Blood pressure readings at rest are normally used to define hypertension. During the late 1950s and the 1960s [4, 5], many studies were done to assess the hemodynamic responses to exercise in known hypertensive persons and, separately, in persons who at that time were thought to be at risk for developing hypertension. The results initiated interest in the pathophysiology and natural history of hypertension. It became apparent that functional testing unmasked subtle abnormalities that were not apparent at rest.

The recent development of ambulatory blood pressure monitoring has been useful in evaluating blood pressure in an environment that is separate from the stress-filled environment of the physician's office [6]. However, ambulatory blood pressure recordings are not easily correlated with the various physical activities that the patient engages in throughout the day.

Standardized exercise testing protocols are now generally available and are an established element of the management of coronary heart disease. These protocols are not routinely used for assessing hypertensive patients unless evidence suggests concurrent myocardial ischemia. In controlled circumstances, formal exercise testing shows that work capacity in asymptomatic hypertensive patients is impaired compared with age- and fitness-matched controls [7]. In this review, we examine the origins and nature of this finding.

Despite progress in the understanding of hypertension in terms of diagnosis and assessment of its complications, it is prudent to reexamine the value of exercise testing. Functional testing with modern techniques can give further insights into the pathophysiology of the disorder and may allow a better evaluation of the prognosis of high-risk normotensive persons or those with borderline hypertension. The blood pressure elevation during exercise is better correlated with end-organ damage [8-11] than are casual measurements. Thus, the response to exercise in hypertensive patients may be a more useful end point to assess the efficacy of antihypertensive therapy than is resting blood pressure. In this review we also evaluate the role of exercise testing as an adjunctive investigative tool in hypertension management and drug assessment.

Methods

Top Methods Summary Author & Article Info References

We searched the MEDLINE database to identify all articles about exercise testing in persons with hypertension published between 1985 and February 1995. We also searched the bibliographies of relevant textbooks and articles. We analyzed data on hemodynamic responses of hypertensive patients and persons thought to be at risk for developing hypertension in terms of correlations with end-organ damage, death, and exercise tolerance. Where appropriate, we highlight the methodologic flaws of the selected studies to obtain a better understanding of experimental exercise testing.

Methodologic Considerations

Unlike exercise stress testing in coronary heart disease, no universally accepted guidelines on exercise testing are directed specifically toward assessing hypertensive patients. Few studies have been published on the comparative merits of different exercise-testing protocols. It has even been assumed that otherwise healthy patients with hypertension have normal exercise oxygen kinetics [12]. Most methodologic comparisons have therefore been done on normal persons and on patients with ischemic heart disease or heart failure. Until this vacuum is filled, we must be guided by these research findings.

Types of Stress Testing

The cardiovascular system can be tested using isometric techniques, isotonic techniques, or both. In clinical practice, hand grip has been used as a beside isometric stress test but is of little use for stressing the whole cardiovascular system. Dynamic or isotonic exercise, which is more often used in research and clinical practice, represents muscle contraction with constant tone and therefore muscle shortening. The cycle ergometer and treadmill are the two commonly used tools. These can be calibrated, and delivered workloads may be accurately defined during staged exercise protocols. The two techniques differ in several respects.

Hemodynamic Effects of Body Position

Supine cycle ergometry is associated with a higher heart rate for a given level of work than is upright cycle ergometry. Accordingly, with supine ergometry, patients with coronary heart disease develop angina at a lower double product, and the ST segment is more depressed at any given work level [13]. Upright treadmill exercise produces lower peak systolic blood pressure than does supine cycle ergometry at the same workload. When the treadmill and upright cycle ergometry are compared at similar workloads, the increase in both heart rate and arterial blood pressure are lower with treadmill testing [14], but the maximal oxygen uptake is 6% to 25% higher [15-17]. Upright cycle ergometry therefore places a greater stress on the cardiovascular system but is less sensitive in eliciting a positive diagnostic response when the patient is being tested for ischemic heart disease [13].

Exercise Testing Protocols

The testing protocol defines the incremental or continuous nature of the test, the duration, and the workload associated with each stage. Measured maximal oxygen uptake differs depending on the protocol used. Maximal oxygen uptake is significantly higher when the test lasts between 8 and 17 minutes [15]. When assessed in patients with heart failure, protocols with slow incremental workloads give reproducible stage indices of metabolic work but result in a longer test and underestimation of maximal oxygen uptake [18]. Protocols that last less than 10 minutes [19] and those with a large incremental increase in workloads [18] are associated with greater variability on repeat testing. The "ramp" concept was created to minimize this variability. A constant workload rate (ramp rate) is set such that maximal exercise for each person can be achieved in approximately 10 minutes. This protocol provides excellent correlations between observed and predicted oxygen uptake calculated from workloads increased at a constant rate [16]. This is a useful protocol in cases in which ventilatory gas analysis is not available, but it is inconvenient: A preliminary maximal exercise test (using a "slow" protocol) is necessary to estimate maximal exercise capacity and set the ramp rate. This protocol is therefore not widely used in clinical practice. Whichever protocol is used for research purposes, many studies have shown that patients must be familiarized with the protocol and the demands of the test. The motivation of the supervising staff and patient remain key features in the outcome and duration of a test [20].

Blood Pressure Measurement during Exercise

Many studies have shown that at-rest blood pressure measurements obtained using sphygmomanometry may differ from those obtained using intraarterial recordings. In a study of 35 patients with early hypertension, Lund-Johansen [5] found that when an indirect method of cuff sphygmomanometry was used compared with intra-arterial measurements, at-rest systolic blood pressure was underestimated by a mean of 4.5 mm Hg and at-rest diastolic blood pressure was overestimated by a mean of 5.1 mm Hg. His findings are not unique [21, 22]. Gould and colleagues [23] compared indirect and direct blood pressure in 25 patients with hypertension during exercise with cycle ergometry and found that systolic blood pressure during exercise was underestimated by 15 to 18 mm Hg. The mean difference in diastolic blood pressure during exercise ranged from – 2 to 4 mm Hg.

Diastolic blood pressure is more difficult to determine during exercise. In a separate study of patients with coronary disease [24], diastolic blood pressure could not be measured with a sphygmomanometer during exercise in as many as 12.5% of patients. During exercise, sphygmomanometry is compromised by movement, respiratory effort, and the noise of the equipment. When cuff methods are compared with intra-arterial methods, intrapatient and interpatient variability may increase as exercise progresses [23].

The use of an automated blood pressure measuring system would be more convenient for clinicians supervising an exercise test. Manual and automated methods of exercise blood pressure measurement have also been compared; unfortunately, automated systems tend to be both inaccurate and unreliable [25-28]. Garcia-Gregory and colleagues [25] compared these methods and found that an automated method (the Blood Pressure Measuring System developed by NASA [National Aeronautic and Space Administration]) overestimated systolic blood pressure and underestimated diastolic blood pressure during exercise. At peak exercise, systolic blood pressure was overestimated by 20 mm Hg; in addition, the blood pressure measurements during exercise were inconsistent, with standard errors 2 to 3 times that of the manual method. The utility of Finapres blood pressure estimations (a finger blood-pressure estimation device), which correlate well with intra-arterial monitoring in pressor dose-response studies [29], has not been fully evaluated during exercise.

A calculated mean arterial blood pressure (by convention, one third of the pulse pressure plus diastolic blood pressure) is only valid while the patient is at rest. During exercise, measured mean intra-arterial pressure shifts to the middle of the pulse pressure with an increase in heart rate and changing vascular resistance [30]. Because blood pressure during exercise decreases rapidly as exercise ceases [23], readings obtained with a sphygmomanometer at the end of exercise are not valid. However, submaximal exercise blood pressure obtained at defined stages by cuff sphygmomanometry may be prognostically useful and is discussed later.

Ventilatory Gas Analysis

Measured maximal oxygen uptake [31], staged submaximal oxygen uptake [32], and ventilatory threshold [33, 34] all have good reproducibility; in contrast, exercise times tend to improve with serial testings through a training effect [35]. These indices are ideal for serial measurements in intervention studies. The advent of rapidly responding expired gas analytical methods has greatly facilitated research on exercise responses. Ventilatory gases can now be analyzed breath by breath. However, short sampling intervals (that is, with each breath) are associated with an increased variability [36]. A balance between precision and variability must therefore be struck. As an index of functional capacity, ventilatory gas analysis is probably the most reproducible end point for research studies.

Normal Response to Exercise

The anticipation of exercise increases the heart rate through activation of the sympathetic nervous system. As exercise begins, the heart rate increases rapidly because of a further reduction of vagal tone, followed later by a further increase in sympathetic tone and circulating catecholamine levels. Stroke volume (about 70 mL at rest), increases early and then levels off at approximately one third of the maximal oxygen uptake, or half of the maximal cardiac output. This is due to an increased venous return, an increased end-diastolic volume, and the sympathetic nervous system stimulation of myocardial contractility and active myocardial relaxation.

As exercise progresses in normal persons, the systolic pressure increases by 50 mm Hg to 70 mm Hg, whereas the diastolic blood pressure remains unchanged or decreases by 4 mm Hg to 8 mm Hg as a result of vasodilatation and decreasing total peripheral resistance. Resting cardiac output (5 to 6 L/min) increases to as high as 20 to 25 L/min during peak exercise, an increase proportional to the workload and exercise demand. Because of the linear relation between oxygen uptake and cardiac output during exercise, oxygen consumption reflects cardiac output. The arteriovenous oxygen difference, which plateaus with increased work, is the amount of oxygen extracted by the tissues. This difference is 4 to 5 mL per 100 mL of blood at rest and increases to a maximum of 13 to 16 mL per 100 mL of blood during exercise. The maximal oxygen uptake at peak exercise is therefore a good indicator of "cardiovascular fitness." The ability to exercise after achieving maximal oxygen uptake merely reflects the will to endure and the musculoskeletal state.

Exercise performance may be affected by many variables. Increasing age [37] is associated with a reduction in the increase in heart rate and cardiac output; the stroke volume either remains the same or declines. With increasing age, blood pressure during exercise increases more steeply and the maximal heart rate decreases. Men have a 10% to 14% greater hemoglobin level, greater muscle mass, and less fat than women. These attributes contribute to the increased exercise capacity in men.

Compared with persons who are not obese, obese persons have greater left ventricular mass, stroke volume, cardiac output at rest, and oxygen uptake but have reduced exercise tolerance at similar workloads [38]. Environmental temperature affects exercise performance, and the increase in blood pressure with exercise may be potentiated with decreasing temperature [39]. Each decrease in the degree centigrade is accompanied by a 1.3-mm Hg increase in the at-rest systolic blood pressure and a 0.6-mm Hg increase in the at-rest diastolic blood pressure. This susceptibility to temperature changes increases with age. Blood pressure also has a well-defined circadian rhythm [40], which clearly may affect the response to exercise at different times of the day. However, no studies have compared exercise performance at different times of the day or night. The concept of "physiologic readiness to perform" is absent at night, but this finding has not been well described quantitatively [41].

Food intake [5, 42] increases metabolism, heart rate, cardiac output, and total peripheral resistance, thereby reducing exercise tolerance; however, food intake is rarely specifically controlled for in routine or research studies. Smoking [43], caffeine intake [44], and alcohol consumption [45] before exercise also affect the exercise response. Conversely, exercise training reduces the heart rate and blood pressure and increases stroke volume, left ventricular chamber size, left ventricular mass, and arteriovenous oxygen difference at rest, with a resultant increase in exercise performance [46].

Exercise Responses in Groups at Risk for Developing Hypertension

The following patient groups are considered to have an increased risk for developing sustained arterial hypertension later in life: persons who may be normotensive but have a strong family history of hypertension [47], adults with "borderline" or "labile" hypertension at rest, and children and young adults with blood pressures greater than the upper 95th percentile for age. Few exercise studies have been done on salt-sensitive normotensive persons.

Hyperkinetic Circulation

Studies in children [48, 49] and young adults [5] have focused on patients with a higher cardiac index and heart rate but a normal stroke index and total peripheral resistance at rest in the supine position. This has been defined as a "hyperkinetic circulation," and the changes disappear during exercise. On exercise, the stroke index and cardiac index increases less, the decline in total peripheral resistance attenuates, and the heart rate is higher than that in normal age-matched controls. These findings gave rise to the hypothesis that a hyperkinetic circulation at rest could be implicated in the pathophysiology of hypertension [50]. A hyperkinetic circulation may cause overperfusion of tissues, which is counteracted by vasoconstriction as an autoregulatory mechanism. The consequent increase in total peripheral resistance contributes to the subsequent development of fixed arterial hypertension. However, this hypothesis is challenged by the absence of hyperkinetic variables in the erect posture at rest. The total peripheral resistance is then prominently increased, the cardiac index returns to normal, and the blood pressure remains the same [51]. Furthermore, oxygen consumption has been found to increase normally and parallel to cardiac output, with a normal arteriovenous oxygen difference [52]. In one study, a group of patients with "true hyperkinetic" circulation and an abnormally low arteriovenous oxygen difference at rest and exercise was identified, but no one in this small group developed hypertension during prolonged follow-up [53].

Prognostic Significance of Exaggerated Systolic Blood Pressure Response during Exercise

An important suggested use of exercise testing is for predicting future hypertension in otherwise normal persons. Many relatively small studies have suggested that an exaggerated increase in systolic blood pressure to absolute values of 200 to 230 mm Hg is associated with an increased risk for hypertension later in life [54-56] (Table 1). The less impressive results of Davidoff and colleagues [56] may be attributed to the fact that they measured blood pressure 30 seconds after completion of exercise. As indicated before, the peak systolic blood pressure decreases rapidly after exercise is stopped. This probably led to an underestimation of the number of hypertensive responders.

Benbassat and From [57] examined follow-up studies that lasted as long as 15 years. Normotensive persons with an exaggerated blood pressure response to exercise had a 2.1- to 3.4-fold greater risk for developing hypertension, but 38% to 89% remained normotensive on follow-up. The wide variation in results are partly due to the different definitions, protocols, and durations of follow-up. These results should be interpreted cautiously because many of these patients may have had borderline hypertension or the white coat response (see below). Initial resting blood pressure was measured only once in each patient, without the benefit of comparison with ambulatory blood pressure monitoring. From the data of Wilson and Meyer [54], we can estimate that 61% of persons who were initially classified as being hypertensive with resting blood pressure greater than 140/90 mm Hg were judged to be normotensive on follow-up.

Children (aged 8 to 10 years) with blood pressures greater than the 95th percentile do not show an exaggerated systolic blood pressure response to exercise compared with children who have resting blood pressures less than the 95th percentile [58]. However, in normotensive children whose parent or parents have essential hypertension, blood pressure [59, 60] during exercise was higher or the systolic blood pressure response to exercise was exaggerated [61-63] compared with normotensive children whose parents were normotensive.

Exercise Systolic Blood Pressure and End-Organ Damage

In addition to a possible role in predicting the evolution of hypertension, the response to exercise testing has also been assessed as a way to predict susceptibility to end-organ damage. The efficacy of exercise testing should ideally be addressed in persons with untreated hypertension at presentation or persons at risk for hypertension to determine the subsequent effect on end-organ function. No controlled studies have been done in this area. In a study of 39 normotensive men, Gottdierer and colleagues [64] found that 14 of 22 men (64%) with exaggerated systolic blood pressure response to exercise but only 1 of 17 controls (6%) had left ventricular hypertrophy on echocardiography. In a larger study of 160 men (100 with blood pressure during exercise more than 220 mm Hg), Polonia and colleagues [65] reported similar findings—54% of patients with an exaggerated blood pressure response had left ventricular hypertrophy. The Framingham Heart Study investigators [66] reported that patients with an exaggerated blood pressure response had a 10% greater left ventricular mass than those who did not. This finding became less statistically significant when age, resting blood pressure, and weight were considered. Smith and colleagues [67] found no relation between exaggerated blood pressure response and left ventricular mass; they indexed the latter to height alone even though the hypertensive patients were significantly heavier than the controls. The prediction of other indices of end-organ damage and hypertensive nephropathy, retinopathy, or atherosclerotic vascular disease have not been examined in detail in normotensive or hypertensive persons.

Systolic Blood Pressure during Exercise and Death from Cardiovascular Disease

In a prospective study using cycle ergometry (progressing from 82 W for 2 minutes to 164 W for 6 minutes to 191 W for 2 minutes), Filipovsky and colleagues [11] followed 4907 middle-aged men in the Paris civil service for an average of 17 years. In the initial screening, the researchers excluded persons with a blood pressure of at least 180/105 mm Hg and an electrocardiogram positive for ischemia during exercise. An exaggerated systolic blood pressure after 5 minutes of exercise was associated with increased mortality after controlling for age, smoking, total cholesterol level, body mass index, left ventricular hypertrophy shown on electrocardiography, and athletic activities. The relative risk for death from cardiovascular disease increased with the elevation in systolic blood pressure during exercise. The relative risk was 1.6 in participants whose systolic blood pressure response to exercise was greater than 230 mm Hg. This large study presents strong evidence that an exaggerated response in normotensive persons may be of prognostic value. Unfortunately, this study did not report the resting blood pressure profiles of these persons during follow-up. Thus, the extent to which this excess in mortality is due to a progression to fixed arterial hypertension alone cannot be assessed.

Increased Diastolic Blood Pressure during Exercise

An epidemiologic study of school children [49] has suggested that those with increased diastolic blood pressure during peak exercise have an increased risk for developing left ventricular hypertrophy on follow-up. In adults, an abnormal increase in diastolic blood pressure (> 15 mm Hg) during exercise has been associated with a high incidence of at least one vessel disease at coronary angiography (incidence as high as 83%) [24]. This percentage, however, may be an overestimation. Compared with the control subgroup, a high prevalence of positive ischemic electrocardiographic findings were seen during exercise in the subgroup who had an exaggerated diastolic blood pressure response and then had coronary angiography.

Exercise Responses in Patients with White Coat Hypertension

The white coat response is a transiently elevated blood pressure associated with the stress of the clinical environment. This effect can be seen in both in normotensive persons (in whom the response is called white coat hypertension) and hypertensive patients [68]. It is present in about 7% of the population and, although estimates vary, may be found in as many as 60% of hypertensive patients [69]. In persons with white coat hypertension, blood pressure measurements are increased in a clinic setting but are normal in ambulatory or home settings. Blood pressure reactivity during exercise on either a treadmill [70] or bicycle [71] is similar to that seen in hypertensive and normotensive persons. However, peak systolic blood pressures are often greater than 200 mm Hg because of higher baseline blood pressures. The clinical significance of a white coat response in terms of illness and death caused by cardiovascular disease is unclear. Compared with normotensive persons, patients with white coat hypertension tend to have a greater left ventricular mass; greater left atrial dimension; abnormal echocardiographic indices of diastolic filling; lower high-density lipoprotein cholesterol levels; and elevated total cholesterol, triglyceride, and insulin levels [69, 71-73]. These findings suggest that white coat hypertension is not an entirely innocuous condition. Persons with this type of hypertension have an exercise capacity similar to that of hypertensive patients [70]. There is little data on the incidence of progression from white coat hypertension with a normal ambulatory blood pressure to fixed hypertension; no clear consensus has yet emerged on whether the condition is associated with increased morbidity or mortality [74, 75].

Exercise Responses in Essential Hypertension

Impaired Exercise Tolerance in Hypertensive Patients

While at rest, patients with essential hypertension have increased total peripheral resistance and a low or normal stroke volume [76]. When patients with hypertension are exercising, stroke volume increases subnormally and the heart rate during peak exercise is lower. The cardiac output is therefore lower, exercise time is decreased, the anaerobic threshold is reached earlier, and maximal oxygen uptake is reduced [4, 7, 77-79]. Exercise capacity in hypertensive patients is reduced as much as 30% compared with age-matched normotensive controls [30, 77, 79-81] (Table 2).

Fagard and colleagues [7] did a matched subgroup analysis comparing younger patients with hypertension (mean age ±SD, 24 ± 5 years) with an older group (mean age, 40 ± 6 years). The blood pressure was equivalent in both groups. The researchers determined that maximal oxygen consumption was reduced 16% during exercise; in the older group, this reduction was attributed to age alone. When the investigators controlled for age and compared groups with resting mean arterial pressures of 125 ± 12 mm Hg (elevated) and 92 ± 11 mm Hg (normal), a similar (16%) reduction in maximal oxygen consumption was seen in the group with high resting blood pressure. In a study of 61 asymptomatic hypertensive patients (44 men and 17 women), Amery and colleagues [4] reported a 35% reduction in maximal oxygen uptake that was attributable to age (age groups of 50 to 69 years and 19 to 34 years). When the investigators used this age correction to compare groups with diastolic blood pressures of 90 to 99 mm Hg (mild hypertension) and greater than 110 mm Hg (moderate to severe hypertension), the 18% reduction in maximal oxygen uptake was attributable to hypertension.

However, the observed reduction in work capacity has not been clearly distinguished from changes in body mass index or insulin resistance, both of which are widely associated with hypertension. It is known that insulin-resistant, normotensive men with a family history of hypertension have lower maximal oxygen uptake [82]. In one isolated observational study that had insufficient statistical power, Dengel and colleagues [83] studied 11 older hypertensive men (mean age, 58 years) and found no association between insulin resistance and maximum oxygen consumption. The importance of body mass and metabolic response to exercise should be further investigated.

Diastolic Blood Pressure Response to Exercise in Hypertensive Patients

Akhras and Jackson [84] analyzed the responses of 541 patients with chest pain or recent uncomplicated myocardial infarction (68 had had hypertension before an acute myocardial infarction) to a symptom-limited treadmill exercise test using the Bruce protocol and reported that 16.8% had an increase in diastolic blood pressure of 15 mm Hg or more. The presence of an abnormal diastolic blood pressure response (elevated pressure during exercise) was associated with multiple coronary artery disease (86% or 24 of 28 patients with an abnormal response) on angiography despite the absence of ischemic electrocardiographic changes during exercise. This blood pressure response was also associated with a reduced ejection fraction on ventriculography, which improved with successful coronary artery bypass grafting. The efficiency of this diastolic blood pressure response in stratifying hypertensive patients is unclear, but the response may be useful if explored more fully. In most conventional analyses, the use of exercise testing to stratify patients with chest pain focuses on electrocardiographic responses or decrease in or failure to elevate systolic blood pressure during exercise [85].

Systolic Blood Pressure Response to Exercise in Hypertensive Patients

In almost all untreated hypertensive patients, the systolic blood pressure during exercise is greater than 200 mm Hg [86]. Several studies have correlated this exaggerated blood pressure response with end-organ damage [10, 87-91] Table 3, but others have found no association [86, 92]. Fagard and colleagues [86] reported that in hypertensive patients, the blood pressure response to controlled exercise is no better associated with end-organ damage than is resting blood pressure. In their study, 169 patients exercised using bicycle ergometry and intra-arterial blood pressure was recorded. Fagard and colleagues' negative findings may be explained by the fact that they included relatively young patients (mean age, 36.3 years) and excluded many more typical patients in older age groups.

Mundal and colleagues [93] recently reported findings linking an exaggerated increase in systolic blood pressure at submaximal exercise with increased mortality in hypertensive patients. Nearly 2000 fit Scandinavian men aged 40 to 59 years were tested on a 100-W and 6-minute cycle ergometer protocol; systolic blood pressure was measured noninvasively when the patients reached the defined stage. The patients were followed for an average of 16 years. In patients with casual blood pressure greater than 140 mm Hg and systolic blood pressure during exercise greater than 200 mm Hg (304 of 520 patients), the mortality rate from cardiovascular disease (myocardial infarction, cerebral stroke, and sudden death) was 16.1%; the rate was 6% in patients with a casual systolic blood pressure greater than 140 mm Hg and a systolic blood pressure during exercise of less than 200 mm Hg (that is, a mortality rate similar to that in persons with normal casual blood pressure). These results were independent of body mass index, smoking habits, serum cholesterol and triglyceride levels, glucose tolerance, and exercise capacity. Unfortunately, the contribution of antihypertensive therapy is unclear. In a smaller subgroup analysis, Fagard and colleagues [94] did not attach independent prognostic significance to this exaggerated systolic blood pressure response above that of elevated resting blood pressure in younger patients (n = 143; mean age, 35.3 years) with a shorter follow-up (10 years).

These studies have important implications. They suggest that blood pressure during exercise may predict prognosis in hypertensive patients. The effect of antihypertensive therapy on this variable should be considered. This question could be answered by a prospective, long-term comparative study of end-organ damage and mortality in treated hypertensive patients with normalized casual blood pressure, those with normalized ambulatory blood pressure, and those with a normalized hemodynamic response to exercise. No such data currently exist to assess the efficacy of this additional target for antihypertensive treatment.

Hypertensive Heart Disease

Hypertensive heart disease is the association of arterial hypertension with left ventricular hypertrophy, increased myocardial fibrosis [95, 96], and accelerated atherosclerotic coronary artery disease. The presence of echocardiographic left ventricular hypertrophy is related to higher resting systolic [90] and diastolic blood pressure [89]. The presence of left ventricular hypertrophy is associated with a higher total peripheral resistance, reduced cardiac output at rest, greater abnormalities in the exercise response, and a greater reduction in exercise capacity [97]. Maximal systolic blood pressure during exercise [10, 87, 89, 90] is also elevated, although an incremental staged increase in systolic blood pressure is blunted because of higher baseline levels [67].

Hypertensive heart disease progresses to hypertensive heart failure. With current policies for detecting and treating hypertension, this progression still occurs but is either rare or unrecognized in Western societies. Heart failure more generally appears in hypertensive patients with concomitant ischemic heart disease caused by intervening myocardial infarction. Elevated resting blood pressure is then lost, and heart failure becomes the dominant clinical presentation.

Exercise studies in black African patients have provided an almost unique opportunity to observe exercise responses in persons with hypertensive heart failure. The prevalence of hypertensive heart failure is almost certainly due to a lack of adequate therapy. The association between hypertension and coronary heart disease in African patients with hypertensive heart failure is much weaker [98, 99]. In a small study, Balogun and colleagues [80] reported a higher resting supine heart rate, systolic blood pressure, and diastolic blood pressure in seven patients with hypertensive heart failure compared with age- and sex-matched normotensive volunteers and patients with uncomplicated hypertension. Both stroke volume and cardiac index were below normal, with high total peripheral resistance and increased arteriovenous oxygen difference at rest [100]. During exercise, the increase in systolic blood pressure and heart rate was markedly impaired, an outcome associated with reduced exercise capacity in patients with hypertensive heart failure. The estimated maximal oxygen uptake in such patients (3.5 METs) was reduced by 50% compared with normotensive persons (7.2 METs) and by 40% compared with persons with uncomplicated hypertension (6.2 METs) [80].

Mechanism of Reduced Exercise Tolerance and Abnormal Exercise Response in Hypertension

Central Hemodynamics and the Natural History of Hypertension

The end-organ effects of hypertension are progressive, and resting blood pressure may not accurately indicate the underlying state of the disease. A hyperkinetic circulation may be present in the early phase, possibly before resting blood pressure is elevated, which may be explained by the relative blood volume expansion [101], reduced venous compliance [102, 103], impaired peripheral vasodilatory capacity [59], and sympathetic hyper-reactivity [104]. At this stage, exercise testing shows a subnormal increase in the stroke volume and the failure of the total peripheral resistance to decrease normally [5]. These factors contribute to the limitation of exercise capacity.

With time, patients with a hyperkinetic circulation may develop fixed hypertension with high mean arterial blood pressure, normal or reduced cardiac output, lower stroke volume, and high total peripheral resistance at rest [105]. These abnormalities are accentuated during exercise. Left ventricular filling is impaired at rest [106-108] with normal systolic function, before left ventricular hypertrophy develops. Impaired diastolic filling is more prominent with exercise, producing reduced stroke volume by the Frank-Starling mechanism [109]. Myocardial reserve is reduced [110], and relative myocardial ischemia may contribute to these findings. With progressing disease, left ventricular hypertrophy is evident on echocardiography and eventually on electrocardiography. The hemodynamic abnormalities are accentuated [110-114]; although the left ventricular end-diastolic volume is normal, end-systolic volume is later increased, further reducing stroke volume, ejection fraction, and peak oxygen uptake during exercise [76, 105, 106]. At this point, left ventricular systolic function is impaired and hypertensive heart failure occurs.

With the onset of hypertensive systolic heart failure, both systolic and diastolic blood pressure can start to decrease at rest, but with markedly increased total peripheral resistance. During exercise, the failing hemodynamic response results in even more limited exercise capacity [115].

Diastolic Dysfunction

Diastolic heart failure may explain symptoms of heart failure despite normal systolic function [116-119]. As many as 65% of patients with diastolic heart failure have hypertension [116]. Hemodynamic studies, echocardiographic investigations, and radionuclide angiography have shown that diastolic dysfunction and high total peripheral resistance play a large part in the abnormal exercise response in hypertension [5, 109].

Left ventricular filling during diastole depends on myocardial relaxation, left ventricular compliance, and atrial contraction. Briefly, left ventricular diastole is divided into four phases: isovolumic relaxation, rapid filling, diastasis, and atrial systole. Myocardial relaxation is energy dependent [120] and involves adenosine triphosphate and calcium transport. In left ventricular hypertrophy, coronary microcirculation is abnormal because of medial hypertrophy of coronary resistance vessels [121]. Animal studies that induced left ventricular hypertrophy have shown an associated reduction in endocardial capillary density [122] and blood flow and impaired thickening of the subendocardial wall, particularly during exercise; these conditions persist for some time after exercise [123]. These abnormal coronary hemodynamics result in reduced flow reserve and consequent exercise stress; these in turn lead to relative ischemia (limitation of adenosine triphosphate production), which impairs diastolic relaxation and reduces ventricular filling.

Left ventricular compliance is related to myocardial collagen composition, and, in hypertension, an increase in collagen deposition is associated with abnormal orientation. This outcome may be secondary to increased aldosterone levels [124, 125]. Reduced left ventricular compliance increases the "stiffness" of the heart, reduces diastolic filling, and elevates atrial pressure and ventricular end-diastolic pressure. Atrial systole becomes more important in ventricular filling in this setting [126], as it does in a normal aging heart [127]. This may be one reason the onset of atrial fibrillation in elderly hypertensive patients can result in acute pulmonary edema despite intact systolic function. The combination of these factors leads to a subnormal increase in stroke volume during exercise that is caused by impaired ventricular filling [109]. Left atrial pressure increases abnormally during exercise, thereby producing pulmonary venous congestion, subjective breath-lessness, and limited exercise tolerance. This finding may explain why exercise capacity is better correlated with diastolic function than systolic function [128].

Pulmonary Circulation and the Right Ventricle

The pulmonary circulation is not spared in hypertension: Both pulmonary artery pressure and pulmonary vascular resistance increase [129]. During exercise, pulmonary circulation resistance has been shown to be significantly higher in hypertensive patients than in matched normal controls [130]. When Doppler echocardiographic indices were used, right ventricular and left ventricular diastolic function were similarly impaired [131, 132]. Right ventricular ejection fraction is consequently reduced, an effect that also occurs with the left-sided circulation [133]. These findings may contribute to the impaired exercise tolerance in hypertensive patients. The mechanism of right ventricular function and pulmonary vascular resistance in hypertension has not been sufficiently studied. This may be a fertile area in which to investigate the cause of reduced exercise capacity and the therapeutic benefits of treatments.

Peripheral Hemodynamics

In the early phases of hypertension, calculated total peripheral resistance is normal at rest but does not decrease during exercise because of a reduced vasodilatory capacity. When fixed hypertension develops, peripheral resistance increases; histopathologic studies have shown an increased wall thick-ness-to-lumen ratio in resistance vessels [134]. These findings may explain the failure of total peripheral resistance to decrease and the exaggerated systolic blood pressure response to exercise. During a 20-year follow-up study of patients with hypertension, total peripheral resistance at rest continued to increase despite treatment [105]. The true efficacy of antihypertensive treatment is difficult to assess (see below).

Sympathetic Nervous System

Plasma catecholamine levels are correlated with increased cardiac output and total peripheral resistance in at least some normal persons at risk for developing hypertension [135]. Such persons develop higher concentrations of catecholamines in response to stress [104, 136] and have a greater response to norepinephrine infusion than controls [137]. This finding has been interpreted as suggesting that the sympathetic nervous system in the early phases of hypertension is overactive and hyper-responsive. The higher concentrations of catecholamines may also partly explain the hyperkinetic circulation and exaggerated blood pressure response to exercise. However, blocking the autonomic nervous system in such persons so that the heart rate during exercise is the same as that in controls results in changes in stroke volume and depressed cardiac output [138]. This discrepancy is probably caused by diastolic dysfunction [95].

Baroreceptors

Baroreceptor activation is a basic physiologic servo control system that reduces sympathetic activity and promotes vagal tone. The threshold of activation or "set point" is higher in hypertension, and responsiveness to autonomic changes is reduced [139]. The modification of the baroreceptor system probably has a role in the hemodynamic response to exercise in prehypertensive patients, hypertensive patients, and patients with hypertensive heart failure in any stage. Antihypertensive treatment may interact with this system (see below).

Hemodynamic Effects of Antihypertensive Medication

Elevated blood pressure at rest is a marker for hypertensive vascular disease that, if left untreated, results in considerable morbidity and mortality. There is little doubt about the benefit of lowering elevated blood pressure using pharmacologic and nonpharmacologic interventions. The incidence of stroke and coronary heart disease can be reduced by as much as 56% and 37%, respectively, by treating hypertension alone [140, 141]. These figures have been established from large studies based on simple sphygmomanometric blood pressure measurement at rest and probably include variable proportions of patients with the white coat response. As mentioned above, patients with white coat hypertension almost certainly do not have the elevated risk for cardiovascular disease that is associated with fixed hypertension. The benefits of treatment may thus be underestimated accordingly.

The measurements used to assess the efficacy of treatment are made at rest. It is now known that end-organ damage progresses despite control of resting blood pressure [105, 142, 143]. Logically, most persons with asymptomatic and uncomplicated hypertension remain normally active and indeed are encouraged to exercise. Abnormal vascular and cardiac responses to activity (submaximal exercise) may persist despite apparently effective antihypertensive treatment at rest in individual patients. Thus, two patients may have the same treated resting blood pressure but different exercise control; the latter depends on the nature of the drugs used and the dose selected. An ideal antihypertensive agent or a combination of agents would therefore normalize at-rest hemodynamics and the response to exercise. Blood pressure control, regardless of the agents used, produces regression of left ventricular hyper-trophy [144-146] and improves left ventricular diastolic function [147]. The extent to which this finding can be explained by control of exercise or activity responses in individual patients is uncertain. This finding may also translate into improved exercise hemodynamics and function, but there is a considerable gap in current knowledge. In this section, we consider the relative profile of several antihypertensive drugs.

Diuretics

The mechanism of action of thiazide diuretics is complex. They act at least in part by reducing blood volume and cardiac output while maintaining a constant heart rate and total peripheral resistance. Total peripheral resistance slightly decreases with long-term treatment [148]. During exercise, hypertensive patients who are receiving a diuretic alone show an attenuated increase in blood pressure and a reduced stroke volume but have a maintained normal heart rate response [148]. More powerful loop-acting or potassium-conserving diuretics are not conventionally used as antihypertensive drugs in patients with normal renal function. The effects of these drugs on exercise hemodynamic responses in hypertension have been poorly described.

ß-Blockers

ß-Blockers reduce heart rate, stroke volume, and cardiac output at rest; increase total peripheral resistance to compensate for this reduction; and result in no immediate change in blood pressure. With time, the total peripheral resistance decreases as blood pressure is reduced [149]. ß-Blockers with intrinsic sympathomimetic activity, such as pindolol, reduce the heart rate to a lesser degree at rest. Selective ß₁-blockers (such as atenolol, metoprolol, and timolol) reduce cardiac output to a lesser extent because of a compensatory increase in stroke volume. During exercise, all ß-blockers reduce heart rate by 20% to 30% and the blood pressure response by 10% to 20% [150]. The increase in cardiac output with exercise is reduced, as is maximal exercise capacity. With longer-term treatment, patients develop tolerance to these effects. Long-term treatment with atenolol has been shown to reduce blood flow to the muscles, but this effect is compensated by an increased oxygen extraction by the tissues [150]. On detailed examination of the literature, little evidence supports the hypothesis that ß-blockers aggravate established but noncritical peripheral vascular disease [151]. The effect of ß-blockade on muscle metabolism during exercise may be important but is poorly differentiated from effects on substrate delivery and blood flow.

-Blockers

₁-Receptor blockers (such as prazosin and doxazosin) cause a marked reduction in total peripheral resistance and therefore reduce blood pressure, with a small increase in heart rate (5%) and cardiac output [152]. Long-term treatment is associated with an increase in both stroke volume and cardiac output and with normalization of central exercise hemodynamics [153]. Paradoxically, in the few studies available, maximal exercise capacity was shown to be reduced by 0.86 METs in young trained male athletes with mild hypertension receiving an -blocker compared with placebo recipients [154].

Dual-Acting Drugs

Drugs with both -and ß-receptor blocking ability (such as labetalol and carvedilol) reduce total peripheral resistance by -receptor blockade and reduce heart rate and cardiac output by ß-receptor blockade (stroke volume unchanged); this results in decreased blood pressure at rest. With long-term treatment, heart rate and total peripheral resistance are significantly reduced, but this effect is associated with an increased stroke volume and cardiac output. With these drugs, exercise hemodynamics returns to normal [155].

Calcium Channel Blockers

The calcium channel antagonists lower blood pressure by reducing total peripheral resistance through peripheral arteriolar dilatation. Verapamil reduces resting blood pressure and may or may not reduce heart rate. During exercise, heart rate decreases by about 10%, with a compensatory increase in stroke volume and thus preservation of cardiac output [156]. In the elderly, diastolic filling of the left ventricle at rest is enhanced with the use of verapamil [157]. It would be of interest to know whether this effect leads to increased exercise performance; however, treating younger hypertensive patients with this drug does not appear to alter exercise capacity [158]. In patients with diastolic heart failure secondary to hypertension, however, exercise time is improved by 33% (15% improvement in placebo arm) with verapamil [159]. In hypertrophic cardiomyopathy, verapamil is known to improve both diastolic function and exercise tolerance [160]. Diltiazem has similar effects, except that heart rate is reduced both at rest and during exercise. With long-term therapy, exercise times are slightly but significantly improved (55 seconds over baseline compared with 18 seconds over baseline for placebo recipients) [161]. Treatment with nifedipine is initially associated with reflex tachycardia, which disappears during long-term treatment [162]. Exercise capacity is not significantly altered with short-term treatment [79].

Angiotensin-Converting Enzyme Inhibitors

Angiotensin-converting enzyme inhibitors act by various potential mechanisms, modulating the renin-angiotensin-aldosterone systems and the kallikrein-kinin and parasympathetic nervous systems. With the use of these inhibitors, total peripheral resistance is reduced and cardiac output and heart rate do not change. Regression of left ventricular hyper-trophy [147] and increased coronary artery blood flow have been associated with use of angiotensin-converting enzyme inhibitors [163], and cardiac out-put is modestly increased with long-term treatment. During exercise, the increase in blood pressure is reduced and heart rate is not affected; blood pressure during exercise is well controlled overall [164, 165]. In a study by Kusaka and colleagues [165], 19 hypertensive elderly patients (> 60 years; 8 men and 11 women) exercised using a treadmill protocol after receiving 8 weeks of captopril therapy; resting blood pressure was well controlled during exercise. Kusaka and colleagues reported an increased exercise tolerance after treatment (13.1 minutes compared with 16.5 minutes). Because this study was not controlled and patients were tested on only two occasions, the improvement may reflect a training effect alone [35].

Centrally Acting Drugs

Clonidine, a centrally acting ₂-receptor stimulant, acts on postsynaptic -receptors in the brain stem. It inhibits the basal efferent sympathetic activity, with reciprocal increase in vagal tone [166]. At rest, blood pressure, cardiac output, and stroke volume are reduced because of a decrease in venous return, which in turn is caused by venodilatation. Although heart rate is reduced by an increase in vagal tone, the total peripheral resistance remains unchanged; however, heart rate does decrease with long-term treatment and improvement of cardiac output to the level before treatment. The oxygen uptake at rest is reduced by 10%; during exercise, the blood pressure increase is not controlled and maximal oxygen uptake is reduced by about 5% [167, 168].

Further Examination of the Effect of Drug Therapy

The implications of drug treatment of hypertension and its complications should be extended to the response to exercise. The adequacy of effective antihypertensive treatment of basal blood pressure for correcting the abnormal hemodynamic response to exercise stress is poorly defined. The functional implications of diastolic dysfunction in hypertensive patients should be more clearly outlined and the effect of the range of antihypertensive strategies, defined. Much circumstantial evidence surrounds the specific use of angiotensin-converting enzyme inhibitors to reverse structural hypertensive changes in the vascular and cardiac interstitium [169, 170]. The effect of angiotensin-converting enzyme inhibitors or the more recently introduced angiotensin-receptor antagonists on the functional exercise responses in hypertensive heart disease and diastolic function require better definition before any specific use of these expensive drugs is justified.

Summary

Top Methods Summary Author & Article Info References

Hypertension is associated with progressive changes in hemodynamics at different stages of its evolution. The early hyperkinetic circulation, with relatively increased central blood volume, high cardiac output, and normal total peripheral resistance, progresses to high total peripheral resistance and normal cardiac output in fixed hypertension and subsequently to low cardiac output with the development of hypertensive heart disease. Ultimately, diastolic or systolic heart failure supervenes with or without intercurrent myocardial infarction, resulting in increased total peripheral resistance and normalization of both systolic and diastolic blood pressure. The treatment of hypertension should be adjusted according to underlying hemodynamic changes. Different antihypertensive agents may be optimal at different stages of hypertension or additions made to existing treatments. It is unclear whether good control of resting blood pressure alone is adequate, particularly when most patients are normally active. It remains to be tested whether treatment of exercise responses alters the progression of the underlying physiologic abnormalities in hypertension or improves the outcome and can produce better care for individual hypertensive patients. Exercise testing can identify persons with exaggerated systolic and diastolic blood pressure response, indicating future risk for developing hypertension and increased mortality. An association between impaired function and left ventricular hypertrophy remains unclear. Impaired exercise tolerance in hypertensive patients defined by functional exercise testing can be used to assess new drugs for hypertension treatment, adequacy of exercise blood pressure control, and the search for appropriate therapy for diastolic dysfunction in relation to exercise.