Clinical Science (1992) 82, 157-162 (Printed in Great Britain)
I57
Comparative study of pressor and heart rate responses to angiotensin II and noradrenaline in pregnant and non-pregnant women Margaret RAMSAY", Fiona BROUGHTON PlPKlNt and Peter RUBIN* Departments of * Therapeutics and fobstetria and Gynaecology, University Hospital, Nottingham, U.K. (Received I I July19 September
I99 I; accepted I 9 September I99 I )
1. Twenty-eight healthy non-pregnant women and 28 women in the first or second trimester of pregnancy were studied. They were given an incremental intravenous infusion of either noradrenaline or angiotensin 11. Pressor and heart rate responses were documented. 2. Dose-pressor response curves were constructed for the two agents in pregnant and non-pregnant women ( n = 14 in each group). The regression parameters of slope and intercept were calculated, and were used to derive the variables of dose required to elicit a 10 mmHg rise in systolic or diastolic blood pressure. 3. The pressor response to angiotensin I1 was diminished in pregnancy, with approximately twice the dose being required to raise the systolic or diastolic arterial blood pressure as in non-pregnant subjects. 4. The systolic pressor response to noradrenaline was slightly diminished in pregnancy, but the diastolic pressor response was unchanged. There were no significant differences between the doses of noradrenaline required to elicit a 10 mmHg rise in systolic or diastolic arterial blood pressure in pregnant or non-pregnant subjects. 5. There was a diminution in the bradycardia evoked in response to both hormones in pregnancy. 6. We conclude that the well-documented pressor insensitivity to angiotensin I1 during pregnancy is a specific phenomenon, not a manifestation of a generalized reduction in vascular reactivity.
INTRODUCTION Insensitivity to the pressor effects of angiotensin I1 (ANG 11)in pregnancy has been well documented following the original description by Abdul-Karim & Assali in 1961 [l].Gant et al. showed that this resistance to ANG I1 has developed by 7-8 weeks of gestation, and reaches its maximum at 28 weeks [ 2 ] .On average, 2.6 times as much ANG I1 is required to elicit a given rise in mean blood pressure in a pregnant as in a non-pregnant woman [31.
However, what is not clear from the literature published to date is whether this insensitivity to ANG I1 is a manifestation of a generalized depression in vascular reactivity in pregnancy or whether it is a phenomenon specific to ANG 11. The pressor responses to noradrenaline (NA) and adrenaline have been studied in pregnancy, but bolus doses or very short-duration infusions of the agents have been used (i.e. steady-state responses have not been considered). The studies have had discrepant results: Morandini & Mangioni [4] found the average dose of NA necessary to raise the diastolic pressure by 20 mmHg to be more than doubled in third-trimester women as compared with non-pregnant control women or women 5 days postpartum. Raab et al. [ 5 ]found no difference in the pressor responses to adrenaline or NA between nonpregnant and third-trimester women. Chesley et al. [3] found no difference in the pressor response to a bolus dose of NA in late pregnancy, but they did find a longer duration of response in the pregnant subjects. Lumbers [6], examining peripheral vascular reactivity with venous occlusion plethysmography, found no differences in response to NA between pregnant and non-pregnant subjects, although those to ANG I1 were blunted in pregnancy. The bradycardia evoked by the infusion of pressor agents was not reported in any of these studies. The purpose of this study was to determine the pressor responses to steady-state infusions of NA and ANG I1 in comparable groups of non-pregnant women, and firstand second-trimester pregnant women, by constructing dose-pressor response curves. We have also studied the baroreflex response to evoked changes in blood pressure in pregnant and non-pregnant women.
MATERIALS AND METHODS Subjects Twenty-eight pregnant women in the first or second trimester were recruited from patients admitted to the
Key words: angiotensin I1 receptors, pre-eclampsia, vascular reactivity. Abbreviations: ACE, angiotensin-converting enzyme; A N G II, angiotensin II; NA, noradrenaline. Correspondence: Dr Margaret Ramsay, Department of Therapeutics, University Hospital, Nottingham NG7 ZUH, U.K.
M. Ramsay et al,
I58
gynaecological ward for termination of pregnancy after counselling by a consultant gynaecologist unconnected with the project. None was having a termination for fetal abnormality or maternal illness; thus they were physically representative of normal pregnancy. Twenty-eight nonpregnant women were recruited from patients admitted for minor gynaecological surgery (e.g. sterilization, laparoscopy, diagnostic curettage), or were members of staff. Details of their ages, weights, basal blood pressures, heart rates, gravidity and, where appropriate, gestational ages are listed in Table 1. No woman was suffering from renal, metabolic or cardiovascular disease, or admitted taking regular medication other than iron and vitamin supplements; in particular, none was taking non-steroidal anti-inflammatory drugs or the contraceptive pill. Fullyinformed written consent was obtained from each subject, and the study was approved by the Ethical Review Boards of the University Hospital and the University of Nottingham. The subjects were divided randomly into two groups and were given an infusion of either NA or ANGII.
Protocol for infusions NA. Subjects were settled in a comfortable, semirecumbent position on a bed in a quiet, darkened room. An infusion of 0.9% (w/v) NaCl (saline)was administered via a Venflon cannula in an antecubital vein at a rate of approximately 1 ml/min. The cuff of an automatic
Table I. Details of subjects included in the study. n = I 4 in each group. Values are means with SEMS in parentheses except where stated otherwise.
sphygmomanometer (Dinamap; Critikon, Tampa, FL, U.S.A.) was placed on the left upper arm. Blood pressure and heart rate (as derived from an ECG) were recorded every 2 min throughout the experiment. The subjects were allowed to rest for at least 30 min, until their diastolic blood pressure was stable (i.e. k 3 mmHg for 20 min). At the end of this resting period, an infusion of N A acid tartrate (Levophed; Sterling-Winthrop), diluted in 5% (w/v) D-glucose to a concentration of 0.01 mg of noradrenaline base/ml, was begun via 'a connection to the intravenous cannula at a rate of 20 ng min-' kg- from an infusion pump (Treonic IP3; Vickers Medical, Basingstoke, Hants, U.K.). The constantly running saline infusion ensured that the NA was flushed rapidly into the circulation. The NA infusion rate was increased to 40,80, 120 and sometimes 160 ng min-' kg-', allowing 18 min at each increment, until a 15-20 mmHg rise in diastolic blood pressure had been achieved; an absolute blood pressure of l S O / l O O mmHg was not exceeded. Blood pressure and heart rate were monitored for 30 min after the NA infusion had been stopped. Basal blood pressure and heart rate values were calculated as the mean of values obtained during the last 20 min of the initial resting period. ANG 11. The established protocol of Broughton Pipkin et af. [7] was followed. A N G 11 amide (Hypertensin; CIBA-Geigy, Horsham, West Sussex, U.K.) was infused at rates of I, 2, 4, 8, 12 and 16 ng min-' kg-' in nonpregnant women until a 20 mmHg rise in diastolic blood pressure was achieved, allowing 10 min at each increment. Earlier studies had shown that it was necessary to use infusion rates of 4, 8, 16, 32 and sometimes 64 ng min-' kg-' in pregnant women to elicit comparable responses.
A N G II study
N A study
Statistical analysis Pregnant women
Non-pregnant Pregnant Non-pregnant women women women
Age (years)
22.9 (1.3)
26.9 (1.5)
26.3 (1.4) 30.9 (1.8)
Weight (kg)
64.9 (3.8)
63.9 (3.6)
66.1 (3.8) 59.4 (1.2)
Primigravidae (no.)
3
-
3
-
Nulligravidae (no.)
-
6
-
6
Median gravidity for the parous patients (min.-max.)
2 (2-4)
3 (1-7)
3 (2-5)
3 (1-4)
Gestation (weeks)
12.9 (0.8)
-
13.0 (0.8) -
17.2 (2.2)
-
Time from last menstrual period (days)
15.7 (2.1)
Initial blood pressure
(mmHg) Systolic Diastolic
112.2(2.1) 56.8 (1.5)
116.5(1.2) 63.5 (2.4)
113.7(2.2)117.5(2.5) 60.7 (1.4) 65.0 (2.5)
Initial heart rate (beatshin)
77.1 (2.3)
74.5 (1.9)
77.9 (2.5) 71.1 (2.5)
The evoked changes in blood pressure and heart rate were tested for skewness and were found to be normally distributed. Arithmetic mean values were thus used as measures of central tendency. Mean responses to each dose of pressor agent were calculated for each subject from the values recorded between 6 and 10 min of every infusion increment. Comparisons were made between the overall responses of the pregnant and non-pregnant women to each agent by analysis of variance or Student's t-test using the SPSSX-3 statistics package computer (Statistics Package for the Social Sciences Inc., Chicago, IL, U.S.A.) on the University mainframe computer. Slopes and intercepts derived from individual dose-pressor response curves using the linear regression equation y = ax + b [where b is the intercept on the y-axis and the slope of the line, a = ( y - b)/x]were used to calculate the dose of pressor agent required to cause a rise in systolic or diastolic blood pressure of 10 mmHg. These data were not normally distributed, so comparisons were made with non-parametric tests (Kruskal-Wallis one-way analysis of variance and Wilcoxon matched-pairs signedrank test).
Cardiovascular responseIS to pressor agents
RESULTS In all subjects, both NA and ANG I1 caused a dosedependent rise in blood pressure, although the responses were qualitatively different (Fig. 1). NA had a progressively greater effect on the systolic blood pressure, such that there was a widening of the pulse pressure with increasing dose in both pregnant and non-pregnant subjects. There was also an associated dose-dependent bradycardia, which often precluded the administration of the highest increment of the infusion. Analysis of variance demonstrated an overall difference in the responses of systolic blood pressure and heart rate between pregnant and non-pregnant subjects given NA
20 L
I
I
0
V Y
>
Y
10
I00 N A infusion rate (ng min-l kg-')
I000
25- (4
I59
( P < 0.037 and P < 0.0001, respectively), but no difference in the response of diastolic blood pressure ( P > 0.1). This can be seen in Fig. l(a), which shows that although the difference between the curves is small, nevertheless the responses in the pregnant women were consistently slightly smaller than those in the non-pregnant women. However, the dose of NA required to elicit a 10 mmHg rise in systolic or diastolic blood pressure did not differ significantly between pregnant and non-pregnant subjects (Fig. 2, Table 2). ANG II had, as expected, a more marked effect on the diastolic blood pressure in both pregnant and nonpregnant subjects. The blood pressure response to ANG II was greater at each dose level in the non-pregnant subjects than in the pregnant ones. Analysis of variance showed statistically highly significant differences between the responses of systolic blood pressure ( P< 0.001), diastolic blood pressure ( P
I57
Comparative study of pressor and heart rate responses to angiotensin II and noradrenaline in pregnant and non-pregnant women Margaret RAMSAY", Fiona BROUGHTON PlPKlNt and Peter RUBIN* Departments of * Therapeutics and fobstetria and Gynaecology, University Hospital, Nottingham, U.K. (Received I I July19 September
I99 I; accepted I 9 September I99 I )
1. Twenty-eight healthy non-pregnant women and 28 women in the first or second trimester of pregnancy were studied. They were given an incremental intravenous infusion of either noradrenaline or angiotensin 11. Pressor and heart rate responses were documented. 2. Dose-pressor response curves were constructed for the two agents in pregnant and non-pregnant women ( n = 14 in each group). The regression parameters of slope and intercept were calculated, and were used to derive the variables of dose required to elicit a 10 mmHg rise in systolic or diastolic blood pressure. 3. The pressor response to angiotensin I1 was diminished in pregnancy, with approximately twice the dose being required to raise the systolic or diastolic arterial blood pressure as in non-pregnant subjects. 4. The systolic pressor response to noradrenaline was slightly diminished in pregnancy, but the diastolic pressor response was unchanged. There were no significant differences between the doses of noradrenaline required to elicit a 10 mmHg rise in systolic or diastolic arterial blood pressure in pregnant or non-pregnant subjects. 5. There was a diminution in the bradycardia evoked in response to both hormones in pregnancy. 6. We conclude that the well-documented pressor insensitivity to angiotensin I1 during pregnancy is a specific phenomenon, not a manifestation of a generalized reduction in vascular reactivity.
INTRODUCTION Insensitivity to the pressor effects of angiotensin I1 (ANG 11)in pregnancy has been well documented following the original description by Abdul-Karim & Assali in 1961 [l].Gant et al. showed that this resistance to ANG I1 has developed by 7-8 weeks of gestation, and reaches its maximum at 28 weeks [ 2 ] .On average, 2.6 times as much ANG I1 is required to elicit a given rise in mean blood pressure in a pregnant as in a non-pregnant woman [31.
However, what is not clear from the literature published to date is whether this insensitivity to ANG I1 is a manifestation of a generalized depression in vascular reactivity in pregnancy or whether it is a phenomenon specific to ANG 11. The pressor responses to noradrenaline (NA) and adrenaline have been studied in pregnancy, but bolus doses or very short-duration infusions of the agents have been used (i.e. steady-state responses have not been considered). The studies have had discrepant results: Morandini & Mangioni [4] found the average dose of NA necessary to raise the diastolic pressure by 20 mmHg to be more than doubled in third-trimester women as compared with non-pregnant control women or women 5 days postpartum. Raab et al. [ 5 ]found no difference in the pressor responses to adrenaline or NA between nonpregnant and third-trimester women. Chesley et al. [3] found no difference in the pressor response to a bolus dose of NA in late pregnancy, but they did find a longer duration of response in the pregnant subjects. Lumbers [6], examining peripheral vascular reactivity with venous occlusion plethysmography, found no differences in response to NA between pregnant and non-pregnant subjects, although those to ANG I1 were blunted in pregnancy. The bradycardia evoked by the infusion of pressor agents was not reported in any of these studies. The purpose of this study was to determine the pressor responses to steady-state infusions of NA and ANG I1 in comparable groups of non-pregnant women, and firstand second-trimester pregnant women, by constructing dose-pressor response curves. We have also studied the baroreflex response to evoked changes in blood pressure in pregnant and non-pregnant women.
MATERIALS AND METHODS Subjects Twenty-eight pregnant women in the first or second trimester were recruited from patients admitted to the
Key words: angiotensin I1 receptors, pre-eclampsia, vascular reactivity. Abbreviations: ACE, angiotensin-converting enzyme; A N G II, angiotensin II; NA, noradrenaline. Correspondence: Dr Margaret Ramsay, Department of Therapeutics, University Hospital, Nottingham NG7 ZUH, U.K.
M. Ramsay et al,
I58
gynaecological ward for termination of pregnancy after counselling by a consultant gynaecologist unconnected with the project. None was having a termination for fetal abnormality or maternal illness; thus they were physically representative of normal pregnancy. Twenty-eight nonpregnant women were recruited from patients admitted for minor gynaecological surgery (e.g. sterilization, laparoscopy, diagnostic curettage), or were members of staff. Details of their ages, weights, basal blood pressures, heart rates, gravidity and, where appropriate, gestational ages are listed in Table 1. No woman was suffering from renal, metabolic or cardiovascular disease, or admitted taking regular medication other than iron and vitamin supplements; in particular, none was taking non-steroidal anti-inflammatory drugs or the contraceptive pill. Fullyinformed written consent was obtained from each subject, and the study was approved by the Ethical Review Boards of the University Hospital and the University of Nottingham. The subjects were divided randomly into two groups and were given an infusion of either NA or ANGII.
Protocol for infusions NA. Subjects were settled in a comfortable, semirecumbent position on a bed in a quiet, darkened room. An infusion of 0.9% (w/v) NaCl (saline)was administered via a Venflon cannula in an antecubital vein at a rate of approximately 1 ml/min. The cuff of an automatic
Table I. Details of subjects included in the study. n = I 4 in each group. Values are means with SEMS in parentheses except where stated otherwise.
sphygmomanometer (Dinamap; Critikon, Tampa, FL, U.S.A.) was placed on the left upper arm. Blood pressure and heart rate (as derived from an ECG) were recorded every 2 min throughout the experiment. The subjects were allowed to rest for at least 30 min, until their diastolic blood pressure was stable (i.e. k 3 mmHg for 20 min). At the end of this resting period, an infusion of N A acid tartrate (Levophed; Sterling-Winthrop), diluted in 5% (w/v) D-glucose to a concentration of 0.01 mg of noradrenaline base/ml, was begun via 'a connection to the intravenous cannula at a rate of 20 ng min-' kg- from an infusion pump (Treonic IP3; Vickers Medical, Basingstoke, Hants, U.K.). The constantly running saline infusion ensured that the NA was flushed rapidly into the circulation. The NA infusion rate was increased to 40,80, 120 and sometimes 160 ng min-' kg-', allowing 18 min at each increment, until a 15-20 mmHg rise in diastolic blood pressure had been achieved; an absolute blood pressure of l S O / l O O mmHg was not exceeded. Blood pressure and heart rate were monitored for 30 min after the NA infusion had been stopped. Basal blood pressure and heart rate values were calculated as the mean of values obtained during the last 20 min of the initial resting period. ANG 11. The established protocol of Broughton Pipkin et af. [7] was followed. A N G 11 amide (Hypertensin; CIBA-Geigy, Horsham, West Sussex, U.K.) was infused at rates of I, 2, 4, 8, 12 and 16 ng min-' kg-' in nonpregnant women until a 20 mmHg rise in diastolic blood pressure was achieved, allowing 10 min at each increment. Earlier studies had shown that it was necessary to use infusion rates of 4, 8, 16, 32 and sometimes 64 ng min-' kg-' in pregnant women to elicit comparable responses.
A N G II study
N A study
Statistical analysis Pregnant women
Non-pregnant Pregnant Non-pregnant women women women
Age (years)
22.9 (1.3)
26.9 (1.5)
26.3 (1.4) 30.9 (1.8)
Weight (kg)
64.9 (3.8)
63.9 (3.6)
66.1 (3.8) 59.4 (1.2)
Primigravidae (no.)
3
-
3
-
Nulligravidae (no.)
-
6
-
6
Median gravidity for the parous patients (min.-max.)
2 (2-4)
3 (1-7)
3 (2-5)
3 (1-4)
Gestation (weeks)
12.9 (0.8)
-
13.0 (0.8) -
17.2 (2.2)
-
Time from last menstrual period (days)
15.7 (2.1)
Initial blood pressure
(mmHg) Systolic Diastolic
112.2(2.1) 56.8 (1.5)
116.5(1.2) 63.5 (2.4)
113.7(2.2)117.5(2.5) 60.7 (1.4) 65.0 (2.5)
Initial heart rate (beatshin)
77.1 (2.3)
74.5 (1.9)
77.9 (2.5) 71.1 (2.5)
The evoked changes in blood pressure and heart rate were tested for skewness and were found to be normally distributed. Arithmetic mean values were thus used as measures of central tendency. Mean responses to each dose of pressor agent were calculated for each subject from the values recorded between 6 and 10 min of every infusion increment. Comparisons were made between the overall responses of the pregnant and non-pregnant women to each agent by analysis of variance or Student's t-test using the SPSSX-3 statistics package computer (Statistics Package for the Social Sciences Inc., Chicago, IL, U.S.A.) on the University mainframe computer. Slopes and intercepts derived from individual dose-pressor response curves using the linear regression equation y = ax + b [where b is the intercept on the y-axis and the slope of the line, a = ( y - b)/x]were used to calculate the dose of pressor agent required to cause a rise in systolic or diastolic blood pressure of 10 mmHg. These data were not normally distributed, so comparisons were made with non-parametric tests (Kruskal-Wallis one-way analysis of variance and Wilcoxon matched-pairs signedrank test).
Cardiovascular responseIS to pressor agents
RESULTS In all subjects, both NA and ANG I1 caused a dosedependent rise in blood pressure, although the responses were qualitatively different (Fig. 1). NA had a progressively greater effect on the systolic blood pressure, such that there was a widening of the pulse pressure with increasing dose in both pregnant and non-pregnant subjects. There was also an associated dose-dependent bradycardia, which often precluded the administration of the highest increment of the infusion. Analysis of variance demonstrated an overall difference in the responses of systolic blood pressure and heart rate between pregnant and non-pregnant subjects given NA
20 L
I
I
0
V Y
>
Y
10
I00 N A infusion rate (ng min-l kg-')
I000
25- (4
I59
( P < 0.037 and P < 0.0001, respectively), but no difference in the response of diastolic blood pressure ( P > 0.1). This can be seen in Fig. l(a), which shows that although the difference between the curves is small, nevertheless the responses in the pregnant women were consistently slightly smaller than those in the non-pregnant women. However, the dose of NA required to elicit a 10 mmHg rise in systolic or diastolic blood pressure did not differ significantly between pregnant and non-pregnant subjects (Fig. 2, Table 2). ANG II had, as expected, a more marked effect on the diastolic blood pressure in both pregnant and nonpregnant subjects. The blood pressure response to ANG II was greater at each dose level in the non-pregnant subjects than in the pregnant ones. Analysis of variance showed statistically highly significant differences between the responses of systolic blood pressure ( P< 0.001), diastolic blood pressure ( P
