INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

AGE-ASSOCIATED CHANGES in cardiovascular reserve mechanisms occur even among healthy individuals (15). The most salient cardiovascular changes with aging occur during stress and include a reduction in maximum heart rate (HR) and increases in left ventricular (LV) afterload and preload (9). An increase in the vascular component of LV afterload with aging consists of four components (12, 24, 25, 32): a modest increase in systemic vascular resistance (9, 25), a more marked reduction in aortic compliance (25), early return of pulse wave reflections from the periphery of the vascular system to the aortic root due to stiffening of the arterial walls and an increase in pulse wave velocity (PWV), and increased inertance due to a larger mass of blood requiring acceleration at the onset of ejection because of aortic dilatation (11, 15). The increase in the vascular component of afterload with aging, together with a decline in maximum intrinsic contractility (34) and in adrenergic augmentation of contractility (10, 16), results in a smaller reduction in LV end-systolic volume (ESV) and a blunted increase in the LV ejection fraction (EF) during vigorous exercise in older vs. younger individuals (2, 9, 27).

Previous studies of this cohort in our laboratory (9, 10, 30) have demonstrated increased LV end-diastolic volume (EDV) in older vs. younger men at seated rest and a greater increase in EDV in older than in younger individuals of both genders during exhaustive, upright exercise (9). However, the stroke volume (SV) elicited by the Frank-Starling mechanism in healthy older persons is limited because of failure of the LV to empty as completely in older persons as it does in younger ones (9, 15), resulting in an age-associated increase in ESV during exercise. Thus, during vigorous exercise, the increased LV vascular afterload and LV preload in older individuals, combined with reduced LV contractility (10, 34) and HR, cause the heart to become dilated relative to the heart in younger individuals throughout the cardiac cycle (9, 15). This cardiac dilatation increases LV wall stress throughout the cardiac cycle in older vs. younger individuals and may contribute to the reduction in maximum LV ejection capacity with aging (28).

The goal of this study was to examine 1) whether unloading the heart in healthy older persons with sodium nitroprusside (nitroprusside), a mixed vasodilator without direct inotropic effects (22, 23, 26), would improve their LV function during vigorous aerobic exercise to that attained by healthy, unmedicated, younger individuals and 2) whether this effect would permit an enhanced maximum cardiac output and maximum exercise capacity in these older individuals.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Study Population

Study Protocol

Carotid-Femoral PWV

Noninvasive Determination of the Carotid Artery AGI

View larger version (16K): [in this window] [in a new window]   Fig. 1.   A representative carotid pulse pressure (PP) recording in an older subject before and during nitroprusside infusion. Inset: method for calculating augmentation index (AGI).

Rest and Exercise Cardiac Volumes

LV volumes were determined by standard methods (20). Briefly, end-diastolic count rate was obtained from a large manually drawn region of interest and was corrected for background activity with the use of a region of interest drawn lateral and inferior to the LV in the end-systolic frame. Attenuation correction was determined individually in each subject by the use of the opposed static view to measure the attenuation distance from a chest wall marker to the count center of the LV, with the assumption that the linear attenuation coefficient was equal to that of water. A blood sample was drawn after cessation of exercise, counted with the same camera-collimator system used for the scintigraphic study, and corrected for the time delay between the scintigram and the counting of the blood sample. LV EDV was obtained from the ratio of the attenuation-corrected end-diastolic count rate from the gated cardiac images to the count rate per milliliter from the sample of venous blood drawn 5-10 min after the completion of exercise. EF was calculated, after background subtraction, by a validated, fully automated algorithm. ESV was calculated from the measured EF and EDV. LV volumes calculated by this method have been validated against invasive measurements (20).

All radionuclide-derived LV volumes were normalized to body surface area, yielding their respective indexes: EDV index (EDVI), ESV index (ESVI), SV index (SVI), and cardiac index (CI). The following blood pressure measurements were measured or derived: SBP, diastolic blood pressure (DBP), MBP [calculated as (2 · DBP + SBP)/3], pulse pressure (PP), and end-systolic pressure [ESP, estimated as (2 · SBP + DBP)/3] (14). Total systemic vascular resistance (TSVR) was calculated as the ratio MBP/CI, and stroke work index (SWI) was calculated as the product of SVI and SBP.

Exercise Protocol

O₂

Statistical Analysis

Furthermore, the percentage of older vs. younger subjects performing aerobic exercise once or twice weekly was similar (36% vs. 44%). Forty-one percent of the younger and 38% of the older group reported no exercise habits.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Baseline Characteristics and Measurement Reproducibility

Effects of Nitroprusside on Hemodynamics and PWV and AGI in the Supine Resting Position

1

1

The central PP was calculated by calibrating the applanation tonometry, with the mean and diastolic pressures from the applanation tonometry waveform set equal to the mean and diastolic brachial artery measurements. The calculated central systolic pressure was 118.40 ± 4.26 and 106.37 ± 4.32 mmHg before and during infusion, respectively. The difference in the amplitude of the calculated central systolic pressure and the cuff pressure in each person varied directly with the amplitude of the cuff pressure. The calculated central PP was 40.79 ± 2.70 and 40.01 ± 3.2 mmHg before and after drug, respectively. The calculated central ESP was 104.80 ± 3.73 and 93.03 ± 3.38 mmHg before and during drug infusion, respectively. The SBP/ESVI calculated from the calibrated tonometry measurement was 5.07 ± 0.42 and 8.43 ± 0.095 before and during drug infusion, respectively. With the use of the calculated central pressure, the estimated SWI was 6,111.85 ± 343.63 and 4,971.77 ± 265.18 before and during drug infusion, respectively.

Effects of Nitroprusside on Hemodynamics at Seated Rest, During Submaximal Exercise Workloads, and at Maximum Exercise

View larger version (36K): [in this window] [in a new window]   Fig. 2.   Systolic blood pressure (A), diastolic blood pressure (B), pulse pressure (C), and total systemic vascular resistance (D) at seated, upright rest, at common submaximal workloads, and at maximum effort (Max). * Statistical significance for paired t-test at rest and at maximum workload. Number of subjects depicted at rest and at maximum workload is 16. ANOVA for repeated measures indicated that drug effect on systolic and diastolic blood pressures at rest was the same at all common workloads. ANOVA for repeated measures indicated that total systemic vascular resistance showed a significant drug-workload interaction.

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Cardiac index (A), heart rate (B), and cardiac volumes [end-diastolic volume index (EDVI), end-systolic volume index (ESVI), and stroke volume index (SVI); C] at seated, upright rest, at common submaximal workloads, and at maximum effort, regardless of absolute workload, in presence (control, C) and absence of nitroprusside (NP) infusion. In subset of subjects who achieved common submaximal workload of 100 W, drug effect to lower ESVI was similar at rest and at all submaximal workloads, whereas drug-workload interactions on EDVI and SVI were observed. * Statistical significance for paired t-test at rest and at maximum workload. ANOVA for repeated measures was used to determine statistical significance.

View larger version (22K): [in this window] [in a new window]   Fig. 4.   A: ejection fraction at seated, upright rest, at intermediate common submaximal workloads, and at maximum effort. ANOVA for repeated measures indicated that magnitude of drug effect was similar at rest and at all common submaximal workloads. B: ventricular function, depicted as stroke work index vs. EDVI relationship at upright, seated rest and during exercise in presence and absence of nitroprusside. Relationship is shifted leftward and downward with nitroprusside, indicating a smaller EDVI and lower stroke work index at any exercise load. * Statistical significance for paired t-test at rest and at maximum workload.

Seated upright rest. At rest in the sitting position, SBP (Fig. 2A), DBP (Fig. 2B), MBP (Fig. 2C), and TSVR (Fig. 2D) were reduced by nitroprusside, as were EDVI and ESVI (Fig. 3C); EF significantly increased (Fig. 4A), whereas HR (Fig. 3B), SVI (Fig. 2A), and CI (Fig. 3A) were not changed significantly. Thus the effect of nitroprusside in the upright position was essentially the same as in the supine position.

Seated upright exercise. In older persons the maximum exercise work rate during drug infusion was not different from that during placebo infusion (Table 3). O₂ and the arteriovenous O₂ difference during graded submaximal exercise and at maximum effort did not significantly differ during control and nitroprusside infusion. At maximum effort, O₂ was 1.59 ± 0.21 and 1.48 ± 0.18 l/min and the arteriovenous O₂ difference was 0.11 ± 0.2 and 0.11 ± 0.01 ml during control and nitroprusside drug infusion, respectively.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Substantial limitations of cardiovascular reserve occur with aging even in healthy individuals. In rigorously screened, healthy, older vs. younger sedentary individuals, a reduction in aerobic work capacity during maximal upright cycle exercise is accompanied by LV dilatation at end diastole and a reduction in maximal exercise HR. The peak exercise SV in older individuals is maintained at the level of their younger counterparts by a larger exercise EDV (9). However, ESV does not decrease as much in older as in younger persons, and this results in a reduction in EF during exercise in the older group: LV dilatation throughout the cardiac cycle occurs during vigorous exercise in healthy older persons, and a given SV is achieved at a higher LV wall stress. Although direct pressure measurements have not been made, a higher LV EDP associated with exercise-induced LV dilatation in older persons also likely occurs and may be another factor that limits their exercise capacity.

An understanding of the mechanisms underlying the age-associated cardiac dilatation during exercise requires an examination of age-associated changes in HR, ventricular preload, afterload, and contractility (31) and the interplay among them (17). The vascular component of afterload greatly influences cardiac pump function. The steady and pulsatile vascular loads increase with aging and limit the heart's ability to eject blood. Although the steady component of vascular load (indexed by TSVR) increases on the order of 20% from the second to the sixth decade of life, pulsatile load increases 140% over the same age span (25, 32) and relates to the stiffening process that takes place in the arterial walls of large- and medium-sized vessels secondary to the progressive loss of elastin fibers and their replacement by less distensible collagen fibers and calcium (15). Arterial stiffening increases the pulsatile vascular load: directly, by reducing arterial compliance, and indirectly, by increasing PWV. The latter accelerates the return of wave reflections from the periphery of the arterial system to the aortic root. The early return of wave reflections further increases systolic pressure in the central arteries and is responsible for the late systolic pressure peak observed in the central pressure waves of older but not younger individuals (Fig. 1). Dilatation of conduit arteries with advancing age also imposes an age-associated increase in the inertance component of vascular afterload. In addition to these increased vascular components of afterload, the increased LV size at end diastole, attributable in part to a longer LV diastolic filling period accompanying a reduced HR (10) and throughout the cardiac cycle, in older individuals during exercise increases the cardiac component of LV afterload in older individuals. Reducing total LV load, i.e., cardiac and vascular components, therefore, might be expected to facilitate ventricular ejection and improve cardiac function in older individuals during exercise.

Vascular stiffness and, in particular, early wave reflection can be dramatically reduced by vasodilators (6, 8, 13, 19, 33). To assess the effect of unloading the older heart, we infused sodium nitroprusside, a balanced vasodilator, at a rate sufficient to lower MBP by ~10% in healthy, older, sedentary community-dwelling subjects. As shown in the supine position (Table 2), this intervention significantly reduced systemic vascular resistance, the AGI and PWV indexes of pulsatile load, and EDVI and ESVI and increased the LV EF. Similar drug effects on cardiac volumes (Fig. 3) and EF (Fig. 4A) occurred at rest in the upright position. In older patients with dilated cardiomyopathy, a reduction in PWV similar in magnitude to that in the present study was observed during low-dose nitroprusside infusion (6). The reduction in PWV in that and in the present study is probably related to a reduction in arterial stiffness and pressure, inasmuch as PWV is known to be pressure dependent (1, 5, 6). However, the relationship between blood pressure and PWV, albeit linear, is steeper in older than in younger individuals (6). Thus, for the same reduction in blood pressure, the potential benefits of pulsatile load reduction are greater in older individuals (6). The most striking effect of nitroprusside was on AGI, which was reduced by 94%. Thus, at supine rest, a modest pharmacological vasodilatation in healthy older individuals markedly attenuates the age-associated increase in pulsatile arterial load and reduces heart size throughout the cardiac cycle.

The present study is the first to demonstrate improved exercise LV function during pharmacological vasodilatation in healthy older adults. The drug effect is manifested by a significant reduction in cardiac volumes (Fig. 3), an augmentation of LVEF (Fig. 4A), and an improvement in cardiac function depicted by a leftward shift in the SWI vs. EDVI relation. EDVI and ESVI, and thus heart size throughout the cardiac cycle, were reduced during exercise (Fig. 3). However, because of the approximately similar reductions in EDVI and ESVI, SVI was not appreciably changed (Fig. 2).

Technical considerations preclude noninvasive assessment of vascular stiffness indexes during exercise. If reflected pulse waves were to reappear in the central arteries during exercise, the systolic stress integral and ESP would be greater than if the AGI were to remain low. However, it is likely that a reduction in arterial stiffness by the drug persisted during exercise for the following reasons. The magnitude of the reduction in SBP by nitroprusside throughout exercise is similar to that observed at rest. The reduction in SBP by nitroprusside is not caused by reduced LV contractility, as evidenced by higher EF and SBP/ESVI before than during drug infusion. The nonsignificant 4 ml/m² reduction in SVI by nitroprusside relative to placebo at peak exercise is identical to that seen at rest and is therefore unlikely to contribute substantially to the decline in SBP. Although the reduction in DBP during exercise by nitroprusside vs. placebo could itself contribute to a lower SBP if SV and contractility were unchanged, this reduction in DBP during exercise is virtually identical to that observed at rest. Thus, although direct proof of reduced large artery stiffness by nitroprusside during exercise is lacking, the parallel changes in the measured determinants of SBP at rest and at peak exercise during drug infusion strongly suggest that arterial stiffness was reduced to a similar extent during exercise by the drug.

Whereas in the absence of drug, older individuals maintain the same peak exercise SVI as their younger counterparts, but with a larger ESVI and reduced EF, nitroprusside infusion permits a higher EF from a smaller EDVI, without compromising SVI or maximum exercise capacity. In fact, the LV EF and ESVI at maximum exercise in older persons during drug infusion do not differ from those in a young, unmedicated BLSA cohort (Table 3) (9, 30). The finding that EF remains higher and the ESVI remains lower during drug infusion than on placebo at maximum exercise, despite the absence of a change in TSVR (Fig. 2D), points to a greater role of pulsatile afterload or preload reduction in the improvement of EF and ESV parameters. The reduction in EDVI induced by nitroprusside appears, in part, to be attributable to a reduction in venous return, because a pure dilatation of the arterial system is not likely to produce a significant change in cardiac chamber filling, since the total volume that can be displaced into the arterial capacitance per se is rather small. It is certainly possible that the reduction in EDVI effected by nitroprusside was due, in part, to the known increase in venous capacitance induced by nitroprusside (21). The fact that a significant decline in EDVI occurred at higher levels of exercise during drug infusion, when TSVR was similar to placebo levels, also suggests a direct drug effect on preload.

As shown in Fig. 4B, nitroprusside caused a leftward shift in the SWI-EDV relation; thus any given SWI was achieved from a smaller EDVI. Additionally, the estimated ESP, which correlates highly with that measured directly (14) and reflects the lumped vascular elastance at end systole, was reduced at any given ESVI. Although it is thus plausible to assume that LV wall stress was reduced by the drug, the latter and some of its determinants, e.g., LV wall thickness and the time course of pressure development, were not measured in the present study. It might be argued that the overall augmentation of LV pump performance during exercise with nitroprusside infusion is attributable, in part, to a sympathetic reflex-mediated increase in LV contractility associated with the reduction in arterial pressure. However, if it were of a substantial magnitude, a reflex increase in LV function would be accompanied by a significant increase in HR, but only a minimal increase in HR occurred in the presence of drug at upright rest or during low-level exercise, and even this minimal increase was blunted as exercise workload increased. In fact, the HR at maximum exercise was less in the presence of nitroprusside than with placebo. Additionally, any difference in reflex sympathetic activity due to the arterial pressure reduction elicited by the drug at rest would be expected to lessen as the exercise workload progresses, because dynamic exercise per se elicits sympathetic and other reflexes to augment cardiovascular function. Additionally, reduced ESVI and increased EF observed at rest during nitroprusside infusion persisted throughout exercise, even at the maximum workload. Such an effect cannot be attributable to reflex cardiac stimulation caused by the drug. Finally, the slope of the change in EDV relative to ESV from rest to peak exercise, which shifts with increases in LV contractility during exercise (28), was unchanged by nitroprusside.

Despite the reduction in cardiac size and improvement in LV function in older persons during nitroprusside infusion, maximum work capacity during this type of cycle exercise did not appreciably change and remained less than in their unmedicated younger counterparts (Table 3). The lack of an effect of nitroprusside to increase maximum work capacity in these older subjects is due to an inability to increase the maximum cardiac output, i.e., SV was not augmented by the drug and the maximum HR was slightly reduced. An increase in SV was likely precluded by the concomitant drug-induced reduction in the preload, manifest by a reduction in EDVI. Thus it might be postulated that reducing vascular afterload during exercise, in the absence of a reduction in preload, would permit an increase in SV in older subjects. Indeed, it has recently been reported that acute lowering of arterial pressure and vascular afterload by verapamil permits older subjects to augment SV during exercise and to improve their exercise capacity (7).

In conclusion, significant reductions in vascular load at rest and cardiac size at rest and during exercise can be effected in healthy, older persons by a low dose of a mixed vasodilator, nitroprusside. At all exercise levels, heart size is reduced and LV EF increases to levels comparable to those in younger, unmedicated individuals. The same SVI at rest and during any level of exercise is attained by an increase in LV EF at reduced LV volume and, thus, at reduced LV wall stress and myocardial O₂. However, maximum cardiac output and exercise capacity are not improved. These findings likely have implications in the treatment of older individuals whose exercise capacity is limited because of symptoms associated with normal or abnormal increases in LV wall stress with activity.