The n e w e ng l a n d j o u r na l of m e dic i n e

special article

Shattuck Lecture

The Hypertension Paradox — More

Uncontrolled Disease despite Improved Therapy
Aram V. Chobanian, M.D.

T
From Boston University Medical Center, he treatment of hypertension has been one of medicine’s major
Boston. Address reprint requests to Dr. successes of the past half-century. The remarkable advances in therapy have
Chobanian at Boston University Medical
Center, 650 Albany St., Boston, MA provided the newfound capability for lowering blood pressure in almost every
02118, or at [email protected]. person with hypertension. Nevertheless, hypertension continues to be a major pub-
lic health problem whose prevalence is increasing worldwide.1 Moreover, the number
N Engl J Med 2009;361:878-87.
Copyright © 2009 Massachusetts Medical Society. of people with uncontrolled blood pressure is also increasing, despite the therapeu-
tic advances. Here, I discuss the factors responsible for this paradox and the strate-
gies required for addressing the growing problem.

E a r ly A pproache s t o Ther a py

Hypertension is a major risk factor for cardiovascular and renal diseases, and early
data indicate that untreated hypertension shortens life expectancy by approximately
5 years.2 Although data linking increased blood pressure to premature death were
long available from insurers,3 the prevailing medical opinion well into the 1950s
was that lowering of elevated blood pressure was detrimental because it would im-
pair perfusion of vital organs and thereby increase the risk of cardiovascular and
renal diseases.4 Three early pioneers who thought otherwise and aggressively pur-
sued blood-pressure lowering were Walter Kempner, Reginald Smithwick, and Rob-
ert Wilkins. Each had a different approach: Kempner used dietary manipulation;
Smithwick, surgery; and Wilkins, drug therapy.
Kempner prescribed a diet composed primarily of rice and fruits. It was low in
calories, fat, protein, and sodium (<30 mmol per day) and caused ketosis, weight
loss, and a decrease in blood pressure.5 The full diet was difficult to follow, but
many patients attended Kempner’s clinic at Duke Medical Center for the treatment
of hypertension or obesity. Some patients’ severe hypertension and renal dysfunc-
tion improved on Kempner’s program, but his work was not taken seriously by the
academic medical community for many years.
Smithwick, who chaired the Department of Surgery at Boston University, devel-
oped a surgical procedure for the treatment of hypertension that involved bilateral
lumbodorsal sympathectomy and splanchnicectomy with resection of the sympa-
thetic ganglia from the lower thoracic through the upper lumbar roots and removal
of much of the greater splanchnic nerve.6 Complications from the surgery, particu-
larly symptomatic orthostatic hypotension, were relatively common, but as with
Kempner’s diet, some patients with severe hypertension benefited and survived
until effective drug therapies became available.
Wilkins, one of my mentors, became chief of cardiology at Boston University in
1945. His interest in hypertension was enhanced by the availability of a large
number of Smithwick’s patients at the institution. Wilkins was asked by James
Shannon, then head of the Squibb Institute and later director of the National In-

878 n engl j med 361;9  nejm.org  august 27, 2009

Downloaded from www.nejm.org on August 26, 2009 . For personal use only. No other uses without permission.
Copyright © 2009 Massachusetts Medical Society. All rights reserved.
 shattuck lecture

stitutes of Health, to work with Squibb on devel- tor blockers (ARBs) — now represent the primary
oping antihypertensive drugs. Around the same treatment options.12 In addition, several clinical
time, Edward Freis joined Wilkins as a research trials were conducted that showed clear-cut ben-
fellow and was assigned to study the hemo­ efits of therapy, beginning with the treatment of
dynamic effects of new antihypertensive drugs. malignant hypertension.13 Subsequently, the land-
One such drug was pentaquine, an antimalarial mark Veterans Administration studies showed
agent that was observed to cause orthostatic impressive reductions in cardiovascular events
hypotension. Although it had to be abandoned among patients with a pretreatment diastolic
because of a high incidence of side effects, penta­ blood pressure of 115 to 129 mm Hg14 and later
quine appeared to cause reversal of malignant among those with a diastolic pressure of 90 to
hypertension in a small number of patients.7 The 114 mm Hg.15 The benefits were so impressive in
development of antihypertensive drugs intensi- the former study that a highly significant effect
fied in the late 1940s and early 1950s, and sev- (in comparison with placebo) was observed with
eral medications, including Rauwolfia serpentina, an intervention group of only 73 patients who
veratrum alkaloids, ganglionic blocking drugs, were treated for 18 months.
and hydralazine, were tested in the laboratories Subsequent placebo-controlled trials showed
of Wilkins and others.8,9 Freis was recruited to the importance of blood-pressure lowering in
Georgetown and the Washington Veterans Ad- elderly patients with isolated systolic hyperten-
ministration Hospital and continued his work sion with therapies based on the use of either
there, competing with his former mentor. diuretics or calcium-channel blockers.16,17 Most
Wilkins became increasingly convinced of the
benefits of antihypertensive therapy and wrote in
Table 1. Advances in the Treatment of Hypertension.
1952, “No case of hypertension associated with
good renal function should be considered impos- Decade and Therapy
sible to treat until proven otherwise.” 8 He and 1940s
his associates began combining drugs with differ- Potassium thiocyanate
ent modes of action when the initial medication Kempner diet
did not adequately lower blood pressure. This Lumbodorsal sympathectomy
“step-care” approach has continued to be a main- 1950s
stay of therapy. The number of patients with Rauwolfia serpentina
hypertension who received treatment expanded Ganglionic blockers
rapidly, particularly when the thiazide diuretics Veratrum alkaloids
became available in the late 1950s. The first use Hydralazine
of chlorothiazide for hypertension was reported Guanethidine
almost simultaneously by the research groups of Thiazide diuretics
Wilkins and Freis. Both rushed to publish, Wilkins 1960s
and Hollander in the Boston Medical Quarterly10 and α2-Adrenergic–receptor agonists
Freis and Wilson in the Medical Annals of the Dis- Spironolactone
trict of Columbia.11 The competition between them β-Adrenergic–receptor agonists
was intense, but the work of each was appropri- 1970s
ately recognized with the Lasker Award, which α1-Adrenergic–receptor antagonists
Wilkins received in 1958 and Freis in 1971. Angiotensin-converting–enzyme inhibitors
1980s
Calcium antagonists
Benefi t s of A n t ih y per tensi v e- 1990s
Drug Ther a py
Angiotensin-receptor blockers
In the 50 years since the introduction of the thi- Endothelin-receptor antagonists*
azide diuretics, many classes of antihyperten- 2000s
sive drugs have been approved for use (Table 1). Renin inhibitors
Five of these — diuretics, beta-receptor blockers,
* This class of drugs has not been approved for clinical use
angiotensin-converting–enzyme (ACE) inhibitors, in patients with hypertension.
calcium-channel blockers, and angiotensin-recep-

n engl j med 361;9  nejm.org  august 27, 2009 879

Downloaded from www.nejm.org on August 26, 2009 . For personal use only. No other uses without permission.
Copyright © 2009 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

recently, the studies were expanded to include tions, including salt restriction.20,21 The groups
patients over the age of 80 years, among whom who were assigned to the sodium-reduction inter-
treatment with a diuretic and an ACE inhibitor vention had significantly fewer cardiovascular
was associated with a substantial reduction in events during long-term follow-up than the usual-
mortality and morbidity from cardiovascular dis­ care group.22 The urinary sodium-to-potassium
eases.18 ratio correlated more strongly with the cardio-
Such reductions that have been achieved with vascular-event rate than urinary sodium alone,
antihypertensive therapy have been truly impres- suggesting that a higher potassium intake had a
sive. In placebo-controlled trials, the incidence of beneficial effect.23
stroke has been reduced by an average of 35 to
40%, the incidence of coronary events by 20 to Drugs
25%, and the incidence of congestive heart fail- In most patients with hypertension, drug therapy
ure by more than 50%. Malignant hypertension is required to achieve target blood-pressure levels.
has become a rare entity, and acute hypertensive Several excellent agents are available.
heart failure and hemorrhagic stroke are now Thiazide-type diuretics, beta-blockers, ACE
uncommon. inhibitors, calcium-channel blockers, and ARBs
are considered to be the most useful classes of
drugs, since they have been shown in clinical
E volv ing A pproache s
t o T r e atmen t trials to reduce cardiovascular complications in
patients with hypertension.12,24 All these classes
The management of hypertension continues to have broadly similar average effects on blood
evolve as newer antihypertensive medications and pressure,25 although there are differences among
clinical trial data become available. On the basis patients. In the majority of patients, two or more
of the current information, I would recommend antihypertensive drugs are required to achieve
the following approaches. target blood-pressure levels. As a result, several
two-drug fixed-dose combinations have been in­
Lifestyle Modifications troduced,26 and recently even a three-drug prep-
The adoption of certain lifestyle modifications aration that combines a calcium-channel block-
has been shown to be effective in lowering blood er, an ARB, and a thiazide diuretic has become
pressure and should be recommended for all pa- available.
tients with hypertension. These modifications in- To determine whether a given drug or drug
clude weight control, exercise, dietary sodium combination is superior to any other, several com-
restriction and potassium enhancement, modera- parison trials have been performed. In general,
tion of alcohol intake, and adoption of the Dietary these studies have shown minimal differences in
Approaches to Stop Hypertension (DASH) eating primary outcomes among the drug classes, as
plan, which emphasizes a high intake of fruits, long as equivalent reduction in blood pressure
vegetables, complex carbohydrates, and low-fat has been achieved (Table 2).27-33 Even blood-
dairy products and restriction of saturated fats.19 pressure differences as small as 3/2 mm Hg be-
The reductions in systolic blood pressure that are tween treatment groups have been associated with
achieved with these approaches have averaged 5 to significant differences in certain outcomes.30,34
10 mm Hg for a weight decrease of 10 kg, 8 to A few trials have shown superiority of one drug or
14 mm Hg for the DASH eating plan, 2 to 8 mm Hg a given combination over another. In the Losartan
for dietary sodium reduction, 4 to 9 mm Hg for Intervention for Endpoint Reduction in Hyperten-
increased physical activity, and 2 to 4 mm Hg sion (LIFE) study (NCT00338260), losartan-based
for moderation of alcohol consumption. therapy was associated with fewer cardiovascular
No clinical trial has examined directly the ef- events than atenolol-based treatment.31 In the
fects of lifestyle interventions on cardiovascular Second Australian National Blood Pressure Study
outcomes. However, indirect evidence supporting (ANBP2), ACE inhibition was associated with a
a favorable effect has been reported from 10-to-15- somewhat lower incidence of cardiovascular com-
year follow-up of patients in the Trials of Hyper- plications than thiazide-based therapy in men but
tension Prevention I and II (TOHP I and TOHP not in women.32 In the Avoiding Cardiovascular
II; ClinicalTrials.gov number, NCT00000528), Events through Combination Therapy in Patients
which studied the effects of lifestyle modifica- Living with Systolic Hypertension (ACCOMPLISH)

880 n engl j med 361;9  nejm.org  august 27, 2009

Downloaded from www.nejm.org on August 26, 2009 . For personal use only. No other uses without permission.
Copyright © 2009 Massachusetts Medical Society. All rights reserved.
 shattuck lecture

Table 2. Comparative Drug Trials in Patients with Hypertension.*

Trial Name Drug Comparison Primary Outcome

STOP-227 Thiazide-type diuretic plus beta-blocker vs. No significant difference
ACE inhibitor plus calcium-channel blocker
ALLHAT28 Thiazide-type diuretic vs. ACE inhibitor vs. No significant difference
­calcium-channel blocker
INVEST29 Thiazide-type diuretic plus beta-blocker vs. No significant difference
­calcium-channel blocker plus ACE inhibitor
ASCOT30 Thiazide-type diuretic plus beta-blocker vs. No significant difference
­calcium-channel blocker plus ACE inhibitor
LIFE31 Angiotensin-receptor blocker vs. beta-blocker Angiotensin-receptor blocker superior
32
ANBP2 Thiazide-type diuretic vs. ACE inhibitor ACE inhibitor superior in men only
ACCOMPLISH33 ACE inhibitor plus thiazide-type diuretic vs. ACE inhibitor plus calcium-channel blocker
ACE inhibitor plus calcium-channel blocker superior

* ACCOMPLISH denotes Avoiding Cardiovascular Events through Combination Therapy in Patients Living with Systolic
Hypertension, ALLHAT Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial, ANBP2 Second
Australian National Blood Pressure Study, ASCOT Anglo-Scandinavian Cardiac Outcomes Trial, INVEST International
Verapamil-Trandolapril Study, LIFE Losartan Intervention for Endpoint Reduction in Hypertension, and STOP-2
Swedish Trial in Old Patients with Hypertension 2.

study (NCT00170950), the benazepril–amlodipine complications to a degree similar to that associ-

combination was superior to a benazepril–hydro- ated with diuretics. In addition, since generic
chlorothiazide-based regimen.33 However, the preparations for each of these drug classes are or
bulk of evidence indicates that by far the most will soon become available, the cost advantage of
critical aspect of therapy is the lowering of blood diuretics has become much less of an issue.
pressure, regardless of how this is achieved. The new treatment algorithm that I would
There are situations, however, in which the recommend is illustrated in Figure 1. The ap-
data demonstrate compelling indications for the proach varies slightly, depending on the severity
use of certain classes of antihypertensive drugs. of hypertension. In stage 1 hypertension (blood
These include the use of ACE inhibitors and pressure, 140–159/90–99 mm Hg), it is appropri-
ARBs in patients with chronic renal disease, ate to begin with lifestyle modifications before
diabetes, congestive heart failure, or recent myo- drug therapy is initiated. Thiazide-type diuretics,
cardial infarction and beta-blockers in those with ACE inhibitors, calcium-channel blockers, or
angina pectoris, recent myocardial infarction, ARBs can be considered initially, depending on
arrhythmias, or heart failure.12 Selection can also the physician’s experience, the patient’s accep-
be based on coexisting conditions for which a tance, and the presence of compelling or coexist-
given drug may provide added benefit (e.g., ing conditions. Because of recent data indicating
calcium-channel blockers or beta-blockers for that traditional beta-blockers, such as atenolol
patients with both hypertension and migraine and metoprolol, are not as effective in reducing
headache). the risk of stroke as the other four classes,35 their
use as first-line agents, particularly in the elderly,
New Treatment Algorithm should be restricted to patients with the compel-
Although thiazide-type diuretics were recom- ling indications described above. It is uncertain
mended in the Joint National Committee Guide- whether the use of newer beta-blockers with
lines of 2003 as the preferred initial drug thera- vasodilator properties, such as carvedilol, should
py for most patients with hypertension, subsequent be similarly restricted.
data from the LIFE study, the Anglo-Scandina- In stage 2 hypertension (blood pressure,
vian Cardiac Outcomes Trial (ASCOT),30 ANBP2, >160/100 mm Hg), drug treatment should be
and ACCOMPLISH point to the need for a more initiated promptly, along with lifestyle approach-
flexible approach. Taken together, the studies es. Two-drug combinations may be used as ini-
show that several drug classes with reasonable tial therapy in some patients. Evaluation for
side-effect profiles can reduce cardiovascular secondary hypertension should be considered

n engl j med 361;9  nejm.org  august 27, 2009 881

Downloaded from www.nejm.org on August 26, 2009 . For personal use only. No other uses without permission.
Copyright © 2009 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

Stage 1 Hypertension Stage 2 Hypertension

(blood pressure, 140–159/90–99 mm Hg) (blood pressure, ≥160/100 mm Hg)

Step 1 Lifestyle modification Two-drug regimen for most

Alcohol restriction patients plus lifestyle
DASH diet modification
Exercise ACE inhibitor
Salt reduction ARB
Weight control Beta-blocker
Calcium-channel blocker
Diuretic

Step 2 Single drug treatment Add third drug of different class

ACE inhibitor Assess adherence
ARB Optimize doses
Calcium-channel blocker
Diuretic
Control other
cardiovascular disease
risk factors

Step 3 Add second drug of different class Add fourth drug of different class
ACE inhibitor Assess alcoholic excess
ARB Assess salt retention
Beta-blocker
Calcium-channel blocker
Diuretic

Step 4 Add third drug of different class Evaluate for secondary hyper-
Assess adherence tension
Optimize doses

Compelling Indications Drug Treatment

Post–myocardial infarction ACE inhibitor, beta-blocker
Angina pectoris Beta-blocker, calcium-channel blocker
Heart failure ACE inhibitor, ARB, beta-blocker, diuretic,
aldosterone antagonist
Chronic renal disease ACE inhibitor, ARB
Diabetes ACE inhibitor, ARB, and others
High coronary heart disease ACE inhibitor, ARB, beta-blocker, calcium-
risk channel blocker, diuretic

Figure 1. Algorithm for Management of Hypertension.

ACE denotes angiotensin-converting enzyme, ARB angiotensin-receptor blocker, and DASH Dietary Approaches to
Stop Hypertension. ICM
AUTHOR: Chobanian RETAKE 1st
FIGURE: 1 of 3 2nd
REG F
3rd
CASE Revised

when three or more antihypertensive

EMail drugs
ARTIST: ts of Line
H/T
4-C
Although
H/T
several
SIZE excellent antihypertensive
33p9
different classes do not control blood pressure. Combo
Enon
drugs are available, the search for new agents
It is likely that in the future, pharmacogenetic continues.
AUTHOR, PLEASE NOTE: Several interesting new therapies for
data will help guide drug choice,Figure buthas been redrawn and type has been reset.
little such hypertension
Please check carefully. are under development, including
information is currently available. There is also the endothelin receptor type A antagonist daru­
a growing interest in instituting antihypertensive- sentan,37ISSUE:
JOB: 36109 which may be approved soon for the
08-27-09
drug therapy on the basis of total cardiovascular treatment of resistant hypertension. New treat-
risk rather than absolute blood-pressure levels,36 ments in early stages of clinical testing include
but direct evidence justifying such an approach a vaccine to block the activity of angiotensin II,38
is lacking. cannabinoid-1 receptor antagonists,39 and alage-

882 n engl j med 361;9  nejm.org  august 27, 2009

Downloaded from www.nejm.org on August 26, 2009 . For personal use only. No other uses without permission.
Copyright © 2009 Massachusetts Medical Society. All rights reserved.
 shattuck lecture

brium, which interferes with cross-linkages of

collagen and elastin and thereby reduces arterial Awareness Treatment Control
80
and myocardial stiffness.40 73 72
70 68
61
H y per tension C on t rol 60

Percent of Patients
55 54
51
50
The control of hypertension continues to be in-
40 35
adequate despite the excellent array of effective, 31 29
30 27
well-tolerated medications. Recent data indicate
that approximately 28% of Americans with hyper- 20
10
tension are unaware of their hypertension, 39% are 10
not receiving therapy, and 65% do not have their 0
NHANES II NHANES III NHANES III NHANES
blood pressure controlled to levels below 140/90 1976–1980 (phase 1) (phase 2) 1999–2004
mm Hg (Fig. 2).41 The control rates are even worse 1988–1991 1991–1994
among patients with chronic kidney disease, dia-
Figure 2. Rates of Awareness, Treatment, and Control of High Blood Pressure
betes, stable angina, the acute coronary syn- AUTHOR: Chobanian RETAKE 1st
in the United States
ICM (1976–2004).
drome, or left ventricular dysfunction, in whom FIGURE: 2 of 3 2nd
High blood pressure is defined as a reading of 140/90 mm Hg3rd
REG F or more for
target blood-pressure levels of 130/80 mm Hg or persons between
CASEthe ages of 18 and 74 years. Despite Revisedmajor improvements
lower are recommended.12,42 In large part, the low in blood-pressure Line some
EMail therapies in recent years, 4-C 28% of SIZE
Americans with
ARTIST: ts H/T H/T
control rates can be attributed to poor manage- hypertension do
Enonnot know they have the condition, 39% are
Combo
22p3 receiving no
ment of elevated systolic blood pressure. therapy, and 65% have insufficient blood-pressure control. Data are from
AUTHOR, PLEASE NOTE:
Chobanian et al.12Figure
and Cutler al.41 NHANES
et redrawn denotes National Health
Race, ethnic background, and income status has been and type has been reset.
and Nutrition Examination Survey.
Please check carefully.
affect control rates of hypertension and other car-
diovascular risk factors. The 1999–2000 Nation­ JOB: 36109 ISSUE: 08-27-09
al Health and Nutrition Examination Survey through 1994 to 65 million in the period from
(NHANES) showed a reduction in blood pressure 1999 through 2004.41 The prevalence of hyper-
to below 140/90 mm Hg with treatment in 33% tension worldwide is projected to increase from
of white patients but only in 28% of black pa- approximately 1.0 billion in 2000 to 1.5 billion by
tients and 18% of Hispanic patients.43 During 25 2025.2 The total number of persons with uncon-
years of follow-up to the Multiple Risk Factor trolled hypertension in the United States has in-
Intervention Trial (MRFIT) (NCT00000487), the creased from 37 million to 42 million, despite
rate of death from cardiovascular disease among improvements in treatment and in control rates
black patients was 25% higher than that among during the past two decades (Fig. 3).
white patients. This difference could be explained What can be done to reverse this trend? To
by differences in blood-pressure levels, the pres- reflect the high risk of hypertension in persons
ence of diabetes, smoking prevalence, and income with blood pressures of 120–130/80–89 mm Hg,
status.44 The reasons behind these health dispari- the seventh report of the Joint National Commit-
ties are complex and involve such factors as the tee on Prevention, Detection, Evaluation, and
availability of health insurance, access to high- Treatment of High Blood Pressure reclassified
quality health care, and cultural and attitudinal blood pressures in this range as prehyperten-
differences between patients and their physicians. sion.12 The rate of progression from prehyperten-
The societal costs from such disparities are high, sion to hypertension can be relatively rapid, and
and there is an urgent need to deal with this hypertension ultimately develops in most persons
problem on a broad basis. if they live long enough. In the Framingham
Heart Study population, approximately 90% of
persons who had normal blood pressure at 55 or
Incr e a se in the Pr e va l ence
of H y per tension 65 years of age became hypertensive in the sub-
sequent 20 years.45 In the international Ather­
The prevalence of hypertension continues to in- osclerosis Risk in Communities (ARIC) study
crease worldwide. NHANES data indicate that (NCT00005131), which followed more than 15,000
prevalence has increased among U.S. adults, from patients between 45 and 64 years of age for
approximately 50 million in the period from 1988 9 years, the average 5-year age-adjusted in-

n engl j med 361;9  nejm.org  august 27, 2009 883

Downloaded from www.nejm.org on August 26, 2009 . For personal use only. No other uses without permission.
Copyright © 2009 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

intake in the United States has increased dur-

70 65 ing the past 25 years and is currently approxi-
60
Uncontrolled mately 150 to 170 mmol per day (3.5 to 4.0 g of
Controlled
50 sodium).53,54 The increase in dietary salt may
Millions of Patients
50 also have contributed to the growing obesity
42
40 problem during this period by causing increased
37
intake of fluids, particularly of high-calorie soft
30
drinks.55
20 The major factors involved in the sensitivity
10 23 of blood pressure to salt are poorly understood.
13 Several single-gene mutations that directly affect
0
1988–1994 1999–2004 renal sodium reabsorption can cause hyperten-
sion, but such variants have been observed in only
Figure 3. Changes in the Prevalence and Control of Hyper- a few persons.56 Blacks, the elderly, and persons
tension
ICM in the United States (1988–2004).
AUTHOR: Chobanian RETAKE 1st
with chronic renal disease have increased salt
FIGURE:of3 of 3 2nd
TheREGtotal
F number persons with uncontrolled hyper- 3rd sensitivity.57 Dietary salt reduction lowers blood
tension
CASE has increased from 37 million to 42 million
Revised dur-
ing EMail
the past two decades, even 4-Cthe rateSIZE
Line though of con-
pressure in persons with both normotension and
trol Enon
ARTIST: ts
has increased from 27%H/T H/T
to 35% during the 16p6
same hypertension and in children as well as adults.58
Combo
period. Data are from Chobanian et al.12 and Cutler et al.41 Extrapolation of data from the Intersalt study
AUTHOR, PLEASE NOTE:
Figure has been redrawn and type has been reset. (NCT00005763) suggests that a salt intake of 50
Please check carefully. mmol below the current mean level (a decrease
crease in systolic blood pressure ranged from 4 to of about one third) would reduce blood pressure
JOB: 36109 ISSUE: 08-27-09
7 mm Hg.46 by an average of 4.0/2.5 mm Hg among persons
Persons with prehypertension — more than with hypertension and by 2.0/1.0 mm Hg among
30% of the U.S. adult population — are at persons with normotension.59
higher-than-average risk for cardiovascular dis- The ratio of dietary sodium to potassium cor-
ease.47 From the prehypertensive range upward, relates better with blood pressure than the level
the effect of blood pressure on cardiovascular risk of either cation alone.60 Diets that are composed
is progressive and continuous, with an average primarily of processed foods are not only high
doubling of risk for every increase of 20/10 mm Hg in sodium but also low in potassium. U.S. di-
in pressure.48 Patients with prehypertension also etary potassium intake averages between 50 and
have a higher prevalence of other cardiovascular 60 mmol per day.60 An Institute of Medicine
risk factors, such as dyslipidemias, diabetes, insu- panel has recommended a sodium intake for
lin resistance, obesity, and the metabolic syn­ adults of 50 to 65 mmol per day and a potassium
drome,47 and have earlier evidence of target or- intake of 120 mmol per day,61 but we are far
gan damage than normotensive persons.49 Thus, from reaching such goals.
patients with prehypertension should be targeted Some countries, such as Finland and Great
for lifestyle interventions that reduce blood pres- Britain, have achieved significant reductions in
sure or delay the onset of hypertension and that dietary sodium through aggressive efforts. In
help to control other cardiovascular risk factors. Finland, average salt intake has decreased by one
Several factors can contribute to the develop- third during the past 30 years, and there has
ment of hypertension (Table 3).26 Recent genome­ been an associated population-wide decrease in
wide analyses have revealed several single-nucle- blood pressure.62 Finland’s intensive program has
otide polymorphisms (SNPs) of genetic loci that included broad educational efforts and coop-
are associated with blood pressure.50,51 Whether eration by the food industry in developing low-
any of these SNPs will prove to be targets for the sodium products and attaching warning labels
prevention or treatment of hypertension remains to high-sodium foods. Once salt intake is reduced
to be determined. Salt intake and body weight for a few weeks, most people appear to readjust
are particularly important in the age-related in- their taste threshold to become more sensitive to
crease in blood pressure. Such increases are un- salt,63 which facilitates long-term reductions in
common in societies in which sodium chloride intake.
intake does not exceed 50 mmol per day.52 Salt The increase in body weight in the population

884 n engl j med 361;9  nejm.org  august 27, 2009

Downloaded from www.nejm.org on August 26, 2009 . For personal use only. No other uses without permission.
Copyright © 2009 Massachusetts Medical Society. All rights reserved.
 shattuck lecture

The success in reducing tobacco use was facili-

Table 3. Risk Factors for Hypertension.*
tated greatly by the passage of legislation against
Genetic predisposition or family history smoking in public areas and the workplace, ex-
Black race pansion of educational efforts and public service
Diagnosis of prehypertension announcements, increased taxation of cigarettes,
Increasing age
elimination of cigarette vending machines in cer-
tain areas, limitation of cigarette sales to minors,
Obesity
and risk labeling of cigarette packs.
High sodium–low potassium intake To combat obesity, support from families,
Excessive alcohol intake schools, community and religious organizations,
Low socioeconomic status government, insurers, food and beverage indus-
Sleep apnea tries, health care providers, and the general
Use of certain illegal drugs or over-the-counter medications public will be essential. Population-wide strate-
gies — such as redesigning of roads and walk-
* Data are from Chobanian et al.26 ways to promote cycling and walking and the ex-
pansion of school health education and physical
education programs — should be combined with
is a critical factor in the increase in the preva- individually targeted interventions to alter dietary
lence of hypertension. Weight loss in overweight intake and increase physical activity. Recent prog-
or obese persons can prevent or delay the onset ress provides some cause for optimism: public
of hypertension.64 Unfortunately, the prevalence awareness of the major contributors to child-
of obesity continues to rise.65 hood obesity and the health risks involved has
Societal changes during the past 30 years have increased,68 and support has grown for making
had a major negative effect on dietary habits and a number of essential changes, some of which
preferences. The rapid growth of the fast-food in­ are beginning to be mandated in several states.
dustry and in the intake of commercially prepared It is paradoxical that despite the enormous
foods has meant an increased consumption of cal­ advances in antihypertensive-drug therapy, the
ories, saturated fat, and salt and a reduced intake number of people with uncontrolled hyperten-
of fruits, vegetables, and complex carbohydrates.55 sion has continued to rise. The failure to adopt
Daily caloric intake in adults has increased an healthy lifestyles has been a critical factor in this
average of 300 kcal during this period as portion increase and must be addressed urgently. To make
sizes have grown and marketing of high-calorie, the necessary changes on a broad basis will be
less-nutritious foods has increased.55,66,67 difficult, but the benefits will be well worth the
Deeply ingrained lifestyle habits cannot be effort.
changed without a national strategy that creates No potential conflict of interest relevant to this article was
broad acceptance of the need for such change. reported.

References
1. Kearney PM, Whelton M, Reynolds K, 6. Smithwick RH. Surgical treatment of nation with antihypertensive agents: pre-
Muntner P, Whelton PK, He J. Global bur- hypertension. Am J Med 1948:4:744-59. liminary report. Med Ann Dist Columbia
den of hypertension: analysis of world- 7. Freis ED, Wilkins RW. The effects of 1957;26:468-71.
wide data. Lancet 2005;365:217-23. pentaquine in patients with hypertension. 12. Chobanian AV, Bakris GL, Black HR,
2. Franco OH, Peeters A, Bonneux L, de Proc Soc Exp Biol Med 1947;64:455-8. et al. The Seventh Report of the Joint Na-
Laet C. Blood pressure in adulthood and 8. Wilkins RW. New drug therapies in tional Committee on Prevention, Detec-
life expectancy with cardiovascular disease arterial hypertension. Ann Intern Med tion, Evaluation, and Treatment of High
in men and women. Hypertension 2005; 1952;37:1144-55. Blood Pressure: the JNC 7 report. JAMA
46:280-6. 9. Wilkins RW, Judson WE. The use of 2003;289:2560-72. [Erratum, JAMA 2003;
3. Hunter ARO. Mortality study of im- Rauwolfia serpentina in hypertensive pa- 290:197.]
paired lives. Actuarial Soc 1923;24:453-6. tients. N Engl J Med 1953;248:48-53. 13. Dustan HP, Schneckloth RE, Cor­
4. Perera GA. Hypertensive vascular 10. Hollander W, Wilkins RW. Chlorothi- coran AC, Page IH. The effectiveness of
disease: description and natural history. azide: a new type of drug for the treat- long-term treatment of malignant hyper-
J Chronic Dis 1955;1:33-42. ment of arterial hypertension. BMQ 1957; tension. Circulation 1958;18:644-51.
5. Kempner W. Treatment of hyperten- 8:69-75. 14. Veterans Administration Cooperative
sive vascular disease with rice diet. Am J 11. Freis ED, Wilson IM. Potentiating ef- Study Group on Antihypertensive Agents.
Med 1948;4:545-77. fect of chlorothiazide (Diuril) in combi- Effects of treatment on morbidity in hy-

n engl j med 361;9  nejm.org  august 27, 2009 885

Downloaded from www.nejm.org on August 26, 2009 . For personal use only. No other uses without permission.
Copyright © 2009 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

pertension: results in patients with dia- Evaluation and Treatment of High Blood from a randomized, double-blind placebo-
stolic blood pressures averaging 115 Pressure. Hypertension 2003;42:1206-52. controlled dose-ranging study. J Clin Hy-
through 129 mm Hg. JAMA 1967;202: 27. Hansson L, Lindholm LH, Ekbom T, pertens (Greenwich) 2007;9:760-9.
1028-34. et al. Randomised trial of old and new 38. Tissot AC, Maurer P, Nussberger J, et
15. Idem. Effects of treatment on morbid- antihypertensive drugs in elderly patients: al. Effect of immunisation against angio-
ity in hypertension. II. Results in patients cardiovascular mortality and morbidity in tensin II with CYT006-AngQb on ambula-
with diastolic blood pressure averaging the Swedish Trial in Old Patients with tory blood pressure: a double-blind, ran-
90 through 114 mm Hg. JAMA 1970;213: Hypertension–2 study. Lancet 1999;354: domised, placebo-controlled phase IIa
1143-52. 1751-6. study. Lancet 2008;371:821-7.
16. Prevention of stroke by antihyperten- 28. Major outcomes in high-risk hyperten- 39. Bátkai S, Pacher P, Osei-Hyiaman D,
sive drug treatment in older persons with sive patients randomized to angiotensin- et al. Endocannabinoids acting at cannab-
isolated systolic hypertension: final re- converting enzyme inhibitor or calcium inoid-1 receptors regulate cardiovascular
sults of the Systolic Hypertension in the channel blocker vs diuretic: the Antihy- function in hypertension. Circulation 2004;
Elderly Program (SHEP). JAMA 1991;265: pertensive and Lipid-Lowering Treatment 110:1996-2002.
3255-64. to Prevent Heart Attack Trial (ALLHAT). 40. Bakris GL, Bank A, Kass DA, Neutel
17. Staessen JA, Fagard R, Thijs L, et al. JAMA 2002;288:2981-97. [Errata, JAMA JM, Preston RA, Oparil S. Advanced gly-
Randomised double-blind comparison of 2003;289:178, 2004;291:2196.] cation end-product cross-link breakers:
placebo and active treatment for older pa- 29. Pepine CJ, Handberg EM, Cooper- a novel approach to cardiovascular pathol-
tients with isolated systolic hypertension. DeHoff RM, et al. A calcium antagonist ogies related to the aging process. Am J
Lancet 1997;350:757-64. vs a non-calcium antagonist hypertension Hypertens 2004;17:23S-30S.
18. Beckett NS, Peters R, Fletcher AE, et al. treatment strategy for patients with coro- 41. Cutler JA, Sorlie PD, Wolz M, Thom T,
Treatment of hypertension in patients 80 nary artery disease: the International Vera- Fields LE, Roccella EJ. Trends in hyper-
years of age or older. N Engl J Med 2008; pamil-Trandolapril Study (INVEST): a ran- tension prevalence, awareness, treatment,
358:1887-98. domized controlled trial. JAMA 2003;290: and control rates in United States adults
19. Appel LJ, Moore TJ, Obarzanek E, et al. 2805-16. between 1988-1994 and 1999-2004. Hyper-
A clinical trial of the effects of dietary 30. Dahlöf B, Sever PS, Poulter NR, et al. tension 2008;52:818-27.
patterns on blood pressure. N Engl J Med Prevention of cardiovascular events with 42. Rosendorff C, Black HR, Cannon CP,
1997;336:1117-24. an antihypertensive regimen of amlodi­ et al. Treatment of hypertension in the
20. The effects of nonpharmacologic in- pine adding perindopril as required ver- prevention and management of ischemic
terventions on blood pressure of persons sus atenolol adding bendroflumethiazide heart disease: a scientific statement from
with high normal levels: results of the as required, in the Anglo-Scandinavian the American Heart Association Council
Trials of Hypertension Prevention, phase I. Cardiac Outcomes Trial-Blood Pressure for High Blood Pressure Research and the
JAMA 1992;267:1213-20. [Erratum, JAMA Lowering Arm (ASCOT-BPLA): a multi- Councils on Clinical Cardiology and Epi-
1992;267:2320.] centre randomised controlled trial. Lan- demiology and Prevention. Circulation
21. Effects of weight loss and sodium re- cet 2005;366:895-906. 2007;115:2761-88. [Erratum, Circulation
duction intervention on blood pressure 31. Dahlöf B, Devereux RB, Kjeldsen SE, 2007;116(5):e121.]
and hypertension incidence in overweight et al. Cardiovascular morbidity and mor- 43. Hajjar I, Kotchen TA. Trends in preva-
people with high-normal blood pressure: tality in the Losartan Intervention For lence, awareness, treatment, and control
the Trials of Hypertension Prevention, Endpoint reduction in hypertension study of hypertension in the United States,
phase II. Arch Intern Med 1997;157:657- (LIFE): a randomised trial against atenolol. 1988-2000. JAMA 2003;290:199-206.
67. Lancet 2002;359:995-1003. 44. Thomas AJ, Eberly LE, Smith GD,
22. Cook NR, Cutler JA, Obarzanek E, et al. 32. Wing LM, Reid M, Ryan P, et al. A com- Neaton JD, Stamler J. Race/ethnicity, in-
Long term effects of dietary sodium re- parison of outcomes with angiotensin- come, major risk factors, and cardiovascu-
duction on cardiovascular disease out- converting–enzyme inhibitors and diuret- lar disease mortality. Am J Public Health
comes: observational follow-up of the ics for hypertension in the elderly. N Engl 2005;95:1417-23.
Trials of Hypertension Prevention (TOHP). J Med 2003;348:583-92. 45. Vasan RS, Beiser A, Seshadri S, et al.
BMJ 2007;334:885-92. 33. Jamerson K, Weber M, Bakris G, et al. Residual lifetime risk for developing hyper-
23. Cook NR, Obarzanek E, Cutler JA, et Benazepril plus amlodipine or hydrochloro­ tension in middle-aged women and men:
al. Joint effects of sodium and potassium thiazide for hypertension in high-risk pa- the Framingham Heart Study. JAMA 2002;
intake on subsequent cardiovascular dis- tients. N Engl J Med 2008;359:2417-28. 287:1003-10.
ease: the Trials of Hypertension Preven- 34. Weber MA, Julius S, Kjeldsen SE, et al. 46. Diez Roux AV, Chambless L, Merkin
tion follow-up Study. Arch Intern Med Blood pressure dependent and indepen- SS, et al. Socioeconomic disadvantage and
2009;169:32-40. dent effects of antihypertensive therapy change in blood pressure associated with
24. Mancia G, De Backer G, Dominczak A, on clinical events in the VALUE Trial. aging. Circulation 2002;106:703-10.
et al. 2007 Guidelines for the manage- Lancet 2004;363:2049-51. 47. Chobanian AV. Prehypertension revis-
ment of arterial hypertension: Task Force 35. Lindholm LH, Carlberg B, Samuelsson ited. Hypertension 2006;48:812-4.
for the Management of Arterial Hyperten- O. Should beta blockers remain first choice 48. Lewington S, Clarke R, Qizilbash N,
sion of the European Society of Hyperten- in the treatment of primary hypertension? Peto R, Collins R. Age-specific relevance
sion (ESH) and of the European Society of A meta-analysis. Lancet 2005;366:1545-53. of usual blood pressure to vascular mor-
Cardiology (ESC). J Hypertens 2007;25: 36. Prevention of cardiovascular disease: tality: a meta-analysis of individual data
1105-87. guidelines for assessment and manage- for one million adults in 61 prospective
25. Law MR, Wald NJ, Morris JK, Jordan ment of cardiovascular risk. Geneva: studies. Lancet 2002;360:1903-13. [Erra-
RE. Value of low dose combination treat- World Health Organization, 2007. (Ac- tum, Lancet 2003;361:1060.]
ment with blood pressure lowering drugs: cessed July 24, 2009, at http://www.who. 49. Toprak A, Wang H, Chen W, et al. Pre-
analysis of 354 randomised trials. BMJ int/cardiovascular_diseases/guidelines/ hypertension and black-white contrasts in
2003;326:1427-34. Full%20text.pdf.) cardiovascular risk in young adults: Boga-
26. Chobanian AV, Bakris GL, Black HR, 37. Black HR, Bakris GL, Weber MA, et al. lusa Heart Study. J Hypertens 2009;27:
et al. Seventh report of the Joint National Efficacy and safety of darusentan in pa- 243-50.
Committee on the Prevention, Detection, tients with resistant hypertension: results 50. Levy D, Ehret GB, Rice K, et al. Genome-

886 n engl j med 361;9  nejm.org  august 27, 2009

Downloaded from www.nejm.org on August 26, 2009 . For personal use only. No other uses without permission.
Copyright © 2009 Massachusetts Medical Society. All rights reserved.
 shattuck lecture

wide association study of blood pressure sensitivity and resistance of blood pres- Ferrell MF, Prineas RJ, Laing B. Effect of
and hypertension. Nat Genet 2009;41:677- sure: age and race as factors in physiologi- dietary sodium restriction on taste re-
87. cal responses. Hypertension 1991;17:Suppl: sponses to sodium chloride: a longitudi-
51. Newton-Cheh C, Johnson T, Gateva V, I102-I108. nal study. Am J Clin Nutr 1986;44:232-43.
et al. Genome-wide association study 58. He FJ, MacGregor GA. Importance of 64. He J, Whelton PK, Appel LJ, Charles-
identifies eight loci associated with blood salt in determining blood pressure in ton J, Klag MJ. Long-term effects of
pressure. Nat Genet 2009;41:666-76. children: meta-analysis of controlled trials. weight loss and dietary sodium reduction
52. Dahl LK, Love RA. Evidence for rela- Hypertension 2006;48:861-9. on incidence of hypertension. Hyperten-
tionship between sodium (chloride) intake 59. Intersalt Cooperative Research Group. sion 2000;35:544-9.
and human essential hypertension. Arch Intersalt: an international study of electro- 65. Wang Y, Beydoun MA. The obesity
Intern Med 1954;94:525-31. lyte excretion and blood pressure: results epidemic in the United States — gender,
53. Briefel RR, Johnson CL. Secular trends for 24-hour urinary sodium and potassium age, socioeconomic, racial/ethnic, and
in dietary intake in the United States. excretion. BMJ 1988;297:319-28. geographic characteristics: a systematic
Annu Rev Nutr 2004;24:401-31. 60. Adrogué HJ, Madias NE. Sodium and review and meta-regression analysis. Epi-
54. Cohen HW, Hailpern SM, Fang J, Al- potassium in the pathogenesis of hyper- demiol Rev 2007;29:6-28.
derman MH. Sodium intake and mortality tension. N Engl J Med 2007;356:1966-78. 66. McGinnis JM, Gootman JA, Kraak VI,
in the NHANES II follow-up study. Am J 61. Panel on Dietary Reference Intakes eds. Food marketing to children and youth:
Med 2006;119(3):275.e7-275.e14. for Electrolytes and Water, Standing Com- threat or opportunity? Washington, DC:
55. F as in fat: how obesity policies are mittee on the Scientific Evaluation of Di- National Academies Press, 2006.
failing in America. Princeton, NJ: Robert etary Reference Intakes. Dietary reference 67. French SA, Story M, Jeffery RW. Envi-
Wood Johnson Foundation, 2008. (Ac- intakes for water, potassium, sodium ronmental influences on eating and phys-
cessed July 24, 2009, at http://www. chloride, and sulfate. Washington, DC: ical activity. Annu Rev Public Health
healthyeatingresearch.org/images/stories/ National Academies Press, 2004. (Ac- 2001;22:309-35.
her_externalreports/obesity2008report. cessed July 24, 2009, at http://www.nap. 68. Evans WD, Finkelstein EA, Kamerow
pdf.) edu/openbook.php?isbn=0309091691.) DB, Renaud JM. Public perceptions of
56. Lifton RP, Gharavi AG, Geller DS. Mo- 62. Karppanen H, Mervaala E. Sodium childhood obesity. Am J Prev Med 2005;
lecular mechanisms of human hyperten- intake and hypertension. Prog Cardiovasc 28:26-32.
sion. Cell 2001;104:545-56. Dis 2006;49:59-75. Copyright © 2009 Massachusetts Medical Society.
57. Luft FC, Miller JZ, Grim CE, et al. Salt 63. Blais CA, Pangborn RM, Borhani NO,

images in clinical medicine

The Journal welcomes consideration of new submissions for Images in Clinical
Medicine. Instructions for authors and procedures for submissions can be found
on the Journal’s Web site at NEJM.org. At the discretion of the editor, images that
are accepted for publication may appear in the print version of the Journal,
the electronic version, or both.

n engl j med 361;9  nejm.org  august 27, 2009 887

Downloaded from www.nejm.org on August 26, 2009 . For personal use only. No other uses without permission.
Copyright © 2009 Massachusetts Medical Society. All rights reserved.