Cardiorespiratory Physiology of Pregnancy

The first measurement of cardiac output in a normal pregnant woman was performed by Lindhard¹ in 1915. Using a nitrous oxide technique, he found an increase in cardiac output of approximately 50% during pregnancy. In 1926, Gammeltoft² first observed that late in pregnancy the cardiac output tended to decline toward nonpregnant levels. He attributed this change to a decline in maternal physical activity. The careful observations and measurements reported by Burwell and associates^3,4 in 1938 helped greatly to delineate many of the pregnancy-related hemodynamic changes. They noted changes in heart rate, blood pressure, venous pressure, and cardiac output in serial basal measurements in a small group of women studied throughout pregnancy. These changes persist throughout pregnancy and generally regress after delivery^5,6; however, there may not be a return to prepregnancy levels.⁷ Invasive and noninvasive techniques have elucidated further the alterations in cardiovascular physiology during normal pregnancy.

Cardiac Output

Cardiac output increases significantly during pregnancy. The increment of 50% initially reported by Lindhard¹ in 1915 is similar to the values reported in the 1980s and 1990s using more refined methodology, such as cardiac catheterization and dye dilution measurements.^8,9,10 Results of several serial studies suggest that cardiac output rises significantly the first trimester of pregnancy,^10,11,12 continues to rise through 26 to 34 weeks’ gestation,^9,13,14 and is maintained at an elevated level to term if measured while the patient is in lateral recumbency.^9,11 The net result of these hemodynamic alterations is that early in pregnancy the increase in cardiac output at rest is accomplished predominantly through an increase in stroke volume. As pregnancy advances, however, heart rate increases, and stroke volume decreases to nonpregnant levels; the increase in heart rate is responsible for maintaining the elevated cardiac output. A review of the changes in cardiac output during pregnancy suggests that cardiac output varies widely among individuals with singleton pregnancies in the third trimester,¹⁵ and the increase in cardiac output is even greater in a twin pregnancy.^16,17 Ueland and coworkers⁹ found a decline in cardiac output, however, between 38 and 40 weeks’ gestation for all positions studied (i.e., supine, sitting, and lateral recumbency), but the only statistically significant decline was encountered in the patients studied while supine. Doppler and echocardiographic studies have confirmed these changes in cardiac output.⁶

Heart Rate

The rise in heart rate during pregnancy reaches 10 to 15 beats/min above nonpregnant values for most maternal positions.⁹ With the exception of the sitting position, the increase seems to be progressive until term.

Stroke Volume

Stroke volume is elevated to peak levels early in pregnancy (20 to 24 weeks’ gestation) and progressively declines to nonpregnant levels at term.⁹ For the supine position, the stroke volume at term is 20 mL less than that measured 6 to 8 weeks postpartum.

Respiratory Physiology

Respiration includes the transfer of oxygen from the external air to the blood in the maternal pulmonary capillaries and oxygen consumption by the peripheral tissues. The maternal rate (including fetal) of oxygen consumption rises progressively during pregnancy, reaching a peak of 20% above nonpregnant levels.¹⁸ The oxygen consumption of women at rest (but not basal) increases even more during pregnancy than does the basal value.^19,20 The total rate of maternal oxygen consumption in the pregnant woman is the sum of the rates of each of her tissues and that of the fetus and placenta. Approximately two thirds of the total increment in basal oxygen consumption during pregnancy is accounted for by the combined consumption of the fetus, placenta, and uterine muscle. The remaining increment is due to the increased oxygen requirements of the maternal myocardium, kidneys, muscles of ventilation, and mammary glands. As might be predicted, the oxygen consumption is greater when measured in women with multiple gestations compared with singleton gestations.²¹ The onset of labor is associated with a doubling in the average oxygen consumption and a threefold increase when measured during a uterine contraction.²² The discrepancy between increases in cardiac output and oxygen consumption and the concomitant early fall in peripheral vascular resistance were compared by Burwell²³ with the hemodynamic changes of an arteriovenous fistula. From a teleologic standpoint, these alterations seem to anticipate reproductive needs because there is a rich flow of well-oxygenated blood to the uterus during early pregnancy when organogenesis is occurring and before fetoplacental circulation is fully developed. The arteriovenous oxygen difference widens to nonpregnant levels later in pregnancy. The well-known pregnancy-associated changes in maternal ventilation are discussed elsewhere in this volume.

Ventricular Function

The steroid hormones of pregnancy, particularly estrogen, not only affect vascular resistance, but also seem to have a direct effect on myocardial muscle. The mode of action is uncertain, but estrogens are structurally similar to the cardiac glycosides. It has been postulated by Csapo²⁴ that estrogens may alter the actomyosin adenosine triphosphatase (ATPase) relationship in the myocardium, increasing myocardial contractility. More recently, measurements of systolic time intervals have been performed in humans during the first and second trimesters.²⁵ The pre-ejection period of left ventricular systole is reduced, particularly in the second trimester. The cause of this change in ventricular dynamics is unclear, but it could be due to a direct action of steroid hormones on myocardial cells.²⁶ Male transsexuals given high-dose estrogen therapy show increases in myocardial contractility.²⁷

Dramatic changes in cardiovascular function have been shown by echocardiographic studies during the course of normal pregnancy. Gilson and colleagues²⁸ studied indices of ventricular remodeling in primigravid women during gestation compared with the postpartum state. Left ventricular diastolic dimension did not change, whereas there was a significant increase in left ventricular end-diastolic volume. The ventricular diameter-to-length ratio did not change, but the left ventricular posterior wall thickness increased with advancing gestation. The radius-to-wall thickness ratio decreased in mid-to-late gestation. No significant change occurred in left ventricular mass.²⁸ When evaluating indices of ventricular contractility, these investigators did not find any significant changes in ejection fraction, fractional shortening, or velocity of circumferential fiber shortening. The rate-corrected ejection time and left ventricular wall stress were each less, as was the ratio of wall stress to velocity of circumferential fiber shortening (a load-independent assessment of left ventricular function). They concluded that independent of the increased preload and heart rate (reflected in velocity of circumferential fiber shortening) and of the decreased systolic blood pressure (reflected in wall stress), the intrinsic myocardial contractility was enhanced.²⁸ Other investigators studying a smaller number of patients found no change in ventricular contractility.²⁹

Vascular Resistance

Maternal vascular resistance (systemic and pulmonary) decreases progressively throughout gestation.⁶ This decline mirrors that of the decline in systolic and diastolic blood pressures.³⁰ Hypotheses for the cause of this vasodilation include decreased pressor responses to angiotensin II in pregnant women, increased vasodilatory prostaglandin synthesis, and alterations of extracellular matrix proteins in the vasculature. Nitric oxide (NO) has been studied intensively and is a frontrunner for this role.^31,32 NO is a short-lived molecule that is synthesized by NO synthase in the endothelium after stimulation and is released into the interstitium to exert its effect locally. In vascular smooth muscle, NO stimulates cyclic guanosine monophosphate (cGMP), which reduces cellular calcium concentrations and inhibits contraction. NO and cGMP have been found in increased concentrations in pregnant animals and can be reduced by the estrogen receptor antagonist tamoxifen, implying that estrogen modulates these increased levels of NO. Weiner and Thompson³² hypothesized that the varying effects of pregnancy on vascular resistance in different organs could be due to differential levels of estrogen receptors on the endothelium.

The role of steroid hormones on NO levels has been elaborated by studies that have suggested a role of fetal dehydroepiandrosterone sulfate (DHEA-S) in maternal vasodilation.^33,34 DHEA-S is produced by the fetal adrenal gland and is metabolized by the placenta to estrogen, which then enters the maternal circulation. In these studies, pregnant women at term were infused with DHEA-S with monitoring of estrogen and nitrite levels. Nitrite levels doubled after DHEA-S infusion and peaked at 10 minutes after infusion. Estrogen levels increased fivefold and peaked at 60 minutes after infusion. The temporal relationship of the peak levels of nitrites and estrogen suggests that there may be other pathways involved in this process and that the fetus may play a role in regulating maternal blood pressure and volume through DHEA-S production. The statistical power of these studies is limited, however, by the small number of patients studied.

Blood Volume

Blood volume, composed of plasma volume and red blood cell (RBC) mass, increases 20% to 100% during pregnancy.^35,36 The rise begins in the first trimester and continues throughout pregnancy but at a much slower rate during the third trimester.^35,37,38 The total increment seems to depend on several factors, including maternal weight³⁶ and the weight of the products of conception.^36,39,40 Expansion of the plasma volume (40% to 50%) is responsible for most of the increase in blood volume, especially between the 6th and the 24th weeks of pregnancy. Further minimal expansion occurs in the third trimester. RBC mass increases progressively throughout pregnancy. The increment in total RBC volume ranges from 20% to 35% when compared with nonpregnant values.³⁵ The “physiologic anemia of pregnancy” results from the disproportionate expansion of plasma volume compared with RBC mass. The resulting hemodilutional anemia is maximal between 16 and 22 weeks of pregnancy, but hematocrit and hemoglobin values increase in the third trimester because the RBC mass continues to expand, whereas plasma volume increases only minimally.⁴¹ Multiple gestation is associated with a greater expansion of the blood volume; however, the presence of a fetus is not required.^35,42 Even when iron stores are adequate and iron supplementation is provided, there is a decline in hemoglobin and hematocrit values late in pregnancy, but in some instances the fall is not statistically significant.^18,43,44,45

The previously reported decline in plasma and blood volume late in pregnancy was probably in error. In 1971, Chesley and Duffus⁴⁶ showed the effect of posture on plasma volume late in pregnancy and consistently found a lower plasma volume in patients studied in the supine position than in patients in lateral recumbency. They attributed the findings to inferior vena caval occlusion, resulting in improper mixing of the dye (Evans blue) and extravasation of fluid into the lower extremities because of the elevated venous pressure.

Failure of the maternal blood volume to expand normally may be associated with suboptimal maternal and fetal outcomes, including impaired fetal growth,^47,48 and may occur at 5 weeks’ gestation.⁴⁹ Longo⁵⁰ hypothesized that the fetus may regulate the increment in maternal blood volume through the production of DHEA-S by the fetal adrenal gland. As described earlier, DHEA-S is the precursor of placental estrogens, which may play a role in vasodilation through NO production, triggering a compensatory expansion of blood volume through the renin-angiotensin-aldosterone axis. It has been shown that DHEA-S itself may play a role in this vasodilation.³⁴

Structural Alteration of Blood Vessels

The sudden appearance of spider angiomas during pregnancy is common, as is the occurrence of palmar erythema. Both of these events are thought to be hormonally mediated. Burwell and Metcalfe⁵¹ reported the rapid growth of a pre-existing arteriovenous fistula during pregnancy. Rupture of splanchnic artery aneurysms is more common in women than in men before the age of 45. Most of these aneurysms rupture between the seventh and ninth months of pregnancy.⁵² Rupture of cerebral aneurysms also have been reported to occur more frequently during pregnancy. The chance that a subarachnoid hemorrhage will occur increases with each trimester of pregnancy.⁵³

Histologic changes have been reported to occur in the wall of the aorta during pregnancy,⁵⁴ but whether or not these are related to vessel strength and contribute to aortic dissection or rupture is open to question.⁵⁵ The exact mechanisms responsible for the many hemodynamic alterations associated with pregnancy are unclear, but undoubtedly many factors are involved. Neither the role played by the fetus nor that of the hypervolemia of pregnancy is known. From the data presented in the literature, the sex steroid hormones must play an important part. Observations by Ueland and Parer⁵⁶ that the intravenous infusion of natural estrogens in nonpregnant ewes produces an increase in cardiac output similar to that encountered in pregnancy in the same animals support this theory. Similarly the findings by Walters and Lim⁵⁷ of an increase in cardiac output (15%) and in plasma volume (300 mL) in women taking oral contraceptives provide further evidence of the important role of the sex steroids.

Maternal Posture

In 1953, Howard and colleagues⁵⁸ first published their data on the supine hypotensive syndrome of late pregnancy. They showed, by manometric measurements, that the fernoral venous pressure was highest when patients were in the supine position and attributed this to inferior vena caval occlusion by the gravid uterus. These findings later were confirmed by Scott and Kerr,⁵⁹ who studied pressure changes in the inferior vena cava in patients undergoing cesarean section. In 1964, Kerr and associates⁶⁰ convincingly showed, by angiographic techniques, total occlusion of the inferior vena cava late in pregnancy in a patient in the supine position. Two subsequent serial hemodynamic studies during pregnancy, using dye dilution techniques to estimate cardiac output, showed the effect of maternal posture.^9,11 Early in pregnancy (8 weeks) and up to 24 weeks’ gestation, cardiac output remained essentially unchanged when measured in the supine and lateral positions. In both studies, there was a statistically significant decrease in cardiac output (25% to 30%) when measured in the supine position at 38 to 40 weeks’ gestation. The decline was entirely attributable to a drop in stroke volume because heart rate remained relatively constant. Maternal upright positioning causes symptoms of syncope, however, in 8% of pregnant women in the first trimester of pregnancy. In the second trimester, the gravid uterus impairs venous return in standing women and can cause cardiovascular disturbances.⁶¹

Using aortography, Bieniarz and coworkers⁶² showed that the arterial side of the vascular tree also was affected to some degree by the large gravid uterus at term. They showed lateral displacement, attenuation, and elongation of the distal aorta as it was compressed against the maternal spine in supine recumbency. During hypotension, this effect becomes more pronounced.

There is conflicting information in the literature regarding the maternal cardiorespiratory response to exercise. This controversy can be attributed in part to the different ways in which the patients were exercised, the maternal posture, and the techniques used to measure the hemodynamic response.

Standard treadmill exercise (weight bearing) is associated with higher rates of oxygen consumption in pregnant women than in women in the postpartum period.⁶³ A consistent increase in oxygen consumption has not been noted when non–weight-bearing exercise has been studied in pregnant women. In a serial study by Ueland and associates,⁹ the increment in cardiac output during mild, standardized, non–weight-bearing exercise on a bicycle ergometer was found to be the same throughout pregnancy and postpartum. Serial measurements of blood oxygen transport during exercise in this same group of patients⁶⁴ showed that exercise of mild intensity was associated with an increased oxygen requirement and cost more in terms of oxygen consumption during pregnancy. In contrast, Knuttgen and Emerson⁶³ reported a slight decrease in oxygen consumption when a similar group of patients were studied in the pregnant and the nonpregnant state. Metcalfe and colleagues⁶⁵ evaluated the effect of regular exercise before and during pregnancy and noted that neither the oxygen cost of steady-state exercise nor the oxygen debt incurred by exercise was elevated during pregnancy.

With moderate exercise, the rise in cardiac output seems to be progressively smaller as pregnancy advances, suggesting a progressive decline in circulatory reserve. If the data of Robbe⁶⁶ are scrutinized carefully, a similar small increment in cardiac output in response to exercise is found to occur late in pregnancy. M-Mode echocardiography suggests that cardiac output increases secondary to increased fractional shortening early in pregnancy. Increase in left ventricular end-diastolic volume is the primary reason for increased cardiac output within pregnancy, however.⁶⁷ Bruce and Johnson⁶⁸ suggested that the decline in circulatory reserve is not related to impaired cardiac function but rather to peripheral pooling of blood resulting from the occlusion of the inferior vena cava by the uterus. There does not seem to be any appreciable buildup in oxygen debt during pregnancy when compared with nonpregnant controls.

The extent to which human uterine blood flow and oxygen consumption are altered by exercise during pregnancy is unclear. Studies considering the response of cardiac output to exercise suggest, however, that cardiac output is fixed and increases at a rate that parallels oxygen demand.¹² Animal data have shown a reduction in uterine blood flow during exercise, accompanied by an increase in the maternal hematocrit. The augmented hemoglobin level enhances the blood oxygen-carrying capacity and serves to attenuate any possible reduction in oxygen delivery.⁶⁹ Additionally, blood flow is redistributed within the uterus, and the uptake of oxygen by the uterus as a whole, and by the fetus, remains constant.⁷⁰ It is likely that fetal oxygen consumption is unaffected during mild-to-moderate exercise in human pregnancy as well. Fetal heart tracings that were obtained from women during recovery from moderately strenuous exercise revealed no evidence of fetal distress (although a consistent fetal tachycardia was noted). This group of women also had fetal nonstress tests before and after exercise. All tests were reactive, and there was no significant difference in the monitoring time required to obtain a reactive nonstress test before and after exercise.⁷¹ The effect of exercise on the uterine artery blood flow velocity shows conflicting results with reports of no change and an increase and a decrease in systolic-to-diastolic ratios during maternal exercise.⁶⁷

The hemodynamic responses to uterine contractions depend on the maternal posture in which they are studied. When measured in the supine position, a uterine contraction produces a 25% increase in cardiac output and a 15% decrease in heart rate. The resultant increment in stroke volume is 33%.⁷² If the same measurements are taken with the patient in lateral recumbency, the respective changes are only +7.6% (cardiac output), −0.7% (heart rate), and +7.7% (stroke volume). Pulse pressure while supine increases more than 26% in response to a contraction, whereas the increase is only 6% in lateral recumbency. There seems to be a remarkable degree of hemodynamic stability when the patient is on her side. Table 1 shows the similarities in peak hemodynamic values during a contraction in both positions studied. The significant differences between contractions are attributable to the inferior vena caval occlusion caused by the gravid uterus. In addition to venous obstruction, contractions while in the supine position produce complete occlusion of the distal aorta or the common iliac arteries⁷³; the blood ejected by the left ventricle is distributed transiently only to the upper half of the vascular tree, resulting in an augmented response when measurements are taken in the upper extremities. Arterial pressure measurements obtained from the femoral artery while in the supine position during a contraction show a marked decrease.⁶² During contractions, there is a rise in central venous pressure and an increase in cardiopulmonary blood volume of approximately 300 to 500 mL.^73,74

TABLE 1. Hemodynamic Measurements During a Contraction

Modified from Ueland K, Hansen JM: Maternal cardiovascular dynamics: II. Posture and uterine contractions. Am J Obstet Gynecol 103:1, 1969.

Anesthesia plays a significant role in modifying the cardiovascular response to labor and delivery. In a series of patients undergoing hemodynamic studies throughout labor and delivery, substantial differences were noted between the group receiving local and paracervical block anesthetics and the group receiving a caudal anesthetic (Table 2).⁷⁵ In the patients receiving local and paracervical block anesthetics, the cardiac output increased progressively during labor and reached a peak value of 80% above supine prelabor values immediately after delivery. Heart rate showed little change, but stroke volume rose by approximately 75%. In the patients receiving a caudal anesthetic, the cardiac output remained relatively stable throughout the first stage of labor, increased only slightly during the second stage, and reached a peak value of 60% above supine prelabor values immediately after delivery. Anesthesia did not modify the profound changes accompanying delivery because the large increments in cardiac output for both groups from second stage to delivery were similar.

TABLE 2. Maternal Hemodynamics During Labor^*

NC, no change; CO, cardiac output; P, pulse; SV, stroke volume.
^*All values expressed as % change from supine prelabor and recorded between contractions.
Modified from Ueland K, Hansen JM: Maternal cardiovascular dynamics: III. Labor and delivery under local and caudal analgesia. Am J Obstet Gynecol 103:8, 1969.

The repetitive hemodynamic changes of uterine contraction during labor can be circumvented by cesarean section delivery. The significant increment in cardiac output accompanying delivery cannot be prevented, although it can be modified to some extent by the anesthetic technique employed. Table 3 summarizes the changes in hemodynamics associated with surgical delivery using spinal anesthesia, balanced general anesthesia (thiopental, succinylcholine, nitrous oxide), and epidural anesthesia with and without epinephrine.^76,77,78,79

Some similarities should be stressed. The marked decrease in cardiac output and blood pressure after the induction of anesthesia was equivalent in the patients receiving a spinal anesthetic and in the patients receiving an epidural anesthetic with epinephrine.^76,78 The hypotension was so profound in the latter group that 10 of the 12 patients required presurgical treatment to correct the blood pressure, which was not responsive to uterine displacement. In the balanced general anesthesia group,⁷⁹ there was good hemodynamic stability throughout surgery except at the time of intubation and extubation and when awakening the patient postoperatively. During these procedures, there was a marked but transient elevation in cardiac output, heart rate, and blood pressure. The patients who received an epidural anesthetic without epinephrine^76,78 showed remarkable hemodynamic stability throughout surgery and had the smallest increment in cardiac output at delivery, reaching only 25% above preanesthesia values.

The documented blood loss at vaginal delivery is approximately 500 mL, and the loss at cesarean section is approximately 1000 mL.^35,36 The loss of this large volume of blood is well tolerated, owing in part to the hypervolemia of pregnancy and in part to the fact that there is a marked diminution in the vascular space at the time of delivery and that the lost volume is no longer needed. Table 4 summarizes the data of serial measurements of blood volume and hematocrit after delivery in a large number of patients.³⁶ By the third postpartum day, the total decline in blood volume was similar for all patients, whether delivered vaginally or by cesarean section (16.2%). The only difference was found in the hematocrit values: The patients delivering vaginally had a 6% rise in hematocrit by the third postpartum day, whereas patients have cesarean section showed a decline of similar magnitude.

TABLE 3. Maternal Hemodynamics at Cesarean Section^*

NC, no change; CO, cardiac output; P, pulse; SV, stroke volume.
^*All values expressed as % change from supine preanesthesia values.
Data from Ueland K: Personal data; Ueland K, Akamatsu TJ, Eng M, et al: Maternal cardiovascular dynamics: VI. Cesarean section under epidural anesthesia without epinephrine. Am J Obstet Gynecol 114:775, 1972; Ueland K, Gills RE, Hansen J: Maternal cardiovascular dynamics: I. Cesarean section under subarachnoid block anesthesia. Am J Obstet Gynecol 100:52, 1968; Ueland K, Hansen JM, Eng M, et al: Maternal cardiovascular dynamics: V. Cesarean section under thiopental, nitrous oxide, and succinylcholine anesthesia. Am J Obstet Gynecol 108:615, 1970.

TABLE 4. Postpartum Blood Volume and Hematocrit Changes^*

NC, no change; BV, blood volume; Hct, hematocrit.
^*Values expressed as % change from predelivery values.
Modified from Ueland K: Maternal cardiovascular dynamics: IIV. Intrapartum blood volume changes. Am J Obstet Gynecol 126:671, 1976.

This chapter is dedicated in loving memory to Kent Ueland, MD, who was a wonderful physician, friend, and mentor.