Effects of human pregnancy on fluid regulation responses to short-term exercise

J Appl Physiol 95: 2321–2327, 2003.
First published September 5, 2003; 10.1152/japplphysiol.00984.2002.
Effects of human pregnancy on fluid regulation
responses to short-term exercise
Aaron P. Heenan,1 Larry A. Wolfe,1,2 Gregory A. L. Davies,1,3 and Michael J. McGrath3
1
School of Physical and Health Education, and Departments of 2Physiology, and 3Obstetrics
and Gynaecology, Queen’s University, Kingston, Ontario, Canada K7L 3N6
Submitted 24 October 2002; accepted in final form 8 August 2003
Heenan, Aaron P., Larry A. Wolfe, Gregory A. L.
Davies, and Michael J. McGrath. Effects of human pregnancy on fluid regulation responses to short-term exercise.
J Appl Physiol 95: 2321–2327, 2003. First published
September 5, 2003; 10.1152/japplphysiol.00984.2002.—
This study tested the hypothesis that human pregnancy
alters fluid and electrolyte regulation responses to acute
short-term exercise. Responses of 22 healthy pregnant
women (PG; gestational age, 37.0 ⫾ 0.2 wk) and 17 nonpregnant controls (CG) were compared at rest and during
cycling at 70 and 110% of the ventilatory threshold (VT).
At rest, ANG II concentration was significantly (P ⬍ 0.05)
higher in PG vs. CG, whereas plasma osmolality and
concentrations of AVP, sodium, and potasium were significantly lower. Atrial natriuretic peptide concentration at
rest was similar between groups. ANG II and AVP concentrations increased significantly from rest to 110% VT in CG
only, whereas increases in atrial natriuretic peptide concentration were similar between groups. Increases in osmolality, and total protein and albumin concentrations
from rest to both work rates were similar between the two
groups. PG and CG exhibited similar shifts in fluid during
acute short-term exercise, but the increases in ANG II and
AVP were absent and attenuated, respectively, during
pregnancy. This was attributed to the significantly augmented fluid volume state already present at rest in late
gestation.
angiotensin II; arginine vasopressin; atrial natriuretic peptide; progesterone; osmolality
BOTH PREGNANCY AND STRENUOUS EXERCISE have striking
effects on body fluid homeostasis, but, to our knowledge, their interactive effects have not been reported in
any published study. Therefore, this study was conducted to examine the effects of human pregnancy on
plasma electrolytes, osmolality, and circulating fluid
balance hormone responses to short-term exercise
above and below the ventilatory anaerobic threshold.
In early pregnancy, a reduction in systemic vascular
tone attributed to gestational hormones activates “volume-restoring mechanisms,” an increase in blood volume, and a reduction in plasma osmolality (9, 23). The
reduction in mean arterial pressure caused by the
decrease in vascular tone activates the renin-ANGaldosterone (Aldo) system (RAAS) (9, 21, 32), causing
Address for reprint requests and other correspondence: L. A. Wolfe,
School of Physical and Health Education, Queen’s Univ., Kingston, Ontario, Canada K7L 3N6 (E-mail: [email protected]).
http://www.jap.org
sodium retention. Plasma renin concentration and
plasma renin activity (PRA) remain elevated throughout pregnancy (9, 32). The elevated circulating ANG II
concentration ([ANG II]) also stimulates AVP release,
leading to water retention (5). The osmotic threshold
for AVP release is also lowered in pregnancy along
with the threshold for thirst (17).
The most important purpose of atrial natriuretic
peptide (ANP) is to protect against excessive increases
in central blood volume, and ANP is secreted in response to increased venous return and the resulting
atrial distention. In human pregnancy, circulating
plasma ANP concentrations ([ANP]) have been variously reported to decrease (29), increase (25), or remain
unchanged (9) during pregnancy. Although blood volume and venous return are increased substantially by
late gestation (20, 22, 23), this is also accommodated by
the cardiac dilation mediated by gestational hormones
that occurs in early pregnancy (20, 33). This may
prevent an increase in resting left ventricular enddiastolic pressure (20) and presumably atrial stretch so
that the stimulus to increase ANP release is not significantly affected.
In response to exercise, plasma osmolality, PRA,
AVP concentration ([AVP]), [ANP], and Aldo concentration ([Aldo]) increase in healthy men (6, 7, 10, 11)
and nonpregnant women (8, 12, 15, 28). Stimulation of
the RAAS and AVP release is intensity-dependent,
requiring an exercise intensity of ⱖ50–60% of maximal
oxygen uptake (V̇O2 max) (7, 11), although others have
reported a significant rise in PRA (6) and plasma [Aldo]
(11) at 40% V̇O2 max. Plasma [ANP] increases significantly at exercise intensities as low as 25% V̇O2 max
(11).
Activation of the RAAS during exercise is attributed
to sympathetic stimulation of juxtaglomerular cells
and reduced renal blood flow (6). An increase in plasma
osmolality, due to the shifting of water out of the
vasculature, stimulates AVP release from the neurohypophysis and promotes water retention. Increased
circulating [ANG II] also stimulates AVP secretion
via type 1 (AT1) angiotensin receptors located outside the blood-brain barrier (5). Stimulation of these
fluid retention mechanisms during exercise helps to
The costs of publication of this article were defrayed in part by the
payment of page charges. The article must therefore be hereby
marked ‘‘advertisement’’ in accordance with 18 U.S.C. Section 1734
solely to indicate this fact.
8750-7587/03 $5.00 Copyright © 2003 the American Physiological Society
2321
2322
FLUID REGULATION DURING EXERCISE IN PREGNANCY
preserve plasma volume to maintain cardiovascular
function and thermoregulation. Also during dynamic
exercise, enhanced venous return and atrial stretch
result in augmented ANP secretion. However, because
strenuous exercise is associated with antidiuresis and
antinatriuresis, there is a dissociation between
changes in plasma ANP levels and the normal effect of
ANP to excrete sodium and promote fluid loss (11).
Acute expansion of plasma volume in men results in
a blunted PRA (24), Aldo (13, 24), and AVP response to
exercise (24). The greater the plasma volume expansion, the less pronounced the Aldo response (13).
Plasma volume expansion also has been observed to
result in lower plasma [AVP] (13) and higher [ANP]
(13, 24) during exercise as a result of changes in resting
levels of these hormones (13, 24).
This study compared the fluid regulation responses
of pregnant and nonpregnant women to short-term (7
min) exercise at intensities corresponding to 70 and
110% of the ventilatory threshold (VT). Based on the
findings of Grant et al. (13) and Roy et al. (24) on the
effects of acute plasma volume expansion in men, we
hypothesized that the plasma volume expansion of
pregnancy would attenuate the normal response for
ANG II and AVP to exercise.
METHODS
Subjects. Subjects were 22 healthy, nonsmoking, physically active pregnant women [pregnant group (PG); aged
30.0 ⫾ 0.8 yr, mean gestational age 37.0 ⫾ 0.2 wk] and 17
healthy nonpregnant women [control group (CG); aged
27.1 ⫾ 1.4 yr] with similar physical and demographic characteristics. Prospective subjects were recruited from local
prenatal fitness classes and from the general population via
media advertisements, posters, flyers, and contact with local
obstetricians. Medical clearance for pregnant subjects was
obtained from the physician or midwife monitoring their
pregnancy with the Physical Activity Readiness Medical Examination for Pregnancy. Nonpregnant subjects completed
the Revised Physical Activity Readiness Questionnaire. Both
forms are available online from the Canadian Society for
Exercise Physiology (www.csep.ca/forms.asp).
Written, informed consent was obtained from all subjects
before entry into the study. The study protocol and consent
form were approved by the Research Ethics Board, Faculty of
Medicine, Queen’s University and the US Army Medical
Research and Materiel Command, Human Subjects Protection Branch.
Basic physical measurements included body height, body
mass, and resting blood pressure. Body mass index was
calculated as body mass/body height2 (kg/m2). PG and CG
values were equated for mean age, body height, prepregnant
body mass (as reported by subject), parity, and aerobic fitness. Members of PG performed the two exercise tests for the
study between 34 and 38 wk of gestation. Subjects in CG
were not using oral contraceptives.
Exercise testing protocols. Subjects performed two exercise
tests on a Sensor Medics (model 800S; Sensor Medics, Yorba
Linda, CA) constant work rate cycle ergometer on separate
days, at least 1 day apart. To help standardize nutritional
status and to ensure normal hydration, subjects consumed a
standard meal (350 kcal, 40% carbohydrate, 40% fat, 20%
protein) 1–2 h before both tests and avoided strenuous physical activity and caffeine on the day of testing. The first test
J Appl Physiol • VOL
was used to determine VT and to assess aerobic working
capacity (14). The protocol involved 5 min of resting data
collection and a 4-min warm-up at 20 W, followed by an
increase in work rate of 20 W/min until a heart rate (HR) of
170 beats/min was reached (14). Respiratory responses and
oxygen uptake (V̇O2) were measured on a breath-by-breath
basis by using a computerized system (First Breath, St.
Agatha, Ontario, Canada) that incorporates a respiratory
mass spectrometer (MGA 1100; Marquette Electronics, Milwaukee, WI) with a volume turbine (VMM-2; Interface Associates, Aliso Viejo, CA) as described previously (14). The
V-slope method (2) was used to identify VT. Heart rate (HR)
was monitored with both a Polar Vantage monitor (Polar
Electro, Woodbury, NY) and Marquette Max-1 electrocardiograph (Marquette Electronics). Oxygen pulse (V̇O2/HR) at a
HR of 170 beats/min was calculated as an index of aerobic
working capacity (14).
The second exercise test involved a 10-min resting data
collection period, followed by a 3-min warm-up at 0 W and a
ramp increase in work rate, over a 30-s time period, to a work
rate corresponding to 70 or 110% VT (14). Both work rates
were continued for 7 min after achievement of the prescribed
work rate. Subjects rested for 20 min between levels. Before
this test, an indwelling catheter was inserted into a dorsal
hand vein situated as far from the thumb as possible. The
hand and lower arm were soaked in a warm water bath
before insertion of the catheter and then placed in a Plexiglas
heating box (45°C) to promote vasodilation. Arterialized
blood samples were then collected at rest and during the
minute 6 of exercise at 70 or 110% VT.
Blood samples for the determination of oxygen tension,
arterial PCO2 and H⫹ concentration were collected in a syringe containing lyophilized heparin and analyzed immediately by using a Radiometer ABL 30 acid-base analyzer at a
standard temperature of 37°C. Tympanic temperature measurements confirmed no significant deviation from 37°C with
pregnancy or exercise with the use of the present protocol
(14). Quality control checks using four control liquids were
done on all testing days. The remaining blood was used for
plasma electrolyte, total protein (TP), and albumin analysis.
Blood for plasma osmolality determination was collected in a
4.5-ml syringe containing lithium heparin (Monovette,
Sarstedt). Blood samples for plasma analysis of [ANG II],
[AVP], and [ANP] were collected in 10-ml chilled vacutainers
containing EDTA (Becton-Dickinson) to which aprotinin (100
␮l of 100,000 KIU/ml solution) was immediately added to the
whole blood and pepstatin A (20 ␮l of 1 mg/ml solution) and
O-phenanthroline (20 ␮l of 50 mM solution) added to 2 ml of
fresh plasma for AII determination. Blood for serum determination of progesterone concentration was collected in a
vacutainer containing no additive and placed in ice for 0.5–1
h to allow time to clot. All blood collected for plasma and
serum analysis was centrifuged for 10 min at 2,500 rpm (532
g), and the plasma or serum was frozen at ⫺80°C for later
analysis, as described below.
Plasma concentrations of sodium ([Na⫹]), potassium ([K⫹])
and chloride ([Cl⫺]) were analyzed by using ion-selective
electrodes. Total calcium concentration was measured using
a dye-binding assay. Ionized calcium concentration ([Ca2⫹])
was calculated from total calcium concentration. Plasma
osmolality was determined by using the freezing-point depression technique (Osmette A, model 5002, Precision Systems, Natick, MA). TP concentration ([TP]) was measured by
using the direct Biuret method. Albumin concentration
([Alb]) was determined by using a conventional dye-binding
method.
95 • DECEMBER 2003 •
www.jap.org
2323
FLUID REGULATION DURING EXERCISE IN PREGNANCY
ANG II was measured with a radioimmunoassay kit using
I-labeled ANG II (Nichols Institute Diagnostics, San Juan
Capistrano, CA) (30). The extraction recovery was 70%. AVP
and ANP were measured by radioimmunoassay developed by
Sarda et al. (26, 30). Intra- and interassay coefficients of
variation were 11 and 10% for AVP and 3.8 and 9.1% for
ANP. The assays for AVP and ANP have a sensitivity of
within 0.2 and 15 pg/ml (1.5 pg/tube), respectively. Progesterone was measured by radioimmunoassay by the Core Lab
at Kingston General Hospital.
Statistical analyses. Physical characteristics, V̇O2 at VT,
and oxygen pulse at 170 beats/min were compared between
groups by using Student’s t-statistics for independent samples. Data at rest and during exercise at 70 and 110% VT
were compared within and between subjects by using a twoway ANOVA (groups vs. rest/exercise level) with repeated
measures on the second factor. When a significant betweengroup main effect was observed, a Tukey’s post hoc test was
used to identify significant differences between group means
at rest, at 70% VT, and at 110% VT. When a significant
within-subjects main effect or group ⫻ time interaction was
observed, Tukey’s post hoc tests were also used to detect
significant differences between rest, 70% VT, and 110% VT
within each group. The critical alpha for significance was
P ⬍ 0.05.
125
RESULTS
By design, there were no significant differences between groups in age, body height, or parity (Table 1).
As expected, body mass and body mass index of the PG
was significantly greater than that of the CG at the
time of testing. However, the prepregnancy body mass
and body mass index of the PG were similar to those of
the CG. There were also no significant differences
between the groups in V̇O2 at VT or O2 pulse at 170
beats/min, confirming that the two groups had a similar level of aerobic fitness. Of the 22 subjects in the PG
and the 17 subjects in the CG, not all subjects produced
complete data sets for [ANG II], [AVP], [ANP], [progesterone], and [TP] due to hemolysis of individual blood
samples or individual blood sample acquisition problems. Only subjects with complete data sets (i.e., data
at rest, 70% VT, and 110% VT) were included in a
variables analysis. The sample sizes used in the analysis of each variable are shown in Tables 2–5 and
Figs. 1–3.
Table 2. V̇O2 and heart rate at rest and
at 2 work rates
Variable
Rest
PG
CG
70% VT
PG
CG
110% VT
PG
CG
V̇O2, l/min
Heart rate,
beats/min
0.38 ⫾ 0.01
0.34 ⫾ 0.01
89 ⫾ 2*
70 ⫾ 2
1.28 ⫾ 0.05†
1.33 ⫾ 0.04†
131 ⫾ 2*†
120 ⫾ 3†
1.91 ⫾ 0.08†‡
2.00 ⫾ 0.06†‡
160 ⫾ 3†‡
156 ⫾ 4†‡
Values are means ⫾ SE. PG, pregnant group (n ⫽ 22); CG, control
group (n ⫽ 17). * Significant difference (P ⬍ 0.05) between groups.
† Significant change (P ⬍ 0.05) within group from rest. ‡ Significant
change (P ⬍ 0.05) within group from cycling at 70% VT to cycling at
110% VT.
HR was significantly higher in the PG vs. the CG at
rest, whereas V̇O2 was not (Table 2). There was no
difference between groups in V̇O2 at 70 or 110% VT. HR
was significantly higher in the PG vs. the CG at 70%
VT but not at 110% VT. Both V̇O2 and HR increased in
both groups from rest to 70% VT and 70 to 110% VT.
Osmolality was significantly lower in the PG vs. CG
at rest and at both work rates. Osmolality increased
significantly from rest to 70% VT and from 70 to 110%
VT in both groups (Table 3). There was a significant
interaction between group and exercise level for osmolality. [Na⫹] and [K⫹] were significantly lower in the
PG vs. the CG at rest and both work rates. [Na⫹] and
[K⫹] increased significantly from rest to 70% VT and
from 70 to 110% VT in both groups, except for [Na⫹] in
the CG from 70 to 110% VT. There was no significant
interaction between group and exercise level for [Na⫹]
or [K⫹]. [Ca2⫹] was not significantly different between
groups at rest but was significantly higher in the PG
vs. the CG at 70 and 110% VT. [Ca2⫹] increased from
rest to 70% VT in the PG and from 70 to 110% VT in
both groups. There was a significant interaction be-
Table 1. Physical characteristics of subjects
Variable
Pregnant
Group
(n ⫽ 22)
Control
Group
(n ⫽ 17)
Age, yr
Gestational age, wk
Height, cm
Body mass, kg
Body mass index, kg/m2
Pre-pregnant body mass, kg
Pre-pregnant body mass index, kg/m2
Parity
V̇O2 at VT, l/min
O2 pulse at 170 beats/min, ml/beat
30.0 ⫾ 0.8
37.0 ⫾ 0.2
163 ⫾ 2
75.8 ⫾ 2.0*
28.4 ⫾ 0.6*
62.4 ⫾ 2.0
23.3 ⫾ 0.6
0.5 ⫾ 0.2
1.67 ⫾ 0.06
12.1 ⫾ 0.5
27.1 ⫾ 1.4
n/a
165 ⫾ 1
61.4 ⫾ 1.7
22.5 ⫾ 0.5
n/a
n/a
0.4 ⫾ 0.2
1.80 ⫾ 0.04
13.0 ⫾ 0.4
Values are means ⫾ SE. V̇O2, O2 uptake; VT, ventilatory threshold.
* Significant difference between groups (P ⬍ 0.05).
J Appl Physiol • VOL
Fig. 1. Plasma ANG II concentration ([ANG II]) at rest and at 2 work
rates. VT, ventilatory threshold. *Significant difference (P ⬍ 0.05)
between groups. #Significant change (P ⬍ 0.05) within group from
rest.
95 • DECEMBER 2003 •
www.jap.org
2324
FLUID REGULATION DURING EXERCISE IN PREGNANCY
Table 3. Plasma electrolytes and osmolality at rest and at 2 work rates
Variable
Rest
PG
CG
70% VT
PG
CG
110% VT
PG
CG
[Na⫹], mM
[K⫹], mM
[Ca⫹], mM
[Cl⫺], mM
Osmolality,
mosmol/kgH2O
136 ⫾ 0.3*
140 ⫾ 0.4
3.9 ⫾ 0.1*
4.2 ⫾ 0.1
1.18 ⫾ 0.01
1.16 ⫾ 0.01
105 ⫾ 0.5
106 ⫾ 0.6
277 ⫾ 0.4*
284 ⫾ 0.8
138 ⫾ 0.4*†
141 ⫾ 0.4†
4.5 ⫾ 0.1*†
4.7 ⫾ 0.1†
1.21 ⫾ 0.01*†
1.14 ⫾ 0.01
105 ⫾ 0.6†
106 ⫾ 0.6
280 ⫾ 0.5*†
288 ⫾ 0.7†
139 ⫾ 0.4*†‡
142 ⫾ 0.4†
4.9 ⫾ 0.1*†‡
5.2 ⫾ 0.1†‡
1.23 ⫾ 0.01*†‡
1.17 ⫾ 0.01‡
105 ⫾ 0.6†
107 ⫾ 0.6†
284 ⫾ 0.7*†‡
292 ⫾ 1.0†‡
Values are means ⫾ SE; n ⫽ 22 for PG and n ⫽ 17 for CG. * Significant difference (P ⬍ 0.05) between groups. † Significant change (P ⬍
0.05) within group from rest. ‡ Significant change (P ⬍ 0.05) within group from cycling at 70% VT to cycling at 110% VT.
tween group and exercise level for [Ca2⫹]. [Cl⫺] increased from rest to 70% VT in the PG and rest to 110%
VT in both groups. There was no significant interaction
between group and exercise level for [Cl⫺].
Both TP and albumin levels were significantly lower
in the PG compared with the CG at all measurement
times (Table 4). Both TP and albumin levels increased
significantly from rest to 70% VT and from 70 to 110%
VT in both groups. There was no significant interaction
between group and exercise level for TP or albumin
levels.
[ANG II] was significantly higher in the PG vs. CG at
rest but not at either work rate. In the transition from
rest to 110% VT, [ANG II] increased significantly in the
CG but not in the PG (Fig. 1). There was a significant
group ⫻ time interaction for [ANG II]. Plasma [AVP]
was significantly lower in the PG vs. CG at rest and
both work rates. Arginine vasopressin concentration
increased significantly from rest and 70% VT to 110%
VT in the CG (Fig. 2). There was no significant interaction between group and exercise level for [AVP].
ANP was not significantly different at rest or either
work rate between groups. [ANP] increased significantly from rest to 70% VT and from 70 to 110% VT in
the PG and was significantly higher at 110% VT than
at rest in both groups (Fig. 3). There was no significant
interaction between group and exercise level for [ANP].
As expected, values for circulating plasma progesterone were substantially higher in the PG vs. CG under
all experimental conditions (Table 5). Values within
the CG reflected the fact that subjects were tested in
varying phases of the menstrual cycle. Progesterone
levels increased significantly in the PG from rest to
110% VT and from 70 to 110% VT. There was a significant interaction between group and exercise level for
progesterone.
DISCUSSION
This study tested the hypothesis that pregnancy
attenuates ANG II and AVP responses to short-term
dynamic exercise. As postulated, the main finding was
that the normal response for the activation of the
renin-angiotensin system and the secretion of AVP in
response to exercise was blunted in late gestation. In
contrast, ANP responses were not attenuated.
Although changes in blood volume were not measured directly in this study, the use of plasma protein
measurements to reflect fluid volume shift in response
to exercise has been used successfully in previous studies of pregnant (3) and nonpregnant women (4). This
practice is based on the assumption that macromolecules such as albumin and other plasma proteins do
Table 4. Plasma total protein and albumin at rest
and at 2 work rates
Variable
Rest
PG
CG
70% VT
PG
CG
110% VT
PG
CG
Total Protein, g/l
Albumin, g/l
61.2 ⫾ 0.9*
69.3 ⫾ 1.3
29.5 ⫾ 0.4*
42.0 ⫾ 0.7
64.5 ⫾ 1.0*†
71.5 ⫾ 1.0†
31.0 ⫾ 0.5*†
43.6 ⫾ 0.6†
67.7 ⫾ 1.3*†‡
74.7 ⫾ 1.3†‡
32.5 ⫾ 0.5*†‡
45.2 ⫾ 0.6†‡
Values are means ⫾ SE. For total protein, n ⫽ 21 for PG and n ⫽
15 for CG. For albumin, n ⫽ 22 for PG and n ⫽ 17 for CG.
* Significant difference (P ⬍ 0.05) between groups. † Significant
change (P ⬍ 0.05) within group from rest. ‡ Significant change (P ⬍
0.05) within group from cycling at 70% VT to cycling at 110% VT.
J Appl Physiol • VOL
Fig. 2. Plasma AVP concentration ([AVP]) at rest and at 2 work
rates. *Significant difference (P ⬍ 0.05) between groups. #Significant
change (P ⬍ 0.05) within group from rest. &Significant change (P ⬍
0.05) within group from cycling at 70% VT to cycling at 110% VT.
95 • DECEMBER 2003 •
www.jap.org
2325
FLUID REGULATION DURING EXERCISE IN PREGNANCY
Fig. 3. Plasma atrial natriuretic peptide concentration ([ANP]) at
rest and at 2 work rates. #Significant change (P ⬍ 0.05) within group
from rest. &Significant change (P ⬍ 0.05) within group from cycling
at 70% VT to cycling at 110% VT.
not cross vascular barriers in important amounts in
response to hydrostatic and osmotic forces during exercise. In this regard, Pivarnik et al. (23) reported
moderate increases in exercise-induced loss of total
plasma protein and albumin at 29–33 wk of gestation;
values in late gestation (33–36 wk) were similar to
postpartum control. These losses were small under all
experimental conditions.
Resting measurements from this study reflected the
substantial expansion of plasma volume and hemodilution that occurs in normal human pregnancy. In this
regard, values for TP and albumin were significantly
lower in the PG vs. nonpregnant subjects in the resting
state. In response to exercise, hemoconcentration was
observed in both groups at both work rates as TP and
albumin concentrations and osmolality increased significantly from rest. The two groups were similar in the
extent of the hemoconcentration as indicated by similar changes in TP and albumin concentration. Both TP
and albumin concentrations increased in the PG from
rest to 70% VT by ⬃5% and from rest to 110% VT by
⬃10–11%, with the CG increasing by 3–4 and ⬃8%,
respectively. This is consistent with the findings of
Pivarnik et al. (23), who observed similar decreases in
plasma volume in women 33–36 wk pregnant compared with 4 wk postpartum in response to 5 min of
cycling at 75 W.
The significantly higher [ANG II] of the PG vs. CG at
rest is consistent with the findings of other studies,
which have found elevated plasma renin concentration
(9), PRA (32), and [ANG II] (21) in pregnancy. The
lower circulating [AVP] in the PG vs. the CG at rest
may be the result of a large increase in the metabolic
clearance rate of AVP during pregnancy (16). Progesterone may also contribute to the lowering of [AVP] as
progesterone administration has been demonstrated to
decrease AVP levels in women (31). However, elevated
circulating estrogen concentration (not measured in
this study) is the probable stimulus for the decrease in
J Appl Physiol • VOL
the osmotic threshold for AVP release during pregnancy (17) because this hormone decreases the osmotic
threshold for AVP release in nonpregnant women (27).
The level of ANP at rest was not different between the
PG and CG. Atrial enlargement during pregnancy (33)
may reduce atrial stretch from the expanded blood
volume and contribute to the maintenance of normal
ANP levels.
As hypothesized, the ANG II response to exercise
was attenuated in the PG but not in the CG as reflected
by a significant group ⫻ time interaction. Although the
rise in [ANG II] in the CG group in response to exercise
is consistent with the literature on the response of the
RAAS to dynamic exercise in the CG (6, 7, 11) values in
the PG did not rise at either intensity. This may indicate that blood flow to the kidney was not reduced in
pregnancy during the exercise, and thus renin production was not stimulated. If renin concentration did not
change, then ANG I levels and subsequently [ANG II]
would not rise. The large increase in blood volume
during pregnancy (18, 23) may have resulted in improved renal perfusion during exercise. Grant et al.
(13) and Roy et al. (24) also hypothesized that the
blunted plasma [Aldo] response to exercise observed in
men after acute plasma volume expansion was related
to improved renal blood flow during exercise.
Attenuation of the RAAS response to exercise in
pregnancy was similar qualitatively to that reported by
Grant et al. (13) and Roy et al. (24) in response to acute
plasma volume expansion in men. The size of this effect
appears to be related to the extent of hypervolemia.
Since plasma volume is usually increased by 45–55%
in pregnancy (18, 23) our results may be viewed as an
extension of the investigations of Grant et al. (13) and
Roy et al. (24). Grant et al. used plasma volume expansion of 14 and 21% in a study of responses of untrained
men to prolonged exercise (120 min at 46% of maximal
aerobic power). The subsequent study by Roy et al.
examined the effects of a 16% plasma volume expansion on responses of similar subjects to 90 min of
exercise at 62% of maximal aerobic power. Thus the
results from the study of the larger plasma volume
expansion in the pregnant state continued the trend
for a decreasing RAAS response with increasing
plasma volume expansion, established by Grant et al.
and Roy et al., by using smaller percentage increases in
plasma volume.
A reduced sympathetic drive (i.e., lower epinephrine
and norepinephrine levels) was also observed by Grant
Table 5. Serum progesterone at rest
and at 2 work rates
Group
Rest
70% VT
110% VT
PG
CG
716 ⫾ 55*
15 ⫾ 4
756 ⫾ 59*
15 ⫾ 4
828 ⫾ 53*†‡
19 ⫾ 5
Values are means ⫾ SE in nmol/l; n ⫽ 20 for PG and n ⫽ 16 for CG.
* Significant difference (P ⬍ 0.05) between groups. † Significant
change (P ⬍ 0.05) within group from rest. ‡ Significant change (P ⬍
0.05) within group from cycling at 70% VT to cycling at 110% VT.
95 • DECEMBER 2003 •
www.jap.org
2326
FLUID REGULATION DURING EXERCISE IN PREGNANCY
et al. (13) and Roy et al. (24) and was hypothesized to
be a potential contributor to the observed lower [Aldo]
(13). Sympathetic activity reduces renal blood flow and
causes the juxtaglomerular cells of the kidney to increase their renin secretion. This in turn results in
elevated circulating [ANG II] and [Aldo]. In our laboratory, a reduction in sympathetic activity has also
been observed during exercise in similar pregnant subjects in late gestation at work rates corresponding to 60
and 110% VT as indicated by lower epinephrine levels
at 60% VT and lower epinephrine and norepinephrine
levels at 110% VT (1). Thus the reduced [ANG II]
during exercise in the PG may also be the consequence
of a reduction in sympathetic drive. Regardless of the
cause, the lack of rise in [ANG II] means that Aldo
secretion will not be stimulated to increase Na⫹ retention during exercise in pregnancy.
The AVP response to exercise in the PG differed from
that of the CG. Again, the CG responded as expected,
increasing significantly in response to strenuous exercise, but not in response to the less intense exercise (7,
11). Values in the PG, however, did not increase in
response to either work rate. Both groups had similar
increases in osmolality at the 110% VT work rate, and
thus it would be expected that the stimulus for AVP
release would be similar in both groups. However,
because the [ANG II] response of the PG was blunted,
there may have been a lesser stimulus for AVP release.
Roy et al. (24) observed a similar blunting of the AVP
response to exercise after plasma volume expansion,
especially during the first hour of their protocol.
Whether other cardiovascular pressure receptors
and reflexes are involved in the blunted ANG II and
AVP responses to exercise is an open question. Although blood pressure during exercise was not measured during the present study, an earlier study from
this laboratory (1), which used a very similar cycling
test protocol and many of the same subjects, observed
no significant differences in systolic blood pressure
immediately before exercise (sitting) or during short
bouts of exercise (7 min) at 60 and 110% VT, respectively, in late pregnancy vs. the nonpregnant state.
Thus there is no solid basis at this time to suggest that
cardiovascular pressure or volume receptors are involved in these blunted hormone responses. However,
further study is needed to clarify this issue.
Despite minor statistical differences, [ANP] rose in
both groups at both work rates, and responses were
similar from a physiological viewpoint. The increase in
ANP in response to exercise in both groups in the
present study was expected at both the 70 and 110%
VT work rates because relative intensity exceeded the
usual threshold to evoke an ANP response. This is
consistent with the earlier findings of Grant et al. (13)
and Roy et al. (24), who showed that the response in
normal subjects to increase ANP in response to exercise was also present in subjects with acute plasma
volume expansion. The fact that, in the present study,
ANP levels at rest and in response to exercise were
generally similar in the pregnant vs. nonpregnant
state may be that the augmented venous return obJ Appl Physiol • VOL
served at rest and during standard submaximal exercise in pregnancy is offset by atrial enlargement (33) so
that the degree of atrial distention and stimulus for
ANP release is not affected. Thus the acute plasma
volume expansion in the Grant et al. and Roy et al.
studies, which resulted in higher ANP values compared with controls at rest, with the higher levels also
persisting into exercise, may have resulted because the
acute plasma volume expansion (resulting in enhanced
venous return) would not be offset by atrial enlargement.
The fact that the RAAS and AVP were affected by
pregnancy during exercise and ANP responses were
not can be explained by the fact that the stimuli for
activation of the two systems are different (19). In this
regard, the RAAS-AVP axis responds primarily to
changes in blood volume, renal perfusion pressure, and
sympathetic stimulation (with osmotic changes also
being a major stimulus for AVP), whereas the ANP
secretion is stimulated by increased venous return and
atrial distention (19).
In conclusion, our results support our original hypothesis that pregnant subjects exhibit blunted RAAS
and AVP responses to short-term exercise. This effect
may have been the result of reduced sympathetic drive
to the kidney and/or reduced circulating norepinephrine resulting in an attenuation of the normal decrease
in renal blood flow in response to exercise and reduced
renin output of juxtaglomerular cells of the kidney.
Lower ANG II levels may then account for the observed
blunting of AVP responses to exercise. In contrast, the
ANP response to exercise above VT were similar to the
CG. Despite the differences in the ANG II and AVP
responses to exercise, the two groups exhibited similar
changes in osmolality and TP and albumin concentrations, suggesting similar overall water balance adjustments during the exercise bouts.
Because this study examined the effects of human
pregnancy responses to short-term exercise, future
studies should be conducted to determine the effects of
pregnancy on responses to prolonged exercise (⬎40min duration). The study of renal blood flow during
exercise in pregnancy would also be useful to understand the mechanisms responsible for possible discrepancies in hormonal responses.
DISCLOSURES
This study was supported by US Army Medical Research and
Materiel Command Contract no. DAMD17-96-C-6112, Ontario Thoracic Society, Ontario Respiratory Disease Foundation (Block-Term
Grant Funding), and Natural Sciences and Engineering Research
Council of Canada. A. P. Heenan was supported by an Ontario
Graduate Scholarship for Science and Technology and an Ontario
Thoracic Society training award.
REFERENCES
1. Avery ND, Wolfe LA, Amara CE, Davies GAL, and McGrath
MJ. Effects of human pregnancy on cardiac autonomic function
above and below the ventilatory threshold. J Appl Physiol 90:
321–328, 2001.
2. Beaver WL, Wasserman K, and Whipp BJ. A new method for
detecting anaerobic threshold by gas exchange. J Appl Physiol
60: 2020–2027, 1986.
95 • DECEMBER 2003 •
www.jap.org
FLUID REGULATION DURING EXERCISE IN PREGNANCY
3. Bonen A, Campagna P, Gilchrist L, Young DC, and Beresford P. Substrate and endocrine responses during exercise at
selected stage of pregnancy. J Appl Physiol 73: 134–142, 1992.
4. Bonen A, Haynes FJ, Watson-Wright W, Sopper MM,
Pierce GN, Low MP, and Graham TE. Effects of menstrual
cycle on metabolic responses to exercise. J Appl Physiol 55:
1506–1513, 1983.
5. Chiodera P, Volpi R, Caiazza A, Giuliani N, Magotti MG,
and Coiro V. Arginine vasopressin and oxytocin responses to
angiotensin II are mediated by AT1 receptor subtype in normal
men. Metabolism 47: 893–896, 1998.
6. Convertino VA, Keil LC, Bernauer EM, and Greenleaf JE.
Plasma volume, osmolality, vasopressin, and renin activity during graded exercise in man. J Appl Physiol 50: 123–128, 1981.
7. Convertino VA, Keil LC, and Greenleaf JE. Plasma volume,
renin and vasopressin responses to graded exercise after training. J Appl Physiol 54: 508–514, 1983.
8. De Souza MJ, Maresh CM, Maguire MS, Kraemer WJ,
Flora-Ginter G, and Goetz KL. Menstrual status and plasma
vasopressin, renin activity, and aldosterone exercise responses.
J Appl Physiol 67: 736–743, 1989.
9. Duvekot JJ, Cheriex EC, Pieters FAA, Menheere PPCA,
and Peters LLH. Early pregnancy changes in hemodynamics
and volume homeostasis are consecutive adjustments triggered
by a primary fall in systemic vascular tone. Am J Obstet Gynecol
169: 1382–1392, 1993.
10. Freund BJ, Claybaugh JR, Dice MS, and Hashiro GM.
Hormonal and vascular fluid responses to maximal exercise in
trained and untrained males. J Appl Physiol 63: 669–675, 1987.
11. Freund BJ, Shizuru EM, Hashiro GM, and Claybaugh JR.
Hormonal, electrolyte, and renal responses to exercise are intensity dependent. J Appl Physiol 70: 900–906, 1991.
12. Galliven EA, Singh A, Michelson D, Bina S, Gold PW, and
Deuster PA. Hormonal and metabolic responses to exercise
across time of day and menstrual cycle phase. J Appl Physiol 83:
1822–1831, 1997.
13. Grant SM, Green HJ, Phillips SM, Enns DL, and Sutton
JR. Fluid and electrolyte hormonal responses to exercise and
acute plasma volume expansion. J Appl Physiol 81: 2386–2392,
1996.
14. Heenan AP and Wolfe LA. Plasma acid-base regulation above
and below ventilatory threshold in late gestation. J Appl Physiol
88: 149–157, 2000.
15. Kinugawa T, Ogino K, Miyakoda H, Saitoh M, Hisatome I,
Fujimoto Y, Yoshida A, Shigemasa C, and Sato R. Responses of catecholamines, renin-angiotensin system, and atrial
natriuretic peptide to exercise in untrained men and women.
Gen Pharmacol 28: 225–228, 1997.
16. Lindheimer MD, Barron WM, and Davison JM. Osmotic and
volume control of vasopressin release in pregnancy. Am J Kidney
Dis 17: 105–111, 1991.
17. Lindheimer MD and Davison JM. Osmoregulation, the secretion of arginine vasopressin and its metabolism during pregnancy. Eur J Endocrinol 132: 133–143, 1995.
18. Longo LD. Maternal blood volume and cardiac output during
pregnancy: a hypothesis of endocrinologic control. Am J Physiol
Regul Integr Comp Physiol 245: R720–R729, 1983.
J Appl Physiol • VOL
2327
19. Mannix ET, Palange P, Aronoff GR, Manfredi F, and Farber MO. Atrial natriuretic peptide and the renin-aldosterone
axis during exercise in man. Med Sci Sports Exerc 22: 785–789,
1990.
20. Morton MJ, Paul MS, and Metcalfe J. Exercise during pregnancy. Med Clin North Am 69: 97–108, 1985.
21. Pedersen EB, Johannesen P, Rasmussen AB, and
Danielsen H. The osmoregulatory system and the renin-angiotensin-aldosterone system in pre-eclampsia and normotensive
pregnancy. Scand J Lab Invest 45: 627–633, 1985.
22. Pivarnik JM, Lee W, Clark SL, Cotton DB, Spillman HT,
and Miller JF. Cardiac output responses of primigravid women
during exercise determined by the direct Fick technique. Obstet
Gynecol 75: 954–959, 1990.
23. Pivarnik JM, Lee W, Miller JF, and Werch J. Alterations in
plasma volume and protein during cycle exercise throughout
pregnancy. Med Sci Sports Exerc 22: 751–755, 1990.
24. Roy BD, Green HJ, Grant SM, and Tarnopolsky MA. Acute
plasma volume expansion in the untrained alters the hormonal
response to prolonged moderate-intensity exercise. Horm Metab
Res 33: 238–245, 2001.
25. Sala C, Campise M, Ambroso G, Motta T, Zanchetti A, and
Morganti A. Atrial natriuretic peptide and hemodynamic
changes during normal human pregnancy. Hypertension 25:
631–636, 1995.
26. Sarda RI, De Bold ML, and De Bold AJ. Optimization of
atrial natriuretic factor radioimmunoassay. Clin Biochem 22:
11–15, 1989.
27. Stachenfeld NS, Silva C, Keefe DL, Kokoszka CA, and
Nadel ER. Effects of oral contraceptives on body fluid regulation. J Appl Physiol 87: 1016–1025, 1999.
28. Stephenson LA, Kolka MA, Francesconi R, and Gonzalez
RR. Circadian variations in plasma renin activity, catecholamines and aldosterone during exercise in women. Eur
J Appl Physiol 58: 756–764, 1989.
29. Thomsen JK, Fogh-Anderson N, Jaszczak P, and Giese J.
Atrial natriuretic peptide (ANP) decrease during normal pregnancy as related to hemodynamic changes and volume regulation. Acta Obstet Gynecol Scand 72: 103–110, 1993.
30. Walker JL and Jennings DB. Angiotensin mediates stimulation of ventilation after vasopressin V1 receptor blockade. J Appl
Physiol 76: 2517–2526, 1994.
31. Watanabe H, Lau DCW, Guyn HL, and Wong NLM. Effect of
progesterone therapy on arginine vasopressin and atrial natriuretic factor in premenstrual syndrome. Clin Invest Med 20:
211–223, 1997.
32. Wilson M, Morganti AA, Zervoudakis I, Letcher RL, Romney BM, Von Oeyon P, Papera S, Sealey JE, and Laragh
JH. Blood pressure, the renin-aldosterone system and sex steroids throughout normal pregnancy. Am J Med 68: 97–104,
1980.
33. Wolfe LA, Preston RJ, Burggraf GW, and McGrath MJ.
Effects of pregnancy and chronic exercise on maternal cardiac
structure and function. Can J Physiol Pharmacol 77: 909–917,
1999.
95 • DECEMBER 2003 •
www.jap.org