Medical Physiology, 3rd Edition

Matching of Venous Return and Cardiac Output

Venous return is the blood flow returning to the heart. Most often, the term is used to mean the systemic venous return, that is, return of blood to the right heart. Because the input of the right heart must equal its output in the steady state, and because the cardiac outputs of the right and left sides of the heart are almost exactly the same, the input to the right heart must equal the output of the left heart. Thus, the systemic venous return must match the systemic cardiac output.

Increases in cardiac output cause right atrial pressure to fall

The right atrial pressure (RAP) determines the extent of ventricular filling and indeed was the first of the three determinants of EDV discussed above in the chapter. In turn, RAP depends on the venous return of blood to the heart. Imagine that we replace the heart and lungs with a simple pump—with adjustable flow—that delivers blood into the aorta and simultaneously takes it back up from the right atrium (Fig. 23-9A). Using this simple pump, we can study the factors that govern venous return from the peripheral systemic circulation to the right heart by varying flow (the independent variable) while recording RAP (the dependent variable).


FIGURE 23-9 Vascular function. In A, a heart-lung machine replaces the right heart, the pulmonary circulation, and the left heart. We can set the pump flow to predetermined values and examine how cardiac output determines CVP and RAP. CV and CA are vascular compliances. B shows the dependent variable (RAP) on the y-axis, whereas C shows RAP on the x-axis.

We start our experiment with the pump generating a normal cardiac output of 5 L/min; the RAP is 2 mm Hg. This situation is represented by point A in Figure 23-9B, which is a plot of RAP versus venous return (i.e., our pump rate). When we reverse the axes, the plot is known as a vascular function curve. Note that we express venous return (i.e., cardiac input) in the same units as cardiac output. We now turn off the pump, a situation analogous to cardiac arrest. For several seconds, blood continues to flow from the arteries and capillaries into the veins because the arterial pressure continues to exceed the venous pressure. Remember that the aortic pressure does not fall to zero during diastole because of the potential energy stored in the elastic recoil of the arterial walls (see Box 22-2). Similarly, after a cardiac arrest, this potential energy continues to push blood to the venous side until all pressures are equal throughout the vascular tree, and flow stops everywhere.

When blood flow finally ceases after a cardiac arrest, the pressures in the arteries, capillaries, veins, and right atrium are uniform. This pressure is called the mean systemic filling pressure (MSFP) and is ~7 mm Hg (point B in Fig. 23-9B). Why is the MSFP not zero? When you fill an elastic container (e.g., a balloon) with fluid, the pressure inside depends on how much fluid you put in as well as the compliance of the container (see Equation 19-5). In the case of the cardiovascular system, MSFP depends on the blood volume and the overall compliance of the entire vascular system. Now that blood flow has ceased, there is obviously no longer any force (i.e., no ΔP) driving the blood from the veins into the pump. Therefore, RAP equals MSFP.

Now, starting with the system at a standstill (point B in Fig. 23-9B), we increase pump rate to some level, measure the steady-state RAP, and repeat this procedure several times at different pump rates. As the pump rate increases, RAP falls (see Fig. 23-9B) because the inflow to the pump draws blood out of the right atrium, the site closest to the inlet of the pump. Because the upstream CVP exceeds RAP, blood flows from the large veins into the right atrium. At a pump rate of 5 L/min, the RAP falls to about +2 mm Hg, its original value. At even higher pump rates, RAP eventually falls to negative values.

Cardiovascular physiologists usually work with the vascular function curve with the axes reversed from those shown in Figure 23-9B, treating RAP as the independent variable and plotting it on the x-axis (see Fig. 23-9C). As RAP becomes less positive, it provides a greater driving pressure (i.e., greater ΔP = CVP − RAP) for the return of blood from the periphery to the right atrium, as it must for cardiac output to increase. Thus, the cardiac output steadily rises as RAP falls. However, as RAP becomes increasingly negative, the transmural pressure of the large veins becomes negative, so that the large veins feeding blood to the right atrium collapse. No further increment in venous return can occur, even though the driving pressure, ΔP, is increasing. Therefore, the vascular function curve plateaus at negative values of RAP, around −1 mm Hg (see Fig. 23-9C).

Two different theories imageN23-11 attempt to explain the shape of the vascular function curve in Figure 23-9C. One explanation for the curve's steepness emphasizes the high compliance of the venous capacitance vessels; the other focuses on their small viscous resistance. In reality, both compliance and resistance are responsible for the steepness of the vascular function curve.


How Does Cardiac Output Determine Right Atrial Pressure?

Contributed by Emile Boulpaep

Imagine that we replace the right heart, the pulmonary circulation, and the left heart with a single pump (see Fig. 23-9A), similar to the heart-lung machine used for cardiopulmonary bypass during open heart surgery. In explaining the shape of the vascular-function curve in Figure 23-9C, we can stress either that large veins have a high compliance or that they have a low resistance.

Model of High Venous Compliance

In the high-venous-compliance model, we ignore the small axial pressure gradient along the lumen of the venous reservoir that is a result of the small viscous resistance between the large veins and right atrium. Instead, we assume that the CVP is the same as the RAP and that we can lump the compliance of all large-capacitance veins into a single compliance (CV). The model in Figure 23-9A has two reservoirs—an aortic and a venous reservoir—separated by the peripheral resistance of the microcirculation. As we noted in our discussion of the “Windkessel” function of the aorta (see Fig. 22-4C), the compliance of the aorta (CA) is far less than that of the large veins (CV). During normal pumping of the heart, the aortic pressure (PA) is much higher than the CVP (PV). Let us consider what happens immediately after a cardiac arrest. At that moment, PA still exceeds PV, and blood continues to flow from the aortic reservoir to the venous reservoir until PA and PV both become equal to MSFP. How much blood volume shifts from the aortic to the venous reservoir depends on the relative compliances of the two reservoirs. For the aorta,

image (NE 23-1)

and for the veins,

image (NE 23-2)

CV is ~20 times larger than CA. In the equilibrium state, all the blood lost by the aorta ends up in the veins (i.e., ΔVV = –ΔVA). Therefore, because CV is ~20 times larger than CA, and because the magnitudes of the volume changes are identical, the fall in the aortic pressure (ΔPA) must be ~20-fold greater than the rise in venous pressure (ΔPV). Thus, at equilibrium, when all the pressures are the same, we must end up at a pressure (MSFP = ~7 mm Hg) that is much closer to the physiological CVP (CVP = ~2 mm Hg) than the original mean aortic pressure (PA = ~95 mm Hg).

When the heart starts to pump again, it depletes the volume in the highly distensible venous reservoir, so that PV progressively falls to the original steady-state value of 2 mm Hg. At this point, cardiac output and venous return are once again matched.

Model of Low Venous-Resistance

In the low-venous-resistance model, we ignore the capacitance of the veins, and instead focus on the resistance between the large veins and right atrium. The hydrodynamic equivalent of Ohm's law says that the driving pressure equals flow times resistance:

image (NE 23-3)

Because the venous return equals F, it should depend on the axial pressure gradient (ΔP) between the CVP upstream and the RAP downstream. According to this model, RAP is the force that opposes flow into the atria. Venous return should also depend on the resistance to blood flow between the large systemic (i.e., “central”) veins that serve as the blood reservoir, on the one hand, and the right atrium, on the other. This resistance to venous return (RVR) resides mostly in these large-capacitance veins. From Ohm's law, the venous return depends on the CVP, RAP, and RVR:

image (NE 23-4)

RVR is usually considered to be the reciprocal of the slope of the linear portion of the vascular function curve in Figure 23-9C. However, this line is not really a plot of Ohm's law. Ohm's law should be a plot of the venous return versus the driving pressure (ΔP), as suggested by Equation NE 23-4. Instead, in Figure 23-9C, we plot the venous return versus RAP, which is not quite the same as ΔP. Changes in RAP in Figure 23-9C represent ΔP only if the “upstream pressure” (i.e., CVP) remains constant, which can only be true over a limited range of flows. The CVP remains constant only as long as sufficient blood is present on the venous side of the circulation. If flow increases too much, not enough blood can enter the veins, CVP falls, and the change in RAP is no longer equal to the change in the driving pressure ΔP. Despite this caveat, cardiovascular physiologists often take the reciprocal of the linear portion of Figure 23-9C as RVR.

According to this model, the negative intravascular pressures at high venous return (i.e., at high cardiac output) cause the large intrathoracic veins to collapse, which effectively increases the RVR to infinity. Hence, the curve plateaus.


Although both the high-capacitance and low-resistance models provide an intuitive understanding of the vascular function curve in Figure 23-9C, both compliance and viscous resistive properties of the venous system affect the relationship between pressure and flow.

Changes in blood volume shift the vascular function curve to different RAPs, whereas changes in arteriolar tone alter the slope of the curve

Because the vascular function curve depends on how full the capacitance vessels are, changing the blood volume affects the vascular function curve. An increase in blood volume (e.g., a transfusion) shifts the curve to the right. In the example in Figure 23-10A, the intercept with the x-axis (i.e., MSFP) moves from 7 to 9 mm Hg, as would be expected if more blood were put into a distensible container, whether it is a balloon or the circulatory system. A decrease in blood volume (e.g., hemorrhage) shifts the curve and the x-intercept to the left. Thus, MSFP increases with transfusion and decreases with hemorrhage. However, changes in blood volume do not affect the slope of the linear portion of the vascular function curve as long as there is no change in either vessel compliance or resistance.


FIGURE 23-10 Effect of changes in blood volume and vasomotor tone on the vascular function curve. The purple curves are the same as the vascular function curve in Figure 23-9C. In B, dilation of the arterioles increases CVP and thus raises the driving force (CVP − RAP) for venous return. Constriction of arterioles has the opposite effects.

Change in the venomotor tone, by constriction or dilation of only the veins, is equivalent to change in the blood volume. Returning to our balloon analogy, even if we hold the amount of blood constant, we can increase pressure inside the balloon by increasing the tension in the wall. Because most of the blood volume is in the veins, a pure increase in venomotor tone would be equivalent to a blood transfusion (see Fig. 23-10A). Conversely, a pure decrease in venomotor tone would reduce the tension in the wall of the container, shifting the curve to the left, just as in a hemorrhage.

Change in the tone of the arterioles has a very different effect on the vascular function curve. Because the arterioles contain only a minor fraction of the blood volume, changes in the arteriolar tone have little effect on MSFP and thus on the x-intercept. However, changes in the arteriolar tone can have a marked effect on the CVP, which, along with the RAP, determines the driving force for the venous return. Arteriolar constriction flattens the vascular function curve; arteriolar dilation has the opposite effect (see Fig. 23-10B). We can understand this effect by examining the vertical line through the three curves at an RAP of 2 mm Hg. For the middle or normal curve, the difference between a CVP of, say, 5 mm Hg and an RAP of 2 mm Hg (ΔP = CVP − RAP = 3 mm Hg) produces a venous return of 5 L/min. Arteriolar constriction might lower the CVP from 5 to 4 mm Hg, thereby reducing ΔP from 3 to 2 mm Hg. Because the driving pressure (ΔP) drops to image of normal, venous return also falls to image of normal. Thus, arteriolar constriction flattens the vascular function curve. Similarly, arteriolar dilation might raise CVP from 5 to 6 mm Hg, so that the ΔP increases from 3 to 4 mm Hg, producing a venous return that is image of normal. Thus, arteriolar dilation steepens the vascular function curve.

In the preceding analysis, we assumed “pure” changes in blood volume or vessel tone. In real life, things can be more complicated. For example, hemorrhage is typically followed by arteriolar constriction for maintenance of the systemic arterial pressure. Thus, real situations may both shift the x-intercept and change the slope of the vascular function curve.

Because vascular function and cardiac function depend on each other, cardiac output and venous return match at exactly one value of RAP

Just as there is a vascular function curve, there is also a cardiac function curve that in effect is Starling's law. The classical Starling's law relationship (see Fig. 22-12B), which is valid for both ventricles, is a plot of developed pressure versus EDV. However, we already expressed Starling's law as a cardiac performance curve in Figure 22-12C, plotting stroke work on the y-axis and atrial pressure on the x-axis. Because stroke work—at a fixed arterial pressure and heart rate—is proportional to cardiac output, we can replace stroke work with cardiac output on the y-axis of the cardiac performance curve. The result is the red cardiac function curve in Figure 23-11A. Note that the cardiac function curve is a plot of cardiac output versus RAP, and it has the same units as the vascular function curve. We can plot the two relationships on the same graph (see Fig. 23-11A). However, the y-axis for the cardiac function curve is cardiac output, whereas the y-axis for the vascular function curve is venous return. Of course, cardiac output and venous return (i.e., cardiac input) must be the same in the steady state.


FIGURE 23-11 Matching of cardiac output with venous return. The purple curves are the same as the vascular function curve in Figure 23-9C. The red cardiac function curves represent Starling's law. In A, point A is the single RAP at which venous return and cardiac output match. A transient increase in RAP from 2 to 4 mm Hg causes an initial mismatch between cardiac output (point B) and venous return (point B′), which eventually resolves (B′C′A and BCA). In B, permanently increasing blood volume (transfusion) shifts the vascular function curve to the right (as in Fig. 23-10A), so that a match between the cardiac output and venous return now occurs at a higher RAP (point B). In C, permanently increasing cardiac contractility shifts the cardiac function curve upward, so that a match between the cardiac output and the venous return now occurs at a lower RAP (point B).

We can now ask, Do the cardiac output and venous return depend on RAP? Does RAP depend on the cardiac output and venous return? The answer to both questions is an emphatic yes! They all depend on each other. There is no absolute dependent or independent variable in this closed circuit because, in the steady state, venous return and cardiac output must be equal. Venous return (from the vascular function curve) and cardiac output (from the cardiac function curve) can be equal only at the single point where the two curves intersect (point A in Fig. 23-11A). Only transient and small deviations in these two curves are possible unless either or both of the function curves change in shape.

Imagine what would happen if RAP transiently increased from 2 to 4 mm Hg (see Fig. 23-11A). Starling's law states that the increased ventricular filling initially increases cardiac output from starting point A (5 L/min) to point B (7.3 L/min). Simultaneously, the increase in RAP causes a decrease in the driving pressure ΔP for venous return, that is, the difference CVP − RAP. Thus, the increase in RAP would initially reduce the venous return from point A (5 L/min) to point B′ (3 L/min). This imbalance between cardiac output and venous return cannot last very long. The transiently elevated cardiac output will have two effects. First, by sucking the right atrium dry, it will tend to lower RAP. Second, by pumping blood out of the heart and toward the veins, it will raise CVP. As a result, the difference CVP − RAP increases, and venous return moves from point B′ to C′. Because RAP is lower at point C′ than at B′, the cardiac output must also be less at point C than at point B. The imbalance between C and C′ (6.5 versus 4 L/min) is now less than that between B and B′ (7.3 versus 3 L/min). In this way, venous return gradually increases, cardiac output gradually decreases, and RAP gradually decreases until once again they all come into balance at point A. Thus, the cardiovascular system has an intrinsic mechanism for counteracting small, transient imbalances between cardiac input and output.

The only way to produce a permanent change in cardiac output, venous return, and RAP is to change at least one of the two function curves. The vascular function curve may be any one of a large family of such curves (see Fig. 23-10A, B), depending on the precise blood volume, venomotor tone, and arteriolar tone. Thus, a transfusion of blood shifts the vascular function curve to the right, establishing a new steady-state operating point at a higher RAP (point A → point B in Fig. 23-11B). Similarly, vasodilation would rotate the vascular function curve to a steeper slope, also establishing a new steady-state operating point at a higher RAP.

The cardiac function curve also may be any one of a large family (see Fig. 22-12C), depending on afterload, heart rate, and, above all, the heart's contractile state. Therefore, increasing contractility by adding a cardiac glycoside such as digitalis (see p. 530) shifts the cardiac function curve upward and to the left, establishing another steady-state operating point (point A → point B in Fig. 23-11C). Both physiological and pathological conditions can reset the vascular and cardiac function curves, resulting in a wide range of operating points for the match between venous return and cardiac output.