Arterial-ventricular coupling: mechanistic insights into cardiovascular performance at rest and during exercise

Abstract

Understanding the performance of the left ventricle (LV) requires not only examining the properties of the LV itself, but also investigating the modulating effects of the arterial system on left ventricular performance. The interaction of the LV with the arterial system, termed arterial-ventricular coupling (E_A/E_LV), is a central determinant of cardiovascular performance and cardiac energetics. E_A/E_LV can be indexed by the ratio of effective arterial elastance (E_A; a measure of the net arterial load exerted on the left ventricle) to left ventricular end-systolic elastance (E_LV; a load-independent measure of left ventricular chamber performance). At rest, in healthy individuals, E_A/E_LV is maintained within a narrow range, which allows the cardiovascular system to optimize energetic efficiency at the expense of mechanical efficacy. During exercise, an acute mismatch between the arterial and ventricular systems occurs, due to a disproportionate increase in E_LV (from an average of 4.3 to 13.2, and 4.7 to 15.5 mmHg·ml⁻¹·m⁻² in men and women, respectively) vs. E_A (from an average of 2.3 to 3.2, and 2.3 to 2.9 mmHg·ml⁻¹·m⁻² in men and women, respectively), to ensure that sufficient cardiac performance is achieved to meet the increased energetic requirements of the body. As a result E_A/E_LV decreases from an average of 0.58 to 0.34, and 0.52 to 0.27 in men and women, respectively. In this review, we provide an overview of the concept of E_A/E_LV, and examine the effects of age, hypertension, and heart failure on E_A/E_LV and its components (E_A and E_LV) in men and women. We discuss these effects both at rest and during exercise and highlight the mechanistic insights that can be derived from studying E_A/E_LV.

aging and cardiovascular (CV) diseases, such as hypertension and congestive heart failure (HF), modify the structure and function of both the central arteries and the left ventricle (LV). Left ventricular performance is influenced by the arterial load (33), and arterial properties are, in turn, influenced by left ventricular performance (33, 77). Indeed, the interaction between the LV and the arterial system, known as arterial-ventricular coupling (E_A/E_LV), is an important determinant of net CV performance (33) and cardiac energetics (74). In this paper, we will review the concept of E_A/E_LV and its components, effective arterial elastance (E_A; a measure of the net arterial load) and left ventricular end-systolic elastance (E_LV; a measure of left ventricular performance), and we will discuss the novel mechanistic insights E_A/E_LV provides into the effects of aging, hypertension, and HF on CV performance at rest and during exercise.

ARTERIAL-VENTRICULAR COUPLING AND ITS COMPONENTS

Traditionally, arterial load was characterized in the frequency domain as impedance spectra (49, 54), whereas left ventricular performance was characterized in the time domain by indexes of pressure and volume (77). This difference hindered the ability to examine the cross talk, or the interactions, between the LV and the central arteries. Studying these interactions required that novel methods be developed, which assess left ventricular and arterial properties in the same domain.

Effective Arterial Elastance

Sunagawa's pioneering work (77), using a three-element windkessel model in the isolated canine heart, showed that the arterial load could globally be characterized in the time domain as E_A (elastance is the change in pressure for a change in volume). E_A is not a measure of a specific arterial property; rather, it is an integrative index that incorporates the principal elements of arterial load, including peripheral vascular resistance (PVR), total arterial compliance, characteristic impedance, and systolic and diastolic time intervals (77). E_A can, therefore, be considered a measure of the net arterial load that is imposed on the LV. E_A is determined from pressure-volume (P-V) loops as the negative slope of the line joining the end-diastolic volume (EDV) and end-systolic pressure (ESP) points (Fig. 1) and can be approximated by the ratio of ESP to stroke volume (SV) (77). ESP can be estimated from the formula [2 × (systolic BP + diastolic BP)]/3 (37), where BP is blood pressure, or from the formula ESP = 0.9 × systolic BP (37). Kelly et al. (37) showed that E_A measured invasively as ESP/SV closely approximated the arterial load obtained from aortic input impedance and arterial compliance data based on a three-element windkessel model [E_A = 1.0 × E_A(windkessel) − 0.11, SE of estimate = 0.12; coefficient of variation = 0.97, P < 0.001] (37). One limitation of the three-element windkessel model is that it does not include the effects of the reflected pressure waves, which originate from areas of major impedance mismatches or major bifurcations. These reflected waves arrive earlier in the cardiac cycle when the velocity of the pulse wave is increased, as occurs with aging or with hypertension, and they can substantially augment the systolic load on the heart. However, the net effects of the reflected waves are functionally accounted for in P-V loops (37). Thus E_A can be considered as a surrogate measure of aortic input impedance whose advantage is that it can be related to measures of E_LV, thus allowing the study of arterial and ventricular interactions (see below).

Fig. 1.

Ventricular pressure-volume diagram from which effective arterial elastance (E_A) and left ventricular (LV) end-systolic elastance (E_LV) are derived. E_A represents the negative slope of the line joining the end-diastolic volume (EDV) and the end-systolic pressure (ESP) points. E_LV represents the slope of the end-systolic pressure-volume relationship passing through the volume intercept (V0). The shaded area represents the cardiac stroke work (SW), and the hatched area represents the potential energy (PE). LV ESP is the LV pressure at the end of systole. EDV is the LV volume at the end of diastole. End-systolic volume (ESV) is the LV volume at the end of systole. Stroke volume (SV) is the volume of blood ejected by the LV with each beat and is obtained from subtracting ESV from EDV. BP, blood pressure; EF, ejection fraction; PVA, pressure-volume area.

Left Ventricular End-Systolic Elastance

Sagawa et al. (67), in an isolated canine heart model, showed that left ventricular contractility (or end-systolic stiffness) could be indexed by E_LV. E_LV is determined from the slope of the end-systolic P-V relationship (Fig. 1), which can be obtained from a series of P-V loops recorded while the preload of the heart is altered. An increase in contractility is depicted by an increase in the slope and a shift in the end-systolic P-V relationship to the left, which allows the ventricle to generate more pressure for a given LV volume. E_LV can be calculated as ESP/[end-systolic volume (ESV) − V0], where V0 is the x-axis volume intercept of the end-systolic P-V relationship (obtained from the linear extrapolation of the end-systolic P-V relationship). The calculation of E_LV assumes that the end-systolic P-V relationship is independent of load, that its slope is linear, and that V0 is insensitive to inotropic influences. Under physiological loading conditions, these assumptions are reasonable approximations (12, 34).

Although E_LV is widely regarded as a load-independent index of left ventricular contractility (66, 67), it is also influenced by the geometric and biochemical properties that underlie left ventricular end-systolic stiffness (9). Thus caution should be exercised in interpreting the significance of an elevated E_LV, particularly when other measures of left ventricular systolic function are normal (32). It is likely that acute changes in E_LV (e.g., with inotropic agents or exercise) reflect acute alterations in left ventricular contractility, whereas baseline values of E_LV represent an index that integrates intrinsic left ventricular contractility as well as the modulating effects of the geometric, structural, and functional properties of the LV (32). E_LV should, therefore, be considered an integrated measure of left ventricular chamber performance that can be related to an integrated measure of arterial load (i.e., E_A).

Arterial-Ventricular Coupling

Importantly, E_A shares common units with E_LV, and their ratio E_A/E_LV is a measure of the interaction between the LV and the arterial system (77). E_A/E_LV is an important determinant of net cardiac performance (33) and cardiac energetics (74). Appropriate matching between the LV and the arterial system at rest results in an optimal transfer of blood from the LV to the periphery without excessive changes in pressure; an optimal or near-optimal stroke work (SW); and energetic efficiency, i.e., the energy consumed by the heart to achieve the required SW (9). In healthy men and women, in the resting state, mean ± SD values of E_A/E_LV, E_A, and E_LV measured invasively are ≈1.0 ± 0.36, 2.2 ± 0.8 mmHg/ml, and 2.3 ± 1.0 mmHg/ml, respectively (18).

Noninvasive Measures of Arterial-Ventricular Coupling and Its Components

The initial studies from Sagawa's group (67, 77–79) were performed in isolated canine hearts, which were instrumented to obtain P-V loops. Subsequently, E_A and E_LV were measured in humans from P-V loops acquired in the cardiac catheterization laboratory (11, 36). However, the invasive nature of these techniques limited their use in humans to a few skilled research groups. Fortunately, noninvasive estimates of E_A and E_LV were subsequently developed, which are discussed below (19, 48, 60, 61, 64).

As noted above, E_A can be calculated as ESP/SV. SV can be readily measured noninvasively (e.g., by echocardiography or gated blood pool scans) (25, 73). Chen et al. (17) found that the calculation of ESP from 0.9 × brachial systolic BP reasonably approximated ESP measured invasively: the correlation coefficient between the two variables was 0.75, and the regression line had a slope of 1.01 (P < 0.0001).

Importantly, the equation E_A = ESP/SV can be algebraically rearranged to show that E_A is proportional to the sum of HR (heart rate) × PVR and [(ESP − mean arterial pressure)/SV], suggesting that the main determinants of E_A include HR, a resistive component (PVR), and a stiffness component (change in pressure/change in volume) (20).

For the noninvasive assessment of E_LV, two approaches have been developed. The first is based on the equation noted above (E_LV = ESP/ESV − V0), and assumes that V0 is negligible compared with ESV. Thus, by measuring ESV noninvasively (e.g., by echocardiography or by gated blood pool scans) and by calculating ESP as 0.9 × systolic BP, E_LV can be noninvasively estimated.

The second noninvasive approach to assess E_LV attempts to derive the slope of the end-systolic P-V relationship without altering the loading conditions of the heart. This approach takes advantage of the finding that, under physiological loading conditions, the time-varying elastance curves, from which the end-systolic P-V relationships are derived, are fairly independent of loading conditions, left ventricular contractile state, and HR, particularly when they are normalized to peak amplitude (end-systolic stiffness) and time to peak amplitude (70). As a result, the end-systolic P-V relationship can be estimated from a single heartbeat (70). This method requires the measurement of systolic and diastolic BPs, ejection fraction (EF), and SV, preejection period, and total systolic ejection period on Doppler echocardiography. This single-beat elastance approach has been validated against invasively measured E_LV with a correlation coefficient of 0.81 (P < 0.001) [single-beat E_LV = 0.78 × E_LV(invasive) + 0.55, SE of estimate = 0.6] (17).

From these noninvasive determinations of E_A and E_LV, the E_A/E_LV ratio can be calculated. The noninvasively obtained values of E_A/E_LV closely approximate those obtained invasively (21, 65). E_A/E_LV is inversely related to EF [E_A/E_LV ≈ (1/EF) − 1] (20). The advantage of E_A/E_LV over EF is that examining the components of E_A/E_LV allows us to evaluate whether alterations in E_A/E_LV are due to alterations in arterial properties, left ventricular properties, or both. Importantly, the ability to noninvasively assess E_A/E_LV has expanded the range of clinical trials and the scope of conditions in which E_A/E_LV could be investigated. For example, E_A/E_LV has been examined in epidemiological studies (61) and in studies evaluating the effects of exercise (50).

Scaling of Effective Arterial Elastance and Left Ventricular End-Systolic Elastance

Because body size is an important determinant of CV structure and function, adjusting for body size (or for left ventricular mass) is often needed when CV variables are being characterized or compared among groups (26). However, there is no consensus as to the best strategy for normalizing E_A or E_LV. Both SV and ESV vary directly with body size. Thus some investigators have scaled SV to body surface area (BSA) (5, 50, 65), whereas others have scaled E_A (i.e., ESP/SV) to BSA (61). For E_LV, a broader spectrum of scaling variables has been utilized, including BSA (29, 47), BSA to the power 1.19 (24), body mass (28), left ventricular mass (7), left ventricular EDV (30), and left ventricular mass/EDV (42). For the most part, the scaling technique has consisted of dividing E_LV by the variable of interest. However, this approach does not necessarily provide adequate adjustment for the variable of interest, because E_LV may not necessarily be linearly related to the variable. Instead, a preferred approach might be the allometric scaling method (6, 26, 52), whereby E_LV is divided by the variable of interest raised to a scalar exponent, which is derived from an equation that linearly relates the two. In the following sections, CV variables that are normalized to body size will be depicted by the suffix “I” (e.g., E_AI and E_LVI).

Arterial-Ventricular Coupling and Its Components During Exercise

Exercise provides a powerful tool to examine the response of the CV system to stress and to assess its functional reserve. During exercise, the goal of the CV system is to prioritize cardiac efficacy over energetic efficiency (50). It uses a complex combination of alterations in HR, LV contractility, preload (EDV and SV), and afterload to ensure adequate blood supply to the tissues.

Few studies have examined the changes in E_A/E_LV and its components during exercise. In adult dogs, Little and Cheng (43) reported that E_A/E_LV decreased by ∼25% from rest to submaximal exercise. In healthy human subjects undergoing supine cycle ergometry, Asanoi et al. (3) observed that E_A/E_LV decreased by 35 and 54% at workloads corresponding to 30% below and 30% above the anaerobic threshold, respectively; and Chantler et al. (15) found that E_AI/E_LVI decreased by ≈65% (from an average of 0.58 to 0.34, and 0.52 to 0.27 in men and women, respectively) from rest to peak exercise, which was attributed to a larger increase in E_LVI (from an average of 4.3 to 13.2, and 4.7 to 15.5 mmHg·ml⁻¹·m⁻² in men and women, respectively) vs. E_AI (from an average of 2.3 to 3.2, and 2.3 to 2.9 mmHg·ml⁻¹·m⁻² in men and women, respectively) during exercise.

The change in E_A during exercise results from a complex interplay among the changes in BP, PVR, arterial stiffness, and HR. At rest, the sensitivity of E_A to a change in PVR/HR is approximately three times higher than to a similar change in arterial stiffness (16, 69). In contrast, during exercise, arterial stiffness has a progressively and intensity-dependent greater impact on E_A than PVR (57, 58), such that E_A (on average) increases during exercise, paralleling the increase in arterial stiffness, despite a reduction in PVR (57).

Some of the methodological issues pertaining to the noninvasive assessment of E_A/E_LV and its components during exercise should be highlighted. At rest, the single-beat elastance approach is regarded as the preferred noninvasive method to measure E_LV (17). However, the single-beat elastance approach may be technically challenging during exercise, because of the difficulties in measuring cardiac volumes and systolic time intervals with echocardiography during exercise (63). If imaging modalities other than echocardiography are available for measuring ESV and SV during exercise (e.g., gated blood pool scans), then estimating E_LV from the formula E_LV = ESP/ESV may be more feasible. However, the noninvasive measurement of E_LV from ESP/ESV has, in turn, its own limitations. 1) It assumes that V0 is negligible compared with ESV. V0 has not been well characterized in humans, particularly during exercise. In healthy adult dogs, Little and Cheng (43) found that, although the absolute values of V0 did not significantly change during exercise, V0 as a percentage of ESV increased by 9%. In contrast, in healthy subjects, Starling (74) found that V0, both in absolute values and as a percentage of ESV, did not appreciably change during dobutamine infusion. 2) The formula used to noninvasively estimate ESP (ESP = 0.9 × systolic BP) has not been validated during exercise. In this regard, methodologies that use radial applanation tonometry may be of help as they allow noninvasive and accurate estimations of central SBP at rest and during exercise, at least in the supine position and at low intensities of exercise (71, 72).

In the subsequent sections, we will examine the effects of age, hypertension, and HF on E_A/E_LV and its components, both at rest and during exercise, and we will highlight important sex differences when appropriate. Advancing age, hypertension, and HF all result in alterations in E_A/E_LV and its components, both at rest and during exercise. These changes are summarized in Table 1.

Table 1.

Summary of the alterations in arterial-ventricular coupling and its components with age, hypertension, and heart failure at rest and during exercise

	At Rest					At Peak Exercise					Reserve
		EA/ELV	EA	ELV	EA/ELV	EA	ELV	EA/ELV	EA	ELV
Old vs. young
Men	↔	↑	↑	↑	↔	⇊	↓	↔	↓
Women	↓	↑	⇈	↑	↑	↓	↓	↔	↓
HTN vs. NT
Men	↔	↑	↑	↔	↑	↑	↔	↔	↔
Women	↓	↑	⇈	↔	↔	↔	↓	↓	↔
SHF vs. healthy controls
Men and Women	↑	↑	⇊	↑	⇊	⇊	⇊	⇊	⇊
HFpEF vs. healthy controls
Men and Women	↔	↑	↑	?	?	⇊	?	?	⇊

Arrows indicate directionality of the comparison. E_A, effective arterial elastance; E_LV, LV end-systolic elastance; E_A/E_LV, arterial-ventricular coupling; HTN, hypertensive; NT, normotensive; SHF, systolic heart failure; HFpEF, heart failure with a preserved ejection fraction.

AGE

Arterial-Ventricular Coupling and Its Components at Rest

Aging influences the structural and functional properties of both the arterial system and the LV. With advancing age, the central arteries dilate, and their walls become thicker and stiffer (51). E_A increases with advancing age in individuals without known CV diseases (18, 21, 65). For example, in 2,042 individuals from Olmsted County, Minnesota, Redfield et al. (61) observed an age-associated increase in E_AI in both men (r = 0.28, P < 0.001) and women (r = 0.28, P < 0.001) (Fig. 2). Similar results were obtained in a subset of the population (n = 623) that was free from CV diseases (61). The age-associated increase in E_AI was principally attributed to the age-associated increase in arterial stiffness, because pulse pressure increased with age, whereas PVR and HR did not change.

Fig. 2.

The association between age and E_A indexed to body surface area (E_AI) (A) and E_LV (Ees = E_LV) (B) in men (solid line) and women (dashed line) at rest. Pearson correlation coefficients, and probability values for each association are shown. With advancing age, both E_aI and Ees increase. However, the increase in Ees with age is significantly greater in women than men. Furthermore, E_aI and Ees are higher in women vs. men at all ages. [Modified from Redfield et al. (61).]

Advancing age is also associated with alterations in left ventricular structure and function. Most notably, there is a reduction in myocyte number in men, but not in women, and there is an increase in left ventricular wall thickness and collagen deposition (40, 55). These alterations are accompanied by an increase in resting E_LV with advancing age (18, 21, 61). For example, healthy 75-yr-old men and women had a ≈10% (P < 0.01) and ≈15% (P < 0.01) higher resting E_LV, respectively, compared with their 55-yr-old counterparts (Fig. 2) (61). These findings are unlikely to be due to differences in left ventricular chamber size, as similar results were obtained when E_LV was normalized to left ventricular EDV (61). Sex-related differences were also noted, whereby E_LV was higher in women than in men at all ages and increased more steeply with advancing age (61). Other markers of left ventricular systolic function, such as stress-corrected fractional shortening (61), circumferential end-systolic stress/ESV index (8), and preload-recruitable SW (30), have also been shown to be higher in women than in men. In addition, female rats have an enhanced systolic function compared with male rats (14, 81), suggesting that the higher E_LV in women may reflect a higher left ventricular systolic function.

Because a higher E_LV represents a steeper slope of the end-systolic P-V relationship, it results in an increased sensitivity of systolic pressure to changes in volume (36). Kass et al. (35) reported, in an isolated canine heart model, larger reductions in E_LV after a myocardial infarction in hearts with a higher resting E_LV. This greater mechanical vulnerability to an ischemic insult may help to explain why older individuals, who have an increased E_LV [particularly older women and hypertensive (HTN) subjects, see below] experience worse outcomes following a myocardial infarction.

Even though E_A and E_LV both increase with advancing age, their ratio, E_A/E_LV, remains relatively unchanged across the age spectrum in men (18, 21, 50), suggesting that the increases in E_A and E_LV are matched. In contrast, E_A/E_LV declines slightly with advancing age in healthy women (61), reflecting a disproportionate increase in E_LV compared with E_AI. This suggests a greater impact of aging on ventricular vs. arterial properties in women compared with men. Furthermore, women have a higher resting E_A, and an even higher resting E_LV than men (50, 61), but a lower resting E_A/E_LV, suggesting that the sex differences in E_A/E_LV are predominantly related to sex differences in ventricular properties. These sex differences in E_A and E_LV persist even after adjusting for age, body size, and HR (61).

Arterial-Ventricular Coupling and Its Components During Exercise

It is well known that there is an age-associated deficit in CV performance at peak exercise that is evident from age-associated reductions in HR, EF, and oxygen consumption at peak exercise (41). Najjar et al. (50) reported that the decline in E_AI/E_LVI from rest to peak upright bicycle exercise is blunted in older (>60 yr) compared with younger (<40 yr) healthy men and women (Fig. 3A). This blunted E_AI/E_LVI response was attributed to a blunted increase in E_LVI during exercise in older compared with younger individuals (Fig. 3C), as the increase in E_AI during exercise did not differ between young and older subjects (Fig. 3B) (50). Consequently, advancing age is associated with a smaller E_AI/E_LVI reserve (E_AI/E_LVI peak − E_AI/E_LVI rest) and E_LVI reserve (E_LVI peak − E_LVI rest) (50). Thus examining E_A/E_LV and its components during exercise provides additional insights into the age-associated deficits in EF reserve and suggests that it is due to age-associated deficits in left ventricular contractile reserve, more so than to age-associated differences in arterial properties.

Fig. 3.

Arterial ventricular coupling (E_aI/E_LVI) (A), E_aI (B), and E_LV indexed to body surface area (E_LVI) (C) in men (dashed lines) and women (solid lines) <40 yr of age (triangles) and >60 yr of age (squares) in the supine and seated positions, at 50% of maximal workload, and at peak exercise. E_aI/E_LVI decreases during exercise in both young and older men and women (P < 0.0001). However, older men and women have a blunted decline in E_aI/E_LVI (P < 0.001). E_AI increases during exercise in both young and older men and women (P < 0.0001). At maximal exercise, E_AI is greater in older vs. younger women (P < 0.002). In contrast, E_AI does not differ between young and older men. E_LVI increases during exercise in both young and older men and women (P < 0.0001). At maximal exercise, E_LVI is greater in younger vs. older men (P < 0.001) and tended to be greater in younger than older women (P = 0.07). [From Najjar et al. (50).]

Najjar et al. (50) also observed age by sex interactions in E_AI/E_LVI in their normotensive (NT) cohort. In younger (<40 yr) individuals, peak E_AI/E_LVI was lower in men than in women, whereas in older (>60 yr) individuals, peak E_AI/E_LVI was higher in men than in women (Fig. 3A). This was attributed to an age by sex interaction in E_LVI, whereby, at peak exercise, E_LVI was higher in younger men vs. younger women, but did not differ between older men and women (Fig. 3C). As for E_AI, it was higher in both young and old men compared with women (Fig. 3B).

Only one study has examined the effects of exercise training on E_A/E_LV and its components during exercise. In 26 patients with coronary artery disease, Rinder et al. (62) found that 12 mo of aerobic endurance exercise training lead to a 37% increase in E_LV and a 23% decrease in E_A/E_LV during handgrip exercise performed at 30% of maximal voluntary contraction. However, the change in E_A during handgrip exercise remained unaltered after the exercise training.

HYPERTENSION

Arterial-Ventricular Coupling and Its Components at Rest

The prevalence of hypertension markedly increases with advancing age, such that it affects 66% of individuals over 60 yr of age (56). Hypertension is an important risk factor for mortality, stroke, and HF (23, 31). The age-associated changes in arterial and left ventricular structure and function are accelerated in the presence of hypertension. HTN patients exhibit greater carotid wall thickness (2), central arterial wall stiffness (1), and reflected waves (53) than NT subjects, even after adjusting for age. Furthermore, hypertension is associated with LV remodeling and fibrosis (46).

Few studies have examined the impact of hypertension on E_A/E_LV and its components. Cohen-Solal et al. (20) showed that HTN individuals have a 60 and 95% higher E_A and E_LV, respectively, compared with NT controls. Saba et al. (64) found that HTN individuals have a 23 and 29% higher E_A and E_LV, respectively, compared with NT controls. However, E_A/E_LV did not differ between HTN and NT men (20, 64), suggesting that the increases in E_A and E_LV in HTN men was matched. In contrast, E_AI/E_LVI was ≈23% lower in systolic HTN compared with NT women (15), a finding that persisted even after adjusting for age (Fig. 4A). The lower E_AI/E_LVI in systolic HTN women was due to a disproportionate increase in E_LVI (Fig. 4B) compared with E_AI (45 vs. 16%) (Fig. 4C), suggesting an adaptation by these women to limit the impact of systolic hypertension on the vasculature or, alternatively, a more pronounced impact of systolic hypertension on ventricular vs. arterial elastance.

Fig. 4.

E_AI/E_LVI (A), E_LVI (B), and E_AI (C), measured at rest in normotensive (NT) and systolic hypertensive (SH) men and women. E_AI/E_LVI does not differ at rest between NT and SH men due to tandem increases in E_AI and E_LVI in SH vs. NT men. In contrast, resting E_AI/E_LVI is lower in SH women vs. NT women, due to a disproportionate increase in E_LVI vs. E_AI. *P < 0.05, **P < 0.01, ***P < 0.001, comparing NT to SH after adjusting for age. [From Chantler et al. (15).]

Arterial-Ventricular Coupling and Its Components During Exercise

Only one study has examined the effects of systolic hypertension on the changes in E_A/E_LV and its components during exercise. E_AI/E_LVI did not differ between NT and HTN men and women at 50% of maximal workload or at peak upright bicycle exercise (Fig. 5) (15). In men, this was because E_AI and E_LVI were proportionally higher at peak exercise in HTN compared with NT, whereas, in women, this was because E_AI and E_LVI did not differ at peak exercise between HTN and NT. Thus E_AI/E_LVI reserve did not differ between HTN and NT men, but it was 61% lower in HTN compared with NT women, because HTN women have a lower E_AI/E_LVI at rest. This was accompanied by a 60% lower E_AI reserve in HTN compared with NT women (15).

Fig. 5.

The change in E_AI/E_LVI (A), E_AI (B), and E_LVI (C) in NT (solid symbols) and SH (open symbols) men (solid lines) and women (dashed lines). At rest, E_AI/E_LVI is similar between NT and SH men and is lower in SH vs. NT women (P < 0.01). E_aI/E_LVI decreases during exercise in both NT and SH men and women (P < 0.01). There are no differences between NT and SH men and women at 50% maximal workload (MWL) or at peak exercise. At rest, E_AI is higher in SH vs. NT men and women (P < 0.001). E_AI increases during exercise in both NT and SH men and women (P < 0.05). However, only SH men have a higher E_AI at 50% MWL and at peak exercise vs. NT men (P < 0.001), as no differences are found between NT and SH women at 50% MWL or peak exercise. At rest, E_LVI is higher in SH vs. NT men (P < 0.05). E_LVI increases during exercise in both NT and SH men and women. However, only SH men have a higher E_LVI at 50% MWL or at peak exercise vs. NT men (P < 0.001), as no differences are found between NT and SH women at 50% MWL or at peak exercise. [Modified from Chantler et al. (15).]

HEART FAILURE

Arterial-Ventricular Coupling and Its Components at Rest

HF is a syndrome that is characterized by an inability of the heart to pump a sufficient amount of blood to meet the demands of the metabolizing tissues, or can do so only at the expense of elevated filling pressures (30a). We will only provide a brief summary of the alterations in E_A/E_LV and its components in HF, as insights they provide into the pathophysiology of HF and response to treatments have recently been reviewed (9, 27).

Systolic HF.

HF patients with systolic dysfunction are characterized by a diminished resting EF and left ventricular contractility (5). Systolic HF patients have a downward and rightward shift of the end-systolic P-V relationship (Fig. 6) and thus a reduced E_LVI (range 0.6–2.6 mmHg·ml⁻¹·m⁻²) (5). Systolic HF patients have an augmented E_AI (range 1.7–3.7 mmHg·ml⁻¹·m⁻²) due to a decrease in SVI and to increases in HR and PVR (5). The increase in E_AI and decrease in E_LVI result in marked, up to three- to fourfold, increases in E_A/E_LV (range 1.3–4.3) (5, 68). This suboptimal coupling reflects diminished CV performance and efficiency of the failing heart.

Fig. 6.

Pressure-volume relationships comparing the slope of the end-systolic pressure-volume relationship (E_LV) and the slope of the line joining the ESP and EDV points (E_A) between healthy controls (left) and systolic heart failure (HF) patients (right). Systolic HF patients have a downward and rightward shift of the end-systolic pressure-volume relationship, reflecting a reduced LV contractility. In addition, systolic HF patients have a higher E_A than healthy controls. Thus, systolic HF patients have a higher resting arterial-ventricular coupling ratio than healthy controls.

HF with a preserved EF.

Approximately 40% of patients with HF have an EF ≥ 50% (59). These individuals are classified as HF patients with a preserved EF (HFpEF). HFpEF is more prevalent in older women than men (39), and systolic HTN is the most common risk factor for this syndrome (39). HFpEF patients have an 18 and 20% higher resting E_A and E_LV, respectively, compared with healthy controls (42). One study found that patients with HFpEF had a disproportionately higher E_LV than E_A compared with young and old NT controls, and thus a lower E_A/E_LV (36). However, in a larger sample population, Lam et al. (42) noted that patients with HFpEF had matched increases in E_A and E_LV; thus their E_A/E_LV did not differ from E_A/E_LV of NT controls or individuals with HTN but without HF. These conflicting findings may be due to differences in the characteristics of the control groups with which the HFpEF patients were compared, or to the heterogeneity among patients classified as having HFpEF. For example, Maurer et al. (45) noted that some patients with HFpEF who were NT had values of E_A and E_LV that did not differ from those of healthy NT controls.

Arterial-Ventricular Coupling and Its Components During Exercise

Systolic HF.

The baseline alterations in E_A/E_LV and its components in patients with systolic HF are also evident during exercise (22). Indeed, E_A, E_LV, and E_A/E_LV do not appreciably change during exercise in systolic HF patients, whereas they markedly change in healthy subjects (22). Thus the limited capacity of systolic HF patients to augment their CV function during times of stress (such as exercise) involve marked deficits in both LV and arterial elastance reserves.

HF with a preserved EF.

To date, there are no studies that have examined the change in E_A/E_LV or E_A during exercise in HFpEF. Borlaug et al. (10) examined the change in E_LV and EF in HFpEF during upright bicycle exercise. Compared with HTN controls with LV hypertrophy, HFpEF patients had a threefold smaller increase in E_LV and a reduced ability to lower their PVR and increase their HR during exercise (10). They also had a 50% smaller increase in EF during exercise. As EF is inversely related to E_A/E_LV (20), this suggests that the change in E_A/E_LV during exercise in HFpEF may also be severely blunted. Since female sex and systolic hypertension are risk factors for HFpEF (39), and as the pathophysiology of HFpEF involves a limited CV reserve (38), the diminished E_A/E_LV reserve observed in systolic HTN women (15) suggests that they may be exhibiting signs of subclinical (Stage B) HF. This raises the possibility that they may be on a trajectory to progressive exercise intolerance and perhaps functional limitations.

CARDIAC ENERGETICS

E_A/E_LV is an important determinant of cardiac energetics (13, 74). From P-V loops, total mechanical energy can be quantified as the pressure-volume area (PVA), which is composed of SW (shaded area in Fig. 1) and potential energy (hatched area in Fig. 1) (76). PVA and myocardial oxygen demand are linearly related (75); therefore, PVA has been used as an index of the oxygen cost of performing a given SW (74). Thus the amount of SW performed by the LV and the amount of oxygen the LV consumes while performing this SW vary with the loading conditions imposed on the LV and the contractile state of the LV (76).

E_A/E_LV in the range of 0.6–1.2 in the resting state (9) is thought to represent the optimal interaction between the arterial and LV systems (74) and confers the near-optimal balance between mechanical efficacy and energetic efficiency (74). Interestingly, this range is also observed across species (44, 80), suggesting that it has been conserved through mammalian evolution.

During exercise, energetic efficiency is sacrificed in favor of cardiac efficacy to augment CV performance. To date, very few studies in humans have examined the impact of age, hypertension, and HF on cardiac energetics during exercise. Chantler et al. (15) found that HTN women, but not men, had a higher SWI and PVA at peak exercise compared with NT controls. Furthermore, an exaggerated increase in E_A during exercise, as noted in older women (50) and in women with hypertension (60), increases the pulsatile load on the LV, which may increase the cardiac energy costs to provide blood flow by raising myocardial oxygen consumption.

CONCLUSIONS

Examination of the alterations in E_A/E_LV and its components, E_A and E_LV with aging, hypertension, and HF at rest and during exercise is often overlooked, yet it can yield mechanistic insights into the pathophysiology of these conditions and how they differ between men and women. Future studies should examine the effects of lifestyle changes (e.g., physical conditioning), other medical conditions (e.g., diabetes), and the various classes of medications used to treat hypertension and HF on E_A/E_LV and its components, E_A and E_LV. Furthermore, longitudinal studies are needed to evaluate whether alterations in E_A/E_LV, E_A, and E_LV provide any prognostic information for adverse outcomes, such as HF.

GRANTS

This work was supported (in part) by the Intramural Research Program of the National Institutes of Health, National Institute on Aging.