Steven A. Dosh, MD, MS

Hypertension is arbitrarily defined as diastolic blood pressure (DBP) of 90 mm Hg or higher, systolic blood pressure (SBP) of 140 mm Hg or higher, or both, on 3 separate occasions. Essential hypertension is hypertension without an identifiable cause. Essential hypertension, also known as primary or idiopathic hypertension, accounts for at least 95% of all cases of hypertension.

According to the third National Health and Nutrition Examination Survey (NHANES III), approximately 60% of the 50 million Americans with hypertension are at increased risk for cardiovascular disease resulting from uncontrolled hypertension. This is because only 53% of hypertensive patients are being treated and only 24% have their hypertension under control.¹ Physicians must play an active role in identifying and treating hypertension.

In an earlier Applied Evidence article² an approach to the diagnosis of hypertension was presented. This article reviews the treatment of essential hypertension in adults and the prognosis of untreated hypertension. Risk stratification, alternative therapies, lifestyle modification, drug therapy, and prognosis will each be reviewed sequentially.

Risk stratification

The decision to treat hypertension and the choice of treatment is affected by the patient’s risk of morbidity and mortality if the blood pressure remains untreated or under-treated. According to the recommendations of the sixth report of the Joint National Committee on the Prevention, Diagnosis, Evaluation, and Treatment of High Blood Pressure (JNC-VI), the first step in planning treatment of a patient with essential hypertension is to categorize the patient’s risk status.³ The patient is placed in 1 of 9 treatment categories according to his or her blood pressure category, cardiovascular risk factors, and evidence of end-organ damage found during the initial evaluation (Table 1). Once the treatment category is identified, initial treatment should begin (Figure 1). Subsequent treatment depends on the patient’s response to initial treatment (Figure 2).

Patients should be monitored regularly to be sure they do not develop signs and symptoms that would place them in a different category and mandate more aggressive treatment. After a patient’s blood pressure has been controlled for 1 year, it may be possible to decrease the dose or the number of antihypertensive drugs—especially among patients who make significant lifestyle changes.⁴

The effectiveness of therapy varies depending on the patient’s cardiovascular risk. The New Zealand Guidelines Group has developed a helpful risk calculator based on the Framingham Heart Study for estimating a patient’s cardiovascular risk. This calculator incorporates sex, age, systolic blood pressure, smoking status, total cholesterol, high-density lipoprotein cholesterol, presence or absence of diabetes, and presence or absence of electrocardiogram evidence of left ventricular hypertrophy. This helpful risk calculator may be downloaded from the Web site of the New Zealand Guidelines Group at http://www.nzgg.org.nz/library/gl_complete/bloodpressure/appendix.cfm#app3. Alternatively, the University of Sheffield Medical School has developed tables to estimate an individual’s risk of heart disease based on cardiovascular risk factors including age, sex, cholesterol level, and presence or absence of smoking, hypertension, and diabetes—Sheffield tables.⁵ Software for handheld computers (Palm and PocketPC) that helps you estimate risk is available at www.jfponline.com.

Regardless of the method used, the benefit of treatment increases steadily as the patient’s current cardiovascular risk increases. With a 5-year cardiovascular risk of less than 2.5%, more than 120 patients have to be treated for 5 years to prevent 1 cardiovascular event; this number decreases to 25 patients with a risk of between 5% and 10%, and only 13 with a risk of between 20% and 24%.⁶ It is tempting to assume that the benefit of hypertension treatment is related to reduction in blood pressure whether achieved by drug therapy, lifestyle modification, or alternative therapy. However, this has not been established and it is important to consider the evidence supporting the benefit of each of these therapeutic options (Table 2).

TABLE 1
HYPERTENSION RISK STRATIFICATION AND TREATMENT CATEGORIES

FIGURE 1
ALGORITHM FOR INITIAL MANAGEMENT OF HYPERTENSION

FIGURE 2
ALGORITHM FOR SUBSEQUENT MANAGEMENT OF HYPERTENSION

TABLE 2
NUMBER NEEDED TO TREAT (NNT) FOR SPECIFIC ANTIHYPERTENSIVE TREATMENTS

Treatment

Drug Therapy

Patients who require drug treatment for hypertension should begin with a low dose of the initial medication, and that dose should be slowly titrated upward every 1 to 2 months (Figure 2). The JNC-VI recommends a diuretic or a ß-blocker with once daily dosing and 24-hour efficacy as the initial treatment for most hypertensive patients. However, the choice of initial medication will be affected by concomitant illnesses: (1) ß-blockers are recommended for the initial treatment of patients with hypertension and a history of coronary artery disease; (2) diuretics are suggested for the initial treatment of isolated systolic hypertension; (3) and angiotensin-converting enzyme (ACE) inhibitors are recommended for hypertensive patients who have systolic dysfunction after myocardial infarction, diabetic nephropathy, or congestive heart failure. Angiotensin II receptor blockers may be used in patients who cannot tolerate ACE inhibitors because of cough or rash. Alpha-adrenergic blockers should be used with caution in light of evidence that they may increase the risk of cardiovascular events (especially congestive heart failure).⁷

Among patients who do not have concomitant illness, the choice of drug therapy is controversial. A case-control study and a meta-analysis suggested that short-acting calcium channel blockers (CCBs) increase cardiovascular mortality.^8,9 Unfortunately, these studies were not designed to establish a causal relationship. A recent nonsystematic review suggested that short-acting CCBs should be avoided and that conventional therapies were more effective than long-acting CCBs.¹⁰ An earlier non-systematic review suggested that short- and intermediate-acting CCBs were associated with increased cardiovascular mortality and morbidity. However, a well-designed cohort study of patients with coronary artery disease failed to reveal an increase in adverse effects among patients taking short-acting CCBs.¹¹ Furthermore, randomized controlled trials suggest that diuretics, ß-blockers, and long-acting CCBs are equally effective in preventing cardiovascular mortality and morbidity.^12,13 Physicians who treat hypertension must choose the best initial treatment for patients who do not have concomitant illness. Fortunately, safety, tolerability, efficacy, price, and simplicity can guide the physician to an ideal drug for most hypertensive patients.

Low-dose thiazide diuretics (the equivalent of 25 to 50 mg of hydrochlorothiazide) appear better tolerated than ß-blockers or CCBs.¹⁴ Treatment with ß-blockers, CCBs, and ACE inhibitors is also more expensive (75% to 85% more) than diuretic therapy.¹⁵ The cost savings offered by diuretics complement the fact that diuretics are safe, effective, and may be dosed once daily. In short, in addition to being the drug of choice for isolated systolic hypertension, low-dose thiazide diuretics are the ideal initial drug treatment of patients without concomitant illness. It should be noted that higher doses of thiazide diuretics offer proportionately less blood pressure reduction and greater risk of hypokalemia.¹⁶

Antihypertensive treatment reduces morbidity and mortality for all stages of hypertension, but people with the greatest baseline cardiovascular risk (eg, older patients and patients with higher levels of blood pressure) have the most to gain from treatment.^17,18 There is no conclusive evidence to suggest that lowering blood pressure to below 140/80 reduces morbidity or mortality in most patients. However, patients who have diabetes or renal failure benefit from more aggressive management of blood pressure.^19,20 Therefore, the JNC-VI recommends a target blood pressure of less than 130/85 for these patients.

The JNC-VI recommendation to start with a low-dose diuretic is supported by the evidence across a spectrum of patient-oriented outcomes. The effectiveness of diuretics and ß-blockers as first-line agents has been confirmed by long-term clinical trials.^14,21 However, low-dose thiazides appear effective against a broader range of outcomes than high-dose thiazides and ß-blockers (Table 3). There is also evidence to suggest that CCBs and ACE inhibitors may be effective first-line agents, but fewer patients have been studied who take CCBs and ACE inhibitors than those who take diuretics and ß-blockers.^22,23

Most patients with hypertension will respond to 1 (approximately 50%) or 2 (approximately 30%) antihypertensive medications.^19,24,25 As noted earlier, failure to respond to treatment suggests an identifiable cause of hypertension. Among patients who do not have a secondary cause of hypertension, inadequate drug treatment (often failure to start a diuretic) and noncompliance are among the most common causes of resistant hypertension.²⁶

When patients who are receiving drug therapy fail to reach the target blood pressure goal or fail to maintain the blood pressure goal, they should have the initial drug dose increased until the goal is reached (Figure 2). Those who fail initial drug therapy at full doses should have a second drug added and increased until the blood pressure goal is reached. Those who fail initial and second drug therapy at full doses should have a third drug added and increased until the pressure goal is reached. Patients who fail to reach the goal on maximal doses of 3 drugs have, by definition, resistant hypertension and will require evaluation by a physician with expertise in managing resistant hypertension. A cause should be sought each time a patient fails to respond to a drug or fails to maintain blood pressure control on a drug that had previously controlled the pressure (Table 3).

TABLE 3
PARTIAL LISTING OF CAUSES OF FAILURE TO REACH OR MAINTAIN TARGET BLOOD PRESSURE

Lifestyle modifications

Several lifestyle modifications are recommended in all treatment categories. Aerobic exercise (45 to 60 minutes at least 3 days per week), low-salt, low-fat, and high fruit and vegetable diet, limited alcohol consumption (less than 3 drinks per day), and modest weight loss (3% to 9% of total body weight) have been demonstrated to yield modest blood pressure reductions, but there is insufficient evidence to suggest that these measures alone reduce morbidity or mortality in hypertensive patients.^27-33 A systematic review of randomized controlled trials found an average 4.4/2.5 mm Hg reduction in blood pressure with no evidence of harm (among patients who were not at risk for hyperkalemia) when diet was supplemented with about 2000 mg of potassium daily.³⁴ A comparable reduction in blood pressure was seen with a daily supplement of more than 3 grams of fish oil.³⁵ Research concerning the value of calcium and magnesium supplementation is conflicting and insufficient for supplementation to be considered standard therapy at this time.

Alternative therapy

The number and the quality of studies evaluating acupuncture, biofeedback, herbal medicine, transcendental meditation, and yoga are, for the most part, limited. They have focused on reduction in blood pressure, not patient-oriented outcomes, such as a reduction in morbidity and mortality. Acupuncture does not appear to have a significant effect on blood pressure levels.^36,37 Biofeedback and other behavioral techniques have not been demonstrated to reduce blood pressure.^38,39 The effect of garlic on blood pressure is unclear with mixed study results.^40,41 Transcendental meditation and yoga may reduce blood pressure, but studies of these modalities are small and the experimental designs have a limited capacity to detect an independent treatment effect or a placebo effect.^42,43

Therefore, physicians who include any of these modalities in their hypertension treatment plan should carefully monitor each patient for adequacy of blood pressure control, development of risk factors, and evidence of end-organ damage. At this time, alternative therapies should be considered experimental adjuncts to lifestyle modification and medical therapy that have not been shown to improve patient-oriented outcomes.

Follow-up of patients with hypertension

Follow-up visits should be designed to identify new risk factors, evidence of end-organ damage, and adequacy of blood pressure control. Follow-up visits may include an interval history, limited physical examination, radiologic evaluation, and laboratory testing. The frequency and nature of follow-up hypertension evaluations will vary according to the presence or absence of preexisting risk factors, evidence of end-organ damage, the nature of the treatment the patient is receiving, and the stability of blood pressure control. Unfortunately, there is little evidence to support specific recommendations for the frequency and nature of follow-up hypertension evaluations.

In the absence of evidence, several general principles may be suggested. Patients should be seen within 2 months of initiation of treatment. Follow-up history should focus on the cardiovascular and neurologic review of systems. The examination should include a focused cardiovascular work-up (eg, retinopathy, carotid bruits). Consideration should be given to periodic laboratory testing for diabetes, renal insufficiency, and hyperlipidemia. Periodic (but less frequent) chest x-rays and electrocardiograms may be helpful to detect cardiomegaly, but there is no evidence to support such testing in the absence of symptoms.

Follow-up visits should be more frequent among patients who have marginal blood pressure control, preexisting risk factors, or end-organ damage. Evaluations may be less frequent among those with good control and no preexisting risk factors or endorgan damage. Office visits and testing should be more frequent whenever changes are made in treatment. The frequency and nature of follow-up testing will also depend on the nature of treatment. Patients taking diuretics should have their potassium levels checked periodically. Renal function and potassium should be monitored in patients who are taking ACE inhibitors, especially during the first few weeks of therapy.

Patients willing to regularly monitor their blood pressure at home may require less frequent follow-up than those who leave it to the physician to check. Patients who monitor their blood pressure at home should have their sphygmomanometers validated initially and periodically.⁴⁴ It is important to remember that home blood pressure measurements are consistently lower and more closely correlated with cardiovascular outcomes than are clinic blood pressure measurements.^45-47

Prognosis

It is difficult to estimate the precise impact blood pressure control has on morbidity and mortality, but it is clear that high blood pressure, if unrecognized or untreated, substantially increases the morbidity and mortality associated with coronary disease, heart failure, renal failure, and stroke.¹⁷ In an early study of untreated hypertension there was a close relationship between blood pressure level and cardiovascular morbidity over 14 years of observation. This study revealed that hypertensive patients (those with a blood pressure of 160/95) had cardiovascular morbidity rates (coronary artery disease, claudication, stroke, and congestive heart failure) 2 to 3 times higher than normotensive patients.⁴⁸ The impact of inadequately controlled blood pressure on morbidity and mortality among patients with diabetes is especially problematic.⁴⁹ Over 9 years, when compared with diabetic patients with less tight control (< 180/105 mm Hg), those with tight blood pressure control (< 150/85 mm Hg) had a 24% reduction in sudden death, hyperglycemic or hypoglycemic death, fatal or nonfatal myocardial infarction, angina, heart failure, fatal or nonfatal stroke, renal failure, amputation, vitreous hemorrhage, and retinal hemorrhage.

Finally, renal function deteriorates more rapidly when blood pressure control is inadequate in patients with chronic renal disease of diverse causes.²⁰ Over 2 years, when compared with patients with renal failure who had less tight control (mean arterial pressure 107 mm Hg), renal failure patients with tight control (mean arterial blood pressure 92 mm Hg) had significantly less proteinuria and lower rates of decline in renal function. Whether this translates into a significant improvement in the risk of end-stage renal disease is unknown.

Hypertension is arbitrarily defined as diastolic blood pressure (DBP) of 90 mm Hg or higher, systolic blood pressure (SBP) of 140 mm Hg or higher, or both, on 3 separate occasions. Essential hypertension is hypertension without an identifiable cause. Essential hypertension, also known as primary or idiopathic hypertension, accounts for at least 95% of all cases of hypertension.

According to the third National Health and Nutrition Examination Survey (NHANES III), approximately 60% of the 50 million Americans with hypertension are at increased risk for cardiovascular disease resulting from uncontrolled hypertension. This is because only 53% of hypertensive patients are being treated and only 24% have their hypertension under control.¹ Physicians must play an active role in identifying and treating hypertension.

In an earlier Applied Evidence article² an approach to the diagnosis of hypertension was presented. This article reviews the treatment of essential hypertension in adults and the prognosis of untreated hypertension. Risk stratification, alternative therapies, lifestyle modification, drug therapy, and prognosis will each be reviewed sequentially.

Risk stratification

The decision to treat hypertension and the choice of treatment is affected by the patient’s risk of morbidity and mortality if the blood pressure remains untreated or under-treated. According to the recommendations of the sixth report of the Joint National Committee on the Prevention, Diagnosis, Evaluation, and Treatment of High Blood Pressure (JNC-VI), the first step in planning treatment of a patient with essential hypertension is to categorize the patient’s risk status.³ The patient is placed in 1 of 9 treatment categories according to his or her blood pressure category, cardiovascular risk factors, and evidence of end-organ damage found during the initial evaluation (Table 1). Once the treatment category is identified, initial treatment should begin (Figure 1). Subsequent treatment depends on the patient’s response to initial treatment (Figure 2).

Patients should be monitored regularly to be sure they do not develop signs and symptoms that would place them in a different category and mandate more aggressive treatment. After a patient’s blood pressure has been controlled for 1 year, it may be possible to decrease the dose or the number of antihypertensive drugs—especially among patients who make significant lifestyle changes.⁴

The effectiveness of therapy varies depending on the patient’s cardiovascular risk. The New Zealand Guidelines Group has developed a helpful risk calculator based on the Framingham Heart Study for estimating a patient’s cardiovascular risk. This calculator incorporates sex, age, systolic blood pressure, smoking status, total cholesterol, high-density lipoprotein cholesterol, presence or absence of diabetes, and presence or absence of electrocardiogram evidence of left ventricular hypertrophy. This helpful risk calculator may be downloaded from the Web site of the New Zealand Guidelines Group at http://www.nzgg.org.nz/library/gl_complete/bloodpressure/appendix.cfm#app3. Alternatively, the University of Sheffield Medical School has developed tables to estimate an individual’s risk of heart disease based on cardiovascular risk factors including age, sex, cholesterol level, and presence or absence of smoking, hypertension, and diabetes—Sheffield tables.⁵ Software for handheld computers (Palm and PocketPC) that helps you estimate risk is available at www.jfponline.com.

Regardless of the method used, the benefit of treatment increases steadily as the patient’s current cardiovascular risk increases. With a 5-year cardiovascular risk of less than 2.5%, more than 120 patients have to be treated for 5 years to prevent 1 cardiovascular event; this number decreases to 25 patients with a risk of between 5% and 10%, and only 13 with a risk of between 20% and 24%.⁶ It is tempting to assume that the benefit of hypertension treatment is related to reduction in blood pressure whether achieved by drug therapy, lifestyle modification, or alternative therapy. However, this has not been established and it is important to consider the evidence supporting the benefit of each of these therapeutic options (Table 2).

TABLE 1
HYPERTENSION RISK STRATIFICATION AND TREATMENT CATEGORIES

FIGURE 1
ALGORITHM FOR INITIAL MANAGEMENT OF HYPERTENSION

FIGURE 2
ALGORITHM FOR SUBSEQUENT MANAGEMENT OF HYPERTENSION

TABLE 2
NUMBER NEEDED TO TREAT (NNT) FOR SPECIFIC ANTIHYPERTENSIVE TREATMENTS

Treatment

Drug Therapy

Patients who require drug treatment for hypertension should begin with a low dose of the initial medication, and that dose should be slowly titrated upward every 1 to 2 months (Figure 2). The JNC-VI recommends a diuretic or a ß-blocker with once daily dosing and 24-hour efficacy as the initial treatment for most hypertensive patients. However, the choice of initial medication will be affected by concomitant illnesses: (1) ß-blockers are recommended for the initial treatment of patients with hypertension and a history of coronary artery disease; (2) diuretics are suggested for the initial treatment of isolated systolic hypertension; (3) and angiotensin-converting enzyme (ACE) inhibitors are recommended for hypertensive patients who have systolic dysfunction after myocardial infarction, diabetic nephropathy, or congestive heart failure. Angiotensin II receptor blockers may be used in patients who cannot tolerate ACE inhibitors because of cough or rash. Alpha-adrenergic blockers should be used with caution in light of evidence that they may increase the risk of cardiovascular events (especially congestive heart failure).⁷

Among patients who do not have concomitant illness, the choice of drug therapy is controversial. A case-control study and a meta-analysis suggested that short-acting calcium channel blockers (CCBs) increase cardiovascular mortality.^8,9 Unfortunately, these studies were not designed to establish a causal relationship. A recent nonsystematic review suggested that short-acting CCBs should be avoided and that conventional therapies were more effective than long-acting CCBs.¹⁰ An earlier non-systematic review suggested that short- and intermediate-acting CCBs were associated with increased cardiovascular mortality and morbidity. However, a well-designed cohort study of patients with coronary artery disease failed to reveal an increase in adverse effects among patients taking short-acting CCBs.¹¹ Furthermore, randomized controlled trials suggest that diuretics, ß-blockers, and long-acting CCBs are equally effective in preventing cardiovascular mortality and morbidity.^12,13 Physicians who treat hypertension must choose the best initial treatment for patients who do not have concomitant illness. Fortunately, safety, tolerability, efficacy, price, and simplicity can guide the physician to an ideal drug for most hypertensive patients.

Low-dose thiazide diuretics (the equivalent of 25 to 50 mg of hydrochlorothiazide) appear better tolerated than ß-blockers or CCBs.¹⁴ Treatment with ß-blockers, CCBs, and ACE inhibitors is also more expensive (75% to 85% more) than diuretic therapy.¹⁵ The cost savings offered by diuretics complement the fact that diuretics are safe, effective, and may be dosed once daily. In short, in addition to being the drug of choice for isolated systolic hypertension, low-dose thiazide diuretics are the ideal initial drug treatment of patients without concomitant illness. It should be noted that higher doses of thiazide diuretics offer proportionately less blood pressure reduction and greater risk of hypokalemia.¹⁶

Antihypertensive treatment reduces morbidity and mortality for all stages of hypertension, but people with the greatest baseline cardiovascular risk (eg, older patients and patients with higher levels of blood pressure) have the most to gain from treatment.^17,18 There is no conclusive evidence to suggest that lowering blood pressure to below 140/80 reduces morbidity or mortality in most patients. However, patients who have diabetes or renal failure benefit from more aggressive management of blood pressure.^19,20 Therefore, the JNC-VI recommends a target blood pressure of less than 130/85 for these patients.

The JNC-VI recommendation to start with a low-dose diuretic is supported by the evidence across a spectrum of patient-oriented outcomes. The effectiveness of diuretics and ß-blockers as first-line agents has been confirmed by long-term clinical trials.^14,21 However, low-dose thiazides appear effective against a broader range of outcomes than high-dose thiazides and ß-blockers (Table 3). There is also evidence to suggest that CCBs and ACE inhibitors may be effective first-line agents, but fewer patients have been studied who take CCBs and ACE inhibitors than those who take diuretics and ß-blockers.^22,23

Most patients with hypertension will respond to 1 (approximately 50%) or 2 (approximately 30%) antihypertensive medications.^19,24,25 As noted earlier, failure to respond to treatment suggests an identifiable cause of hypertension. Among patients who do not have a secondary cause of hypertension, inadequate drug treatment (often failure to start a diuretic) and noncompliance are among the most common causes of resistant hypertension.²⁶

When patients who are receiving drug therapy fail to reach the target blood pressure goal or fail to maintain the blood pressure goal, they should have the initial drug dose increased until the goal is reached (Figure 2). Those who fail initial drug therapy at full doses should have a second drug added and increased until the blood pressure goal is reached. Those who fail initial and second drug therapy at full doses should have a third drug added and increased until the pressure goal is reached. Patients who fail to reach the goal on maximal doses of 3 drugs have, by definition, resistant hypertension and will require evaluation by a physician with expertise in managing resistant hypertension. A cause should be sought each time a patient fails to respond to a drug or fails to maintain blood pressure control on a drug that had previously controlled the pressure (Table 3).

TABLE 3
PARTIAL LISTING OF CAUSES OF FAILURE TO REACH OR MAINTAIN TARGET BLOOD PRESSURE

Lifestyle modifications

Several lifestyle modifications are recommended in all treatment categories. Aerobic exercise (45 to 60 minutes at least 3 days per week), low-salt, low-fat, and high fruit and vegetable diet, limited alcohol consumption (less than 3 drinks per day), and modest weight loss (3% to 9% of total body weight) have been demonstrated to yield modest blood pressure reductions, but there is insufficient evidence to suggest that these measures alone reduce morbidity or mortality in hypertensive patients.^27-33 A systematic review of randomized controlled trials found an average 4.4/2.5 mm Hg reduction in blood pressure with no evidence of harm (among patients who were not at risk for hyperkalemia) when diet was supplemented with about 2000 mg of potassium daily.³⁴ A comparable reduction in blood pressure was seen with a daily supplement of more than 3 grams of fish oil.³⁵ Research concerning the value of calcium and magnesium supplementation is conflicting and insufficient for supplementation to be considered standard therapy at this time.

Alternative therapy

The number and the quality of studies evaluating acupuncture, biofeedback, herbal medicine, transcendental meditation, and yoga are, for the most part, limited. They have focused on reduction in blood pressure, not patient-oriented outcomes, such as a reduction in morbidity and mortality. Acupuncture does not appear to have a significant effect on blood pressure levels.^36,37 Biofeedback and other behavioral techniques have not been demonstrated to reduce blood pressure.^38,39 The effect of garlic on blood pressure is unclear with mixed study results.^40,41 Transcendental meditation and yoga may reduce blood pressure, but studies of these modalities are small and the experimental designs have a limited capacity to detect an independent treatment effect or a placebo effect.^42,43

Therefore, physicians who include any of these modalities in their hypertension treatment plan should carefully monitor each patient for adequacy of blood pressure control, development of risk factors, and evidence of end-organ damage. At this time, alternative therapies should be considered experimental adjuncts to lifestyle modification and medical therapy that have not been shown to improve patient-oriented outcomes.

Follow-up of patients with hypertension

Follow-up visits should be designed to identify new risk factors, evidence of end-organ damage, and adequacy of blood pressure control. Follow-up visits may include an interval history, limited physical examination, radiologic evaluation, and laboratory testing. The frequency and nature of follow-up hypertension evaluations will vary according to the presence or absence of preexisting risk factors, evidence of end-organ damage, the nature of the treatment the patient is receiving, and the stability of blood pressure control. Unfortunately, there is little evidence to support specific recommendations for the frequency and nature of follow-up hypertension evaluations.

In the absence of evidence, several general principles may be suggested. Patients should be seen within 2 months of initiation of treatment. Follow-up history should focus on the cardiovascular and neurologic review of systems. The examination should include a focused cardiovascular work-up (eg, retinopathy, carotid bruits). Consideration should be given to periodic laboratory testing for diabetes, renal insufficiency, and hyperlipidemia. Periodic (but less frequent) chest x-rays and electrocardiograms may be helpful to detect cardiomegaly, but there is no evidence to support such testing in the absence of symptoms.

Follow-up visits should be more frequent among patients who have marginal blood pressure control, preexisting risk factors, or end-organ damage. Evaluations may be less frequent among those with good control and no preexisting risk factors or endorgan damage. Office visits and testing should be more frequent whenever changes are made in treatment. The frequency and nature of follow-up testing will also depend on the nature of treatment. Patients taking diuretics should have their potassium levels checked periodically. Renal function and potassium should be monitored in patients who are taking ACE inhibitors, especially during the first few weeks of therapy.

Patients willing to regularly monitor their blood pressure at home may require less frequent follow-up than those who leave it to the physician to check. Patients who monitor their blood pressure at home should have their sphygmomanometers validated initially and periodically.⁴⁴ It is important to remember that home blood pressure measurements are consistently lower and more closely correlated with cardiovascular outcomes than are clinic blood pressure measurements.^45-47

Prognosis

It is difficult to estimate the precise impact blood pressure control has on morbidity and mortality, but it is clear that high blood pressure, if unrecognized or untreated, substantially increases the morbidity and mortality associated with coronary disease, heart failure, renal failure, and stroke.¹⁷ In an early study of untreated hypertension there was a close relationship between blood pressure level and cardiovascular morbidity over 14 years of observation. This study revealed that hypertensive patients (those with a blood pressure of 160/95) had cardiovascular morbidity rates (coronary artery disease, claudication, stroke, and congestive heart failure) 2 to 3 times higher than normotensive patients.⁴⁸ The impact of inadequately controlled blood pressure on morbidity and mortality among patients with diabetes is especially problematic.⁴⁹ Over 9 years, when compared with diabetic patients with less tight control (< 180/105 mm Hg), those with tight blood pressure control (< 150/85 mm Hg) had a 24% reduction in sudden death, hyperglycemic or hypoglycemic death, fatal or nonfatal myocardial infarction, angina, heart failure, fatal or nonfatal stroke, renal failure, amputation, vitreous hemorrhage, and retinal hemorrhage.

Finally, renal function deteriorates more rapidly when blood pressure control is inadequate in patients with chronic renal disease of diverse causes.²⁰ Over 2 years, when compared with patients with renal failure who had less tight control (mean arterial pressure 107 mm Hg), renal failure patients with tight control (mean arterial blood pressure 92 mm Hg) had significantly less proteinuria and lower rates of decline in renal function. Whether this translates into a significant improvement in the risk of end-stage renal disease is unknown.

Hypertension is arbitrarily defined as diastolic blood pressure (DBP) of 90 mm Hg or higher, systolic blood pressure (SBP) of 140 mm Hg or higher, or both, on 3 separate occasions. Essential hypertension is hypertension without an identifiable cause. Essential hypertension, also known as primary or idiopathic hypertension, accounts for at least 95% of all cases of hypertension.

According to the third National Health and Nutrition Examination Survey (NHANES III), approximately 60% of the 50 million Americans with hypertension are at increased risk for cardiovascular disease resulting from uncontrolled hypertension. This is because only 53% of hypertensive patients are being treated and only 24% have their hypertension under control.¹ Physicians must play an active role in identifying and treating hypertension.

In an earlier Applied Evidence article² an approach to the diagnosis of hypertension was presented. This article reviews the treatment of essential hypertension in adults and the prognosis of untreated hypertension. Risk stratification, alternative therapies, lifestyle modification, drug therapy, and prognosis will each be reviewed sequentially.

Risk stratification

The decision to treat hypertension and the choice of treatment is affected by the patient’s risk of morbidity and mortality if the blood pressure remains untreated or under-treated. According to the recommendations of the sixth report of the Joint National Committee on the Prevention, Diagnosis, Evaluation, and Treatment of High Blood Pressure (JNC-VI), the first step in planning treatment of a patient with essential hypertension is to categorize the patient’s risk status.³ The patient is placed in 1 of 9 treatment categories according to his or her blood pressure category, cardiovascular risk factors, and evidence of end-organ damage found during the initial evaluation (Table 1). Once the treatment category is identified, initial treatment should begin (Figure 1). Subsequent treatment depends on the patient’s response to initial treatment (Figure 2).

Patients should be monitored regularly to be sure they do not develop signs and symptoms that would place them in a different category and mandate more aggressive treatment. After a patient’s blood pressure has been controlled for 1 year, it may be possible to decrease the dose or the number of antihypertensive drugs—especially among patients who make significant lifestyle changes.⁴

The effectiveness of therapy varies depending on the patient’s cardiovascular risk. The New Zealand Guidelines Group has developed a helpful risk calculator based on the Framingham Heart Study for estimating a patient’s cardiovascular risk. This calculator incorporates sex, age, systolic blood pressure, smoking status, total cholesterol, high-density lipoprotein cholesterol, presence or absence of diabetes, and presence or absence of electrocardiogram evidence of left ventricular hypertrophy. This helpful risk calculator may be downloaded from the Web site of the New Zealand Guidelines Group at http://www.nzgg.org.nz/library/gl_complete/bloodpressure/appendix.cfm#app3. Alternatively, the University of Sheffield Medical School has developed tables to estimate an individual’s risk of heart disease based on cardiovascular risk factors including age, sex, cholesterol level, and presence or absence of smoking, hypertension, and diabetes—Sheffield tables.⁵ Software for handheld computers (Palm and PocketPC) that helps you estimate risk is available at www.jfponline.com.

Regardless of the method used, the benefit of treatment increases steadily as the patient’s current cardiovascular risk increases. With a 5-year cardiovascular risk of less than 2.5%, more than 120 patients have to be treated for 5 years to prevent 1 cardiovascular event; this number decreases to 25 patients with a risk of between 5% and 10%, and only 13 with a risk of between 20% and 24%.⁶ It is tempting to assume that the benefit of hypertension treatment is related to reduction in blood pressure whether achieved by drug therapy, lifestyle modification, or alternative therapy. However, this has not been established and it is important to consider the evidence supporting the benefit of each of these therapeutic options (Table 2).

TABLE 1
HYPERTENSION RISK STRATIFICATION AND TREATMENT CATEGORIES

FIGURE 1
ALGORITHM FOR INITIAL MANAGEMENT OF HYPERTENSION

FIGURE 2
ALGORITHM FOR SUBSEQUENT MANAGEMENT OF HYPERTENSION

TABLE 2
NUMBER NEEDED TO TREAT (NNT) FOR SPECIFIC ANTIHYPERTENSIVE TREATMENTS

Treatment

Drug Therapy

Patients who require drug treatment for hypertension should begin with a low dose of the initial medication, and that dose should be slowly titrated upward every 1 to 2 months (Figure 2). The JNC-VI recommends a diuretic or a ß-blocker with once daily dosing and 24-hour efficacy as the initial treatment for most hypertensive patients. However, the choice of initial medication will be affected by concomitant illnesses: (1) ß-blockers are recommended for the initial treatment of patients with hypertension and a history of coronary artery disease; (2) diuretics are suggested for the initial treatment of isolated systolic hypertension; (3) and angiotensin-converting enzyme (ACE) inhibitors are recommended for hypertensive patients who have systolic dysfunction after myocardial infarction, diabetic nephropathy, or congestive heart failure. Angiotensin II receptor blockers may be used in patients who cannot tolerate ACE inhibitors because of cough or rash. Alpha-adrenergic blockers should be used with caution in light of evidence that they may increase the risk of cardiovascular events (especially congestive heart failure).⁷

Among patients who do not have concomitant illness, the choice of drug therapy is controversial. A case-control study and a meta-analysis suggested that short-acting calcium channel blockers (CCBs) increase cardiovascular mortality.^8,9 Unfortunately, these studies were not designed to establish a causal relationship. A recent nonsystematic review suggested that short-acting CCBs should be avoided and that conventional therapies were more effective than long-acting CCBs.¹⁰ An earlier non-systematic review suggested that short- and intermediate-acting CCBs were associated with increased cardiovascular mortality and morbidity. However, a well-designed cohort study of patients with coronary artery disease failed to reveal an increase in adverse effects among patients taking short-acting CCBs.¹¹ Furthermore, randomized controlled trials suggest that diuretics, ß-blockers, and long-acting CCBs are equally effective in preventing cardiovascular mortality and morbidity.^12,13 Physicians who treat hypertension must choose the best initial treatment for patients who do not have concomitant illness. Fortunately, safety, tolerability, efficacy, price, and simplicity can guide the physician to an ideal drug for most hypertensive patients.

Low-dose thiazide diuretics (the equivalent of 25 to 50 mg of hydrochlorothiazide) appear better tolerated than ß-blockers or CCBs.¹⁴ Treatment with ß-blockers, CCBs, and ACE inhibitors is also more expensive (75% to 85% more) than diuretic therapy.¹⁵ The cost savings offered by diuretics complement the fact that diuretics are safe, effective, and may be dosed once daily. In short, in addition to being the drug of choice for isolated systolic hypertension, low-dose thiazide diuretics are the ideal initial drug treatment of patients without concomitant illness. It should be noted that higher doses of thiazide diuretics offer proportionately less blood pressure reduction and greater risk of hypokalemia.¹⁶

Antihypertensive treatment reduces morbidity and mortality for all stages of hypertension, but people with the greatest baseline cardiovascular risk (eg, older patients and patients with higher levels of blood pressure) have the most to gain from treatment.^17,18 There is no conclusive evidence to suggest that lowering blood pressure to below 140/80 reduces morbidity or mortality in most patients. However, patients who have diabetes or renal failure benefit from more aggressive management of blood pressure.^19,20 Therefore, the JNC-VI recommends a target blood pressure of less than 130/85 for these patients.

The JNC-VI recommendation to start with a low-dose diuretic is supported by the evidence across a spectrum of patient-oriented outcomes. The effectiveness of diuretics and ß-blockers as first-line agents has been confirmed by long-term clinical trials.^14,21 However, low-dose thiazides appear effective against a broader range of outcomes than high-dose thiazides and ß-blockers (Table 3). There is also evidence to suggest that CCBs and ACE inhibitors may be effective first-line agents, but fewer patients have been studied who take CCBs and ACE inhibitors than those who take diuretics and ß-blockers.^22,23

Most patients with hypertension will respond to 1 (approximately 50%) or 2 (approximately 30%) antihypertensive medications.^19,24,25 As noted earlier, failure to respond to treatment suggests an identifiable cause of hypertension. Among patients who do not have a secondary cause of hypertension, inadequate drug treatment (often failure to start a diuretic) and noncompliance are among the most common causes of resistant hypertension.²⁶

When patients who are receiving drug therapy fail to reach the target blood pressure goal or fail to maintain the blood pressure goal, they should have the initial drug dose increased until the goal is reached (Figure 2). Those who fail initial drug therapy at full doses should have a second drug added and increased until the blood pressure goal is reached. Those who fail initial and second drug therapy at full doses should have a third drug added and increased until the pressure goal is reached. Patients who fail to reach the goal on maximal doses of 3 drugs have, by definition, resistant hypertension and will require evaluation by a physician with expertise in managing resistant hypertension. A cause should be sought each time a patient fails to respond to a drug or fails to maintain blood pressure control on a drug that had previously controlled the pressure (Table 3).

TABLE 3
PARTIAL LISTING OF CAUSES OF FAILURE TO REACH OR MAINTAIN TARGET BLOOD PRESSURE

Lifestyle modifications

Several lifestyle modifications are recommended in all treatment categories. Aerobic exercise (45 to 60 minutes at least 3 days per week), low-salt, low-fat, and high fruit and vegetable diet, limited alcohol consumption (less than 3 drinks per day), and modest weight loss (3% to 9% of total body weight) have been demonstrated to yield modest blood pressure reductions, but there is insufficient evidence to suggest that these measures alone reduce morbidity or mortality in hypertensive patients.^27-33 A systematic review of randomized controlled trials found an average 4.4/2.5 mm Hg reduction in blood pressure with no evidence of harm (among patients who were not at risk for hyperkalemia) when diet was supplemented with about 2000 mg of potassium daily.³⁴ A comparable reduction in blood pressure was seen with a daily supplement of more than 3 grams of fish oil.³⁵ Research concerning the value of calcium and magnesium supplementation is conflicting and insufficient for supplementation to be considered standard therapy at this time.

Alternative therapy

The number and the quality of studies evaluating acupuncture, biofeedback, herbal medicine, transcendental meditation, and yoga are, for the most part, limited. They have focused on reduction in blood pressure, not patient-oriented outcomes, such as a reduction in morbidity and mortality. Acupuncture does not appear to have a significant effect on blood pressure levels.^36,37 Biofeedback and other behavioral techniques have not been demonstrated to reduce blood pressure.^38,39 The effect of garlic on blood pressure is unclear with mixed study results.^40,41 Transcendental meditation and yoga may reduce blood pressure, but studies of these modalities are small and the experimental designs have a limited capacity to detect an independent treatment effect or a placebo effect.^42,43

Therefore, physicians who include any of these modalities in their hypertension treatment plan should carefully monitor each patient for adequacy of blood pressure control, development of risk factors, and evidence of end-organ damage. At this time, alternative therapies should be considered experimental adjuncts to lifestyle modification and medical therapy that have not been shown to improve patient-oriented outcomes.

Follow-up of patients with hypertension

Follow-up visits should be designed to identify new risk factors, evidence of end-organ damage, and adequacy of blood pressure control. Follow-up visits may include an interval history, limited physical examination, radiologic evaluation, and laboratory testing. The frequency and nature of follow-up hypertension evaluations will vary according to the presence or absence of preexisting risk factors, evidence of end-organ damage, the nature of the treatment the patient is receiving, and the stability of blood pressure control. Unfortunately, there is little evidence to support specific recommendations for the frequency and nature of follow-up hypertension evaluations.

In the absence of evidence, several general principles may be suggested. Patients should be seen within 2 months of initiation of treatment. Follow-up history should focus on the cardiovascular and neurologic review of systems. The examination should include a focused cardiovascular work-up (eg, retinopathy, carotid bruits). Consideration should be given to periodic laboratory testing for diabetes, renal insufficiency, and hyperlipidemia. Periodic (but less frequent) chest x-rays and electrocardiograms may be helpful to detect cardiomegaly, but there is no evidence to support such testing in the absence of symptoms.

Follow-up visits should be more frequent among patients who have marginal blood pressure control, preexisting risk factors, or end-organ damage. Evaluations may be less frequent among those with good control and no preexisting risk factors or endorgan damage. Office visits and testing should be more frequent whenever changes are made in treatment. The frequency and nature of follow-up testing will also depend on the nature of treatment. Patients taking diuretics should have their potassium levels checked periodically. Renal function and potassium should be monitored in patients who are taking ACE inhibitors, especially during the first few weeks of therapy.

Patients willing to regularly monitor their blood pressure at home may require less frequent follow-up than those who leave it to the physician to check. Patients who monitor their blood pressure at home should have their sphygmomanometers validated initially and periodically.⁴⁴ It is important to remember that home blood pressure measurements are consistently lower and more closely correlated with cardiovascular outcomes than are clinic blood pressure measurements.^45-47

Prognosis

It is difficult to estimate the precise impact blood pressure control has on morbidity and mortality, but it is clear that high blood pressure, if unrecognized or untreated, substantially increases the morbidity and mortality associated with coronary disease, heart failure, renal failure, and stroke.¹⁷ In an early study of untreated hypertension there was a close relationship between blood pressure level and cardiovascular morbidity over 14 years of observation. This study revealed that hypertensive patients (those with a blood pressure of 160/95) had cardiovascular morbidity rates (coronary artery disease, claudication, stroke, and congestive heart failure) 2 to 3 times higher than normotensive patients.⁴⁸ The impact of inadequately controlled blood pressure on morbidity and mortality among patients with diabetes is especially problematic.⁴⁹ Over 9 years, when compared with diabetic patients with less tight control (< 180/105 mm Hg), those with tight blood pressure control (< 150/85 mm Hg) had a 24% reduction in sudden death, hyperglycemic or hypoglycemic death, fatal or nonfatal myocardial infarction, angina, heart failure, fatal or nonfatal stroke, renal failure, amputation, vitreous hemorrhage, and retinal hemorrhage.

Finally, renal function deteriorates more rapidly when blood pressure control is inadequate in patients with chronic renal disease of diverse causes.²⁰ Over 2 years, when compared with patients with renal failure who had less tight control (mean arterial pressure 107 mm Hg), renal failure patients with tight control (mean arterial blood pressure 92 mm Hg) had significantly less proteinuria and lower rates of decline in renal function. Whether this translates into a significant improvement in the risk of end-stage renal disease is unknown.

Hypertension is arbitrarily defined as a diastolic blood pressure (DBP) of 90 mm Hg or higher, a systolic blood pressure (SBP) equal to or higher than 140 mm Hg, or both, on 3 separate occasions. It affects 24% of the population of the United States and is common among black (28%), white (24%), and Hispanic (14%) Americans. The prevalence of hypertension increases with age and is more than 70% among people 65 years and older. Among principal diagnoses given by family physicians for outpatient visits, only acute respiratory tract infection (7%) is more common than hypertension (6%). The annual direct medical cost of caring for hypertension exceeds $10 billion.

This article will discuss the pathophysiology and diagnosis of hypertension from an evidence-based perspective. An upcoming Applied Evidence article will cover treatment of hypertension and prognosis.

Pathophysiology

Idiopathic, or essential, hypertension accounts for more than 95% of cases and appears to be caused by a complex interaction between genetic predisposition and environmental factors. The predisposition to essential hypertension is polygenic in origin and may find full expression when combined with environmental factors, such as obesity, low physical activity levels, high stress levels, high alcohol consumption, high dietary sodium, and low dietary potassium, calcium, and magnesium. The complex interaction of genetics and environment may affect sodium, catecholamines, the renin-angiotensin system, insulin, and cell membrane function, causing elevation of the blood pressure.

The more common identifiable causes of hypertension include chronic renal disease (2%-5%), renovascular disease—including renal artery atherosclerosis and fibromuscular dysplasia—(0.2%-0.7%), Cushing syndrome (0.1%-0.6%), pheochromocytoma (0.04%-0.1%), and primary hyperaldosteronism (0.01%-0.3%). Although obesity, excessive alcohol consumption, oral contraceptive therapy, and sleep apnea may cause hypertension, they are not typically included as identifiable causes of hypertension. The prevalence of the latter conditions as identifiable causes of hypertension remains to be defined.

Diagnosis

The presence of hypertension must be confirmed by blood pressure measurements obtained with proper technique. The blood pressure of all patients 18 years and older should be measured at each health care visit because of the high prevalence of hypertension. Patients should be encouraged to abstain from nicotine and caffeine for at least 30 minutes before the measurement of the blood pressure. Measurement should be made with a mercury sphygmomanometer or a recently calibrated aneroid device. The bladder of the blood pressure cuff should encircle 80% of the arm. The pressure should be taken after at least 5 minutes of rest with the patient sitting, back supported, and arm bared and supported at heart level. The first sound heard (phase 1) is the SBP, and the last sound heard (phase 5) is the DBP. Two readings separated by 2 minutes should be averaged. Hypertension is present when an accurately measured blood pressure is high on 3 separate occasions.

A major consensus report, the Sixth Report of the Joint National Committee on Prevention Detection, Evaluation, and Treatment of High Blood Pressure,¹¹ designates 6 categories of blood pressure:

• Optimal—SBP less than 120 mm; DBP less than 80 mm
• Normal—SBP less than 130 mm; DBP less than 85 mm
• High normal—SBP is 130 to 139 mm; DBP is 85 to 89 mm
• Stage 1 hypertension—SBP is 140 to 159 mm; DBP is 90 to 99 mm
• Stage 2 hypertension—SBP is 160 to 179 mm; DBP is 100 to 109 mm
• Stage 3 hypertension—SBP is 180 mm or higher; DBP is 110 mm or higher

It is important to note that the recommended diagnostic evaluation is based on a consensus and should not be considered evidence-based. A complete history, physical examination, and limited diagnostic testing (urinalysis, complete blood count, potassium, sodium, fasting glucose, creatinine, total cholesterol, high-density cholesterol, and electrocardiogram) are recommended once the presence of hypertension has been confirmed. This evaluation has 3 purposes:

1. Identify other cardiovascular risk factors. Most patients with hypertension have multiple cardiovascular risk factors at the time of initial evaluation. Risk factors include smoking, hyperlipidemia, diabetes, age older than 60 years, sex (men or postmenopausal women), and family history of cardiovascular disease in a female relative before age 65 years or a male relative before 55 years.
2. Identify end-organ damage. Evidence of end-organ damage includes left ventricular hypertrophy, angina, previous myocardial infarction, previous angioplasty or coronary revascularization, heart failure, stroke or transient ischemic attack, nephropathy, peripheral arterial disease, and retinopathy.
3. Identify secondary causes of hypertension. Estimating the pretest probability of a secondary (identifiable) cause of hypertension is problematic, because referral bias is a major problem in hypertension prevalence studies; patients are typically included in these studies only after being referred to a study center by their primary care physician for resistant or difficult to control hypertension. On the basis of the best available estimates, it would be reasonable to assume that patients presenting to primary care physicians have a 5% probability of an identifiable cause of hypertension.

The diagnostic evaluation serves to identify other cardiovascular risk factors and end-organ damage in patients with high-normal blood pressure or hypertension. This information is used to identify 3 risk groups: Group A includes patients with no other cardiovascular risk factor, cardiovascular disease, or evidence of end-organ damage; group B includes patients who do not have cardiovascular disease or end-organ damage but have 1 or more of the major risk factors other than diabetes mellitus; and group C includes patients who have cardiovascular disease, other end-organ damage, or diabetes mellitus. The risk associated with hypertension and the intensity of recommended treatment increases progressively as a person moves from risk group A through risk group C.

The diagnostic evaluation may also reveal patients who are more likely to have an identifiable cause of hypertension. The probability of an identifiable cause of hypertension is increased by the onset of hypertension outside the normal age for essential hypertension (30-55 years), sudden onset or worsening of hypertension, stage 3 hypertension, and blood pressure that responds poorly to treatment. Elevated creatinine levels suggest hypertension caused by renal parenchymal disease. Abdominal or flank bruits, hypokalemia, or a significant rise in the serum creatinine level after an angiotensin-converting enzyme inhibitor is started suggests renovascular hypertension. Osteoporosis, truncal obesity, moon face, purple striae, muscle weakness, easy bruising, hirsutism, hyperglycemia, hypokalemia, and hyperlipidemia suggest Cushing syndrome. Labile hypertension, orthostatic hypotension, headache, palpitations, pallor, and diaphoresis suggests pheochromocytoma. Isolated hypokalemia may be caused by hyperaldosteronism.

Unfortunately, the accuracy of the history, physical examination, and preliminary diagnostic testing for patients presenting with hypertension has not been adequately studied. Therefore, estimating the pretest probability of a secondary cause of hypertension in a patient with specific clinical characteristics must be considered crude at best. The best available evidence is shown in Table 1.

Diagnostic Strategy

Patients whose initial history, physical, and laboratory evaluation suggest the possibility of a secondary cause of hypertension should undergo additional testing. The search for the secondary cause of hypertension should focus on chronic renal disease, renovascular hypertension, pheochromocytoma, Cushing syndrome, and primary aldosteronism, depending on the clinical scenario Table 2.

Chronic renal disease will be evident from the blood urea nitrogen, creatinine, and the urinalysis results. The diagnostic approach to other causes of hypertension is more complicated.

Although renal artery stenosis is suggested by the presence of an abdominal or flank bruit, it is an insensitive test (sensitivity=65%; specificity=90%). It is useful when positive (positive likelihood ratio=6.5) but does not rule out renal artery stenosis when negative (negative likelihood ratio=0.4). A clinical decision rule has been developed and validated that integrates several findings from the history and physical examination. Software to implement this decision rule in clinical practice, using Palm or PocketPC hand-held computers, is available at no charge from the JFP Web site at PC download.

Duplex sonography is very accurate (sensitivity=98%; specificity=98%) when the study is adequate but is often nondiagnostic in obese patients. For these patients, magnetic resonance angiography (MRA) is better (sensitivity=93%; specificity=95%). Captopril renal scanning (CRS) is less sensitive and less specific than either sonography or MRA. Renal artery stenosis is confirmed by the highly accurate but more invasive reference standard test of conventional angiography.

Pheochromocytoma is rare even in the presence of suggestive symptoms (headache, palpitations, and excessive and inappropriate perspiration), but failure to identify this disease can have disastrous consequences. Therefore, patients who have suggestive signs and symptoms should be screened for pheochromocytoma. However, the standard for screening pheochromocytoma remains controversial. A 24-hour urinary metanephrine (cutoff point of >3.70 nmol/day ) is highly sensitive and specific when done well, but urine collection is inconvenient and may be incomplete. Plasma metanephrines (metanephrine >0.66 nmol/L or normetanephrine >0.30 nmol/L) are easy to obtain, 100% sensitive, and may represent a good screening test for pheochromocytoma. Because they have limited specificity (85%), a positive plasma metanephrine should be confirmed by the 24-hour urinary metanephrine-to-creatinine ratio (cutoff point of >0.354; specificity=98%) before proceeding to anatomical localization of the tumor.

Two imaging studies are commonly used to localize pheochromocytomas. Metaiodobenzylguanidine (MIBG) scintigraphy is more specific but less sensitive than computed tomography (CT). Relying on CT to guide surgery is less likely to miss tumors than MIBG scintigraphy (CT sensitivity = 100% vs MIBG=88%) but is more likely to result in unnecessary surgery because of the lower specificity (CT specificity = 50%; MIBG=89%).

The 24-hour urinary free cortisol (cutoff point >90 mg/day; sensitivity=100%; specificity=98%) is a useful screening test for Cushing syndrome. It is very sensitive, but false-positives may be seen in patients with depression and polycystic ovarian syndrome. The single-dose (1 mg) overnight dexamethasone suppression test is equally sensitive but is a little less specific than the 24-hour urinary cortisol. However, this test is relatively simple for patients. The patient takes 1 mg of dexamethasone at midnight, and the plasma cortisol level is drawn in the morning (cutoff point >100 nmol). The combined dexamethasone and corticotropin-releasing hormone (CRH) suppression test, which has both a sensitivity and a specificity of almost 100%, can be used to confirm the diagnosis of Cushing syndrome. However, it is a little more complicated for the patient. The patient takes 0.5 mg of dexamethasone at noon on the first day and repeats this dose every 6 hours for a total of 8 doses (ending at 6 am on the third day). Two hours after the last dose the patient is given an intravenous bolus of CRH (1 μg/kg), and 15 minutes later a plasma cortisol is drawn. A cortisol level greater than 38 nmol is the cutoff point for this test.

The coexistence of hypertension and spontaneous or diuretic-induced hypokalemia is strongly suggestive of primary aldosteronism. However, it is important to remember that many (if not most) patients with primary aldosteronism do not have hypokalemia. In the past, screening for primary aldosteronism was accomplished by measuring urinary aldosterone levels after oral or intravenous salt loading. The sensitivity of these tests is 90% to 95%, and they carry a risk of precipitous elevation of blood pressure due to volume expansion or hypokalemia. Measuring the plasma renin and aldosterone levels can be used to test for hyperaldosteronism. Various cut points and ratios have been suggested, but the plasma aldosterone-to-renin ratio (cutoff point >25) is currently the most useful screening test for hyperaldosteronism. For this test the patient is asked to rise at 6 am and remain ambulatory for 2 hours, at which time the plasma aldosterone and renin levels are drawn. Beta-blockers and dihydropyridine calcium channel blockers must be stopped for 2 weeks, and spironolactone and loop diuretics must be stopped for 6 weeks before the test. Primary aldosteronism can be confirmed by the fludrocortisone suppression test.

Acknowledgments

Special thanks to Kathleen Dosh, MS; Greg Tan, MD; and Mark Povich, DO, for help during the initial editing of this paper.

Hypertension is arbitrarily defined as a diastolic blood pressure (DBP) of 90 mm Hg or higher, a systolic blood pressure (SBP) equal to or higher than 140 mm Hg, or both, on 3 separate occasions. It affects 24% of the population of the United States and is common among black (28%), white (24%), and Hispanic (14%) Americans. The prevalence of hypertension increases with age and is more than 70% among people 65 years and older. Among principal diagnoses given by family physicians for outpatient visits, only acute respiratory tract infection (7%) is more common than hypertension (6%). The annual direct medical cost of caring for hypertension exceeds $10 billion.

This article will discuss the pathophysiology and diagnosis of hypertension from an evidence-based perspective. An upcoming Applied Evidence article will cover treatment of hypertension and prognosis.

Pathophysiology

Idiopathic, or essential, hypertension accounts for more than 95% of cases and appears to be caused by a complex interaction between genetic predisposition and environmental factors. The predisposition to essential hypertension is polygenic in origin and may find full expression when combined with environmental factors, such as obesity, low physical activity levels, high stress levels, high alcohol consumption, high dietary sodium, and low dietary potassium, calcium, and magnesium. The complex interaction of genetics and environment may affect sodium, catecholamines, the renin-angiotensin system, insulin, and cell membrane function, causing elevation of the blood pressure.

The more common identifiable causes of hypertension include chronic renal disease (2%-5%), renovascular disease—including renal artery atherosclerosis and fibromuscular dysplasia—(0.2%-0.7%), Cushing syndrome (0.1%-0.6%), pheochromocytoma (0.04%-0.1%), and primary hyperaldosteronism (0.01%-0.3%). Although obesity, excessive alcohol consumption, oral contraceptive therapy, and sleep apnea may cause hypertension, they are not typically included as identifiable causes of hypertension. The prevalence of the latter conditions as identifiable causes of hypertension remains to be defined.

Diagnosis

The presence of hypertension must be confirmed by blood pressure measurements obtained with proper technique. The blood pressure of all patients 18 years and older should be measured at each health care visit because of the high prevalence of hypertension. Patients should be encouraged to abstain from nicotine and caffeine for at least 30 minutes before the measurement of the blood pressure. Measurement should be made with a mercury sphygmomanometer or a recently calibrated aneroid device. The bladder of the blood pressure cuff should encircle 80% of the arm. The pressure should be taken after at least 5 minutes of rest with the patient sitting, back supported, and arm bared and supported at heart level. The first sound heard (phase 1) is the SBP, and the last sound heard (phase 5) is the DBP. Two readings separated by 2 minutes should be averaged. Hypertension is present when an accurately measured blood pressure is high on 3 separate occasions.

A major consensus report, the Sixth Report of the Joint National Committee on Prevention Detection, Evaluation, and Treatment of High Blood Pressure,¹¹ designates 6 categories of blood pressure:

• Optimal—SBP less than 120 mm; DBP less than 80 mm
• Normal—SBP less than 130 mm; DBP less than 85 mm
• High normal—SBP is 130 to 139 mm; DBP is 85 to 89 mm
• Stage 1 hypertension—SBP is 140 to 159 mm; DBP is 90 to 99 mm
• Stage 2 hypertension—SBP is 160 to 179 mm; DBP is 100 to 109 mm
• Stage 3 hypertension—SBP is 180 mm or higher; DBP is 110 mm or higher

It is important to note that the recommended diagnostic evaluation is based on a consensus and should not be considered evidence-based. A complete history, physical examination, and limited diagnostic testing (urinalysis, complete blood count, potassium, sodium, fasting glucose, creatinine, total cholesterol, high-density cholesterol, and electrocardiogram) are recommended once the presence of hypertension has been confirmed. This evaluation has 3 purposes:

1. Identify other cardiovascular risk factors. Most patients with hypertension have multiple cardiovascular risk factors at the time of initial evaluation. Risk factors include smoking, hyperlipidemia, diabetes, age older than 60 years, sex (men or postmenopausal women), and family history of cardiovascular disease in a female relative before age 65 years or a male relative before 55 years.
2. Identify end-organ damage. Evidence of end-organ damage includes left ventricular hypertrophy, angina, previous myocardial infarction, previous angioplasty or coronary revascularization, heart failure, stroke or transient ischemic attack, nephropathy, peripheral arterial disease, and retinopathy.
3. Identify secondary causes of hypertension. Estimating the pretest probability of a secondary (identifiable) cause of hypertension is problematic, because referral bias is a major problem in hypertension prevalence studies; patients are typically included in these studies only after being referred to a study center by their primary care physician for resistant or difficult to control hypertension. On the basis of the best available estimates, it would be reasonable to assume that patients presenting to primary care physicians have a 5% probability of an identifiable cause of hypertension.

The diagnostic evaluation serves to identify other cardiovascular risk factors and end-organ damage in patients with high-normal blood pressure or hypertension. This information is used to identify 3 risk groups: Group A includes patients with no other cardiovascular risk factor, cardiovascular disease, or evidence of end-organ damage; group B includes patients who do not have cardiovascular disease or end-organ damage but have 1 or more of the major risk factors other than diabetes mellitus; and group C includes patients who have cardiovascular disease, other end-organ damage, or diabetes mellitus. The risk associated with hypertension and the intensity of recommended treatment increases progressively as a person moves from risk group A through risk group C.

The diagnostic evaluation may also reveal patients who are more likely to have an identifiable cause of hypertension. The probability of an identifiable cause of hypertension is increased by the onset of hypertension outside the normal age for essential hypertension (30-55 years), sudden onset or worsening of hypertension, stage 3 hypertension, and blood pressure that responds poorly to treatment. Elevated creatinine levels suggest hypertension caused by renal parenchymal disease. Abdominal or flank bruits, hypokalemia, or a significant rise in the serum creatinine level after an angiotensin-converting enzyme inhibitor is started suggests renovascular hypertension. Osteoporosis, truncal obesity, moon face, purple striae, muscle weakness, easy bruising, hirsutism, hyperglycemia, hypokalemia, and hyperlipidemia suggest Cushing syndrome. Labile hypertension, orthostatic hypotension, headache, palpitations, pallor, and diaphoresis suggests pheochromocytoma. Isolated hypokalemia may be caused by hyperaldosteronism.

Unfortunately, the accuracy of the history, physical examination, and preliminary diagnostic testing for patients presenting with hypertension has not been adequately studied. Therefore, estimating the pretest probability of a secondary cause of hypertension in a patient with specific clinical characteristics must be considered crude at best. The best available evidence is shown in Table 1.

Diagnostic Strategy

Patients whose initial history, physical, and laboratory evaluation suggest the possibility of a secondary cause of hypertension should undergo additional testing. The search for the secondary cause of hypertension should focus on chronic renal disease, renovascular hypertension, pheochromocytoma, Cushing syndrome, and primary aldosteronism, depending on the clinical scenario Table 2.

Chronic renal disease will be evident from the blood urea nitrogen, creatinine, and the urinalysis results. The diagnostic approach to other causes of hypertension is more complicated.

Although renal artery stenosis is suggested by the presence of an abdominal or flank bruit, it is an insensitive test (sensitivity=65%; specificity=90%). It is useful when positive (positive likelihood ratio=6.5) but does not rule out renal artery stenosis when negative (negative likelihood ratio=0.4). A clinical decision rule has been developed and validated that integrates several findings from the history and physical examination. Software to implement this decision rule in clinical practice, using Palm or PocketPC hand-held computers, is available at no charge from the JFP Web site at PC download.

Duplex sonography is very accurate (sensitivity=98%; specificity=98%) when the study is adequate but is often nondiagnostic in obese patients. For these patients, magnetic resonance angiography (MRA) is better (sensitivity=93%; specificity=95%). Captopril renal scanning (CRS) is less sensitive and less specific than either sonography or MRA. Renal artery stenosis is confirmed by the highly accurate but more invasive reference standard test of conventional angiography.

Pheochromocytoma is rare even in the presence of suggestive symptoms (headache, palpitations, and excessive and inappropriate perspiration), but failure to identify this disease can have disastrous consequences. Therefore, patients who have suggestive signs and symptoms should be screened for pheochromocytoma. However, the standard for screening pheochromocytoma remains controversial. A 24-hour urinary metanephrine (cutoff point of >3.70 nmol/day ) is highly sensitive and specific when done well, but urine collection is inconvenient and may be incomplete. Plasma metanephrines (metanephrine >0.66 nmol/L or normetanephrine >0.30 nmol/L) are easy to obtain, 100% sensitive, and may represent a good screening test for pheochromocytoma. Because they have limited specificity (85%), a positive plasma metanephrine should be confirmed by the 24-hour urinary metanephrine-to-creatinine ratio (cutoff point of >0.354; specificity=98%) before proceeding to anatomical localization of the tumor.

Two imaging studies are commonly used to localize pheochromocytomas. Metaiodobenzylguanidine (MIBG) scintigraphy is more specific but less sensitive than computed tomography (CT). Relying on CT to guide surgery is less likely to miss tumors than MIBG scintigraphy (CT sensitivity = 100% vs MIBG=88%) but is more likely to result in unnecessary surgery because of the lower specificity (CT specificity = 50%; MIBG=89%).

The 24-hour urinary free cortisol (cutoff point >90 mg/day; sensitivity=100%; specificity=98%) is a useful screening test for Cushing syndrome. It is very sensitive, but false-positives may be seen in patients with depression and polycystic ovarian syndrome. The single-dose (1 mg) overnight dexamethasone suppression test is equally sensitive but is a little less specific than the 24-hour urinary cortisol. However, this test is relatively simple for patients. The patient takes 1 mg of dexamethasone at midnight, and the plasma cortisol level is drawn in the morning (cutoff point >100 nmol). The combined dexamethasone and corticotropin-releasing hormone (CRH) suppression test, which has both a sensitivity and a specificity of almost 100%, can be used to confirm the diagnosis of Cushing syndrome. However, it is a little more complicated for the patient. The patient takes 0.5 mg of dexamethasone at noon on the first day and repeats this dose every 6 hours for a total of 8 doses (ending at 6 am on the third day). Two hours after the last dose the patient is given an intravenous bolus of CRH (1 μg/kg), and 15 minutes later a plasma cortisol is drawn. A cortisol level greater than 38 nmol is the cutoff point for this test.

The coexistence of hypertension and spontaneous or diuretic-induced hypokalemia is strongly suggestive of primary aldosteronism. However, it is important to remember that many (if not most) patients with primary aldosteronism do not have hypokalemia. In the past, screening for primary aldosteronism was accomplished by measuring urinary aldosterone levels after oral or intravenous salt loading. The sensitivity of these tests is 90% to 95%, and they carry a risk of precipitous elevation of blood pressure due to volume expansion or hypokalemia. Measuring the plasma renin and aldosterone levels can be used to test for hyperaldosteronism. Various cut points and ratios have been suggested, but the plasma aldosterone-to-renin ratio (cutoff point >25) is currently the most useful screening test for hyperaldosteronism. For this test the patient is asked to rise at 6 am and remain ambulatory for 2 hours, at which time the plasma aldosterone and renin levels are drawn. Beta-blockers and dihydropyridine calcium channel blockers must be stopped for 2 weeks, and spironolactone and loop diuretics must be stopped for 6 weeks before the test. Primary aldosteronism can be confirmed by the fludrocortisone suppression test.

Acknowledgments

Special thanks to Kathleen Dosh, MS; Greg Tan, MD; and Mark Povich, DO, for help during the initial editing of this paper.

Hypertension is arbitrarily defined as a diastolic blood pressure (DBP) of 90 mm Hg or higher, a systolic blood pressure (SBP) equal to or higher than 140 mm Hg, or both, on 3 separate occasions. It affects 24% of the population of the United States and is common among black (28%), white (24%), and Hispanic (14%) Americans. The prevalence of hypertension increases with age and is more than 70% among people 65 years and older. Among principal diagnoses given by family physicians for outpatient visits, only acute respiratory tract infection (7%) is more common than hypertension (6%). The annual direct medical cost of caring for hypertension exceeds $10 billion.

This article will discuss the pathophysiology and diagnosis of hypertension from an evidence-based perspective. An upcoming Applied Evidence article will cover treatment of hypertension and prognosis.

Pathophysiology

Idiopathic, or essential, hypertension accounts for more than 95% of cases and appears to be caused by a complex interaction between genetic predisposition and environmental factors. The predisposition to essential hypertension is polygenic in origin and may find full expression when combined with environmental factors, such as obesity, low physical activity levels, high stress levels, high alcohol consumption, high dietary sodium, and low dietary potassium, calcium, and magnesium. The complex interaction of genetics and environment may affect sodium, catecholamines, the renin-angiotensin system, insulin, and cell membrane function, causing elevation of the blood pressure.

The more common identifiable causes of hypertension include chronic renal disease (2%-5%), renovascular disease—including renal artery atherosclerosis and fibromuscular dysplasia—(0.2%-0.7%), Cushing syndrome (0.1%-0.6%), pheochromocytoma (0.04%-0.1%), and primary hyperaldosteronism (0.01%-0.3%). Although obesity, excessive alcohol consumption, oral contraceptive therapy, and sleep apnea may cause hypertension, they are not typically included as identifiable causes of hypertension. The prevalence of the latter conditions as identifiable causes of hypertension remains to be defined.

Diagnosis

The presence of hypertension must be confirmed by blood pressure measurements obtained with proper technique. The blood pressure of all patients 18 years and older should be measured at each health care visit because of the high prevalence of hypertension. Patients should be encouraged to abstain from nicotine and caffeine for at least 30 minutes before the measurement of the blood pressure. Measurement should be made with a mercury sphygmomanometer or a recently calibrated aneroid device. The bladder of the blood pressure cuff should encircle 80% of the arm. The pressure should be taken after at least 5 minutes of rest with the patient sitting, back supported, and arm bared and supported at heart level. The first sound heard (phase 1) is the SBP, and the last sound heard (phase 5) is the DBP. Two readings separated by 2 minutes should be averaged. Hypertension is present when an accurately measured blood pressure is high on 3 separate occasions.

A major consensus report, the Sixth Report of the Joint National Committee on Prevention Detection, Evaluation, and Treatment of High Blood Pressure,¹¹ designates 6 categories of blood pressure:

• Optimal—SBP less than 120 mm; DBP less than 80 mm
• Normal—SBP less than 130 mm; DBP less than 85 mm
• High normal—SBP is 130 to 139 mm; DBP is 85 to 89 mm
• Stage 1 hypertension—SBP is 140 to 159 mm; DBP is 90 to 99 mm
• Stage 2 hypertension—SBP is 160 to 179 mm; DBP is 100 to 109 mm
• Stage 3 hypertension—SBP is 180 mm or higher; DBP is 110 mm or higher

It is important to note that the recommended diagnostic evaluation is based on a consensus and should not be considered evidence-based. A complete history, physical examination, and limited diagnostic testing (urinalysis, complete blood count, potassium, sodium, fasting glucose, creatinine, total cholesterol, high-density cholesterol, and electrocardiogram) are recommended once the presence of hypertension has been confirmed. This evaluation has 3 purposes:

1. Identify other cardiovascular risk factors. Most patients with hypertension have multiple cardiovascular risk factors at the time of initial evaluation. Risk factors include smoking, hyperlipidemia, diabetes, age older than 60 years, sex (men or postmenopausal women), and family history of cardiovascular disease in a female relative before age 65 years or a male relative before 55 years.
2. Identify end-organ damage. Evidence of end-organ damage includes left ventricular hypertrophy, angina, previous myocardial infarction, previous angioplasty or coronary revascularization, heart failure, stroke or transient ischemic attack, nephropathy, peripheral arterial disease, and retinopathy.
3. Identify secondary causes of hypertension. Estimating the pretest probability of a secondary (identifiable) cause of hypertension is problematic, because referral bias is a major problem in hypertension prevalence studies; patients are typically included in these studies only after being referred to a study center by their primary care physician for resistant or difficult to control hypertension. On the basis of the best available estimates, it would be reasonable to assume that patients presenting to primary care physicians have a 5% probability of an identifiable cause of hypertension.

The diagnostic evaluation serves to identify other cardiovascular risk factors and end-organ damage in patients with high-normal blood pressure or hypertension. This information is used to identify 3 risk groups: Group A includes patients with no other cardiovascular risk factor, cardiovascular disease, or evidence of end-organ damage; group B includes patients who do not have cardiovascular disease or end-organ damage but have 1 or more of the major risk factors other than diabetes mellitus; and group C includes patients who have cardiovascular disease, other end-organ damage, or diabetes mellitus. The risk associated with hypertension and the intensity of recommended treatment increases progressively as a person moves from risk group A through risk group C.

The diagnostic evaluation may also reveal patients who are more likely to have an identifiable cause of hypertension. The probability of an identifiable cause of hypertension is increased by the onset of hypertension outside the normal age for essential hypertension (30-55 years), sudden onset or worsening of hypertension, stage 3 hypertension, and blood pressure that responds poorly to treatment. Elevated creatinine levels suggest hypertension caused by renal parenchymal disease. Abdominal or flank bruits, hypokalemia, or a significant rise in the serum creatinine level after an angiotensin-converting enzyme inhibitor is started suggests renovascular hypertension. Osteoporosis, truncal obesity, moon face, purple striae, muscle weakness, easy bruising, hirsutism, hyperglycemia, hypokalemia, and hyperlipidemia suggest Cushing syndrome. Labile hypertension, orthostatic hypotension, headache, palpitations, pallor, and diaphoresis suggests pheochromocytoma. Isolated hypokalemia may be caused by hyperaldosteronism.

Unfortunately, the accuracy of the history, physical examination, and preliminary diagnostic testing for patients presenting with hypertension has not been adequately studied. Therefore, estimating the pretest probability of a secondary cause of hypertension in a patient with specific clinical characteristics must be considered crude at best. The best available evidence is shown in Table 1.

Diagnostic Strategy

Patients whose initial history, physical, and laboratory evaluation suggest the possibility of a secondary cause of hypertension should undergo additional testing. The search for the secondary cause of hypertension should focus on chronic renal disease, renovascular hypertension, pheochromocytoma, Cushing syndrome, and primary aldosteronism, depending on the clinical scenario Table 2.

Chronic renal disease will be evident from the blood urea nitrogen, creatinine, and the urinalysis results. The diagnostic approach to other causes of hypertension is more complicated.

Although renal artery stenosis is suggested by the presence of an abdominal or flank bruit, it is an insensitive test (sensitivity=65%; specificity=90%). It is useful when positive (positive likelihood ratio=6.5) but does not rule out renal artery stenosis when negative (negative likelihood ratio=0.4). A clinical decision rule has been developed and validated that integrates several findings from the history and physical examination. Software to implement this decision rule in clinical practice, using Palm or PocketPC hand-held computers, is available at no charge from the JFP Web site at PC download.

Duplex sonography is very accurate (sensitivity=98%; specificity=98%) when the study is adequate but is often nondiagnostic in obese patients. For these patients, magnetic resonance angiography (MRA) is better (sensitivity=93%; specificity=95%). Captopril renal scanning (CRS) is less sensitive and less specific than either sonography or MRA. Renal artery stenosis is confirmed by the highly accurate but more invasive reference standard test of conventional angiography.

Pheochromocytoma is rare even in the presence of suggestive symptoms (headache, palpitations, and excessive and inappropriate perspiration), but failure to identify this disease can have disastrous consequences. Therefore, patients who have suggestive signs and symptoms should be screened for pheochromocytoma. However, the standard for screening pheochromocytoma remains controversial. A 24-hour urinary metanephrine (cutoff point of >3.70 nmol/day ) is highly sensitive and specific when done well, but urine collection is inconvenient and may be incomplete. Plasma metanephrines (metanephrine >0.66 nmol/L or normetanephrine >0.30 nmol/L) are easy to obtain, 100% sensitive, and may represent a good screening test for pheochromocytoma. Because they have limited specificity (85%), a positive plasma metanephrine should be confirmed by the 24-hour urinary metanephrine-to-creatinine ratio (cutoff point of >0.354; specificity=98%) before proceeding to anatomical localization of the tumor.

Two imaging studies are commonly used to localize pheochromocytomas. Metaiodobenzylguanidine (MIBG) scintigraphy is more specific but less sensitive than computed tomography (CT). Relying on CT to guide surgery is less likely to miss tumors than MIBG scintigraphy (CT sensitivity = 100% vs MIBG=88%) but is more likely to result in unnecessary surgery because of the lower specificity (CT specificity = 50%; MIBG=89%).

The 24-hour urinary free cortisol (cutoff point >90 mg/day; sensitivity=100%; specificity=98%) is a useful screening test for Cushing syndrome. It is very sensitive, but false-positives may be seen in patients with depression and polycystic ovarian syndrome. The single-dose (1 mg) overnight dexamethasone suppression test is equally sensitive but is a little less specific than the 24-hour urinary cortisol. However, this test is relatively simple for patients. The patient takes 1 mg of dexamethasone at midnight, and the plasma cortisol level is drawn in the morning (cutoff point >100 nmol). The combined dexamethasone and corticotropin-releasing hormone (CRH) suppression test, which has both a sensitivity and a specificity of almost 100%, can be used to confirm the diagnosis of Cushing syndrome. However, it is a little more complicated for the patient. The patient takes 0.5 mg of dexamethasone at noon on the first day and repeats this dose every 6 hours for a total of 8 doses (ending at 6 am on the third day). Two hours after the last dose the patient is given an intravenous bolus of CRH (1 μg/kg), and 15 minutes later a plasma cortisol is drawn. A cortisol level greater than 38 nmol is the cutoff point for this test.

The coexistence of hypertension and spontaneous or diuretic-induced hypokalemia is strongly suggestive of primary aldosteronism. However, it is important to remember that many (if not most) patients with primary aldosteronism do not have hypokalemia. In the past, screening for primary aldosteronism was accomplished by measuring urinary aldosterone levels after oral or intravenous salt loading. The sensitivity of these tests is 90% to 95%, and they carry a risk of precipitous elevation of blood pressure due to volume expansion or hypokalemia. Measuring the plasma renin and aldosterone levels can be used to test for hyperaldosteronism. Various cut points and ratios have been suggested, but the plasma aldosterone-to-renin ratio (cutoff point >25) is currently the most useful screening test for hyperaldosteronism. For this test the patient is asked to rise at 6 am and remain ambulatory for 2 hours, at which time the plasma aldosterone and renin levels are drawn. Beta-blockers and dihydropyridine calcium channel blockers must be stopped for 2 weeks, and spironolactone and loop diuretics must be stopped for 6 weeks before the test. Primary aldosteronism can be confirmed by the fludrocortisone suppression test.

Acknowledgments

Special thanks to Kathleen Dosh, MS; Greg Tan, MD; and Mark Povich, DO, for help during the initial editing of this paper.