Hypertension: JNC 7 and Beyond




Key Points





  • Reducing high blood pressure, an important and the most common cardiovascular risk factor, prevents or delays the onset of many types of cardiovascular disease.



  • Measurement of blood pressure as an independent risk factor for cardiovascular disease can be easily and inexpensively accomplished in the medical office, in the home, or by sophisticated devices.



  • Dietary sodium restriction reduces the long-term risk of cardiovascular disease, but weight loss is the most effective short-term lifestyle modification to lower blood pressure.



  • Drug therapy for hypertension reduces the risk of cardiovascular disease more than placebo or no treatment. Different expert panels do not agree on a universally applicable treatment algorithm, but all agree that a hypertensive drug that improves prognosis can be given to the patient with a compelling indication for that drug.



  • The traditional treatment target for blood pressure is <140/90 mm Hg in uncomplicated hypertensives; <130/80 mm Hg has been recommended for diabetics, people with chronic kidney disease, or those with established heart disease, but the evidence for these targets is controversial.



  • Despite the widespread availability of many effective and inexpensive antihypertensive drugs, hypertension control rates are suboptimal. A concerted effort by the health care provider, patient, and health care delivery system is required to control blood pressure in the long term and to reduce the risk of cardiovascular disease.



Elevated blood pressure (BP) is responsible for more deaths than any other risk factor for cardiovascular disease (CVD), which is already the leading cause of death and disability in the developed world and is expected to become the leading cause of death and disability worldwide by the year 2025. Historically, “hypertension” (BP ≥140/90 mm Hg) was thought to be the major driver of these problems, but BP levels between 120-139/80-89 mm Hg (now called prehypertension) not only are more prevalent but also elevate cardiovascular risk. Compared with many other risk factors for stroke, acute myocardial infarction, and heart failure, hypertension is among the simplest to diagnose, has the widest variety of treatment options, and (particularly in high-risk individuals) is the most cost-effective preventive strategy. Because of its high prevalence (e.g., ∼29% of adults; Fig. 9-1 ) in the United States, hypertension ranks first among the chronic conditions for which Americans visit a health care provider. One of the major reasons for the impressive reduction in age-adjusted stroke mortality (∼62%) and coronary heart disease (CHD) mortality (∼45%) in the United States since 1972 is the widespread acceptance of the need to treat hypertension and our increased ability to reduce BP effectively.




FIGURE 9-1


Age- and gender-specific prevalence of hypertension in the National Health and Nutrition Examination Survey (NHANES) 2005-2006. As in all previous NHANES data sets going back to 1971, hypertension was defined as BP ≥140/90 mm Hg or taking antihypertensive medication. The overall prevalence of hypertension was about 29% (no significant change since NHANES 1999-2000), but the prevalence of the awareness, treatment, and control (to <140/90 mm Hg) of hypertension has improved to 78%, 68%, and 44%, respectively.




Hypertension Guidelines


Because of the global public health impact of hypertension, many countries and organizations have developed guidelines for its diagnosis and treatment. There are major differences in how frequently these are updated, and there is little consensus among them regarding risk stratification, initial therapy, or BP targets ( Table 9-1 ). Nonetheless, these interesting recommendations summarize our current knowledge but come to very different conclusions.



TABLE 9–1

Recommendations of Different Sets of Hypertension Guidelines


































JNC 7, 2003 ASHWG, 2005 NICE, 2006 ESH/ESC, 2007
Risk stratification Based on BP only Based on tests for target organ damage, not BP Based on absolute risk Extensive system based on absolute risk
Default initial therapy Low-dose diuretic Calcium channel blocker or angiotensin-converting enzyme inhibitor (age dependent) Does not matter; most need two drugs anyway
Beta blockers Second-line No comment Fourth-line Not with diuretics in patients with metabolic syndrome
BP targets <130/80 mm Hg for diabetics, chronic kidney disease Lower for diabetics <130/80 mm Hg for diabetics, even lower if renal dysfunction and proteinuria >1 g/day

JNC 7, Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure ; ASHWG, American Society of Hypertension Writing Group ; NICE, National Institute for Health and Clinical Excellence (British National Guidelines) ; ESH/ESC, European Society of Hypertension/European Society of Cardiology.




Pathophysiology of Hypertension


Although several genetic forms of hypertension have been discovered and many secondary causes of hypertension identified, most hypertensive individuals have primary (or essential) hypertension, for which no specific cause can be found. Many neurohormonal systems affect BP, and many of these have been manipulated pharmacologically to assist with its control. Perhaps chief among these are the renally regulated “BP–vascular volume” concept, the renin-angiotensin-aldosterone system (RAAS), and the sympathetic nervous system (SNS). Much recent work has developed the concept of dysregulation of nitric oxide (and oxidative stress) as a potential contributor to CVD, including hypertension. Although animal studies, epidemiologic evidence, small cohort trials, and a few larger randomized clinical trials have implicated nitric oxide, we have few pharmacologic agents specific to this system that are clinically useful in large populations of free-living hypertensive patients. Further research is therefore necessary before the Science Magazine “molecule of the year” in 1998 achieves the status of the RAAS or SNS as a major contributor to elevated BP.




Definition and Classification of Hypertension


From 1974 to 1993, hypertension in the United States was defined only by diastolic BP (≥90 mm Hg) and was classified as mild, moderate, or severe. Since then, burgeoning evidence from both epidemiologic studies and clinical trials has demonstrated that systolic BP is a better predictor of future CVD and renal events than diastolic or pulse pressure is. Therefore, the focus has shifted to systolic BP, particularly because most hypertensives (especially those older than 55 years) have pretreatment systolic BP ≥ 140 mm Hg rather than diastolic BP ≥ 90 mm Hg. Similarly, the classification system has evolved into “stages” of hypertension ( Table 9-2 ), of which JNC 7 recognizes only two: stage 1 (systolic BP 140-159 mm Hg or diastolic BP 90-99 mm Hg) and stage 2 (systolic BP ≥ 160 mm Hg or diastolic BP ≥ 100 mm Hg). This scheme is independent of gender and age, although many authorities have suggested that higher risk individuals (e.g., African Americans, diabetics, kidney and disease patients) should have BP-lowering treatment initiated at threshold BPs lower than 140/90 mm Hg. This suggestion has been carried to its logical extreme by several sets of guidelines that base diagnosis and treatment decisions on the absolute risk of CVD for a given individual and not on any specific cutoff for BP. Proponents of the “polypill” (which contains three antihypertensive agents at moderate doses) have recommended that not only classification of BP but also measurement of BP should be abandoned in favor of BP lowering in demographically defined populations at high risk for CVD. This approach denies the benefits of therapy to those who are likely to suffer target organ damage and worsened hypertension, rather than CVD events, if left untreated.



TABLE 9–2

Some Useful “Cut Points” for Blood Pressure































JNC 7 Threshold Setting
Office Home (self) 24-Hour Ambulatory
Prehypertension 120/80 mm Hg 115/75 mm Hg About 114/72 mm Hg
Target for “high-risk” patients 130/80 mm Hg 130/80 mm Hg Probably 123/77 mm Hg
Stage 1 hypertension (for diagnosis) 140/90 mm Hg 135/85 mm Hg 133/82 mm Hg
Stage 2 hypertension (for diagnosis) 160/100 mm Hg 155/95 mm Hg About 152/93 mm Hg

Modified from JNC 7, the American Heart Association Scientific Statement, and the European Society of Hypertension.


Prehypertension was defined in JNC 7 as BPs between 120-139/80-89 mm Hg, but the term has been largely ignored by other guidelines. Some believe that the term is too pessimistic and deterministic, but in the Framingham cohort, more than 90% of individuals who are not already hypertensive at the age of 55 years become so during the next 25 years of follow-up. Individuals with this level of BP clearly have increased CVD risk compared with normotensives (BP < 120/80 mm Hg). In the most recent population-based survey of American adults, prehypertension (43% prevalence in men, 30% in women) was more common than hypertension (only 30% and 28%, respectively). Because there are many more prehypertensive than hypertensive individuals in most populations, the burden of BP-related illness would soar if this “disease label” was widely accepted. Although lifestyle modifications are generally recommended for these individuals, and pharmacologic therapy with a moderate-dose angiotensin receptor blocker (ARB) was “feasible” (and significantly prevented the transition to frank hypertension), outcomes from large clinical trials are currently lacking to address the question of whether lowering of BP in prehypertensives is effective in preventing CVD. The intent of JNC 7 in introducing the term was to highlight the elevated CVD risk and to motivate affected individuals to pursue strategies to lower their BP, not (as some have suggested) to increase the sales of antihypertensive drugs.




Measurement of Blood Pressure


Although the recommendation to eschew measurement of BP in favor of its lowering has recently been made, terminology first introduced by Nicolai Korotkoff in 1905 is still used in recording of indirectly determined BPs. Systolic BP is recognized when clear and repetitive tapping sounds are heard; diastolic BP is recorded when the sounds disappear. Only when audible sounds are heard down to 0 mm Hg is the “muffling” of sound (Korotkoff phase IV) recorded, between the systolic reading and zero (e.g., 178/72/0 mm Hg).


Techniques of Measuring Blood Pressure


The proper technique of accurate BP measurement is typically taught very early during medical training but seldom followed thereafter. Experience in many clinical trials has shown that retraining and at least yearly certification in BP measurement are often required to obtain meaningful BP data. One reason for including a placebo arm in registration trials of new antihypertensive drugs is to account for the many potential confounders in BP measurement, including observer expectation bias. The current push for “pay-for-performance” (which includes BP as one indicator of quality of care) is likely to lead to an overabundance of “8” as the terminal digit preference of BPs recorded by health care professionals in many office settings to ensure the greatest possible proportion of customers who are below national BP thresholds. Interestingly, current Healthcare Effectiveness Data and Information Set (HEDIS) guidelines have made achievement of BP goals more likely. Before 2007, the lowest recorded BP (systolic and diastolic) obtained simultaneously at an office visit was accepted as “the” BP for that visit. In the most recent revision, however, the lowest recorded systolic and the lowest recorded diastolic (even if obtained many minutes apart) are accepted, biasing the probability of a below-threshold BP higher.


Home Blood Pressure Measurements


Because of these and many similar challenges, more emphasis is being placed on measurements outside the medical office. Home (or self) BP monitoring is the least expensive and most widely applicable to large populations. Many convenient, inexpensive, and relatively accurate machines are now available. Some authorities think that such devices should be provided to every person with elevated BP and that their physicians should be paid for interpreting home BP data, but others are concerned about their widespread use because clinical trials have seldom based their treatment decisions solely on home readings.


Home BP readings are typically lower than measurements taken in the traditional medical environment (by about 12/7 mm Hg on average), even in normotensive subjects. Home readings are better correlated with both the extent of target organ damage and the risk of future CVD events or mortality than are readings taken in the health care provider’s office. Home readings can also be helpful in evaluating symptoms suggestive of hypotension, especially if they are intermittent or infrequent. During treatment, reliable home readings can lower costs by substituting for multiple visits to health care providers.


Current recommendations advocate use of validated oscillometric devices with an appropriately sized cuff around the upper arm. The device should be calibrated against a standard sphygmomanometer (using a Y tube) and the technique of the measurer checked. At least 12 (a week of twice-daily duplicate or triplicate) readings are the minimum on which to base treatment decisions. Home BP monitoring, coupled with remote monitoring and feedback by a health care professional, has improved BP control rates, perhaps by improving medication adherence.


Ambulatory Blood Pressure Monitoring


Automatic recorders are now available that measure BP frequently in a 24-hour period, during a person’s usual daily activities (including sleep). In the United States, devices that measure BP indirectly (i.e., without arterial cannulation) use either an auscultatory or an oscillometric technique. The auscultatory type uses a microphone placed over the artery to detect Korotkoff sounds in the traditional fashion. The oscillometric technique measures biophysical oscillations of the brachial artery, which are compared (by use of a proprietary algorithm) with those observed with a mercury sphygmomanometer; systolic BP is determined directly from the threshold oscillation, mean arterial pressure is estimated, and diastolic BP is calculated. Both types of monitors are lightweight (<450 g), simple to apply and to use, accurate, relatively quiet and tolerable, and powered by two to four small batteries. Data from 80 to 120 measurements of BP and pulse rate (usually every 15 to 20 minutes during waking hours and every 30 minutes during the night) typically are stored in a small microprocessor and then downloaded into a desktop computer, which then edits the readings and prints the report. Much research with the use of ambulatory blood pressure monitoring (ABPM) has suggested that this technique is the most accurate method of measuring BP, correlates most closely with target organ damage, and best predicts future cardiovascular events (even independently of office BP measurements). In the research setting, ABPM readings have therefore become accepted as the “gold standard” of BP measurements. Accordingly, several expert panels have provided both correlations between ABPM results and BP measurements in other settings (see Table 9-2 ) and recommendations for its use ( Box 9-1 ). Although this method of measuring BP is becoming “standard of care” in many settings, its use is limited in the United States by restrictions on reimbursement for the procedure, which currently in the Medicare age group amounts to $60 to $90, only when the result is a new diagnosis of white coat hypertension.



BOX 9-1

Indications for Ambulatory Blood Pressure Monitoring


Diagnosis and Prognosis





  • Evaluation of suspected white coat hypertension



  • Evaluation of refractory or resistant hypertension



  • Evaluation of circadian pattern of blood pressure



Symptoms





  • Evaluation of dizziness, presyncope, and syncope



  • Evaluation of relationship of blood pressure to clinical symptoms



Evaluation of Antihypertensive Agents (research based)





  • Evaluation of trough-to-peak ratios (and determine optimal dosing intervals)



  • Evaluation of antihypertensive efficacy



  • Evaluation of effects of timing of dosing of antihypertensive agents



Modified from Pickering et al and O’Brien et al.


ABPM makes it possible to measure BP routinely during sleep and has reawakened interest in the circadian variation of heart rate and BP. Most normotensives and perhaps 80% of hypertensives have at least a 10% drop in BP during sleep compared with the daytime average. Although there may be some important demographic confounders (African Americans and the elderly have less prominent “dips”), several prospective studies have shown an increased risk of cardiovascular events (and proteinuria in type 1 diabetics) among those with a nocturnal “nondipping” BP or pulse pattern. Several Japanese studies have raised concern that elderly persons with more than a 20% difference between nighttime and daytime average BPs (“excessive dippers”) may suffer unrecognized ischemia in “watershed areas” (of the brain and other organs) during sleep if their BP declines below the autoregulatory threshold.


White Coat and Masked Hypertension


Approximately 10% to 20% of American hypertensives have substantially lower BP measurements outside the health care provider’s office than in it (so-called white coat hypertension). The white coat itself is unlikely to be the only factor that increases BP. Careful studies originally done in Italy (and now corroborated elsewhere) show that BP rises in response to an approaching physician who is not previously known to the subject. The acute elevation in BP apparently is less marked if a nurse approaches the subject, even if the nurse is wearing a white coat. The pathophysiologic and psychological “reasons” for this exaggerated BP response are unclear.


The clinical consequences, prognostic significance, and amenability of white coat hypertension to drug treatment are controversial. One point of view suggests that if a person has an acute rise in BP due to “stress” from an approaching physician, similar elevations in BP are likely whenever any stressful stimulus is encountered. In several convenience samples and population-based studies, people with white coat hypertension had a greater prevalence of subclinical risk factors for CVD, including left ventricular hypertrophy, family history of hypertension and heart disease, hypertriglyceridemia, elevated fasting insulin levels, and lower HDL-cholesterol levels.


A minority view, based on more conservative definitions of the “white coat effect,” proposes that some individuals consistently show a similar and marked elevation in BP in response to the health care environment. Several long-term observational studies have shown a much reduced risk of either target organ damage or major CVD sequelae among people with lower BPs measured either at home or by 24-hour BP monitoring compared with measurements taken in the same person in the physician’s office. A recent meta-analysis of the seven published observational studies showed that the 1550 patients with white coat hypertension had a significantly lower risk of CVD than the 4819 with sustained hypertension and a nonsignificant and only slightly higher risk than the 3827 with hypertension controlled in both the office and at home ( Fig. 9-2 ). An intermediate viewpoint is that white coat hypertension is merely regression to the mean among the subset of patients with high BP variability.




FIGURE 9-2


Comparison of cardiovascular risk in white coat and masked hypertension with sustained and controlled hypertension. Each box corresponds to the 95% confidence limits (also given numerically within each box), and the dark vertical line within each box corresponds to the adjusted relative risk (also given numerically within each box). This meta-analysis summarizes the 912 first cardiovascular events in seven published observational studies involving 11,502 subjects with an average baseline age of 63 ± 6.5 years, during 8.0 years of mean follow-up. All subjects had their blood pressures monitored in the office (where control was <140/90 mm Hg) and outside the office (where control was <135/85 mm Hg). There was no significant difference in prognosis between those whose blood pressures were controlled both at and outside the office (n = 3827, or 33.3% of the population studied) and those whose blood pressures were controlled only at home but not at the office (n = 1550, or 13.5% of the population studied). These observations support the concept that white coat hypertension has a better prognosis than sustained hypertension. The risk of individuals with masked hypertension (n = 1306, or 11.4% of the population studied: controlled blood pressure at the office but not at home) was not significantly different from that of sustained hypertensives in both locales (n = 4819, or 41.9% of the population studied), and both of these groups had a worse prognosis than that of those with controlled hypertension.


Masked hypertension is said to be present when the in-office BPs are significantly lower than those measured in other settings. Originally thought to be more common in young women with large childcare responsibilities, this condition is now found more widely, in perhaps 6% to 12% of the general population. As such patients would normally not be offered drug treatment (as their in-office BPs are, by definition, below treatment thresholds), masked hypertension is associated with a higher prevalence of target organ damage and incidence of CVD events than in true normotensives. In the seven published observational studies, the 1306 patients with this form of hypertension had only a slightly (and nonsignificant) reduced risk of future CVD events compared with the 4819 patients with sustained hypertension. Many have therefore recommended ABPM (or at least home readings) for all people at risk for hypertension as this is the only way to detect masked hypertension.




Blood Pressure and Cardiovascular Risk


Initial estimates of how much elevated BP increases the risk of heart attack, heart failure, stroke, and other CVD events were derived from prospective epidemiologic surveys. Perhaps the most well known of these in the United States is the Framingham Heart Study, in which 5209 healthy men and women were extensively evaluated initially and then observed over time. After a sufficient number of subjects had events, a quantitative estimate could be made of the importance of hypertension in the development of these events, even after adjusting statistically for the presence of other risk factors (e.g., elevated plasma lipid levels or smoking). Data from Framingham and 60 other observational and epidemiologic data bases have been pooled ( Fig. 9-3 ) and clearly show a strong, positive, and continuous relationship between the level of initial BP and the future risk of death from CHD. Within each decade of life, for each BP increase of 20/10 mm Hg, beginning at 115/75 mm Hg, the risk of death from CHD, stroke, or CVD doubles . These data also show that systolic BP is a much better predictor than diastolic BP or even pulse pressure of CVD and CHD outcomes.




FIGURE 9-3


Relationship of risk of death from CHD (on a logarithmic scale) and the initial blood pressures (systolic at left, diastolic at right) measured at the beginning of each decade of life (given at the top of each line, from 40 to 89 years) in nearly 1 million participants in 61 epidemiologic studies. Larger boxes indicate estimates with smaller 95% confidence intervals, which are indicated for smaller boxes by the vertical lines. The total number of fatal CHD deaths was 30,143. The best-fit regression line for each decade of life ignores the point corresponding to the lowest blood pressure (always <115/70 mm Hg), which includes some individuals with very low blood pressures.

(Modified from Prospective Studies Collaborative. Lancet 360:1903, 2002.)


Perhaps more important than epidemiologic and observational studies of large numbers of people that correlate risk of death from heart disease and BP levels many years earlier are the results of prospective, randomized clinical trials that show, separately and in aggregate, that antihypertensive drug therapy reduces the risk of CHD and other CVD events during a ≤6-year time frame. Most impressive are the results of studies comparing placebo or no treatment with antihypertensive drugs (typically diuretics and beta blockers in older studies, although a few studies with angiotensin-converting enzyme [ACE] inhibitors or calcium antagonists exist). Figure 9-4 shows the results of meta-analyses of 32 clinical trials that compared placebo or no treatment with an initial diuretic, beta blocker, calcium antagonist, or ACE inhibitor in the prevention of CVD in hypertensive subjects. These data indicate highly significant benefit of BP-lowering drugs in hypertensive individuals across all types of CVD and all-cause mortality.




FIGURE 9-4


Results of meta-analyses of clinical trials involving placebo or no treatment compared with an initial low-dose diuretic, beta blocker, ACE inhibitor, or calcium antagonist in the prevention of cardiovascular events in hypertensive subjects. The results, all of which are highly significant (with no significant inhomogeneity), change only slightly if trials that gave the randomized agent as add-on therapy are included. The numbers above or below the line for each event are the numbers of subjects with that event or at risk for that event. *Pbo, placebo or no treatment; OR (95% CI), odds ratio (95% confidence interval); CV, cardiovascular; CHD, coronary heart disease; CVD, cardiovascular disease (first CHD event, stroke, or CV death); HF, heart failure.

(Updated from Elliott WJ: Cardiovascular events in clinical trials of antihypertensive drugs vs. placebo/no treatment: a meta-analysis [abstract], J Hypertens 23[Suppl 2]:S273, 2005.)


Perhaps the most direct and persuasive evidence in favor of the link between BP lowering and CVD (especially CHD) events in clinical trials can be shown in one of several meta-regression analyses correlating the difference in achieved systolic BP across randomized arms in large numbers of clinical trials involving hundreds of thousands of subjects with the relative risk for the specific CVD event ( Fig. 9-5 ). With the exception of heart failure, these analyses generally show that a larger difference in achieved systolic BP (rather than the change in BP or initial BP) is associated with a larger difference between the randomized arms in CVD endpoints. For example, in the first such report, the differences in CHD across randomized treatment groups in trials was not at all well explained by the initial BPs of the participants in each trial ( r 2 = 0.02; P = 0.37) but was instead highly significantly correlated with differences across groups in achieved systolic BP ( r 2 = 0.53; P = 0.0005). Such analyses also have been used to claim “benefit beyond BP control” for specific antihypertensive drug classes in preventing specific endpoints (e.g., ACE inhibitors preventing CHD), but this is controversial.




FIGURE 9-5


Meta-regression of the relative risk for CHD ( y -axis, test agent–control) on the achieved difference in systolic blood pressure between randomized groups (control–test agent) in 27 clinical trials involving ACE inhibitors (placebo control in open squares, active control in filled squares). The regression line and its 95% confidence intervals are based on data from 27 older clinical trials. ADVANCE, Action in Diabetes and Vascular disease: preterAx and diamicroN-MR Controlled Evaluation; ALLHAT-D, Antihypertensive and Lipid-Lowering treatment to prevent Heart Attack Trial–Diuretic comparison; ALLHAT-C, Antihypertensive and Lipid-Lowering treatment to prevent Heart Attack Trial–Calcium antagonist comparison; ANBP-2, Australian National Blood Pressure trial 2; CAMELOT, Comparison of AMlodipine vs. Enalapril to Limit Occurrences of Thrombosis; CAPPP, CAptopril Primary Prevention Project; DIABHYCAR, DIABetes, HYpertension, microalbuminuria or proteinuria, Cardiovascular events And Ramipril; EUROPA, EURopean Reduction Of cardiac events with Perindopril in stable coronary Artery disease; HOPE, Heart Outcomes Prevention Evaluation; ONTARGET, ONgoing Telmisartan Alone and in combination with Ramipril Global Endpoint Trial; PART-2, Prevention of Atherosclerosis with Ramipril Trial 2; PEACE, Prevention of Events with Angiotensin-Converting Enzyme inhibition; PROGRESS, Perindopril pROtection aGainst REcurrent Stroke Study; SCAT, Simvastatin/enalapril Coronary Atherosclerosis Trial; STOP-2-C, Swedish Trial in Old Patients with hypertension 2–Calcium antagonist comparison; STOP-2-D, Swedish Trial in Old Patients with hypertension 2–Diuretic or beta blocker comparison; UKPDS, United Kingdom Prospective Diabetes Study.

(Updated from Elliott WJ, Jonsson MC, Black HR: Management of hypertension: is it the pressure or the drug? Circulation 113:2763, 2006.)


These data (both epidemiologic and those derived from clinical trials) have been interpreted in several different ways. Several groups have pointed out that elevated BP is but one (relatively minor) predictor of CVD ; age, for example, is a much more powerful risk factor. The sensitivity and specificity of BP are low; not everyone with elevated BP will eventually have an event, just as, regrettably, not every person with a “normal” BP will be spared. Any strategy that attempts to fix a value above which everyone should receive treatment is unlikely to be successful and cost-effective. These authorities recommend instead that treatment decisions about BP should be based on a person’s absolute risk of CVD, which can be easily estimated by one of several country-specific risk estimators ; the Framingham risk equation is the most widely used of these in the United States and has been recently updated with a risk algorithm for total CVD risk. For these reasons, their opinion about BP can be easily summarized: “Hypertension: Time to move on.” Advocates of the polypill deem the results of BP-lowering trials so compelling that universal treatment for everyone above a gender-specific age with three moderate-dose antihypertensive drugs (and other agents) is recommended, without even measuring BP. These two different approaches lead to attempts to improve the cost-effectiveness of BP treatment. The first approach would provide treatment for individuals above a certain 10-year risk level (and those below it would fund their own medical care for hypertension). The second approach uses only generically available drugs (combined in a single pill) and avoids expensive interactions with health care providers. It will be very interesting to see how JNC 8 deals with these suggestions.




Predictors of Hypertension


In addition to demographic factors, many features of modern American life appear to increase the risk of becoming hypertensive. As shown in Figure 9-1 , older age is probably the most powerful predictor. For those younger than 55 years, men are at greater risk than women, but this reverses after the age of 55 years. African Americans have a greater risk of hypertension as well as more target organ damage when it is diagnosed and a greater burden of heart disease, stroke, and end-stage renal disease, although these have been improving during the last 5 to 10 years. Uncontrolled hypertension appears to be even more common in Mexican Americans than in other racial and ethnic groups, perhaps because of limited access to health care. Although only a few clinical syndromes are clearly heredofamilial, a single BP measurement is about 40% predictive of inherited BP levels; longitudinal measurements increase this to about 55%. Prospective longitudinal studies of the heritability of BP in male medical students show about equal inheritance from the mother or father.


Although lower educational attainment, lower socioeconomic status, lower physical activity scores, smoking, dyslipidemia, inflammation, psychosocial stressors, sleep apnea, and dietary factors (including higher consumption of dietary fats, sodium, alcohol, and calories or lower potassium intake) have all been implicated in many observational studies as significant predictors of hypertension, much emphasis has recently been focused on obesity, particularly in childhood and adolescence. A study of the NHANES 1988-2004 data suggests that 20% to 80% of the increase in hypertension prevalence in adults during this period could be accounted for by the alarming increase in the average body mass index. Many intervention trials have also demonstrated the benefit of a low-sodium diet in preventing hypertension, particularly one that is rich in fruits and vegetables and low in saturated fat. Observational studies have also shown long-term benefit in preventing hypertension (and CVD events) of dietary patterns similar to those in the Dietary Approaches to Stop Hypertension (DASH) cookbooks, which are among the most popular downloads from the website of the National Institutes of Health.


As with all diagnoses that depend on crossing a threshold of a continuous variable (e.g., BP, fasting glucose concentration), probably the most important predictor of hypertension is prehypertension (BP 120-139/80-90 mm Hg), simply because individuals with it are closer to BP ≥ 140/90 mm Hg than are people with normal BPs (<120/80 mm Hg). The TRial Of Prevention of HYpertension (TROPHY) was organized to test, in humans (as had been demonstrated in rodents), whether short-term drug administration could prevent hypertension, even after the drug had been discontinued. After 2 years of candesartan, the risk of hypertension was reduced by 66% compared with placebo (only three people needed treatment to prevent one from progressing); but after 2 years of receiving placebo, the benefit was much reduced (to 15.6%). Although adverse events were not significantly different, serious adverse events (including cardiovascular events: 1 versus 6) were less common in the group given candesartan. These data clearly showed that drug treatment of candesartan was feasible, but whether it is worth the cost is still debatable, particularly as a shorter Danish study showed no benefit after discontinuation of candesartan.


Another important predictor of hypertension is the use of nonsteroidal anti-inflammatory drugs (NSAIDs), particularly among individuals with above-normal BPs. Hypertension incidence was originally thought to be less with the selective cyclooxygenase 2 inhibitors, but increased BP has been seen both in large observational studies and in meta-analyses of clinical trials. Many attribute the increased CVD risk seen with rofecoxib in several clinical trials (against naproxen or placebo) to its higher risk of BP elevation in the long term, but whether the remaining agent in this class shares these adverse effects is the topic of an ongoing, large, multicenter trial.




Evaluation of the Hypertensive Patient


Four key issues must be addressed during the initial evaluation of a person with elevated BP readings:




  • Documenting an accurate diagnosis of hypertension.



  • Stratifying the person’s risk for CVD disease, which involves (1) defining the presence or absence of existing CVD or renal disease or target organ damage related to hypertension and (2) screening for other CVD risk factors that often accompany hypertension to obtain an estimate of global cardiac risk.



  • Assessing whether the person is likely to have an identifiable cause for the elevated BP (secondary hypertension) and should have further diagnostic testing for it.



  • Obtaining information that may be helpful in choosing appropriate therapy.



Although extensive and expensive laboratory studies are rarely necessary in the evaluation of hypertensive patients, the health care provider must be able to recognize when additional studies or consultation with a specialist is appropriate and warranted. Delaying discovery of a potentially curable form of hypertension or failing to properly assess whether target organ damage or comorbidities are present puts the patient at unnecessary risk, delays implementation of specific treatment, prolongs the time to BP control, and increases CVD risk.


Documenting the Diagnosis


BP should be measured under relaxed and controlled conditions after appropriate rest (typically 5 minutes) and taken by someone whose ability to perform the measurement accurately has been certified. Because ∼10% to 15% of Americans cannot properly hear and interpret Korotkoff sounds, it is unlikely that measurements reported by every observer are accurate. Excellent programs are available that can train, validate, and recertify the competency of the person performing BP measurements. Careful attention should be given to proper, standardized technique.


Before the diagnosis of hypertension is made, an individual usually should have elevated BP measurements documented at least twice, at visits separated by a week or more. Each measurement should be an average of two or three readings differing by <5 mm Hg from each other, taken a few minutes apart in the seated or supine position. Patients who exhibit wide fluctuations in BP or who are hypertensive at some evaluations but normotensive at others may need additional measurements, in the office, at home, or by ABPM, to confirm that they are indeed hypertensive. Treatment should generally not be instituted until the diagnosis is clearly proven. In some circumstances, such as when target organ damage is present, treatment may need to be started after a single set of measurements.


Stratifying Risk for Cardiovascular Disease


Before a treatment program directed at lowering of BP is begun, a thorough assessment of the person’s risk for development of CVD is warranted. Little guidance is provided in JNC 7 about this, but many other schemes are available to assist the physician, most of which have been developed for or adapted to European populations. Many of these are semiquantitative, similar to those of the Framingham Heart Study’s risk score, and allow a reasonable estimate of the patient’s 10-year risk of CVD to be calculated. The Sheffield tables, based on the Scottish national health surveys, may be the most interesting of these. These reasonably well validated tables (one for men, another for women) use age and the ratio of total cholesterol to HDL-cholesterol (with 22 choices for men, 18 for women) and simply dichotomize hypertension (defined as ≥140/90 mm Hg or taking treatment), smoking, and diabetes. Perhaps surprisingly, their data indicate that hypertension (whether treated or untreated) affects overall risk of CHD only slightly; when other risk factors are incorporated into the risk equation, the absolute level of blood BP (especially at diagnosis) is nearly irrelevant.


The determination of an individual’s absolute risk has important implications for the selection of antihypertensive agents, the BP target, and the time to goal attainment. Individuals with the highest short-term risk of a stroke or heart attack are those who already have established CVD or renal disease (e.g., history of a recent transient ischemic attack or previous myocardial infarction). These individuals should be treated promptly, intensively, and (in general, although controversy exists ) to a lower BP goal than for uncomplicated hypertensive people. The search for evidence of concomitant CVD or renal disease need not be extensive or expensive, however. Typically, a complete medical history, a directed physical examination, and a few routine laboratory tests (including an electrocardiogram, urinalysis, and serum chemistry panel) are sufficient.


Hypertensive people with target organ damage are also at substantial risk for cardiovascular events. Target organ damage encompasses many subclinical features of the physical examination or laboratory test results indicating that there has already been an alteration of structure or function in the eyes, heart, kidneys, or blood vessels related to hypertension. Although such individuals may not have as yet suffered an irreversible hypertension-related event (e.g., stroke), some are at substantial risk for these sequelae (e.g., those with chronic kidney disease [CKD]), and the presence of target organ damage usually indicates that hypertension has been present for some time. These people also should receive prompt and intensive efforts to lower BP, typically to a lower than usual goal. Risk calculations are useful only for reasonably low-risk individuals. People with prior CVD, diabetes, or CKD are all assumed to have a 10-year risk of CVD >20%.


Other CVD risk factors (tobacco use, family history of premature CVD) are often found in hypertensive people, and central obesity, dyslipidemia, diabetes, and hypertension tend to cluster. Because other risk factors tend to be additive (if not multiplicative) in increasing the probability of CVD events, it is important to screen a newly diagnosed hypertensive person for these other risk factors to more accurately estimate cardiovascular and renal risks.


Even though age is the most important (nonmodifiable) predictor of CVD risk (see Fig. 9-3 ), the treatment scheme and BP goals recommended in JNC 7 are independent of the patient’s age. There are now good data, several meta-analyses, and an important recent placebo-controlled clinical trial showing that older people, even past the age of 80 years, benefit greatly from BP-lowering drug therapy.


Considering Secondary Hypertension


More than 95% of Americans with hypertension have no specific cause of their elevated BPs (i.e., idiopathic, essential, or primary hypertension). There are three reasons to consider the possibility that hypertension in a newly diagnosed patient might have a specific cause. First, BP control is often difficult to achieve in those with secondary causes of hypertension; diagnosis of it early is likely to get BP to goal more quickly. Second, and particularly important in younger people, use of specific modalities to cure the underlying disease will reduce the future burden of treatment (medical care costs, adverse effects of therapy, and quality of life). Last, routine consideration of secondary causes of hypertension when the diagnosis is originally made will ensure that these diagnostic possibilities will be entertained, and the pros and cons of further testing critically evaluated, and that the clinician will not miss a secondary cause when it is present (see later for details).


Guiding Therapy


The more than 120 antihypertensive agents and fixed-dose combinations currently available in the United States differ in BP-lowering efficacy in various situations. It is often helpful to discuss these potential confounders of treatment with the patient in an effort to “individualize” treatment according to the patient’s specific dietary, medical, and personal considerations. For example, diuretics and calcium antagonists are more effective than ACE inhibitors and angiotensin II receptor antagonists when dietary sodium is excessive. JNC VI and JNC 7 both recommended treating hypertension and a concomitant illness or condition with a specific antihypertensive drug when that drug has been shown in clinical trials to improve CVD morbidity and mortality (so-called compelling indication for a specific class of antihypertensive drug). Therefore, even though ACE inhibitors were not routinely recommended as initial therapy for uncomplicated hypertensives, if the patient has heart failure of the systolic type, an ACE inhibitor could be prescribed. It would be expected not only to lower the BP but also to provide the impressive benefits seen in many long-term studies in every stage of heart failure. Last, some patients are particularly fearful of specific potential adverse effects of certain antihypertensive drugs (e.g., male sexual dysfunction). If the prescriber knows this information, efforts may be taken to avoid medications with a high incidence of this particular problem.


Medical History


In addition to assessment of the risk of future CVD and renal disease, a careful medication, environmental, and nutritional history should be obtained during the initial evaluation and intermittently thereafter. It is particularly important to ascertain whether the patient is taking any agent (either by prescription or over-the-counter) or other substance that might elevate BP ( Box 9-2 ). Of particular concern are the NSAIDs, which are widely used and available over-the-counter and are sometimes not recognized as “medicines” by many patients. Sympathomimetic amines (once commonly found in weight loss, cold, and allergy preparations) have been associated with an increase in both BP and risk of intracerebral hemorrhage and stroke. Hypertensive people should avoid both NSAIDs and sympathomimetic amines and attempt to obtain relief of pain with acetaminophen and of the symptoms of nasal congestion with antihistamines, if possible. When these modalities are ineffective, short-term use of the usually prescribed drugs may be condoned, but with the recognition that BP control is likely to be suboptimal during and immediately after their consumption.



BOX 9-2

Substances That Can Raise Blood Pressure (Partial List)





  • Nonsteroidal anti-inflammatory drugs (including the newer COX-2 inhibitor celecoxib)



  • Corticosteroids



  • Sympathomimetic amines



  • Oral contraceptive hormones



  • Methylxanthines (including theophylline and caffeine *


    * Short duration (minutes to hours).

    )



  • Cyclosporine



  • Erythropoietin



  • Cocaine



  • Nicotine


    Very short duration (seconds to minutes).




  • Phencyclidine (PCP)



COX-2, cyclooxgygenase 2; PCP, phenylcyclohexyl piperidine (“angel dust”).


Abridged from Elliott WJ: Drug interactions and drugs affecting blood pressure control. J Clin Hypertens (Greenwich) 8:731, 2006.


Oral contraceptive pills containing estrogens and progestins may raise BP in some women, although this is much less of a problem with the lower doses in common use today. If a newly diagnosed hypertensive woman uses these pills, discontinuation for 6 months and observation of the BP may allow a decision to be made about whether the pills are the cause of hypertension. Conjugated estrogens (with or without progesterone), typically given for postmenopausal hormone replacement therapy, seldom raise BP, although they have been associated with a wide variety of other problems, including increased rates of CVD events.


Other prescription drugs can either elevate BP or interfere with certain antihypertensive agents. Of the former, cyclosporine, erythropoietin, corticosteroids, cocaine, and theophylline are perhaps the most widely recognized. Of the latter, monoamine oxidase inhibitors, NSAIDs, and tricyclic antidepressants are the most common. It is important to ascertain whether a hypertensive patient has taken any of these agents as well as several other illicit drugs (e.g., phencyclidine). Some chemical elements, particularly lead and chromium, may elevate BP long after exposure; questioning about these and other environmental toxins may sometimes be helpful.


A focused dietary history is important because the most effective lifestyle modifications involve limiting either calories or sodium or both. Dietary salt and saturated fat intake can be estimated from an informal survey of dietary habits and preferences. Many processed foods, “fast foods,” “diet foods,” condiments, and snack items are concentrated, often-unrecognized sources of salt. Now that most foodstuffs bear labels attesting to their high sodium content, many patients are more easily able to choose healthier foods. A sensible target (now validated in several clinical trials, notably the DASH-Sodium substudy and PREMIER ) is 100 mEq (2.4 g or 2400 mg) of sodium per day; this can usually be achieved if the high-salt items mentioned before are avoided and the patient does not add salt either at the table or in cooking. On occasion, it is useful (and relatively inexpensive compared with formal dietary counseling) to have the patient collect a 24-hour urine sample for sodium, particularly when the patient claims to be avoiding salt but the physician is doubtful. Although not all hypertensive patients will experience a reduction in BP on a low-salt diet or an increase in BP on a high-salt diet, individuals who are “salt sensitive” (∼60% of the U.S. population) will benefit from reducing dietary sodium. In general, African Americans and elderly, obese, and diabetic patients are more likely to be salt sensitive, with BPs that are more responsive to dietary salt restriction.


The nutritional history should also include questions about saturated fat consumption, dairy intake, and whether any mineral or vitamin supplements are being used. Because obesity is a major problem for many hypertensive patients, the calorie intake, eating pattern, and changes in weight should be included. Weight loss remains the most successful of all short-term lifestyle modifications for hypertension and should be a part of the therapeutic plan in all overweight hypertensives from the outset.


Social History


Although alcohol in moderation (one drink for women and two drinks for men maximum, with each drink being 12 ounces of beer, 4 ounces of wine, or 1 ounce of spirits) appears to protect against CHD in hypertensives, excessive alcohol intake (four or more drinks per day) raises both BP and all-cause mortality. In some patients, reducing or stopping alcohol ingestion can have salutary effects on BP. A clinical trial in veterans who consumed approximately six drinks per day when enrolled was unsuccessful in demonstrating a significant reduction in BP in those receiving a cognitive-behavioral alcohol reduction program, despite consuming significantly fewer drinks per day (by 1.3, on average) than the control group, who received a much less intensive educational program. Nonetheless, a meta-analysis suggests that there is a place for alcohol restriction in the nonpharmacologic therapy for hypertension.


Large populations of smokers have, on average, lower BP than nonsmokers do, probably because smokers tend to be less obese than nonsmokers. Consuming tobacco has both acute and chronic adverse effects on BP and hypertensive patients. Smoking a single cigarette raises BP and heart rate acutely (within seconds to minutes) because of nicotine’s stimulation of catecholamine secretion. This effect disappears in about 15 minutes, so BP should be measured at least 15 to 30 minutes after the most recent cigarette is extinguished. Chronic tobacco abuse roughly doubles the long-term risk of CVD and has an even larger effect on peripheral arterial disease (including renovascular hypertension). Inquiry about tobacco abuse and advice to discontinue it (if present) should be a part of every encounter with a health care professional. Hypertensive patients should also be questioned about a sedentary lifestyle and whether there is willingness or ability to engage in regular physical activity. Even limited aerobic exercise, including brisk walking for 30 minutes on most days, can reduce BP and the risk of all-cause and CVD mortality. Snoring, daytime sleepiness, and other clinical features of obstructive sleep apnea, especially in obese hypertensives, should lead to a Berlin Questionnaire and consideration of a formal evaluation for this underappreciated and underdiagnosed form of secondary hypertension.


Physical Examination


The directed physical examination of the hypertensive patient should pay special attention to weight, target organ damage, and features consistent with secondary hypertension. It should focus on items that were suggested by the medical history.




  • The pattern of fat distribution should be noted. Android obesity (waist-to-hip ratio >0.95) is associated with increased CVD risk, whereas gynecoid obesity (waist-to-hip ratio <0.85) is not. Men whose waist is ≥102 cm (40 inches) and women whose waist is ≥88 cm (35 inches) are at increased risk. The International Diabetes Federation recommends lower waist circumference cut points for other populations, such as European whites and Asians, to define abdominal obesity.



  • The patient’s skin should be carefully examined for café au lait spots (suggesting neurofibromatosis and possible pheochromocytoma), acanthosis nigricans (suggesting insulin resistance), xanthomas at tendons, or xanthelasma (indicating dyslipidemia). Other skin signs suggesting pheochromocytoma (axillary freckles, ash-leaf patches, port-wine stains in the trigeminal distribution, and adenoma sebaceum) are uncommon, except in patients with phakomatoses.



  • The many physical signs associated with other secondary causes should be sought, particularly if they are suggested by the medical history. The signs of Cushing syndrome (purple striae, moon facies, dorsocervical fat pad, atrophic skin changes) or thyroid disease (abnormal Achilles reflexes, hair quality, and eye signs) are typically difficult to ignore.



  • The funduscopic examination is important in assessing the duration and severity of hypertension. The presence of hypertensive retinopathy (grade 1: arterial tortuosity, silver wiring; grade 2: arteriovenous crossing changes [“nicking”]; grade 3: hemorrhages or exudates; grade 4: papilledema) provides definitive evidence of target organ damage.



  • The neck should be examined for an enlarged thyroid gland, abnormalities of the venous circulation (e.g., jugular venous distention, abnormal or cannon a waves), and carotid bruits.



  • The chest should be auscultated for evidence of heart failure or bronchospasm; bronchospasm would likely make beta blockers contraindicated.



  • The heart should be examined carefully for cardiomegaly, murmurs, and extra sounds.



  • The abdominal examination is one of the most important parts of the directed physical examination because the finding of an abdominal bruit is one of the most cost-effective ways to screen for renovascular hypertension. All four abdominal quadrants and the back should be auscultated, typically with use of the pulse at the wrist as the synchronizing stimulus. Diastolic or continuous bruits are common in renovascular hypertension, but systolic bruits in young and especially thin hypertensive subjects may not be indicative of renal artery stenosis. Abdominal masses can sometimes be palpated in patients with pheochromocytoma or polycystic kidney disease.



  • The groin and legs should be examined for evidence of peripheral arterial disease, which often is manifested as bruits, absent or decreased pulses, and abnormal hair growth patterns. Edema can be a sign of heart failure, renal disease, or high doses of dihydropyridine calcium antagonists.



  • The neurologic examination need not be extensive in a hypertensive patient with no history of cerebrovascular disease, but it should be complete if a history of stroke or transient ischemic attack is present.



Laboratory Testing


For most hypertensives, only a few simple and inexpensive laboratory tests are needed initially. In selected patients, however, more extensive testing is not only appropriate but also necessary to diagnose secondary hypertension and to avoid delaying proper treatment. The laboratory tests that are recommended for all hypertensive persons are shown in Box 9-3 and can be divided into those that are done to assess risk, to establish etiology, to screen for important common diseases, and, finally, to guide the choice of initial therapy.


Jul 10, 2019 | Posted by in CARDIOLOGY | Comments Off on Hypertension: JNC 7 and Beyond

Full access? Get Clinical Tree

Get Clinical Tree app for offline access