Approximately 76 million individuals in the United States and more than 1 billion worldwide are affected by hypertension, and these numbers are increasing with the passage of time. There is much epidemiologic evidence that links age with the development of hypertension. The National Health and Nutrition Examination Survey (NHANES III),1 which provides a cross-sectional examination of the adult population in the United States, shows an inexorable increase in systolic blood pressure (SBP) with aging; about 26% of untreated hypertensive patients are between the ages of 18 and 49 years and 74% are 50 years and older.2 These trends of age-related increases in hypertension occur in both sexes and various ethnic groups, suggesting that changes in blood pressure (BP) occur as a pathophysiologic manifestation of the aging process and the result of both lifestyle factors and genetic predisposition.
J Clin Hypertens (Greenwich). 2012;14:779–786. ©2012 Wiley Periodicals, Inc.
Once considered an inconsequential part of the aging process, the development of isolated systolic hypertension represents a late manifestation of increased elastic artery stiffness and is the predominant hypertensive subtype in the middle-aged and elderly populations. Its inherent increased risk for vascular events, such as coronary heart disease, stroke, heart failure, peripheral artery disease, chronic kidney disease, and dementia, highlights the importance of its control. The purpose of this short review is to summarize how hypertension is different in the elderly when compared with “essential hypertension” in younger adults. The emphasis will be on the multiple ways that increased artery stiffness affects the natural history and clinical manifestations of hypertension in the elderly.
Greater Prevalence of Isolated Systolic Hypertension in the Elderly
Until recently, SBP was the underappreciated component of BP. Early treatment guidelines emphasized control of diastolic BP (DBP). The reasoning behind the recommendation that DBP made a greater contribution to cardiovascular risk than SBP stemmed from the widespread belief that SBP was a normal and inconsequential part of the aging process, obliterating the need to treat all but severe elevations in systolic hypertension. Isolated systolic hypertension (ISH) is defined by an SBP ≥140 mm Hg and DBP <90 mm Hg. In the United States, approximately 80% of patients 50 years or older with untreated hypertension have the ISH subtype (Figure 1).3 Projecting the estimate to the 2020 census, there will not only be an overall increase in the number of hypertensive patients nationally, but also an increase in number of older hypertensives: an increase from 74% to 80% will be 50 years or older with ISH. Furthermore, a recent Framingham Heart Study showed that normotensive persons reaching age 65 had a 90% lifetime risk of developing hypertension—almost exclusively of the ISH subtype—if they lived another 20 to 25 years.4
De Novo Origins of ISH in the Elderly
As suggested by their age-dependent divergent patterns of onset, diastolic hypertension and ISH may be two distinct disorders with significant overlap. The conversion from diastolic hypertension to ISH in the older age group has been termed “burned-out” diastolic hypertension. While some people who have had untreated or poorly treated diastolic hypertension at a younger age develop ISH as they become older, data from the Framingham Study suggest that only about 40% of patients acquire ISH in this manner.5 Surprisingly, 6 of 10 people who developed ISH did so without going through a stage of elevated DBP, but developed de novo ISH from normal and high-normal SBP.5 An incomplete list of de novo causes of ISH include: (1) impaired synthesis of elastin as observed in intrauterine fetal growth retardation6 and successfully repaired coarctation of the aorta;7 (2) the presence of advanced glycation end products that accompanies poorly controlled type 1 diabetes8 and occasionally type 2 diabetes; and (3) increased aortic calcification as found in chronic kidney disease,9 osteoporosis,10 and advanced aging.11
Elastic Artery Stiffness Exceeds Arteriolar Resistance in the Elderly With ISH
Pathophysiology of Hypertension in the Elderly
Aging affects BP hemodynamics. The Framingham Heart Study12 tracked BP in patients aged 30 to 84 who did not take antihypertensive drug therapy or have cardiovascular disease. The almost parallel rise in SBP, DBP, and mean arterial pressure (MAP) up to age 50 to 55 could best be explained by an increase in peripheral vascular resistance. By the time a person reaches age 55, however, the heart has contracted about 2 billion times and the elastic protein in the central conduit vessels—the thoracic aorta and its branches—begins to show wear and tear. Now the heart has to pump against the increased stiffness in these large conduit arteries. After 60 years of age, DBP declines, pulse pressure rises steeply, and MAP values level off, while SBP continues to show a linear increase in most individuals throughout the geriatric years. The BP pattern from age 50 to 55 onward is best explained by a predominance of large artery stiffness.12 With age-related stiffening of the aorta, there is a decreased elasticity and a greater peripheral runoff of stroke volume during systole. With less blood remaining in the aorta at the beginning of diastole, and with diminished elastic recoil, DBP decreases and the diastolic decay curve becomes steeper. The age-related changes in pulse pressure suggest an interaction between vascular aging and hypertension. In fact, hypertension left untreated at an early age can accelerate the rate of vascular aging by as many as 15 to 20 years.12 Moreover, in the presence of diseases that accelerate arterial stiffness, such as diabetes or chronic kidney disease, ISH can develop at an earlier age. In summary, the rate of rise in pulse pressure not only reflects physiological aging, but also various superimposed disease conditions that produce pathologic aging.
Increased Pulse Pressure and Decreased DBP Increase Cardiovascular Risk in the Elderly
Using almost the same Framingham cohort as the previous study, coronary heart disease risk was more related to the pulsatile stress of elastic artery stiffness during systole, as reflected in a rise in pulse pressure, than the steady-state stress of resistance during diastole, as reflected in a parallel rise in SBP and DBP.13 Considerable evidence now favors the superiority of increased pulse pressure and decreased DBP to that of elevated SBP in predicting cardiovascular risk in the elderly. Indeed, increased pulse pressure predicts cardiac complications of left ventricular hypertrophy, atrial fibrillation, systolic/diastolic dysfunction, and heart failure. In addition, increased pulse pressure predicts large artery complications of acute myocardial infarction and thrombotic and hemorrhagic stroke in the elderly.
Ventricular-Vascular Uncoupling in the Elderly
In addition to arterial stiffening, the left ventricle itself becomes stiff, perhaps as an adaptive change to facilitate cardiac ejection and maintain matched coupling of heart to arteries. This is particularly notable in hearts that develop left ventricular hypertrophy, a common occurrence in the elderly and particularly in patients with ISH. A stiffer left ventricle coupled with a stiffer arterial system can contribute to increased cardiovascular risk in several ways, as has been shown by Kass and colleagues.14 There is increased late systolic wall stress and increased cardiac energy costs imposed on the heart to deliver cardiac output to the systemic circulation. Furthermore, the imposition of high late-systolic load that often rises markedly during stress demand slows cardiac relaxation rates, potentially leading to incomplete diastolic relaxation, elevated diastolic filing pressures, and compromised cardiac reserve. This appears to be a frequent pattern in patients with heart failure symptoms who have apparent preservation of left ventricular function. Many of these disturbances in cardiovascular function characterize the elderly person with longstanding ISH and markedly elevated pulse pressure. In brief, diastolic dysfunction and heart failure result from the combination of an elevated cardiac afterload presented to a compromised left ventricle, which is unable to handle the load. Thus, cardiovascular risk is defined by both increased SBP—a marker of cardiac afterload—and concordant decreased DBP, resulting in increased pulse pressure—a marker of vascular-cardiac stiffness and predictor of diastolic dysfunction.14
Decreased Baroreceptor Buffering and Increased BP Variability in the Elderly
Increased large artery stiffness involving the carotid arteries results in decreased baroreceptor sensitivity and eventually increased BP variability—a new cardiovascular risk factor that tracks with increased pulse pressure in elderly patients with ISH.15,16 Indeed, patients with ISH presenting with either white-coat hypertension or white-coat effect are more likely to have an “alerting” or “white-coat” response on the measured BP as a result of stiffened arteries and a concomitant reduction in arterial buffering capacity.17,18 In a recent international population study19 of patients with ISH, BP was measured in the office and by 24-hour ambulatory pressure monitors. Surprisingly, white-coat and masked hypertension comprised 73% of elderly patients with ISH, almost 3 of 4 persons, and would not have been diagnosed accurately with exclusive use of conventional clinic or office BP measurements. Thus, the use of conventional office or clinic BP measurements to identify patients with ISH at risk would have resulted in over-treatment of white-coat hypertension and underdiagnosis and under-treatment of masked hypertension.19 Twenty-four–hour ambulatory BP measurement is the ideal method of diagnosing both masked and white-coat hypertension,20 but other options are available that are less expensive and more easily repeatable for the additional assessment of the response to treatment, including home BP monitoring20 or the use of a repeated automated office BP device with multiple recordings on a single visit.21
Increased Pulsatility to the Microcirculation in the Elderly With ISH
Arteries serve a dual role of transmitting blood to the peripheral tissues and buffering peak SBP pulsations that result from intermittent ventricular contraction. The diminution of this buffering function, reflected in increased pulse pressure, not only adds to the load on the heart and large central arteries, but also increases pulsatility to the periphery arteries.22 When the proximal aortic stiffness reaches and exceeds the stiffness of muscular arteries, wave reflection may reverse and be directed forward in association with increased pulse pressure and increased pulse wave velocity.22 This develops most often in elderly persons with de novo ISH and minimal or no antecedent diastolic hypertension. High-flow organs such as the brain and kidneys are especially impacted with transmission of increased pulsatility to the microcirculation; the resulting remodeling and damage of the small arteries can lead to flow-sensitive ischemia.22 Since elderly persons with ISH have more labile BP, they are prone to repeated episodes of transient ischemia that can damage target organs: in the brain, this results in white matter lesions, cognitive impairment, and eventual dementia; and damage to the kidneys leads to proteinuria and progressive chronic renal disease. Furthermore, excessive transmission of pulsatility contributes to artery damage and inflammation, which can result in atherosclerosis, eventual risk of plaque rupture, and acute coronary syndromes.22
Pseudohypertension in the Elderly With ISH: A Flawed Concept
There is a mistaken belief that systolic pseudohypertension, secondary to brachial artery stiffness, is an occasional cause of pseudo-resistant hypertension in older adults.23 Contrary to this view, the majority of studies in elderly populations show no significant difference (or a slight decrease) in sphygmomanometer-measured brachial systolic BP as compared with simultaneous intra-arterial brachial pressure. In contrast, there is higher diastolic BP of approximately 10 mm Hg in sphygmomanometer-measured brachial artery vs intra-arterial pressure (Figure 2).24 Thus, the true intra-arterial pulse pressure is wider and, therefore, associated with increased cardiovascular risk in older patients with ISH. By the same token, failure to respond to effective antihypertensive therapy, often in association with treatment-induced orthostatic hypotension in the absence of hypertensive target organ damage, should not be labeled systolic pseudohypertension; most likely this represents white-coat hypertension or grade 1 ISH (SBP 140–159 mm Hg) with marked white-coat effect.23
Greater Absolute Cardiovascular Benefit From Therapy in the Elderly With ISH: Results of Randomized Controlled Trials
There has been doubt as to the age-limit beyond which there was no benefit or even increased risk for treating geriatric ISH. In 1991, the landmark double-blinded, placebo-controlled Systolic Hypertension in the Elderly Program (SHEP) study25 first established that older patients with ISH and a mean age of 72 years benefited from antihypertensive lowering of stage 2 ISH (≥160 mm Hg). The Systolic Hypertension in Europe (Syst-Eur)26 and the Systolic Hypertension in China (Syst-China)27 trials corroborated these findings. These results clearly demonstrated that antihypertensive treatment in patients older than 60 years reduced morbidity and mortality. Furthermore, these studies negated prior assumptions that age-related changes in BP are safe and reinforced the emerging paradigm that treatment will benefit patients with elevated systolic pressure, even when they have normal diastolic pressure.
The Hypertension in the Very Elderly Trial (HYVET) study,28 voted the outstanding BP trial of 2008, involved the very old (from 80 to 105 years of age, mean age of 83 years), utilized a randomized, double-blind, placebo-controlled protocol in 3845 patients with sustained SBP ≥160 mm Hg. This study of stage 2 ISH and systolic-diastolic hypertension showed that there was a significant reduction in both fatal and nonfatal strokes (−34%), heart failure (−72%), and reduction in both cardiovascular (−27%) and all-cause mortality (−28%).28 Therefore, there was overwhelming evidence that effective pharmacologic lowering of BP by a mean decrease of 12/4 mm Hg reduced cardiovascular events in this elderly population.28
On the other hand, there are questions that remain unanswered in the HYVET study, which was stopped prematurely at 18 months.28 The majority of patients recruited in HYVET were healthy and robust; would an increasing number of frail elderly give different results? Secondly, what is the optimal BP target goal for maximizing therapeutic benefit in the HYVET study? Importantly, recruitment for intervention trials should use out-of-office rather than clinic BPs in order to exclude patients with white-coat hypertension and to include patients with masked hypertension so as to obtain a more representative population of true cardiovascular disease risk in elderly patients with ISH.
ISH in the Elderly: More Difficult to Control
Analyzing treatment failures revealed an age-related discrepancy in the successful control of BP in NHANES III patients.3 Approximately 50% of younger patients who failed treatment had both SBP and DBP that were not at target goals, representing a concordant failure.3 In stark contrast, older patients who failed to achieve treatment goals had discordant failure: only 17% of patients 50 years and older were above their DBP goal, but 82% were above their SBP target goal and by a significantly greater SBP margin than their younger counterparts.3
While several aspects involving treatment approaches may explain the failure to optimally control ISH overall, including physician inertia and patient nonadherence, one must conclude that there exists a substantial number of patients with systolic hypertension that are truly resistant to currently available medications, even when used properly. That the target SBP apparently becomes more difficult to achieve with aging might be explained by the presence of large artery stiffness, which is not only the result of altered elastic tissue, but also other major changes in the artery wall. McEniery and colleagues11 have shown that aortic calcification, as measured by quantitative high-resolution computed tomographic imaging at the ascending, descending, and abdominal aorta, correlated with aortic stiffness, as measured by carotid-to-femoral pulse wave velocity (after correcting for age and MAP) in patients with ISH who are otherwise apparently healthy.11 Furthermore, the magnitude of aortic calcification correlated with the severity of ISH and to the resistance of SBP control with antihypertensive therapy after adjusting for potential confounding influences.11 This resistance to controlling SBP may be directly the result of using traditional drugs that largely produce vasodilatation rather than decrease arterial stiffness.
Different Therapeutic BP Targets in the Elderly With ISH?
The guidelines for initiating and defining target goal of antihypertensive therapy in the elderly have been updated in a new consensus statement.29 Importantly, to date, there have been no intervention trials involving the elderly that used systolic BP (SBP) target goals of <140 mm Hg or <150 mm Hg, but instead used SBP target goals of <160 mm Hg; surprisingly, national and international guidelines used target goals of <140 mm Hg for all ages, largely on the basis of expert opinion rather than on outcome of randomized controlled trials. The new consensus statement29 confirms that antihypertensive therapy should be started in uncomplicated hypertension in persons aged 65 to 79 years with an SBP ≥140 mm Hg or a diastolic BP (DBP) ≥90 mm Hg with a target goal <140/90 mm Hg (again, based on expert opinion only); however, in persons 80 years and older, the threshold for starting therapy and target goal was raised to an SBP ≥150 mm Hg.29 Unfortunately, there is no agreement on how to define elderly: should it be 60, 70, or 80 years? Should it be based on chronological or physiological age? Are we dealing with vigorous or frail elderly with competing illnesses? This could become a slippery slope in deciding when to change guidelines on the basis of geriatric age vs overall health.
Interpreting the J Curves of Increased Cardiovascular Risk in the Elderly
Controversy persists regarding the presence and significance of BP J curves of increased cardiovascular disease risk as they relate to older people with ISH.30 At present, there are three postulated explanations for the DBP J curve that can occur independently of each other: (1) increased cardiovascular disease risk secondary to increased arterial stiffness, (2) an epiphenomenon related to an underlying chronic debilitating illness and/or cardiac dysfunction (this is called reversed causality), and (3) antihypertensive therapy-induced lowering of DBP, which leads to myocardial ischemia and increased risk for an acute coronary event.
The Framingham Heart Study31 shed some light on DBP J curves that are the result of artery stiffness. Using logistic regression analysis, models containing both SBP and DBP were superior to single BP components in predicting cardiovascular events (coronary heart disease, heart failure, and stroke); risk increased at both the low and high extremes of DBP when combined with elevated SBP, and these findings were independent of antihypertensive therapy and antecedent cardiovascular disease events.31 Furthermore, when cardiovascular risk events in patients with ISH were stratified by DBP <70 mm Hg vs ≥70 to 89 mm Hg to evaluate risk according to the Sixth Report of the Joint National Committee on Prevention, Detection, Evaluation and Treatment of High Blood Pressure Joint National Committee (JNC 6) classification, those individuals with DBP <70 mm Hg had an increased cardiovascular event rate that was equivalent to an increase of 20 mm Hg of SBP (Table).31 In the presence of a DBP J curve, there was not only an increase in coronary heart disease risk, but also a similar increased risk for stroke and heart failure risk, supporting the concept of increased artery stiffness as the cause of this generalized increased risk.31
|JNC 6 Classification||BP Limits, mm Hg||SBP/DBP mm Hg||Odds Ratio (95% CI)|
|Optimal BP||SBP <120 |
|Normal/high-normal BP||SBP 120–139 |
|Stage 1 isolated systolic HTN||SBP 140–159 |
|Stage 1 isolated systolic HTN||SBP 140–159 |
|Stage 2 isolated systolic HTN||SBP ≥160 |
|Stage 2 systolic-diastolic HTN||SBP ≥160 |
|Stage 2 systolic-diastolic HTN||SBP 160–179 |
Importantly, among untreated persons with ISH in the NHANES study,32 there was a 30% prevalence of DBP <70 mm Hg and these individuals were at increased cardiovascular risk compared with those with DBP of 70 mm Hg to 89 mm Hg. Advanced age, female sex, and diabetes mellitus, but not treatment status, were associated with low DBP. Thus, DBP J curves in untreated elderly patients with ISH are common, associated with considerable cardiovascular risk, and most likely representative of increased arterial stiffness.
Reverse causality occurs when a disease event causes a reduction in the primary risk factor (in the present example, BP), making the event rate appear higher than it should be at lower risk levels.33 An example of reverse causation was the high 1-year mortality rate seen in US hemodialysis center patients associated with reduction in SBP, due mainly to failing cardiac function.34 The defining characteristic of reverse causality is the presence of a combination of DBP and SBP J curves.
Finally, in the presence of high-grade stenosis of coronary arteries, increased risk of myocardial infarction with antihypertensive therapy–induced decrease in BP may well occur, as implied in post hoc analysis of intervention trials,35,36 but is probably the least common occurrence of the J-curve phenomenon. Indeed, the risk of plaque disruption that leads to acute coronary syndromes depends more on plaque composition, plaque vulnerability (plaque type), and the degree of pulsatile stress than on the degree of coronary artery stenosis (plaque size).37 Not surprisingly, therefore, the majority of myocardial infarctions (>70%) occur from plaque rupture in coronary arteries that have <50% stenosis.37
In view of the ongoing debate regarding origins of the treatment-induced DBP J curve, Wang and colleagues38 conducted a post hoc meta-analysis to answer the question: to what extent does DBP reduction or achieved DBP contribute to cardiovascular disease outcome in patients with a larger-than-median reduction in SBP while receiving antihypertensive drug therapy? The Individual Data Analysis of Antihypertensive Intervention Trials (INDANA) data set was used to select 11 random control trials totaling 26,000 patients that included systolic-diastolic hypertension in the young (30–49 years), ISH in the old (60–79 years), and ISH in the very old (80 years and older). Importantly, in this meta-analysis of the INDANA database, antihypertensive therapy reduced the risk of stroke and myocardial infarctions in all three age groups to a similar extent.38 Moreover, in patients who had a larger-than-median reduction in SBP, active therapy vs placebo control reduced the cardiovascular disease risk of all outcomes irrespective of the decrease in DBP or the achieved DBP in the young, old, and very old hypertensive patients (Figure 3).38 Therefore, this post hoc meta-analysis showed no evidence of a therapeutic DBP J curve at values <70 mm Hg in the elderly population with ISH.
In summary, a solitary DBP J curve with elevated SBP and pulse pressure suggests increased large artery stiffness in the presence or absence of antihypertensive therapy; risk is defined by increased pulse pressure that results in decreased DBP and increased SBP.30 In contrast, the combination of J curves of SBP and DBP, with normal-to-low pulse pressure and low MAP, suggests reverse causality; risk is defined by decreased DBP and decreased SBP, rather than by increased pulse pressure.29 The presence of a therapeutic DBP J curve is controversial; if present, the frequency is low. Because of the many factors that result in DBP J-curve risks, only a prospective trial with baseline and pre-event BP determinations can establish the presence and frequency of a therapeutic-induced DBP J curve. On the other hand, the optimal therapeutic reduction in SBP and DBP in elderly pateitns with ISH that maximizes benefit is a separate question from the presence of a therapeutic J curve of increased cardiovascular risk.
Is There Less Benefit of Therapy in the Frail Elderly With ISH?
Many previous community-based studies of persons 85 years and older found that survival worsens in association with lower SBP and DBP, an observation that has been attributed to reversed causality.39,40 Unlike the solitary DBP J-shaped curve indicative of increased arterial stiffness, reversed causality explains the combined DBP and SBP J-shaped curves often seen with competing comorbidities, disabilities, or orthostatic hypotension.30 Furthermore, one cannot make a distinction between those cardiovascular events that are naturally occurring vs those that are treatment induced.30 Indeed, frail elderly persons (often institutionalized) may have DBP <60 mm Hg and SBP <120 mm Hg in association with reduced survival and often without abnormal left ventricular function or antihypertensive drug therapy.41
There is overwhelming evidence that middle-aged and elderly persons have the highest prevalence of hypertension, predominantly of the ISH subtype, as compared with their younger counterparts, and this represents a significant burden for future cardiovascular events. Paradoxically, elderly with ISH not only have a greater cardiovascular risk than younger adults, but they also profit more from effective lowering of SBP with antihypertensive therapy. Poor control of ISH in the older population may result in part from the degraded elastic tissue and the deposition of calcium in large conduit arteries. The identification of severity of ISH in the elderly requires measuring out-of-office BP in order not to over-treat white-coat hypertension, nor to overlook the diagnosis of masked hypertension. There is evidence supporting a DBP J curve of increased cardiovascular disease risk in both treated and untreated elderly patients with ISH, but evidence of a specific DBP J curve secondary to therapeutic lowering of DBP is controversial. Interpreting the causes and therapeutic implications of J curves risk in the elderly represents a major clinical challenge. The therapeutic target goal in reducing SBP and DBP depends in part on competing comorbidities, disabilities, and presence of possible orthostatic hypotension. Future randomized controlled trials, based on out-of-office BP monitoring, are necessary to determine how low to go for maximum benefit of antihypertensive therapy.