Risk and benefit of treatment of isolated systolic
hypertension in the elderly: evidence from the Systolic
Hypertension in Europe Trial
Hilde Celis, MD, Robert H. Fagard, MD, PhD, Jan A. Staessen, MD, PhD, and
Lutgarde Thijs, BSc
The Syst-Eur trial investigated whether active treatment
starting with the dihydropyridine calcium channel blocker
(CCB) nitrendipine, could reduce the cardiovascular
complications of isolated systolic hypertension (ISH) in the
elderly. The intention-to-treat analysis showed that active
treatment improved outcome. The per-protocol analysis largely
confirmed these results. The effect of treatment on total and
cardiovascular mortality might be attenuated in very old
patients. Further analysis also suggested benefit in those
patients who remained on nitrendipine monotherapy. Active
treatment was more beneficial in patients with diabetes as
compared with those without diabetes at entry and reduced
the incidence of dementia by 50%. Analyses of data from the
Ambulatory Blood Pressure Monitoring (ABPM) Side Project
suggested that most of the benefit of treatment was seen in
patients with a daytime systolic BP ⱖ 160 mm Hg. Finally, a
meta-analysis partly based on Syst-Eur data showed that in
older hypertensive patients pulse pressure and not mean
pressure is the major determinant of cardiovascular risk.
Curr
Opin Cardiol 2001, 16:342–348 © 2001 Lippincott Williams & Wilkins, Inc.
The presence of isolated systolic hypertension (ISH)
rises curvilinearly with age. Among septagenarians it av-
erages 8% and beyond 80 years it rises to more than 25%
[2]. Isolated systolic hypertension thus affects a large
proportion of the elderly population. Against this back-
ground, the European Working Party on High Blood
Pressure in the Elderly started the placebo-controlled
double-blind Syst-Eur (Systolic Hypertension in Eu-
rope) trial [3]. In 1991 the Systolic Hypertension in the
Elderly (SHEP) trial showed that diuretic-based treat-
ment prevented stroke, myocardial infarction, and con-
gestive heart failure [4]. However, because of the re-
maining uncertainties about the treatment of ISH in the
elderly [5], the Syst-Eur trial continued after the publi-
cation of the SHEP results [4]. Furthermore, the contro-
versy on the role of calcium channel blockers (CCBs) as
first-line antihypertensive agents [6–9], highlighted the
lack of evidence regarding the reduction of cardiovascu-
lar risk by these agents. The Syst-Eur trial stopped on
February 14
th
, 1997, after the second of four planned
interim analyses, because the primary endpoint of a sig-
nificant benefit for stroke was reached [3].
Patients and methods
The protocol of the Syst-Eur trial, described in detail
elsewhere [3], was approved by the Ethics Committees
of all participating centers. Eligible patients had to be
60 years or older and have a sitting systolic BP of 160 to
219 mm Hg with sitting diastolic BP below 95 mm Hg
during a run-in phase on single-blind placebo. After
stratification by center, sex, and previous cardiovascular
complications, patients were randomly assigned double-
blind treatment with active treatment or placebo. Active
treatment was initiated with nitrendipine (10–40 mg per
day), if necessary combined or replaced by enalapril (5–
20 mg per day) or hydrochlorothiazide (12.5–25 mg per
day) or both drugs. In the control group matching place-
bos were used. The study medications were stepwise
titrated and combined to reduce the systolic BP by 20
mm Hg or more to below 150 mm Hg. Patients with-
drawing from double-blind treatment remained in open
follow-up.
Overview of previously published
Syst-Eur results
Of 4695 randomized patients, 2398 were randomly as-
signed to active treatment [1,10•,11,12,13••,14,15••,16•].
This review was written for the Systolic Hypertension in Europe Trial Investigators,
Hypertension and Cardiovascular Rehabilitation Unit, Department of Molecular
and Cardiovascular Research, Faculty of Medicine, University of Leuven K.U.
Leuven, Leuven, Belgium. A complete list of the investigators appears in reference
[1].
Correspondence to Hilde Celis, MD, Studiecoördinatiecentrum, Laboratorium
Hypertensie, Campus Gasthuisberg, Herestraat 49, B-3000 Leuven, Belgium;
E-mail: hilde.celis@med.kuleuven.ac.be
Current Opinion in Cardiology 2001, 16:342–348
342
In the intention-to-treat analysis [1], the between-group
difference in BP was 10.1 mm Hg systolic and 4.5 mm
Hg diastolic at the median follow-up of 2.0 years. Car-
diovascular mortality was slightly lower on active treat-
ment (−26%, P = 0.08), but all-cause mortality was not
significantly changed (−13%, P = 0.28) [10]. Active treat-
ment reduced the incidence of total stroke by 42%
(P = 0.002) and of nonfatal stroke by 45% (P = 0.004)
[10]. All fatal and nonfatal cardiac endpoints decreased
by 25% (P = 0.03) and all fatal and nonfatal cardiovascular
endpoints by 30% (P < 0.001) [10]. The incidence of fatal
and nonfatal cancer (−12%, P = 0.42) and of bleeding
(excluding cerebral and retinal hemorrhages; −9%,
P = 0.75) was not different in the two treatment groups
[10]. In terms of absolute benefit, treating 1000 elderly
patients with ISH for 5 years could prevent 29 strokes or
53 major cardiovascular events.
Subgroup analysis [11] showed that male sex, previous
cardiovascular complications, older age, higher systolic
BP, and smoking at randomization were positively and
independently correlated with cardiovascular risk. The
relative benefit of active treatment was different accord-
ing to sex or to the presence of cardiovascular complica-
tions at entry. Furthermore, for total (P = 0.009) and
cardiovascular (P = 0.09) mortality, the benefit of anti-
hypertensive treatment weakened with advancing age
(Fig. 1). For total mortality (P = 0.05), it increased with
higher systolic BP at entry (Fig. 1) and for fatal and
nonfatal stroke (P = 0.01), it was most evident in non-
smokers (92.5% of all patients).
The results of the intention-to-treat [1,10] and per-
protocol [11] (including only the endpoints occurring
during the double-blind phase) analysis were largely
similar. Active treatment reduced total mortality by 26%
(P = 0.05); similar but nonsignificant trends were ob-
served for cardiac and cerebrovascular mortality. Cardio-
vascular, cardiac, and cerebrovascular events declined by
respectively, 32% (P < 0.001), 26% (P = 0.05), and 44%
(P = 0.004). In terms of absolute benefit, the per-protocol
analysis suggested that treating 1000 patients for 5 years
would prevent 24 deaths, 29 strokes, 25 cardiac end-
points, or 54 major cardiovascular events.
The relative benefits of antihypertensive treatment in
the Syst-Eur trial were largely similar to those of other
trials in older patients with combined systolic and dia-
stolic hypertension [17–22] or with isolated systolic hy-
pertension [4,23,24].
In view of the persistent controversy about the possible
adverse effects of CCBs [8,9,25–28], the question was
raised whether treatment with nitrendipine alone would
also influence outcome. Further analyses [12] also sug-
gested similar benefit in the patients who remained on
nitrendipine monotherapy. In these analyses, 1327 pa-
tients who remained on single nitrendipine treatment
throughout the whole trial, were matched by sex, age,
previous cardiovascular complications, and systolic BP at
entry, with an equal number of placebo patients from the
control group. Compared with this matched control
group, treatment with nitrendipine reduced cardiovascu-
lar mortality by 41% (P = 0.05), all cardiovascular end-
points by 33% (P = 0.01), fatal and nonfatal cardiac end-
points by 33% (P = 0.05) and fatal and nonfatal heart
failure by 48% (P = 0.05).
These data from the Syst-Eur trial invalidate circumstan-
tial evidence based on case-controlled and observational
studies [6–9,29–33], which according to the investigators’
interpretation left a margin of uncertainty regarding the
occurrence of potentially dangerous side effects. Several
studies investigated the effects of dihydropyridine CCBs
in Chinese hypertensive patients [23,24,34–36]. Al-
though some of these trials used unorthodox designs
[34,35], they also demonstrated a positive influence on
outcome.
At randomization, 492 patients (10.5%) had diabetes
mellitus [13]. After adjustment for possible confounders,
active treatment reduced all-cause mortality by 55%, car-
Figure 1. Risk according to age and initial systolic blood
pressure
Adjusted relative hazard rates of total and cardiovascular mortality according to
age and initial systolic blood pressure. The hazard rates (placebo/active
treatment), calculated by intention-to-treat, are presented as continuous risk
functions with 95% confidence intervals. P values refer to the interaction terms
between treatment and independent predictor variable. Published with
permission [11].
ISH in the elderly Celis et al. 343
diovascular mortality by 76%, all cardiovascular end-
points by 69%, fatal and nonfatal stroke by 73% and all
cardiac endpoints by 63% in the group of diabetic pa-
tients. In the nondiabetic patients, all cardiovascular
endpoints were reduced by 26% and all fatal and nonfatal
strokes by 38%. On active treatment, the reductions in
total mortality (P = 0.04), cardiovascular mortality
(P = 0.02) and all cardiovascular endpoints (P = 0.01)
were significantly larger in diabetic patients than in non-
diabetic patients.
These data from the Syst-Eur trial [13,37] were the first
to prove that antihypertensive drug treatment starting
with a dihydripyridine CCB was particularly beneficial in
diabetic patients. These results contradicted the sugges-
tion that some (second-generation) dihydropyrine CCBs
might be harmful, particularly in hypertensive patients
with diabetes mellitus [29–33,38,39].
Systolic hypertension is associated with an increased risk
of dementia in elderly people. The primary hypothesis
of the Syst-Eur Vascular Dementia Substudy, was that a
reduction in BP would protect against vascular dementia
[14]. Two thousand four hundred and eighteen patients
were enrolled [15]. At the median follow-up of 2.0 years,
active treatment reduced the incidence of dementia by
50% (P = 0.05) from 7.7 to 3.8 cases per 1000 patient-
years. Active treatment prevented mainly Alzheimer’s
dementia (8 versus 15 cases), but also vascular (0 versus
2) and mixed (3 versus 4) dementia. According to the
intention-to-treat analysis, treating 1000 elderly patients
with ISH for 5 years might prevent 19 cases of dementia.
The prevention of degenerative dementia was somewhat
unexpected, although recent studies indicate that vascu-
lar factors, in particular hypertension, might also play a
role in the development of degenerative dementia [40].
An alternative explanation, although still unproven,
could be that lipophilic CCBs, which cross the blood-
brain barrier and bind to brain receptors located in areas
affected by Alzheimer’s disease, might confer a specific
neuroprotective effect [41,42].
Overview of recent Syst-Eur results
Response to antihypertensive therapy in patients
with sustained and nonsustained isolated systolic
hypertension (ISH)
Patients with nonsustained hypertension (also called
white coat or isolated clinic hypertension) have normal
ambulatory blood pressure (ABP), but elevated clinic
blood pressure (CBP) [43–45]. Target organ damage [46–
50] and cardiovascular risk [49,51,52] is reported to be
lower in patients with nonsustained as compared with
patients with sustained (ABP and CBP both elevated)
hypertension, but data on nonsustained hypertension in
elderly patients with ISH are generally lacking.
Therefore, the data from the Ambulatory Blood Pressure
Monitoring (ABPM) Substudy [53,54] of the Syst-Eur
trial were analyzed. The objectives of this analysis were
to evaluate the consequences of nonsustained systolic
hypertension as well as the impact of antihypertensive
treatment. Patients enrolled in the ABPM Substudy
were classified according to baseline daytime systolic
ABP in 1 of 3 subgroups: nonsustained ISH (< 140 mm
Hg), mild sustained ISH (140 to 159 mm Hg), and mod-
erate sustained ISH (ⱖ 160 mm Hg).
At baseline, diastolic daytime and systolic and diastolic
nighttime and clinic BPs were higher in sustained than in
nonsustained hypertensives. Patients with nonsustained
ISH had smaller ECG voltages than patients with mild
and moderate sustained ISH. The differences in ECG
voltages between these three groups remained signifi-
cant after controlling for systolic CBP, but significance
was lost after adjusting for daytime systolic ABP.
During follow-up, active treatment reduced daytime and
nighttime ABP significantly in the patients with sus-
tained ISH but not in patients with nonsustained ISH.
CBP, however, decreased significantly in the three sub-
groups. Active treatment reduced ABP and CBP more
than placebo in patients with mild and moderate sus-
tained ISH. By contrast in patients with nonsustained
ISH, the changes in ABP between the treatment groups
were not significantly different (except for daytime
SBP), whereas the changes in CBP were also more pro-
nounced on active treatment as compared with placebo.
Patients with nonsustained ISH had a lower incidence of
stroke (P < 0.05) and of cardiovascular events (P = 0.01)
during follow-up than patients with sustained ISH. Ac-
tive treatment significantly reduced (P < 0.05) the ECG
voltages in patients with sustained ISH, but not in these
with nonsustained ISH. The influence of active treat-
ment on the incidence of stroke (P < 0.05) and cardio-
vascular events (P = 0.06) (Fig. 2) and on ECG voltages
was significantly more beneficial than that of placebo
only in patients with moderate sustained ISH.
These results confirm previous studies in other hyper-
tensive populations [49,51,52,55,56], which suggested
that white coat hypertension is associated with a better
outcome [49,51,52] or that the predictive value of ABP
persists after controlling for CBP [55,56]. A previous
analysis of data from the ABPM Substudy, in which BP
was treated as a continuous variable, showed that the
systolic ABP was a significant predictor of cardiovascular
complications over and beyond CBP [57]. The night/day
ratio had a predictive value independent of the 24 hour
ABP [57]. In this analysis diastolic ABP and CBP were
not related to outcome [57]. The previously reported
findings on surrogate (ECG voltages) and hard (out-
come) endpoints [43] allow us to conclude that sustained
344 Hypertension
ISH is more harmful than nonsustained ISH, in particu-
lar when the daytime systolic ABP ⱖ 160 mm Hg. More-
over, the benefit of antihypertensive therapy is also
mainly seen in the latter patients, and becomes less evi-
dent when daytime systolic ABP is less than 160 mmHg.
Pulse pressure as cardiovascular risk factor in
elderly hypertensives
For the moment, guidelines for the management of hy-
pertension are mainly based on the measurement of sys-
tolic and diastolic BP and these are usually considered as
being two isolated variables [44,45,58,59]. However, BP
is more correctly described as consisting of a pulsatile
(pulse pressure [PP]) and a steady (mean pressure [MP])
component [50]. Ventricular ejection, arterial stiffness
and the timing of the wave reflections are the major
determinants of PP, while MP mainly depends on pe-
ripheral vascular resistance and cardiac output. Several
observational studies suggested that in elderly people PP
might be a better predictor of cardiovascular risk than
MP [60–69]. Since PP widens with advancing age [2], the
results of outcome trials in older hypertensives might
provide an ideal database to test this hypothesis. There-
fore the data of three placebo-controlled trials in older
patients with hypertension (EWPHE [17], Syst-Eur [1],
Syst-China [23]) were pooled to assess the independent
roles of PP and MP as determinants of cardiovascular risk
with sufficient statistical power. The meta-analysis was
based on individual patient data. Relative hazard rates
associated with PP and MP were calculated using Cox
regression analysis and adjustments were made for sex,
age, previous cardiovascular complications, smoking, and
active treatment. Furthermore the relative hazard rates
for PP were also adjusted for MP and vice versa.
In the overall Cox regression analysis with stratification
for the three trials and with adjustment for the previously
mentioned covariates, a wider PP at baseline increased
the risk of major cardiovascular complications. The in-
crease in risk associated with a 10 mm Hg wider PP,
ranged from about 13% for all coronary end points
(P = 0.02) to nearly 20% for cardiovascular mortality
(P = 0.001). In a similar analysis, MP could only be iden-
tified as a significant predictor of risk, after removal of PP
as an explanatory variable from the model. The 2-year
probability of a major cardiovascular endpoint increased
with higher systolic BP (P < 0.001) (Fig. 3). Furthermore,
at any given level of systolic BP, the risk also increased
with lower diastolic BP (P = 0.001). This observation sug-
gests that the wider PP was driving the risk of major
complications.
The role of PP as a significant predictor of cardiovascular
risk had already been identified by several investigators
and in different groups of patients [60,61,63–67,69–71].
Moreover, a recent meta-analysis [45] including data of
more than 15 000 patients with ISH from 8 different
outcome trials also confirmed the role of PP as a risk
Figure 2. Risk according to hypertension status and treatment
group
Number of strokes and cardiovascular events per 1000-patient-years and
absolute number of events during follow-up in patients with Non-SH
(nonsustained hypertension), MiSH (mild sustained hypertension), and MoSH
(moderate sustained hypertension), divided according to treatment group (open
columns, placebo treatment; filled columns, active treatment). Results are from
intention-to-treat analysis. P value refers to comparison of rates between 2
treatment groups within each subgroup according to daytime systolic BP.
Published with permission [43].
Figure 3. Risk estimates for all cardiovascular endpoints
based on three trials
Risk associated with increasing systolic blood pressure at fixed levels of diastolic
blood pressure. The 2-year probability of a cardiovascular end point was
adjusted for active treatment, sex, age, previous cardiovascular complications,
and smoking by Cox multiple regression with stratification for trial (EWPHE [17],
Syst-Eur [1], Syst-China [23]). Published with permission [44].
ISH in the elderly Celis et al. 345
factor. Furthermore, the authors also reported that active
antihypertensive treatment was particularly beneficial in
men, in patients aged 70 years or more, and in those with
previous cardiovascular complications or wider PP at
baseline. If PP at baseline was 90 mm Hg or greater, 63
patients had to be treated to prevent one cardiovascular
death, whereas 119 patients had to be treated if pulse
pressure was less than 90 mm Hg.
From the present data it can be concluded that in older
hypertensives, PP, but not MP is the major determinant
of cardiovascular risk. These findings must however be
interpreted with caution. The positive and independent
association between PP and the incidence of cardiovas-
cular complications does not automatically imply a causal
or reversible relationship. Whether these data in the el-
derly can be extrapolated to younger or middle-aged
patients also remains to be proven. Furthermore, the
present findings may also have important clinical impli-
cations [72]. They suggest that the prediction of cardio-
vascular complications might be improved by accounting
for both the pulsatile (PP) and the steady (MP) compo-
nent of BP. This should however be further investigated
in randomized clinical trials in which the pulsatile com-
ponent of BP is differently affected by antihypertensive
drug treatment.
Conclusion
From the previous data some important conclusions
emerge. First, that stepwise antihypertensive treatment
starting with the CCB nitrendipine improves prognosis
in elderly patients with ISH and may particularly be in-
dicated in diabetic patients with ISH or in those at risk
of dementia. Second, that most of the benefit of
treatment is seen in patients with a daytime systolic
ABP ⱖ 160 mm Hg. And finally, that in elderly hyper-
tensive patients pulse pressure and not mean pressure is
the major determinant of cardiovascular risk.
Acknowledgment
The Syst-Eur trial, initiated by Antoon Amery, MD (died on November 2, 1994), was
a concerted action of the BIOMED Research Program sponsored by the European
Union. The trial was carried out in consultation with the World Health Organization,
the International Society of Hypertension, the European Society of Hypertension,
and the World Hypertension League. The trial was sponsored by Bayer AG (Wup-
pertal, Germany). The National Fund for Scientific Research (Brussels, Belgium)
provided additional support. Study medication was donated by Bayer AG and
Merck Sharp and Dohme Inc (West Point, PA, USA).
References and recommended reading
Papers of particular interest, published within the annual period of review,
have been highlighted as:
• Of special interest
•• Of outstanding interest
1 Staessen JA, Fagard R, Thijs L, et al., for the Systolic Hypertension in Europe
(Syst-Eur) Trial Investigators: Randomised double-blind comparison of pla-
cebo and active treatment for older patients with isolated systolic hyperten-
sion. Lancet 1997, 350:757–764.
2 Staessen JA, Amery A, Fagard R: Editorial review. Isolated systolic hyperten-
sion in the elderly. J Hypertens 1990, 8:393–405.
3 Amery A, Birkenhager W, Bulpitt CJ, et al.: Syst-Eur. A multicentre trial on the
treatment of isolated systolic hypertension in the elderly: objestives, protocol,
and organization. Aging 1991, 3:287–302.
4 SHEP Cooperative Research Group: Prevention of stroke by antihyperten-
sive drug treatment in older persons with isolated systolic hypertension. Final
results of the Systolic Hypertension in the Elderly Program (SHEP). JAMA
1991, 265:3255–3264.
5 Staessen JA, Fagard R, Amery A: Isolated systolic hypertension in the elderly:
implications of SHEP for clinical practice and for the ongoing trials. J Hum
Hypertens 1991, 5:469–474.
6 Pahor M, Guralnik JM, Salive ME, et al.: Do calcium channel blockers increase
the risk of cancer? Am J Hypertens 1996, 9:695–699.
7 Pahor M, Guralnik JM, Furberg CD, et al.: Risk of gastrointestinal haemor-
rhage with calcium antagonists in hypertensive persons over 67 years old.
Lancet 1996, 347:1061–1065.
8 Furberg CD, Psaty BM, Meyer JV: Nifedipine - dose-related increase in mor-
tality in patients with coronary heart disease. Circulation 1995, 92:1326–
1331.
9 Psaty BM, Heckbert SR, Koepsell TD, et al.: The risk of myocardial infarction
associated with antihypertensive drug therapies. JAMA 1995, 274 :620–
625.
•
10 Staessen JA, Thijs L, Birkenhager W, et al., on behalf of the Syst-Eur investi-
gators: Update on the Systolic Hypertension in Europe (Syst-Eur) Trial. Hy-
pertension 1999, 33:1476–1477.
This short paper provides updated outcome results based on the final Syst-Eur
database. Although a slightly greater number of endpoints was available, the con-
clusions of the initial Syst-Eur report were not affected.
11 Staessen JA, Fagard R, Thijs L, et al., for the Systolic Hypertension in Europe
Trial Investigators: Subgroup and per-protocol analysis of the randomized
European trial on isolated systolic hypertension in the elderly. Arch Intern Med
1998, 158:1681–1691.
12 Staessen JA, Thijs L, Fagard R, et al., for the Systolic Hypertension in Europe
(Syst-Eur) Trial Investigators: Calcium channel blockade and cardiovascular
prognosis in the European Trial on Isolated Systolic Hypertension. Hyperten-
sion 1998, 32:410–416.
••
13 Tuomilehto J, Rastenyte D, Birkenhager W, et al.: Effects of calcium-channel
blockade in older patients with diabetes and systolic hypertension. N Engl J
Med 1999, 340:677–684.
Ten percent of all 4695 Syst-Eur patients had diabetes mellitus. This post-hoc
analysis compared outcome in the diabetic and nondiabetic patients. It demon-
strated that nitrendipine-based antihypertensive treatment is particularly beneficial
in older diabetic patients with ISH and refuted the hypothesis that the use of CCB
might be harmful in diabetic patients.
14 Forette F, Amery A, Staessen JA, et al.: Is prevention of vascular dementia
possible? The Syst-Eur Vascular Dementia Project. Aging Clin Exp Res 1991,
3:373–382.
••
15 Forette F, Seux ML, Staessen JA, et al., on behalf of the Syst-Eur investiga-
tors: Prevention of dementia in randomised double-blind placebo- controlled
Systolic Hypertension in Europe (Syst-Eur) trial. Lancet 1998,352:1347–
1351.
Analysis of data from the Vascular Dementia Side Project of the Syst-Eur Trial
showed a reduction in the incidence of dementia of 50%. Treatment mainly pre-
vented Alzheimer’s dementia, a somewhat unexpected finding. The authors sug-
gest that specific neuroprotection by CCBs might partly explain this result.
•
16 Staessen JA, Wang JG, Thijs L, et al., for the Systolic Hypertension in Europe
Trial Investigators: Use of dihydropyridines for antihypertensive treatment in
older patients: evidence from the Systolic Hypertension in Europe Trial. J Clin
Basic Cardiol 2000, 3:15–21.
This paper reviews the main Syst-Eur results published before 2000. These results
show that stepwise antihypertensive drug treatment starting with the CCB nitren-
dipine, improves prognosis in older patients with ISH.
17 Amery A, Brixko P, Clement D, et al.: Mortality and morbidity results from the
European Working Party on High Blood Pressure in the Elderly Trial. Lancet
1985, 1349–1354.
•
18 Coope J, Warrender TS: Randomised trial of treatment of hypertension in
elderly patients in primary care. BMJ 1986, 293:1145–1151.
19 The Management Committee of the National Blood Pressure Study: Treat-
ment of mild hypertension in the elderly. A study initiated and administered by
the National Heart Foundation of Australia. Med J Aust 1981, 2:398–402.
20 Kuramoto K, Matsushita S, Kuwajima I, Murakam M: Prospective study on the
treatment of mild hypertension in the aged. Jpn Heart J 1981, 22:75–85.
21 Dahlof B, Lindholm LH, Hansson L, et al.: Morbidity and mortality in the Swed-
ish Trial in Old Patients with Hypertension (STOP-Hypertension). Lancet
1991, 338:1281–1285.
346 Hypertension
