| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Exercise Science, The University of Iowa, Iowa City, Iowa
Submitted 11 May 2005 ; accepted in final form 17 August 2005
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
venous physiology; venous return; cardiac filling pressure
Previous investigations have examined the effects of changes in cardiac output distribution on right atrial pressure in three basic ways. One approach has been to simulate changes in cardiac output distribution using theoretical computer models of the circulation (10). Another approach has been to alter cardiac output distribution by evoking reflexes (8) or infusing drugs (16). The third approach has been to impose stresses that alter cardiac output distribution such as exercise (27, 28) and combined exercise and heat stress (9, 22). There are important limitations to each of these approaches. Computer simulations are useful in providing insight into how the circulatory system may operate, but knowledge gained from computer simulations needs to be tested in vivo. A limitation of employing reflex- and drug-evoked changes in cardiac output distribution is that these interventions can directly mobilize blood toward the heart (2, 29) and thus raise right atrial pressure, independently of a change in the distribution of cardiac output. For example, these interventions can also cause venous smooth muscle to contract and actively expel blood toward the heart (11). Likewise, stresses such as hyperthermia or exercise evoke reflexes that can alter blood volume in ways other than by changing the distribution of cardiac output (23). Exercise also activates the muscle pump (12, 26, 28, 32), which has a profound affect on cardiac filling pressure (28, 34). To date, the effect on right atrial pressure caused by redistribution of cardiac output in isolation has received little attention. One approach has been to mechanically alter the distribution of cardiac output by occluding the descending thoracic aorta (1, 30). Although this approach sheds considerable light on integrative circulatory function, it clearly represents a nonphysiological magnitude of redistribution.
In the present study we sought to directly test the hypothesis that a diversion of blood flow from a relatively noncompliant vasculature (muscle) to compliant vasculature (splanchnic organs and skin) acts to reduce right atrial pressure. The approach was to occlude terminal aortic blood flow for 30 s in one of two modes of ventricular pacing in atrioventricularly blocked dogs. Experiments were carried out after blockade of autonomic reflexes with hexamethonium to isolate the passive mechanical effects of imposing alterations in the distribution of cardiac output on right atrial pressure. In one mode, cardiac output was maintained constant (35), and in the other mode arterial pressure was maintained constant. The rationale for this approach is as follows. When cardiac output is maintained constant, cuff inflation will cause a portion of terminal aortic blood flow to be diverted from the hindlimbs (a noncompliant circulation) to the splanchnic and skin circulations (compliant circulations). The increase in splanchnic blood flow is expected to cause splanchnic venous transmural pressure to rise, which in turn will cause blood volume to accumulate within the splanchnic circulation (4). Any accumulation of blood volume in the splanchnic organs and skin is expected to lower central venous blood volume and thus right atrial pressure. Conversely, when arterial pressure is maintained constant by ventricular pacing, cardiac output must fall in proportion to the fall in hindlimb blood flow. Thus splanchnic perfusion pressure and splanchnic blood flow are expected to remain constant (no redistribution), meaning that splanchnic blood volume should remain constant. Thus any difference in the response of right atrial pressure under the two conditions can be largely attributed to the influence of the distribution of blood flow between compliant and noncompliant vasculature on right atrial pressure.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Experiments were carried out using five 18- to 24-kg mongrel houndtype dogs (Oak Hill Genetics, Ewing, IL).
Surgical preparation.
The animals were prepared in the following series of two to three aseptic surgical procedures as described previously (23, 26, 27, 28). A right thoracotomy was performed, and pacing leads were sutured to the apex of the left ventricle. Atrioventricular block was induced by injecting a small volume of formalin near the atrioventricular node, and a blood flow transducer (Transonic, Ithaca, NY) was placed on the ascending aorta. The probe cable and pacing leads were tunneled subcutaneously to an exit site on the back. A skin patch delivering 50 µg/h of fentanyl was placed on the dog for 72 h after surgery to control postoperative pain. In a second procedure, a blood flow transducer (Transonic) was placed around the terminal aorta via a midabdominal incision. A pneumatic vascular occluder cuff (In Vivo Metric, Healdsburg, CA) was placed on the aorta distal to the blood flow transducer. The probe cable and the cuff-actuating tube were tunneled subcutaneously to an exit site on the back. When feasible, catheters were implanted during this procedure as described below. A skin patch delivering 25 µg/h of fentanyl was placed on the dog for 72 h after surgery to control postoperative pain. When needed, a third procedure was performed during which a catheter was inserted into a sidebranch of the femoral artery and advanced into the abdominal aorta, a catheter was inserted into a sidebranch of the femoral vein and advanced into the abdominal vena cava, and a catheter was inserted into the right jugular vein and advanced to the caval-right atrial junction. A skin patch delivering 25 µg/h of fentanyl was placed on the dog for 24–72 h after surgery to control postoperative pain. The animals were treated with cephazolin (500 mg iv) during each surgical procedure and with cephalexin for 1 wk postoperatively. The animals were allowed at least 1 wk for recovery between surgical procedures. All experiments were performed after the animals had recovered from the surgery and were afebrile, active, and of good appetite. A pacemaker carried by the dog paced the heart at 
70 beats/min between experiments.
Experimental procedures. The animals were sedated with acepromazine (20–30 mg iv) in order that the results would not be complicated by the influence of muscle contraction on blood volume distribution and thus right atrial pressure. The animals were treated with hexamethonium (10 mg/kg iv), a dose previously shown to provide effective inhibition of autonomic ganglionic neurotransmission (26, 28). The animals were lightly restrained in lateral recumbency on a padded table.
Experimental protocol. The target cardiac output or mean arterial pressure was selected, and the measured variables were allowed to achieve steady levels. The occluder cuff was then inflated for 30 s after which it was slowly deflated in order that the resulting reactive hyperemia would not overwhelm the computer control of ventricular pacing. Time was allowed for the measured variables to return to baseline values, the mode of pacing was changed, and the foregoing procedure was repeated. The modes of pacing were imposed in no regular order.
Data collection. The arterial catheter and the jugular catheter were connected to pressure transducers (PE-10 EZ, Ohmeda, Madison, WI) secured at the level of the right atrium. The flow transducers were connected to flowmeters (T106, Transonic, Ithaca, NY). All signals were digitized at 1 kHz, and beat-by-beat mean values were calculated and stored on a microcomputer for subsequent analysis. Also, cardiac output and terminal aortic flow and arterial and right atrial pressure were simultaneously digitized at 250 Hz (as depicted in Fig. 1) and stored on a second microcomputer using a commercially available data-collection system (Sonometrics, London, Ontario, Canada).
|
Statistical analysis. Statistical comparisons were done using paired t-tests, and P < 0.05 was considered significant. Data are presented as means ± SE.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
The group mean responses, averaged from the five dogs, are presented in Fig. 2, and steady-state values averaged over the time periods depicted in Fig. 2A (black bars) are presented in Table 1. The thick line Fig. 2A (constant arterial pressure) and the thin line Fig. 2D (constant cardiac output), as well as the related data in Table 1, demonstrate that the ventricular pacing control of arterial pressure and cardiac output was successful. Terminal aortic occlusion had essentially identical affects on hindlimb blood flow (Fig. 2E) and hindlimb vascular conductance (Fig. 2G) between the two trials. The magnitude and timing of the changes in total peripheral conductance (Fig. 2F) were also similar in the two trials. Little change in heart rate was required to maintain cardiac output constant, whereas a sizable drop in rate was needed to maintain arterial pressure constant during terminal aortic occlusion (Fig. 2H, Table 1). Occlusion raised arterial pressure 13 mmHg when cardiac output was maintained constant (Fig. 2A, Table 1), and cardiac output needed to be lowered by 14% to maintain arterial pressure constant during occlusion. Right atrial pressure was little affected by terminal aortic occlusion when cardiac output was maintained constant (thin line of Fig. 2B; Table 1), whereas right atrial pressure rose during occlusion when arterial pressure was maintained constant (thick line in Fig. 2B; Table 1). This difference in the response of right atrial pressure under the two conditions achieved statistical significance (Fig. 3).
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
The mechanical properties of the venous system play a critical role in determining how much blood is available to the heart and play an important role in determining the manner and extent to which various stresses and disease processes impact the circulation (3, 17–19, 33). For example, a change in cardiac output produced by a change in cardiac pump performance elicits an opposite change in right atrial pressure owing to the resistive and capacitive properties of the peripheral circulation (3, 14, 28). A rise in cardiac output increases peripheral pressures because blood vessels provide resistance to blood flow, and the rise in peripheral pressures leads to a rise in peripheral blood volume because blood vessels are elastic. The rise in peripheral blood volume comes at the expense of central blood volume, and right atrial pressure falls accordingly. Opposite responses occur with a fall in cardiac output produced by a fall in cardiac pump performance. The opposite and proportional changes in right atrial pressure that accompany a change in cardiac output provide a passive hydraulic regulation of cardiac output (20, 25). Furthermore, a change in the distribution of cardiac output between relatively compliant and noncompliant vasculature is proposed to induce a passive change in right atrial pressure even in the absence of a change in cardiac output, and that is the focus of the present paper.
The influence of organ blood flow on organ blood volume has been addressed in two main ways. One approach has been to assess and directly report the volume of blood that accumulates or is displaced from an organ when its blood flow increases or decreases (20). For example, the volume of blood in the splanchnic circulation of dogs is reported to fall by 0.19 ml/kg for each 1 ml·min–1·kg–1 decrease in blood flow (4). The other approach has been to evaluate the time constant of venous drainage (equal to the product of venous compliance and venous resistance) of an organ system. For example, the time constant of muscle is reported to be 4 s (15), which corresponds to a volume sensitivity to flow of 0.07 ml/kg for each 1 ml·min–1·kg–1 decrease in blood flow, whereas the time constant of skin is reported to be 28 s (7), which corresponds to a volume sensitivity to flow of 0.47 ml/kg for each l ml·min–1·kg–1 decrease in blood flow. Thus it appears that skin is seven times more compliant than muscle.
A consequence of terminal aortic occlusion that does not occur during physiologically induced diversion of blood flow is the large fall in arterial pressure distal to the occluder cuff. The fall in distal arterial pressure will cause arterial blood volume to be released to the central circulation, which will work to raise right atrial pressure. Importantly, distal pressure is expected to fall similarly in both trials, meaning that an equal volume of blood will be released in both trials. Thus, although discharge of hindlimb arterial blood volume is expected to raise right atrial pressure in both circumstances, it should do so equally and therefore does not provide an explanation for the difference in the response in right atrial pressure observed between the two trials.
To maintain mean arterial pressure constant during terminal aortic occlusion, cardiac output was reduced by an amount equal to the reduction in terminal aortic flow (Table 1). When arterial pressure was maintained constant during occlusion, right atrial pressure rose by 1 mmHg, likely due to the central discharge of hindlimb blood volume to the central circulation. As a consequence of maintaining mean arterial pressure constant, there would be little or no alteration in the flow to the other (nonhindlimb) regions, meaning little or no blood volume would be shifted into or out of these regions.
When cardiac output was maintained constant during terminal aortic occlusion, the blood flow that the hindlimbs had been receiving was suddenly redirected to the other (nonhindlimb) regions. Arterial pressure rose in proportion to the reduction in total peripheral conductance caused by occlusion owing to the absence of the arterial baroreflexes, which prevented the normal compensatory vasodilation (31). In contrast to the rise in right atrial pressure observed when arterial pressure was maintained constant, right atrial pressure was unchanged when cardiac output was maintained constant (Fig. 2B; Table 1). Because presumably a similar volume of hindlimb blood volume was discharged centrally under both conditions, why did right atrial pressure not rise when cardiac output was maintained constant? There are likely two explanations. First, arterial pressure rose when cardiac output was maintained constant, and thus blood volume would accumulate within the arterial system of the nonhindlimb regions. However, this is expected to contribute little to the difference in the response of right atrial pressure between the two trials because the veins are far more compliant than arteries. Total vascular compliance is reported to be 3.00 ml·mmHg–1·kg–1 (2), and arterial compliance is reported to be 0.06 ml·mmHg–1·kg–1 (5). Thus the compliance of the venous system is 2.94 ml·mmHg–1·kg–1 (equal to total compliance less arterial compliance), meaning that venous compliance is nearly 50-fold (2.94 vs. 0.06) greater than arterial compliance. Thus the 13-mmHg rise in arterial pressure when cardiac output was maintained constant would be expected to reduce right atrial pressure by 1/50th of this or 0.26 mmHg, which is only 25% of the total observed difference (Fig. 3).
The second and likely most important explanation for the difference in the response of right atrial pressure between the two trials was the diversion of blood flow from noncompliant to compliant vasculature. Of the regions supplied by the terminal aorta, 85% is estimated to go to skeletal muscle, a relatively noncompliant circulation. When cardiac output was maintained constant, the blood flow diverted from the hindlimbs is redirected to the nonhindlimb regions in proportion to the vascular conductance of each of the nonhindlimb regions. The reduction in blood flow imposed on the hindlimb muscles will cause a small volume of muscle venous blood to be released to the central circulation. However, a portion of the diverted terminal aortic flow will be directed to upper body muscle, which will lead to a rise in the blood volume of these muscles that is directly proportional to the volume released from the hindlimb muscles. For example, if all of the diverted terminal aortic flow were redirected to a similar proportion of muscle, skin, and bone in the upper body, there would be no effect on right atrial pressure; the blood volume lost from the hindlimb tissues would be matched by the rise in volume in the upper body tissues. The same would be true if regional compliances did not differ; diversion of blood flow from one region to another would be without affect on right atrial pressure. For example, diversion of blood flow from muscle to the renal and/or cerebral circulation would have little effect on right atrial pressure because all of these regions are relatively noncompliant. Only when flow is diverted from a region of low compliance to a region of high compliance, or vice versa, is there an impact on right atrial pressure.
To what extent did this occur in the present study? Changes in skin blood flow were likely unimportant in the present study. For example, the skin receives 5% of the cardiac output (or 0.18 l/min in our dogs), and a sizable fraction (perhaps one-third) is perfused by the terminal aorta. Thus the upper body skin would receive 0.12 l/min, which would be expected to rise by 17% (0.02 l/min) during terminal aortic occlusion when cardiac output was maintained constant given that the arterial-venous pressure gradient rose by 17%. The 0.02 l/min increase in upper body skin blood flow would precisely equal the decrease in hindlimb skin blood flow during occlusion if skin receives 4% of the terminal aortic flow.
The diversion of blood flow from hindlimb muscle to the splanchnic circulation likely constituted an important mechanism causing the difference observed in the response of right atrial pressure between the two trials in the present study. We estimate that 0.15 l/min of blood flow from hindlimb muscle was redirected to the splanchnic circulation. For example, the splanchnic circulation receives 25% of cardiac output (0.88 l/min in our dogs), and splanchnic flow would be expected to rise by 17% (0.15 l/min) during terminal aortic occlusion when cardiac output was maintained constant given that the arterial-venous pressure gradient rose by 17%. The blood volume contained within the splanchnic circulation in dogs changes by 0.19 ml/kg for each 1 ml·min–1·kg–1 reduction in splanchnic blood flow (4), meaning 1.3 ml/kg of blood would accumulate within this organ system were its flow to rise by 6.8 ml·min–1·kg–1 as in the present study. Based on an extrasplanchnic venous compliance of 2.25 ml·mmHg–1·kg–1 (21), a redistribution of 1.3 ml of volume to the splanchnic circulation would reduce right atrial pressure by 0.6 mmHg, i.e., a large fraction of the 0.75-mmHg venous component of the difference in right atrial pressure we observed between the two trials in the present study.
Under the conditions of the present experiment, we found that flow diversion produced a net change in right atrial pressure of 0.75 mmHg in excess of the change attributable to the difference in mean arterial pressure between the two trials. This is comparable to the rise induced by bilateral carotid artery occlusion in a similar animal model after vagal block (2) but is considerably less than the 2.3-mmHg rise induced by the muscle chemoreflex (23). By comparison, all of the foregoing mechanisms for raising right atrial pressure are dwarfed by the muscle pump, which can raise filling pressure by 5 mmHg (28, 12). However, the extent to which right atrial pressure could be altered by flow diversion in the present study is limited by the relatively small diversion of flow from noncompliant to compliant regions. That is, the 0.15 l/min of blood flow we estimate was diverted to the splanchnic circulation represents only 4% of the resting cardiac output. Finally, the opposite changes in heart rate in the two trials would work to lessen the difference in the response of right atrial pressure between the two trials, owing to the capacitive function of the heart (24). That is, the heart maintains a certain time-averaged volume of blood that can increase or decrease, meaning that blood volume can accumulate within the heart or can be displaced from the heart, just as blood volume is altered in other organs. When arterial pressure was maintained constant, some of the blood volume released from the hindlimbs was absorbed by the heart inasmuch as its capacitance rises with falling heart rate (24). This will work to reduce right atrial pressure. Conversely, when cardiac output was maintained constant, heart rate rose during occlusion, meaning that blood volume was expelled from the heart. This will work to raise right atrial pressure. Thus the difference in the change in right atrial pressure between the two trials shown in Fig. 3 would have been larger if the capacitance of the heart had not changed.
Release of occlusion in the present study, wherein the hindlimbs suddenly "demand" an increase in blood flow, mimics physiological conditions such as exercise, where there is an increased demand for muscle blood flow, and heat stress, where there is an increased demand for skin blood flow. When arterial pressure was maintained constant, cuff deflation led to a fall in right atrial pressure, likely due to refilling of the hindlimbs inasmuch as blood flow to other regions would not change because cardiac output was raised to match the rise in hindlimb flow. This mimics the normal response of dogs to mild-to-moderate exercise, in which cardiac output rises to meet the demands of active skeletal muscle with little or no compensatory splanchnic vasoconstriction except that, in exercise, the muscle pump expels muscle blood volume. When cardiac output was maintained constant, blood flow from other regions was diverted to the hindlimbs. This mimics the normal response of humans to exercise and to heat stress, in both of which there is significant compensatory splanchnic vasoconstriction. The reduction in blood flow to compliant regions induces a release of blood volume from these regions to the central circulation. In the present study, right atrial pressure was little changed, i.e., it was effectively maintained constant, despite the requirement to refill the hindlimbs with blood volume. Again, a modest portion (25%) of the blood volume needed to refill the hindlimbs on cuff deflation likely came from the arterial system, stemming from the fall in mean arterial pressure.
In the two-reservoir (arterial and central venous) model provided by Levy (14), arterial pressure rises and central venous pressure falls with rising cardiac output, owing to the translocation of blood volume from the central venous to the arterial reservoir. In this model a rise in total peripheral resistance causes arterial pressure to rise more steeply and central venous pressure to fall more steeply with rising cardiac output, signifying that a greater arterial-venous pressure gradient is created by a given cardiac output when resistance rises. The relationship between cardiac output and arterial pressure in the present study was in accordance with this model, namely, vascular occlusion (increased resistance) caused arterial pressure to be greater at a given cardiac output. In the constant arterial pressure trials, the fall in cardiac output required to keep arterial pressure constant after occlusion was accompanied by a rise in right atrial pressure. This response is reminiscent of the influence of cardiac output on right atrial pressure in the two-compartment model of Levy (14), in which the decrease in cardiac output is the cause and the increase in right atrial pressure is the effect. Thus it could be construed that the observed increase in right atrial pressure during these trials might have been due in part to the decrease in cardiac output, per se. However, strictly applying cause and effect as identified by Levy's analysis to the present study is problematic given the stark methodological differences between the two approaches. For example, in Levy's approach, the decrease in cardiac output was imposed across the entire peripheral circulation, whereas in the present study the flow reduction was restricted to the hindlimbs.
In summary, even a relatively small (4% of cardiac output) diversion of blood flow from noncompliant to compliant vasculature leads to a moderate (0.75 mmHg) fall in right atrial pressure in awake dogs.
|
GRANTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
