- Email: [email protected]
Autonomic Imbalance in Preeclampsia: Evidence for Increased Sympathetic Tone in Response to the Supine-Pressor Test
Autonomic Imbalance in Preeclampsia: Evidence for Increased Sympathetic Tone in Response to the Supine-Pressor Test REUVEN M. LEWINSKY, MD, AND SHLOMIT RISKIN-MASHIAH, MD
Objective: To examine whether an increase in sympathetic nervous tone contributes to the augmented response to cardiovascular reflex testing in preeclamptic women. Methods: Maternal electrocardiograms were recorded from 11 nonpregnant women and 25 normotensive and 15 preeclamptic nulliparous women at term, during periods of quiet respiration in the left-lateral position and after shifting to the supine position. Power spectral analysis was applied to epochs of 512 consecutive beat-to-beat intervals to determine the contribution of sympathetic tone, parasympathetic tone, and respiratory sinus arrhythmia to heart rate variability. Results: Both normotensive and preeclamptic pregnant women showed a significant decrease in respiratory sinus arrhythmia and an increase in sympathetic tone compared with nonpregnant women. In nonpregnant and in normotensive pregnant women, shifting from the left-lateral to the supine position did not cause any change in autonomic characteristics. In contrast, preeclamptic women demonstrated a marked increase in power within the very low– frequency range representing sympathetic tone, from 288 6 214 to 556 6 322 second2/Hz, in response to the same challenge (P < .05). Conclusion: Third-trimester pregnancy is characterized by sympathetic overactivity. When complicated by preeclampsia, sympathetic overreactivity to cardiovascular reflex testing is observed. Our data support the notion that the pathophysiologic phenomena that characterize preeclampsia are mediated not only by circulating or locally acting vasoactive substances, but also, at least in part, by an increase in sympathetic nervous tone. (Obstet Gynecol 1998;91:935–9. © 1998 by The American College of Obstetricians and Gynecologists.)
From the Department of Obstetrics and Gynecology, Bnai Zion Medical Center, Faculty of Medicine, Technion Institute of Technology, Haifa, Israel. The authors thank Dr. Zeev Rogovski, Dr. Meir Steinkuler, and Prof. Gideon Inbar from the Faculty of Electrical Engineering, Technion Institute of Technology, for their contribution to the signal processing procedure.
VOL. 91, NO. 6, JUNE 1998
Preeclampsia, a disorder of the second half of pregnancy, is characterized by a marked increase in peripheral vascular resistance.1 This condition is peculiar to pregnancy and resolves shortly after delivery. Most research investigating the mechanisms active in the pathophysiology of preeclampsia has concentrated on an imbalance between vasoconstricting and vasodilating substances, either circulating or locally-acting in the endothelium. Much less attention has been directed towards a possible role of the autonomic nervous system in preeclampsia. Recently, an increase in sympathetic vasoconstrictor activity in the peripheral muscle of preeclamptic women was found compared with that of normotensive pregnant women.2 However, this may not reflect sympathetic activity in other organs, as suggested by previous studies, which did not show a generalized increase in sympathetic activity in preeclampsia.3,4 Gant and co-workers5 found a subsequent development of preeclampsia in the majority of pregnant women who at 28 –32 weeks’ gestation had an increase of at least 20 mmHg in diastolic blood pressure when shifting from the left recumbent to the supine position. This maneuver is known as the supine-pressor test or the “roll-over test.” The mechanism by which assuming the supine position incites a rise in blood pressure is not clear, but an augmented response to the decrease in cardiac preload and cardiac output secondary to inferior vena cava compression provides a partial explanation. We used power spectral analysis of maternal heart rate variability, a powerful noninvasive technique for assessing cardiovascular regulation,6 –9 to test the hypothesis that sympathetic nervous activity contributes significantly to the pathophysiologic mechanisms of preeclampsia, such as the augmented response to the supine-pressor test.
0029-7844/98/$19.00 PII S0029-7844(98)00105-7
935
Table 1. Clinical Data
No. of women Age (y) Gestational age (wk) Weight (kg)
Pregnant normotensive
Pregnant preeclamptic
25 25 6 4 33 6 3 73 6 8
15 24 6 5 35 6 4 76 6 11
Nonpregnant 11 29 6 6 65 6 6*
* P , .05 compared with pregnant normotensive and pregnant preeclamptic women.
Materials and Methods We studied 51 white women during the period December 1994 to May 1997 (Table 1). Fifteen nulliparous women in the third trimester of pregnancy fulfilled the criteria of preeclampsia. None had a history of hypertension or renal disease before the 20th week of pregnancy. All had blood pressure values exceeding 140/90 mmHg on two occasions at least 6 hours apart and proteinuria exceeding 300 mg per day. Measurements in the preeclamptic women were taken before labor and before the start of magnesium or antihypertensive therapy. Three of the patients with preeclampsia were reexamined 4 to 6 weeks after delivery, at which time their blood pressure returned to normal. The pregnant normotensive group consisted of 25 nulliparous women with comparable age, week of gestation, and body weight. These women were hospitalized for premature uterine contractions that subsided without requiring tocolytic treatment. None was diabetic or taking any medications other than iron supplementation. All were observed until delivery, and none developed hypertension. We also studied 11 nonpregnant normotensive women of similar age recruited from the medical staff. All subjects gave their informed consent to participate in the study, which was approved by the hospital review board for clinical studies. Two standard electrodes were applied to the patients’ chest, and the electrocardiogram (ECG) was digitized at a sampling rate of 500 Hz. A series of beat-to-beat intervals was stored for subsequent analysis. The first measurements were taken with the subjects in a 30degree left lateral position after at least 10 minutes of rest. The patients were then asked to turn to the supine position, and after 10 minutes of stabilization, another set of measurements was taken. Maternal blood pressure and respiratory rate were measured at the start and end of each recording in each position, and the mean value was taken for analysis. Blood pressure was measured by the standard auscultatory Riva Rocci method from the right upper arm. Diastolic blood pressure was measured at the disappearance of Korotkoff’s sounds (phase V), which was audible in all women. Power spectral analysis was performed by applying a
936 Lewinsky and Riskin-Mashiah
fast Fourier transform to the series of beat-to-beat intervals. To use the convenience of methods applicable to equally spaced time series, the heart rate was interpolated linearly to equal time intervals. To minimize the spurious effects from the data sampling process, and to avoid the shape deformation of power spectrum in the low-frequency region, all data were passed through a cosine (Hanning) window. Epochs of 512 beats were subjected to an averaging process by use of five sections of 256 beats, each offset by 64 points. An extensive description of our system for power spectral analysis of heart rate variability has been published previously.10 Each recording was observed first as a tachogram presenting equally-spaced sets of 256 beat-to-beat intervals. Sections found to include respiratory excursions indicative of the performance of the Valsalva maneuver or deep breaths (sighs) were excluded from the analysis because these respiratory maneuvers have been shown to increase sympathetic tone. The area under the curve of an individual peak represents the power or the amount of variance in the data within that frequency range. Power changes within set frequency bands are used to compare different physiologic and pharmacologic influences. We measured the power content within three predefined frequency bands: 1) very low frequency (0 – 0.05 Hz), representing vasomotor control and sympathetic tone, 2) midfrequency (0.05– 0.15 Hz), reflecting baroreceptor feedback activity mediated jointly by the sympathetic and parasympathetic systems, and 3) high frequency (0.2– 0.5 Hz), the major contributor to which is respiratory sinus arrhythmia solely under parasympathetic control. The range of 0 – 0.15 Hz is sometimes referred to as the low-frequency band, representing predominantly sympathetic tone. We elected to subdivide it into very low frequency and midfrequency because midfrequency may contain a substantial parasympathetic component. This subdivision enables comparison with previous reports using power spectral analysis of heart rate variability as well as better understanding of underlying physiologic mechanisms. Within groups, we used a paired t test, using each woman as her own control, comparing measurements of total power as well as power content within selected frequency ranges, taken at both positions. Differences between groups were compared by analysis of variance, followed by pair-wise testing using Tukey’s test. Because absolute values of power may show a high variation between individuals, differences between groups were studied by comparing the proportion of power content within frequency ranges. Power analysis revealed that if an increase of at least 30% in total power or in very low–frequency power after change to the supine position is considered significant, to detect such
Preeclampsia and Sympathetic Tone
Obstetrics & Gynecology
Table 2. Changes in Heart Rate Power Spectra and Cardiorespiratory Indices Associated With the Supine-Pressor Test Pregnant normotensive
TP (sec2/Hz) VLF (sec2/Hz) MF (sec2/Hz) HF (sec2/Hz) Systolic BP (mmHg) Diastolic BP (mmHg) Heart rate (beats/min) Respiratory rate (breaths/min)
Pregnant preeclamptic
Nonpregnant
Lateral
Supine
Lateral
Supine
Lateral
Supine
472 6 341 281 6 225 100 6 58 52 6 52 101 6 11 64 6 10 85 6 10 19 6 3
475 6 291 278 6 194 109 6 78 49 6 59 115 6 12† 76 6 9† 85 6 9 18 6 3
544 6 322 288 6 214 141 6 121 78 6 79 147 6 16‡ 93 6 8‡ 77 6 8 22 6 3
878 6 397* 556 6 322* 161 6 118 78 6 78 161 6 14† 111 6 7† 81 6 9 23 6 5
884 6 646 312 6 289 229 6 176 257 6 181 103 6 9 62 6 10 65 6 8§ 14 6 2§
1023 6 828 327 6 284 306 6 291 317 6 277 107 6 11 68 6 7 64 6 7§ 14 6 3§
TP 5 total power; VLF 5 very low frequency; MF 5 midfrequency; HF 5 high frequency; BP 5 blood pressure. * P , .05 compared with lateral position in preeclamptic pregnancies. † P , .01 compared with lateral position in same pregnancies. ‡ P , .001 compared with lateral position in pregnant normotensive group. § P , .05 compared with same position in pregnant normotensive and pregnant preeclamptic women.
a difference with a significance of P , .05 at a power level of 0.9, a sample size of at least ten cases is required in each group. Values are presented as mean 6 standard deviation.
Results Both pregnant groups had a significantly elevated heart rate and respiratory rate compared with the nonpregnant women (Table 2). A significant difference in the relative contribution to variability was found between nonpregnant women and pregnant women, unrelated to their blood pressure status (Table 3). Pregnant women showed a marked increase in very low– frequency component, concomitant with a significant reduction in the high-frequency range. The changes in heart rate power spectra and cardiorespiratory indices associated with the supine-pressor test in the three groups of subjects are shown in Table 2. As expected, systolic and diastolic blood pressure values were significantly higher in the patients with preeclampsia compared with the pregnant normotensive women. Nonpregnant women did not show any change in cardiorespiratory indices or in heart rate power spectra when shifting from the left-lateral to the supine position. In pregnant normotensive women, shifting to the supine position caused a rise in systolic and dia-
stolic blood pressure with no concomitant change in heart rate or respiratory rate. There was no change in total spectral power or its distribution between frequency bands. In contrast, in addition to the increase in blood pressure, preeclamptic women showed a significant increase in total power and in power content within the very low–frequency range. The change in very low frequency was a net increase and not a result of a shift from either the mid-frequency or the highfrequency range. The three preeclamptic women who were reexamined 4 – 6 weeks after delivery were all normotensive at the time of examination. The mean value of total power at the left-lateral position was 986 6 417 sec2/Hz, with a relative power content of 34 6 11%, 28 6 10%, and 32 6 14% for very low– mid- and high-frequency ranges, respectively. These values did not change with a shift to the supine position and are similar to those in the nonpregnant control group.
Discussion We found that preeclamptic women show a marked increase in cardiac sympathetic tone in response to noninvasive cardiovascular reflex testing with the supine-pressor test. This phenomenon did not occur in pregnant normotensive women or in nonpregnant
Table 3. Proportions of Power Content Within Frequency Ranges, in the Three Groups of Subjects Pregnant normotensive
VLF (%) MF (%) HF (%)
Pregnant preeclamptic
Nonpregnant
Lateral
Supine
Lateral
Supine
Lateral
Supine
57 6 14* 24 6 9 10 6 8*
59 6 14* 24 6 10 9 6 8*
53 6 15* 27 6 12 12 6 7*
62 6 15* 20 6 15 9 6 8*
36 6 15 26 6 8 29 6 15
32 6 17 30 6 10 31 6 17
Abbreviations as in Table 3. * P , .01 compared with same position in nonpregnant group.
VOL. 91, NO. 6, JUNE 1998
Lewinsky and Riskin-Mashiah
Preeclampsia and Sympathetic Tone
937
women. Furthermore, we found that the sympathetic overreactivity in preeclamptic women is superimposed on an increased sympathetic tone characteristic for pregnancy, whether complicated by preeclampsia or not, when compared with the nonpregnant state. These findings suggest that the augmented response to the supine-pressor test is mediated not only by circulating or locally-acting vasoactive substances, as previously suggested, but also, at least in part, by an increase in sympathetic nervous tone. Power spectral analysis of heart rate variability is a well-established technique that has been used widely for study of the autonomic modulation of cardiac function in both health and disease.6 Studies in both humans7 and conscious dogs8,9 using blocking agents specific to the renin-angiotensin, sympathetic, and parasympathetic system, provided the basis for the association with the very low–, low-, and mid-frequency peaks, respectively. Respiratory sinus arrhythmia is the major contributor to power in the high-frequency range greater than 0.15 Hz. We found that pregnancy itself, even when uncomplicated by preeclampsia, is characterized by a marked shift of power from the high-frequency to the very low–frequency range, indicating an increase in sympathetic tone and a decrease in respiratory sinus arrhythmia. Assali et al11 stated that increased sympathetic tone prevails during the second half of pregnancy, on the basis of the observation that the action of ganglionic blocking agents on blood pressure is exceptionally potent when compared with that of nonpregnant subjects. The significantly elevated heart rate observed in pregnant women also supports the existence of an underlying increase in sympathetic cardiac activity. Therefore, it appears that a central resetting of autonomic balance with a trend towards sympathetic dominance occurs in normal pregnancy, with the decrease in high-frequency oscillations a relative rather than an absolute change. When subjected to the supine-pressor test, preeclamptic women showed a significant increase in total power, most of which was the result of a rise in the power content of very low–frequency oscillations, representing an increase in sympathetic modulation of the heart. The change in total power or its distribution was not observed in nonpregnant or in pregnant normotensive women. The supine-pressor test, also known as the roll-over test, was first described by Gant et al5 as a clinical test to predict preeclampsia. As expected, perhaps, those women who demonstrated a positive supine-pressor response also were abnormally sensitive to infused angiotensin II, whereas those without a hypertensive response were normally refractory. The exact mechanism by which assumption of the supine position
938 Lewinsky and Riskin-Mashiah
incites a rise in blood pressure is not clear. It is postulated that the enlarged pregnant uterus compresses the abdominal vena cava, thus interfering with venous return to the heart, causing a decrease in cardiac output. This in turn causes diminished renal perfusion, leading to activation of the renin-angiotensin system in the juxtaglomerular apparatus. Although we performed the supine-pressor test, our study differs from the typical test in three major aspects: 1) we studied the mechanism of preeclampsia once the disease is already established; 2) the gestational age of our subjects was more advanced than the usual range of 28 –32 weeks at which the supine-pressor test is usually performed, the larger uterus probably exerting more pressure on the vena cava; and 3) our end result was not a change in blood pressure, which may be partly a result of a change in hydrostatic pressure, but rather an objective measure of heart rate variability. Our results indicate that the type of stressful stimulus influences the nature of sympathetic response in both a quantitative and qualitative fashion. There was no significant difference between pregnant normotensive and pregnant hypertensive women in their blood pressure, heart rate, plasma noradrenaline levels, or vascular resistance when subjected to an isometric handgrip test or to the cold-pressor test.3 Conversely, when subjected to tilting, hypertensive patients had higher arterial epinephrine and norepinephrine responses than control subjects during pregnancy.4 When our data are viewed in the perspective of previous observations, it appears that the provocation generated by somatic cardiovascular reflexes not dependent on volume (low pressure) receptor activation, ie, venous return, is insufficient to elucidate the difference between preeclamptic and normotensive pregnant women. The lesspronounced pregnancy-induced increase in blood volume in patients with preeclampsia appears to explain their increased response to a preload-affecting cardiovascular challenge such as the supine-pressor test; therefore, this test appears to be more adequate for the study of this specific disease state. In a recent study, Schobel et al2 measured sympathetic-nerve activity in blood vessels of skeletal muscle. At the resting state, a significantly higher level was found in preeclamptic women when compared with normotensive pregnant women. The authors concluded that “preeclampsia is a state of sympathetic overactivity.” We found that pregnancy itself, whether complicated by preeclampsia or not, is a state of cardiac sympathetic overactivity. A limitation of the study by Schobel et al,2 as stated by the authors themselves, is the fact that sympathetic outflow to skeletal muscle may not reflect sympathetic activity in other organs. Nissel et al3,4 have stated that there is no evidence to support the conten-
Preeclampsia and Sympathetic Tone
Obstetrics & Gynecology
tion of a generalized increase in sympathetic nerve activity in preeclampsia. Our findings support the notion that sympathetic nerve activity occurs in a highly differentiated, organ-specific fashion, as previously discussed for essential hypertension by Folkow et al,12 who found that although sympathetic activity to the heart is increased, muscle sympathetic activity and muscle vascular resistance are unchanged or reduced. As already mentioned, Akselrod et al8,9 have shown that angiotensin-converting enzyme inhibitors cause an increase in very low–frequency power. While the juxtaglomerular apparatus in the kidney was believed originally to be the main site for release of renin and initiation of the renin-angiotensin cascade, recently, more evidence is accumulating of the presence of all the components of the system in various tissues including the brain, blood vessels, and heart.13 The pathophysiologic phenomena in preeclampsia and the activation of the renin-angiotensin system may well be organspecific. If proven to be so, central-nervous or local mechanisms in the heart may explain our observation of changes in sympathetic modulation of the heart in normal and preeclamptic pregnancies.
8.
9.
10. 11.
12.
13.
Address reprint requests to:
References 1. Cunningham FG, Lindheimer MD. Hypertension in pregnancy. N Engl J Med 1992;326:927–32. 2. Schobel HP, Fischer T, Heuszer K, Geiger H, Schmieder RE. Preeclampsia—a state of sympathetic overactivity. N Engl J Med 1996;335:1480 –5. 3. Nissel H, Hjemdahl P, Linde B, Olov-Lunell N. Sympathoadrenal and cardiovascular reactivity in pregnancy-induced hypertension. I. Responses to isometric exercise and a cold pressor test. Br J Obstet Gynaecol 1985;92:722–31. 4. Nissel H, Hjemdahl P, Linde B, Olov-Lunell N. Sympathoadrenal and cardiovascular reactivity in pregnancy-induced hypertension. I. Responses to tilting. Am J Obstet Gynecol 1985;152:554 – 60. 5. Gant NF, Chand S, Worley RJ, Whalley PJ, Crosby UD, MacDonald PC. A clinical test useful for predicting the development of acute hypertension in pregnancy. Am J Obstet Gynecol 1974;120:1–7. 6. van Ravenswaaij-Arts CMA, Kolle´e LAA, Hopman JCW, Stoelinga
VOL. 91, NO. 6, JUNE 1998
7.
GBA, van Geijn HP. Heart rate variability. Ann Intern Med 1993;118:436 – 47. Pomeranz B, Macaulay RJB, Caudill MA, Kutz I, Adam D, Gordon D. Assessment of autonomic function in humans by heart rate spectral analysis. Am J Physiol 1985;248:H151–3. Akselrod S, Gordon D, Ubel FA, Shannon DC, Barger AC, Cohen RJ. Power spectrum analysis of heart rate fluctuation: A quantitative probe of beat-to-beat cardiovascular control. Science 1981;213: 220 –3. Akselrod S, Gordon D, Madwed JB, Snidman NC, Shannon DC, Cohen RJ. Hemodynamic regulation: Investigation by spectral analysis. Am J Physiol 1985;249:H867–75. Oppenheimer LW, Lewinsky RM. Power spectral analysis of fetal heart rate. Baillieres Clin Obstet Gynecol 1994;8:643– 61. Assali NS, Vergon JM, Tada Y, Garber ST. Studies on autonomic blockade. VI. The mechanisms regulating the hemodynamic changes in the pregnant woman and their relation to the hypertension of toxemia of pregnancy. Am J Obstet Gynecol 1952;63: 978 – 88. Folkow B, Di Bona GF, Hjemdahl P, Toren PH, Wallin BG. Measurements of plasma norepinephrine concentrations in human primary hypertension. A word of caution on their applicability for assessing neurogenic contributions. Hypertension 1983;5:399 – 403. Phillips MI, Spealman EA, Kimure B. Levels of angiotensin and molecular biology of the tissue renin angiotensin system. Regul Pept 1993;43:1–20.
Reuven M. Lewinsky, MD Department of Obstetrics and Gynecology Bnai Zion Medical Center 47 Golomb Street PO Box 4940 Haifa 31048 Israel
Received October 15, 1997. Received in revised form February 10, 1998. Accepted February 26, 1998.
Copyright © 1998 by The American College of Obstetricians and Gynecologists. Published by Elsevier Science Inc.
Lewinsky and Riskin-Mashiah
Preeclampsia and Sympathetic Tone
939
Objective: To examine whether an increase in sympathetic nervous tone contributes to the augmented response to cardiovascular reflex testing in preeclamptic women. Methods: Maternal electrocardiograms were recorded from 11 nonpregnant women and 25 normotensive and 15 preeclamptic nulliparous women at term, during periods of quiet respiration in the left-lateral position and after shifting to the supine position. Power spectral analysis was applied to epochs of 512 consecutive beat-to-beat intervals to determine the contribution of sympathetic tone, parasympathetic tone, and respiratory sinus arrhythmia to heart rate variability. Results: Both normotensive and preeclamptic pregnant women showed a significant decrease in respiratory sinus arrhythmia and an increase in sympathetic tone compared with nonpregnant women. In nonpregnant and in normotensive pregnant women, shifting from the left-lateral to the supine position did not cause any change in autonomic characteristics. In contrast, preeclamptic women demonstrated a marked increase in power within the very low– frequency range representing sympathetic tone, from 288 6 214 to 556 6 322 second2/Hz, in response to the same challenge (P < .05). Conclusion: Third-trimester pregnancy is characterized by sympathetic overactivity. When complicated by preeclampsia, sympathetic overreactivity to cardiovascular reflex testing is observed. Our data support the notion that the pathophysiologic phenomena that characterize preeclampsia are mediated not only by circulating or locally acting vasoactive substances, but also, at least in part, by an increase in sympathetic nervous tone. (Obstet Gynecol 1998;91:935–9. © 1998 by The American College of Obstetricians and Gynecologists.)
From the Department of Obstetrics and Gynecology, Bnai Zion Medical Center, Faculty of Medicine, Technion Institute of Technology, Haifa, Israel. The authors thank Dr. Zeev Rogovski, Dr. Meir Steinkuler, and Prof. Gideon Inbar from the Faculty of Electrical Engineering, Technion Institute of Technology, for their contribution to the signal processing procedure.
VOL. 91, NO. 6, JUNE 1998
Preeclampsia, a disorder of the second half of pregnancy, is characterized by a marked increase in peripheral vascular resistance.1 This condition is peculiar to pregnancy and resolves shortly after delivery. Most research investigating the mechanisms active in the pathophysiology of preeclampsia has concentrated on an imbalance between vasoconstricting and vasodilating substances, either circulating or locally-acting in the endothelium. Much less attention has been directed towards a possible role of the autonomic nervous system in preeclampsia. Recently, an increase in sympathetic vasoconstrictor activity in the peripheral muscle of preeclamptic women was found compared with that of normotensive pregnant women.2 However, this may not reflect sympathetic activity in other organs, as suggested by previous studies, which did not show a generalized increase in sympathetic activity in preeclampsia.3,4 Gant and co-workers5 found a subsequent development of preeclampsia in the majority of pregnant women who at 28 –32 weeks’ gestation had an increase of at least 20 mmHg in diastolic blood pressure when shifting from the left recumbent to the supine position. This maneuver is known as the supine-pressor test or the “roll-over test.” The mechanism by which assuming the supine position incites a rise in blood pressure is not clear, but an augmented response to the decrease in cardiac preload and cardiac output secondary to inferior vena cava compression provides a partial explanation. We used power spectral analysis of maternal heart rate variability, a powerful noninvasive technique for assessing cardiovascular regulation,6 –9 to test the hypothesis that sympathetic nervous activity contributes significantly to the pathophysiologic mechanisms of preeclampsia, such as the augmented response to the supine-pressor test.
0029-7844/98/$19.00 PII S0029-7844(98)00105-7
935
Table 1. Clinical Data
No. of women Age (y) Gestational age (wk) Weight (kg)
Pregnant normotensive
Pregnant preeclamptic
25 25 6 4 33 6 3 73 6 8
15 24 6 5 35 6 4 76 6 11
Nonpregnant 11 29 6 6 65 6 6*
* P , .05 compared with pregnant normotensive and pregnant preeclamptic women.
Materials and Methods We studied 51 white women during the period December 1994 to May 1997 (Table 1). Fifteen nulliparous women in the third trimester of pregnancy fulfilled the criteria of preeclampsia. None had a history of hypertension or renal disease before the 20th week of pregnancy. All had blood pressure values exceeding 140/90 mmHg on two occasions at least 6 hours apart and proteinuria exceeding 300 mg per day. Measurements in the preeclamptic women were taken before labor and before the start of magnesium or antihypertensive therapy. Three of the patients with preeclampsia were reexamined 4 to 6 weeks after delivery, at which time their blood pressure returned to normal. The pregnant normotensive group consisted of 25 nulliparous women with comparable age, week of gestation, and body weight. These women were hospitalized for premature uterine contractions that subsided without requiring tocolytic treatment. None was diabetic or taking any medications other than iron supplementation. All were observed until delivery, and none developed hypertension. We also studied 11 nonpregnant normotensive women of similar age recruited from the medical staff. All subjects gave their informed consent to participate in the study, which was approved by the hospital review board for clinical studies. Two standard electrodes were applied to the patients’ chest, and the electrocardiogram (ECG) was digitized at a sampling rate of 500 Hz. A series of beat-to-beat intervals was stored for subsequent analysis. The first measurements were taken with the subjects in a 30degree left lateral position after at least 10 minutes of rest. The patients were then asked to turn to the supine position, and after 10 minutes of stabilization, another set of measurements was taken. Maternal blood pressure and respiratory rate were measured at the start and end of each recording in each position, and the mean value was taken for analysis. Blood pressure was measured by the standard auscultatory Riva Rocci method from the right upper arm. Diastolic blood pressure was measured at the disappearance of Korotkoff’s sounds (phase V), which was audible in all women. Power spectral analysis was performed by applying a
936 Lewinsky and Riskin-Mashiah
fast Fourier transform to the series of beat-to-beat intervals. To use the convenience of methods applicable to equally spaced time series, the heart rate was interpolated linearly to equal time intervals. To minimize the spurious effects from the data sampling process, and to avoid the shape deformation of power spectrum in the low-frequency region, all data were passed through a cosine (Hanning) window. Epochs of 512 beats were subjected to an averaging process by use of five sections of 256 beats, each offset by 64 points. An extensive description of our system for power spectral analysis of heart rate variability has been published previously.10 Each recording was observed first as a tachogram presenting equally-spaced sets of 256 beat-to-beat intervals. Sections found to include respiratory excursions indicative of the performance of the Valsalva maneuver or deep breaths (sighs) were excluded from the analysis because these respiratory maneuvers have been shown to increase sympathetic tone. The area under the curve of an individual peak represents the power or the amount of variance in the data within that frequency range. Power changes within set frequency bands are used to compare different physiologic and pharmacologic influences. We measured the power content within three predefined frequency bands: 1) very low frequency (0 – 0.05 Hz), representing vasomotor control and sympathetic tone, 2) midfrequency (0.05– 0.15 Hz), reflecting baroreceptor feedback activity mediated jointly by the sympathetic and parasympathetic systems, and 3) high frequency (0.2– 0.5 Hz), the major contributor to which is respiratory sinus arrhythmia solely under parasympathetic control. The range of 0 – 0.15 Hz is sometimes referred to as the low-frequency band, representing predominantly sympathetic tone. We elected to subdivide it into very low frequency and midfrequency because midfrequency may contain a substantial parasympathetic component. This subdivision enables comparison with previous reports using power spectral analysis of heart rate variability as well as better understanding of underlying physiologic mechanisms. Within groups, we used a paired t test, using each woman as her own control, comparing measurements of total power as well as power content within selected frequency ranges, taken at both positions. Differences between groups were compared by analysis of variance, followed by pair-wise testing using Tukey’s test. Because absolute values of power may show a high variation between individuals, differences between groups were studied by comparing the proportion of power content within frequency ranges. Power analysis revealed that if an increase of at least 30% in total power or in very low–frequency power after change to the supine position is considered significant, to detect such
Preeclampsia and Sympathetic Tone
Obstetrics & Gynecology
Table 2. Changes in Heart Rate Power Spectra and Cardiorespiratory Indices Associated With the Supine-Pressor Test Pregnant normotensive
TP (sec2/Hz) VLF (sec2/Hz) MF (sec2/Hz) HF (sec2/Hz) Systolic BP (mmHg) Diastolic BP (mmHg) Heart rate (beats/min) Respiratory rate (breaths/min)
Pregnant preeclamptic
Nonpregnant
Lateral
Supine
Lateral
Supine
Lateral
Supine
472 6 341 281 6 225 100 6 58 52 6 52 101 6 11 64 6 10 85 6 10 19 6 3
475 6 291 278 6 194 109 6 78 49 6 59 115 6 12† 76 6 9† 85 6 9 18 6 3
544 6 322 288 6 214 141 6 121 78 6 79 147 6 16‡ 93 6 8‡ 77 6 8 22 6 3
878 6 397* 556 6 322* 161 6 118 78 6 78 161 6 14† 111 6 7† 81 6 9 23 6 5
884 6 646 312 6 289 229 6 176 257 6 181 103 6 9 62 6 10 65 6 8§ 14 6 2§
1023 6 828 327 6 284 306 6 291 317 6 277 107 6 11 68 6 7 64 6 7§ 14 6 3§
TP 5 total power; VLF 5 very low frequency; MF 5 midfrequency; HF 5 high frequency; BP 5 blood pressure. * P , .05 compared with lateral position in preeclamptic pregnancies. † P , .01 compared with lateral position in same pregnancies. ‡ P , .001 compared with lateral position in pregnant normotensive group. § P , .05 compared with same position in pregnant normotensive and pregnant preeclamptic women.
a difference with a significance of P , .05 at a power level of 0.9, a sample size of at least ten cases is required in each group. Values are presented as mean 6 standard deviation.
Results Both pregnant groups had a significantly elevated heart rate and respiratory rate compared with the nonpregnant women (Table 2). A significant difference in the relative contribution to variability was found between nonpregnant women and pregnant women, unrelated to their blood pressure status (Table 3). Pregnant women showed a marked increase in very low– frequency component, concomitant with a significant reduction in the high-frequency range. The changes in heart rate power spectra and cardiorespiratory indices associated with the supine-pressor test in the three groups of subjects are shown in Table 2. As expected, systolic and diastolic blood pressure values were significantly higher in the patients with preeclampsia compared with the pregnant normotensive women. Nonpregnant women did not show any change in cardiorespiratory indices or in heart rate power spectra when shifting from the left-lateral to the supine position. In pregnant normotensive women, shifting to the supine position caused a rise in systolic and dia-
stolic blood pressure with no concomitant change in heart rate or respiratory rate. There was no change in total spectral power or its distribution between frequency bands. In contrast, in addition to the increase in blood pressure, preeclamptic women showed a significant increase in total power and in power content within the very low–frequency range. The change in very low frequency was a net increase and not a result of a shift from either the mid-frequency or the highfrequency range. The three preeclamptic women who were reexamined 4 – 6 weeks after delivery were all normotensive at the time of examination. The mean value of total power at the left-lateral position was 986 6 417 sec2/Hz, with a relative power content of 34 6 11%, 28 6 10%, and 32 6 14% for very low– mid- and high-frequency ranges, respectively. These values did not change with a shift to the supine position and are similar to those in the nonpregnant control group.
Discussion We found that preeclamptic women show a marked increase in cardiac sympathetic tone in response to noninvasive cardiovascular reflex testing with the supine-pressor test. This phenomenon did not occur in pregnant normotensive women or in nonpregnant
Table 3. Proportions of Power Content Within Frequency Ranges, in the Three Groups of Subjects Pregnant normotensive
VLF (%) MF (%) HF (%)
Pregnant preeclamptic
Nonpregnant
Lateral
Supine
Lateral
Supine
Lateral
Supine
57 6 14* 24 6 9 10 6 8*
59 6 14* 24 6 10 9 6 8*
53 6 15* 27 6 12 12 6 7*
62 6 15* 20 6 15 9 6 8*
36 6 15 26 6 8 29 6 15
32 6 17 30 6 10 31 6 17
Abbreviations as in Table 3. * P , .01 compared with same position in nonpregnant group.
VOL. 91, NO. 6, JUNE 1998
Lewinsky and Riskin-Mashiah
Preeclampsia and Sympathetic Tone
937
women. Furthermore, we found that the sympathetic overreactivity in preeclamptic women is superimposed on an increased sympathetic tone characteristic for pregnancy, whether complicated by preeclampsia or not, when compared with the nonpregnant state. These findings suggest that the augmented response to the supine-pressor test is mediated not only by circulating or locally-acting vasoactive substances, as previously suggested, but also, at least in part, by an increase in sympathetic nervous tone. Power spectral analysis of heart rate variability is a well-established technique that has been used widely for study of the autonomic modulation of cardiac function in both health and disease.6 Studies in both humans7 and conscious dogs8,9 using blocking agents specific to the renin-angiotensin, sympathetic, and parasympathetic system, provided the basis for the association with the very low–, low-, and mid-frequency peaks, respectively. Respiratory sinus arrhythmia is the major contributor to power in the high-frequency range greater than 0.15 Hz. We found that pregnancy itself, even when uncomplicated by preeclampsia, is characterized by a marked shift of power from the high-frequency to the very low–frequency range, indicating an increase in sympathetic tone and a decrease in respiratory sinus arrhythmia. Assali et al11 stated that increased sympathetic tone prevails during the second half of pregnancy, on the basis of the observation that the action of ganglionic blocking agents on blood pressure is exceptionally potent when compared with that of nonpregnant subjects. The significantly elevated heart rate observed in pregnant women also supports the existence of an underlying increase in sympathetic cardiac activity. Therefore, it appears that a central resetting of autonomic balance with a trend towards sympathetic dominance occurs in normal pregnancy, with the decrease in high-frequency oscillations a relative rather than an absolute change. When subjected to the supine-pressor test, preeclamptic women showed a significant increase in total power, most of which was the result of a rise in the power content of very low–frequency oscillations, representing an increase in sympathetic modulation of the heart. The change in total power or its distribution was not observed in nonpregnant or in pregnant normotensive women. The supine-pressor test, also known as the roll-over test, was first described by Gant et al5 as a clinical test to predict preeclampsia. As expected, perhaps, those women who demonstrated a positive supine-pressor response also were abnormally sensitive to infused angiotensin II, whereas those without a hypertensive response were normally refractory. The exact mechanism by which assumption of the supine position
938 Lewinsky and Riskin-Mashiah
incites a rise in blood pressure is not clear. It is postulated that the enlarged pregnant uterus compresses the abdominal vena cava, thus interfering with venous return to the heart, causing a decrease in cardiac output. This in turn causes diminished renal perfusion, leading to activation of the renin-angiotensin system in the juxtaglomerular apparatus. Although we performed the supine-pressor test, our study differs from the typical test in three major aspects: 1) we studied the mechanism of preeclampsia once the disease is already established; 2) the gestational age of our subjects was more advanced than the usual range of 28 –32 weeks at which the supine-pressor test is usually performed, the larger uterus probably exerting more pressure on the vena cava; and 3) our end result was not a change in blood pressure, which may be partly a result of a change in hydrostatic pressure, but rather an objective measure of heart rate variability. Our results indicate that the type of stressful stimulus influences the nature of sympathetic response in both a quantitative and qualitative fashion. There was no significant difference between pregnant normotensive and pregnant hypertensive women in their blood pressure, heart rate, plasma noradrenaline levels, or vascular resistance when subjected to an isometric handgrip test or to the cold-pressor test.3 Conversely, when subjected to tilting, hypertensive patients had higher arterial epinephrine and norepinephrine responses than control subjects during pregnancy.4 When our data are viewed in the perspective of previous observations, it appears that the provocation generated by somatic cardiovascular reflexes not dependent on volume (low pressure) receptor activation, ie, venous return, is insufficient to elucidate the difference between preeclamptic and normotensive pregnant women. The lesspronounced pregnancy-induced increase in blood volume in patients with preeclampsia appears to explain their increased response to a preload-affecting cardiovascular challenge such as the supine-pressor test; therefore, this test appears to be more adequate for the study of this specific disease state. In a recent study, Schobel et al2 measured sympathetic-nerve activity in blood vessels of skeletal muscle. At the resting state, a significantly higher level was found in preeclamptic women when compared with normotensive pregnant women. The authors concluded that “preeclampsia is a state of sympathetic overactivity.” We found that pregnancy itself, whether complicated by preeclampsia or not, is a state of cardiac sympathetic overactivity. A limitation of the study by Schobel et al,2 as stated by the authors themselves, is the fact that sympathetic outflow to skeletal muscle may not reflect sympathetic activity in other organs. Nissel et al3,4 have stated that there is no evidence to support the conten-
Preeclampsia and Sympathetic Tone
Obstetrics & Gynecology
tion of a generalized increase in sympathetic nerve activity in preeclampsia. Our findings support the notion that sympathetic nerve activity occurs in a highly differentiated, organ-specific fashion, as previously discussed for essential hypertension by Folkow et al,12 who found that although sympathetic activity to the heart is increased, muscle sympathetic activity and muscle vascular resistance are unchanged or reduced. As already mentioned, Akselrod et al8,9 have shown that angiotensin-converting enzyme inhibitors cause an increase in very low–frequency power. While the juxtaglomerular apparatus in the kidney was believed originally to be the main site for release of renin and initiation of the renin-angiotensin cascade, recently, more evidence is accumulating of the presence of all the components of the system in various tissues including the brain, blood vessels, and heart.13 The pathophysiologic phenomena in preeclampsia and the activation of the renin-angiotensin system may well be organspecific. If proven to be so, central-nervous or local mechanisms in the heart may explain our observation of changes in sympathetic modulation of the heart in normal and preeclamptic pregnancies.
8.
9.
10. 11.
12.
13.
Address reprint requests to:
References 1. Cunningham FG, Lindheimer MD. Hypertension in pregnancy. N Engl J Med 1992;326:927–32. 2. Schobel HP, Fischer T, Heuszer K, Geiger H, Schmieder RE. Preeclampsia—a state of sympathetic overactivity. N Engl J Med 1996;335:1480 –5. 3. Nissel H, Hjemdahl P, Linde B, Olov-Lunell N. Sympathoadrenal and cardiovascular reactivity in pregnancy-induced hypertension. I. Responses to isometric exercise and a cold pressor test. Br J Obstet Gynaecol 1985;92:722–31. 4. Nissel H, Hjemdahl P, Linde B, Olov-Lunell N. Sympathoadrenal and cardiovascular reactivity in pregnancy-induced hypertension. I. Responses to tilting. Am J Obstet Gynecol 1985;152:554 – 60. 5. Gant NF, Chand S, Worley RJ, Whalley PJ, Crosby UD, MacDonald PC. A clinical test useful for predicting the development of acute hypertension in pregnancy. Am J Obstet Gynecol 1974;120:1–7. 6. van Ravenswaaij-Arts CMA, Kolle´e LAA, Hopman JCW, Stoelinga
VOL. 91, NO. 6, JUNE 1998
7.
GBA, van Geijn HP. Heart rate variability. Ann Intern Med 1993;118:436 – 47. Pomeranz B, Macaulay RJB, Caudill MA, Kutz I, Adam D, Gordon D. Assessment of autonomic function in humans by heart rate spectral analysis. Am J Physiol 1985;248:H151–3. Akselrod S, Gordon D, Ubel FA, Shannon DC, Barger AC, Cohen RJ. Power spectrum analysis of heart rate fluctuation: A quantitative probe of beat-to-beat cardiovascular control. Science 1981;213: 220 –3. Akselrod S, Gordon D, Madwed JB, Snidman NC, Shannon DC, Cohen RJ. Hemodynamic regulation: Investigation by spectral analysis. Am J Physiol 1985;249:H867–75. Oppenheimer LW, Lewinsky RM. Power spectral analysis of fetal heart rate. Baillieres Clin Obstet Gynecol 1994;8:643– 61. Assali NS, Vergon JM, Tada Y, Garber ST. Studies on autonomic blockade. VI. The mechanisms regulating the hemodynamic changes in the pregnant woman and their relation to the hypertension of toxemia of pregnancy. Am J Obstet Gynecol 1952;63: 978 – 88. Folkow B, Di Bona GF, Hjemdahl P, Toren PH, Wallin BG. Measurements of plasma norepinephrine concentrations in human primary hypertension. A word of caution on their applicability for assessing neurogenic contributions. Hypertension 1983;5:399 – 403. Phillips MI, Spealman EA, Kimure B. Levels of angiotensin and molecular biology of the tissue renin angiotensin system. Regul Pept 1993;43:1–20.
Reuven M. Lewinsky, MD Department of Obstetrics and Gynecology Bnai Zion Medical Center 47 Golomb Street PO Box 4940 Haifa 31048 Israel
Received October 15, 1997. Received in revised form February 10, 1998. Accepted February 26, 1998.
Copyright © 1998 by The American College of Obstetricians and Gynecologists. Published by Elsevier Science Inc.
Lewinsky and Riskin-Mashiah
Preeclampsia and Sympathetic Tone
939