pressure reduction in the active treat-ment group was associated with a lower incidence oftotal strokes of 36% and of cardiovascular and coro-nary end points of 32% and 27%, respectively.

This was the first study to demonstrate the efficacyof antihypertensive drug treatment in lowering theincidence of cerebrovascular and coronary events inelderly patients with isolated systolic hypertension.

ANTIHYPERTENSIVE TREATMENT TRIALS42 THE AMERICAN JOURNAL OF GERIATRIC CARDIOLOGY 2002 VOL. 11 NO. 1

Table I. Principal Baseline Characteristics of Patients and Active Treatments

ACTIVE DRUGS

TRIAL PATIENTS (N) AGE (Y) BLOOD PRESSURE 1ST STEP 2ND STEP

(MM HG)

SHEP4 4736 71.6 170/77 Chlorthalidone AtenololSyst-Eur6 4695 70.2 174/85 Nitrendipine EnalaprilSyst-China7 2394 66.5 171/86 Nitrendipine CaptoprilEWPHE9 840 72.0 182/101 HCTZ + triamterene MethyldopaSTOP-Hyper10 1627 76.0 195/102 Diuretic or β blocker Diuretic or β blockerCoope11 884 68.7 196/99 Atenolol DiureticMRC12 4396 70.0 185/91 Diuretic or β blocker Diuretic or β blockerSTONE13 1632 66.0 168/100 Nifedipine Captopril

Trial acronyms are defined within text; HCTZ=hydrochlorothiazide


Systolic Hypertension in Europe (Syst-Eur). Thissecond study on isolated systolic hypertension in el-derly patients6 included 4695 patients ≥60 years ofage, with systolic blood pressure of 160–219 mm Hgand diastolic blood pressure of <95 mm Hg. Theywere randomly allocated to either a dihydropyridinecalcium antagonist, nitrendipine, 10–40 mg/day, withpossible addition of enalapril and hydrochloro-thiazide, or to matching placebo. The primary endpoint was the incidence of stroke, and secondary endpoints were cardiac and all cardiovascular events.

At a median follow-up of 2 years, systolic bloodpressure had fallen by 13 and 23 mm Hg and di-astolic blood pressure by 2 and 6 mm Hg in theplacebo and active treatment groups, respectively.The mean difference between the groups was10.1/4.5 mm Hg. As a consequence of the lowerblood pressure values in the actively treatedgroup, the incidence of stroke, the primary endpoint of the trial, was reduced by 12%, and therates of all cardiovascular end points and of car-diac complications were reduced by 31% and26%, respectively.

Systolic Hypertension in China (Syst-China).This was the first trial on isolated systolic hyper-tension in China.7 The inclusion criteria were thesame as in Syst-Eur,6 with two differences: 1) afterstratification for center, gender, and previous car-diovascular complications, the patients were alter-natively assigned (rather than randomlyallocated) to active drug or placebo; and 2) thestarting agent was the same (nitrendipine), butthe second-step agent was captopril instead ofenalapril. The active drug was given to 1253 pa-tients, and 1141 received matching placebo.

There were no statistically significant differ-ences between the two groups in mean age or sys-

tolic and diastolic blood pressure at entry. Themedian follow-up was 3.0 years. The primary endpoint was the incidence of stroke. The systolicblood pressure was reduced by 20.0 and 10.9 mmHg in the active and placebo groups, while dias-tolic values decreased by 5.0 and 1.9 mm Hg, re-spectively. There was a statistically significantreduction of 38% in the incidence of stroke andof 37% for all cardiovascular events with activetreatment vs. placebo.

Combined Results. Staessen and Wang8 pooledthe results of these three trials and calculated theodds ratios for active vs. placebo treatments forthe major cardiovascular end points. Overall, ac-tive treatment, compared with placebo, reducedall-cause mortality by 17%, cardiovascular mortal-ity by 25%, all cardiovascular end points by 32%,total strokes by 37%, and myocardial infarctions,including sudden death, by 25%: all of the reduc-tions were statistically significant.

SYSTOLIC AND DIASTOLIC HYPERTENSIONSystolic/diastolic hypertension has been evaluatedin two trials.

European Working Party on High Blood Pres-sure in the Elderly trial (EWPHE). This double-blind, placebo-controlled trial9 included patientsover the age of 60 years with an average diastolicblood pressure in the range of 90–119 mm Hg andsystolic values between 160 and 239 mm Hg. Eighthundred forty patients were randomly allocated tothe active drug (a combination of hydrochloro-thiazide plus triamterene) or to matching placebo.If predeterminated blood pressure goal levels werenot attained, α-methyldopa was added in the ac-

Table II. Effects of Treatment on Blood Pressure and Cardiovascular Events

TRIAL PLACEBO BP ACTIVE BP ∆ BP STROKE CHD CV EVENTS CV MORTALITY TOTAL MORTALITY

SHEP4 155/71 144/68 –11/–3 –36%* –27% –32%* –20% –13%Syst-Eur6 161/83 151/78 –10/–5 –42%* –26%* –31%* –27% –14%Syst-China7 159/84 151/81 –9/–3 –38%* –37% –37%* –39%* –39%*EWPHE9 167/90 148/85 –19/–5 –32% –47%* –38%* –27%* –9%STOP-Hyper10 186/96 167/87 –19/–8 –47%* –13% –40%* NA –43%*Coope11 180/89 162/78 –18/–11 –42%* +3% NA –22% –3%MRC12 168/85 152/76 –16/–9 –25%* –19% –17%* –9% –3%STONE13 156/90 146/87 –9/–5 –57%* –6% –60%* –26% –45%*

BP=blood pressure; CHD=coronary heart disease; CV=cardiovascular; NA=not available; *active significantlylower than placebo Trial acronyms are defined within the text.


tively treated group and matching placebo in thecontrol group. The follow-up period averaged 4.6years; the primary end point was total, cardiac,and cardiovascular mortality.

The two groups were well matched for age,gender, and blood pressure values at randomiza-tion. During the trial, the blood pressure was con-sistently and significantly reduced in theactive-drug group compared to the placebogroup, and at the end of the study there was a sta-tistically significant difference of 19/5 mm Hg be-tween the two. According to intention-to-treatanalysis, total mortality was not significantly re-duced in the actively treated group (–9%), al-though the reductions in cardiac (–38%) andcardiovascular (–27%) mortality did reach statisti-cal significance.

Swedish Trial in Old Patients With Hyper-tension (STOP-Hypertension). This trial10 inves-tigated the effects of antihypertensive treatment inpatients in the age range of 70–84 years—olderthan those in previous trials. Blood pressure crite-ria for inclusion in the trial were systolic pressureof 180–230 mm Hg and diastolic pressure of atleast 90 mm Hg, or diastolic blood pressure of105–120 mm Hg, regardless of the systolic values.Patients (n=1627) were randomly allocated to anactive treatment based on thiazide diuretics or βblockers or to matching placebo. The primary endpoint was the combination of stroke, myocardialinfarction, and cardiovascular death.

There were no statistically significant differ-ences in baseline characteristics between the ac-tive and control groups. The mean age, asexpected from entry criteria, was 76±4 years—well above that of the other trials in the elderly—and the average blood pressure at randomizationwas 195/102 mm Hg. The total duration of fol-low-up was 65 months and the average time inthe study was 25 months.

Active treatment reduced blood pressure signifi-cantly more than did placebo; the difference be-tween the two groups was 19.5/8.1 mm Hg.Compared with placebo, active treatment signifi-cantly reduced the incidence of the combined endpoint by 40%, and even total mortality, althoughnot an official end point, was noted to be signifi-cantly reduced (–43%) in the active-drug group.

SYSTOLIC AND/OR DIASTOLIC HYPERTENSIONThree trials have been devoted to a populationwith systolic and/or diastolic hypertension.

Coope & Warrender Trial in Elderly Hyper-tensives. Coope and Warrender11 conducted a ran-domized trial of arterial hypertension treatment in884 patients aged 60–79 years at the onset of thestudy, with the aim of determining whether pharma-cologic intervention could reduce the incidence ofstroke or coronary events or affect cardiovascular ortotal mortality. At that time, there was no evidenceconcerning the benefit of treating elderly patients.

Blood pressure of ≥170/105 mm Hg was a crite-rion for entry. The patients were randomized50:50, without stratification, to either the β blockeratenolol (the diuretic bendrofluazide could beadded if systolic and/or diastolic blood pressuredid not decrease) or placebo. The mean baselineblood pressures were 198/100 and 196/97 mm Hgin the active-drug and placebo groups, respective-ly, and the difference between the two groups atthe end of follow-up, which averaged 4.4 years, was18/11 mm Hg. The incidence of stroke was signifi-cantly reduced by 42% in treated patients, whilethe rates of myocardial infarction and total mortal-ity were unaffected.

Medical Research Council (MRC) Trial inElderly Hypertensive Adults. This trial12 com-pared active and placebo treatment in 4396 pa-tients aged 65–74 years, with mean systolicpressure of 160–209 mm Hg and mean diastolicpressure of <115 mm Hg recorded during an 8-week run-in period without antihypertensiveagents. The patients were randomly allocated toone of two active treatment regimens based on ei-ther a diuretic or a β blocker, or to placebo. Theywere followed single-blind for an average of 5.8years. The primary end points were stroke, coro-nary events, and mortality from all causes.

The patients’ characteristics at entry were compa-rable in all three treatment groups. The mean agewas 70 years and average blood pressure 183/91 mmHg in men and 184/90 mm Hg in women. Systolicand diastolic blood pressure fell in all groups, withgreater, and similar, blood pressure reduction in thetwo active-treatment groups. Compared with theplacebo group, the diuretic- and β blocker-treatedpatients combined had a statistically significant re-duction in stroke (–25%) and all cardiovascular events(–17%) and a nonsignificant reduction in coronaryevents (–19%). However, when the active treatmentswere analyzed separately, stroke, coronary complica-tions, and all cardiovascular complications were sig-nificantly reduced only in the diuretic-treated group;the β blocker group showed no significant reductionin the combined end points.


Shangai Trial of Nifedipine in the Elderly(STONE). This single-blind trial13 included 1632Chinese hypertensives with an age range of 60–79years and systolic blood pressure of 160–220 mmHg and/or diastolic pressure between 96 and 125mm Hg. They were alternately allocated to long-acting nifedipine or to matching placebo; in non-responders (systolic blood pressure of 160 mm Hgand/or a diastolic value of >90 mm Hg), captopriland hydrochlorothiazide (or matching placebo)were added. The mean follow-up was 30 months.The clinical events included stroke, congestiveheart failure, myocardial infarction, severe ar-rhythmia, and sudden death (considered com-bined cardiovascular events) and other events,such as hospitalization, cancer, and severe illness.

The mean age was 66±5 years and the averagebaseline blood pressure was 169/98 mm Hg; therewas no statistically significant difference in any ofthe principal baseline characteristics between thetwo treatment groups, including blood pressures(168/99 mm Hg in the nifedipine-treated patientsand 169/97 mm Hg in the control group). After amean follow-up of 30 months, systolic and diastolicblood pressures in the active-treatment group werereduced significantly more than in the placebogroup (systolic blood pressure, –21.6 vs. –12.3 mmHg; diastolic blood pressure, –13.1 vs. –7.6 mmHg, respectively). Relative risks were significantlyreduced—by 62% for combined cardiovascularevents and by 59% for all clinical events.

In conclusion, trials of antihypertensive drugtherapy in the elderly have consistently indicatedbenefit and no evidence of harm, whether patientshave both systolic-diastolic hypertension or whenthey have isolated systolic hypertension.

CHOICE OF ANTIHYPERTENSIVEAGENTAlthough the recent guidelines of the JointNational Committee on Prevention, Detection,Evaluation, and Treatment of High Blood Pressure(JNC)14 and of the World Health Organization/International Society of Hypertension Committee(WHO/ISH)15 have many points in common, the in-dications for the pharmacologic approach to treat-ment differ to some extent. The JNC based itsrecommendations on the results of available trialsand recommended diuretics or β blockers for elder-ly patients with systolic-diastolic hypertension, anddiuretics or dihydropyridine calcium antagonists inelderly patients with isolated systolic hypertension.On the other hand, the WHO/ISH guidelines sug-gest that all available antihypertensive agents

(diuretics, β blockers, calcium antagonists, an-giotensin-converting enzyme [ACE] inhibitors, α1blockers, and angiotensin II antagonists) can beconsidered first-choice treatment in all hyperten-sive patients, including the elderly, bearing in mindthe specific indications and contraindications. Inaddition, an indirect comparison16 of the benefitsobtained in nondiabetic patients on diuretic-basedtherapy in the SHEP trial and in those on calciumantagonist-based therapy in the Syst-Eur trial re-vealed entirely superimposable results (Figure 1).

After the publication of both major guidelines,the results of an additional trial in elderly hyper-tensives, the STOP-Hypertension 2 trial,17 were re-ported. These investigators compared the effectson cardiovascular morbidity and mortality of con-ventional therapy, i.e., therapy based on diureticsor β blockers, and of newer antihypertensiveagents, i.e., ACE inhibitors and calcium antago-nists of the dihydropyridine group.

In this prospective, randomized trial, 6614 hy-pertensives aged 70–84 years (≥180 mm Hg sys-tolic and ≥105 mm Hg diastolic, or both) wererandomly assigned to conventional antihyperten-sive drugs (atenolol, metoprolol, pindolol, or hy-drochlorothiazide plus amiloride) or to newerdrugs (enalapril, lisinopril, felodipine, or isradi-pine). Combined fatal stroke, fatal myocardial in-farction, and other fatal cardiovascular eventsconstituted the primary end point.

Patient characteristics did not differ among thethree groups (mean age, 76 years; mean supineblood pressure, 194/98 mm Hg). Blood pressurewas decreased similarly in all treatment groups.The primary end point of fatal stroke, fatal my-ocardial infarction, and other fatal cardiovascularevents occurred in 19.8% of conventionally treat-

Figure 1. Differences in all deaths and cardiovascularevents between active-drug treatment and placebo ad-ministration in nondiabetic hypertensive patients in theSHEP and Syst-Eur trials.


ed patients and in 19% of the patients treatedwith newer drugs. No significant differencesamong treatment groups were found when com-bined fatal and nonfatal cardiovascular eventswere compared. Even when the ACE inhibitorand calcium antagonist groups were separatelycompared with conventional drug groups, therewere no clinically or statistically significant differ-ences in the incidence of fatal or combined fataland nonfatal cardiovascular events.

An indirect comparison in elderly patients withdiabetes mellitus suggests that calcium antago-nist-based treatment may be superior to diuretic-based therapy. Indeed, active treatment appearedto induce greater reduction in all cardiovascularend points in the subset of diabetic patients in theSyst-Eur trial16 (in which treatment was startedwith a dihydropyridine) than in the subset of dia-betic patients in the SHEP trial18 (in which activetreatment was based on a diuretic) (Figure 2).However, a recent subgroup analysis of theSTOP-Hypertension 2 trial19 showed a similar re-duction in the primary end points in elderly hy-pertensive patients with diabetes mellitus,irrespective of the drugs used. There was a trendtoward a greater benefit with ACE inhibitors inprevention of coronary events, but the power ofthe subgroup analysis was not sufficient to pro-vide definitive results.

AGE AND ANTIHYPERTENSIVE THERAPYThe randomized, controlled trials of antihyper-tensive treatment in the elderly have shown bene-fits comparable to those in younger ormiddle-aged hypertensives, but, as the baselinecardiovascular risk is higher in the elderly, theabsolute benefit of treatment (expressed as num-ber of events prevented per 1000 patient-years) isdefinitely higher in the elderly. It is not clear,however, if these benefits are also present in thevery elderly, i.e., patients over age 80, who havebeen a minority in the different trials; no singletrial has had the power to answer this question.Gueyffier et al.20 performed a meta-analysis ofdata from subgroups of elderly patients includedin randomized, controlled trials in order to assessthe evidence for and against antihypertensivetreatment in persons over age 80. The primaryoutcome of the meta-analysis was fatal and nonfa-tal stroke, and secondary end points were totaland cardiovascular death, fatal and nonfatalmajor coronary and cerebrovascular events, andheart failure. The author selected four over-80

subgroups from double-blind trials4,6,10,19 andthree from single-blind trials11,21,22 comprising atotal of 1670 patients. The data indicate thattreatment prevented 33% of strokes (the primaryend point). Non-uniform results were obtainedfor secondary end points: although the rates ofmajor cardiovascular events and heart failurewere significantly reduced by 22% and 39%, re-spectively, there was a nonsignificant trend to-ward reduction in major coronary events (22%)and a nonsignificant 6% increase in total mortali-ty. When the effect of treatment on fatal and non-fatal stroke was analyzed separately, thetreatment benefit was limited to nonfatal strokes.Overall, the results of this meta-analysis suggest atreatment benefit for the primary outcome ofstroke and for the two secondary outcomes ofmajor cardiovascular events and heart failure.However, more definitive information will beavailable when an ongoing trial that includes pa-tients aged 80 years or older (Hypertension in theVery Elderly Trial [HYVET])23 is completed.

RELATIONSHIP BETWEEN MORTALITYAND TREATED BLOOD PRESSURESome retrospective analyses of intervention stud-ies24,25 have suggested a J-curve relationship betweenthe risk of myocardial infarction and treated bloodpressure. However, none of these studies was place-bo-controlled and other long, placebo-controlled in-tervention trials have either not confirmed26–28 or notreported29 this J-shaped relationship.

In a post-hoc analysis of the EWPHE trial,30 theincidence of mortality was evaluated separately forthe active-drug (n=352) and placebo (n=339)groups according to the distribution, in tertiles, ofon-treatment blood pressure. The two groups were

Figure 2. Differences in all deaths and cardiovascularevents between active-drug treatment and placebo ad-ministration in diabetic hypertensive patients in the SHEPand Syst-Eur trials


continued on page 57

similar in terms of age, gender, systolic and dias-tolic blood pressure, and percentage with cardio-vascular complications at randomization.

From this retrospective analysis, it appears that inelderly hypertensives given active treatment, totalmortality had a U-shaped relationship with on-treat-ment systolic blood pressure, with a nadir at about150 mm Hg, whereas total mortality increased gradu-ally with decreasing diastolic blood pressure, from theupper tertile of 98 mm Hg. However, a U curve be-tween mortality and diastolic blood pressure, with anadir at 95 mm Hg, was also found in the patientstaking placebo, and this suggests that the increasedmortality in the lower tertile of the active-treatmentpatients may not have been drug-induced. The in-creased mortality is likely to have been the expressionof some deterioration in general health, as suggestedby the concomitant decrease in body weight and he-moglobin concentration.

There is no doubt that conclusive inferences can-not be drawn from retrospective analyses, but, on thebasis of the results obtained in the eight trials so farcompleted, in elderly hypertensives an average sys-tolic blood pressure reduction to the range of140–160 mm Hg and an average diastolic bloodpressure between 80–90 mm Hg can be recom-mended. In the Hypertension Optimal Treatment(HOT) Study,31 about one third of the more than18,000 hypertensive patients enrolled were olderthan 65 years, and it was found that the optimalblood pressure for the lowest incidence of cardiovas-cular events was about 138 mm Hg for systolic andabout 83 mm Hg for diastolic pressure. There wasno significant rise in cardiovascular end points in hy-pertensive patients with blood pressure treated tolower levels. The intention-to-treat analysis showed asimilar pattern in the incidence of cardiovascularevents in the adults and in the older patients, sug-gesting optimal blood pressure reductions are simi-lar and independent of age.

CONCLUSIONThe aim of this review of the therapeutic trials wasto answer the second question presented in the in-troduction: whether drug-induced reduction ofhigh blood pressure in elderly individuals can sig-nificantly reduce the incidence of hypertension-re-lated cardiovascular events.

The eight studies analyzed are predominantlyrandomized, placebo-controlled, parallel-grouptrials and fulfill the requirements of “evidence-based medicine.” All of the trials are concordant inshowing that the greater blood pressure reductionsobtained in treated patients, relative to those on

placebo, does not cause harm to elderly hyperten-sives and indeed decreases the incidence of fataland nonfatal cardiovascular, cerebrovascular, andcardiac events. Stroke is the most frequent andthreatening complication of hypertension in the el-derly, and in all of the trials, the reduction instroke incidence has been consistently statisticallysignificant in the treated patients. Therefore, thereply to the question is positive: the pharmacologiclowering of hypertension in elderly persons im-proves their prognosis.

A few decades ago, the pharmacologic arma-mentarium for lowering blood pressure was verylimited, and severe adverse events were frequent.Now, however, several different classes of antihy-pertensive agents can be used, all of which are veryeffective and have a good profile of safety and tol-erability. The randomized, controlled trials thathave included elderly hypertensives have em-ployed diuretics, β blockers, dihydropyridines, cal-cium antagonists, and ACE inhibitors, and there isno doubt that their use in the elderly is justified bymedical evidence. However, monotherapy can nor-malize blood pressure in only about 40%–50% ofhypertensive patients and therefore a combinationof two or more drugs is often required to achiverecommended blood pressure goals.

Bearing in mind that the human lifespan is be-coming longer and longer, and that arterial hyper-tension is an age-dependent phenomenon, ournext inquiry is whether there is an upper age limitto the benefit of antihypertensive treatment. Onlya small number of patients 80 years or older wereadmitted to previous trials, and subgroup analysesof patients in that age group are not sufficientlypowered to provide definite conclusions. The sub-group meta-analysis of very elderly hypertensivepatients suggests a benefit in this population, butthe final conclusion must probably await the resultsof the ongoing HYVET trial.

Both the JNC and the WHO/ISH guidelinesrecommend lowering blood pressure in elderly hy-pertensives to below 140/90 mm Hg. While we be-lieve that there are sufficient data that a diastolicpressure between 80 and 90 mm Hg is associatedwith a clear benefit in elderly patients, the data insupport of a systolic pressure reduction below 140mm Hg require further direct confirmation.

REFERENCES

1 Fagard RH. Epidemiology of hypertension in the el-derly. Am J Geriatr Cardiol. 2002;11:24–27.

2 Veterans Administration Cooperative Study Groupon Antihypertensive Agents. Effects of treatment on

morbidity in hypertension. Results in patients withdiastolic blood pressure averaging 115 through 125mm Hg. JAMA. 1967;202:1028–1034.

3 Collins R, Peto R, MacMahon S, et al. Blood pres-sure, stroke, and coronary heart disease. Part 2.Short-term reductions in blood pressure: overview ofrandomised drug trials in their epidemiological con-text. Lancet. 1990;335:827–838.

4 SHEP Cooperative Research Group. Prevention ofstroke by antihypertensive drug treatment in olderpersons with isolated systolic hypertension. Final re-sults of the Systolic Hypertension in the ElderlyProgram (SHEP). JAMA. 1991;265(24):3255–3264.

5 Veterans Administration Cooperative Study Groupon Antihypertensive Agents. Effects of treatment onmorbidity in hypertension II: results in patients withdiastolic blood pressure averaging 90 through 114mm Hg. JAMA. 1970;213:827–838.

6 Staessen JA, Fagard R, Thijs L, et al. Randomiseddouble-blind comparison of placebo and active treat-ment for older patients with isolated systolic hyper-tension. Lancet. 1997;350:757–764.

7 Liu L, Wang JG, Gong L, et al. Comparison of activetreatment and placebo in older Chinese patientswith isolated systolic hypertension. J Hypertens.1998;16:1823–1829.

8 Staessen JA, Wang JG. Benefit of antihypertensivetreatment in older patients with isolated systolic hy-pertension. Eur Heart J. 1999;11(suppl P):3–8.

9 Amery A, Brixko P, Clement D, et al. Mortality andmorbidity results from the European Working Partyon High Blood Pressure in the Elderly trial. Lancet.1985;1:1349–1354.

10 Dahlof B, Lindholm LH, Hansson L, et al. Morbidityand mortality in the Swedish Trial in Old Patientswith Hypertension (STOP-Hypertension). Lancet.1991;338:1281–1285.

11 Coope J, Warrender TS. Randomised trial of treat-ment of hypertension in elderly patients in primarycare. BMJ (Clin Res Ed). 1986;293:1145–1151.

12 MRC Working Party. Medical Research Council trialof treatment of hypertension in older adults: princi-pal results. BMJ. 1992;304:405–412.

13 Gong L, Zhang W, Zhu Y, et al. Shanghai trial ofnifedipine in the elderly (STONE). J Hypertens.1996;14:1237–1245.

14 Joint National Commitee. The sixth report of theJoint National Committee on prevention, detection,evaluation, and treatment of high blood pressure.Arch Intern Med. 1997;157:2413–2446.

15 Guidelines Subcommitee. 1999 World HealthOrganization-International Society of HypertensionGuidelines for the Management of Hypertension. J Hypertens. 1999;17:151–183.

16 Tuomilehto J, Rastenyte D, Birkenhager WH, et al.Effects of calcium-channel blockade in older patientswith diabetes and systolic hypertension. SystolicHypertension in Europe (Syst-Eur) Trial Investigators.N Engl J Med. 1999;340:677–684.

17 Hansson L, Lindholm LH, Ekbom T, et al.

Randomised trial of old and new antihypertensivedrugs in elderly patients: cardiovascular mortality andmorbidity in the Swedish Trial in Old Patients withhypertension 2 study. Lancet. 1999;354:1751–1756.

18 Curb JD, Pressel SL, Cutler JA, et al. Effect of diuret-ic-based antihypertensive treatment on cardiovascu-lar disease risk in older diabetic patients withisolated systolic hypertension. Systolic Hypertensionin the Elderly Program Cooperative ResearchGroup. JAMA. 1996;276:1886–1892.

19 Lindholm LH, Hansson L, Ekbom T, et al.Comparison of antihypertensive treatments in pre-venting cardiovascular events in elderly diabetic pa-tients: results from the Swedish Trial in Old Patientswith Hypertension-2. STOP Hypertension-2 StudyGroup. J Hypertens. 2000;18:1671–1675.

20 Gueyffier F, Bulpitt C, Boissel JP, et al. Anti-hypertensive drugs in very old people: a subgroupmeta-analysis of randomised controlled trials. IN-DANA group. Lancet. 1999;353:793–796.

21 Perry HM Jr, Smith WM, McDonald RH. Morbidity andmortality in the Systolic Hypertension in the ElderlyProgram (SHEP) pilot study. Stroke. 1989;20:4–13.

22 Casiglia E, Spolaore P, Mazza A, et al. Effect of twodifferent therapeutic approaches on total and cardio-vascular mortality in a Cardiovascular Study in theElderly (CASTEL). Jpn Heart J. 1994;35:589–600.

23 Bulpitt CJ, Fletcher AE, Beckett N, et al. The Hyper-tension in the Very Elderly Trial (HYVET). Protocolfor the main trial. Drugs Aging. 2001;18:151–166.

24 Cruickshank JM. Coronary flow reserve and the Jcurve relation between diastolic blood pressure andmyocardial infarction. BMJ. 1988;297:1227–1230.

25 Beevers DG. Overtreating hypertension. BMJ.1988;297:1212.

26 Management Commitee. The Australian therapeutictrial in mild hypertension. Lancet. 1980;1:1261–1267.

27 Management Commitee of the Australian TherapeuticTrial in Mild Hypertension. Untreated mild hyperten-sion. Lancet. 1982;1:185–191.

28 IPPPSH Collaborative Group. Cardiovascular risk andrisk factors in a randomized trial of treatment based onthe beta-blocker oxprenolol: the InternationalProspective Primary Prevention Study in Hypertension(IPPPSH). J Hypertens. 1985;3:379–392.

29 In: Miall WE, Greenberg G, eds. Mild Hypertension: IsThere Pressure to Treat? Cambridge, UK: CambridgeUniversity Press; 1987:101–108.

30 Staessen J, Bulpitt C, Clement D, et al. Relation be-tween mortality and treated blood pressure in elder-ly patients with hypertension: report of theEuropean Working Party on High Blood Pressure inthe Elderly. BMJ. 1989;298:1552–1556.

31 Hansson L, Zanchetti A, Carruthers SG, et al., for theHOT Study Group. Effects of intensive blood-pressurelowering and low-dose aspirin in patients with hyper-tension: principal results of the Hypertension OptimalTreatment (HOT) randomised trial. Lancet.1998;351:1755–1762.


results of antihypertensive treatment trials in the elderly

Documents