Author + information
- Received September 6, 2001
- Revision received April 16, 2002
- Accepted May 20, 2002
- Published online August 21, 2002.
- Bronwyn A Kingwell, PhD*,* (, )
- Tamara K Waddell, PhD*,
- Tanya L Medley, BAppSc*,
- James D Cameron, MD, MengSc† and
- Anthony M Dart, BChBM, DPhil*
- ↵*Reprint requests and correspondence:
Dr. Bronwyn Kingwell, Alfred and Baker Medical Unit, Baker Medical Research Institute, P.O. Box 6492, St. Kilda Road Central, Melbourne, Victoria, 8008, Australia.
Objectives The goal of this study was to determine whether large artery stiffness contributes to exercise-induced myocardial ischemia in patients with coronary artery disease (CAD).
Background Large artery stiffness is an independent predictor of cardiovascular mortality and a major determinant of pulse pressure and, thus, cardiac afterload and coronary perfusion. Clinical relevance of the hemodynamic consequences of large artery stiffening has not previously been demonstrated in relation to myocardial ischemia.
Methods We hypothesized that stiffer large arteries would reduce myocardial ischemic threshold as assessed by time to ST-segment depression of 0.15 mV during a treadmill exercise test in patients with CAD. Ninety-six patients with CAD (78 men) age 62 ± 9 years (mean ± SD) were classified as having single (52 patients), double (31 patients), or triple (13 patients) coronary vessel disease, based on angiographically confirmed stenoses >50%. Systemic arterial compliance, distensibility index, aortic pulse wave velocity, and carotid augmentation index were measured using carotid applanation tonometry and Doppler velocimetry of the ascending aorta, at rest.
Results In univariate analysis, all large artery stiffness/compliance indexes correlated with time to ischemia (p = 0.01 to 0.009). Both carotid (p = 0.007) and brachial (p = 0.001) pulse pressure also correlated inversely with time to ischemia. In multivariate analysis including other major risk factors plus severity of coronary stenosis, indexes of arterial stiffness were significant independent predictors of ischemic threshold.
Conclusions Within a patient group with moderate CAD, large artery stiffness was a major determinant of myocardial ischemic threshold.
Large artery stiffness has recently been related to cardiovascular mortality (1,2); however, the mechanisms underlying this relationship have not been established. While several mechanisms could contribute, an unfavorable hemodynamic profile caused by cardiac ejection into a stiff proximal circulation may promote myocardial ischemia in patients with coronary artery disease (CAD).
Pulse pressure (PP) elevation is an important consequence of increased large artery stiffness. Clinical studies have demonstrated a strong association between PP at baseline and future cardiovascular, including coronary, events (3,4). Elevated PP could affect coronary outcomes through increased systolic pressure and, thus, afterload and diminished coronary perfusion secondary to diastolic pressure reduction (5,6). Thus, aortic stiffening tightens the link between cardiac systolic performance and myocardial perfusion (7). Aortic stiffness may, thus, be an important determinant of ischemic threshold in patients with CAD; however, the clinical relevance of this mechanism has not been investigated.
The primary objective of the current study was to determine whether large artery stiffness was an independent determinant of myocardial ischemic threshold in patients with CAD. The underlying hypothesis was that, for any degree of coronary stenosis, patients with stiffer large vessels would have a lower ischemic threshold. To address this hypotheses in humans noninvasively, time to ST-segment depression of 0.15 mV during a graded exercise test was used as a measure of myocardial ischemic threshold in patients with angiographically confirmed CAD. Arterial compliance, distensibility index, pulse wave velocity (PWV), and augmentation index (AI) were used to assess different aspects of large artery stiffness at rest in relation to ischemic threshold.
Subjects and study design
Ninety-six sequential patients (78 men) age 62 ± 9 years (mean ± SD) meeting the inclusion criteria were recruited. All had positive exercise stress tests (at least 0.1 mV depression) and angiographically confirmed CAD (see Angiography section). Participants gave informed consent for the study, which was approved by the Alfred Hospital Ethics Committee and conducted in accordance with the Declaration of Helsinki (2000) of the World Medical Association.
All patients attended the laboratory <4 weeks after their angiogram and before revascularization or changes to their medical therapy. Fasting blood samples were drawn for the measurement of lipids and glucose. Patients then rested in a quiet room for 10 min or until blood pressure had stabilized. Systemic arterial compliance (SAC), distensibility index (DI), PWV, and AI were then assessed noninvasively (see the following text), and an echocardiogram was performed for assessment of left ventricular mass. Finally, patients underwent a treadmill exercise test under standardized conditions (this was additional to their diagnostic stress test).
Before entry in the study, all patients had undergone coronary angiography. Angiograms were analyzed independently as part of routine hospital procedure by an experienced cardiologist before recruitment into the study. Angiographically identified stenoses ≥50% in the major coronary vessels were used to classify patients as having single- (52 patients), double- (31 patients), or triple- (13 patients) vessel disease. The major coronary vessels were the left anterior descending, the left circumflex, or its marginal branch when the branch constituted the main continuation of the circumflex, and the right coronary artery. Percentage narrowing was determined from views of the vessels in two planes. The greatest narrowing of the most severe lesion in each vessel was reported. The most severe lesion in the three major vessels was reported as the maximum coronary stenosis.
Resting brachial arterial blood pressure and heart rate were measured at 3-min intervals using a Dinamap vital signs monitor (1846 SX, Critikon, Tampa, Florida), with subjects remaining undisturbed in the supine position for 10 min. The mean of three values was taken to represent resting levels.
Systemic arterial compliance was measured using calculations based on the area method of Liu et al. (8), as described previously (9,10). To obtain an estimate of mechanical properties independent of vessel size, a DI was calculated as SAC normalized to left ventricular outflow tract area.
Pulse wave velocity was measured by simultaneous recordings of arterial pressure waves at two sites using applanation tonometry (SPT-301, Millar Instruments, Houston, Texas), as described previously (10,11). Pulse wave velocity was measured between the carotid and femoral arteries and from the femoral to dorsalis pedis arteries.
Carotid pressure waveforms were used to calculate AI, defined as the difference between the first and second systolic peaks of the central arterial waveform, expressed as a percentage of the PP (12).
Left ventricular mass
Left ventricular mass normalized for body surface area (left ventricular mass index) was estimated using echocardiography (Hewlett-Packard Sonos 1500, Hewlett-Packard, Andover, Massachusetts) (13).
Exercise stress test
All patients underwent a standard modified Bruce protocol exercise stress test as part of the study protocol. Beta-adrenergic blocking agents were discontinued at least 24 h before the test, and at least 4 h elapsed between use of nitrates and exercise. A positive response in the electrocardiogram was defined as horizontal or downsloping ST-segment depression of ≥0.1 mV (1 mm), measured 80 ms after the J point, occurring in at least six consecutive complexes, in at least three different leads. The lead showing the greatest ST-segment depression was analyzed every minute by an experienced cardiac technician. The time at 0.15 mV (1.5 mm) ST-segment depression was calculated using linear regression. This method shows a close correlation with the number of ischemic episodes detected during ambulatory monitoring (r = −0.86) (14).
Myocardial work calculations
Rate pressure product (RPP) (systolic blood pressure [SBP]·heart rate) and pressure work index (15) were calculated at rest as indexes of myocardial work. Rate pressure product was also calculated at peak exercise (pressure work index, which incorporates stroke volume, could not be calculated at maximum exercise). As a measure of myocardial reserve, the difference between peak and resting RPP was also calculated.
Total, low-density lipoprotein and high-density lipoprotein cholesterol, triglycerides, and glucose were determined (16).
Pearson correlation coefficient was used for univariate analysis, whereas multiple regression analysis was performed to examine determinants of the time to ischemia. Multiple regression was by stepped backward and forward entry. Variables were entered if the respective F probability was <0.05 and were removed if it was >0.1. All data were analyzed using SPSS for Windows Version 9.0.1. (SPSS Inc., Chicago, Illinois). Unless otherwise stated, group results are presented as mean ± SD, and statistical significance was deemed to have been achieved when p < 0.05.
The cohort studied had an average age of 62 ± 9 years, a body mass index of 28 ± 4 kg/m−2 and a left ventricular mass index of 104 ± 30 g/m−2. Both blood pressure and lipid levels were reasonably controlled (Table 1), and patients had a moderate level of CAD, as indicated by both the magnitude of the maximum coronary stenosis and the average number of major coronary vessels with a stenosis ≥50% (Table 1). Mean fasting plasma glucose was within the normal range (Table 1), although 10 patients were diagnosed with type 2 diabetes. Of the 96 patients, 5 were current smokers, 33 were nonsmokers, and 58 were ex-smokers. With regard to medication, 20 patients were taking angiotensin-converting enzyme inhibitors, 48 were on beta antagonists, 25 on calcium antagonists, 64 on lipid-lowering therapy, 33 on nitrates, 7 on diuretics, 2 on hormone replacement therapy, and 76 were taking aspirin.
CAD severity and time to ischemia
Predictably, the magnitude of the maximum stenosis in the major coronary vessels was inversely related to time to ST-segment depression (Table 2). There was, however, no relationship between the number of major vessels with a stenosis ≥50% and ischemic threshold. These findings are consistent with ischemic threshold being determined by a critical stenosis (maximum stenosis) and not being related to the extent of CAD (number of diseased vessels).
Age and time to ischemia
Patients’ ages spanned 33 years (from 44 to 77 years). Because age will relate to both disease severity and large artery stiffness, it would also be expected to correlate with ischemic threshold. In this cohort age was a strong univariate predictor of time to ST-segment depression (Table 2).
Large artery stiffness and time to ischemia
Mean values for all arterial stiffness indexes are shown in Table 3. All indexes of large artery stiffness, including SAC, DI, carotid-femoral PWV, and AI, correlated in univariate analysis with time to ischemia, with r values ranging from −0.25 to 0.27 (p < 0.05 for all; Table 2). Both carotid and brachial PPs were similarly related to ischemic time (p < 0.01 for all; Table 2). There was no relationship, however, to peripheral arterial stiffness as measured by femoral-dorsalis pedis PWV.
With time to ischemia as the dependent variable, indexes of arterial stiffness were entered into separate multivariate analyses also incorporating age, gender, body mass index, number of diseased vessels, magnitude of maximum stenosis, smoking status, diabetic status, mean arterial pressure, low-density lipoprotein cholesterol, heart rate, and left ventricular mass index (Table 4). As expected, in all these analyses age was an independent predictor of time to ischemia. In addition to age, in separate analyses, SAC, DI, and AI were the only other independent determinants of time to ischemia (Table 4). These relationships were unchanged when the various medication classes were entered into the multivariate analyses. The univariate correlation between time to ischemia and carotid-femoral PWV was not independent of age in multivariate analysis (Table 4). This was also true for systolic and pulse pressure (Table 4). Figure 1 shows the relations between arterial stiffness indexes and time to ischemia. All data are adjusted to the mean age of the group (62 years) using the slope of the univariate relation between each variable and age.
Myocardial work and time to ischemia
Mean values for myocardial work indexes are shown in Table 3. Indexes of resting cardiac work, including RPP and pressure work index (15), were inversely related to time to ST-segment depression (Table 2). In addition, RPP at maximal exercise was positively related to time to ST-segment depression suggesting that large artery compliance may have influenced exercise capacity through effects on cardiac work at maximal exercise (Table 2). Even stronger was the positive relationship of the change in RPP (peak-resting) and time to ST-segment depression (Table 2). Furthermore, in multivariate analysis, this relationship was independent of all major covariants in predicting time to ST-segment depression (Table 4). These data are supportive of the hypothesis that the ability to increase cardiac work is an important determinant of ischemic threshold. Thus, lower cardiac work at rest (which is associated with higher SAC [r = −0.23, p = 0.03] and DI [r = −0.26, p = 0.01] and lower carotid-femoral PWV [r = 0.47, p < 0.001]) is predictive of a higher ischemic threshold.
The major new finding of this study is that large artery stiffness is an important independent determinant of myocardial ischemic threshold during exercise in patients with CAD. This mechanism may underlie the relation between large artery stiffness and cardiovascular and coronary mortality (1,2). Experimental animal data suggest that the hemodynamic consequences of large artery stiffening, including elevated systolic and reduced diastolic pressure, tighten the relation between cardiac systolic performance and myocardial perfusion, particularly in the setting of a coronary occlusion (6,7). This occurs partly as a result of increased PWV, which shifts pressure wave reflections from diastole to systole, thus augmenting systolic pressure. In addition, a stiff aorta has a diminished capacity to serve as a blood reservoir during cardiac ejection such that blood is available for coronary perfusion during diastole. The current study provides the first clinical evidence to suggest that this mechanism may be important in patients with CAD. Specifically, for any given level of stenosis severity, large artery stiffness at rest was inversely related to ischemic threshold.
The end point in the current study was time to ST-segment depression of 0.15 mV. Because all participants had been shown to have significant CAD on angiography, it is unlikely that such depression represented a false positive test, even in female participants in whom this can most commonly occur. All patients had previously undergone a diagnostic stress test and were, therefore, familiar with the procedure. Serial monitoring of ST-segment deviation allowed precise identification of the exercise time to a common end point. Previous studies have demonstrated a close relationship between ischemic ST-segment change and impaired contractility during exercise in patients with CAD (17).
Determinants of myocardial ischemic threshold
Factors affecting both coronary blood supply and myocardial blood demand will contribute to myocardial ischemic threshold. On the “demand” side of the equation, myocardial efficiency and cardiac work are the most important factors. Large artery stiffening does not affect myocardial efficiency (5); however, systolic pressure elevation increases the energetic cost to the heart to maintain adequate cardiac output. While this would not be expected to result in any major functional decrement at rest, reserve capacity would be expected to be limited. Myocardial blood supply depends largely on diastolic pressure and duration and on coronary vascular resistance, the latter being partly dependent on the severity of coronary lesions. Pulse pressure elevation, secondary to large artery stiffening, would, therefore, be expected to negatively influence both sides of the coronary blood supply and demand equation, through both higher systolic and lower diastolic blood pressure. Aortic stiffening induced in dogs, through either aortic bandaging (6) or diverting aortic outflow through a stiff bypass (7), has verified this experimentally. The relationships of PP and indexes of cardiac work with time to ischemia suggest that this mechanism is probably also operative in humans with CAD. Pulse pressure elevation was not, however, an independent predictor of time to ischemia in multivariate analysis. Thus, factors additional to PP appear to be important. These could include parallel effects on cellular metabolism caused by aging or diet in both the myocardium and the large arteries, resulting in impaired efficiency in one and increased stiffness in the other. Finally, the importance of a compliant circulation in relation to myocardial reserve was highlighted by the positive independent correlation between the change in RPP (peak-resting) and time to ST-segment depression. Thus, patients with a more compliant circulation have a greater maximum myocardial work capacity, due, in part, to favorable hemodynamics, which permit a greater heart rate to be achieved before ischemia is experienced.
In this patient cohort, arterial stiffness indexes were stronger determinants of time to ischemia than angiographic assessment of severity of coronary stenoses. That is, while CAD severity may be the primary coronary flow determinant and, therefore, of time to ischemia, the stiffness of the large arteries becomes the major independent influence on ischemic threshold within a group of patients of relatively homogeneous age and disease severity. This finding implicates large artery stiffness as a potentially important target for therapy in patients with CAD. Therapies including antihypertensive agents, estrogen-containing hormone replacement therapy, and nonpharmacologic treatments, such as aerobic exercise training, have all been shown to reduce arterial stiffness and may confer particular benefit in CAD by improving ischemic threshold (18). Future therapies may target the large arteries more directly. Thiazolium compounds have shown promise as selective breakers of glycated protein crosslinks, which have been associated with vessel stiffening. In primates (19) and rats (20), these compounds reduce vascular stiffness.
Factors affecting large artery stiffening
Age, gender, and mean blood pressure are all major determinants of large artery stiffness (11), whereas heart rate is a lesser influence (21). Despite the fact that all these variables would be expected to contribute to the univariate relationships between stiffness indexes and ischemic threshold, multivariate analysis indicated that large artery stiffness (SAC, DI, AI) related to ischemic threshold independently. Furthermore, both gender and mean pressure were not independently related to ischemic threshold in multivariate analysis. It should also be noted that, due to the parallel relationship between large artery stiffness and age, multivariate analysis would actually underestimate the importance of stiffness in relation to myocardial ischemia. This factor likely accounts for the univariate correlation between time to ischemia and carotid-femoral PWV ceasing to be significant in multivariate analysis. The same explanation also applies to PP and SBP.
Aortic atherosclerosis is another major influence on large artery stiffness (22). Because aortic and coronary diseases are known to develop in parallel (23), it is likely that aortic atherosclerosis was a major determinant of large artery stiffness in this cohort. Indeed, our group has previously shown that large artery stiffness is higher in patients with CAD (24–26). It was not possible in this study to determine the relative contribution of aortic atherosclerosis to large artery stiffness and, therefore, ischemic threshold. It is evident, however, that atherosclerosis promotes myocardial ischemia not only via coronary artery obstruction but also via large artery stiffening.
Genetic variation will also contribute to large artery stiffening and underlie the influences previously discussed. Genetic modulation of both the renin-angiotensin system (27) and extracellular matrix homeostasis (28) are likely to be important.
In accordance with their disease status, all patients were taking a variety of medications. Both beta antagonists (24 h) and nitrates (4 h) were withdrawn before the study; however, other drug classes were taken as normal. While some of these drugs may affect large artery stiffness or ischemic threshold, it was the relationship between the prevailing level of stiffness and ischemic threshold, regardless of acute influences on large artery stiffness that was under investigation. Furthermore, none of the medication classes when entered into the multivariate analyses were significantly related to time to ischemia.
Large artery stiffness was associated with a reduction in myocardial contractile reserve (peak-resting RPP) and ischemic threshold. Indeed, within a patient group with moderate CAD, large artery stiffness was a key determinant of myocardial ischemic threshold. Because ischemic threshold is the major determinant of functional capacity in this patient group, large artery stiffness also stands to impact significantly on quality of life. These data highlight large artery stiffness as an ischemic mechanism and potential therapeutic target in patients with CAD.
The authors are grateful for the technical expertise of Melissa Formosa.
☆ Supported by research grants from VicHealth, Australia; the National Heart Foundation of Australia; and the National Health and Medical Research Council of Australia.
- augmentation index
- coronary artery disease
- distensibility index
- pulse pressure
- pulse wave velocity
- rate pressure product
- systemic arterial compliance
- systolic blood pressure
- Received September 6, 2001.
- Revision received April 16, 2002.
- Accepted May 20, 2002.
- American College of Cardiology Foundation
- Domanski M.J.,
- Mitchell G.F.,
- Norman J.E.,
- Exner D.V.,
- Pitt B.,
- Pfeffer M.A.
- Kelly R.P.,
- Tunin R.,
- Kass D.A.
- Watanabe H.,
- Ohtsuka S.,
- Kakihana M.,
- Sugishita Y.
- Kass D.A.,
- Saeki A.,
- Tunin R.S.,
- Recchia F.A.
- Liu Z.,
- Brin K.,
- Yin F.
- Cameron J.,
- Dart A.
- Rajkumar C.,
- Kingwell B.A.,
- Cameron J.D.,
- et al.
- Vaitkevicius P.,
- Fleg J.,
- Engel J.,
- et al.
- Murgo J.P.,
- Westerhof N.,
- Giolma J.P.,
- Altobelli S.A.
- Panza J.A.,
- Quyyumi A.A.,
- Diodati J.G.,
- Callahan T.S.,
- Epstein S.E.
- Rooke G.A.,
- Feigl E.O.
- Kingwell B.,
- Sherrard B.,
- Jennings G.,
- Dart A.
- Parker J.O.,
- West R.O.,
- Case R.B.,
- Chiong M.A.
- Dart A.M.,
- Kingwell B.A.
- Vaitkevicius P.V.,
- Lane M.,
- Spurgeon H.,
- et al.
- Wolffenbuttel B.H.,
- Boulanger C.M.,
- Crijns F.R.,
- et al.
- Farrar D.J.,
- Bond M.G.,
- Riley W.A.,
- Sawyer J.K.
- Benetos A.,
- Gautier S.,
- Ricard S.,
- et al.
- Medley T.L.,
- Cole T.J.,
- Gatzka C.D.,
- Wang W.Y.S.,
- Dart A.M.,
- Kingwell B.A.