Journal of Human (1999) 13, 179–183  1999 Stockton Press. All rights reserved 0950-9240/99 $12.00 http://www.stockton-press.co.uk/jhh ORIGINAL ARTICLE Baroreflex control of rate is impaired in pre-eclampsia

P Molino1, F Veglio1, G Cat Genova1, R Melchio1, C Benedetto2, L Chiarolini2, F Rabbia1, T Grosso1, P Mulatero1 and L Chiandussi1 1Department of Medicine and Experimental Oncology, and 2Department of Obstetric and Gynecology, University of Turin, Italy

Autonomic nervous dysfunction, such as parasympa- cross-spectral analysis of systolic and thetic and sympathetic impairment, has been suggested interval. The result was that baroreflex gain at rest as possible cause of pre-eclampsia, but the studies are was lower in pre-eclamptic women both compared to not conclusive. Our purpose was to assess non-invas- non-pregnant and healthy pregnant subjects (P Ͻ 0.05). ively if pre-eclampsia is associated with a decreased Moreover, a decrease of the baroreflex sensitivity was baroreflex function. Nine women with pre-eclampsia present in all pregnant women in the orthostatic pos- (PE), eight normotensive pregnant women, and seven ition (P Ͻ 0.05). In conclusion pregnancy per se is asso- healthy normotensive non-pregnant women were stud- ciated with a decrease in the baroreflex control of the ied. Continuous finger blood pressure was recorded by heart, whereas in pre-eclampsia, the baroreflex sensi- a Portapres device in the left lateral recumbent position tivity is impaired further. and active standing. Baroreflex gain was evaluated by

Keywords: baroreflex; ; pre-eclampsia

Introduction The interest in baroreflex activity is due to the blood pressure lability and enhanced responsive- Pre-eclampsia is a multisystemic disorder of preg- ness to that characterises pre-eclampsia, nancy characterised by proteinuria, and a marked which could reflect, at least in part, baroreflex dys- increase in peripheral , which, in function.7 Furthermore, in pre-eclampsia the endo- turn, causes an increase in blood pressure. The thelial dysfunction may represent a possible cause pathophysiological mechanism is still unknown, of the baroreflex sensitivity impairment.8,9 although some reports suggest an autonomic ner- The aim of our study was to assess non-invasively vous dysfunction such as sympathetic and parasym- the baroreflex sensitivity in pre-eclamptic patients pathetic impairment.1–3 Most studies have investi- gated autonomic nervous function by non-invasive (PE) compared to normal pregnant, and normotens- evaluations using standard cardiovascular tests (ie, ive non-pregnant women. The study was performed deep breathing, standing up, cold pressure tests and with the Portapres device, a portable battery-oper- ated device for the continuous non-invasive Valsalva manoeuvre), or by measuring the sympath- 10 etic nerve activity in the blood vessels of the skeletal measurement of blood pressure in the finger. muscle with intraneural microelectrodes.1 Few stud- Power spectral analysis of arterial blood pressure ies used spectral analysis of blood pressure and and interbeat interval was performed in the fre- variability to evaluate autonomic nervous quency domain; this represents a powerful and non- system function in pre-eclampsia and pregnancy- invasive tool which appears to provide a quantitat- induced hypertension, showing conflicting results ive evaluation of the efferent arc of the autonomic ranging from decreased vagal control of the heart to nervous system activity modulating cardiac function increased sympathetic and parasympathetic control and systemic circulation, allowing the dissection of of heart rate and blood pressure.2,4,5 In addition, the effects of the sympathetic and parasympathetic baroreflex control has not been investigated to date sections of the autonomic nervous system on the in a dynamic way, by the frequency domain cardiovascular function. When applied to the heart approach, such as cross-spectrum analysis and cal- rate, power spectral analysis allows for the detection culation of ␣-index, which has become a useful and of the low frequency component (LF, marker of sym- non-invasive assessment of the spontaneous barore- pathetic activity), and of the high frequency compo- flex gain.5,6 nents (HF, related to the respiratory activity, which is mediated by the ).11–14 The ratio between the LF and HF components (LF/HF) is, Correspondence: Dr Franco Veglio, Cattedra di Medicina Interna, therefore, an index of the sympatho-vagal balance.15 Ospedale San Vito, Strada San Vito 34, 10133 Torino, Italy Received 5 November 1998; revised and accepted 31 December 1998 Baroreflex in pre-eclampsia P Molino et al 180 Materials and methods beat basis, according to the arterial volume-clamp method of Penaz.17–19 Subjects

Twenty-four women were analysed: nine PE Computations of haemodynamic parameters women; eight normotensive pregnant women matched for age, week of gestation and body weight Data were analysed by the Beatfast program (TNO, with the pre-eclamptic women; and seven healthy Amsterdam, NL) to identify systolic and diastolic normotensive non-pregnant women of similar age blood pressure values and interbeat intervals for (Table 1). each : after digital low pass filtering at Pre-eclampsia was defined according to the cri- 17 Hz, systolic blood pressure is stored as the high- teria of the International Society for the Study of est pressure level during the ejection. Diastolic Hypertension in Pregnancy:16 no previous history of blood pressure is arterial end-diastolic pressure just hypertension, cardiovascular or renal disease before before the next upstroke begins. Interbeat interval is pregnancy; blood pressure values exceeding 140/90 the time difference between two following systolic mm Hg after the 20th week of gestation, confirmed pressures. by two consecutive readings at least 6 h apart; pro- teinuria Ͼ0.300 g/dei. Spectral analysis and baroreflex gain The PE women were hospitalised because of an increase in blood pressure in the third trimester; A stationary sequence of 5 min (about 300 to 500 none of these received antihypertensive medication. data points), for left lateral recumbent and upright The patients were normotensive before the actual position, was analysed.15 Mean values of systolic pregnancy and during eventual previous pregnancy. and diastolic blood pressure, and interbeat interval Normotensive pregnant and all normotensive non- were calculated. Power spectral analysis of these pregnant women were members of the medical staff. parameters was performed in the frequency domain None of these women had histories of hypertension, by an autoregressive method (Burg method) using cardiovascular or renal disease, or were taking anti- custom software implemented on MATLAB pro- hypertensive medication. gram, after linear interpolation of data at 1 Hz fre- There was no statistical difference of the body quency (to obtain equispaced data) and detrending. mass index (BMI) in all studied patients before preg- The order of the autoregressive method was chosen nancy. All pregnant women gave their informed by Akaike criterion for values Ͼ15, whereas it was consent to participate in the study, that was considered =15 if lower than 15. approved by our Ethical Commitee. The total power was divided into low frequency (0.025–0.14 Hz) and high frequency (0.14–0.4 Hz) components, markers respectively of sympathetic Protocol and signal recordings and parasympathetic activity. The power is The study was performed in a quiet room between expressed in normalised units, which are obtained 2 pm and 5 pm, at least 2 hours after a meal. At first by dividing the power of each component by total the subjects were studied in the left lateral recum- variance from which direct current has been sub- bent position: after 20 min of bed rest, continuous tracted, and multiplying this value by 100.15 finger blood pressure was recorded for 15 min by To assess the statistical link between heart rate Portapres Model-2 device (TNO, Amsterdam, NL) at and systolic blood pressure series at any given fre- 200 Hz sampling frequency. Then the procedure was quency, coherence was measured. Cross-spectral repeated, using the same finger, during 10 min of analysis was performed by means of bivariate auto- orthostatic position. regressive identification and was used to compute Portapres is a portable, battery-operated device for a squared coherence function (ie, the square cross- the continuous, non-invasive, 24-h ambulatory spectrum amplitude normalised by the product of measurement of blood pressure in the finger.10 It is the spectra of the two signals). Coherence is based on the Finapres, a stationary device for the expressed as a number between 0 and 1 and only measurement of finger arterial pressure on a beat-to- values Ͼ0.5 were considered significant. In our

Table 1 Clinical data of the studied groups

Non-pregnant Healthy pregnant Pre-eclamptic pregnant (n = 7) (n = 8) (n = 9)

Age (years) 27 (26–33) 31 (30–34) 32 (29–33) BMI (kg/m2) 20.43 (19.69–22.10) 27.23 (24.14–31.21)* 27.11 (26.56–30.29)* Week of gestation — 32.5 (28.5–36.5) 35.0 (32.0–36.0) Proteinuria (g/d) 0 0 2.4 (1.8–3.4) Primigravida (No. of women) — 6 7 Systolic blood pressure at rest (mm Hg) 118.30 (112.60–124.58) 109.85 (95.15–117.22) 151.80 (142.00–154.90)*† Diastolic blood pressure at rest (mm Hg) 61.23 (51.40–67.80) 56.41 (48.11–64.44) 94.21 (83.38–100.06)*† Resting heart rate (bpm) 69.9 (64.46–72.74) 87.35 (84.11–89.59)* 88.67 (77.60–89.50)*

Data are expressed as Median (25th centile–75th centile). *P Ͻ 0.05 vs non-pregnant, †P Ͻ 0.05 vs healthy pregnant by Mann– Whitney test. Baroreflex in pre-eclampsia P Molino et al 181 patients coherence was always Ͼ0.5, either in rest towards lower ␣-index values in the healthy preg- and upright position. nant group at rest compared to non-pregnant The baroreflex gain was computed as alpha index, women. The ␣-index decreased in all groups after ie the square root of the ratio of the powers of standing, but this reduction was significant only in interbeat interval in the low frequency range to cor- non-pregnant and normal pregnant women. In all responding spectral components of systolic blood pregnant women the ␣-index in the upright position pressure, if coherence was Ͼ0.5: was significantly lower than in non-pregnant ␣-index = women. ͱ low frequencyinterbeat interval/low frequencysystolic blood pressure Spectral components of IBI variability Low and high frequency, expressed in normalised Statistical analysis units, did not differ between the groups, in the Data are expressed as median (25th–75th centile). resting position. Statistical evaluation was performed by Statistical In orthostatic position, all pregnant women Analysis System (SAS Institute Inc, Cary, NC, USA). showed significantly lower low frequency values For comparison of unpaired data, a non-parametric than non-pregnant women, whereas high frequency analysis of variance (Kruskal–Wallis test) was first were significantly higher compared to non-pregnant used followed by the Mann–Whitney test in case of women. Low frequency increased and high fre- significance. For comparison of paired data, a Stud- quency decreased significantly after standing only ent’s t-test for paired data was used. Differences in non-pregnant women. were considered significant at P Ͻ 0.05. The LF/HF ratio was similar in all groups at rest; it significantly increased after standing only in non- Results pregnant women, whereas it was unchanged com- pared to the resting position in all pregnant groups Main clinical and haemodynamic parameters are and significantly lower compared to non-pregnant shown in Table 1. Systolic and diastolic blood press- women. ure were significantly higher in PE women, com- pared to non-pregnant and normal pregnant women. Heart rate was significantly higher in normal preg- Spectral components of blood pressure variability nant and PE women compared to non-pregnant A comparison between the non-pregnant women women. BMI was significantly lower in non-preg- and the two pregnant groups, showed no significant nant women compared to other groups. differences in systolic and diastolic blood pressure Table 2 summarises data on baroreflex gain and variability with respect to high frequency or low fre- spectral analysis of pulse interval and blood press- quency bands, both in the recumbent and ortho- ure variability. static position. Baroreflex gain Discussion The ␣-index was significantly lower in PE women at rest if compared both to normal pregnant and non- The major finding of the present study is the pregnant women (Table 2). There was a trend decrease of the baroreflex gain, measured as ␣-index,

Table 2 Baroreflex gain and power spectral analysis of interbeat interval and blood pressure variabilty

Non-pregnant (n = 7) Healthy pregnant (n = 8) Pre-eclamptic pregnant (n = 9)

Rest Standing Rest Standing Rest Standing

BRG (␣-index) (ms/mmHg) 12.16 (9.02–19.17) 10.22 (8.21–15.36)‡ 7.98 (6.71–9.93) 5.70 (5.24–7.18)*‡ 5.60 (5.25–6.90)*† 4.07 (3.70–6.92)*

IBI variability LF power (ms2) 1235 (754–1799) 3181 (1664–6052) 388 (268–620) 686 (310–1016)* 488 (361–769) 690 (455–715)* HF power (ms2) 906 (194–1084) 346 (290–462) 371 (155–492) 323 (170–547) 194 (101–486) 336 (281–456) LF (NU) 0.62 (0.50–0.76) 0.89 (0.69–0.91)‡ 0.48 (0.35–0.76) 0.62 (0.47–0.76)* 0.62 (0.47–0.69) 0.56 (0.48–0.66)* HF (NU) 0.35 (0.22–0.48) 0.08 (0.06–0.19)‡ 0.31 (0.16–0.56) 0.29 (0.15–0.45)* 0.30 (0.23–0.40) 0.27 (0.27–0.45)* LF/HF 1.80 (1.05–3.42) 10.94 (3.78–13.27)‡ 1.53 (0.63–5.08) 2.23 (1.12–5.03)* 2.03 (1.14–2.88) 2.04 (1.04–2.44)*

SBP variability LF power (mm Hg2) 6.9 (3.2–15.9) 24.6 (17.7–29.5) 8.2 (6.4–13.4) 14.1 (11.6–23.7) 14.0 (9.6–16.1) 37.3 (29.7–41.8)‡ HF power (mm Hg2) 1.5 (1.0–2.5) 4.3 (3.5–11.1)‡ 2.6 (1.5–3.6) 7.0 (3.5–10.5) 1.5 (1.1–3.8) 7.9 (5.1–9.1)‡

DBP variability LF power (mm Hg2) 7.5 (3.8–8.3) 13.0 (5.1–14.3) 2.9 (2.5–3.7) 6.7 (4.6–9.8) 4.8 (32.–6.9) 15.1 (10.8–15.3)‡ HF power (mm Hg2) 0.7 (0.2–1.0) 2.4 (1.1–2.9)‡ 0.9 (0.6–1.4) 1.3 (1.1–3.3) 1.1 (0.9–1.5) 3.0 (2.1–3.9)‡

Data are expressed as Median (25th centile–75 centile). *P Ͻ 0.05 vs non-pregnants, †P Ͻ 0.05 vs healthy pregnants by Mann–Whitney test; ‡P Ͻ 0.05 vs rest by Student’s t-test. BRG: baroreflex gain; IBI: interbeat interval; SBP: systolic blood pressure; DBP: diastolic blood pressure; HF: high frequencies; LF: low frequencies; NU: normalised Units. Baroreflex in pre-eclampsia P Molino et al 182 in pre-eclampsia, compared to non-pregnant and quence rather than a precursor of severe pre-eclamp- healthy pregnant women. Furthermore, in pregnant sia, which cannot be demonstrated as an impair- groups we showed a reduction of the baroreflex gain ment of spectral components in early stages of the in the orthostatic position compared to non-preg- disease. This could explain why our patients nant women. These results provide evidence that showed an impaired sympathovagal balance only pregnancy per se is associated with a decrease of the after standing: this simulus might point out an alter- baroreflex sensitivity, but this is further impaired in ation which is not still evident in basal conditions. pre-eclamptic subjects. Previous attempts to study autonomic nervous Several studies concerning baroreflex function in system function by means of plasma and urinary cat- pregnant women reported conflicting results.20–23 echolamine concentrations did not give conclusive On the contrary, many animal studies24,25 showed results.23,30,31 A substantial increase in sympathetic pregnancy to be associated with an impaired barore- vasoconstrictor activity has been found by Schobel flex sensitivity, and an attenuated ability to increase et al,1 measuring the postganglionic sympathetic- sympathetic outflow above baseline levels during a nerve activity in the blood vessels of skeletal mus- hypotensive challenge. cle. Since five of the nine pre-eclamptic patients On the other hand, it is known that in hyperten- received dihydralazine at the time of the study, we sive humans and animals there is a diminished baro- have to assume that pre-eclampsia was at a greater reflex control of the heart,26 which may be due to a level of severity than in our patients. Moreover, the reduction in the vagal range.27 Furthermore, endo- use of a direct vasodilator per se may be a stimulus thelial dysfunction and altered release of paracrine to enhance sympathetic activity. factors contribute significantly to the decreased The method we used in this study allows the sensitivity in hypertension and athero- dynamic assessment of the baroreceptor heart rate 28 sclerosis. Previous studies have described a vascu- reflex, and evidence in both animals and humans lar endothelial dysfunction in pre-eclampsia, with indicates that this reflex function may not always decreased vasodilatatory prostacyclin and increased reflect baroreceptor control of peripheral circulation 7 vasoconstrictor thromboxane synthesis. In and blood pressure.32,33 addition, there is evidence for placental release of Also in normal pregnancy, several human and circulating substances able to alter endothelial 8 animal studies have demonstrated an impairement cells, with a reduced production of vasodilatatory of both vagal and sympathetic functions, which autocrine and paracrine factors such as nitric oxide 34 29 have been found to be variably altered. and prostacyclin. Hence a loss of the excitatory In our study, non-pregnant women showed an influence of endogenous prostanoids in pre-eclamp- increase in low frequency and a decrease in high tic women could represent a possible mechanism for frequency components of with the impaired baroreflex gain found in our study. standing indicating a stimulation of the sympathetic Studies concerning autonomic nervous function nervous system, whereas in all pregnant women the in pre-eclampsia have reported inconclusive results. increase was not significant. Moreover, in ortho- A study which used spectral analysis showed a sig- static position heart rate and diastolic blood press- nificant reduction of high frequency of interbeat ure variability of healthy pregnants were lower com- interval variability, indicating that pre-eclampsia is associated with decreased vagal control of the pared to non-pregnant women, confirming the 2 4 possibility of changes in the baroreflex control and heart. On the contrary, Ekholm et al showed 20 increased heart rate and blood pressure variability increased maternal to be involved. in women with pregnancy-induced hypertension Nevertheless, in our study, normal pregnant and and proteinuria, suggesting increased of both sym- pre-eclamptic women showed significantly pathetic and parasympathetic components. The dis- increased heart rate with standing, also if the cordant results could be due both to the different increase was lower compared to non-pregnant conditions of breathing (spontaneous and controlled women (9.8% and 13.8%, vs 21.9%, respectively). breathing respectively) and to the different order of In conclusion, the innovative aspect of our study the autoregressive method chosed for the spectral was the assessment of the baroreflex function in nor- analysis; besides the two groups of pre-eclamptic mal pregnancy compared to pre-eclamptic women, ␣ women studied showed unlike values of resting by a non-invasive method, such as -index calcu- heart rate. lated on finger blood pressure measurement. This In our study no differences between the groups method provides a global quantification of the spon- were present at rest regarding blood pressure and taneous baroreflex control of heart rate, avoiding heart rate variability. When the orthostatic position most of the limitations of pharmacological methods, was assumed pre-eclamptic women showed a and requiring little subject cooperation or stress. reduction in the low-frequency and an increase in Therefore, it is of great utility in patients such as the high-frequency components of the heart rate pre-eclamptic women, in which the avoidance of variability and the LF/HF ratio was also reduced stress is an important part of the therapy. compared to non-pregnant women. Finally, we think that longitudinal studies will be Airaksinen et al3 showed a decreased resting vagal useful to evaluate if the baroreflex function is activity in normal pregnancy, and only in severe restored after delivery and if it could provide prog- pre-eclampsia were both the sympathetic and para- nostic information to enable the early identification sympathetic reflexes impaired. The authors suggest of pregnant women at greater or probable risk to that autonomic nervous dysfunction is a conse- develop pre-eclampsia. Baroreflex in pre-eclampsia P Molino et al 183 Study limitations ment, physiological interpretation, and clinical use. Circulation 1996; 93: 1043–1065. This study has some potential limitations: although 16 Davey DA, MacGillivray I. The classification and pletysmographic finger blood pressure recording has definition of the hypertensive disorders of pregnancy. been used for a non-invasive determination of spon- Am J Ostet Gynecol 1988; 158: 892–898. taneous cardiac baroreflex,10,18,19 actually this meth- 17 Penaz J. Photoelectric measurement of blood pressure, volume and flow in the finger. Digest 10th Int Conf odology has not been formally cross-validated Med Bio Engen, 1973, Dresden, pp 104. against invasive techniques in pregnancy; besides 18 Parati G et al. Comparison of finger and intrabrachial our study was conducted on a relatively small num- blood pressure monitoring in rest and during labora- ber of patients, therefore further study are required tory tests. Hypertension 1989; 13: 647–655. to better define the results found. 19 Imholz BPM et al. Non-invasive continuous finger blood pressure measurement during orthostatic stress compared to intrabrachial pressure. Cardiovasc Res 1990; 24: 214–221. References 20 Ekholm EMK, Piha SJ, Erkkola RU, Antila KJ. Auto- nomic cardiovascular reflexes in pregnancy. A longi- 1 Schobel HP et al. Preeclampsia—A state of sympath- tudinal study. Clin Autonom Res 1994; 4: 161–165. etic overactivity. N Engl J Med 1996; 335: 1480–1485. 21 Ramsay MM, Broughton Pipkin F, Rubin PC. Pressor, 2 Eneroth-Grimfors E et al. Autonomic cardiovascular heart rate and plasma catecholamine responses to nor- control in normal and pre-eclamptic pregnancy. Acta adrenaline in pregnant and non-pregnant women. Br J Obstet Gynecol Scand 1994; 73: 680–684. Obstet Gynaecol 1993; 100: 170–176. 3 Airaksinen KEJ, Kirkinen P, Takkunen JT. Autonomic 22 Castren O, Saarikoski S, Siimes A. Vascular reactivity nervous dysfunction in severe pre-eclampsia. Euro J in normal pregnancy and in pregnancies complicated Obstet Gynecol Reprod Biol 1985; 19: 269–276. by hypertension. Int J Gynaecol Obstet 1973; 11: 4 Ekholm EMK, Tahvanainen KUO, Metsala T. Heart rate 236–246. and blood pressure variabilities are increased in preg- 23 Nissel H, Hjemdahl P, Linde B, Lunell N-O. Sympa- nancy-induced hypertension. Am J Ostet Gynecol thoadrenal and cardiovascular reactivity in pregnancy- 1997; 177: 1208–1212. induced hypertension. II. Responses to tilting. Am J 5 Robbe HWJ et al. Assessment of baroreceptor reflex Obstet Gynecol 1985; 152: 554–560. sensitivity by means of spectral analysis. Hypertension 24 Quesnell RR, Brooks VL. Alterations in the baroreflex 1987; 10: 538–543. occur late in pregnancy in conscious rabbits. Am J 6 Parati G, Saul JP, Di Rienzo M, Mancia G. Spectral Obstet Gynecol 1997; 176: 692–694. analysis of blood pressure and heart rate variability in 25 Heesch CM, Rogers RC. Effects of pregnancy and pro- evaluating cardiovascular regulation. A critical gesterone metabolites on regulation of sympathetic appraisal. Hypertension 1995; 25: 1276–1286. outflow. Clin Exp Pharmacol Physiol 1995; 22: 136– 7 Wasserstrum N et al. Implications of sino-aortic baro- 142. receptor reflex dysfunction in severe preeclampsia. 26 Gribbin B, Pickering TG, Sleight P, Smyth HS. Effect Obstet Gynecol 1989; 74: 34–39. of age and high blood pressure on baroreflex sensi- 8 Speroff L. Toxiemia of pregnancy. Mechanism and tivity in man. Circulation Res 1971; 29: 424–431. therapeutic menagement. Am J Cardiol 1973; 32: 27 Head GA. Baroreflex and cardiovascular regulation in 582–586. hypertension. J Cardiovasc Pharmacol 1995; 26 9 Roberts JM, Redman CWG. Preeclampsia: more than (Suppl 2): S7–S16. pregnancy-induced hypertension. Lancet 1993; 341: 28 Chapleau MW et al. Structural versus functional 1447–1451. modulation of the arterial baroreflex. Hypertension 10 Imholz BPM et al. Feasibility of ambulatory, continu- 1995; 26: 341–347. ous 24-hour finger arterial pressure recording. Hyper- 29 McCarthy AL, Raju KS, Poston L. Endothelium depen- tension 1993; 21: 65–73. dent relaxation in preeclampsia [Abstract]. J Hypertens 11 Akselrod S et al. Power spectrum analysis of heart rate 1992; 10 (Suppl): S45. fluctuations:a quantitative probe of beat-to-beat cardio- 30 Zuspan FP. Catecholamines: their role in pregnancy vascular control. Science 1981; 213: 220–223. and the development of pregnancy-induced hyperten- 12 Malliani A, Pagani M, Lombardi F. Methods for assess- sion. J Reprod Med 1979; 23: 143–150. ment of sympatho-vagal balance:power spectral analy- 31 Davey DA, Macnab MF. Plasma adrenaline, noradrena- sis. In: Levy MN, Schwartz PJ (eds). Vagal control of line, dopamine in pregnancy hypertension. Br J Obstet the heart: experimental basis and clinical impli- Gynaecol 1981; 88: 611–618. cations. Futura Publishing Company: New York, 1994, 32 Mancia G, Mark AL. Arterial baroreflexes in humans. pp 433–454. In: Shepherd JT, Abboud FM (eds). Handbook of 13 Malliani A, Pagani M, Lombardi F. Physiology and Hypertension. The Cardiovascular System IV. Vol 3, clinical implications of variability of cardiovascular sect 2. American Physiological Society: Washington, parameters whith focus on heart rate and blood press- DC, pp 759–794. ure. Am J Cardiol 1994; 73: 3C–9C. 33 Guo KL, Thames MD, Abboud FM. Preservation of 14 Pomeranz B et al. Assessment of autonomic function baroreflex control in lumbar sympathetic nerve in humans by heart rate spectral analysis. Am J Physiol activity in renal hypertensive rabbits. Clin Res 1981; 1985; 248: H151–H153. 29: 751A. Abstract. 15 Task Force of the European Society of Cardiology and 34 Ekholm EM et al. Changes in autonomic cardiovascu- the North American Society of Pacing and Electrophy- lar control in mid-pregnancy. Clin Physiol 1992; 12: siology. Heart rate variability: standards of measure- 527–536.