INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

REGULATION OF CARDIOVASCULAR and neuroendocrine variables in heart failure (HF) has primarily been investigated by utilizing orthostatic stress to decrease central blood volume. Results of these investigations (19, 20, 27, 30, 33, 34, 43) indicate that the neuroendocrine and vasoconstrictor responses to orthostatic maneuvers are abnormal in HF. Therefore, baroreflex regulation of vascular resistance and sympathetic nervous activity and modulation of hormonal vasoconstrictor activity [renin-angiotensin system, arginine vasopressin (AVP), and endothelins] may also be compromised in HF patients in response to central blood volume expansion and baroreceptor loading (21).

Seated thermoneutral (34.5-35.0°C) water immersion (WI) is a recognized model to induce isosmotic central volume expansion in humans (13). In healthy subjects, WI induces a prompt and sustained increase in central blood volume (13, 41) accompanied by loading of cardiopulmonary and arterial baroreceptors (16, 39, 41). Concomitantly, sympathetic nervous activity and systemic vascular resistance decrease and AVP release and the renin-angiotensin-aldosterone axis are suppressed (13, 23, 36, 41).

It is generally accepted that elevated plasma levels of AVP, angiotensin II (ANG II), endothelin-1 (ET-1), and sympathetic nervous activity induce vasoconstriction and sodium and water retention in HF (40), leading to volume overload and an increase in cardiac filling pressures. If baroreflex function is impaired in HF, this could prevent or attenuate the modulation of sympathetic nervous activity and release of vasoconstrictor hormones and the decrease in systemic vascular resistance in response to an increase in central blood volume.

Medical treatment of patients with severe HF often causes symptomatic relief and improves cardiac function. These patients can then be kept in a compensated state with a more moderate degree of cardiac dysfunction. When treatment is withdrawn for longer periods, these patients again develop disturbances of neuroendocrine control mechanisms. The question is whether the neuroendocrine dysregulation in response to an increase in central blood volume persists in a compensated state of HF and whether chronic impairment of the baroreflex-mediated responses to central blood volume expansion contributes to the abnormalities. Therefore, the purpose of this investigation was to assess the cardiovascular and neuroendocrine [AVP, plasma renin activity (PRA), ANG II, ET-1, and catecholamines] responses to an increase in central blood volume in compensated HF.

Apparently, thermoneutral WI has not previously been used as a tool to investigate baroreflex-mediated control of the circulation and the release of vasoconstrictor hormones in compensated HF. Therefore, we utilized this approach to test the hypothesis that suppression of the neuroendocrine vasoconstrictor systems is attenuated in compensated HF in response to WI-induced central hypervolemia.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects. Nine outpatients with HF (evaluated for cardiac transplantation) in a stable compensated phase and with low ejection fractions [New York Heart Association functional class (NYHA) II, n = 7; NYHA III, n = 2] and 10 age- and gender-matched control subjects were included in the study (Table 1). All patients exhibited dilated cardiomyopathy (idiopathic, n = 6; ischemic, n = 3). In addition to routine clinical and echocardiographic evaluation, the diagnosis was based on pretransplant invasive cardiac evaluation in eight of the patients. The average duration of HF symptoms was 51 mo (range 9-150 mo). The patients had no present history of angina pectoris or recent (<3 mo) history of myocardial infarction and were without signs of decompensation (pulmonary rales and peripheral edema) on the day of the study. They were on standard medical treatment with angiotensin-converting enzyme (ACE) inhibitors (n = 8), diuretics (n = 8), and digoxin (n = 7). In addition, three HF patients received nitrate vasodilator therapy, one received an /-adrenoreceptor blocker (carvedilol), and one received a ₁-adrenoreceptor blocker (metoprolol). ACE inhibitors were discontinued 24 h before the study, and all other medications were withheld on the day of the study. All subjects were in sinus rhythm at the time of the study, except one HF patient with atrial fibrillation. Another HF patient had short-lasting non-insulin-dependent diabetes mellitus (<2 mo) treated by dietary restraint. Control subjects were healthy as indicated by medical history, clinical examination, normal blood pressure (<140/95 mmHg), normal hematocrit, and urine dip stick test. All participants gave written informed consent.

Experimental protocol. Within 1 wk before the study, echocardiographic examination was carried out to determine cardiac functional status. For 3 days before the investigation, the participants ingested a standardized sodium-restricted diet containing 1 mmol sodium · kg body wt¹ · day¹. For 24 h before the experiment, urine was collected for determinations of sodium and potassium excretion. The subjects were instructed to abstain from eating, drinking, and smoking from 2200 on the evening before the experiment until its termination.

Measurements. Central venous pressure (CVP; HF, n = 8; controls, n = 10) was measured using a fluid-filled catheter (Cavafix, Braun), with the transducer (model DT-XX, ROSE, BOC Ohmeda) reference set at the sternal border of the fourth intercostal space during all times of the experiment. The intrathoracic location of the catheter tip was demonstrated by observing characteristic waveforms and responses to respiratory maneuvers.

1

1

1

1

Statistical analysis. Multifactor ANOVA corrected for repeated measures and a post hoc multiple range (Newman-Keuls) test were used to detect changes within groups from the average of the two baseline measurements over time during the WI or control sessions, respectively. During the 90-min seated control session, pooled data were used for comparison between groups. A two-sample t-test was used to detect differences in subject characteristics and the responses to WI for selected variables. Statistical evaluation of forearm subcutaneous and skeletal muscle vascular resistance was performed on absolute values but presented as relative changes. Data were logarithmically transformed before analysis if heterogeneity of variance was present. P < 0.05 was considered statistically significant. Values are means ± SE.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Central hemodynamic responses to WI. During the seated baseline period, no differences in arterial pressures were observed between controls and HF patients (Table 2). In response to WI, the control subjects exhibited an increase in systolic arterial pressure and pulse pressure (P < 0.05) compared with baseline values, whereas mean and diastolic arterial pressures did not change. In HF, systolic and mean arterial pressure increased during the last 15 min of immersion (P < 0.05) compared with baseline values with no changes in diastolic or pulse pressure.

View larger version (13K): [in this window] [in a new window]   Fig. 1.   Changes in central venous pressure (CVP), left atrial diameter (LAD), and stroke volume index (SVI) at baseline (0-30 min), during water immersion (WI) to the xiphoid process (30-60 min; enclosed in brackets), and during recovery (60-90 min) in patients with heart failure (HF) and control subjects (Con). Values are means ± SE; n = 8 HF patients and n = 10 controls for CVP, n = 9 HF patients and n = 8 controls for LAD, and n = 6 HF patients and n = 8 controls for SVI. * Statistically significant change (P < 0.05) from the average of the baseline measurements;  statistically significant difference (P < 0.05) between average HF values of baseline, WI, or recovery and Con values.

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Changes in systemic, forearm subcutaneous, and skeletal muscle vascular resistance at baseline (0-30 min), during WI (30-60 min; enclosed in brackets), and during recovery (60-90 min) in patients with HF and control subjects. Forearm subcutaneous and skeletal muscle vascular resistances are depicted as relative changes from the average of the 2 baseline measurements. Values are means ± SE of 9 HF patients and 10 controls for forearm subcutaneous and skeletal muscle vascular resistance and 6 HF patients and 8 controls for systemic vascular resistance. * Statistically significant change (P < 0.05) from the average of the baseline measurements.

1

2

1

2

5

Forearm vascular responses to WI. In the control group, there was a decrease (P < 0.05) in forearm subcutaneous vascular resistance during WI, and forearm muscle vascular resistance increased during recovery (P < 0.05; Fig. 2, Table 2). Forearm subcutaneous and muscle vascular resistance did not change in HF in response to WI or recovery.

Neuroendocrine responses to WI. Plasma concentration of AVP was higher (P < 0.05) in HF patients than in controls (Fig. 3). During WI, plasma concentration of AVP decreased (P < 0.05) in both groups and returned to baseline levels during recovery. Plasma concentrations of ANG II were significantly higher in HF patients (P < 0.05) than in controls (Fig. 3). In the control group, plasma ANG II concentrations were suppressed (P < 0.05) during WI, whereas WI did not cause a statistically significant suppression of plasma ANG II concentrations from baseline levels in HF patients (P = 0.16). PRA was higher in HF patients (P < 0.05) than in controls and was suppressed in response to WI in both groups (P < 0.05; Fig. 3). Group differences were not observed for plasma concentrations of NE or Epi, and plasma concentrations of NE were suppressed in both groups in response to WI. Plasma concentrations of Epi were suppressed in controls during WI, whereas no statistically significant suppression of Epi from baseline levels was observed in HF patients (P = 0.14).

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Changes in arginine vasopressin (AVP), norepinephrine (NE), and plasma renin activity (PRA) at baseline (0-30 min), during WI (30-60 min; enclosed in brackets), and during recovery (60-90 min) in patients with HF and control subjects. Values are means ± SE of 9 HF patients and 10 controls for AVP, NE, and PRA. * Statistically significant change (P < 0.05) from the average of the baseline measurements;  statistically significant difference (P < 0.05) between average HF values of baseline, WI, or recovery and control values.

Plasma composition during WI. Plasma sodium concentrations did not change in response to WI but were higher in controls (from 139 ± 0.5 to 140 ± 0.6 mmol/l) than in HF patients (from 136 ± 1.4 to 137 ± 1.2 mmol/l, P < 0.05 vs. controls).

Responses to 90 min of seated control. In the control subjects, systolic, diastolic, and mean arterial pressure increased slightly from 117 ± 2, 80 ± 3, and 92 ± 3 mmHg, respectively, at the beginning of the 90-min seated control to 119 ± 3, 83 ± 3, and 96 ± 3 mmHg (P < 0.05), respectively, during the last 15 min (Table 3). In HF patients, systolic, diastolic, and mean arterial pressures varied insignificantly throughout the control session from 109 ± 8 to 113 ± 8 mmHg, from 75 ± 4 to 78 ± 4 mmHg, and from 86 ± 5 to 89 ± 5 mmHg, respectively. Cardiac index did not change during the 90 min of seated control (controls, from 1.82 ± 0.1 to 2.00 ± 0.1 l/m²; HF, from 1.95 ± 0.2 to 2.23 ± 0.3 l/m²), and no differences were observed between the groups (Table 3). Heart rate did not change in controls (from 70 ± 3 to 71 ± 3 beats/min) and was lower than in HF (P < 0.05; Table 3). In HF patients, heart rate increased during the last 45 min of seated control from 86 ± 5 to 91 ± 5 beats/min (P < 0.05). Plasma sodium and potassium concentrations and plasma osmolality did not change over time. Plasma protein concentration increased slightly at the end of seated control in control subjects from 73.5 ± 1.2 to 74.6 ± 1.2 g/l (P < 0.05) but did not change in HF patients (from 80.1 ± 2.0 to 80.9 ± 2.2 g/l). Additional hemodynamic and neuroendocrine responses to 90 min of seated control are summarized in Table 3.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The new information of this study is that central blood volume expansion and an increase in cardiac filling within the normal physiological range in compensated HF elicit a reduction in systemic vascular resistance similar to that in normal controls, despite blunted forearm vascular responses. Concomitantly, the release of AVP, renin, and NE is suppressed in HF patients to the same extent as in healthy subjects. Therefore, the baroreflex-mediated decrease in systemic vascular resistance and release of vasoactive hormones are dissociated from the reflex control of forearm vascular beds during WI-induced central blood volume expansion in HF. The hypothesis that suppression of the release of neuroendocrine mediators and the decrease in systemic vascular resistance would be attenuated in compensated HF in response to central blood volume expansion was therefore not confirmed.

The effects of upright seated central blood volume expansion by WI are probably different from those of whole body tilting from upright to supine, because in the supine position the transverse gravitational stress compresses the heart and thorax from back to front (42). This might have a more pronounced effect on cardiac performance of the dilated ventricles in HF patients than in controls. During central volume expansion by seated WI, the gravitational compression of the thorax and heart is absent. We chose the upright seated position, because the upright body posture is the normal posture of humans during the daytime. Therefore, changing central blood volume by WI without changing body position may represent a more physiological stimulus than whole body tilting for investigating effects of central blood volume expansion on baroreflexes and vasoactive hormone release.

In HF patients and controls, central blood volume expansion augmented cardiac filling, improved cardiac performance, and increased stroke volume. The absolute increases in LAD and CVP in response to WI were very similar between HF patients and controls. The relative increase in LAD in response to central blood volume expansion, however, was less in HF patients (16 ± 4%) than in the control subjects (27 ± 5%), because the HF patients exhibited dilated cardiac chambers at baseline (as indicated by larger left atrial and left ventricular end-diastolic diameters). In the control subjects, systolic and arterial pulse pressures increased throughout WI; in HF patients, arterial pressure was unchanged during the initial 15 min, and systolic and mean arterial pressures increased only during the final 15 min of WI. Because stimulation of arterial baroreceptors is related to a change in arterial wall stretch (10), it is difficult to exclude that arterial baroreceptor stimulation occurred, even in the absence of a detectable change in arterial pressures. In this experiment, the increase in stroke volume could have increased pulsatile vascular dimensions in the aorta, leading to an increase in aortic pulsatile stretch, a powerful stimulus to arterial baroreceptors (10). It is therefore conceivable that cardiopulmonary and arterial baroreceptors were subjected to increased stretch in response to WI in HF patients and controls. Whether the receptors were stretched and/or stimulated to the same extent in HF patients and controls, however, cannot be deduced from this investigation.

Central blood volume expansion initiated a decrease in NE release and systemic vascular resistance from similar control levels in the two groups. This observation argues in favor of a preserved baroreflex-mediated inhibition of overall efferent sympathetic nervous activity in HF with a resultant vasodilatation. We expected elevated baseline levels of plasma NE and an attenuated suppression during WI in HF (21, 24). Administration of digitalis glycosides (15) and ACE inhibition reduce efferent sympathetic nervous activity and enhance the sympathoinhibitory response to baroreceptor stimulation (5, 22). Therefore, the normalized levels and responsiveness of plasma NE to WI in HF could be explained by the prior medical treatment and the fact that the patients were clinically stable in a compensated phase (NYHA class II-III).

In addition, AVP and renin release were suppressed to the same extent in HF patients and controls in response to WI. The HF patients exhibited elevated baseline levels of plasma AVP, whereas plasma osmolality did not differ between HF patients and healthy controls. Therefore, nonosmotic release of AVP was apparently enhanced (2). Central blood volume expansion initiated a normalized suppression of AVP release in HF, indicating that baroreceptor-mediated control of AVP release was well preserved in these HF patients.

The HF patients exhibited substantially elevated baseline levels of PRA and ANG II. The prior medical treatment, a relatively low sodium intake, increased efferent renal sympathetic nerve activity (8, 24), and reduced renal blood flow (8) could have contributed to the elevated PRA. Suppression of PRA during WI in HF patients was similar to that of the control subjects, indicating that the response to an increase in central blood volume was preserved. The reduction in PRA during WI in HF did not cause a statistically significant reduction in plasma concentration of ANG II. Residual ACE inhibition or increases in pulmonary blood flow during WI may have perturbed the relationship between renin activity and generation of ANG II. This notion, however, remains to be investigated.

In contrast to the preserved decrease in systemic vascular resistance and suppression of NE, AVP, and renin release, the HF patients exhibited blunted reflex control of the forearm vascular beds and heart rate during WI and recovery. This is in compliance with previous observations that muscle sympathetic nerve activity and limb vascular responses are abnormal in HF in response to phlebotomy (33), head-up tilt, or lower body negative pressure (19, 20, 27, 30, 33, 34, 43) and that HF patients demonstrate impaired control of heart rate in response to changes in baroreceptor stimulation (5, 10).

There are several possible explanations for the coexistence of blunted forearm vascular and heart rate responses and the preserved decrease in systemic vascular resistance and release of neuroendocrine mediators. First, arterial baroreceptor modulation of efferent sympathetic activity seems to be well preserved in humans and experimental dog models of heart failure (1, 5, 6), whereas cardiopulmonary modulation of efferent sympathetic outflow is markedly blunted (5, 7). In healthy humans, cardiopulmonary baroreflexes modulate muscle sympathetic nerve activity and limb vascular resistance (10, 25, 32, 39). Furthermore, several investigations indicate that simultaneous changes in cardiopulmonary and arterial baroreceptor stimulation must be present to modulate AVP and renin release (35, 39). Therefore, the blunted forearm but preserved systemic vascular and neuroendocrine responses to WI in HF might be explained by a combination of relatively blunted low-pressure cardiopulmonary and preserved high-pressure arterial baroreflexes.

Second, if cardiopulmonary baroreflexes initiated the decrease in systemic vascular resistance and release of AVP and renin in HF, selective dysfunction of one cardiopulmonary baroreflex control mechanism (forearm) with the preservation of others (systemic vascular resistance and neuroendocrine responses) must have been present (33). Further investigations, however, are needed to clarify this issue.

Third, the relative change in LAD in response to WI was less in HF patients than in controls. This could have caused less of a distension stimulus and resulted in an attenuated cardiopulmonary baroreflex engagement compared with controls and could explain the blunted forearm vascular responses.

A further understanding, however, of the degree of involvement of cardiopulmonary and arterial baroreceptors during WI is necessary to characterize their relative contributions to cardiovascular and neuroendocrine readjustments, respectively. Therefore, a definitive conclusion regarding the mechanism(s) of the dissociation of forearm vascular responses vs. the preserved neuroendocrine responses and decrease in systemic vascular resistance cannot be deduced from this study.

It has been recognized that increased parasympathetic activity to the heart elicited by arterial baroreceptor stimulation is the predominant mediator of the decrease in heart rate in response to a rapid hypervolemic/hypertensive stimulus (10, 31). Because of the more prolonged (30 min) character of the hypervolemic immersion stimulus, it is conceivable, however, that inhibition of efferent sympathetic activity to the heart also contributes to the decrease in heart rate during immersion in healthy subjects. Therefore, the blunted decrease in heart rate observed in response to central volume expansion by WI in HF was probably caused by attenuated baroreflex control of parasympathetic (6) and sympathetic modulation of heart rate.

In HF patients, the forearm vascular responses to WI were more heterogeneous (increased, unchanged, and decreased) than in the control subjects, who exhibited uniform directional changes. Therefore, the variability (error bars in Fig. 2) of the measurements is different from that of the control subjects in response to WI. The variable response was probably due to the different degrees of HF, which are related to the degree of abnormal peripheral vascular response (43). In contrast, systemic vascular resistance decreased in all HF patients and controls in response to WI. The variability between HF patients and controls (error bars in Fig. 2) was due to differences in individual levels of systemic vascular resistance. Furthermore, the CV of all these measurements did not differ in HF patients and controls during the 90-min seated control study (see METHODS; Table 3).

In the control subjects, the increase in central blood volume would have been expected to cause a decrease in forearm skeletal muscle vascular resistance in addition to that of the subcutaneous vascular bed. In a recent investigation from our laboratory (unpublished observations), a decrease in forearm muscle vascular resistance could not be detected during WI to the xiphoid process, whereas an additional increase in central blood volume by WI to the neck caused skeletal muscle vasodilatation. Thus the volume stimulus by WI to the xiphoid process was apparently not sufficient to cause a detectable vasodilatation. Furthermore, the control subjects exhibited an increase in skeletal muscle vascular resistance during recovery from WI, whereas HF patients did not. These findings indicate that recovery from WI in controls was associated with an increase in forearm muscle sympathetic vasoconstrictor activity, whereas this was not observed in HF patients. The absence of skeletal muscle vasodilatation during WI may seem in contrast to the pronounced vasoconstriction during recovery. Baroreflex adaptation and/or resetting, however, may occur during the 30 min of WI. Therefore, WI and recovery may not elicit identical reciprocal changes in forearm muscle vasoconstrictor activity.

Baseline levels of ET-1 in HF patients were not different from those in controls. During WI and recovery, plasma ET-1 concentrations were slightly higher in HF patients than in controls. The (patho)physiological importance, however, of this minor difference is not clear. In HF, treatment with the ACE inhibitor fosinopril reduces elevated plasma ET-1 levels to normal values (18). Therefore, treatment with ACE inhibitors could have normalized plasma levels of ET-1 in HF. Neither control subjects nor HF patients, however, exhibited any changes in plasma ET-1 in response to WI. Thus changes in plasma concentrations of ET-1 do not seem to explain the systemic vasodilatation in response to WI in HF patients and controls.

Limitations. The HF patients received prior medical treatment and were well compensated in NYHA class II-III without signs of elevated cardiac filling pressures, as indicated by the normal CVP at baseline. Because we measured CVP only in the upright seated position, it cannot be excluded that these patients exhibited elevated CVP in the supine position or elevated left-sided filling pressures. It is possible that HF patients with elevated cardiac filling pressures in the upright seated position and/or without the prior medical treatment may have exhibited a response to WI that was different from that of the subjects investigated in this study.

Conclusions. The results of this study demonstrate that central blood volume expansion in compensated HF elicits a reduction in systemic vascular resistance similar to that of normal control subjects, despite blunted forearm vascular responses. Concomitantly, release of AVP, renin, and NE is suppressed in HF patients to the same extent as in healthy subjects. Thus the baroreflex-mediated decrease in systemic vascular resistance and in release of vasoactive hormones is dissociated from the reflex control of forearm vascular beds and heart rate in compensated HF during WI-induced central blood volume expansion.

Perspectives. The increase in central blood volume improved cardiac performance, decreased sympathetic nerve activity and systemic vascular resistance, and suppressed the release of vasoactive and sodium- and water-retaining hormones in compensated HF. In fact, WI tended to normalize the levels of these variables. Therefore, the effects of WI on renal sodium and water handling in HF should be addressed in future studies. Such investigations may provide further insight into the pathophysiology of extracellular fluid volume control and may have implications for the treatment of HF.