Sie sind auf Seite 1von 21

Blood Pressure and Stroke: An Overview of Published Reviews

Carlene M.M. Lawes, Derrick A. Bennett, Valery L. Feigin and Anthony Rodgers
Stroke. 2004;35:776-785; originally published online February 19, 2004;
doi: 10.1161/01.STR.0000116869.64771.5A
Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright 2004 American Heart Association, Inc. All rights reserved.
Print ISSN: 0039-2499. Online ISSN: 1524-4628

The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://stroke.ahajournals.org/content/35/3/776

An erratum has been published regarding this article. Please see the attached page for:
http://stroke.ahajournals.org/content/35/4/1024.full.pdf

Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Stroke can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.
Once the online version of the published article for which permission is being requested is located, click
Request Permissions in the middle column of the Web page under Services. Further information about this
process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Stroke is online at:
http://stroke.ahajournals.org//subscriptions/

Downloaded from http://stroke.ahajournals.org/ by guest on November 3, 2014

Comments, Opinions, and Reviews


A CORRECTED VERSION OF THIS ARTICLE HAS BEEN PUBLISHED

Blood Pressure and Stroke


An Overview of Published Reviews
Carlene M.M. Lawes, MBChB, FAFPHM, MPH; Derrick A. Bennett, MSc, PhD, Cstat;
Valery L. Feigin, MD, MSc, PhD; Anthony Rodgers, MBChB, FAFPHM, PhD
BackgroundThe last few years have seen a considerable increase in the amount of information available concerning
blood pressure (BP) and stroke associations. This article provides an overview of published reviews of the effects on
stroke seen in trials of BP-lowering drugs and compares these with the results available from cohort studies.
Summary of ReviewWe present a review of major overviews of prospective cohort studies and an updated meta-analysis
of 40 randomized controlled trials of BP lowering, which included 188 000 participants and approximately 6800
stroke events. Cohort studies now indicate that in the Asia Pacific region as well as in North America and Western
Europe, each 10 mm Hg lower systolic BP is associated with a decrease in risk of approximately one third in subjects
aged 60 to 79 years. The association is continuous down to levels of at least 115/75 mm Hg and is consistent across
sexes, regions, and stroke subtypes and for fatal and nonfatal events. The proportional association is age dependent but
is still a strong and positive association in those aged 80 years. Data from randomized controlled trials, in which mean
age at event was approximately 70 years, indicate that a 10 mm Hg reduction in systolic BP is associated with a
reduction in risk of stroke of approximately one third. Per mm Hg systolic BP reduction, the benefits for stroke appear
similar between agents, by baseline BP levels, and whether or not individuals have a past history of cardiovascular
disease. There is, however, evidence of greater benefit with a larger BP reduction.
ConclusionsThe epidemiologically expected benefits of BP lowering for stroke risk reduction are broadly consistent
across a range of different population subgroups. There are greater benefits from larger BP reductions, and initiating and
maintaining BP reduction for stroke prevention is a more important issue than choice of initial agent. (Stroke. 2004;35:
776-785.)
Key Words: blood pressure cohort studies epidemiology meta-analysis randomized controlled trials stroke

he risk of stroke increases continuously above blood


pressure (BP) levels of approximately 115/75 mm Hg.
Since the association is steep, and BP levels are high in most
adult populations, almost two thirds of stroke burden globally
is attributable to nonoptimal BP (ie, 115/75 mm Hg).1
Approximately two thirds of this burden occurs in middleaged subjects (45 to 69 years), and approximately two thirds
occurs in developing regions.
In 1990 it was demonstrated that in North American and
Western European populations, a 5 mm Hg lower diastolic
blood pressure (DBP) was associated with a 30% to 40%
lower stroke risk,2 and this was reversed within a few years of
BP lowering.3 Since then, further evidence has become
available, especially in the elderly and in the Asia-Pacific
region and from many more trials comparing different agents,
more intensive and less intensive BP-lowering regimens, and
BP lowering in those without hypertension. This article
reviews published trial meta-analyses of the impact of pharmacological BP-lowering therapies on stroke across all population groups. The results are compared with overviews of

cohort study data that assessed the association between BP


and stroke in a range of populations. The main gaps in current
research and the clinical and public health implications of the
association of BP with stroke are also discussed.

Methods
Prospective Cohort Studies
Published overviews of cohort studies were identified by a search of
MEDLINE and EMBASE databases with key words overview,
meta-analysis, blood pressure, stroke or cerebrovascular or cardiovascular disease, cohort study or prospective study, and comparative
study. Overviews were included if they were published or data were
accessible by May 1, 2003, and if they provided data on inclusion
criteria, number and regions of the cohort studies included, number
of participants and demographics, duration of follow-up, disease end
points, and methods of analyses. Results were independently extracted and summarized by 2 reviewers, but no additional analyses
were conducted.

Randomized Controlled Trials


Meta-analyses of trials of BP-lowering agents were identified by a
MEDLINE and EMBASE search with search terms overview,

Received August 14, 2003; final revision received October 7, 2003; accepted November 19, 2003.
From the Clinical Trials Research Unit, Department of Medicine, University of Auckland, Auckland, New Zealand.
Correspondence to Carlene Lawes, Clinical Trials Research Unit, Department of Medicine, University of Auckland, Private Bag 92019, Auckland, New
Zealand. E-mail c.lawes@ctru.auckland.ac.nz
2004 American Heart Association, Inc.
Stroke is available at http://www.strokeaha.org

DOI: 10.1161/01.STR.0000116869.64771.5A

776
Downloaded from http://stroke.ahajournals.org/
by guest on November 3, 2014

Lawes et al
meta-analysis, blood pressure or hypertension, stroke or cerebrovascular or cardiovascular disease, and randomized controlled trial.
Data from individual trials included in these meta-analyses were
included if they were published before May 1, 2003, and if they
provided data on inclusion criteria, total number of participants and
demographics, randomization method, trial interventions, trial end
points, duration of follow-up, and net reduction in BP (ie, treatment
effect). Data from individual studies, from both the trial and
meta-analytic publications, were independently extracted by 2 reviewers. In some cases, existing meta-analyses of trials311 included
the most recent data on individual trials, and therefore these were
included in the final meta-analyses.
Core baseline and outcome data were extracted, and trials included
in the meta-analysis were stratified into the following 3 subgroups to
reduce clinical heterogeneity: (1) drug versus placebo or no treatment, (2) more intensive dose of drug versus less intensive dose of
drug, and (3) drug versus drug. The net difference in BP between, for
example, the treatment and placebo groups was calculated; this
reflects the true difference in BP achieved as the influence of
regression to the mean (which would be evident in both groups) is
removed. The updated meta-analysis was based on summary data
rather than individual participant data. The relative risk was computed as the measure of effect size for each trial and subgroup of

TABLE 1.

Blood Pressure and Stroke: An Overview

777

trials. Statistical heterogeneity was assessed with the use of the


standard Q statistic for heterogeneity, with a probability value of
0.1 used to indicate evidence of heterogeneity. A less stringent
probability value of 0.1 was used because it is well known that tests
of heterogeneity have low statistical power. If there was no evidence
of statistical heterogeneity when this criterion was used, then the
results were combined with the use of the standard inverse-variance
weighted fixed effect approach.12 If there was evidence of statistical
heterogeneity between the trials, attempts were made to explore
potential sources of heterogeneity. If this was not possible, the
random effects approach of DerSimonian and Laird13 was used as
part of a sensitivity analysis to assess the robustness of the conclusions. When there were sufficient studies, evidence of publication
bias was checked with the use of standard funnel plots.14
For each subgroup of trials, the weighted mean age, weighted
mean follow-up, and weighted mean BP reduction were estimated.
Trials were weighted by sample size and, in a sensitivity analysis, by
number of stroke events.

Results
Prospective Cohort Studies
Five overviews of cohort studies assessing the association
between BP and cardiovascular disease were identified. The

Major Overviews of Prospective Cohort Studies

No. of cohorts
Countries included
Inclusion criteria

Prospective Studies Collaboration (2002)

Asia Pacific Cohort Studies Collaboration (2003)

61

37

Europe, United States, Asia, Australia

Mainland China, Japan, Hong Kong, Taiwan, Singapore,


South Korea, New Zealand, Australia

At least 5000 person-years of follow-up


recorded or planned

Population from the Asia Pacific region

Date of birth or age, sex, blood pressure,


and cholesterol recorded at baseline

At least 5000 person-years of follow-up recorded


or planned

Date of death or age at death recorded


during follow-up

Date of birth or age, sex, and blood pressure


recorded
at baseline
Date of death or age at death recorded during
follow-up

958 074

425 251

% male

No. of participants

NR

57% (range 34%-100%)

Mean age (range) at baseline

NR

47 (20107) years

Range 417 years

Range 227 years

Follow-up duration

Stroke end points

Mean 12 years

Mean 7 years

12.7 million person years

3.0 million person years

11 960 strokes

5556 strokes

100% fatal

50% fatal

RR with a 10-mm Hg decrease in SBP

NR

0.63 (95% CI 0.610.64)

Fatal versus nonfatal

NR

(37% decrease in stroke risk)


No significant difference

Association by sex

No significant difference 60 years of age

No significant difference

Association by age

For a 10-mm Hg lower SBP

For a 10-mm Hg lower SBP

4049 years RR0.60 (0.570.63)

60 years RR0.46 (0.430.47)

5059 years RR0.62 (0.590.63)

6069 years RR0.64 (0.620.66)

6069 years RR0.66 (0.640.67)

70 years RR0.75 (0.720.77)

7079 years RR0.71 (0.690.72)


8089 years RR0.82 (0.790.84)
IHD indicates ischemic heart disease; CVD, cardiovascular disease; DBP, diastolic blood pressure; SBP, systolic blood pressure; NR, not reported; RR, relative risk;
CI, confidence interval.
Prospective Studies Collaboration (2002)16 includes MacMahon et al (1990)2 and Prospective Studies Collaboration (1995),15 and Asia Pacific Cohort Studies
Collaboration (2003)18 includes Eastern Collaborative Research Group (1998).17

Downloaded from http://stroke.ahajournals.org/ by guest on November 3, 2014

778

Stroke

March 2004

Figure 1. Usual SBP and risk of stroke by


age, with data from prospective cohort study
overviews. Data are shown from the
APCSC18 and PSC16 demonstrating the
associations between usual SBP and risk of
stroke plotted by age groups defined by age
at risk on a log scale. The x-axis coordinate
for each group is the mean follow-up SBP.
The y-axis coordinate is the hazard ratio
computed by the floating absolute risk technique,93 which enabled the comparison
between pairs of exposure groups rather
than comparisons between each exposure
group and a reference group. The solid
squares are larger where there are more
events because their size is proportional to
the inverse variance; vertical lines represent
95% CI. (Figures reprinted with permission
from Elsevier [Lancet] and Lippincott Williams & Wilkins [J Hypertens.])

MacMahon2 and Prospective Studies Collaboration (PSC)15,16


overviews included cohorts predominantly from the United
Kingdom, United States, and Europe. Cohorts in the MacMahon2 and first PSC15 review were subsequently included in
the later PSC review.16 The Eastern Stroke and Coronary
Heart Disease Collaborative Research17 overview is a subset
of the Asia Pacific Cohort Studies Collaboration (APCSC),18
which comprises cohorts from Mainland China, Hong Kong,
Taiwan, Japan, Singapore, South Korea, Australia, and New
Zealand. The PSC and APCSC overviews included almost 1
million and 420 000 individuals, respectively; both analyzed individual participant data and corrected for regression
dilution bias to reflect true or usual BP levels.16,18 In total,
7 cohorts were included in both the PSC and APCSC
overviews, representing 5% and 10% of participants,
respectively. The main characteristics and results of the
overviews are summarized in Table 1.
An important finding in all overviews was that the association
between BP and risk of stroke was continuous and log linear.
Therefore, a given absolute difference in usual BP was
associated with a similar relative risk reduction of stroke at all
levels of BP. There was no evidence of a threshold below
which levels of BP were no longer associated with lower
relative risk of stroke, down to approximately 115 mm Hg
systolic blood pressure (SBP) or 75 mm Hg DBP.16,18 All
overviews commented on the statistical heterogeneity between studies, and age appeared to be an important factor,16,18
explaining approximately 80% of the between-study heterogeneity in APCSC.18 The strength of the overall association
was not altered by restricting analyses to those with and
without a history of coronary heart disease at baseline or by
controlling for additional cardiovascular risk factors.16,18
The overall risk estimates provided by APCSC,18 PSC,16 and
previous meta-analyses2,15,17 (a 5 mm Hg lower DBP or
10 mm Hg lower SBP was associated with approximately a 30%
to 40% lower risk of stroke) were influenced by the age of the
cohorts, as the association of BP with stroke varied with age.
The proportional change in stroke risk per unit change in BP was
less extreme in old age than in middle age (Figure 1), but the
relationship remained strong and positive for all age groups.16,18
Age attenuation limits the ability to directly compare the

summary relative risk estimates across the overviews, but the 2


recent overviews that have published age-specific relative risks
had similar results (Table 1)16,18; a 10 mm Hg lower SBP was
associated with approximately a 40% to 50%, 30% to 40%, and
20% to 30% lower risk of stroke in those aged 60 years, 60 to
69 years, and 70 years, respectively. There has been no
evidence in any of the overviews that the strength of association
between BP and stroke varied by sex2,15,16,18 or for fatal and
nonfatal stroke events.2,18 Regional comparisons in APCSC and
PSC suggested that the association between BP and stroke in
Asia was similar or even slightly steeper than in Australia and
New Zealand,18 Europe,16 and the Unites States16 after standardization for age. In these 2 overviews the associations of BP with
stroke subtypes were broadly similar,16,18 although low rates of
CT/MRI confirmation would tend to obscure any differences.

Randomized Controlled Trials


Randomized controlled trials provide data on the impact of
BP lowering on short-term stroke outcomes, which is relevant
in assessing the extent to which the epidemiologically expected reductions in stroke are produced.2,3
Randomized Controlled Trials Comparing Antihypertensive
Drugs With a Placebo (or No Treatment)
Seventeen trials compared the effects of -blocker and/or
diuretic with a placebo or no treatment; in 6 trials the comparison
drug was angiotensin-converting enzyme (ACE) inhibitors, and
in 2 it was calcium antagonists (Table 2).50 79 In total, there were
73 500 participants and almost 2900 stroke events recorded in
these trials. Weighted mean follow-up time for the 3 groups of
trials ranged from 2 to 4.6 years, and the weighted mean age of
participants ranged from 57 to 68 years. There was clear
evidence of a reduction in stroke risk with BP lowering with
each drug versus placebo comparison (P0.005), with pooled
relative risk reductions for -blocker and/or diuretic, ACE
inhibitors, and calcium antagonists of 35%, 28%, and 39%,
respectively. These broadly similar relative risk reductions
were achieved with weighted mean reductions in BP
(SBP/DBP) of 13/6, 5/2, and 10/5 mm Hg, respectively.
While the BP reductions appeared less for the ACE
inhibitor versus placebo trials, there is also greater uncer-

Downloaded from http://stroke.ahajournals.org/ by guest on November 3, 2014

Lawes et al
TABLE 2.

Blood Pressure and Stroke: An Overview

779

Randomized Controlled Trials Comparing Antihypertensive Drugs to a Placebo (or No Treatment)


Treatment

Trial

Placebo

Net Difference*
Follow-Up,
y

Mean Age,
y

%
Male

SBP

DBP

RR and 95% CI

Beta-blocker and/or diuretic vs placebo or no treatment


ANBPS50

13

1721

22

1706

50

63

NR

0.59 (0.301.16)

58

58

56

43

15

10

0.99 (0.049.56)

Carter et al52

10

49

21

48

NA

57

NR

NR

0.47 (0.250.88)

53

Barraclough51
Coope

20

419

39

465

4.4

69

31

18

11

0.57 (0.340.96)

Dutch TIA54

52

732

62

741

2.6

65

64

0.85 (0.601.21)

EWPHE55

32

416

48

424

4.7

72

30

21

0.68 (0.441.04)

HDFP5660

102

5485

158

5455

51

55

NR

0.64 (0.500.82)
0.78 (0.541.11)

HSCSG I II III61

43

233

52

219

2.3

59

41

25

12

MRC older62

101

2183

134

2213

5.8

70

42

14

0.76 (0.590.98)

MRC young63

60

8700

109

8654

52

52

11

0.55 (0.400.75)

406

379

5.5

45

100

17

10

0.08 (0.001.53)

Oslo64,65
66

SHEP

105

2365

162

2371

4.5

72

43

12

0.65 (0.510.83)

STOP H67

30

812

55

815

2.1

76

37

20

0.55 (0.350.85)

USPHS68

193

196

44

80

15

10

0.17(0.021.39)

VA6970

254

23

257

3.8

51

100

17

10

0.26 (0.110.64)

VA-NHLBI71,72

508

504

1.5

38

81

NR

0.99 (0.0249.9)

Wolff et al73

45

42

1.4

49

32

NR

20

1.87 (0.1819.8)

577

24 579

897

24 547

4.6

57

53

13

0.65 (0.590.73)

156

4654

226

4652

66

73

0.69 (0.570.84)

308

309

61

82

1.76 (0.525.94)

307

3051

420

3054

64

70

0.73 (0.640.84)

QUIET

878

872

58

82

NR

NR

0.99 (0.0615.8)

SCAT77

229

231

61

89

0.22 (0.051.03)

Summary
ACE inhibitors vs placebo
HOPE74
PART 275
PROGRESS27
76

TEST78
Summary

372

69

348

2.6

70

60

1.00 (0.751.35)

547

9492

729

9466

4.3

64

73

0.72 (0.680.83)

Calcium antagonists vs placebo


PREVENT79

417

408

57

80

0.98 (0.293.35)

Syst-Eur47

47

2398

77

2297

70

33

11

0.58 (0.410.84)

Summary

52

2815

82

2705

68

40

10

0.61 (0.440.85)

nnumber of events; Nnumber of participants. SBP indicates systolic blood pressure; DBP, diastolic blood pressure; RR, relative risk; 95% CI, 95% confidence
interval; NR, not recorded.
*Net BP differenceblood pressure difference between the randomized groups in mm Hg (ie, BP change in treatment groupBP change in placebo group).
Summary follow-up, mean age, % male, and net SBP and DBP change are weighted by study size for those studies that provided data on net changes in both
SBP and DBP. Weighting by number of events made negligible difference. The pooled RR and 95% CI (using the fixed effects approach) for each of the 3 trial subgroups
are presented, and the test for heterogeneity (Q statistic and P value) are Q14.23, P0.36; Q8.90, P0.11; and Q0.61, P0.43, respectively.
These trials also included older classes of drugs such as methyldopa and alkaloids.
There were no stroke events in the Barraclough or VA-NHLBI trial; however, for the purposes of meta-analyses, 0.5 was added to all cells in the 22 table.

tainty over the BP reductions in this subgroup.19 There was


no evidence of small study bias20 as assessed by a funnel
plot14 (not shown) and no significant statistical heterogeneity in any of the 3 subgroup comparisons.
The trials comparing antihypertensive drugs with a placebo
(or no treatment) were stratified by several trial characteristics such as mean age at entry, baseline BP, and history of
cardiovascular disease (Figure 2). Overall, there was a reduction in risk of stroke of 30% and a risk reduction of

approximately 30% to 40% in most subgroups. However, in


several groups the risk reduction appeared to be greater with
larger net reduction in BP, for example, in those with no past
history of stroke/transient ischemic attack or vascular disease.
Randomized Controlled Trials Comparing More Intensive
Versus Less Intensive Antihypertensive Drug Regimens
Three individually inconclusive clinical trials including a
total of 20 408 participants and 384 stroke events compared
treatment regimens of different intensities2124 (Table 3). The

Downloaded from http://stroke.ahajournals.org/ by guest on November 3, 2014

780

Stroke

March 2004
results were of borderline statistical significance. The latter
subgroup was the only one in which there was evidence of
statistical heterogeneity (P0.09), and a random effects
model gave a nonsignificant relative risk reduction of 5%.

Figure 2. Randomized controlled trials comparing antihypertensive drugs with a placebo (or no treatment) by subgroup. The
meta-analyses of BP-lowering trials are presented stratified into
subgroups on the basis of mean age of trial participants at
entry, baseline SBP level, and whether trial participants predominantly had a history of stroke/transient ischemic attack (TIA) or
vascular disease. The diamonds are centered on the pooled
estimate of effect and represent 95% CI. The solid diamond
represents the pooled relative risk and 95% CI for all contributing trials.

meta-analysis suggested that additional benefit may be gained


from a more intensive treatment regimen (relative risk reduction of 20%; P0.04). There was no significant statistical
heterogeneity.
Randomized Controlled Trials Comparing Different
Antihypertensive Drugs
Fifteen trials compared the effects of different types of
antihypertensive drugs, with 2 trials including several drug
versus drug comparisons (Table 4).80 92 In total, there were
96 000 participants and almost 3600 stroke events recorded
over a mean follow-up time of 4 to 5 years. The weighted
mean reduction in BP in many of the drug versus drug trials
was small, often 1 mm Hg SBP and DBP (Table 4).
Overall, these trials indicated that there was little difference
between the drug classes, with relative risk reductions of 9%
(-blockers and/or diuretics versus ACE inhibitors), 8%
(-blockers and/or diuretics versus calcium antagonists), and
11% (calcium antagonists versus ACE inhibitors). These
TABLE 3.

More

ABCD22
21

UKPDS-HDS
Summary

Discussion
Cohort studies have demonstrated that the strong association
between BP and stroke appears to be continuous down to
levels of at least 115/75 mm Hg. The strength of the associations was similar for men and women and for fatal and
nonfatal events but attenuated with age.2,1518 Despite a
potentially weaker relative association for older age groups,
the higher overall stroke rate in this group means that in
absolute terms, the benefits of a given reduction in BP are
likely to be greater. The analyses have not consistently
demonstrated a difference in the association between intracerebral hemorrhage and ischemic stroke; however, only approximately one half of stroke subtypes were confirmed (CT,
MRI, or autopsy). Regional comparisons are very limited

Randomized Controlled Trials Comparing More Intensive vs Less Intensive Antihypertensive Drug Regimens

More Intensive vs
Less Intensive Trials
HOT

Net Reduction in BP and Relative Risk Reduction in Stroke


A meta-regression analysis was used to assess the direct
relationship between the net reduction in SBP in the 7 trial
meta-analyses. The net reduction in SBP was plotted against
the relative risk reduction in stroke for each meta-analysis.
This plot suggested that a dose-response relationship may
exist (Figure 3). A weighted linear regression line was
superimposed in which weights are equal to the inverse of the
variance for each trial meta-analysis. The slope of this
weighted regression line indicated that a 10 mm Hg greater
reduction in SBP would be associated with a reduction in risk
of stroke of 31% (R0.71). The mean weighted follow-up
duration for all these trials was 4.5 years, and the mean
weighted age of participants at entry into trials was 63 years
(the mean age at event is likely to be approximately a decade
later). Age-specific analyses from the 2 cohort study overviews16,18 estimated that a 10 mm Hg lower SBP in participants aged 60 to 69 years at event was associated with a risk
reduction in stroke of approximately 35%; for those aged
70 years the risk reduction was 25% to 29% (Table 5). The
results from cohort studies and randomized controlled trials
are therefore broadly consistent, suggesting that most of the
epidemiologically expected benefit of BP lowering is
achieved within a few years for stroke.

23,24

Less
N

Net Difference*
N

Follow-Up,
y

Mean Age,
y

%
Male

SBP

DBP

RR and 95% CI

58

67

0.98 (0.402.43)

237

233

89

6262

205

12528

62

53

0.87 (0.681.11)

38

758

34

390

56

55

10

0.58 (0.370.90)

136

7257

248

13 151

4.2

62

53

0.80 (0.650.99)

nno. of events, Nno. of participants. SBP indicates systolic blood pressure; DBP, diastolic blood pressure; RR, relative risk; 95% CI, 95% confidence interval.
* Net BP differenceblood pressure difference between the randomized groups (mm Hg) (ie, BP change in more intensive drug treatment groupBP change in
less intensive drug treatment group).
Summary follow-up, mean age, % male, and net SBP and DBP change are weighted by study size for those studies that provided data on net changes in both
SBP and DBP. Weighting by number of events made negligible difference. The pooled RR and 95% confidence interval (using the fixed effects approach) are presented,
and the test for heterogeneity (Q statistic and P value) are Q2.70, P0.26.

Downloaded from http://stroke.ahajournals.org/ by guest on November 3, 2014

Lawes et al
TABLE 4.

Blood Pressure and Stroke: An Overview

781

Randomized Controlled Trials Comparing Different Blood PressureLowering Drugs


First Drug

Drug vs Drug
Trials

Second Drug
N

Net Difference*

Follow-Up,
y

Mean Age,
y

%
Male

SBP

DBP

RR and 95% CI

Beta-blockers and/or diuretics vs ACE inhibitors


ALLHAT80

675

15255

457

9054

4.9

67

53

0.98(0.780.98)

ANBPS81

107

3039

112

3044

4.1

72

49

0.95(0.741.23)

CAPPP82

148

5493

189

5492

53

53

0.78(0.630.97)

237

2213

215

2205

76

33

1.10(0.931.32)

17

358

21

400

56

54

0.90(0.491.69)

1184

26 358

994

20 195

5.1

65

51

0.91(0.830.99)

1.06(0.931.20)

46

STOP-2

UKPDS-HDS
Summary

23,24

Beta-blockers and/or diuretics vs calcium antagonists


ALLHAT80

675

15255

377

9048

67

53

CASTEL83

205

146

73

51

0.71(2.440.21)

118

8361

133

8241

66

44

0.88(0.681.12)

ELSA85

14

1157

1177

56

55

1.59(0.693.70)

INSIGHT86

84

3164

79

3157

65

46

1.06(0.781.43)

MIDAS87

441

442

59

78

0.50(0.132.00)

NICS-EH88

214

215

70

33

1.33(0.473.85)

89

NORDIL

196

5471

159

5410

60

49

1.22(0.991.49)

STOP-246

237

2213

207

2196

76

33

1.14(0.951.35)

VHAS90,91

707

707

54

49

0.80(0.222.94)

1344

37 188

986

30 739

4.2

65

48

1.08(0.991.16)

CONVINCE84

Summary

Calcium antagonists vs ACE inhibitors


ABCD22
ALLHAT80
92

FACET

STOP-246
Summary

11

235

235

58

67

1.57(0.623.98)

377

9048

457

9054

67

53

0.83(0.720.94)

10

191

189

207

2196

215

2205

11 683

11 683

11 683

11 683

3.5

63

41

2.47(0.797.75)

76

33

0.97(0.811.16)

4.9

68

49

0.89(0.800.99)

nno. of events, Nno. of participants; SBP indicates systolic blood pressure; DBP, diastolic blood pressure; RR, relative risk; 95% CI, 95% confidence interval.
*Net BP differenceblood pressure difference between the randomized groups (mm Hg). Presented as the BP change in (first listed drug)(second listed drug),
ie, a negative value means that the second listed drug achieved a greater blood pressure reduction than the first listed drug.
Summary follow-up, mean age, % male, and net SBP and DBP change are weighted by study size for those studies that provided data on net changes in both
SBP and DBP. Weighting by number of events made negligible difference. The pooled RR and 95% confidence interval (using the fixed effects approach) for each
of the 3 trial subgroups are presented, and the test for heterogeneity (Q statistic and P value) are Q6.91, P0.14; Q7.36, P0.59; and Q6.53, P0.09,
respectively. The pooled RR and 95% CI using a random effects approach for the last trial subgroup was 0.95 (0.76 1.19).

because most of the cohort studies have been conducted


primarily in the developed parts of the world; however, the
broad consistency of the results across overviews was more
evident than any differences. This suggests that the proportional associations are likely to be reasonably generalizable to
the many populations in the world (eg, Africa, South America) for which there are limited data on stroke
epidemiology.25
Data from trials provide more reliable data on the likely
short-term benefits of BP lowering on stroke. As with the cohort
data, randomized controlled trials have been conducted predominantly in populations of Western countries, and data for developing countries (where approximately three quarters of stroke
deaths occur worldwide26) are very scarce. The results of the
updated meta-analysis of trials are consistent with those of
several previous meta-analyses and confirmed the risk reduction

of stroke with BP lowering compared with placebo (regardless


of the agent used) as being between 30% and 40%. Previous
analyses have also found no evidence of a difference in effect
size for men and women or for fatal and nonfatal events.3 8
Limited data were available on treatment effects by age, and the
treatment effects by age within individual trials27 and between
trials4,5,9,10,28 have not been consistent. However, the trials were
often not powered for age subgroup analyses and often included
only a narrow age band. Few data are available on stroke
subtypes, limiting the scope for meta-analyses. The limited data
from the Perindopril Protection Against Recurrent Stroke Study
(PROGRESS)27 and the Heart Outcome Prevention Evaluation
(HOPE) study29 are insufficient to draw conclusions regarding
the impact of BP lowering on stroke subtypes. Analyses have
indicated, however, that a more intensive BP-lowering regimen
with a given agent may produce a greater risk reduction than a

Downloaded from http://stroke.ahajournals.org/ by guest on November 3, 2014

782

Stroke

March 2004
Heart Attack Trial (ALLHAT)30 and the Losartan Intervention
For Endpoint reduction in hypertension study (LIFE)31,32 were
not included in the meta-analysis because no other trials included the same classes of drugs. They indicated that there was
a higher risk of stroke in the -adrenergic blocker group
compared with diuretics30 and that angiotensin II blockers may
have greater impact on stroke than -blockers.31,32

Shape and Size of Association Between BP


and Stroke

Figure 3. Net reduction in SBP and relative risk reduction in


stroke in randomized controlled trials of BP lowering. In a metaregression, the direct relationship between the net reduction in
SBP was plotted against the relative risk reduction in stroke for
each of the 7 trial meta-analyses. The slope of the weighted linear regression line (where weights are equal to the inverse of
the variance for each trial meta-analysis) indicated that a
10 mm Hg greater reduction in SBP would be associated with a
reduction in risk of stroke of 31% (R0.71). From left to right,
the diamonds represent the meta-analyses for -blockers
and/or diuretics vs calcium antagonists, calcium antagonists vs
ACE inhibitors, and -blockers and/or diuretics vs ACE inhibitors; the circle represents the more intensive vs less intensive
antihypertensive drug regimens; and the squares are ACE inhibitors vs placebo, calcium antagonists vs placebo, and
-blockers and/or diuretics vs placebo or no treatment. The
sizes of the diamonds, circle, and squares are larger where
there are more events because their size is proportional to the
inverse variance of that comparison; vertical lines represent
95% CI.

less intensive regimen and that overall the trials demonstrate a


dose-response relationship between magnitude of BP reduction
and reduction in risk of stroke.
When different classes of BP-lowering drugs were directly
compared, there was very little difference in either the magnitude of the BP reduction achieved or the impact of different
agents on stroke. The similarities between the effects of different
drug classes are more striking than the differences, but there is
uncertainty over the extent to which any differences are explained by differences in BP reduction and chance. One arm of
the Antihypertensive and Lipid-Lowering Treatment to Prevent

There has been considerable debate over the shape of the


association between BP and stroke. The data presented here
from cohort studies suggested a continuous log-linear association. However, some other cohort data appear to have
demonstrated a J-shaped curve,3337 and it has been suggested
that this relationship indicates that BP has been lowered too
far and that cerebral blood flow is compromised, leading to
ischemia. It is more likely to reflect deteriorating health,
which accounts for both falling pressure and a greater
likelihood of a cardiovascular disease event, but the 2 factors
may not be causally related.38 43 Several trials have found no
evidence of a J-curve association despite including subgroups
of patients susceptible to reduced perfusion.44 47 The large
PROGRESS48 and HOPE29 trials found no evidence of a
J-curve association across a wide range of BP levels. Overall,
these results support a log-linear association and indicate that
a potential J-curve relationship should not detract attention
from the major benefits of BP lowering.
A key finding from this article is that the 7 meta-analyses
of randomized controlled trials of BP lowering and stroke
suggest a dose-response relationship. At mean age at baseline
of approximately 63 years, a 10 mm Hg reduction in SBP was
associated with a risk reduction in stroke of 31%. Given that
most of these individuals would not have experienced their
stroke event until approximately a decade later, the results are
consistent with the predicted reduction in stroke risk from the
cohort study data. This indicates that most of the epidemiologically expected benefit of BP lowering may be achieved
within a few years for stroke.

Future Research and Implications for Prevention


Research and analyses from cohort study collaborations such
as APCSC and meta-analysis of individual participant trial

TABLE 5. Relative Risk Reduction in Stroke With a Reduction in Systolic Blood Pressure Predicted From
Cohort Studies and Observed From Randomized Controlled Trials
Observed Effects on Stroke With a Reduction
in SBP of 10 mm Hg

Predicted Effects on Stroke of Lowering SBP 10 mm Hg*


Mean Age at Event

Prospective Studies Collaboration (2002)16


18

Asia Pacific Cohort Studies Collaboration (2003)

Estimated Mean Age at Event

6069 Years

7079 Years

Approximately 73 Years

34%

29%

31%

36%

25%

SBP indicates systolic blood pressure.


*Relative risk reduction in stroke from cohort studies with a 10-mm Hg lower SBP, by age at event.
Relative risk reduction in stroke from randomized controlled trials with a 10-mm Hg reduction in SBP based on the slope of the
regression line in Figure 3. Mean age at entry into trials was 63 years, and mean age at event is likely to approximately a decade
later, ie, 73 years.

Downloaded from http://stroke.ahajournals.org/ by guest on November 3, 2014

Lawes et al
data improve the understanding of the associations between
BP and stroke and the potential benefits of BP lowering in
different population subgroups. Additional analyses are able
to examine the combined impact of multiple risk factors on
stroke risk. BP intervention strategies would be greatly
enhanced if these data were also used in the improvement of
risk prediction frameworks to better identify those at higher
absolute risk of cardiovascular disease. Development of
simple versions of these risk prediction tools would be of
particular value in developing regions where much of the
future cardiovascular disease burden is likely to occur.1,49
Future trial analyses should focus on the presence or
absence of age attenuation in the effect of BP lowering. This
would require comparison of age-specific, follow-up time
specific, and end pointspecific analyses of individual participant data from both observational studies (such as APCSC)
and clinical trials (such as the Blood Pressure Lowering
Treatment Trialists Collaboration11). Given that most individuals require combination therapy, it is also important to
shift the focus of trials from comparisons of the short-term
benefits of individual drugs. Instead, future trials should aim
to determine the optimal combination therapy and the potential long-term benefits of different antihypertensive regimens
in different population subgroups. While there may be some
differences between BP-lowering agents, the relative differences between these agents are minor compared with the
relative benefits of any BP-lowering agent versus no
treatment.
Finally, it is important to acknowledge that treatment of
high-risk groups will only partially address the growing
cardiovascular disease epidemic. Data presented in this article
indicated that the relative benefits of BP lowering are not
only limited to those classified as hypertensive but would also
be evident in those with normal or below normal BP. These
relative benefits appear broadly consistent across many population subgroups. Other important areas of research therefore include exploration of cost-effective strategies to bring
about population changes in cardiovascular risk factors such
as BP. Such initiatives would complement targeted treatment
approaches and would not only benefit the developed world
but could potentially arrest the projected increase in cardiovascular disease in developing countries.

Acknowledgments
This study was supported by a Health Research Council (New
Zealand) fellowship to Dr Lawes. Dr Rodgers is a National Health
Foundation (New Zealand) Senior Research Fellow. We thank
Varsha Parag and Ruey-Bin Lin for statistical assistance and Clarissa
Could-Thorpe for secretarial support.

Blood Pressure and Stroke: An Overview

4.

5.

6.

7.

8.
9.

10.

11.

12.
13.
14.
15.

16.

17.

18.

19.

20.

21.

22.

References
1. World Health Organisation. The World Health Report 2002: Reducing
Risks, Promoting Healthy Life. Geneva, Switzerland: World Health
Organisation; 2002.
2. MacMahon S, Peto R, Cutler J, Collins R, Sorlie P, Neaton J, Abbott R,
Godwin J, Dyer A, Stamler J, et al. Blood pressure, stroke, and coronary
heart disease, part I: prolonged differences in blood pressure: prospective
observational studies corrected for the regression dilution bias. Lancet.
1990;335:765774.
3. Collins R, Peto R, MacMahon S, Hebert P, Fiebach NH, Eberlein KA,
Godwin J, Qizilbash N, Taylor JO, Hennekens CH, et al. Blood pressure,
stroke, and coronary heart disease, part 2: short-term reductions in blood

23.

24.

25.

26.

783

pressure: overview of randomised drug trials in their epidemiological


context. Lancet. 1990;335:827 838.
MacMahon S, Rodgers A. The effects of anti-hypertensive treatment on
vascular disease: reappraisal of the evidence in 1994. J Vasc Med Biol.
1993;4:265271.
MacMahon S, Rodgers A. The effects of blood pressure reduction in older
patients: an overview of five randomized controlled trials in elderly
hypertensives. Clin Exp Hypertens. 1993;15:967978.
Insua JT, Sacks HS, Lau TS, Lau J, Reitman D, Pagano D, Chalmers TC.
Drug treatment of hypertension in the elderly: a meta-analysis. Ann Intern
Med. 1994;121:355362.
Gueyffier F, Boutitie F, Boissel JP, Pocock S, Coope J, Cutler J, Ekbom
T, Fagard R, Friedman L, Perry M, et al, for the INDANA Investigators.
Effect of antihypertensive drug treatment on cardiovascular outcomes in
women and men: a meta-analysis of individual patient data from randomized, controlled trials. Ann Intern Med. 1997;126:761767.
He J, Whelton PK. Elevated systolic blood pressure as a risk factor for
cardiovascular and renal disease. J Hypertens Suppl. 1999;17:S7S13.
Gueyffier F, Boutitie F, Boissel JP, Coope J, Cutler J, Ekbom T, Fagard
R, Friedman L, Perry HM, Pocock S. INDANA: a meta-analysis on
individual patient data in hypertension: protocol and preliminary results.
Therapie. 1995;50:353362.
Gueyffier F, Froment A, Gouton M. New meta-analysis of treatment trials
of hypertension: improving the estimate of therapeutic benefit. J Hum
Hypertens. 1996;10:1 8.
Neal B, MacMahon S, Chapman N, for the Blood Pressure Lowering
Treatment Trialists Collaboration. Effects of ACE inhibitors, calcium
antagonists, and other blood-pressure-lowering drugs: results of prospectively designed overviews of randomised trials. Lancet. 2000;356:
19551964.
Hedges LV, Olkin I. Statistical Methods for Meta-analysis. London, UK:
Academic Press; 1985.
DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin
Trials. 1986;7:177188.
Light RJ, Pillemer DB. Summing Up: The Science of Reviewing Research.
Cambridge, Mass: Harvard University Press; 1984.
Prospective Studies Collaboration. Cholesterol, diastolic blood pressure,
and stroke: 13,000 strokes in 45,000 people in 45 prospective cohorts.
Lancet. 1995;346:16471653.
Prospective Studies Collaboration. Age-specific relevance of usual blood
pressure to vascular mortality: a meta-analysis of individual data for one
million adults in 61 prospective studies. Lancet. 2002;360:19031913.
Eastern Stroke and Coronary Heart Disease Collaborative Research
Group. Blood pressure, cholesterol, and stroke in eastern Asia. Lancet.
1998;352:18011807.
Asia Pacific Cohort Studies Collaboration. Blood pressure and cardiovascular disease in the Asia Pacific region. J Hypertens. 2003;21:
707716.
Svensson P, de Faire U, Sleight P, Yusuf S, Ostergren J. Comparative
effects of ramipril on ambulatory and office blood pressures: a HOPE
substudy. Hypertension. 2001;38:E28 E32.
Sterne JA, Egger M, Smith GD. Systematic reviews in health care:
investigating and dealing with publication and other biases in meta-analysis. BMJ. 2001;323:101105.
Hansson L, Zanchetti A, Carruthers SG, Dahlof B, Elmfeldt D, Julius S,
Menard J, Rahn KH, Wedel H, Westerling S, for the HOT Study Group.
Effects of intensive blood-pressure lowering and low-dose aspirin in
patients with hypertension: principal results of the Hypertension Optimal
Treatment (HOT) randomised trial. Lancet. 1998;351:17551762.
Estacio RO, Jeffers BW, Hiatt WR, Biggerstaff SL, Gifford N, Schrier
RW. The effect of nisoldipine as compared with enalapril on cardiovascular outcomes in patients with non-insulin-dependent diabetes and
hypertension. N Engl J Med. 1998;338:645 652.
UK Prospective Diabetes Study Group. Tight blood pressure control and
risk of macrovascular and microvascular complications in type 2 diabetes:
UKPDS 38. BMJ. 1998;317:703713.
UK Prospective Diabetes Study Group. Efficacy of atenolol and captopril
in reducing risk of macrovascular and microvascular complications in
type 2 diabetes: UKPDS 39. BMJ. 1998;317:713720.
Feigin VL, Lawes CMM, Bennett DA, Anderson CS. Stroke epidemiology: a review of population-based studies of incidence, prevalence, and
case-fatality in the late 20th century. Lancet Neurol. 2003;2:4353.
Truelsen T, Bonita R, Jamrozik K. Surveillance of stroke: a global
perspective. Int J Epidemiol. 2001;30(suppl 1):S11S16.

Downloaded from http://stroke.ahajournals.org/ by guest on November 3, 2014

784

Stroke

March 2004

27. PROGRESS Collaborative Group. Randomised trial of a


perindopril-based blood-pressure-lowering regimen among 6,105 individuals with previous stroke or transient ischaemic attack. Lancet. 2001;
358:10331041.
28. Gueyffier F, Bulpitt C, Boissel J-P, Schron E, Ekbom T, Fagard R,
Casiglia E, Kerlikowske K, Coope J, for the INDANA Group. Antihypertensive drugs in very old people: a subgroup meta-analysis of randomised controlled trials. Lancet. 1999;353:793796.
29. Bosch J, Yusuf S, Pogue J, Sleight P, Lonn E, Rangoonwala B, Davies R,
Ostergren J. Use of ramipril in preventing stroke: double blind randomised trial. BMJ. 2002;324:699 702.
30. ALLHAT Collaborative Research Group. Major cardiovascular events in
hypertensive patients randomized to doxazosin vs chlorthalidone: the
Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack
Trial (ALLHAT). JAMA. 2000;283:19671975.
31. Dahlof B, Devereux RB, Kjeldsen SE, Julius S, Beevers G, Faire U,
Fyhrquist F, Ibsen H, Kristiansson K, Lederballe-Pedersen O, et al.
Cardiovascular morbidity and mortality in the Losartan Intervention For
Endpoint reduction in hypertension study (LIFE): a randomised trial
against atenolol. Lancet. 2002;359:9951003.
32. Lindholm LH, Ibsen H, Dahlof B, Devereux RB, Beevers G, de Faire U,
Fyhrquist F, Julius S, Kjeldsen SE, Kristiansson K, et al. Cardiovascular
morbidity and mortality in patients with diabetes in the Losartan Intervention For Endpoint reduction in hypertension study (LIFE): a randomised trial against atenolol. Lancet. 2002;359:1004 1010.
33. Stewart IM. Relation of reduction in pressure to first myocardial
infarction in patients receiving treatment for severe hypertension. Lancet.
1979;1:861 865.
34. Kannel WB, DAgostino RB, Silbershatz H. Blood pressure and cardiovascular morbidity and mortality rates in the elderly. Am Heart J. 1997;
134:758 763.
35. Cruickshank JM. J curve in antihypertensive therapy: does it exist? A
personal point of view. Cardiovasc Drugs Ther. 1994;8:757760.
36. DAgostino RB, Belanger AJ, Kannel WB, Cruickshank JM. Relation of
low diastolic blood pressure to coronary heart disease death in presence
of myocardial infarction: the Framingham Study. BMJ. 1991;303:
385389.
37. Farnett L, Mulrow CD, Linn WD, Lucey CR, Tuley MR. The J-curve
phenomenon and the treatment of hypertension: is there a point beyond
which pressure reduction is dangerous? JAMA. 1991;265:489 495.
38. Alderman MH. Blood pressure J-curve: is it cause or effect? Curr Opin
Nephrol Hypertens. 1996;5:209 213.
39. MacMahon S, Rodgers A, Neal B, Chalmers J. Blood pressure lowering
for the secondary prevention of myocardial infarction and stroke. Hypertension. 1997;29:537538.
40. Sleight P. Lowering of blood pressure and artery stiffness. Lancet. 1997;
349:362.
41. Sleight P. Lowering of blood pressure and artery stiffness. Lancet. 1997;
349:955956.
42. Neaton JD, Wentworth D, for the Multiple Risk Factor Intervention Trial
Research Group. Serum cholesterol, blood pressure, cigarette smoking,
and death from coronary heart disease: overall findings and differences by
age for 316,099 white men. Arch Intern Med. 1992;152:56 64.
43. Flack JM, Neaton J, Grimm R Jr, Shih J, Cutler J, Ensrud K, MacMahon
S, for the Multiple Risk Factor Intervention Trial Research Group. Blood
pressure and mortality among men with prior myocardial infarction.
Circulation. 1995;92:24372445.
44. McMurray J, McInnes GT. The J-curve hypothesis. Lancet. 1992;339:
561562.
45. Pfeffer MA. Angiotensin-converting enzyme inhibition in congestive
heart failure: benefit and perspective. Am Heart J. 1993;126(pt
2):789 793.
46. Hansson L, Lindholm LH, Ekbom T, Dahlof B, Lanke J, Schersten B,
Wester PO, Hedner T, de Faire U. Randomised trial of old and new
antihypertensive drugs in elderly patients: cardiovascular mortality and
morbidity: the Swedish Trial in Old Patients with Hypertension-2 study.
Lancet. 1999;354:17511756.
47. Staessen J, Fagard R, Thijs L, Celis H, Arabidze G, Birkenhager W,
Bulpitt C, Leeuw PD, Dollery C, Fletcher AE, et al. Randomised
double-blind comparison of placebo and active treatment for older
patients with isolated systolic hypertension. Lancet. 1997;350:757764.
48. PROGRESS Collaborative Group. The lowering of blood-pressure after
stroke. Lancet. 2001;358:1994 1995.
49. Murray CJL, Lopez AD. The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability From Diseases, Injuries,

50.
51.
52.
53.
54.

55.

56.

57.

58.

59.

60.

61.
62.
63.
64.
65.
66.

67.

68.
69.

70.

71.
72.

73.
74.

75.

and Risk Factors in 1990 and Projected to 2020. Cambridge, Mass:


Harvard University Press; 1996.
Report by the Management Committee. The Australian therapeutic trial in
mild hypertension. Lancet. 1980;1:12611267.
Control of moderately raised blood pressure: report of a co-operative
randomized controlled trial. BMJ. 1973;3:434 436.
Carter AB. Hypotensive therapy in stroke survivors. Lancet. 1970;1:
485 489.
Coope J, Warrender TS. Randomised trial of treatment of hypertension in
elderly patients in primary care. BMJ. 1986;293:11451151.
The Dutch TIA Trial Study Group. Trial of secondary prevention with
atenolol after transient ischemic attack or nondisabling ischemic stroke.
Stroke. 1993;24:543548.
Amery A, Birkenhager W, Brixko P, Bulpitt C, Clement D, Deruyttere M,
De Schaepdryver A, Dollery C, Fagard R, Forette F. Mortality and
morbidity results from the European Working Party on High Blood
Pressure in the Elderly trial. Lancet. 1985;1:1349 1354.
Hypertension Detection and Follow-up Program Cooperative Group.
Five-year findings of the Hypertension Detection and Follow-up
Program, I: reduction in mortality of persons with high blood pressure,
including mild hypertension. JAMA. 1979;242:25622571.
Hypertension Detection and Follow-up Program Cooperative Group.
Five-year findings of the Hypertension Detection and Follow-up
Program, II: mortality by race-sex and age. JAMA. 1979;242:25722577.
Hypertension Detection and Follow-up Program Cooperative Group.
Five-year findings of the Hypertension Detection and Follow-up
Program, III: reduction in stroke incidence among persons with high
blood pressure. JAMA. 1982;247:633 638.
The effect of treatment on mortality in mild hypertension: results of the
Hypertension Detection and Follow-up Program. N Engl J Med. 1982;
307:976 980.
Effect of stepped care treatment on the incidence of myocardial infarction
and angina pectoris: 5-year findings of the Hypertension Detection and
Follow-up Program. Hypertension. 1984;6(pt 2):I198 I206.
Hypertension-Stroke Cooperative Study Group. Effect of antihypertensive treatment on stroke recurrence. JAMA. 1974;229:409 418.
MRC Working Party. Medical Research Council trial of treatment of
hypertension in older adults: principal results. BMJ. 1992;304:405 412.
Medical Research Council Working Party. MRC trial of treatment of mild
hypertension: principal results. BMJ. 1985;291:97104.
Helgeland A. Treatment of mild hypertension: a five year controlled drug
trial: the Oslo study. Am J Med. 1980;69:725732.
Leren P, Helgeland A. Oslo Hypertension Study. Drugs. 1986;31(suppl
1):41 45.
SHEP Cooperative Research Group. Prevention of stroke by antihypertensive drug treatment in older persons with isolated systolic hypertension: final results of the Systolic Hypertension in the Elderly Program
(SHEP). JAMA. 1991;265:32553264.
Dahlof B, Lindholm LH, Hansson L, Schersten B, Ekbom T, Wester PO.
Morbidity and mortality in the Swedish Trial in Old Patients with Hypertension (STOP-Hypertension). Lancet. 1991;338:12811285.
Smith WM. Treatment of mild hypertension: results of a ten-year intervention trial. Circ Res. 1977;40(suppl 1):I98 I105.
Effects of treatment on morbidity in hypertension: results in patients with
diastolic blood pressures averaging 115 through 129 mm Hg. JAMA.
1967;202:1028 1034.
Effects of treatment on morbidity in hypertension, II: results in patients
with diastolic blood pressure averaging 90 through 114 mm Hg. JAMA.
1970;213:11431152.
Perry HM Jr. Treatment of mild hypertension: preliminary results of a
two-year feasibility trial. Circ Res. 1977;40(suppl 1):I180 I187.
Veterans AdministrationNational Heart, Lung, and Blood Institute Study
Group for Evaluating Treatment in Mild Hypertension. Evaluation of
drug treatment in mild hypertension: VA-NHLBI feasibility trial: plan
and preliminary results of a two-year feasibility trial for a multicenter
intervention study to evaluate the benefits versus the disadvantages of
treating mild hypertension. Ann N Y Acad Sci. 1978;304:267292.
Wolff FW, Lindeman RD. Effects of treatment in hypertension: results of
a controlled study. J Chron Dis. 1966;19:227240.
Yusuf S, Sleight P, Pogue J, Bosch J, Davies R, Dagenais G, for the Heart
Outcomes Prevention Evaluation Study Investigators. Effects of an angiotensin-converting-enzyme inhibitor, ramipril, on cardiovascular events
in high-risk patients. N Engl J Med. 2000;342:145153.
MacMahon S, Sharpe N, Gamble G, Clague A, Mhurchu CN, Clark T,
Hart H, Scott J, White H, for the PART-2 Collaborative Research Group.

Downloaded from http://stroke.ahajournals.org/ by guest on November 3, 2014

Lawes et al

76.

77.

78.
79.

80.

81.

82.

83.

84.

Randomized, placebo-controlled trial of the angiotensin-converting


enzyme inhibitor, ramipril, in patients with coronary or other occlusive
arterial disease: Prevention of Atherosclerosis with Ramipril. J Am Coll
Cardiol. 2000;36:438 443.
Cashin-Hemphill L, Holmvang G, Chan RC, Pitt B, Dinsmore RE, Lees
RS, for the QUIET Investigators. Angiotensin-converting enzyme inhibition as antiatherosclerotic therapy: no answer yet: QUinapril Ischemic
Event Trial. Am J Cardiol. 1999;83:43 47.
Teo KK, Burton JR, Buller CE, Plante S, Catellier D, Tymchak W, et al.
Long-term effects of cholesterol lowering and angiotensin-converting
enzyme inhibition on coronary atherosclerosis: the Simvastatin/Enalapril
Coronary Atherosclerosis Trial (SCAT). Circulation. 2000;102:
1748 1754.
Eriksson S, Olofsson BO, Wester PO. Atenolol in secondary prevention
after stroke. Cerebrovasc Dis. 1995;5:2125.
Pitt B, Byington RP, Furberg CD, Hunninghake DB, Mancini GB, Miller
ME, Riley W. Effect of amlodipine on the progression of atherosclerosis
and the occurrence of clinical events. Circulation. 2000;102:15031510.
The ALLHAT Officers and Coordinators for the ALLHAT Collaborative
Research Group. Major outcomes in high-risk hypertensive patients randomized to angiotensin-converting enzyme inhibitor or calcium channel
blocker vs diuretic: the Antihypertensive and Lipid-Lowering Treatment
to Prevent Heart Attack Trial (ALLHAT). JAMA. 2002;288:29812997.
Wing LM, Reid CM, Ryan P, Beilin LJ, Brown MA, Jennings GL,
Johnston CI, McNeil JJ, Macdonald GJ, Marley JE, et al. Second Australian National Blood Pressure Study: a comparison of outcomes with
angiotensin-converting enzyme inhibitors and diuretics for hypertension
in the elderly. N Engl J Med. 2003;348:583592.
Hansson L, Lindholm LH, Niskanen L, Lanke J, Hedner T, Niklason A,
Luomanmaki K, Dahlof B, de Faire U, Morlin C, et al. Effect of angiotensin-converting-enzyme inhibition compared with conventional therapy
on cardiovascular morbidity and mortality in hypertension: the Captopril
Prevention Project (CAPPP) randomised trial. Lancet. 1999;353:
611 616.
Casiglia E, Spolaore P, Mazza A, Ginocchio G, Colangeli G, Onesto C,
Di Menza G, Pegoraro L, Ambrosio GB. Effect of two different therapeutic approaches on total and cardiovascular mortality in a Cardiovascular Study in the Elderly (CASTEL). Jpn Heart J. 1994;35:589 600.
Black HR, Elliott WJ, Grandits G, Grambsch P, Lucente T, White WB,
Neaton JD, Grimm RH Jr, Hansson L, Lacourciere Y, et al. Principal

Blood Pressure and Stroke: An Overview

85.

86.

87.

88.

89.

90.

91.

92.

93.

785

results of the Controlled Onset Verapamil Investigation of Cardiovascular


End Points (CONVINCE) trial. JAMA. 2003;289:20732082.
Zanchetti A, Bond MG, Hennig M, Neiss A, Mancia G, Dal Palu C,
Hansson L, Magnani B, Rahn KH, Reid JL, et al. Calcium antagonist
lacidipine slows down progression of asymptomatic carotid atherosclerosis: principal results of the European Lacidipine Study on Atherosclerosis (ELSA), a randomized, double-blind, long-term trial. Circulation.
2002;106:24222427.
Brown MJ, Palmer CR, Castaigne A, de Leeuw PW, Mancia G, Rosenthal
T, Ruilope LM. Morbidity and mortality in patients randomised to
double-blind treatment with a long-acting calcium-channel blocker or
diuretic in the international nifedipine GITS study: Intervention as a Goal
in Hypertension Treatment (INSIGHT). Lancet. 2000;356:366 372.
Borhani NO, Mercuri M, Borhani PA, Buckalew VM, Canossa-Terris M,
Carr AA, Kappagoda T, Rocco MV, Schnaper HW, Sowers JR, Bond
MG. Final outcome results of the Multicenter Isradipine Diuretic Atherosclerosis Study (MIDAS): a randomized controlled trial. JAMA. 1996;
276:785791.
National Intervention Cooperative Study in Elderly Hypertensives Study
Group. Randomized double-blind comparison of a calcium antagonist and
a diuretic in elderly hypertensives. Hypertension. 1999;34:1129 1133.
Hansson L, Hedner T, Lund-Johansen P, Kjeldsen SE, Lindholm LH,
Syvertsen JO, Lanke J, de Faire U, Dahlof B, Karlberg BE. Randomised
trial of effects of calcium antagonists compared with diuretics and betablockers on cardiovascular morbidity and mortality in hypertension: the
Nordic Diltiazem (NORDIL) study. Lancet. 2000;356:359 365.
Zanchetti A, Rosei EA, Dal Palu C, Leonetti G, Magnani B, Pessina A.
The Verapamil in Hypertension and Atherosclerosis Study (VHAS):
results of long-term randomized treatment with either verapamil or
chlorthalidone on carotid intima-media thickness. J Hypertens. 1998;16:
16671676.
Rosei EA, Dal Palu C, Leonetti G, Magnani B, Pessina A, Zanchetti A,
for the VHAS Investigators. Clinical results of the Verapamil in Hypertension and Atherosclerosis Study. J Hypertens. 1997;15:13371344.
Tatti P, Pahor M, Byington RP, Di Mauro P, Guarisco R, Strollo G,
Strollo F. Outcome results of the Fosinopril Versus Amlodipine Cardiovascular Events Randomized Trial (FACET) in patients with hypertension and NIDDM. Diabetes Care. 1998;21:597 603.
Easton DF, Peto J, Babiker AG. Floating absolute risk: an alternative to
relative risk in survival and case-control analysis avoiding an arbitrary
reference group. Stat Med. 1991;10:10251035.

Downloaded from http://stroke.ahajournals.org/ by guest on November 3, 2014

Correction
The article entitled Blood Pressure and Stroke: An Overview of Published Reviews, by Lawes
et al1 contained incorrect data resulting from a production error. The corrected article appears
below. The publisher regrets the error.

Blood Pressure and Stroke


An Overview of Published Reviews
Carlene M.M. Lawes, MBChB, FAFPHM, PhD; Derrick A. Bennett, MSc, PhD, CStat;
Valery L. Feigin, MD, MSc, PhD; Anthony Rodgers, MBChB, FAFPHM, PhD
BackgroundThe last few years have seen a considerable increase in the amount of information available concerning
blood pressure (BP) and stroke associations. This article provides an overview of published reviews of the effects on
stroke seen in trials of BP-lowering drugs and compares these with the results available from cohort studies.
Summary of ReviewWe present a review of major overviews of prospective cohort studies and an updated meta-analysis
of 40 randomized controlled trials of BP lowering, which included 188 000 participants and approximately 6800
stroke events. Cohort studies now indicate that in the Asia Pacific region as well as in North America and Western
Europe, each 10 mm Hg lower systolic BP is associated with a decrease in risk of stroke of approximately one third in
subjects aged 60 to 79 years. The association is continuous down to levels of at least 115/75 mm Hg and is consistent
across sexes, regions, and stroke subtypes and for fatal and nonfatal events. The proportional association is age
dependent but is still strong and positive in those aged 80 years. Data from randomized controlled trials, in which mean
age at event was approximately 70 years, indicate that a 10 mm Hg reduction in systolic BP is associated with a
reduction in risk of stroke of approximately one third. Per mm Hg systolic BP reduction, the relative benefits for stroke
appear similar between agents, by baseline BP levels, and whether or not individuals have a past history of
cardiovascular disease. There is, however, evidence of greater benefit with a larger BP reduction.
ConclusionsThe epidemiologically expected benefits of BP lowering for stroke risk reduction are broadly consistent
across a range of different population subgroups. There are greater benefits from larger BP reductions, and initiating and
maintaining BP reduction for stroke prevention is a more important issue than choice of initial agent. (Stroke. 2004;35:
1024-1033.)
Key Words: blood pressure cohort studies epidemiology meta-analysis randomized controlled trials stroke

he risk of stroke increases continuously above blood


pressure (BP) levels of approximately 115/75 mm Hg.
Since the association is steep, and BP levels are high in most
adult populations, almost two thirds of stroke burden globally
is attributable to nonoptimal BP (ie, 115/75 mm Hg).1
Approximately two thirds of this burden occurs in middleaged subjects (45 to 69 years), and approximately two thirds
occurs in developing regions.
In 1990 it was demonstrated that in North American and
Western European populations, a 5 mm Hg lower diastolic
blood pressure (DBP) was associated with a 30% to 40%
lower stroke risk,2 and this was reversed within a few years of

BP lowering.3 Since then, further evidence has become


available, especially in the elderly and in the Asia-Pacific
region and from many more trials comparing different agents,
more intensive and less intensive BP-lowering regimens, and
BP lowering in those without hypertension. This article
reviews published trial meta-analyses of the impact of pharmacological BP-lowering therapies on stroke across all population groups. The results are compared with overviews of
cohort study data that assessed the association between BP
and stroke in a range of populations. The main gaps in current
research and the clinical and public health implications of the
association of BP with stroke are also discussed.

Received August 14, 2003; final revision received October 7, 2003; accepted November 19, 2003.
From the Clinical Trials Research Unit, Department of Medicine, University of Auckland, Auckland, New Zealand.
Correspondence to Carlene Lawes, Clinical Trials Research Unit, Department of Medicine, University of Auckland, Private Bag 92019, Auckland, New
Zealand. E-mail c.lawes@ctru.auckland.ac.nz
1
[Correction for Vol 35, Number 3, March 2004. Pages 776 785.]
2004 American Heart Association, Inc.
Stroke is available at http://www.strokeaha.org

DOI: 10.1161/01.STR.0000126208.14181.DD

1024

Lawes et al

Methods
Prospective Cohort Studies
Published overviews of cohort studies were identified by a search of
MEDLINE and EMBASE databases with key words overview,
meta-analysis, blood pressure, stroke or cerebrovascular or cardiovascular disease, cohort study or prospective study, and comparative
study. Overviews were included if they were published or data were
accessible by May 1, 2003, and if they provided data on inclusion
criteria, number and regions of the cohort studies included, number
of participants and demographics, duration of follow-up, disease end
points, and methods of analyses. Results were independently extracted and summarized by 2 reviewers, but no additional analyses
were conducted.

Randomized Controlled Trials


Meta-analyses of trials of BP-lowering agents were identified by a
MEDLINE and EMBASE search with search terms overview,
meta-analysis, blood pressure or hypertension, stroke or cerebrovascular or cardiovascular disease, and randomized controlled trial.
Data from individual trials included in these meta-analyses were
included if they were published before May 1, 2003, and if they
provided data on inclusion criteria, total number of participants and
demographics, randomization method, trial interventions, trial end
points, duration of follow-up, and net reduction in BP (ie, treatment

TABLE 1.

Blood Pressure and Stroke: An Overview

1025

effect). Data from individual studies, from both the trial and
meta-analytic publications, were independently extracted by 2 reviewers. In some cases, existing meta-analyses of trials311 included
the most recent data on individual trials, and therefore these were
included in the final meta-analyses.
Core baseline and outcome data were extracted, and trials included
in the meta-analysis were stratified into the following 3 subgroups to
reduce clinical heterogeneity: (1) drug versus placebo or no treatment, (2) more intensive dose of drug versus less intensive dose of
drug, and (3) drug versus drug. The net difference in BP between, for
example, the treatment and placebo groups was calculated; this
reflects the true difference in BP achieved as the influence of
regression to the mean (which would be evident in both groups) is
removed. The updated meta-analysis was based on summary data
rather than individual participant data. The relative risk was computed as the measure of effect size for each trial and subgroup of
trials. Statistical heterogeneity was assessed with the use of the
standard Q statistic for heterogeneity, with a probability value of
0.1 used to indicate evidence of heterogeneity. A less stringent
probability value of 0.1 was used because it is well known that tests
of heterogeneity have low statistical power. If there was no evidence
of statistical heterogeneity when this criterion was used, then the
results were combined with the use of the standard inverse-variance
weighted fixed effect approach.12 If there was evidence of statistical
heterogeneity between the trials, attempts were made to explore

Major Overviews of Prospective Cohort Studies

No. of cohorts
Countries included
Inclusion criteria

Prospective Studies Collaboration (2002)

Asia Pacific Cohort Studies Collaboration (2003)

61

37

Europe, United States, Asia, Australia

Mainland China, Japan, Hong Kong, Taiwan, Singapore,


South Korea, New Zealand, Australia

At least 5000 person-years of follow-up


recorded or planned

Population from the Asia Pacific region

Date of birth or age, sex, blood pressure,


and cholesterol recorded at baseline

At least 5000 person-years of follow-up recorded


or planned

Date of death or age at death recorded


during follow-up

Date of birth or age, sex, and blood pressure


recorded at baseline
Date of death or age at death recorded during
follow-up

No. of participants

958 074

425 325

% male

NR

57% (range 34%-100%)

Mean age (range) at baseline

NR

47 (20107) years

Range 417 years

Range 227 years

Mean 12 years

Mean 7 years

12.7 million person years

3.0 million person years

11 960 strokes

5178 strokes

Follow-up duration

Stroke end points

100% fatal

50% fatal

RR with a 10-mm Hg decrease in SBP

NR

0.63 (95% CI 0.610.64)

Fatal versus nonfatal

NR

(37% decrease in stroke risk)


No significant difference

Association by sex

No significant difference 60 years of age

No significant difference

Association by age

For a 10-mm Hg lower SBP

For a 10-mm Hg lower SBP

4049 years RR0.60 (0.570.63)

60 years RR0.46 (0.440.47)

5059 years RR0.62 (0.590.63)

6069 years RR0.64 (0.620.66)

6069 years RR0.66 (0.640.67)

70 years RR0.75 (0.720.78)

7079 years RR0.71 (0.690.72)


8089 years RR0.82 (0.790.84)
CVD indicates cardiovascular disease; DBP, diastolic blood pressure; SBP, systolic blood pressure; NR, not reported; RR, relative risk; CI, confidence interval.
Prospective Studies Collaboration (2002)16 includes MacMahon et al (1990)2 and Prospective Studies Collaboration (1995),15 and Asia Pacific Cohort Studies
Collaboration (2003)18 includes Eastern Collaborative Research Group (1998).17

1026

Stroke

April 2004

Figure 1. Usual SBP and risk of stroke by


age, with data from prospective cohort study
overviews. Data are shown from the
APCSC18 and PSC16 demonstrating the
associations between usual SBP and risk of
stroke plotted by age groups defined by age
at risk on a log scale. The x-axis coordinate
for each group is the mean follow-up SBP.
The y-axis coordinate is the hazard ratio
computed by the floating absolute risk technique,93 which enabled the comparison
between pairs of exposure groups rather
than comparisons between each exposure
group and a reference group. The solid
squares are larger where there are more
events because their size is proportional to
the inverse variance; vertical lines represent
95% CI. (Figures reprinted with permission
from Elsevier [Lancet] and Lippincott Williams & Wilkins [J Hypertens.])

potential sources of heterogeneity. If this was not possible, the


random effects approach of DerSimonian and Laird13 was used as
part of a sensitivity analysis to assess the robustness of the conclusions. When there were sufficient studies, evidence of publication
bias was checked with the use of standard funnel plots.14
For each subgroup of trials, the weighted mean age, weighted
mean follow-up, and weighted mean BP reduction were estimated.
Trials were weighted by sample size and, in a sensitivity analysis, by
number of stroke events.

Results
Prospective Cohort Studies
Five overviews of cohort studies assessing the association
between BP and cardiovascular disease were identified. The
MacMahon2 and Prospective Studies Collaboration (PSC)15,16
overviews included cohorts predominantly from the United
Kingdom, United States, and Europe. Cohorts in the MacMahon2 and first PSC15 review were subsequently included in
the later PSC review.16 The Eastern Stroke and Coronary
Heart Disease Collaborative Research17 overview is a subset
of the Asia Pacific Cohort Studies Collaboration (APCSC),18
which comprises cohorts from Mainland China, Hong Kong,
Taiwan, Japan, Singapore, South Korea, Australia, and New
Zealand. The PSC and APCSC overviews included almost 1
million and 420 000 individuals, respectively; both analyzed individual participant data and corrected for regression
dilution bias to reflect true or usual BP levels.16,18 In total,
7 cohorts were included in both the PSC and APCSC
overviews, representing 5% and 10% of participants,
respectively. The main characteristics and results of the
overviews are summarized in Table 1.
An important finding in all overviews was that the association between BP and risk of stroke was continuous and log
linear. Therefore, a given absolute difference in usual BP was
associated with a similar relative risk reduction of stroke at all
levels of BP. There was no evidence of a threshold below
which levels of BP were no longer associated with lower
relative risk of stroke, down to approximately 115 mm Hg
systolic blood pressure (SBP) or 75 mm Hg DBP.16,18 All
overviews commented on the statistical heterogeneity between studies, and age appeared to be an important factor,16,18
explaining approximately 80% of the between-study hetero-

geneity in APCSC.18 The strength of the overall association


was not altered by restricting analyses to those with and
without a history of coronary heart disease at baseline or by
controlling for additional cardiovascular risk factors.16,18
The overall risk estimates provided by APCSC,18 PSC,16 and
previous meta-analyses2,15,17 (a 5 mm Hg lower DBP or
10 mm Hg lower SBP was associated with approximately a 30%
to 40% lower risk of stroke) were influenced by the mean age of
the cohorts, as the association of BP with stroke varied with age.
The proportional change in stroke risk per unit change in BP was
less extreme in old age than in middle age (Figure 1), but the
relationship remained strong and positive for all age groups.16,18
Age attenuation limits the ability to directly compare the
summary relative risk estimates across the overviews, but the 2
recent overviews that have published age-specific relative risks
had similar results (Table 1)16,18; a 10 mm Hg lower SBP was
associated with approximately a 40% to 50%, 30% to 40%, and
20% to 30% lower risk of stroke in those aged 60 years, 60 to
69 years, and 70 years, respectively. There has been no
evidence in any of the overviews that the strength of association
between BP and stroke varied by sex2,15,16,18 or for fatal and
nonfatal stroke events.2,18 Regional comparisons in APCSC and
PSC suggested that the association between BP and stroke in
Asia was similar or even slightly steeper than in Australia and
New Zealand,18 Europe,16 and the Unites States16 after standardization for age. In these 2 overviews the associations of BP with
stroke subtypes were broadly similar,16,18 although low rates of
CT/MRI confirmation would tend to obscure any differences.

Randomized Controlled Trials


Randomized controlled trials provide data on the impact of
BP lowering on short-term stroke outcomes, which is relevant
in assessing the extent to which the epidemiologically expected reductions in stroke are produced.2,3
Randomized Controlled Trials Comparing Antihypertensive
Drugs With a Placebo (or No Treatment)
Seventeen trials compared the effects of -blocker and/or
diuretic with a placebo or no treatment; in 6 trials the
comparison drug was angiotensin-converting enzyme (ACE)
inhibitors, and in 2 it was calcium antagonists (Table

Lawes et al
TABLE 2.

Blood Pressure and Stroke: An Overview

1027

Randomized Controlled Trials Comparing Antihypertensive Drugs to a Placebo (or No Treatment)


Treatment

Trial

Placebo

Net Difference*
N

Follow-Up,
y

Mean Age,
y

%
Male

SBP

DBP

RR and 95% CI

Beta-blocker and/or diuretic vs placebo or no treatment


ANBPS50

13

1721

22

1706

4.0

50

63

NR

0.59 (0.301.16)

58

58

2.0

56

43

15

10

0.99 (0.049.56)

Carter et al52

10

49

21

48

4.0

NA

57

NR

NR

0.47 (0.250.88)

53

Coope

20

419

39

465

4.4

69

31

18

11

0.57 (0.340.96)

Dutch TIA54

52

732

62

741

2.6

65

64

0.85 (0.601.21)

EWPHE55

32

416

48

424

4.7

72

30

21

0.68 (0.441.04)

HDFP5660

102

5485

158

5455

5.0

51

55

NR

0.64 (0.500.82)
0.78 (0.541.11)

Barraclough51

HSCSG I II III61

43

233

52

219

2.3

59

41

25

12

MRC older62

101

2183

134

2213

5.8

70

42

14

0.76 (0.590.98)

MRC young63

60

8700

109

8654

5.0

52

52

11

0.55 (0.400.75)

406

379

5.5

45

100

17

10

0.08 (0.001.53)
0.65 (0.510.83)

Oslo64,65
66

SHEP

105

2365

162

2371

4.5

72

43

12

STOP H67

30

812

55

815

2.1

76

37

20

0.55 (0.350.85)

USPHS68

193

196

7.0

44

80

15

10

0.17 (0.021.39)

VA6970

254

23

257

3.8

51

100

17

10

0.26 (0.110.64)

508

504

1.5

38

81

NR

0.99 (0.0249.9)

45

42

1.4

49

32

NR

20

1.87 (0.1819.8)

577

24 579

897

24 547

4.6

57

53

13

0.65 (0.590.73)

156

4654

226

4652

4.5

66

73

0.69 (0.570.84)

308

309

4.7

61

82

1.76 (0.525.94)

307

3051

420

3054

3.9

64

70

0.73 (0.640.84)

QUIET

878

872

2.3

58

82

NR

NR

0.99 (0.0615.8)

SCAT77

229

231

4.0

61

89

0.22 (0.051.03)

71,72

VA-NHLBI

Wolff et al73
Summary
ACE inhibitors vs placebo
HOPE74
PART 275
PROGRESS27
76

TEST78
Summary

74

372

69

348

2.6

70

60

1.00 (0.751.35)

547

9492

729

9466

4.3

64

73

0.72 (0.680.83)

Calcium antagonists vs placebo


PREVENT79

417

408

3.0

57

80

0.98 (0.293.35)

Syst-Eur47

47

2398

77

2297

2.6

70

33

11

0.58 (0.410.84)

Summary

52

2815

82

2705

2.8

68

40

10

0.61 (0.440.85)

nnumber of events; Nnumber of participants. SBP indicates systolic blood pressure; DBP, diastolic blood pressure; RR, relative risk; 95% CI, 95% confidence
interval; NR, not recorded.
*Net BP differenceblood pressure difference between the randomized groups in mm Hg (ie, BP change in treatment groupBP change in placebo group).
Summary follow-up, mean age, % male, and net SBP and DBP change are weighted by study size for those studies that provided data on net changes in both
SBP and DBP. Weighting by number of events made negligible difference. The pooled RR and 95% CI (using the fixed effects approach) for each of the 3 trial subgroups
are presented, and the test for heterogeneity (Q statistic and P value) are Q14.23, P0.36; Q8.90, P0.11; and Q0.61, P0.43, respectively.
These trials also included older classes of drugs such as methyldopa and alkaloids.
There were no stroke events in the Barraclough or VA-NHLBI trial; however, for the purposes of meta-analyses, 0.5 was added to all cells in the 22 table.

2).27,47,50 79 In total, there were 73 500 participants and


almost 2900 stroke events recorded in these trials. Weighted
mean follow-up time for the 3 groups of trials ranged from
2.8 to 4.6 years, and the weighted mean age of participants
ranged from 57 to 68 years. There was clear evidence of a
reduction in stroke risk with BP lowering with each drug
versus placebo comparison (P0.005), with pooled relative
risk reductions for -blocker and/or diuretic, ACE inhibitors,
and calcium antagonists of 35%, 28%, and 39%, respec-

tively. These broadly similar relative risk reductions were


achieved with weighted mean reductions in BP (SBP/DBP)
of 13/6, 5/2, and 10/5 mm Hg, respectively. While the BP
reductions appeared less for the ACE inhibitor versus
placebo trials, there is also greater uncertainty over the BP
reductions in this subgroup.19 There was no evidence of
small study bias20 as assessed by a funnel plot14 (not
shown) and no significant statistical heterogeneity in any
of the 3 subgroup comparisons.

1028

Stroke

April 2004

Figure 2. Randomized controlled trials comparing antihypertensive drugs with a placebo (or no treatment) by subgroup. The
meta-analyses of BP-lowering trials are presented stratified into
subgroups on the basis of mean age of trial participants at
entry, baseline SBP level, and whether trial participants predominantly had a history of stroke/transient ischemic attack (TIA) or
vascular disease. The diamonds are centered on the pooled
estimate of effect and represent 95% CI. The solid diamond
represents the pooled relative risk and 95% CI for all contributing trials.

The trials comparing antihypertensive drugs with a placebo


(or no treatment) were stratified by several trial characteristics such as mean age at entry, baseline BP, and history of
cardiovascular disease (Figure 2). Overall, there was a reduction in risk of stroke of 30% and a risk reduction of
approximately 30% to 40% in most subgroups. However, in
several groups the risk reduction appeared to be greater with
larger net reduction in BP, for example, in those with no past
history of stroke/transient ischemic attack or vascular disease.
Randomized Controlled Trials Comparing More Intensive
Versus Less Intensive Antihypertensive Drug Regimens
Three individually inconclusive clinical trials including a
total of 20 408 participants and 384 stroke events compared
treatment regimens of different intensities2124 (Table 3). The
meta-analysis suggested that additional benefit may be gained
from a more intensive treatment regimen (relative risk reduction of 20%; P0.04). There was no significant statistical
heterogeneity.
TABLE 3.

More

ABCD22
21

UKPDS-HDS
Summary

Net Reduction in BP and Relative Risk Reduction in Stroke


A meta-regression analysis was used to assess the direct
relationship between the net reduction in SBP in the 7 trial
meta-analyses. The net reduction in SBP was plotted against
the relative risk reduction in stroke for each meta-analysis.
This plot suggested that a dose-response relationship may
exist (Figure 3). A weighted linear regression line was
superimposed in which weights are equal to the inverse of the
variance for each trial meta-analysis. The slope of this
weighted regression line indicated that a 10 mm Hg greater
reduction in SBP would be associated with a reduction in risk
of stroke of 31% (R 20.71). The mean weighted follow-up
duration for all these trials was 4.5 years, and the mean
weighted age of participants at entry into trials was 63 years
(the mean age at event is likely to be approximately a decade
later). Age-specific analyses from the 2 cohort study overviews16,18 estimated that a 10 mm Hg lower SBP in participants aged 60 to 69 years at event was associated with a risk
reduction in stroke of approximately 35%; for those aged
70 years the risk reduction was 25% to 29% (Table 5). The
results from cohort studies and randomized controlled trials
are therefore broadly consistent, suggesting that most of the
epidemiologically expected benefit of BP lowering is
achieved within a few years for stroke.

Randomized Controlled Trials Comparing More Intensive vs Less Intensive Antihypertensive Drug Regimens

More Intensive vs
Less Intensive Trials
HOT

Randomized Controlled Trials Comparing Different


Antihypertensive Drugs
Fifteen trials compared the effects of different types of
antihypertensive drugs, with 2 trials including several drug
versus drug comparisons (Table 4).2224,46,80 92 In total, there
were 96 000 participants and almost 3600 stroke events
recorded over a mean follow-up time of 4 to 5 years. The
weighted mean reduction in BP in many of the drug versus
drug trials was small, often 1 mm Hg SBP and DBP (Table
4). Overall, these trials indicated that there was little difference between the drug classes, with relative risk reductions of
9% (-blockers and/or diuretics versus ACE inhibitors),
8% (-blockers and/or diuretics versus calcium antagonists), and 11% (calcium antagonists versus ACE inhibitors).
These results were of borderline statistical significance. The
latter subgroup was the only one in which there was evidence
of statistical heterogeneity (P0.09), and a random effects
model gave a nonsignificant relative risk reduction of 5%.

23,24

Less
N

Net Difference*
N

Follow-Up,
y

Mean Age,
y

%
Male

SBP

DBP

RR and 95% CI

5.3

58

67

0.98 (0.402.43)

237

233

89

6262

205

12528

3.8

62

53

0.87 (0.681.11)

38

758

34

390

8.4

56

55

10

0.58 (0.370.90)

136

7257

248

13 151

4.2

62

53

0.80 (0.650.99)

nno. of events, Nno. of participants. SBP indicates systolic blood pressure; DBP, diastolic blood pressure; RR, relative risk; 95% CI, 95% confidence interval.
* Net BP differenceblood pressure difference between the randomized groups (mm Hg) (ie, BP change in more intensive drug treatment groupBP change in
less intensive drug treatment group).
Summary follow-up, mean age, % male, and net SBP and DBP change are weighted by study size for those studies that provided data on net changes in both
SBP and DBP. Weighting by number of events made negligible difference. The pooled RR and 95% confidence interval (using the fixed effects approach) are presented,
and the test for heterogeneity (Q statistic and P value) are Q2.70, P0.26.

Lawes et al
TABLE 4.

Blood Pressure and Stroke: An Overview

1029

Randomized Controlled Trials Comparing Different Blood PressureLowering Drugs


First Drug

Drug vs Drug
Trials

Second Drug
N

Net Difference*
Follow-Up,
y

Mean Age,
y

%
Male

SBP

DBP

RR and 95% CI

Beta-blockers and/or diuretics vs ACE inhibitors


ALLHAT80

675

15255

457

9054

4.9

67

53

0.88 (0.780.98)

ANBPS81

107

3039

112

3044

4.1

72

49

0.95 (0.741.23)

CAPPP82

148

5493

189

5492

6.1

53

53

0.78 (0.630.97)

237

2213

215

2205

5.0

76

33

1.10 (0.931.32)

17

358

21

400

8.4

56

54

0.90 (0.491.69)

1184

26 358

994

20 195

5.1

65

51

0.91 (0.830.99)

46

STOP-2

UKPDS-HDS
Summary

23,24

Beta-blockers and/or diuretics vs calcium antagonists


ALLHAT80

675

15255

377

9048

4.9

67

53

1.06 (0.931.20)

CASTEL83

205

146

7.0

73

51

0.71 (2.440.21)

118

8297

133

8179

3.0

66

44

0.88 (0.681.12)

ELSA

14

1157

1177

4.0

56

55

1.59 (0.693.70)

INSIGHT86

74

3164

67

3157

4.0

65

46

1.06 (0.781.43)

MIDAS87

441

442

3.0

59

78

0.50 (0.132.00)

NICS-EH88

1.33 (0.473.85)

CONVINCE84
85

214

215

5.0

70

33

89

NORDIL

196

5471

159

5410

5.0

60

49

1.22 (0.991.49)

STOP-246

237

2213

207

2196

5.0

76

33

1.14 (0.951.35)

VHAS90,91

707

707

2.0

54

49

0.80 (0.222.94)

1334

37 124

974

30 677

4.2

65

48

1.08 (0.991.16)

Summary

Calcium antagonists vs ACE inhibitors


ABCD22
ALLHAT80
92

FACET

11

235

235

5.3

58

67

1.57 (0.623.98)

377

9048

457

9054

4.9

67

53

0.83 (0.720.94)
2.47 (0.797.75)

10

191

189

3.5

63

41

STOP-246

207

2196

215

2205

5.0

76

33

0.97 (0.811.16)

Summary

605

11 670

683

11 683

4.9

68

49

0.89 (0.800.99)

nno. of events, Nno. of participants; SBP indicates systolic blood pressure; DBP, diastolic blood pressure; RR, relative risk; 95% CI, 95% confidence interval.
*Net BP differenceblood pressure difference between the randomized groups (mm Hg). Presented as the BP change in (first listed drug)(second listed drug),
ie, a negative value means that the second listed drug achieved a greater blood pressure reduction than the first listed drug.
Summary follow-up, mean age, % male, and net SBP and DBP change are weighted by study size for those studies that provided data on net changes in both
SBP and DBP. Weighting by number of events made negligible difference. The pooled RR and 95% confidence interval (using the fixed effects approach) for each
of the 3 trial subgroups are presented, and the test for heterogeneity (Q statistic and P value) are Q6.91, P0.14; Q7.36, P0.59; and Q6.53, P0.09,
respectively. The pooled RR and 95% CI using a random effects approach for the last trial subgroup was 0.95 (0.76 1.19).

Discussion
Cohort studies have demonstrated that the strong association
between BP and stroke appears to be continuous down to
levels of at least 115/75 mm Hg. The strength of the associations was similar for men and women and for fatal and
nonfatal events but attenuated with age.2,1518 Despite a
potentially weaker relative association for older age groups,
the higher overall stroke rate in this group means that in
absolute terms, the benefits of a given reduction in BP are
likely to be greater. The analyses have not consistently
demonstrated a difference in the association between intracerebral hemorrhage and ischemic stroke; however, only approximately one half of stroke subtypes were confirmed (CT,
MRI, or autopsy). Regional comparisons are very limited
because most of the cohort studies have been conducted
primarily in the developed parts of the world; however, the

broad consistency of the results across overviews was more


evident than any differences. This suggests that the proportional
associations are likely to be reasonably generalizable to the
many populations in the world (eg, Africa, South America) for
which there are limited data on stroke epidemiology.25
Data from trials provide more reliable data on the likely
short-term benefits of BP lowering on stroke. As with the cohort
data, randomized controlled trials have been conducted predominantly in populations of Western countries, and data for developing countries (where approximately three quarters of stroke
deaths occur worldwide26) are very scarce. The results of the
updated meta-analysis of trials are consistent with those of
several previous meta-analyses and confirmed the risk reduction
of stroke with BP lowering compared with placebo (regardless
of the agent used) as being between 30% and 40%. Previous
analyses have also found no evidence of a difference in effect

1030

Stroke

April 2004

Figure 3. Net reduction in SBP and relative risk reduction in


stroke in randomized controlled trials of BP lowering. In a metaregression, the direct relationship between the net reduction in
SBP was plotted against the relative risk reduction in stroke for
each of the 7 trial meta-analyses. The slope of the weighted linear regression line (where weights are equal to the inverse of
the variance for each trial meta-analysis) indicated that a
10 mm Hg greater reduction in SBP would be associated with a
reduction in risk of stroke of 31% (R 20.71). From left to right,
the diamonds represent the meta-analyses for -blockers
and/or diuretics vs calcium antagonists, calcium antagonists vs
ACE inhibitors, and -blockers and/or diuretics vs ACE inhibitors; the circle represents the more intensive vs less intensive
antihypertensive drug regimens; and the squares are ACE inhibitors vs placebo, calcium antagonists vs placebo, and
-blockers and/or diuretics vs placebo or no treatment. The
sizes of the diamonds, circle, and squares are larger where
there are more events because their size is proportional to the
inverse variance of that comparison; vertical lines represent
95% CI.

size for men and women or for fatal and nonfatal events.3 8
Limited data were available on treatment effects by age, and the
treatment effects by age within individual trials27 and between
trials4,5,9,10,28 have not been consistent. However, the trials were
often not powered for age subgroup analyses and often included
only a narrow age band. Few data are available on stroke
subtypes, limiting the scope for meta-analyses. The limited data
from the Perindopril Protection Against Recurrent Stroke Study
(PROGRESS)27 and the Heart Outcomes Prevention Evaluation
(HOPE) study29 are insufficient to draw definitive conclusions
regarding the impact of BP lowering on stroke subtypes. Anal-

yses have indicated, however, that a more intensive BP-lowering


regimen with a given agent may produce a greater risk reduction
than a less intensive regimen and that overall the trials demonstrate a dose-response relationship between magnitude of BP
reduction and reduction in risk of stroke.
When different classes of BP-lowering drugs were directly
compared, there was very little difference in either the magnitude of the BP reduction achieved or the impact of different
agents on stroke. The similarities between the effects of different
drug classes are more striking than the differences, but there is
uncertainty over the extent to which any differences are explained by differences in BP reduction and chance. One arm of
the Antihypertensive and Lipid-Lowering Treatment to Prevent
Heart Attack Trial (ALLHAT)30 and the Losartan Intervention
For Endpoint reduction in hypertension study (LIFE)31,32 were
not included in the meta-analysis because no other trials included the same classes of drugs. They indicated that there was
a higher risk of stroke in the -adrenergic blocker group
compared with diuretics30 and that angiotensin II blockers may
have greater impact on stroke than -blockers.31,32

Shape and Size of Association Between BP


and Stroke
There has been considerable debate over the shape of the
association between BP and stroke. The data presented here
from cohort studies suggested a continuous log-linear association. However, some other cohort data appear to have
demonstrated a J-shaped curve,3337 and it has been suggested
that this relationship indicates that BP has been lowered too
far and that cerebral blood flow is compromised, leading to
ischemia. It is more likely to reflect deteriorating health,
which accounts for both falling pressure and a greater
likelihood of a cardiovascular disease event, but the 2 factors
may not be causally related.38 43 Several trials have found no
evidence of a J-curve association despite including subgroups
of patients susceptible to reduced perfusion.44 47 The large
PROGRESS48 and HOPE29 trials found no evidence of a
J-curve association across a wide range of BP levels. Overall,
these results support a log-linear association and indicate that
a potential J-curve relationship should not detract attention
from the major benefits of BP lowering.
A key finding from this article is that the 7 meta-analyses
of randomized controlled trials of BP lowering and stroke

TABLE 5. Relative Risk Reduction in Stroke With a Reduction in Systolic Blood Pressure Predicted From
Cohort Studies and Observed From Randomized Controlled Trials
Observed Effects on Stroke With a Reduction
in SBP of 10 mm Hg

Predicted Effects on Stroke of Lowering SBP 10 mm Hg*


Mean Age at Event

Prospective Studies Collaboration (2002)16


18

Asia Pacific Cohort Studies Collaboration (2003)

Estimated Mean Age at Event

6069 Years

7079 Years

Approximately 73 Years

34%

29%

31%

36%

25%

SBP indicates systolic blood pressure.


*Relative risk reduction in stroke from cohort studies with a 10-mm Hg lower SBP, by age at event.
Relative risk reduction in stroke from randomized controlled trials with a 10-mm Hg reduction in SBP based on the slope of the
regression line in Figure 3. Mean age at entry into trials was 63 years, and mean age at event is likely to approximately a decade
later, ie, 73 years.

Lawes et al
suggest a dose-response relationship. At mean age at baseline
of approximately 63 years, a 10 mm Hg reduction in SBP was
associated with a risk reduction in stroke of 31%. Given that
most of these individuals would not have experienced their
stroke event until approximately a decade later, the results are
consistent with the predicted reduction in stroke risk from the
cohort study data. This indicates that most of the epidemiologically expected benefit of BP lowering may be achieved
within a few years for stroke.

Future Research and Implications for Prevention


Research and analyses from cohort study collaborations such
as APCSC and meta-analysis of individual participant trial
data improve the understanding of the associations between
BP and stroke and the potential benefits of BP lowering in
different population subgroups. Additional analyses are able
to examine the combined impact of multiple risk factors on
stroke risk. BP intervention strategies would be greatly
enhanced if these data were also used in the improvement of
risk prediction frameworks to better identify those at higher
absolute risk of cardiovascular disease. Development of
simple versions of these risk prediction tools would be of
particular value in developing regions where much of the
future cardiovascular disease burden is likely to occur.1,49
Future trial analyses should focus on the presence or
absence of age attenuation in the effect of BP lowering. This
would require comparison of age-specific, follow-up time
specific, and end pointspecific analyses of individual participant data from both observational studies (such as APCSC)
and clinical trials (such as the Blood Pressure Lowering
Treatment Trialists Collaboration11). Given that most individuals require combination therapy, it is also important to
shift the focus of trials from comparisons of the short-term
benefits of individual drugs. Instead, future trials should aim
to determine the optimal combination therapy and the potential long-term benefits of different antihypertensive regimens
in different population subgroups. While there may be some
differences between BP-lowering agents, the relative differences between these agents are minor compared with the
relative benefits of any BP-lowering agent versus no
treatment.
Finally, it is important to acknowledge that treatment of
high-risk groups will only partially address the growing
cardiovascular disease epidemic. Data presented in this article
indicated that the relative benefits of BP lowering are not
only limited to those classified as hypertensive but would also
be evident in those with normal or below normal BP. These
relative benefits appear broadly consistent across many population subgroups. Other important areas of research therefore include exploration of cost-effective strategies to bring
about population changes in cardiovascular risk factors such
as BP. Such initiatives would complement targeted treatment
approaches and would not only benefit the developed world
but could potentially arrest the projected increase in cardiovascular disease in developing countries.

Acknowledgments
This study was supported by a Health Research Council (New
Zealand) fellowship to Dr Lawes. Dr Rodgers is a National Health

Blood Pressure and Stroke: An Overview

1031

Foundation (New Zealand) Senior Research Fellow. We thank


Varsha Parag and Ruey-Bin Lin for statistical assistance and Clarissa
Could-Thorpe for secretarial support.

References
1. World Health Organisation. The World Health Report 2002: Reducing
Risks, Promoting Healthy Life. Geneva, Switzerland: World Health
Organisation; 2002.
2. MacMahon S, Peto R, Cutler J, Collins R, Sorlie P, Neaton J, Abbott R,
Godwin J, Dyer A, Stamler J. Blood pressure, stroke, and coronary heart
disease, part I: prolonged differences in blood pressure: prospective
observational studies corrected for the regression dilution bias. Lancet.
1990;335:765774.
3. Collins R, Peto R, MacMahon S, Hebert P, Fiebach NH, Eberlein KA,
Godwin J, Qizilbash N, Taylor JO, Hennekens CH. Blood pressure,
stroke, and coronary heart disease, part 2: short-term reductions in blood
pressure: overview of randomised drug trials in their epidemiological
context. Lancet. 1990;335:827 838.
4. MacMahon S, Rodgers A. The effects of anti-hypertensive treatment on
vascular disease: reappraisal of the evidence in 1994. J Vasc Med Biol.
1993;4:265271.
5. MacMahon S, Rodgers A. The effects of blood pressure reduction in older
patients: an overview of five randomized controlled trials in elderly
hypertensives. Clin Exp Hypertens. 1993;15:967978.
6. Insua JT, Sacks HS, Lau TS, Lau J, Reitman D, Pagano D, Chalmers TC.
Drug treatment of hypertension in the elderly: a meta-analysis. Ann Intern
Med. 1994;121:355362.
7. Gueyffier F, Boutitie F, Boissel JP, Pocock S, Coope J, Cutler J, Ekbom
T, Fagard R, Friedman L, Kerlikowske K, Perry M, Prineas R, Schron E,
for the INDANA Investigators. Effect of antihypertensive drug treatment
on cardiovascular outcomes in women and men: a meta-analysis of
individual patient data from randomized, controlled trials. Ann Intern
Med. 1997;126:761767.
8. He J, Whelton PK. Elevated systolic blood pressure as a risk factor for
cardiovascular and renal disease. J Hypertens Suppl. 1999;17:S7S13.
9. Gueyffier F, Boutitie F, Boissel JP, Coope J, Cutler J, Ekbom T, Fagard
R, Friedman L, Perry HM, Pocock S. INDANA: a meta-analysis on
individual patient data in hypertension: protocol and preliminary results.
Therapie. 1995;50:353362.
10. Gueyffier F, Froment A, Gouton M. New meta-analysis of treatment trials
of hypertension: improving the estimate of therapeutic benefit. J Hum
Hypertens. 1996;10:1 8.
11. Neal B, MacMahon S, Chapman N, for the Blood Pressure Lowering
Treatment Trialists Collaboration. Effects of ACE inhibitors, calcium
antagonists, and other blood-pressure-lowering drugs: results of prospectively designed overviews of randomised trials. Lancet. 2000;356:
19551964.
12. Hedges LV, Olkin I. Statistical Methods for Meta-analysis. London, UK:
Academic Press; 1985.
13. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin
Trials. 1986;7:177188.
14. Light RJ, Pillemer DB. Summing Up: The Science of Reviewing Research.
Cambridge, Mass: Harvard University Press; 1984.
15. Prospective Studies Collaboration. Cholesterol, diastolic blood pressure,
and stroke: 13,000 strokes in 45,000 people in 45 prospective cohorts.
Lancet. 1995;346:16471653.
16. Prospective Studies Collaboration. Age-specific relevance of usual blood
pressure to vascular mortality: a meta-analysis of individual data for one
million adults in 61 prospective studies. Lancet. 2002;360:19031913.
17. Eastern Stroke and Coronary Heart Disease Collaborative Research
Group. Blood pressure, cholesterol, and stroke in eastern Asia. Lancet.
1998;352:18011807.
18. Asia Pacific Cohort Studies Collaboration. Blood pressure and cardiovascular disease in the Asia Pacific region. J Hypertens. 2003;21:
707716.
19. Svensson P, de Faire U, Sleight P, Yusuf S, Ostergren J. Comparative
effects of ramipril on ambulatory and office blood pressures: a HOPE
substudy. Hypertension. 2001;38:E28 E32.
20. Sterne JA, Egger M, Smith GD. Systematic reviews in health care:
investigating and dealing with publication and other biases in meta-analysis. BMJ. 2001;323:101105.
21. Hansson L, Zanchetti A, Carruthers SG, Dahlof B, Elmfeldt D, Julius S,
Menard J, Rahn KH, Wedel H, Westerling S, for the HOT Study Group.
Effects of intensive blood-pressure lowering and low-dose aspirin in

1032

22.

23.

24.

25.

26.
27.

28.

29.

30.

31.

32.

33.

34.

35.
36.

37.

38.
39.

40.
41.
42.

43.

Stroke

April 2004

patients with hypertension: principal results of the Hypertension Optimal


Treatment (HOT) randomised trial. Lancet. 1998;351:17551762.
Estacio RO, Jeffers BW, Hiatt WR, Biggerstaff SL, Gifford N, Schrier
RW. The effect of nisoldipine as compared with enalapril on cardiovascular outcomes in patients with non-insulin-dependent diabetes and
hypertension. N Engl J Med. 1998;338:645 652.
UK Prospective Diabetes Study Group. Tight blood pressure control and
risk of macrovascular and microvascular complications in type 2 diabetes:
UKPDS 38. BMJ. 1998;317:703713.
UK Prospective Diabetes Study Group. Efficacy of atenolol and captopril
in reducing risk of macrovascular and microvascular complications in
type 2 diabetes: UKPDS 39. BMJ. 1998;317:713720.
Feigin VL, Lawes CMM, Bennett DA, Anderson CS. Stroke epidemiology: a review of population-based studies of incidence, prevalence, and
case-fatality in the late 20th century. Lancet Neurol. 2003;2:4353.
Truelsen T, Bonita R, Jamrozik K. Surveillance of stroke: a global
perspective. Int J Epidemiol. 2001;30(suppl 1):S11S16.
PROGRESS Collaborative Group. Randomised trial of a
perindopril-based blood-pressure-lowering regimen among 6,105 individuals with previous stroke or transient ischaemic attack. Lancet. 2001;
358:10331041.
Gueyffier F, Bulpitt C, Boissel J-P, Schron E, Ekbom T, Fagard R,
Casiglia E, Kerlikowske K, Coope J, for the INDANA Group. Antihypertensive drugs in very old people: a subgroup meta-analysis of randomised controlled trials. Lancet. 1999;353:793796.
Bosch J, Yusuf S, Pogue J, Sleight P, Lonn E, Rangoonwala B, Davies R,
Ostergren J. Use of ramipril in preventing stroke: double blind randomised trial. BMJ. 2002;324:699 702.
ALLHAT Collaborative Research Group. Major cardiovascular events in
hypertensive patients randomized to doxazosin vs chlorthalidone: the
Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack
Trial (ALLHAT). JAMA. 2000;283:19671975.
Dahlof B, Devereux RB, Kjeldsen SE, Julius S, Beevers G, de Faire U,
Fyhrquist F, Ibsen H, Kristiansson K, Lederballe-Pedersen O, Lindholm
LH, Nieminen MS, Omvik P, Oparil S, Wedel H, LIFE Study Group.
Cardiovascular morbidity and mortality in the Losartan Intervention For
Endpoint reduction in hypertension study (LIFE): a randomised trial
against atenolol. Lancet. 2002;359:9951003.
Lindholm LH, Ibsen H, Dahlof B, Devereux RB, Beevers G, de Faire U,
Fyhrquist F, Julius S, Kjeldsen SE, Kristiansson K, Lederballe-Pederson
O, Nieminen MS, Omvik P, Oparil S, Wedel H, Aurup P, Edelman J,
Snappin S, LIFE Study Group. Cardiovascular morbidity and mortality in
patients with diabetes in the Losartan Intervention For Endpoint reduction
in hypertension study (LIFE): a randomised trial against atenolol. Lancet.
2002;359:1004 1010.
Stewart IM. Relation of reduction in pressure to first myocardial
infarction in patients receiving treatment for severe hypertension. Lancet.
1979;1:861 865.
Kannel WB, DAgostino RB, Silbershatz H. Blood pressure and cardiovascular morbidity and mortality rates in the elderly. Am Heart J. 1997;
134:758 763.
Cruickshank JM. J curve in antihypertensive therapy: does it exist? A
personal point of view. Cardiovasc Drugs Ther. 1994;8:757760.
DAgostino RB, Belanger AJ, Kannel WB, Cruickshank JM. Relation of
low diastolic blood pressure to coronary heart disease death in presence
of myocardial infarction: the Framingham Study. BMJ. 1991;303:
385389.
Farnett L, Mulrow CD, Linn WD, Lucey CR, Tuley MR. The J-curve
phenomenon and the treatment of hypertension: is there a point beyond
which pressure reduction is dangerous? JAMA. 1991;265:489 495.
Alderman MH. Blood pressure J-curve: is it cause or effect? Curr Opin
Nephrol Hypertens. 1996;5:209 213.
MacMahon S, Rodgers A, Neal B, Chalmers J. Blood pressure lowering
for the secondary prevention of myocardial infarction and stroke. Hypertension. 1997;29:537538.
Sleight P. Lowering of blood pressure and artery stiffness. Lancet. 1997;
349:362.
Sleight P. Lowering of blood pressure and artery stiffness. Lancet. 1997;
349:955956.
Neaton JD, Wentworth D, for the Multiple Risk Factor Intervention Trial
Research Group. Serum cholesterol, blood pressure, cigarette smoking,
and death from coronary heart disease: overall findings and differences by
age for 316,099 white men. Arch Intern Med. 1992;152:56 64.
Flack JM, Neaton J, Grimm R Jr, Shih J, Cutler J, Ensrud K, MacMahon
S, for the Multiple Risk Factor Intervention Trial Research Group. Blood

44.
45.

46.

47.

48.
49.

50.
51.
52.
53.
54.

55.

56.

57.

58.

59.

60.

61.
62.
63.
64.
65.
66.

67.

68.

pressure and mortality among men with prior myocardial infarction.


Circulation. 1995;92:24372445.
McMurray J, McInnes GT. The J-curve hypothesis. Lancet. 1992;339:
561562.
Pfeffer MA. Angiotensin-converting enzyme inhibition in congestive
heart failure: benefit and perspective. Am Heart J. 1993;126(pt
2):789 793.
Hansson L, Lindholm LH, Ekbom T, Dahlof B, Lanke J, Schersten B,
Wester PO, Hedner T, de Faire U. Randomised trial of old and new
antihypertensive drugs in elderly patients: cardiovascular mortality and
morbidity: the Swedish Trial in Old Patients with Hypertension-2 study.
Lancet. 1999;354:17511756.
Staessen J, Fagard R, Thijs L, Celis H, Arabidze G, Birkenhager W,
Bulpitt C, de Leeuw PD, Dollery C, Fletcher AE, Forette F, Leonetti G,
Nachev C, OBrien ET, Rosenfeld J, Rodicio JL, Tuomilehto J, Zanchetti
A. Randomised double-blind comparison of placebo and active treatment
for older patients with isolated systolic hypertension. Lancet. 1997;350:
757764.
PROGRESS Collaborative Group. The lowering of blood-pressure after
stroke. Lancet. 2001;358:1994 1995.
Murray CJL, Lopez AD. The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability From Diseases, Injuries,
and Risk Factors in 1990 and Projected to 2020. Cambridge, Mass:
Harvard University Press; 1996.
Report by the Management Committee. The Australian therapeutic trial in
mild hypertension. Lancet. 1980;1:12611267.
Control of moderately raised blood pressure: report of a co-operative
randomized controlled trial. BMJ. 1973;3:434 436.
Carter AB. Hypotensive therapy in stroke survivors. Lancet. 1970;1:
485 489.
Coope J, Warrender TS. Randomised trial of treatment of hypertension in
elderly patients in primary care. BMJ. 1986;293:11451151.
The Dutch TIA Trial Study Group. Trial of secondary prevention with
atenolol after transient ischemic attack or nondisabling ischemic stroke.
Stroke. 1993;24:543548.
Amery A, Birkenhager W, Brixko P, Bulpitt C, Clement D, Deruyttere M,
De Schaepdryver A, Dollery C, Fagard R, Forette F. Mortality and
morbidity results from the European Working Party on High Blood
Pressure in the Elderly trial. Lancet. 1985;1:1349 1354.
Hypertension Detection and Follow-up Program Cooperative Group.
Five-year findings of the Hypertension Detection and Follow-up
Program, I: reduction in mortality of persons with high blood pressure,
including mild hypertension. JAMA. 1979;242:25622571.
Hypertension Detection and Follow-up Program Cooperative Group.
Five-year findings of the Hypertension Detection and Follow-up
Program, II: mortality by race-sex and age. JAMA. 1979;242:25722577.
Hypertension Detection and Follow-up Program Cooperative Group.
Five-year findings of the Hypertension Detection and Follow-up
Program, III: reduction in stroke incidence among persons with high
blood pressure. JAMA. 1982;247:633 638.
The effect of treatment on mortality in mild hypertension: results of the
Hypertension Detection and Follow-up Program. N Engl J Med. 1982;
307:976 980.
Effect of stepped care treatment on the incidence of myocardial infarction
and angina pectoris: 5-year findings of the Hypertension Detection and
Follow-up Program. Hypertension. 1984;6(pt 2):I198 I206.
Hypertension-Stroke Cooperative Study Group. Effect of antihypertensive treatment on stroke recurrence. JAMA. 1974;229:409 418.
MRC Working Party. Medical Research Council trial of treatment of
hypertension in older adults: principal results. BMJ. 1992;304:405 412.
Medical Research Council Working Party. MRC trial of treatment of mild
hypertension: principal results. BMJ. 1985;291:97104.
Helgeland A. Treatment of mild hypertension: a five year controlled drug
trial: the Oslo study. Am J Med. 1980;69:725732.
Leren P, Helgeland A. Oslo Hypertension Study. Drugs. 1986;31(suppl
1):41 45.
SHEP Cooperative Research Group. Prevention of stroke by antihypertensive drug treatment in older persons with isolated systolic hypertension: final results of the Systolic Hypertension in the Elderly Program
(SHEP). JAMA. 1991;265:32553264.
Dahlof B, Lindholm LH, Hansson L, Schersten B, Ekbom T, Wester PO.
Morbidity and mortality in the Swedish Trial in Old Patients with Hypertension (STOP-Hypertension). Lancet. 1991;338:12811285.
Smith WM. Treatment of mild hypertension: results of a ten-year intervention trial. Circ Res. 1977;40(suppl 1):I98 I105.

Lawes et al
69. Effects of treatment on morbidity in hypertension: results in patients with
diastolic blood pressures averaging 115 through 129 mm Hg. JAMA.
1967;202:1028 1034.
70. Effects of treatment on morbidity in hypertension, II: results in patients
with diastolic blood pressure averaging 90 through 114 mm Hg. JAMA.
1970;213:11431152.
71. Perry HM Jr. Treatment of mild hypertension: preliminary results of a
two-year feasibility trial. Circ Res. 1977;40(suppl 1):I180 I187.
72. Veterans AdministrationNational Heart, Lung, and Blood Institute Study
Group for Evaluating Treatment in Mild Hypertension. Evaluation of
drug treatment in mild hypertension: VA-NHLBI feasibility trial: plan
and preliminary results of a two-year feasibility trial for a multicenter
intervention study to evaluate the benefits versus the disadvantages of
treating mild hypertension. Ann N Y Acad Sci. 1978;304:267292.
73. Wolff FW, Lindeman RD. Effects of treatment in hypertension: results of
a controlled study. J Chron Dis. 1966;19:227240.
74. Yusuf S, Sleight P, Pogue J, Bosch J, Davies R, Dagenais G, for the Heart
Outcomes Prevention Evaluation Study Investigators. Effects of an angiotensin-converting-enzyme inhibitor, ramipril, on cardiovascular events
in high-risk patients. N Engl J Med. 2000;342:145153.
75. MacMahon S, Sharpe N, Gamble G, Clague A, Mhurchu CN, Clark T,
Hart H, Scott J, White H, for the PART-2 Collaborative Research Group.
Randomized, placebo-controlled trial of the angiotensin-converting
enzyme inhibitor, ramipril, in patients with coronary or other occlusive
arterial disease: Prevention of Atherosclerosis with Ramipril. J Am Coll
Cardiol. 2000;36:438 443.
76. Cashin-Hemphill L, Holmvang G, Chan RC, Pitt B, Dinsmore RE, Lees
RS, for the QUIET Investigators. Angiotensin-converting enzyme inhibition as antiatherosclerotic therapy: no answer yet: QUinapril Ischemic
Event Trial. Am J Cardiol. 1999;83:43 47.
77. Teo KK, Burton JR, Buller CE, Plante S, Catellier D, Tymchak W, et al.
Long-term effects of cholesterol lowering and angiotensin-converting
enzyme inhibition on coronary atherosclerosis: the Simvastatin/Enalapril
Coronary Atherosclerosis Trial (SCAT). Circulation. 2000;102:
1748 1754.
78. Eriksson S, Olofsson BO, Wester PO. Atenolol in secondary prevention
after stroke. Cerebrovasc Dis. 1995;5:2125.
79. Pitt B, Byington RP, Furberg CD, Hunninghake DB, Mancini GB, Miller
ME, Riley W. Effect of amlodipine on the progression of atherosclerosis
and the occurrence of clinical events. Circulation. 2000;102:15031510.
80. The ALLHAT Officers and Coordinators for the ALLHAT Collaborative
Research Group. Major outcomes in high-risk hypertensive patients randomized to angiotensin-converting enzyme inhibitor or calcium channel
blocker vs diuretic: the Antihypertensive and Lipid-Lowering Treatment
to Prevent Heart Attack Trial (ALLHAT). JAMA. 2002;288:29812997.
81. Wing LM, Reid CM, Ryan P, Beilin LJ, Brown MA, Jennings GL,
Johnston CI, McNeil JJ, Macdonald GJ, Marley JE, Morgan TO, West
MJ; Second Australian National Blood Pressure Study. A comparison of
outcomes with angiotensin-converting enzyme inhibitors and diuretics
for hypertension in the elderly. N Engl J Med. 2003;348:583592.
82. Hansson L, Lindholm LH, Niskanen L, Lanke J, Hedner T, Niklason A,
Luomanmaki K, Dahlof B, de Faire U, Morlin C, Karlberg BE, Wester
PO, Bjork JE. Effect of angiotensin-converting-enzyme inhibition

Blood Pressure and Stroke: An Overview

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

93.

1033

compared with conventional therapy on cardiovascular morbidity and


mortality in hypertension: the Captopril Prevention Project (CAPPP)
randomised trial. Lancet. 1999;353:611 616.
Casiglia E, Spolaore P, Mazza A, Ginocchio G, Colangeli G, Onesto C,
Di Menza G, Pegoraro L, Ambrosio GB. Effect of two different therapeutic approaches on total and cardiovascular mortality in a Cardiovascular Study in the Elderly (CASTEL). Jpn Heart J. 1994;35:589 600.
Black HR, Elliott WJ, Grandits G, Grambsch P, Lucente T, White WB,
Neaton JD, Grimm RH Jr, Hansson L, Lacourciere Y, Muller J, Sleight P,
Weber MA, Williams G, Wittes J, Zanchetti A, Anders RJ, CONVINCE
Research Group. Principal results of the Controlled Onset Verapamil
Investigation of Cardiovascular End Points (CONVINCE) trial. JAMA.
2003;289:20732082.
Zanchetti A, Bond MG, Hennig M, Neiss A, Mancia G, Dal Palu C,
Hansson L, Magnani B, Rahn KH, Reid JL, Rodicio J, Safar M, Eckes L,
Rizzini P, European Lacidipine Study on Atherosclerosis Investigators.
Calcium antagonist lacidipine slows down progression of asymptomatic
carotid atherosclerosis: principal results of the European Lacidipine Study
on Atherosclerosis (ELSA), a randomized, double-blind, long-term trial.
Circulation. 2002;106:24222427.
Brown MJ, Palmer CR, Castaigne A, de Leeuw PW, Mancia G, Rosenthal
T, Ruilope LM. Morbidity and mortality in patients randomised to
double-blind treatment with a long-acting calcium-channel blocker or
diuretic in the international nifedipine GITS study: Intervention as a Goal
in Hypertension Treatment (INSIGHT). Lancet. 2000;356:366 372.
Borhani NO, Mercuri M, Borhani PA, Buckalew VM, Canossa-Terris M,
Carr AA, Kappagoda T, Rocco MV, Schnaper HW, Sowers JR, Bond
MG. Final outcome results of the Multicenter Isradipine Diuretic Atherosclerosis Study (MIDAS): a randomized controlled trial. JAMA. 1996;
276:785791.
National Intervention Cooperative Study in Elderly Hypertensives Study
Group. Randomized double-blind comparison of a calcium antagonist and
a diuretic in elderly hypertensives. Hypertension. 1999;34:1129 1133.
Hansson L, Hedner T, Lund-Johansen P, Kjeldsen SE, Lindholm LH,
Syvertsen JO, Lanke J, de Faire U, Dahlof B, Karlberg BE. Randomised
trial of effects of calcium antagonists compared with diuretics and betablockers on cardiovascular morbidity and mortality in hypertension: the
Nordic Diltiazem (NORDIL) study. Lancet. 2000;356:359 365.
Zanchetti A, Rosei EA, Dal Palu C, Leonetti G, Magnani B, Pessina A.
The Verapamil in Hypertension and Atherosclerosis Study (VHAS):
results of long-term randomized treatment with either verapamil or
chlorthalidone on carotid intima-media thickness. J Hypertens. 1998;16:
16671676.
Rosei EA, Dal Palu C, Leonetti G, Magnani B, Pessina A, Zanchetti A,
for the VHAS Investigators. Clinical results of the Verapamil in Hypertension and Atherosclerosis Study. J Hypertens. 1997;15:13371344.
Tatti P, Pahor M, Byington RP, Di Mauro P, Guarisco R, Strollo G,
Strollo F. Outcome results of the Fosinopril Versus Amlodipine Cardiovascular Events Randomized Trial (FACET) in patients with hypertension and NIDDM. Diabetes Care. 1998;21:597 603.
Easton DF, Peto J, Babiker AG. Floating absolute risk: an alternative to
relative risk in survival and case-control analysis avoiding an arbitrary
reference group. Stat Med. 1991;10:10251035.