A longitudinal study of maternal hemodynamics during normal pregnancy

A Longitudinal Hemodynamics

Study of Maternal During Normal Pregnancy

A. CARLA C. VAN OPPEN, MD, PhD, INGEBORG VAN DER TWEEL, G. P. JOHAN ALSBACH, MSc, ROBERT M. HEETHAAR, PhD, AND HEIN W. BRUINSE, MD, PhD Objective: To investigate the maternal hemodynamic changes that occur during normal pregnancy. Methods: Serial hemodynamic investigations were performed throughout normal pregnancy by thoracic electrical bioimpedance monitoring in 50 healthy women. Analysis of variance with repeated measurements was used to evaluate the time course of a number of hemodynamic indices. Results: The mean heart rate (2 standard error [SE]) increased gradually from 87 f 2 beats per minute at lo-18 weeks’ gestation to 92 f 1 beats per minute at 34-42 weeks’ gestation. Mean arterial pressure decreased significantly after 14 weeks’ gestation and increased significantly after 29 weeks’ gestation. During the third trimester, mean cardiac output and mean stroke volume decreased, and mean systemic vascular resistance increased significantly. The course of cardiac output during the third trimester was not uniform in all women; it increased in nine and decreased in 41 women. A significantly higher mean cardiac output was found in nulliparous women compared with multiparous women (mean difference *SE 0.76 + 0.33 L/minute). The mean (*SE) cardiac output increased significantly from 6 (5.49 2 0.16 L/minute) to 12 weeks’ postpartum (5.91 f 0.19 L/minute). Conclusion: Mean cardiac output and mean stroke volume decreased in late pregnancy. A significant difference in mean cardiac output was observed between nulliparous and multiparous women. Cardiac output usually, but not invariably, declined during the third trimester. (Obstet Gynecol1996;88: 40-6)

Major adaptations in the maternal cardiovascular system are necessary for the normal course of pregnancy. In the absenceof these adaptations, gestational complications such as fetal growth restriction and pregnancyinduced hypertension reportedly increase.’ There is general agreement that the increase of cardiac output From

the Departments of Obstetrics and Gynecology and Biostatistics, University of Utrecht, Utrecht; and the Department of Medical Physics and Informatics, Vrije Uniuersitcit, Amsterdam, the Netherlands.

40

0029.7844/96/$15.00 PII SOO29-7844(96)00069-S

starts early in pregnancy,2 but opinions differ as to whether this rise persists throughout pregnancy. A third-trimester fall has been considered to occur only with the subject in the supine position, presumably because of impaired venous return.3 However, serial studies in which subjects were examined in the lateral position sometimesreported increasesor no changes,4,5 whereas decreaseshave also been reported2z6J7in the third trimester. To elucidate the exact nature and time course of hemodynamic changesduring normal pregnancy, serial hemodynamic evaluations should be done in the same subjectsby means of a reliable, reproducible, and noninvasive technique. Thoracic electrical bioimpedance monitoring is noninvasive, accurate,8 reproducible,’ inexpensive, and requires few technical skills. Furthermore, it has been validated in pregnancyi”,” and is suitable for serial measurementsin pregnancy.12 For an extensive analysis of the measurement of cardiac output during pregnancy, the reader is referred to a recent review article.13 The intent of this study was to 1) describe the pattern of changes in cardiac output, stroke volume, heart rate, and systemic vascular resistance measured or calculated by thoracic electrical bioimpedance monitoring during the course of normal pregnancy and puerperium, and 2) compare the values of thesehemodynamic characteristics between nulliparous and multiparous women.

Materials and Methods Design The study protocol was approved by the local Human Research Committee and subjects gave written, informed consent. Subjects were recruited from the Ob-

Obstetrics

& Gynecology

stetric Outpatient Department of the Utrecht University Hospital. All women attending our department were invited to participate in the study at their first visit. When informed consent was obtained, subjects were requested to proceed with the first measurement. Each was in the first trimester of a singleton pregnancy, and there were no concurrent obstetric or medical problems. All pregnancies were dated by last menstrual period and ultrasound examination before the 12th week of gestation. Examinations were performed at $-week intervals from the first prenatal visit until delivery, and again at 6 and 12 weeks’ postpartum. After completion of the study, only pregnancies with an entirely uneventful course were selected. None of thesewomen received medication other than iron supplements or vitamins, and none developed pregnancy-induced hypertension. All pregnancies ended in a spontaneous vaginal birth of a single infant with a birth weight between the tenth and 90th percentiles, between 259 and 294 days’ gestation. The Apgar score at 5 minutes was at least 9, and estimated loss of blood during delivery was less than 1000 mL.

Recording

Technique

All measurements were performed between 8:30 AM and noon. Body weight was recorded at each visit and an initial 15-minute period was scheduled, with a view to allow the subject to become familiar with the laboratory environment. During this period, a blood pressure (BP) cuff was fitted, and the subject assumeda comfortable sitting position. Blood pressure was measured with a cuff sphygmomanometer. Diastolic BP was taken at the fourth Korotkoff sound. Subsequently, the noninvasive continuous cardiac output monitor model 3, revision 7 (NCCOM3-R7; BoMed Medical Manufacturing, Ltd, Irvine, CA) was attached to the subject by two opposing pairs of surface electrocardiogram electrodes (Red Dot no. 2249; The 3M Company, Borken, Germany) at the root of the neck, and two opposing pairs of surface electrodes in the midaxillary line at the level of the xiphoid process, all in the midcoronal plane. Height and pre-pregnancy weight of the subject were entered manually into the cardiac output monitor. The equipment then calculated a reading of hemodynamic data in each cardiac cycle on the basis of the electrical impedance signal and the subject’s height and weight. After constant maternal heart rate (variation less than 5%) had been reached, three measurements were taken within 10 minutes. The mean of these three hemodynamic determinations was used for analysis. The instrument was adjusted so that each set of data recorded was the average of 16 cardiac cycles.

VOL.

88, NO.

1, JULY

1996

Measurements

and Calculations

The outer pairs of electrodes injected a 70-kHz, 2.5-mA current into the thoracic tissue. The resulting potential was recorded subsequently at the inner pairs of electrodes. The electrical resistance to the injected current depends on the fluid, tissue, and electrolyte characteristics in the chest. The pulsatile changes in the thoracic resistance (bioimpedance), timed to ventricular electrical depolarization and mechanical systole are used to calculate stroke volume. The cardiac output monitor determines stroke volume (in milliliters) from the Bernstein-Sramek equation,14 Stroke volume = VEPT * VET * (dZ/ dt),,,/Z,, where VEPT is the volume of electrically participating tissue in milliliters, VET is the ventricular ejection time in set, (dZ/ dt),,, is the maximum rate of change of impedance during systole in ohms/ second, and Z, is the basal thoracic impedance in ohms. The ventricular ejection time and (dZ/ df),,,/Z, are measured quantities, whereas VEPT is calculated from body dimensions according to the equation VEPT = 1.23 * L3 * (W, / W,), where L is the height in cm, W, is the observed (actual) weight in kilograms, and Wi is the ideal weight for the given height, as defined by the 1983 Metropolitan Life Insurance Company tables. During pregnancy, calculated VEPT thus varied in proportion to actual weight. The contribution of the observed weight increase during pregnancy to an increase in VEPT may be assumedto be negligible becausemost of the weight increase in pregnancy does not affect the chest. To eliminate the influence of changing body dimensions in the calculation of stroke volume, we calculated stroke volume with a constant VEPT throughout pregnancy. Therefore, pre-pregnancy weight was used instead of the actual weight in the calculation of VEPT for all measurements during pregnancy and postpartum.” Cardiac output in liters/minute was calculated according to the equation Cardiac output = stroke volume (L) X heart rate (beats per minute). Heart rate (beats per minute) was derived from the R-R interval of the simultaneously recorded electrocardiogram. Mean arterial pressure in millimeters of mercury was calculated from systolic (PsYst)and diastolic (Pdiast)BP as follows: Mean arterial pressure = (Psyst- Pdiast)/3 + Pdias..Systemic vascular resistance in dynes * second * centimeterY5 was calculated as follows: Systemic vascular resistance = 80 (mean arterial pressure - 3) /cardiac output.

Data Analysis

Subjects were not measured at predetermined time points, but at 4-week intervals after the first visit. The

van Oppen et al

Hemodynnmics

in Pregnancy

41

1. Main Reasons for Withdrawing Subjectsfrom the

Table

2. Characteristicsof 50 WomenWith Normal

Table

Study

Pregnancies n

Pregnancy-induced Birth weight

~10th

hypertension percentile

23 23

Birth weight Gestational

290th percentile age at delivery

Gestational Medications Postpartum

age at delivery 5259 days started in pregnancy hemorrhage ~1000 mL

Hemoglobin

56

mmol

Initial characteristics Maternal age at last

2294

19 9 13 8

days

3 1

/L

Admission Admission pregnancy

of the infant of the mother

in

5 12

Incomplete

measurement

series

30

Total

146

Mean Neonatal Mean

weight

(kg) surface

area

(m*)

30.7 166.2

k 4.9 -c 7.5

64.0 1.70

2 12.1 t 0.14

27123

weight increase

gestational

(y)

gain (kg) in body surface age at delivery

area

(m’)

(d)

(g)

11.7

% 4.1

0.12 279.1

* 0.04 2 8.3

3341

? 356

deviation or n. lx-e-pregnancy values

and

Red ts Over a 9-month period (March through November 1991), 196 subjectswere enrolled. Data from 116 women were omitted becausecomplications had occurred during the pregnancy; data from another 30 women were omitted because measurement series were incomplete. Table 1 lists the reasonsfor exclusion. The remaining 50 women had entirely uneventful pregnancies and deliveries. General characteristics of these 50 women are listed in Table 2. Alterations

During

Pregnancy

and Postpartum

Mean values C standard deviation of cardiac output, stroke volume, heart rate, systemic vascular resistance, mean arterial pressure, and weight are presented in Table 3. The samehemodynamic indices as a function of gestational age are presented graphically in Figure 1. Mean cardiac output at 34-42 weeks‘ gestation was significantly lower than the average mean cardiac output at lo-18,18-26, and 26-34 weeks’ gestation (mean difference 2 standard error [SE] 1.05 2 0.15 L/minute; P < .OOl). Mean cardiac output at 34-42 weeks’ gestation was significantly higher than the average mean

3. CardiovascularIndicesDuring and After Pregnancy Time Indices

CO (L / min) SV (mL) HR (beats SVR (dyne

/ min) . set . cm-‘)

7.26 t- 1.56 85 -t 21 87 2 14

MAP

(mmHg)

Weight

(kg)

67 2 14

Hemodynamics

in weeks’

gestation 26 -34

Postpartum 34-42

7.38 2 1.63 82 2 21

6.37 k 1.48 70 2 14

92 2 14 932 k 240 84 2 7

92 2 7 1118 2 325 86 2 7

70 2 14

73 t 14

= heart

in Pregnancy

rate;

SVR

= systemic

vascular

MAP

12

5.49 e 1.13 70 2 14 79 t 7 1274 ? 325

5.91

2 1.34 75 2 21 80 2 7 1175 2 304 8.5 2 7

86 -+ 7 67 2 14

76 k 14 resistance;

weeks

6

-t 1.63 85 2 21 90 2 14 901 Z 224 84 2 7

7.60

CO = cardiac output; SV = stroke volume; HR Data are presented as mean 2 standard deviation.

van Oppen et al

period 18-26

lo-18

966 2 226 87 + 7

42

weight

I’re-pregnancy body Nulliparasimultiparas Gestational changes Gestational Gestational

menstruation

Data are presented as mean t standard Gestational changes: changes behveen values at 34-42 weeks’ gestation.

individual changes of hemodynamic characteristics during and after pregnancy were evaluated by multivariate analysis of variance using a model of repeated measurements. Missing values impede the analysis, and therefore we clustered all observations during pregnancy in four equal periods: lo-18 weeks’ gestation (mean 14.4), 18-26 weeks’ gestation (mean 21.9), 26-34 weeks’ gestation (mean 29.4), and 34-42 weeks’ gestation (mean 36.7). When more than one measurement for one subject was available during a given time period, the average was used for analysis. Two time points were studied postpartum: 6 weeks and 12 weeks. Analysis of variance with repeated measurements was used to evaluate the time course of the hemodynamic characteristics, and to study differences between groups of women. When multivariate F tests were significant, univariate t tests were applied to test various contrasts (ie, prespecified differences between measurements at different time periods). The degrees of freedom for these t tests were adjusted using HuynhFeldt’s epsilon to correct for the dependency between measurements at successive time periods (Statistical Products and Service Solutions, version 4.01; SPSS,Inc., Chicago, IL).

Table

Height (cm) I’re-pregnancy

= mean

arterial

67 i 14 pressure.

Obstetrics

& Gynecology

PO., PnYm

I -,-‘:-,

Figure 1. Mean cardiac output, stroke volume, heart rate, systemic vascular resistance (SVR), mean arterial pressure (MAP), and maternal weight during pregnancy at 6 and 12 weeks after delivery. Mean values for all women (n = 50), nulliparous women (n = 27), and multiparous women (tt = 23) are shown.

postpartum value (mean difference % SE 0.67 ? 0.17 L/minute; P < .OOl), and increased significantly from 6-12 weeks’ postpartum (mean difference + SE 0.43 2 0.15 L/minute; P = ,007). Mean stroke volume at 34-42 weeks’ gestation was significantly lower than the average mean value at 10-18, 18-26, and 26-34 weeks‘ gestation (mean difference + SE 14.2 + 1.7 mL; P < .OOl), and increased from 6 to 12 weeks’ postpartum by 5.1 + 1.7 mL (P = .005). The average mean postpartum value for stroke volume was significantly lower than mean stroke volume at lo-18 weeks’ gestation (mean difference ? SE 12.7 ? 2.3 mL; P < ,001). The average mean postpartum value was not significantly different from the value at 34-42 weeks’ gestation. Mean heart rate (+SE) increased significantly from lo-18 to 34-42 weeks‘ gestation by 5.5 + 1.5 beats per minute (P < ,001). This increase was not purely linear but slightly curved (P = ,089). Mean heart rate at 6 weeks’ postpartum was not significantly different from mean heart rate

VOL.

88, NO.

1, JULY

1996

at 12 weeks’ postpartum. The average mean postpartum heart rate was significantly lower than mean heart rate at lo-18 weeks’ gestation (mean difference +- SE 7.3 t 1.5 beats per minute; P < ,001). Mean systemic vascular resistance at 18-26 weeks’ was significantly lower than at lo-18 weeks’ gestation (mean difference ? SE 65.3 + 21.5 dyne * second * cme5; P = .004). Mean systemic vascular resistance at 34-42 weeks’ gestation was significantly higher than the average mean value at lo-18,18-26, and 26-34 weeks’ gestation (mean difference ? SE 184.8 2 30.5 dyne * second * cm-5; P < .OOl). The average mean postpartum value was significantly higher than mean systemic vascular resistance at 34-42 weeks’ gestation (mean difference ? SE 106.8 ? 44.0 dyne * second * cme5; P = .02). The average mean arterial pressure at lo-18,34-42,6 weeks’ postpartum, and 12 weeks’ postpartum was significantly higher than mean arterial pressure at 18-26 and 26-34 weeks’ gestation (mean difference ? SE 2.0 -C 0.6 mmHg; P = .002). Mean arterial pressure at 6 weeks’ postpartum was not significantly different from mean arterial pressure at 12 weeks‘ postpartum. Thirty-two of the 50 women had a cardiac output starting value (the value at lo-18 weeks’ gestation) lower than 7.5 L/minute, whereas 18 women had a starting value at least 7.5 L/minute. The cutoff point of 7.5 L/minute was chosen because it corresponded to approximately the maximum cardiac output value throughout the entire pregnancy for all subjects. A graphic presentation of changes in cardiac output in “low” and “high” starters is given in Figure 2. A significant interaction between starting value (low / high) and the subsequent course of cardiac output was found: in the low group (II = 32) mean cardiac output (*SE) increased significantly from lo-18 to 18-26 weeks’ gestation by 0.56 ? 0.19 L/minute (P = .005), whereas in the high group (n = 18) mean cardiac output at lo-18 weeks was not significantly different from mean cardiac output at 18-26 weeks’ gestation. From 26-34 weeks’ gestation and onward, the time course of mean cardiac output was the same for both groups. Mean cardiac output (+SE) increased by 0.42 5 0.16 L/minute between 6 and 12 weeks’ postpartum. By multivariate analysis, this increase was nonsignificantly related to the infant feeding method (P = .089) and the use of oral contraceptives (OC) (P = .097). Parity Differences The population consisted of 27 nulliparous and 23 multiparous women. Apart from parity, there were no significant differences in general characteristics between both groups (data not shown). No significant interaction was found between the courses of cardiac output,

van Oppen et al

Hemodymvtzics

in Pregnancy

43

~

All - -

5

High starters

I :

Low starters

1

Post partum

. IO-16

16-26

26-34

34-42

Gestational

+6

age

+12

in weeks

Figure 2. Mean cardiac output in all women (n = 50), and in women with “high” (at least 7.5 L/minute; PI = 18) and “low” (below 7.5 L/minute; n = 32) initial cardiac output.

stroke volume, heart rate, systemic vascular resistance, mean arterial pressure, and parity; the time course of these indices during gestation was the same for nulliparous and multiparous women (Figure 1). In absolute terms, only mean cardiac output was significantly greater in nulliparous women (mean difference 2 SE 0.76 i: 0.33 L/minute; P = .027).

Discussion In this study, we measured the hemodynamic changes during normal pregnancy by thoracic electrical bioimpedance monitoring. Mean cardiac output reached its highest value in the second trimester and was subsequently found to decrease. Our data confirm many of the observations made previously in longitudinal studies by invasivei and noninvasive2P6,7 methods. Measurements were carried out with the subjects in the sitting position. We feel that the sitting position is the most comfortable when performing measurements by thoracic electrical bioimpedance during pregnancy. In the lateral position, the electrodes that are placed at the level of the xiphoid process at the midaxillary line can be compressed, resulting in patient discomfort. In the nonpregnant state, the method is applied mostly in the supine position. Clark et al” performed a study in which cardiac index measurements obtained by thoracic electrical bioimpedance

44

van Oppen et al

Hemodynarnics in Pregnancy

monitoring were correlated to measurements obtained by the oxygen extraction (Fick) technique in normal, late third-trimester pregnancies. Measurements obtained in the sitting position in seven subjects showed a poor correlation (r = 0.554). As all individual values for cardiac index were published,” we were able to repeat the regression analysis and calculate the coefficients of correlation. The calculation by Clark et al” was inaccurate. The correct coefficient of correlation in the sitting position is r = 0.794. There are indications toward a lower cardiac output in the sitting position, compared with the commonly applied left lateral position, for hemodynamic studies during pregnancy.16,17An explanation could be the pooling of blood in dependent vessels,reducing venous return and therefore cardiac output. As a result, BP falls, activating baroreflex to maintain pressure by increasing vascular tone and systemic vascular resistance.“j However, in a longitudinal study, in which relative changes over time are investigated, a consistently lower value is of no importance. Becauseof the 23% reduction in mean cardiac output at 12 weeks’ postpartum compared with the mean value at the first measurement in gestation, the rise in mean cardiac output must have appeared early in pregnancy. This has also been found in studies in which measurements were performed before as well as during pregnancy.6,‘3 In our study, mean cardiac output decreased during the third trimester. However, the changes of cardiac output through the course of pregnancy showed substantial interindividual variation. Despite the decrease of mean cardiac output for our total study group of 50 women, nine women actually showed an increase of cardiac output during the third trimester. In these nine women, the increase of cardiac output from 26-34 to 34-42 weeks’ gestation varied from 0.05 to 1.12 L/minute (from 0.9 to 16% of cardiac output at 26-34 weeks’ gestation). General characteristics in this subset of nine women did not differ from the other subjects (Table 2). Eighteen of the 50 women had a higher cardiac output starting value, at or above 7.5 L/minute. In this subgroup, no increase in mean cardiac output was observed during the course of pregnancy (Figure 2). Obviously, some women adapt earlier than others, and a high initial cardiac output level obviates a further increase of cardiac output. In our study population, this initial cardiac output level did not correlate with birth weight and there was no association between parity and the cardiac output value at the first measurement (data not shown). The difference between the “high” and the “low” groups cannot be due to a difference in gestational age at the time of the first estimation of cardiac output. The mean gestational age at which the first

Obstetrics

& Gynecology

estimation was made in the “high” group was 13.9 (range 11.6-16.0) versus 14.4 (range 12.5-17.4) weeks’ gestation in the “low” group. Studies on individual courses of cardiac output in normal pregnancy are few. In a serial study of healthy nulliparous subjects in whom cardiac output was measured by Doppler echocardiography, Easterling et al” established a fall in mean cardiac output from 34 weeks’ gestation until the last measurement before delivery. However, the performance of individual subjects between 34 weeks‘ and delivery was variable: cardiac output fell by at least 1.0 L/minute in 29% of subjects but rose by 1.0 L/minute or more in 9%.” Caton and Banner7 reported changes in cardiac output measured by Doppler echocardiography in 20 study subjects with normal pregnancy. All individual courses of cardiac output were presented in a graph and showed great variation. Lees et al4 reported mean and individual values of cardiac output measurement by the indicator dilution technique in five subjects at each trimester throughout pregnancy. Two of the five subjects showed a decrease between the second and third trimester; three of the five subjects showed an increase. Conflicting results concerning the course of mean cardiac output during the third trimester in longitudinal studies215r7 may be explained by this interindividual variation. Therefore, analysis of the available longitudinal studies on cardiac output changes during pregnancy is not conclusive as to what really happens to cardiac output during the third trimester. Most series were small. Averaging the individual cardiac output values sometimes resulted in a fa112T7and sometimes in a rise5 of mean cardiac output during the third trimester. In our large study population, mean cardiac output decreased during the third trimester. A plausible explanation for the fall in stroke volume and cardiac output during the third trimester seems to be the compression of the inferior vena cava by the uterus, thereby reducing venous return and leading to a reduction in atria1 diastolic filling pressure and cardiac output. Compression of the inferior vena cava by the uterus has been described to happen predominantly in the supine position in late pregnancy.3 However, it has also been demonstrated in the upright position.” Compression of the inferior vena cava may also occur in the left lateral or sitting position and impede venous return flo~.~’ Variation of the course of cardiac output during the third trimester could be due to individual factors such as uterine volume or weight, degree of (paravertebral) venous collateral circulation, anatomy of the lumbar spine, and the position of the inferior vena cava along the lumbar spine. Conflicting results concerning the course of mean cardiac output during the third

VOL.

88, NO.

1, JULY

1996

trimester in longitudinal studies may be explained by these interindividual factors. The effect of parity on the course of cardiac output in normal pregnancy has never been investigated specifically. Our data showed consistent differences between nulliparous and multiparous women. Nulliparous women showed higher absolute values of stroke volume, heart rate, and mean arterial pressure. These differences were not significant, but resulted in a significant difference for cardiac output (mean difference 2 SE 0.76 ? 0.33 L/minute). The pattern of change was similar in the two groups. The major factor determining the cardiac output is the rate of venous return,21 which is determined by arterial pressure divided by total peripheral resistance. Increased blood volume could augment cardiac output by increased venous return.21 There is no evidence of differences in increases in the volumes of plasma and whole blood related to paritye2’ The American Heart Association recommends that the onset of phase V (disappearance of sound), marked by the last sound heard, be used for defining diastolic pressure in (nonpregnant) adults. The onset of phase IV or muffling is harder to recognize and is subject to greater interobserver and intraobserver variability. In Europe, the current consensus on the diastolic end point in pregnancy favors the fourth sound, because in pregnancy sounds may continue until the zero point, whereas the fifth sound is recommended in nonpregnant subjectsz3 In this study, mean cardiac output at 12 weeks postpartum was significantly higher than mean cardiac output at 6 weeks postpartum, A similar difference was found in one study,6 whereas three others5,24,25 found no differences between 6 and 12 weeks postpartum. Because it is known that the use of combined estrogenprogesteron OCs may cause an increase in cardiac output, 26 we examined the effect of the use of OCs on the course of mean cardiac output. The increase between 6 and 12 weeks postpartum could have been caused by the use of OCs (P = ,097) or lactation (P = .089), but the numbers of subjects were small and the differences did not reach statistical significance. It must be noted that there was an association between infant feeding method and contraception: 18 of the 34 women who bottle-fed their infants at 6 and 12 weeks postpartum, versus none of the 11 women who breast-fed their infants at 6 and 12 weeks postpartum, took OCs. A lack of association between breast-feeding and a significant change in cardiac output was found in other reports,24,27 although the number of subjects in the one study was sma11,24 whereas the study period in the other study covered only 14 postnatal daysz7

van

Oppen

et al

Hemodynamics in Pregnancy

45

References

effects of 1988;72:550-2.

1. Rosso P, Donoso E, Braun S, Espinoza Maternal hemodynamic adjustments retardation. 2. Duvekot

Gynecol JJ, Cheriex

pregnancy consecutive

R, Fernandez in idiopathic

C, Salas SP. fetal growth

Obstet Invest 1993;35:162-5. EC, I’ieters FA, Menheere PI’, Peeters

changes in hemodynamics adjustments triggered

and volume by a primary

vascular tone. Am J Obstet Gynecol 1993;169:1382-92. 3. Kerr MG. The mechanical effects of the gravid uterus pregnancy. J Obstet Gynaecol Br Commw 1965;72:513-29. 4. Lees MM, Taylor at rest throughout 1967;74:319-28. 5. Mabie WC, longitudinal

LG, Sibai m normal

Am J Obstet Gynecol 1994;170:849-56. 6. Atkins AF, Watt JM, Milan l’, Davies nal

study

in

late

of cardiovascular

BM, Arheart KL. human pregnancy.

P, Crawford

dynamic

changes

8. Fuller HD. The validity of cardiac output impedance: A meta-analysis. Clin Invest 9. Jewkes C, Sear JW, Verhoeff F, Sanders of cardiac

output

by

JP. A comparison

A

throughout

preg-

12.

Heethaar RM, van Bruinse HW. Thoracic toring stroke volume

Reprod Biol 13. van Oppen normal 310-8.

electrical Obstet

A

RH, critical

14. Bernstein DP. A new stroke bioimpedance: Theory and

Bruinse

HW. Obstet

volume rationale.

Obstet Gynecol TR, 16. Easterling

1969;104:856-64. Schmucker BC,

46

et al

Oppen

normal

term

oxygen

preg-

Cardiac

output

Gynecol

in

Metcalfe J. Maternal cardiovasof gestational age on the to posture and exercise. Am J

Obstet Gynecol Kerr MG, Scott

Hetnod~ynnrt~ics

in Pregnancy

TJ, Schmucker and preeclamptic

partum.

1993;169:183-8. DB, Samuel

E. Studies

Am

J Obstet

BC, Millard SF. Maternal pregnancies: A longiR. An orthostatic study. Am

of the inferior

vena

cava

23.

Obstet Gynecol 1967;98:393-403. l’etrie JC, O’Brien ET, Littler WA, de Swiet M. Recommendations on blood pressure measurement. BMJ 1986;293:611-5.

CJ, Donovan

Robson changes

physiology.

JC. Blood

volume

SC, Hunter S, Moore during the puerperium:

diographic Capeless

study. EL, Clapp

to their 883-6.

preconception

Br J Obstet JF. When

Br J Obstet

Gynaecol

reprint

during

pregnancy.

Am

in

Am

J

W. Haemodynamic and M-mode echocar-

Gynaecol 1987;94:1028-39. do cardiovascular parameters

values?

J

7th ed. Philadelphia:

M, Dunlop A Doppler

26. Walters WAW, Lim YL. Haemodynamic oral contraceptives. J Obstet Gynaecol SC, Dunlop W. Haemodynamic 27. Robson

J Obstet

Gynecol

changes Br Commw effects

return 1991;165:

in women taking 1970;77:1007-12. of breast-feeding.

1989;96:1106-8.

requests

to:

A. Carla C. van Oppen, MD, PhD Department of Obstetrics Utrechf University Hospital P.O. Box 85.500, F05.829 Utrecht, 3508 GA The Netherlands

Received October 4, 1995. Received i,r revised form ]muary 23, 1996. Accepted February 9, 1996. Copyright

TJ. The

post

1996;87:

equation for thoracic electrical Crit Care Med 1986;14:904-9.

Benedetti

and

Gynecol

GD, Benedetti profile during

tudinal study. Obstet Gynecol 1990;76:1061-9. Schneider KTM, Bung P, Weber S, Huch A, Huch uterovascular syndrome: A prospective, longitudinal

Address

extrac-

FA, Brouwer FAS, Suitable for moniEur J Obstet Gynecol

review.

15. Ueland K, Navy MJ, Peterson EN, cular dynamics. IV. The influence maternal cardiovascular response

van

in

Oppen AC, Ottenhoff electrical bioimpedance: durmg pregnancy?

1995;58:183-90. ACC, Stigter

pregnancy:

index

JM, Lee W, Hankins central hemodynamic

pregnancy

Benedetti in normal

Obstet

Lund

25.

bioimpedance and Gynecol 1994;83:669-72.

pregnancy.

22.

measurement by thoracic Med 1992;13:103-12. DJ, FoPx P. Non-invasive bioimped-

DB, Pivarnik change and

third-trimester 1991;164:883-7.

Easterling TR, hemodynamics

during

21.

24.

electrical

stress

late pregnancy. BMJ 1964;1:532-3. Guyton AC. Textbook of medical WB Saunders, 1986:272-86.

during

thoracic

of cardiac

nancy using thoracic tion (Fick) techniques.

19.

20.

ante: A study of reproducibility and comparison with thermodilution. Br J Anaesth 1991;67:788-94. DJ, Greenspoon JS, Ouzounian JG. Measurement of cardiac 10. Masaki output in pregnancy by thoracic electrical bioimpedance and thermodilution. Am J Obstet Gynecol 1989;161:680-4. 17. Clark SL, Southwick J, Pivarnik JM, Cotton DB, Hankins GD, Phelan

18.

JS. A longitudi-

nancy. Eur J Obstet Gynecol Reprod Biol 1981;12:215-24. 7. Caton D, Banner TE. Doppler estimates of cardiac output pregnancy. Bull N Y Acad Med 1987;63:727-31.

measurement

normal Gynecol

SH, Scott DB, Kerr MG. A study of cardiac output pregnancv. J Obstet Gynaecol Br Commw

DiSessa TG, Cracker study of cardiac output

17. Clark SL, Cotton TJ, et al. Position

LL. Early

homeostasis are fall in systemic

orthostatic

hemodynamic

Gynecologists,

0

1996

by

Published

The

American

by Elsevier

College Science

of Obstetricians

and

Inc.

Obstetrics b Gynecology