DOI: 10.1111/j.1471-0528.2006.01056.

x
www.blackwellpublishing.com/bjog
General obstetrics

Uterine blood flow during supine rest and

exercise after 28 weeks of gestation
RM Jeffreys,a W Stepanchak,a B Lopez,a,b J Hardis,a JF Clapp IIIa,b,c
a Department of Obstetrics and Gynecology and the Schwartz Center for Metabolism and Nutrition, MetroHealth Medical Center, Cleveland,

OH, USA b Department of Reproductive Biology, Case Western Reserve University School of Medicine, Cleveland, OH, USA
c Department of Obstetrics and Gynecology, University of Vermont College of Medicine, Burlington, VT, USA

Correspondence: Dr JF Clapp III, Department of Obstetrics and Gynecology, MetroHealth Medical Center, 2500 MetroHealth Drive, Cleveland,
OH 44109, USA. Email jfclapp@metrohealth.org

Accepted 19 July 2006. Published OnlineEarly 15 September 2006.

Objective To test the null hypothesis that, after 28 weeks of
gestation, uterine blood flow during supine rest and supine
exercise is no different than uterine blood flow at left-lateral rest.
Design In vivo experimental study in pregnant women.
Setting Department of Obstetrics, MetroHealth Medical Center,
Cleveland, OH, USA.
Population Fourteen, physically active, late-pregnant women who
continued supine exercise throughout gestation.
Methods Studies were carried out between 29 and 38 weeks of
gestation. Maternal blood pressure, maternal heart rate, and
ultrasound estimates of volume blood flow in the right ascending
branch of the uterine artery were obtained serially at rest in the
left-lateral position, at rest in the supine position, during and
immediately after 10 minutes of supine exercise, and again at rest
in the left-lateral position. Exercise sessions included alternating
60- to 90-second periods of abdominal crunches and leg exercise at
moderate/high intensity (Borgs rating of perceived exertion 14 1).
Main outcome measures Blood pressure, heart rate, and uterine
artery volume flow.
Results Data are presented as the mean SD. Maternal heart rate
and blood pressure were unchanged at supine rest but increased
during supine exercise (heart rate increased from 76 9 to 98 12
beats per minute, mean arterial pressure increased from 81 6 to
102 12 mmHg). Volume flow fell from 410 93 to 267 73 cc/
minute after 5 minutes of supine rest and then, during supine
exercise, increased to 355 125 cc/minute. Uterine artery luminal
diameter and blood flow correlated directly with tissue weights at
birth (r2 values between 0.32 and 0.59).
Conclusions In physically active women, uterine blood flow
decreases during both supine rest and supine exercise but the
decrease in the former is twice that seen in the latter.
Keywords late pregnancy, supine exercise, supine rest, uterine
blood flow

Please cite this paper as: Jeffreys R, Stepanchak W, Lopez B, Hardis J, Clapp III J. Uterine blood flow during supine rest and exercise after 28 weeks of gestation.
BJOG 2006;113:12391247.

decreased venous return and cardiac output coupled with

Introduction
This study is part of a continuing comprehensive assessment
of the interaction between the physiological changes which
occur during pregnancy and those associated with various
forms of exercise. It was designed to determine if the obstetric
recommendation that healthy, physically active, late-pregnant
women discontinue supine floor exercise is warranted.1
The recommendation is based on the results of studies that
have examined the effects of posture and exercise on maternal
haemodynamics in late pregnancy. The studies which have
examined posture indicate that uterine blood flow may fall
dramatically in the supine position due to compression of
the inferior vena cava by the pregnant uterus producing
peripheral vasoconstriction and flow redistribution away
from the splanchnic and uterine circulations. The early xenon
clearance studies of Kauppila et al.2 demonstrated that inter-
villous space blood flow fell 34% when asymptomatic, late-
pregnant women changed from a 45
left-lateral tilt position
to the supine position and the classic studies of Howard et al.3
and Lees et al.4 documented that femoral venous pressure rose
while cardiac output and blood pressure fell when late-
pregnant women changed from the position of left-lateral rest
to supine rest. The studies by Clark et al.5 also demonstrated
that cardiac output fell when healthy, asymptomatic, preg-
nant women changed from a position of lateral to supine rest
late in pregnancy and Ikeda et al.6 observed a 20% decrease in

RCOG 2006 BJOG An International Journal of Obstetrics and Gynaecology 1239

 Jeffreys et al.
the mean flow velocity in the internal carotid artery with the
same positional change in healthy women in the third tri-
mester.
Only one study has assessed the effects of supine exercise
following a period of supine rest. That study was performed
by Morris et al.7 and demonstrated that myometrial clearance
times decreased by an average of 33% from those at supine
rest during moderate intensity supine cycling. Several studies
have examined the effects of upright exercise on flow redis-
tribution away from the splanchnic and uterine circulations
with mixed results. Ultrasound measurements of portal vein
blood flow indicate that treadmill exercise decreases splanch-
nic blood flow by more than 50% during mid and late human
pregnancy,8 while xenon clearance studies indicate that uter-
ine blood flow is unchanged following bicycle exercise.9 How-
ever, multiple studies in pregnant sheep reviewed elsewhere10
demonstrate dramatic decreases in uterine blood flow during
treadmill exercise.
Nonetheless, we had multiple concerns about the use of
these findings to formulate recommendations for active,
healthy, pregnant women. First, our anecdotal experience
indicated that pregnancy outcomes were normal in women
who chose to continue supine exercise throughout their preg-
nancies. Second, the magnitude of the changes in cardiac out-
put and splanchnic blood flow at supine rest in asymptomatic
women was far less than those observed during quiet stand-
ing5,8 or weight-bearing exercise.8 Third, the women who
were studied by Morris et al.7 were presumably sedentary,
and had been at bedrest for some time. These two factors
are known to decrease plasma volume and, therefore, should
alter venous return, preload, cardiac output, and regional
flow responses.5,11,12 Also, the technique used to estimate
uterine blood flow measured the clearance rate of a tracer
injected directly into the myometrium and probably did not
reflect changes in placental perfusion per se.9 Fourth, the only
study that directly assessed the effect of exercise on intervil-
lous blood flow found no change and recent animal studies
document that the uterine vasoconstrictive response to exer-
cise is halved in late pregnancy.9,13
Therefore, we designed this study to directly assess the
effect of supine rest and exercise on uterine blood flow in
healthy, fit women during late pregnancy. The null hypothesis
tested was that uterine blood flow during supine rest and
supine exercise is no different than uterine blood flow at
left-lateral rest after 28 weeks of gestation.
A secondary interest was to determine if there is a relation-
ship between the rate of uterine blood flow and fetoplacental
weight in the human. A direct linear relationship is present in
multiple animal models but the issue has not been examined
in human pregnancy.1417 Therefore, the data were also used
to determine if there was a correlation between the basal or
resting rate of uterine blood flow in the third trimester and
fetoplacental weight at the time of birth.
1240
Methods
General
The protocol was approved by the MetroHealth Human
Experimentation Committee. All 14 subjects were healthy,
non-substance-abusing, physically active women who had
continued regular supine floor exercise beyond the 28th week
of gestation. In addition to regular floor exercise (two to six
times a week), all the women regularly (three to six times
a week) participated in other forms of exercise (30+ minutes
of aerobics, running, kick boxing, yoga, or weight training) at
moderate to high intensity (Borgs18 rating of perceived exer-
tion between 13 and 17). All were carrying singletons, had
experienced an uncomplicated antenatal course, and lived at
altitudes under 100 m. Fourteen initial studies were con-
ducted between 28 and 34 weeks of gestation (31 2 weeks)
and eight of the 14 women were able to return for a second
study between 35 and 38 weeks of gestation (36 1 weeks).
The data from the 14 initial experiments were used to exam-
ine the responses to supine rest and supine exercise and the
eight sets of paired data were used to detect any potential
gestational age effect on the outcome variables.
In all women, dating was confirmed by an ultrasound at
2426 weeks of gestation and neonatal morphometrics and
placental weight were obtained in a standardised fashion by
an independent observer at the time of birth.19 Birthweights
were corrected for gestational age, sex, and race using norma-
tive data from our obstetric populace.19,20 Then, a corrected
total tissue weight was calculated as the sum of corrected
birthweight and placental weight.
Experimental protocol
Each study was conducted in the MetroHealth Fetal Diagnos-
tic Center between 1 and 3 pm, 90120 minutes after eating.
No attempt was made to control hydration prior to study.
Upon arrival, a brief history documenting the subjects ante-
natal course, recent weight, and exercise habits was obtained.
Then she was placed in the left-lateral decubitus position and
an automated sphygmomanometer was placed on her left
arm. During the experiment that followed, multiple measure-
ments of each parameter were obtained under each of the four
experimental conditions (left-lateral decubitus rest, supine
rest, supine exercise, and lateral recovery), and the average
value obtained was used in the statistical calculations.
After 5 minutes of rest in the left-lateral decubitus position,
maternal blood pressure and heart rate were measured in
triplicate at 2-minute intervals while duplicate estimates of
volume flow in the ascending branch of the right uterine
artery were obtained using a modification of the technique
initially described and validated by Palmer et al.21 The mod-
ifications included the use of colour flow Doppler for vessel
identification and localisation of the measurement site as well
as simultaneous measurement of flow velocity and vessel
RCOG 2006 BJOG An International Journal of Obstetrics and Gynaecology

diameter. The fetal heart rate was then recorded before the
subject was asked to shift to the supine position. No attempt
was made to identify the site of placentation.
After 3 minutes of rest in the supine position, measure-
ments of maternal heart rate and blood pressure were begun
and repeated at 2-minute intervals. After 5 minutes, duplicate
estimates of volume flow through the ascending branch of the
uterine artery were obtained at the same point in the vessel as
those obtained at left-lateral rest. Then the fetal heart rate was
briefly recorded.
The subject then began supine exercise. The duration and
sequence of the exercise routine were standardised and were
designed to mimic that used in most health club floor exercise
sessions. The exercise session was 10 minutes in length and
the exercise sequence included 60- to 90-second intervals of
abdominal crunches (flexion of the head, neck, and upper
thorax on the trunk) alternating with 6090 seconds of either
leg lifts (flexion of extended legs on the trunk) or cycling
motions (alternating thigh flexion and extension on the
trunk) at a cadence of 60 moves per minute. Each subject
was asked to exercise at her usual intensity. Maternal
blood pressure and heart rate measurements were begun at
3 minutes and repeated during each set of leg exercises and at
completion of the exercise. Duplicate volume flow estimates
in the ascending branch of the right uterine artery were
obtained at the same point in the vessel during a 10- to 20-
second pause in the exercise at 5 minutes and again within
1020 seconds of exercise cessation at 10 minutes. Then a brief
measurement of fetal heart rate was obtained, the subject
returned to the left-lateral rest position, and a final set of
measurements was obtained after 2 minutes of recovery in
that position.
Throughout the study, care was taken to ensure that the
table was flat and to standardise the positions in which meas-
urements were made. When in the left-lateral position, the
subject lay with her back vertical, knees slightly flexed, left
arm extended, and right arm resting on her right side. When
at supine rest she lay flat with one or two folded towels under
her head, left arm extended laterally, and right arm at her side.
During the abdominal crunches knees were slightly flexed and
during leg exercise the trunk was flat and the left arm
extended laterally.
Measurement techniques
Maternal blood pressure and heart rate were estimated over
a 30- to 45-second interval with an Omron 711 automatic
portable monitor (Omron Matsusaka Co., Bannockburn, IL)
whose accuracy was periodically checked against a mercury
manometer and a portable heart rate monitor. In the left-
lateral position the device was zeroed at the mid-clavicular
line and, in the supine position, at the mid-axillary line. Fetal
heart rate was obtained by ultrasound. Exercise intensity was
assessed using Borgs18 rating of perceived exertion. This
RCOG 2006 BJOG An International Journal of Obstetrics and Gynaecology
Uterine flow during supine rest and exercise
employs a scale of 620. It has been shown to be a valid
estimate of relative exercise intensity in both the pregnant
and nonpregnant state and all the subjects were familiar with
the technique and used it on a daily basis to assess their
exercise intensity.18,22 A metronome was used to establish
and maintain cadence.
Volume flow in the ascending branch of the right uterine
artery was estimated by a single experienced ultrasonographer
who was blinded to both the studys hypothesis and each
subjects exercise history. All measurements were obtained
at rest because initial attempts at measurement during exer-
cise were unsuccessful.
An Acuson/Siemens Sequoia machine (Acuson/Siemens,
Malvern, PA) with a multi-frequency probe set at 4 or 5
MHz was used. Initially, the probe was placed 37 cm above
the inguinal ligament, approximately one-third of the way
between the anterior superior iliac spine and the symphysis
pubis, and the vessel segment to be insonated was identified
by its anatomic location (medial and inferior to the external
iliac artery) and waveform shape (continuous high diastolic
flow velocity) with colour flow Doppler. Then an easily rec-
ognisable segment with a reproducible waveform (23 cm
medial to its crossing over the external iliac artery) was chosen
and pulsed wave Doppler was used to measure the time-averaged
velocity of flow over three cycles (calculated electronically
from pixel densities) with simultaneous measurement of
the vessel luminal diameter at the point of insonation. No
attempt was made to correct the diameter measurement for
changes in the cardiac cycle. Volume flow through the vessel
was calculated as the product of the luminal cross-sectional
area [(diameter/2)2 3.1416] and the time-averaged flow
velocity. The coefficient of variation for the paired measure-
ments of time-averaged velocity, diameter, and estimated
volume flow were 16, 8, and 11%, respectively. Uterine
vascular resistance was calculated from the data in peripheral
resistance units (mmHg/ml/second).
Data management
The level of significance was set at a P value <0.05. The sample
size necessary to detect a change in volume flow of 10% at that
level with a power of 0.80 was calculated to be 8 using the
variance data obtained in our earlier studies of portal vein
blood flow.8 The outcome parameters of interest were mater-
nal blood pressure and heart rate, fetal heart rate, time-
averaged uterine artery flow velocity, uterine artery luminal
diameter, calculated volume flow and vascular resistance, and
the relationships among uterine artery luminal diameter, the
rate of uterine blood flow, and tissue weight at birth.
The ShapiroWilk normality test indicated that the param-
eters of interest were normally distributed in all four posi-
tions. Significant changes in the outcome parameters of
interest with changes in position and exercise were sought
with the data from the initial 14 experiments using repeated
1241
 Jeffreys et al.
measures analysis of variance followed by Tukeys test. Linear
regression was used to examine the relationships among uter-
ine artery luminal diameter, volume flow in the ascending
branch of the right uterine artery, and corrected birthweight,
placental weight, and total corrected weight (corrected birth-
weight + placental weight) in the 14 initial experiments and
separately in the eight repeat studies. Linear regression was
also used to assess relationships among other variables mea-
sured and the between-experiment variability in the flow
response. The paired t test was used to detect changes in the
parameters of interest with advancing gestation.
All data obtained in a single position were averaged and the
mean values were used in the statistical calculations. Statistix
8 software (Statistix, Tallahassee, FL) was used for data anal-
ysis. Blood pressure, heart rates, flow velocities, and luminal
diameters are reported as the mean SD while volume flow
and vascular resistance estimates are reported both numeri-
cally and as percent change from those at either left-lateral
rest or supine rest.
Results
General
Mean maternal age was 34 3 years (range 2938 years),
parity ranged from zero to four, and historical nonpregnant
mean weight and height were 59.1 4.2 kg and 169 5 cm,
respectively. At the time of study, the body mass index ranged
from 18.4 to 29.6 (24.5 2.8). The mean gestational age for
the 14 initial studies was 31 2 and 36 1 weeks for the eight
repeat studies.
Fetal heart rate at left-lateral rest was unchanged after
5 minutes of supine rest (136 10 versus 137 11 beats per
minute [b.p.m.]). However, it increased significantly (P <
0.02) after 10 minutes of supine exercise (144 15 b.p.m.)
and then returned to the initial level by the third minute of
recovery (137 8 b.p.m.). Exercise intensity using Borgs sub-
jective rating of perceived exertion ranged between 13 and 17
(moderate to very hard) with a mean rating of 15 1(hard).
Labour onset occurred at term in all women [range = 268
289 days, (277 6 days)] and was uncomplicated in 13 of 14.
The remaining labour was complicated by shoulder dystocia.
Table 1. Maternal heart rate and blood pressure responses to positional change and supine exercise in the initial 14 studies
Parameter Left-lateral rest
Heart rate (b.p.m.) 76  9
Systolic pressure (mmHg) 112  11
Diastolic pressure (mmHg) 66  6
Mean pressure (mmHg) 81  6
Data are presented as the mean  SD.
*P , 0.001 versus all resting positions.
1242
Mean birthweight was 3658 417 g (range 28104706 g) and
mean corrected birthweight was similar (3719 346 g).
Ponderal indices ranged between 2.30 and 3.12 (2.71
0.26) and placental weights ranged between 365 and 550 g.
Total corrected weight (corrected birthweight + placental
weight) ranged between 3509 and 5106 g (4188 385 g).
Neonatal courses were unremarkable.
Changes in maternal heart rate and
blood pressure
The heart rate and blood pressure responses to positional
change and supine exercise in the initial 14 experiments are
presented in Table 1. They were qualitatively and direction-
ally similar in all subjects at both gestational ages and, in the
eight women who were studied twice, there were no changes
in magnitude with advancing gestation.
Maternal heart rate was unchanged from that recorded in
the left-lateral position after 35 minutes of supine rest but
was significantly elevated (P < 0.0001) during and immedi-
ately after the last 7 minutes of supine exercise. The heart rate
obtained immediately after exercise was similar to the last
measurement obtained during exercise. After 2 minutes of
recovery at left-lateral rest, heart rate fell to a level not signif-
icantly different from the initial value obtained after
5 minutes of left-lateral rest.
Maternal blood pressure followed a similar pattern with no
significant change during supine rest but it increased signif-
icantly (P < 0.0001) during and immediately after supine
exercise. The average blood pressure obtained during exercise
was virtually identical to that recorded immediately after
exercise (1 2 mmHg). After 2 minutes of recovery, blood
pressure had returned to initial levels.
Changes in uterine artery luminal diameter
and flow velocity
The changes in uterine artery luminal diameter and flow
velocity in response to position change and supine exercise
were qualitatively similar in all 14 initial experiments. They
are detailed in Table 2. There were no significant changes
with advancing gestation in the eight women who were stud-
ied twice.
Supine rest Supine exercise Lateral recovery
78  9 98  12* 78  9
110  12 133  13* 111  15
67  10 86  12* 67  10
81  10 102  12* 82  10
RCOG 2006 BJOG An International Journal of Obstetrics and Gynaecology

Table 2. Changes in uterine artery luminal diameter and flow velocity with positional change and supine exercise in the initial 14 studies
Parameter Left-lateral rest
Luminal diameter (mm) 4.9  0.8
Flow velocity (cm/sec) 38  13
Data are presented as the mean  SD.
*P , 0.001 from left-lateral rest.
After 5 minutes in the left-lateral rest position, luminal
diameter of the ascending branch of the uterine artery was
4.9 0.8 mm. After 5 minutes of supine rest, the diameter
consistently decreased (P < 0.0001) to 4.2 0.7 mm and did
not change further after 5 and 10 minutes of supine exercise
(4.3 0.8 mm). After returning to left-lateral rest, luminal
diameter increased (P < 0.05) to a level not significantly dif-
ferent from the baseline value.
After 5 minutes of rest in the left-lateral rest position time-
averaged flow velocity in the vessel was 38 13 cm/second.
After 5 minutes of supine rest, velocity fell consistently (P <
0.009) to a mean value of 32 11 cm/second. However, after
5 and 10 minutes of supine exercise, it rose significantly (P <
0.0006) to a level similar to that obtained initially at left-
lateral rest (43 14 cm/second, P < 0.15). After 2 minutes
of left-lateral recovery, it was 37 12 cm/second, a value not
significantly different from that observed either at baseline
left-lateral rest or during supine exercise.
Estimated volume flow, vascular resistance,
and the systolic/diastolic flow/velocity ratio
in the uterine artery
The responses in the initial 14 experiments are presented in
Table 3. They were directionally similar in all subjects and the
response pattern to changes in position and exercise was not
altered with advancing gestation. However, in the eight
women studied twice, volume flow at left-lateral decubitus
rest rose significantly (394 57 to 504 103 cc/minute,
P < 0.0001) and calculated vascular resistance fell (12.1
1.9 to 10.1 2.4 mmHg/ml/second, P < 0.0001) with advanc-
Table 3. Changes in uterine blood flow and vascular resistance with positional change and supine exercise in the initial 14 studies
Parameter Left-lateral rest
Blood flow (cc/minute) 410  93
Percent change from left-lateral rest 0 234  11*
Vascular resistance (mmHg/ml/sec) 12.4  3.0
Systolic/diastolic ratio 2.0  0.4
Data presented as the mean  SD.
*P , 0.002 from left-lateral rest.
RCOG 2006 BJOG An International Journal of Obstetrics and Gynaecology
Uterine flow during supine rest and exercise
Supine rest Supine exercise Lateral recovery
4.2  0.8* 4.3  0.8* 4.8  0.9
32  11* 43  14 37  13
ing gestation while the systolic/diastolic ratio was unchanged
(2.2 0.4 and 2.1 0.4, respectively).
After 5 minutes of rest in the left-lateral position, mean
volume flow for the initial 14 studies was 410 93 cc/minute.
Five minutes after changing to the supine rest position, vol-
ume flow decreased in all women (34 11%) to a mean value
of 267 73 cc/minute (P < 0.0001). During supine exercise
volume flow rose in all women. The mean value (355 125
cc/minute) was significantly greater than that observed at
supine rest (P < 0.003) but was not significantly different
from that observed during either baseline or recovery left-
lateral rest.
The variability in the magnitude of the flow increase during
supine exercise (88 87 cc/minute or 33 32%) was also of
interest. On five occasions it actually rose above the level
observed at left-lateral rest and on three others the increase
was less than 10% above that obtained at supine rest. These
differences in response did not appear to be technical artefact
as the within-subject coefficient of variation between the 5-
and 10-minute estimates was similar to that for the estimate
itself (14 versus 11%) with an average difference between
the two estimates of 3 31 cc/minute. Therefore, an attempt
was made to identify potential explanatory variables. The
magnitude of the flow increase did not correlate significantly
with subjective exercise intensity (P = 0.54), initial flow rate
(P = 0.92), flow rate at supine rest (P = 0.94), or maternal
pressor response (P = 0.16). However, significant correlations
were present between the magnitude of the flow response and
birthweight, corrected birthweight, and total corrected weight
(r2 values between 0.44 and 0.49, P values between 0.01 and
0.005).
Supine rest Supine exercise Lateral recovery
267  73* 355  125 409  170
216  23 21  30
19.5  6* 18.6  6.5* 13.9  4.8
2.0  0.6 1.9  0.2 1.9  0.3
1243
 Jeffreys et al.
After 5 minutes of rest in the left-lateral position calculated
vascular resistance was 12.4 3.0 mmHg/ml/second. It rose in
all the 14 initial studies by an average of 56% to 19.5 6.0
mmHg/ml/second after 5 minutes of supine rest (P < 0.0001)
and then fell insignificantly (P = 0.40) after 5 and 10 minutes
of supine exercise (18.6 6.5 mmHg/ml/second). After
2 minutes of left-lateral recovery, it fell significantly (P <
0.01) to a level that was not significantly different (P =
0.29) from the baseline value.
After 5 minutes of rest in the left-lateral position the sys-
tolic/diastolic velocity waveform ratio was 2.0 0.4. As shown
in Table 3, it did not change significantly after 5 minutes of
supine rest, 10 minutes of supine exercise, or 2 minutes of
recovery in the left-lateral position. Thus, it did not reflect
either changes in volume flow or calculated vascular resis-
tance in the uterine circulation.
The relationships among uterine artery luminal
diameter, volume flow, and tissue weight at term
As anticipated, the rate of uterine blood flow in the 14 initial
studies correlated directly with the luminal diameter of the
uterine artery both at left-lateral rest (r2 = 0.37) and during
supine exercise (r2 = 0.50). Uterine blood flow also correlated
directly with corrected birthweight, placental weight, and total
corrected tissue weight (r2 values between 0.32 and 0.37).
Similar but stronger relationships were present among luminal
diameter (both at left-lateral rest and during supine exercise)
and corrected birthweight, placental weight, and total cor-
rected tissue weight (r2 values between 0.44 and 0.59, P values
all <0.01). The strongest of these is shown in Figure 1. Signif-
icant correlations were also present among the uterine artery
luminal diameter, flow rate, and corrected birthweight, placen-
tal weight, and total corrected tissue weight (r2 values between
0.34 and 0.58, P values all <0.05) in the eight studies conducted
between the 35th and 38th week.
Figure 1. The relationship between the luminal diameter of the uterine
artery during supine exercise and corrected tissue weight at term in the
initial 14 studies. R2 = 0.5889. The relationship is described by the
following formula: tissue weight = 2600 + 37.4 the uterine artery
luminal diameter in millimetres.
1244
Discussion
The observation that uterine blood flow in the ascending
branch of the uterine artery decreased 34% at supine rest is
not a new finding and is identical in magnitude to that
reported earlier by Kauppila et al.2 using a clearance tech-
nique. However, the ultrasonic technique used in the current
study, also allowed us to determine that two-thirds of the
reduction in flow was the result of a decrease in the vessels
luminal cross-sectional area and only one-third the result of
a fall in velocity. This uterine vasoconstrictive response to
a decrease in venous return and cardiac output35 is a new
finding and the fact that the luminal diameter did not
decrease further during supine exercise suggests that the ini-
tial response was near maximal for this bed in late pregnancy.
Earlier observations made in women with the supine hypo-
tensive syndrome demonstrated that uterine flow rate fell at
supine rest in symptomatic women,3,23 and it was assumed
that the fall was due to the decrease in arterial pressure which
decreased flow velocity. Likewise, the studies of Kauppila
et al.2 and Kinesella et al.24 in asymptomatic women sug-
gested that the fall in intervillous and femoral blood flow
in the supine position was due to aortocaval compression
producing a decrease in perfusion pressure. The possibility
of vasoconstriction was not considered because earlier work
had demonstrated that, in late human pregnancy, general
vascular reactivity is blunted, sympathetic nerve endings in
the uterine arteries are markedly reduced, and uterine blood
flow is pressure passive.13,2527 Thus, the finding of vasocon-
striction in a large uterine vessel during supine rest was
unanticipated but consistent with our earlier findings of
vasoconstriction in the splanchnic circulation at standing rest
in late pregnancy.8
The observation that vasoconstriction occurs during
supine rest in normal pregnant women suggests that vaso-
constriction in the uterine circulation during mid and late
pregnancy probably is a normal response to a variety of hae-
modynamic stimuli that are a part of everyday life. If this is
indeed the case, then the rate of uterine blood flow probably
varies widely throughout any 24-hour period, suggesting that
there is a large margin of reserve within the uteroplacental
circulation. Further experiments will be necessary to provide
objective evidence that this supposition is correct but it is
probable that changes in uterine vascular resistance will fol-
low those in the general systemic circulation in response to
changes in cardiac output.5
The finding that uterine blood flow during supine exercise
increases above the level present at supine rest in asymptom-
atic, physically active women is another new finding which
does not support the specific portion of the American College
of Obstetricians and Gynaecologists guidelines that prohibits
supine exercise in all women.1 Indeed, the increase in uterine
flow rate appears to be proportional to the tissue mass that is
RCOG 2006 BJOG An International Journal of Obstetrics and Gynaecology

being perfused which is clearly fetoprotective. As acute exer-
cise-induced decreases in uterine blood flow of greater mag-
nitude are clearly well tolerated by the fetus in both animals
and man,2,8,10,22 this level of blood flow should not be a reason
for clinical concern in healthy, asymptomatic women who
choose to continue supine exercise in late pregnancy. On
the other hand, the same is not true for supine rest, especially
if protracted, or in either symptomatic or anaesthetised
women.3,4,23,24,28
The increase in uterine flow during supine exercise con-
flicts with the earlier work of Morris et al.7 The reason for the
conflict is unclear but we speculate that it is the result of two
factors. First, as noted earlier, there were marked differences
between both the subject characteristics and the experimental
protocol. Second, different methodologies were used to
assess flow. In the current study the measurement of volume
flow was direct and had a relatively low coefficient of vari-
ation while the earlier study estimated myometrial clearance
rather than uterine flow and required between 6 and 12
minutes. The potential error inherent in the latter is rein-
forced by the findings of Rauramo and Forss9 who, like
Kauppila et al.,2 used an intravenous marker which produced
both a placental and myometrial wash-out curve. They
reported that upright cycle ergometry had no effect on pla-
cental blood flow.
The mechanisms underlying the decrease in uterine blood
flow at supine rest followed by an increase with the onset of
exercise appears to be related to concomitant changes in
venous return, cardiac output, blood pressure, and vascular
resistance in multiple vascular beds.210 The change in posi-
tion to supine rest was associated with concomitant
decreases in venous return and cardiac output which initi-
ated maximal vasoconstriction in the uterine vascular bed
which was maintained during supine exercise. The increase
in flow rate during supine exercise was simply a function of
the exercise-induced increase in venous return, cardiac out-
put, and perfusion pressure that increased flow velocity and
volume flow through the vessel. It is of clinical interest that
none of the changes in flow or resistance were reflected by
changes in the systolic/diastolic velocity flow profile in the
uterine artery.
The differences seen in magnitude of the flow increase
during supine exercise were not related to the exercise regi-
men. Most aspects of the exercise regimen were standardised
and the two that were not (exercise intensity and pressor
response) were unrelated to the magnitude of the flow
response. However, significant direct correlations were pres-
ent between the magnitude of the flow response and both
vessel diameter and term tissue weights. This is surprising
and appears to reflect direct relationships among uterine
artery vessel diameter (both at left-lateral rest and during
supine exercise), volume flow, and tissue weights at term.
While this requires confirmation, it suggests that the rela-
RCOG 2006 BJOG An International Journal of Obstetrics and Gynaecology
Uterine flow during supine rest and exercise
tionship between flow rate and tissue weight at term reflects
the effect of the rate of tissue growth within the uterus on the
luminal diameter and reactivity of the uterine artery.
The strong relationship amongst uterine artery luminal
diameter, the rate of uterine blood flow during the third tri-
mester, and tissue weights at term is a third new finding and is
similar to that seen in ovine pregnancy.1416 This may explain
the marked betweensubject variability in the basal rate of
uterine blood flow seen in the current study. It explains the
between-subject variability in uterine blood flow and suggests
that the measurement of either uterine artery diameter or
uterine blood flow under defined conditions may have clinical
value as screening tools for both fetal growth restriction and
overgrowth. It is well known that diseases which chronically
reduce placental perfusion or placental size are associated
with growth restriction, whereas overgrowth is associated
with large placentas and, in the current study, high rates of
uterine blood flow.15,29 Clearly, more work needs to be done
both to confirm this supposition and to develop normative
standards.
The uterine blood flow values initially obtained at left-
lateral rest in the current study are higher than those reported
by both Palmer et al.21 and Thaler et al.30 In all probability, the
differences are due to a combination of factors including
differences in positioning and time, the equipment used,
the analytic algorithm used, and the subject populace. Palmer
et al.21 collected data over 2 hours in the supine position with
a hip wedge in place which may have been inadequate to
maintain uterine volume flow at the same level as that present
during left-lateral rest over that time interval.24,28 Crawford
et al.28 noted that cardiovascular responses of both mother
and fetus deteriorated over time despite the use of a hip wedge
and Kinsella et al.24 report that lateral tilt using a Crawford
wedge did not alter the cardiovascular effects seen with aor-
tocaval compression at supine rest and the same should be
true for intervillous space blood flow.2 Thaler et al.30 obtained
their measurements transvaginally in a shorter time period
but their subjects were in the supine position and should also
have had aortocaval compression.
Our supine measurements were comparable with those
reported at a similar gestational age by both Palmer et al.21
and Thaler et al.30 (267 versus 275 cc/minute), and the mag-
nitude of the decrease in flow when subjects changed from
a left-lateral rest to supine rest observed in the current study
(34%) was identical to the decrease in intervillous space clear-
ance observed by Kauppila et al.2 (34%). Both suggest that the
differences between our values and theirs were primarily due
to changes with position and time.
The equipment and software used by Palmer et al.21 were
specifically designed to estimate uterine volume flow and
were clearly more reproducible than ours. The fact that the
vessel diameters were 2025% smaller in both studies prob-
ably reflects, in part, a consistent overestimation when the
1245
 Jeffreys et al.
measurement is made using colour flow Doppler but some of
the difference may have been due to vasoconstriction in the
artery as our luminal diameters decreased by an average of 0.7
mm at supine rest.21,30 The consistently higher flow velocities
reported by Palmer et al.21 reflect their decision to analyse
only the largest waveform series obtained over the period of
study. In the current study, rapid, brief, three to six beat series
were obtained in duplicate without specific selection after
a short, prescribed time interval.
Finally, the subjects of Palmer et al.21 were studied at an
altitude of 1600 m while ours were studied at sea level. This
may have decreased both their blood volume and uterine
blood flow as both are reduced at high altitude.31
The current study has several limitations. First, the dura-
tion and intensity of the supine exercise was designed to
mimic the routines commonly used in local health clubs
and it is entirely possible that a longer exercise session at
a lower intensity could produce a different result. Second,
we were unable to obtain measurements during the exercise
due to motion artefact and the values obtained when the
exercise was stopped may have been slightly higher than those
present during the exercise.10 Third, the coefficients of varia-
tion for the measurement of vessel diameters and flow veloc-
ity in the current study were large.
In summary, these data do not support the original
hypothesis. Uterine blood flow falls at supine rest primarily
because of uterine vasoconstriction in response to a decrease
in both venous return and cardiac output. With the onset of
supine exercise venous return, cardiac output, and blood
pressure rise which increases uterine blood flow to a level
significantly above that observed at supine rest in healthy,
fit, late-pregnant women. The luminal diameter of the uterine
artery, the basal rate of uterine blood flow, and the magnitude
of its increase during supine exercise correlate with the tissue
weights obtained at the time of delivery.
Acknowledgements
This study was supported by National Institutes of Health
grants HD21268, RR00080, and RR00109 and by Metro-
Health Medical Center. j
References
1 ACOG Committee on Obstetric Practice. Exercise during pregnancy
and the postpartum period. Obstet Gynecol 2002;99:1713.
2 Kauppila A, Koskinen M, Puolakka J, Tuimala R, Kuikka J. Decreased
intervillous and unchanged myometrial blood flow in supine recum-
bency. Obstet Gynecol 1980;55:2035.
3 Howard BK, Goodson JM, Mengert WF. Supine hypotensive syndrome
in late pregnancy. Obstet Gynecol 1953;1:3718.
4 Lees MM, Taylor SH, Scott DB, Kerr MG. The circulatory effects of
recumbent postural change in late pregnancy. Clin Sci 1967;32:45365.
1246
5 Clark SL, Cotton DB, Pivarnik JM, Lee W, Hankins GD, Benedetti TJ,
et al. Position change and central hemodynamic profile during normal
third-trimester pregnancy and postpartum. Am J Obstet Gynecol
1991;164:8837.
6 Ikeda T, Ohbuchi H, Ikenoue T, Mori N. Maternal cerebral hemodynamics
in the supine hypotensive syndrome. Obstet Gynecol 1992;79:2731.
7 Morris N, Osborn SB, Wright HP, Hart A. Effective uterine blood flow
during exercise in normal and pre-eclamptic pregnancies. Lancet
1956;2:4814.
8 Clapp JF, Stepanchak W, Tomaselli J, Kortan M, Fanselow S. Portal vein
blood floweffects of pregnancy, gravity and exercise. Am J Obstet
Gynecol 2000;183:16772.
9 Rauramo I, Forss M. Effect of exercise on maternal hemodynamics and
placental blood flow in healthy women. Acta Obstet Gynecol Scand
1988;67:215.
10 Lotgering FK, Gilbert RD, Longo LD. Exercise in pregnancy in the exper-
imental animal. In: Mittlemark RA, Wiswell RA, Drinkwater BL, editors.
Exercise in Pregnancy. Baltimore, MD: Williams and Wilkins; 1991.
p. 15774.
11 Pivarnik JM, Mauer MB, Ayres NA, Kirshon B, Dildy GA, Cotton DB.
Effect of chronic exercise on blood volume expansion and hematologic
indices during pregnancy. Obstet Gynecol 1994;83:2659.
12 Saltin B, Blomqvist B, Mitchell JH, Johnson RL, Wildenthal K, Chapman
CB. Response to exercise after bed rest and training. Circulation
1968;38(5 Suppl):VII178.
13 OHagan JP, Alberts JA. Uterine artery blood flow and renal sympa-
thetic nerve activity during exercise in rabbit pregnancy. Am J Physiol
Regul Intergr Comp Physiol 2003;285:R113544.
14 Clapp JF, McLaughlin MK, Larrow RW, Farnham J, Mann LI. The uterine
hemodynamic response to repetitive unilateral vascular embolization in
the pregnant ewe. Am J Obstet Gynecol 1982;144:30918.
15 Lang U, Baker RS, Braems G, Zygmunt M, Kunzel W, Clark KE. Uterine
blood flowa determinant of fetal growth. Eur J Obstet Gynecol
Reprod Biol 2003;110:S5561.
16 Lang U, Baker RS, Khoury J, Clark KE. Effect of chronic reduction in
uterine blood flow on fetal and placental growth in the sheep. Am J
Physiol Regul Integr Comp Physiol 2000;279:R539.
17 Clavero-Nunez JA, Negueruela J, Ortiz L, De los Heros AJ, Modrego PS.
Blood flow in the intervillous space and fetal blood flow. I. Normal
values in human pregnancies at term. Am J Obstet Gynecol
1973;116:3406.
18 Borg GAV. Borgs Perceived Exertion and Pain Scales. Champaign, IL:
Human Kinetics, 1998.
19 Clapp JF, Kim H, Burciu B, Schmidt S, Petry K, Lopez B. Continuing
regular exercise during pregnancy: effect of exercise volume on feto-
placental growth. Am J Obstet Gynecol 2002;189:1427.
20 Amini SB, Catalano PM, Hirsh V, Mann LI. An analysis of birth weight by
gestational age using a computerized perinatal data base, 1975-1992.
Obstet Gynecol 1994;83:34252.
21 Palmer SK, Zamudio S, Coffin C, Parker S, Stamm E, Moore L.
Quantitative estimation of uterine artery blood flow and pelvic
blood flow redistribution in pregnancy. Obstet Gynecol 1992;80:
10006.
22 Clapp JF. Exercising Through your Pregnancy. Omaha, NE: Addicus
Books, 2002.
23 Kinsella SM, Lohmann G. Supine hypotensive syndrome. Obstet Gyne-
col 1994;83:77488.
24 Kinsella SM, Lee A, Spencer JAD. Maternal and fetal effects of supine
and pelvic tilt positions in late pregnancy. Eur J Obstet Gynecol Reprod
Biol 1990;36:1117.
25 Chesley LC, Talledo OE, Bohler CS, Zuspan FP. Vascular reactivity to
angiotensin II and norepinephrine in pregnant and nonpregnant
women. Am J Obstet Gynecol 1965;91:83742.
RCOG 2006 BJOG An International Journal of Obstetrics and Gynaecology

26 Gant NF, Daley GL, Chand S, Whalley PJ, MacDonald PC. A study of
angiotensin II pressor response throughout primigravid pregnancy.
J Clin Invest 1973;52:26829.
27 Greiss FC. Uterine pressure-flow relationships. In: Moawad AF,
Lindheimer MD, editors. Uterine and Placental Blood Flow. New York:
Masson Publishing; 1982. p. 6772.
28 Crawford JS, Burton M, Davies P. Time and lateral tilt at Caesarian
section. Br J Anaesth 1972;44:47784.
RCOG 2006 BJOG An International Journal of Obstetrics and Gynaecology
Uterine flow during supine rest and exercise
29 Clapp JF, Risk KH, Appleby-Wineberg SK, Crass JR. Mid-trimester placental
volumes predict birth weight at term. J Soc Gynecol Investig 1995;2:1922.
30 Thaler I, Manor D, Itskovitz J, Rottem S, Levit N, Timor-Tritsch I, et al.
Changes in uterine blood flow during human pregnancy. Am J Obstet
Gynecol 1990;162:1215.
31 Zamudio S, Palmer SK, Droma T, Staum E, Coffin C, Moore LG. Effect
of altitude on uterine artery blood flow during normal pregnancy.
J Appl Physiol 1995;79:714.
1247