Citation for published version (APA): Lefrandt, J. D. (2006). Autonomic dysfunction in cardiovascular disease. s.n.

Transcription

1 University of Groningen Autonomic dysfunction in cardiovascular disease Lefrandt, Johan Daniël IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below. Document Version Publisher's PDF, also known as Version of record Publication date: 2006 Link to publication in University of Groningen/UMCG research database Citation for published version (APA): Lefrandt, J. D. (2006). Autonomic dysfunction in cardiovascular disease. s.n. Copyright Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons). Take-down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Downloaded from the University of Groningen/UMCG research database (Pure): For technical reasons the number of authors shown on this cover page is limited to 10 maximum. Download date:

2 Autonomic Dysfunction in Cardiovascular Disease Johan Daniël Lefrandt i

3 Financial support by the Dutch Diabetes Research Foundation and the Netherlands Heart Foundation for the publication of this thesis is gratefully acknowledged. The Groningen Hypertension Service is gratefullly acknowledged for there contribution. The printing of this thesis was financially supported by: Boehringer Ingelheim, University of Groningen, NovoNordisk, Bristol-Myers Squibb, MSD, GlaxoSmithKline, Pfizer, Abbo, Astellas Pharma B.V. Nederland, AstraZeneca, Bayer HealthCare, Sanofi Aventis, AMGEN, Novartis, Servier Nederland Farma B.V., Zambon, Huntleigh Healthcare BV Medilog. Copyright 2006 J.D. Lefrandt All rights reserved. No part of this publication may be reproduced, or transmi ed in any form or by any means without wri en permission of the author and the publisher holding the copyright of the published articles. Cover art: Louis Royo, The Vampire Sexte e. Cover design: Taco de Jong, Groningen. Layout: Taco de Jong, Groningen. Printed by: Faciltair Bedrijf RUG, Groningen. ii

4 The artwork of the front cover refers to the VAMPHYRE study, an international multicenter study that supplied data to a considerable part of this thesis. The VAMPHYRE is the acronym of this study: effects on autonomic function of Verapamil SR versus AMlodipine in Patients with mild to moderate HYpertension at Rest and during Exercise. The vampire bat on the back cover is put in an erythrocyte, a red blood cell, that are snacks for vampires. The vampire bats (Desmodus rotundus) have a wingspan of about eight inches and a body about the size of an adult s thumb. If not for their diet, people would not pay much a ention to these small bats. Vampire bats feed on the blood of large birds, ca le, horses, and pigs. However, they don t suck the blood of their victims. Using their sharp teeth, the bats make tiny cuts in the skin of a sleeping animal. The bats saliva contains a chemical that keeps the blood from clo ing. The bats then lap up the blood that oozes from the wound. Another chemical in their saliva numbs the animal s skin and keeps them from waking up. A vampire bat finds its prey with echolocation, smell, and sound. They fly about one meter above the ground. Then they use special heat sensors in their noses to find veins that are close to the skin. Scientists have discovered that vampire bat saliva is be er at keeping blood from clo ing than any known medicine. Vampire bats may one day help prevent heart a acks and strokes. Vampire bats are one of the few bat species that are considered a pest. In Latin America, ca le raising is a growing business, and sleeping ca le a ract vampire bats. In ranching areas, control programs have been started. However, millions of beneficial bats are destroyed by people who mistake them for vampires. (From: h p:// iii

5 iv

6 RIJKSUNIVERSITEIT GRONINGEN Autonomic Dysfunction in Cardiovascular Disease Proefschri ter verkrijging van het doctoraat in de Medische Wetenschappen aan de Rijksuniversiteit Groningen op gezag van de Rector Magnificus, dr. F. Zwarts, in het openbaar te verdedigen op woensdag 21 juni 2006 om uur door Johan Daniël Lefrandt geboren op 13 augustus 1970 te Groningen v

7 Promotor: Prof. dr. R.O.B. Gans Copromotores: Dr. A.J. Smit Dr. K.H. Hoogenberg Beoordelingscommissie: Prof. dr. P.E. de Jong Prof. dr. P.A. de Graeff Prof. dr. D.J. van Veldhuisen Prof. dr. P.W. de Leeuw ISBN (printed edition): ISBN (electronic edition): vi

8 Voor: Mijn Moeder, Theodora Hendrika Lefrandt Meijring (Thea), Omdat haar kinderen haar levenswerk waren, Mijn Vader, Johan Daniël Lefrandt (Joop), Voor de goede start die ik heb meegekregen, Mijn Zussen, Jeanne e Lefrandt en Daniëlle Hoeberechts Lefrandt, Voor het zijn van wie ze zijn, Mijn Vrouw en Kinderen, Astrid Lefrandt Engels, Roos Lefrandt en... Omdat zij het allemaal de moeite waard maken, en het Grootste Geluk in mijn leven zijn. vii

9 Paranimfen: Ymkje Stienstra Peter Luik viii

10 Contents Preface 7 Part I: Aims of the thesis and introduction Chapter 1 11 Aims of the thesis Chapter 2 15 Autonomic Dysfunction in Cardiovascular Disease Submi ed Part II: Diabetes mellitus Chapter 3 51 Baroreflex sensitivity is depressed in Microalbuminuric Type 1 diabetic patients at rest and during sympathetic manoeuvres Diabetologia. 1999;42: Chapter 4 65 Heart rate variability and baroreflex sensitivity in diabetic patients with peripheral neuropathy are indices of both neural and vascular disease. Submi ed. Chapter 5 81 Sympathetic mediated vasomotion and skin capillary permeability in diabetic patients with peripheral neuropathy Diabetologia. 2003;46:40 47

11 Contents 2

12 Contents Part III: Hypertension Chapter Autonomic Function in Hypertensive and Normotensive Subjects Hypertension June;37: Chapter The Effects of Dihydropyridine and Phenylalkylamine Calcium Antagonist Classes on Autonomic Function in Hypertension: The VAMPHYRE Study American Journal of Hypertension Nov;14: Chapter Contrasting effects of verapamil and amlodipine on cardiovascular stress responses in hypertension Britisch Journal of Clinical Pharmacology Dec;52: Chapter Less Adrenergic Response to Mental Task During Verapamil Compared to Amlodipine Treatment in Hypertensive Subjects Blood Pressure Oct;10:111 5 Part IV: For debate Chapter High fasting glucose and QTc duration in a large healthy cohort (Le er) Diabetologia : Chapter Inverse Relationship Between Blood Glucose and Autonomic Function in Healthy Subjects (Le er) Diabetes Care Dec;23:

13 Contents 4

14 Contents Chapter Glycemic Control and Impaired Autonomic Function (Le er: response of Watkins et al. to chapter 10.2) Diabetes Care Dec;23: Chapter Acute Hyperglycemia and Autonomic Function Diabetes Care Nov;24: Chapter Relation Between Autonomic Function and Blood Glucose in the Nondiabetic Range. (Le er: response to Marfella et al. chapter 10.4) Diabetes Care Nov;24:2017 Part V: Summary and future perspectives Chapter Summary Chapter Nederlandse samenva ing Dankwoord 219 Curriculum vitae 225 Bibliography 227 5

15 Contents 6

16 Preface

17 Preface In 1976, D.J. Ewing, I.W. Campbell and B.F. Clarke published a paper in the Lancet, entitled: Mortality in diabetic autonomic neuropathy 1. In this paper, a number of bedside tests based on heart rate and blood pressure variations was performed to diagnose autonomic neuropathy in diabetic patients. A clear survival disadvantage was shown for the patients with autonomic neuropathy. This thesis acknowledges the importance of Ewing s bedside tests and their predictive value for mortality. Yet, two essentials of this thesis challenge Ewing s findings. First, modern techniques of spectral analysis of cardiovascular parameters allow more accurate and reliable evaluation of the cardiovascular autonomic function. There is an ongoing implementation of these modern techniques of evaluating cardiovascular autonomic function in health and disease, to which this thesis contributes. Second, what Ewing really tested with his bedside tests was the integrity of a series of components that make up the cardiovascular autonomic function. Vascular, cardiac and neural properties contribute to the performance of what is called the cardiovascular autonomic nervous system. Did Ewing show a survival disadvantage for diabetic patients with autonomic neuropathy or were the poor test results in these patients a reflection of damage to the blood vessels and the heart? This thesis implements modern techniques that assess the cardiovascular autonomic function and examines the determinants of cardiovascular autonomic function in health and in cardiovascular disease. This is done by studying (1) the physiology in healthy subjects, (2) the pathofysiological changes in hypertension and diabetes and (3) the effects of pharmacological and physical interventions. J.D. Lefrandt. Reference 1. Ewing DJ, Campbell IW, Clarke BF: Mortality in diabetic autonomic neuropathy. Lancet 1: ,

18 Part I: Aims of the thesis and introduction

19

20 Chapter 1 Aims of the thesis

21 Chapter 1 The general aims of this thesis 1. to implement (cross )spectral analysis of heart rate, blood pressure and capillary blood flowmetry as means of assessing cardiovascular autonomic function in a clinical se ing, 2. to study the abnormalities of cardiovascular autonomic function in diabetes mellitus, as assessed by the above (at 1.) mentioned techniques, in relation to vascular and neural complications of diabetes mellitus, 3. to study the abnormalities of cardiovascular autonomic function in hypertension, as assessed by the above (at 1.) mentioned techniques, and the influence of pharmacological and physical interventions, 4. to discuss the interplay between autonomic function and (micro-) vascular properties in health and cardiovascular disease. This will be done by studying cardiovascular autonomic function in health (chapters 6 and 10), in cardiovascular disease (diabetes: chapters 3 to 5, hypertension: chapters 6 to 9) and during interventions by pharmacological (chapters 7 to 9) and physical (chapters 3, 8 and 9) means. 12

22 Aims of the thesis More specifically, the following issues will be addressed in this thesis Chapter 2 reviews the physiology of autonomic control of blood pressure and heart, the techniques to assess autonomic function, the structural and functional changes of the autonomic nervous system in cardiovascular disease, the predictive value of autonomic dysfunction for mortality in health and cardiovascular disease and the interdependence of autonomic function and (micro ) vascular properties. Chapter 3 describes the early changes in autonomic function at rest and during sympathetic maneuvers, assessed by analysis of heart rate variability (HRV) and baroreflex sensitivity (BRS), in diabetic patients complicated by microalbuminuria. Chapter 4 examines the determinants of cardiovascular autonomic function in diabetic patients with peripheral neuropathy. Chapter 5 relates functional properties of the autonomic nervous system to physical characteristics of blood vessels by studying relation between sympathetic mediated vasomotion and skin capillary permeability in diabetic patients with peripheral neuropathy. Chapter 6 reports the abnormalities in HRV and BRS in hypertensive subjects compared with healthy subjects and the gender specificity of these abnormalities. Chapter 7 describes the effects of pharmacological modulation of autonomic function, assessed by HRV, BRS, and plasma catecholamines, by two different classes of calcium channel blockers in hypertension in a multicenter trial (the VAMPHYRE study). Chapter 8 is a report of the VAMPHYRE study, with particular interest in the effects of the two calcium channel blockers on cardiovascular stress responses in the hypertensive subjects. Chapter 9 is substudy of the VAMPHYRE study, that investigates the effects of the two calcium channel blockers on adrenergic responses during mental tasks. Chapter 10 concerns correspondence with fellow scientists on the relation between blood glucose and autonomic function in the non diabetic range, published as le ers. Chapter 10.1 is a le er we wrote in response to a report of Marfella et al. that described the increase in QTc duration during a hyperglycaemic clamp in healthy men. Chapter 10.1 shows support for this association between QTc and plasma glucose 13

23 Chapter 1 in a population based study of 6543 non diabetic subjects (data of the PREVEND study).chapter 10.2 is a report on the inverse relationship between blood glucose and autonomic function, assessed by BRS, in healthy subjects. Chapter 10.3 is the first response to chapter 10.2, where Watkins et al. raises the question if the observed relation is an effect of glucose or insulin. Chapter 10.4 is the second response to the le er of chapter 10.2, where Marfella et al. discuss the putative mechanisms of the relation between BRS and blood glucose. Chapter 10.5 is our reponse to Marfella s response of chapter 10.2, further debating the underlying mechanisms of the relationship between autonomic function and blood glucose in the non diabetic range. Chapter 11 is a summary of this thesis and discusses the conclusions of all chapters in perspective of the above stated aims. Furthermore, the lines of future research arising from the results and the thereby newly generated questions are considered. Chapter 12 is a Dutch translation of chapter

24 Chapter 2 Autonomic Dysfunction in Cardiovascular Disease A consequence of vascular damage Submi ed J.D. Lefrandt 1, A.J. Smit 1, K. H. Hoogenberg 2, R.O.B. Gans 1. 1 University Hospital Groningen, department of Internal Medicine and 2 Martini Hospital Groningen, The Netherlands

25 Chapter 2 Introduction The beat of our heart and our blood pressure are not constant. They vary with respiration, physical and mental stress and are controlled by the autonomic nervous system. In 1876, Mayer described periodical fluctuations in blood pressure with a cycle of once per 10 seconds 1. These Mayer waves are considered markers of sympathetic control of vasomotor tone 2. Heart rate variations with a period shorter than 3 5 seconds are reflections of respiration and vagal modulation, whereas variations of once per 10 seconds represent vagal as well as sympathetic cardiac modulation 3;4. After pharmacological denervation of the heart with atropine and propranolol, nearly all variability of the heart rate is abolished and the intrinsic heart rate of 100 beats per minute reveals in humans 5. The clinical relevance of a diminished heart rate variability was first recognized in 1965 when Hon and Lee perceived that fetal death was preceded by vanishing of heart rate variability. In the 70s, Ewing developed a set of bedside tests to diagnose diabetic cardiovascular autonomic neuropathy and showed that diminished blood pressure and heart rate responses to standard stimuli were predictive for mortality 6. Modern evaluation of cardiovascular autonomic function by analysis of heart rate variability (HRV) and baroreceptor reflex sensitivity (BRS) have proven their value for risk stratification in chronic heart failure and after myocardial infarction 7;8. How should autonomic function be measured? A variety of tests is available, that depend on the integrity of different parts of the aut o- nomic nervous system. The baroreceptors, located in the carotid artery that has its specific vascular properties, the afferent, central, and efferent pathways and the effector organs (heart, blood vessels an adrenal glands) all have their specific contribution to the auto nomic function tests results, that reflect the function of different parts of the autonomic nervous system. Similarly, the factors that are responsible for the impaired autonomic function in cardiovascular disease may be multiple and may differ between diseases. Ewing a ributed abnormal autonomic function tests to diabetic autonomic neuropathy. However, today s evidence shows a strong association and probably dependence of autonomic function on structural and functional (cardio ) vascular properties. Not only in chronic heart failure, where impaired auto- 16

26 Autonomic Dysfunction in Cardiovascular Disease. A consequence of vascular damage nomic control of heart rate is the result from both end organ damage and neurohumoral activation. But also in diabetes the (subclinical) cardiac, micro and macrovascular or endothelial dysfunction may be an important cause of a failing autonomic control. Viceversa, early subclinical autonomic neuropathy is strongly related to insulin resistance and probably precedes development of overt diabetes mellitus. We will review the methods to assess autonomic function, the evidence on the predictive value of impaired autonomic function for mortality and the possible mechanisms involved, and summarize the determinants of autonomic function including recent reports on role of nitric oxide (NO) as a essential modulator of autonomic control. Blood pressure control Our blood pressure is kept between narrow limits. Figure 1 denotes the blood pressure controlling reponses tot a sudden change in blood pressure. Of the short term systems, the baroreceptors respond within seconds.the baroreceptors have to long been thought to lose their effects after hours, after which the baroreflex is reset to a new blood pressure level 9;10. However this has more recently been challenged by observations that damage to baroreceptors for example after bilateral carotid body tumor resection or after radiotherapy has long term effects on heart rate or blood pressure control, especially variability 11. More importantly, Lohmeier et al found that prolonged activation of the baroreflex in dogs led to a sustained fall in blood pressure, and heart rate, also associated with a fall in plama norepinephrine levels 12. Ambulatory blood pressure monitoring has shown that blood pressure tends to be highest during the day and lowest during the night 13;14. Individuals, whose blood pressure does not fall or falls scarcely at night, are called non dippers 15. The mechanism of the non dipping nocturnal pa ern is not well understood. An impairment in autonomic nervous activity may play an important role since under physiological conditions, the heart rate, cardiac output, peripheral resistance and plasma catecholamines are reduced during the night 16;17. Withdrawal of sympathetic activity seems to play a pivotal role. Conditions with abnormal autonomic function, such as diabetes, heart failure and car- 17

27 Chapter 2 diac transplantation are associated with a lack or a enuation of the day night blood pressure changes In hypertension, a non dipping nocturnal blood pressure has been associated with increased cardiovascular morbidity, possibly resulting from the higher haemodynamic load throughout the 24 hours 13; Cardiovascular reflexes Both arterial and cardiopulmonary baroreceptors exert a tonic inhibitory influence on centrally mediated sympathetic neural outflow. The arterial baroreceptors are located in all thoracic large arteries, but are strongly concentrated in the carotid artery and the aorta. Afferent nerves supply the cardiovascular control center in the medulla oblongata. Efferent nerves run to the atria and ventricles, peripheral arterioles and veins and adrenal glands 24. The aim of the baroreflex is to keep blood pressure constant. A sudden rise in blood pressure will increase the inhibition of sympathetic outflow and stimulate vagal outflow, resulting in a decrease of heart rate, myocardial contractility and peripheral resistance (figure 2). The cardiopulmonary baroreceptors are located in the heart (atria and ventricles), lung, and great veins and are activated by increases in cardiac filling pressure, contractility, wall stress and the depth of breathing 24. The recent studies of Lohmeier et al have challenged some of our views on the function of the baroreflex: prolonged stimulation of the carotid baroreceptors was found to be sufficient to lead to a prolonged fall in blood pressure. The technique used to stimulate the carotid baroreceptors suggests that this effect goes beyond that of stretch of the vascular receptors in the wall of the carotid. Moreover, in the effector arm of the reflex indirect effects of the renal nerve on the renin angiotensin aldosterone system seem to be involved 12. In this review, we will focus on the arterial baroreceptor reflex. Techniques to assess the autonomic function Ewing ba ery. The Ewing ba ery consists of five parameters derived 18

28 Autonomic Dysfunction in Cardiovascular Disease. A consequence of vascular damage from four tests evaluating the heart rate responses to the Valsalva maneuver, deep breathing and standing up from the supine position and the blood pressure responses to standing up and sustained handgrip 25. These tests are still considered golden standard to diagnose diabetic neuropathy and have proven their predictive value for mortality in diabetes 26 (figure 3). However, the results are strongly dependent upon performance of the maneuvers and are age dependent. Heart rate variability (HRV). Analysis of heart rate variability of short (minutes) as well as long (24 to 48 hours) ECG recordings has been used as a measure of cardiac autonomic control. The time domain approach consists of statistical analysis of the duration of the RR interval lengths and the differences of successive RR intervals 27. A number of time domain variables is available, including the SDNN, the simplest variable, defined as the standard deviation of all normal RR intervals. Since it reflects all variance it is dependent on total recording length. A distinction has been made between variables that measure high and low frequency variation, representing the fast vagal and slower sympathetic modulations, respectively. For instance the RMSSD, the square root of the mean squared differences of successive RR intervals, is a high frequency parameter and the SDANN, the standard deviation of avarage RR interval length of 5 minute segments, is by definition a low frequency parameter (cycle >5 minutes). An alternative approach is by means of spectral analysis of the RR interval lengths. The high frequency (HF, Hz) components mainly represent vagal modulation and are influenced by respiration, whereas the low frequency (LF, Hz) components are both sympathetically and vagally modulated. This was first shown by parasympathetic and total autonomic blockade with glycopyrrolate and propranolol in dogs 3 and later reproduced in humans 4. Less is known about the physiological basis of the very low frequency ( Hz) fluctuations, that may represent parasympathetic outflow, thermoregulation and the renin angiotensin aldosteron system 28. The ratio between low and high frequency oscillations, the LF/HF ratio, is considered to reflect sympathovagal balance 29, although this has been under dispute

29 Chapter 2 Baroreceptor reflex sensitivity (BRS). The BRS can be assessed by measuring the changes in heart rate to mechanically or pharmacologically induced changes in blood pressure. Series of blood pressure measurements can be obtained invasively, as well as non invasively. The non invasive blood pressure measurements by a photoplethysmographic device (Finapres, Ohmeda, Englewood, Colo, USA) have been validated against intra arterial recordings 31. Classically, the BRS has been determined by measuring the changes in RR interval to intravenous injection of pressor agents (angiontensin II, phenylephrine) 32 or depressor agents (nitroprusside). The BRS is then expressed in msec/ mmhg. However, these drugs might influence the baroreflex itself. Similarly, RR interval length modulation after baroreceptor deactivation by a neck suction chamber can assess BRS 33. Alternatively, cross spectral analysis of spontaneous blood pressure and heart rate variation can be used to estimate BRS 34. Similar to HRV, a spectrum of blood pressure variability (BPV) can be obtained by spectral analysis of blood pressure series. As a measure of BRS, the mean gain between BPV and HRV spectra in the LF and HF bands 34 is calculated, see figure Keeping in mind the goal of a stable blood pressure, a sensitive system responds to small changes in blood pressure with a powerful adaptation of the heart rate. Consequently, there is an expected inverse relationship between the sensitivity of the baroreceptor reflex and the short term variability in blood pressure and a positive relationship between BRS and variability of heart rate. Finally, the sequence method is based on identification of sequences of four or more consecutive heart beats characterized by a progressive rise in systolic blood pressure and RR interval lengthening or a progressive decrease in systolic blood pressure and RR interval shortening. The slope of regression line of the changes in systolic blood pressure and RR interval length is taken as index of BRS 36. The different techniques to assess BRS are highly correlated 35. QTc. The QT interval reflects the total duration of ventricular myocardial depolarization and repolarization. Heart rate adjusted QT interval (QTc) prolongation is regarded as an indicator of an imbalanced distribution of autonomic nervous system activity to the heart 37;38. Congenital of pharmacological perturbation of the autonomic nervous system may result in prolongation of the QTc interval 39;40. Two mechanisms 20

30 Autonomic Dysfunction in Cardiovascular Disease. A consequence of vascular damage have been suggested to explain the increased risk of a prolonged QTc. First, predominance of sympathetic nerve activity or parasympathetic damage results in dispersion of repolarization, resulting in a high risk of ventricular fibrillation 41. Second, disturbed myocardial membrane function may lead to electrical instability and consequently to a high risk of arrhythmia and sudden death 42. Predictive value of autonomic function parameters for mortality in health and cardiovascular disease Healthy subjects. The Framingham Heart Study has shown, in a 30 years follow up, the prognostic value of a high heart rate for all cause and cardiovascular mortality in subjects free of cardiovascular disease 43. Later, a diminished HRV from a 2 hour ambulatory ECG was predictive of cardiac events in the same study, independent from traditional cardiovascular disease risk factors 44. Similar results were seen from a low HRV in a 2 minute rhythm strip in the ARIC study 45. A high heart rate or a diminished HRV could be indicative of inappropriate autonomic arousal or may reflect subclinical cardiac disease or poor health Prolonged QTc interval predicted cardiovascular and all cause mortality in apparently healthy populations 46;47. One must be aware that alterations in QTc intervals may result from physiological manoeuvres and vasodilation in healthy subjects 48. The predictive value of QTc could mean that it is an indicator of subclinical cardiovascular disease, since prolonged QTc is associated with traditional risk factors and with atherosclerotic disease as measured with intima media thickness 49;50. Post myocardial infarction (MI). In the ATRAMI study, depressed autonomic function was a strong independent predictor of mortality; a low BRS (<3.0 ms per mmhg) or low HRV (SDNN <70 ms) carried a multivariate risk or cardiac mortality 3.2 and 2.8 while left ventricular ejection dysfunction and the presence of frequent ventricular premature complexes were included in the model 51. This study confirmed earlier studies of the predictive value of impaired HRV after myocardial infarction, both for the time domain 52 and frequency domain 21

31 Chapter 2 parameters 53. A possible explanation for this may be the increased susceptibility for the development of life threatening arrhythmias in an altered state of autonomic balance (a relative sympathetic predominance and reduced vagal activity) 54. The association of a depressed BRS or reduced HRV with an increased incidence of arrhythmic events in postinfarction patients supports this hypothesis 55. Also, prolonged QTc, as a marker of autonomic disturbance, was predictive of sudden death in patients with myocardial infarction 56. Heart failure. Chronic heart failure (CHF) is characterized by increased sympathetic activation, as mirrored by increased sympathetic nerve activity, and to some extent by high levels of norepinephrine and angiotensin II 57;58. Furthermore, CHF patients have a decreased vagal tone and severe baroreceptor impairment 59. This leads to cessation of the restraining influence on central sympathetic activity, and thus contributes to sympathetic excitation 60. Neurohormonal activation, as reflected by plasma norepinephrine and plasma renin, is a predictor of mortality in CHF Autonomic dysfunction, measured by BRS and HRV, is also a powerful predictor of mortality in CHF. A BRS <1.3 ms per mmhg (lowest quartile) created a relative risk of and the UK Heart trial reported an annual mortality rate of 5.5% in patients with a preserved HRV (SDNN >100 ms) compared to 12.7% in patients with moderately impaired HRV (SDNN 50 to 100 ms) and 51.4% with severely impaired HRV (SDNN <50 ms) 65. In both studies, the autonomic function was an independent predictor of known risk factors and only moderately correlated with left ventricular ejection fraction. 22

32 Autonomic Dysfunction in Cardiovascular Disease. A consequence of vascular damage The autonomic nervous system in hypertension and diabetes mellitus. Hypertension. The role of the sympathetic nervous system in the pathogenesis of hypertension has been acknowledged since decades. The evidence for increased sympathetic drive principally comes from studies of the early phases: borderline and mild hypertension. A hyperkinetic state (i.e. elevated cardiac output and heart rate) as a marker of increased sympathetic tone predicts later development of hypertension 66. Plasma catecholamine levels and norepinephrine spillover rates are increased in young hypertensive patients 67;68. Direct intraneural recordings of sympathetic nerves in borderline hypertensive patients show elevated activity 69. Also analysis of HRV shows sympathetic predominance in hypertension 70. Finally, In the Framingham Heart Study, lower HRV could predict the development of hypertension in normotensive men but not in women 71. Whereas the evidence for the increased sympathetic tone in the development of hypertension is largely concordant, it is less clear whether the cardiovascular reflexes are changed very early on or whether the baroreflexes suffer as a result from structural and functional changes of the peripheral components of the baroreflex arch. Baroreflex function is in part genetically determined, as the presence of a family history of hypertension is the strongest predictor of BRS in hypertensive as well as in normotensive subjects 72 However, since the early work in this field of McCubbin et al demonstrated a marked rese ing of the arterial baroreflex in chronic hypertension 73, skepticism has predominated about a primary role of the baroreflex in the long term control of arterial blood pressure. Other studies also reported rese ing of the baroreceptor reflex setpoint to a higher level 74 and a reduction of the baroreceptor reflex sensitivity in hypertension 75;76. Other reasons that the concept of baroreceptor reflex dysfunction as a cause of hypertension was abandoned was that complete denervation of the baroreceptors does not results in sustained blood pressure elevation 77;78. Rather, following the short term blood pressure stabilizing goal of the baroreceptor reflex, blood pressure variability was increased 77;78. An alternative view argues a resetting of the baroreceptor setpoint while the stimulus response curve remains essentially the same, i.e. the baroreceptor reflex sensitivity is not changed

33 Chapter 2 Nonetheless, considering the importance of the baroreflex in the acute regulation of sympathetic activity and arterial pressure, there has been continued interest in the possibility that it plays a role in the pathogenesis of hypertension. Indeed, many have considered that the excessive sympathetic activation in hypertension might be accounted for by baroreflex dysfunction, an associated finding in some forms of experimental and clinical hypertension. However, whether impaired baroreflex suppression of sympathetic activity plays a role in the hypertensive process depends critically on whether baroreflexes completely reset when exposed to chronic changes in arterial pressure. If resetting is complete and baroreflexes do not chronically alter sympathetic activity, then they could not produce functional changes that influence the severity of hypertension. The fundamental question of whether baroreflexes completely reset and have the capacity to chronically alter sympathetic activity and arterial pressure remained unanswered up till recently due, in large part, to technical limitations that prcluded the assessment of chronic changes in sympathetic nerve activity and the long term effects of alterations in baroreflex activity on arterial pressure. Recently, however, a number of novel observations in chronically instrumented animals have indicated that the baroreflexes do not completely reset and are chronically activated in hypertension. These studies also support the hypothesis that in the effector path of the baroreflex suppression of renal sympathetic nerve activity and a endant increments in renal excretory function are key mediators of the antihypertensive response. The afferent part of the baroreceptor reflex could be impaired secondary to changes in the vascular mechanical and endothelial functional properties at the site of the baroreceptors. The baroreceptor reflex sensitivity is strongly related to carotid artery elasticity 80, which is known to be diminished in hypertension 81. However, functional damage to the endothelium at the baroreceptor region may affect the integrity of the afferent pathway of the baroreceptor reflex as well. Impaired prostacyclin release 82;83, enhanced formation of oxygen free radicals 84 and platelet aggregation 85 suppress baroreceptor activity and could contribute to baroreceptor dysfunction in hypertension. The efferent pathways of the baroreceptor reflex loop could suffer from adaptation to elevated blood pressure as well. Vascular hypertrophy may be responsible for increased adrenergic responsiveness to 24

34 Autonomic Dysfunction in Cardiovascular Disease. A consequence of vascular damage stress 86. However, one could also argue that the hypertrophied vascular tree loses compliance and thereby its buffering capacity. Furthermore, in animal models, development of left ventricular hypertrophy (LVH) impairs baroreflex control of heart rate, whereas treatment with an angiotensin converting enzyme (ACE) inhibitor could reverse these processes 87. In hypertensive patients, the reduction of LVH was confined to those with improved baroreceptor reflex control of heart rate as measured by HRV 88. Thus, the peripheral components of the baroreceptor reflex arch are vulnerable to adaptive changes of an sustained elevated blood pressure. Further, evidence is emerging that in hypertension the renal nerves may be the critical efferent link for baroreceptor induced suppression of central sympathetic output through which long term compensatory reductions in arterial pressure are produced 89. Irrespective whether autonomic imbalance and impaired baroreceptor reflex function precede or follow the development of hypertension, they are closely related to morbidity and mortality in hypertension. As clearly explained by Julius, elevated heart rate as marker of increased sympathetic tone is not only related to hypertension, but also to lipid disorders and insulin resistance, thus contributing to the increased coronary risk 90;91. Finally, in the Framingham Heart Study, heart rate was a predictor for cardiovascular and all cause mortality in hypertensive men and women 86. Diabetes mellitus. Autonomic dysfunction is a common complication of diabetes mellitus. When assessed with Ewing s standard autonomic function tests, abnormal autonomic function is reported in 20 40% of unselected diabetic patients 92. The presence of cardiovascular autonomic dysfunction is associated with an increased risk for mortality in diabetes mellitus 26; The predictive value was first shown by Ewing et al., using the autonomic function tests that have become golden standard 26. However, in practice these tests are strongly dependent upon performance and suffer from poor reproducibility. Initially, autonomic dysfunction was thought to be a late complication, because it was diagnosed in patients with long standing diabetes 101. With the development of more sensitive methods to diagnose autonomic dysfunction, it is earlier recognized. HRV and BRS can already be impaired, even when standard autonomic function tests are still normal The 25

35 Chapter 2 QT interval may also be prolonged already when diabetes mellitus is first diagnosed, and was found to be a be er predictor of cardiac death than ankle brachial pressure index and autonomic function tests 106. The importance of early recognition is underscored by the predictive value of impaired autonomic function for mortality, even at a subclinical level 95;96. What mechanisms are responsible for the autonomic dysfunction in diabetes and how do they contribute to increased mortality? The abnormal cardiovascular autonomic function in the initial reports 26;107 of Ewing was ascribed to the presence of neuropathy, originating from the observation of a lower nerve conduction velocity and longer terminal latency in patients with abnormal heart rate response to the Valsalva maneuver 107. Indeed, cardiovascular autonomic dysfunction is often seen in patients with diabetic peripheral neuropathy, but there is a lower than expected concordance and a variable relationship between them A number of non neural abnormalities, like increased intima media thickness at the site of the baroreceptors, reduced distensibility, impaired cardiac vagal function, left ventricular hypertrophy and endothelial dysfunction could contribute to the autonomic dysfunction in diabetes mellitus. Interestingly, these abnormalities are closely associated with the presence of (micro )albuminuria , which is seen as a reflection of endothelial dysfunction or vascular damage in diabetes mellitus 116. Furthermore, autonomic dysfunction is particularly pronounced when diabetes is complicated by (micro )albuminuria in both type 1 105;113;117 and type diabetic patients. In fact, autonomic dysfunction may explain part of the increased mortality that is associated with (micro )albuminuria in diabetes mellitus 96;119. Conversely, since autonomic dysfunction is a predictor of deterioration of renal function in diabetes mellitus, renal disease may partly account for the increased mortality rate in patients with diabetes mellitus complicated by autonomic neuropathy 120;121. Another mechanism by which autonomic dysfunction may deteriorate prognosis is the increased vulnerability to lethal arrhythmias, by analogy with post myocardial infarction patients. This could explain the high number of unexplained deaths in diabetic patients 114. Finally, the notion that autonomic neuropathy is a result of longstanding hyperglycaemia is incompatible with the recent report of a higher incidence 26

36 Autonomic Dysfunction in Cardiovascular Disease. A consequence of vascular damage in non diabetic offspring of type 2 diabetic patients compared with non diabetic offspring of nondiabetic subjects 122. Clearly, these subjects had not been exposed to longstanding hyperglycaemia. Rather, in this study autonomic neuropathy was associated with features of the metabolic syndrome and microvascular damage (increased urinary albumin excretion). In summary, widespread vascular damage at the endothelial, vascular and cardiac level is an important determinant of autonomic dysfunction. For diabetes mellitus, the impaired cardiovascular reflexes, as assessed by bedside autonomic function tests, in the diabetic patients in the original paper from Ewing 26 may very well have been related to and partly caused by vascular damage, rather than neuropathy per se. With the availability of modern, sensitive measures of autonomic function, the stratification of diabetic patients at risk may improve. Until now, the predictive value of BRS and spectral and time domain parameters of HRV has yet to be proven in diabetes mellitus. Similar to impaired autonomic function as measured by Ewing s tests, HRV and BRS, prolonged QTc duration is a predictor of cardiovascular mortality in type 1 and type 2 diabetes mellitus 93;97;99;100. Prolonged QTc is associated with traditional risk factors in type 1 diabetic patients, underscoring the relation between (subclinical) cardiovascular damage and autonomic dysfunction 98. The interplay between autonomic control and vascular function. Cardiovascular determinants of autonomic function. As discussed earlier, a number of structural and functional changes of the cardiovascular system can contribute to autonomic dysfunction in cardiovascular disease, by modifying the afferent or effector systems. First, the effects of structural damage to the baroreceptors after for example radiotherapy have been discussed earlier. A reduction in arterial compliance may reduce the baroreflex mediated control of cardiac output and blood pressure. An a enuated baroreflex control of cardiac output and arterial blood pressure has been shown after an acute increase in arterial compliance in rats 123. In humans, a diminished BRS 27

37 Chapter 2 and reduced carotid arterial compliance were found in hypertensive subjects compared to normotensive subjects 124. However, although a significant correlation of r=0.53 was found between BRS and arterial compliance in the complete group, this could not be demonstrated in the hypertensive subjects only, suggesting that other factors also contributed to baroreflex dysfunction in these individuals. In a study of 19 healthy young subjects, BRS was significantly related to carotid artery distensibility with r= However, other reports have shown that changes in baroreceptor properties do not simply depend on a decreased distensibility of the carotid sinus arterial wall 126;127. Second, the effects of a defective endothelium at the site of the baroreceptor in the carotid arterial wall on baroreflex function have been reported. Local, paracrine and endocrine effects may play a role. An impaired formation of PGI2 was partly responsible for decreased baroreceptor sensitivity in hypertension and atherosclerosis 128. Furthermore, aggregating platelets at the carotid sinus decreased baroreceptor activity in an animal model, without altering the carotid pressure diameter relation, suggesting a direct inhibitory effect on the baroreceptors 129. The role endothelial derived NO is discussed in some more detail below. Also in an animal model, oxidative stress contributed to baroreceptor dysfunction 130. Possibly, such mechanisms plays a role in atherosclerosis. Third, cardiac changes can reduce effective baroreflex control of heart rate and cardiac output in pathological states, as discussed in more detail above for myocardial infarction and heart failure 7;8; However, the mechanism of baroreflex dysfunction in CHF is probably multifactorial and is not limited to cardiac changes but may be located in all components of the reflex arch The interaction between nitric oxide and the the autonomic nervous system. The interaction between nitric oxide, either endothelium derived or not, and baroreflex mediated sympathetic and cholinergic signaling derserves some special a ention. This subject has recently been extensively reviewed by Sartori et al 139;140. The relationship is complex, as illustrated by the differential dose related effects of modulating NO synthesis for example by l NMMA infusions: at high doses, l NMMA and phenylephrine similarly suppress sympathetic nerve activity, whereas when infused at lower doses, these drugs have differential 28

38 Autonomic Dysfunction in Cardiovascular Disease. A consequence of vascular damage sympathetic effects 141. Sartori proposed that these differential sympathetic effects of low and high dose l NMMA infusion in humans may be partly due to differences in IC50 doses for l NMMA induced blockade of neural vs. endothelial NO synthesis 140. It also seems as if inhibition of NO synthase has central sympathoexcitatory effects that are masked by an inhibitory effect of the baroreflexes 142. Insulin is another player in the interaction between NO and the autonomic nervous system which is relevant in cardiovascular disease. Insulin has a role in the regulation of peripheral vascular tone and arterial blood pressure in humans. In lean, healthy subjects, insulin infusion increases, as a marker of its sympathetic vasoconstrictor action, sympathetic neural outflow to skeletal muscle tissue 143. On the other hand however, insulin has also been reported to stimulate blood flow and decrease vascular resistance in skeletal muscle 144. These effects are not due to beta adrenergic or cholinergic mechanisms, because propranolol and atropine do not a enuate insulin s vasodilation 145. There is now abundant evidence, that NO release accounts for insulin s vasodilator action 146. In line with this concept, insulin induced vasodilation in humans is abolished by l NMMA 147. Thus, insulin induced vasodilation is mediated by stimulation of endothelial and sympathetically mediated NO release. Studies in patients with sympathectomy and auto nomic failure indicate that the sympathetic vasoconstrictor tone restricts this response, thereby preventing exaggerated insulin induced vasodilation and hypotension 148. In normal subjects, the cholinergic system plays a major, hitherto unrecognised role in offse ing pressor effects caused by NO synthase inhibition. Lepori et al. examined the effects of cholinergic blockade on the blood pressure and peripheral vasoconstrictor responses to systemic l NMMA infusion in normal subjects 149. Cholinergic blockade had a dramatic effect on the pressor response to NO synthase inhibition in healthy subjects: the l NMMA induced increase in mean arterial pressure was roughly 3 times larger in the presence than in the absence of atropine infusion. This potentiation is specific for NO dependent vasoconstriction, because atropine did not alter the responses to phenylephrine infusion. As for effects of NO on heart rate, the role of neuronal NO seems to predominate above that of endothelial NO. Studies in intact neuronal NO synthase knockout mice, as well as isolated atria harvested from such animals, suggest that parasym- 29

39 Chapter 2 pathetic control of heart rate is impaired in the absence of neuronal NO synthase 150;151. Interestingly, the exercise training induced bradycardia in mice appears also to be related to upregulation of neuronal NO synthase 152. Studies by Brunner et al. support that endothelial NO synthase derived NO appears to have li le role in the peripheral autonomic regulation of heart rate 153. Chowdhary found in human studies, using heart rate variability analysis suggest that NO augments cardiac vagal control in healthy subjects and patients with heart failure 154;155. In the earlier discussed study by Lepori et al, cholinergic blockade had not only marked vasoconstrictive effects in the vascular responses to NO synthase inhibition in humans, but also alters the reflex decrease in heart rate to l NMMA infusion alone to a sympathetically mediated increase in heart rate (could be blocked by propranolol) 149. The triangle of sympathetic activation, insulin resistance and metabolic syndrome (crossroads of hypertension and diabetes). The metabolic syndrome carries an increase in cardiovascular risk and its occurrence is associated with alterations which might include sympathetic hyperactivity, because the metabolic syndrome is associated with insulin resistance and hyperinsulinemia, which has sympathostimulating effects 156. Sympathetic activation has been considered as a link between insulin resistance, hyperinsulinemia, and hypertension. Hypertension may, however, also be considered as a confounder, because essential hypertension without components of the metabolic syndrome is also associated with sympathetic hyperactivity. However, a recent study by Grassi et al using MSNA recordings in persons with NCEP ATP III criteria metabolic syndrome patients provided direct evidence that sympathetic activation is not limited to individuals in whom the metabolic syndrome is accompanied by hypertension 157. Sympathetic nerve traffic was also greater than in control subjects when the metabolic syndrome was associated with normal BP values, and thus its diagnosis depended on criteria other than the BP elevation. So, a hyperadrenergic state represents an intrinsic feature of this condition. Epidemiologically, in the Framingham Heart Study, an impaired fasting glucose level was also associated with lower HRV, independent of blood pressure levels 158. In the Grassi study, subjects with metabolic syndrome had an a enuation of the sensitivity of baroreceptor sympathetic control, which suggests that impairment of the ability of the baroreflex to restrain adrenergic 30

40 Autonomic Dysfunction in Cardiovascular Disease. A consequence of vascular damage tone is also involved, besides factors like the degree of insulin resistance. Endothelial NO synthesis is defective in insulin resistant states, and, as discussed above, may contribute to the neuronal abnormalities of the metabolic syndrome. Autonomic dysfunction may be an early defect, before the metabolic syndrome has become manifest, as suggested by an increase in sympathovagal balance measured by HRV in response to hyperinsulinemia in nondiabetic offspring of type 2 diabetic patients, compared to controls 159. As for the transitional area between metabolic syndrome and hypertension, the HRV abnormalities in a group of hypertensive patients were particularly pronounced in those with metabolic features of the insulin resistance syndrome 160. Conclusion Analysis of (variations of) heart rate and blood pressure in rest and during stimuli can assess cardiovascular autonomic function. Ewing s tests and analysis of heart variability are now accompanied by modern techniques as baroreceptor reflex sensitivity, QTc interval length, muscle sympathetic nerve activity and paracrine and endocrine markers. The predictive value for morbidity and mortality of these autonomic function parameters has been and is established in health and cardiovascular disease, like heart failure, post myocardial infarction and diabetes mellitus. The cardiovascular autonomic function should be considered as the result of the interplay between many components, consisting of vascular, neural, cardiac and paracrine and endocrine entities. The autonomic dysfunction in cardiovascular disease is therefore most likely the result of defects at multiple sites. Furtheremore, cardiovascular risk factors and abnormalities may cluster in individuals and populations and the cardiovascular autonomic dysfunction must be seen as the outcome of damage at different sites, resulting in a compromised integrity of the cardiovascular reflexes. 31

41 Chapter 2 References 1. Mayer, S. Studien zur Physiologie des Herzens und der Blutgefaesse 6. Abhandlung: ueber spontane Blutdruckschwenkungen. [Studies on the physiology of the heart and the blood vessels 6. Discourse on fluctuations in blood pressure.]. Sitz. Ber. Akad. Wiss.Wien. Mathe Naturwiss. Kl. Anat. 74, Ref Type: Generic 2. Bernardi L, Leuzzi S, Radaelli A, Passino C, Johnston JA, Sleight P: Low frequency spontaneous fluctuations of R R interval and blood pressure in conscious humans: a baroreceptor or central phenomenon? Clin.Sci.(Lond) 87: , Akselrod S, Gordon D, Ubel FA, Shannon DC, Berger AC, Cohen RJ: Power spectrum analysis of heart rate fluctuation: a quantitative probe of beat to beat cardiovascular control. Science 213: , Pomeranz B, Macaulay RJ, Caudill MA, Kutz I, Adam D, Gordon D, Kilborn KM, Barger AC, Shannon DC, Cohen RJ: Assessment of autonomic function in humans by heart rate spectral analysis. Am.J.Physiol 248:H151 H153, Jose AD, Collison D: The normal range and determinants of the intrinsic heart rate in man. Cardiovasc.Res. 4: , Ewing DJ, Campbell IW, Clarke BF: Mortality in diabetic autonomic neuropathy. Lancet 1: , Nolan J, Batin PD, Andrews R, Lindsay SJ, Brooksby P, Mullen M, Baig W, Flapan AD, Cowley A, Presco RJ, Neilson JM, Fox KA: Prospective study of heart rate variability and mortality in chronic heart failure: results of the United Kingdom heart failure evaluation and assessment of risk trial (UK heart). Circulation 98: , La Rovere MT, Bigger JT, Jr., Marcus FI, Mortara A, Schwartz PJ: Baroreflex sensitivity and heart rate variability in prediction of total cardiac mortality after myocardial infarction. ATRAMI (Autonomic Tone and Reflexes After Myocardial Infarction) Investigators. Lancet 351: , Scher, A. M. Cardiovascular control. Textbook of physiology, volume W.B. Saunders Company. Ref Type: Generic 32

42 Autonomic Dysfunction in Cardiovascular Disease. A consequence of vascular damage 10. Karemaker, J. M. Neurophysiology of the baroreceptor reflex. The beat by beat investigation of cardiovascular function Oxford, Clarendon Press. Ref Type: Generic 11. Smit AA, Timmers HJ, Wieling W, Wagenaar M, Marres HA, Lenders JW, van Montfrans GA, Karemaker JM: Long term effects of carotid sinus denervation on arterial blood pressure in humans. Circulation 105: , Lohmeier TE, Irwin ED, Rossing MA, Serdar DJ, Kieval RS: Prolonged activation of the baroreflex produces sustained hypotension. Hypertension 43: , Pickering TG: The clinical significance of diurnal blood pressure variations. Dippers and nondippers. Circulation 81: , Mancia G, Di Rienzo M, Parati G: Ambulatory blood pressure monitoring use in hypertension research and clinical practice. Hypertension 21: , Mann S, Altman DG, Raftery EB, Bannister R: Circadian variation of blood pressure in autonomic failure. Circulation 68: , Millar Craig MW, Bishop CN, Raftery EB: Circadian variation of blood pressure. Lancet 1: , Somers VK, Dyken ME, Mark AL, Abboud FM: Sympathetic nerve activity during sleep in normal subjects. N.Engl.J.Med. 328: , Ikeda T, Matsubara T, Sato Y, Sakamoto N: Circadian blood pressure variation in diabetic patients with autonomic neuropathy. J.Hypertens. 11: , Caruana MP, Lahiri A, Cashman PM, Altman DG, Raftery EB: Effects of chronic congestive heart failure secondary to coronary artery disease on the circadian rhythm of blood pressure and heart rate. Am.J.Cardiol. 62: , Reeves RA, Shapiro AP, Thompson ME, Johnsen AM: Loss of nocturnal decline in blood pressure after cardiac transplantation. Circulation 73: , Shimada K, Kawamoto A, Matsubayashi K, Nishinaga M, Kimura S, Ozawa T: Diurnal blood pressure variations and silent cerebrovascular damage in elderly patients with hypertension. J.Hypertens. 10: ,

43 Chapter Verdecchia P, Porcellati C, Schillaci G, Borgioni C, Ciucci A, Battistelli M, Guerrieri M, Ga eschi C, Zampi I, Santucci A,.: Ambulatory blood pressure. An independent predictor of prognosis in essential hypertension. Hypertension 24: , Devereux RB, Pickering TG: Relationship between the level, pattern and variability of ambulatory blood pressure and target organ damage in hypertension. J.Hypertens.Suppl 9:S34 S38, Guyton, A. C. and Hall, J. E. Textbook of Medical Physiology, Ninth Edition Ref Type: Generic 25. Ewing DJ, Clarke BF: Diagnosis and management of diabetic autonomic neuropathy. Br.Med.J.(Clin.Res.Ed) 285: , Ewing DJ, Campbell IW, Clarke BF: Mortality in diabetic autonomic neuropathy. Lancet 1: , Heart rate variability. Standards of measurement, physiological interpretation, and clinical use. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Eur.Heart J. 17: , Taylor JA, Carr DL, Myers CW, Eckberg DL: Mechanisms underlying very low frequency RR interval oscillations in humans. Circulation 98: , Pagani M, Lombardi F, Guzze i S, Rimoldi O, Furlan R, Pizzinelli P, Sandrone G, Malfa o G, Dell Orto S, Piccaluga E,.: Power spectral analysis of heart rate and arterial pressure variabilities as a marker of sympatho vagal interaction in man and conscious dog. Circ.Res. 59: , Eckberg DL: Sympathovagal balance: a critical appraisal. Circulation 96: , Imholz BP, Wieling W, van Montfrans GA, Wesseling KH: Fifteen years experience with finger arterial pressure monitoring: assessment of the technology. Cardiovasc.Res. 38: , Smyth HS, Sleight P, Pickering GW: Reflex regulation of arterial pressure during sleep in man. A quantitative method of assessing baroreflex sensitivity. Circ.Res. 24: , Eckberg DL, Cavanaugh MS, Mark AL, Abboud FM: A simplified neck suction device for activation of carotid baroreceptors. J.Lab Clin.Med. 85: ,

44 Autonomic Dysfunction in Cardiovascular Disease. A consequence of vascular damage 34. deboer RW, Karemaker JM, Strackee J: Hemodynamic fluctuations and baroreflex sensitivity in humans: a beat to beat model. Am.J.Physiol 253:H680 H689, Robbe HW, Mulder LJ, Ruddel H, Langewitz WA, Veldman JB, Mulder G: Assessment of baroreceptor reflex sensitivity by means of spectral analysis. Hypertension 10: , Parati G, Castiglioni P, Di Rienzo M, Omboni S, Pedo i A, Mancia G: Sequential spectral analysis of 24 hour blood pressure and pulse interval in humans. Hypertension 16: , Yanowitz F, Preston JB, Abildskov JA: Functional distribution of right and left stellate innervation to the ventricles. Production of neurogenic electrocardiographic changes by unilateral alteration of sympathetic tone. Circ.Res. 18: , Schwartz PJ, Snebold NG, Brown AM: Effects of unilateral cardiac sympathetic denervation on the ventricular fibrillation threshold. Am.J.Cardiol. 37: , Schwartz PJ, Periti M, Malliani A: The long Q T syndrome. Am.Heart J. 89: , Ahnve S, Vallin H: Influence of heart rate and inhibition of autonomic tone on the QT interval. Circulation 65: , Schartz, PJ and Priori, SG. Sympathetic nervous system and cardiac arrhythmias. Zipes, DP and Jalife, J. Cardiac electrophysiology: from cell to bedside Philadelphia, Saunders. Ref Type: Generic 42. Zipes DP: The long QT interval syndrome. A Rose a stone for sympathetic related ventricular tachyarrhythmias. Circulation 84: , Kannel WB, Kannel C, Paffenbarger RS, Jr., Cupples LA: Heart rate and cardiovascular mortality: the Framingham Study. Am. Heart J. 113: , Tsuji H, Larson MG, Vendi i FJ, Jr., Manders ES, Evans JC, Feldman CL, Levy D: Impact of reduced heart rate variability on risk for cardiac events. The Framingham Heart Study. Circulation 94: ,

45 Chapter Dekker JM, Crow RS, Folsom AR, Hannan PJ, Liao D, Swenne CA, Schouten EG: Low heart rate variability in a 2 minute rhythm strip predicts risk of coronary heart disease and mortality from several causes: the ARIC Study. Atherosclerosis Risk In Communities. Circulation 102: , de Bruyne MC, Hoes AW, Kors JA, Hofman A, van Bemmel JH, Grobbee DE: Prolonged QT interval predicts cardiac and all cause mortality in the elderly. The Ro erdam Study. Eur.Heart J. 20: , Schouten EG, Dekker JM, Meppelink P, Kok FJ, Vandenbroucke JP, Pool J: QT interval prolongation predicts cardiovascular mortality in an apparently healthy population. Circulation 84: , Berger E, Patel K, Anwar S, Davies W, Sheridan DJ: Investigation of the effects of physiological and vasodilation induced autonomic activation on the QTc Interval in healthy male subjects. Br. J.Clin.Pharmacol. 60:17 23, Lefrandt JD, Diercks GF, van Boven AJ, Crijns HJ, van Gilst WH, Gans RO: High fasting glucose and QTc duration in a large healthy cohort. Diabetologia 43: , Festa A, D Agostino R, Jr., Rautaharju P, O Leary DH, Rewers M, Mykkanen L, Haffner SM: Is QT interval a marker of subclinical atherosclerosis in nondiabetic subjects? The Insulin Resistance Atherosclerosis Study (IRAS). Stroke 30: , La Rovere MT, Bigger JT, Jr., Marcus FI, Mortara A, Schwartz PJ: Baroreflex sensitivity and heart rate variability in prediction of total cardiac mortality after myocardial infarction. ATRAMI (Autonomic Tone and Reflexes After Myocardial Infarction) Investigators. Lancet 351: , Kleiger RE, Miller JP, Bigger JT, Jr., Moss AJ: Decreased heart rate variability and its association with increased mortality after acute myocardial infarction. Am.J.Cardiol. 59: , Bigger JT, Jr., Fleiss JL, Steinman RC, Rolnitzky LM, Kleiger RE, Ro man JN: Frequency domain measures of heart period variability and mortality after myocardial infarction. Circulation 85: ,

46 Autonomic Dysfunction in Cardiovascular Disease. A consequence of vascular damage 54. Schwartz PJ, La Rovere MT, Vanoli E: Autonomic nervous system and sudden cardiac death. Experimental basis and clinical observations for post myocardial infarction risk stratification. Circulation 85:I77 I91, Farrell TG, Odemuyiwa O, Bashir Y, Cripps TR, Malik M, Ward DE, Camm AJ: Prognostic value of baroreflex sensitivity testing after acute myocardial infarction. Br.Heart J. 67: , Schwartz PJ, Wolf S: QT interval prolongation as predictor of sudden death in patients with myocardial infarction. Circulation 57: , Francis GS, Benedict C, Johnstone DE, Kirlin PC, Nicklas J, Liang CS, Kubo SH, Rudin Toretsky E, Yusuf S: Comparison of neuroendocrine activation in patients with left ventricular dysfunction with and without congestive heart failure. A substudy of the Studies of Left Ventricular Dysfunction (SOLVD). Circulation 82: , Packer M: The neurohormonal hypothesis: a theory to explain the mechanism of disease progression in heart failure. J.Am.Coll. Cardiol. 20: , Nolan J, Flapan AD, Capewell S, MacDonald TM, Neilson JM, Ewing DJ: Decreased cardiac parasympathetic activity in chronic heart failure and its relation to left ventricular function. Br.Heart J. 67: , Eckberg DL, Drabinsky M, Braunwald E: Defective cardiac parasympathetic control in patients with heart disease. N.Engl.J.Med. 285: , Cohn JN, Levine TB, Olivari MT, Garberg V, Lura D, Francis GS, Simon AB, Rector T: Plasma norepinephrine as a guide to prognosis in patients with chronic congestive heart failure. N.Engl. J.Med. 311: , Francis GS, Cohn JN, Johnson G, Rector TS, Goldman S, Simon A: Plasma norepinephrine, plasma renin activity, and congestive heart failure. Relations to survival and the effects of therapy in V HeFT II. The V HeFT VA Cooperative Studies Group. Circulation 87:VI40 VI48,

47 Chapter Swedberg K, Eneroth P, Kjekshus J, Wilhelmsen L: Hormones regulating cardiovascular function in patients with severe congestive heart failure and their relation to mortality. CONSENSUS Trial Study Group. Circulation 82: , Mortara A, La Rovere MT, Pinna GD, Prpa A, Maestri R, Febo O, Pozzoli M, Opasich C, Tavazzi L: Arterial baroreflex modulation of heart rate in chronic heart failure: clinical and hemodynamic correlates and prognostic implications. Circulation 96: , Nolan J, Batin PD, Andrews R, Lindsay SJ, Brooksby P, Mullen M, Baig W, Flapan AD, Cowley A, Presco RJ, Neilson JM, Fox KA: Prospective study of heart rate variability and mortality in chronic heart failure: results of the United Kingdom heart failure evaluation and assessment of risk trial (UK heart). Circulation 98: , Julius S, Krause L, Schork NJ, Mejia AD, Jones KA, van d, V, Johnson EH, Sekkarie MA, Kjeldsen SE, Petrin J,.: Hyperkinetic borderline hypertension in Tecumseh, Michigan. J.Hypertens. 9:77 84, Goldstein DS: Plasma norepinephrine in essential hypertension. A study of the studies. Hypertension 3:48 52, Esler M, Jennings G, Lambert G: Noradrenaline release and the pathophysiology of primary human hypertension. Am.J.Hypertens. 2:140S 146S, Anderson EA, Sinkey CA, Lawton WJ, Mark AL: Elevated sympathetic nerve activity in borderline hypertensive humans. Evidence from direct intraneural recordings. Hypertension 14: , Guzze i S, Piccaluga E, Casati R, Ceru i S, Lombardi F, Pagani M, Malliani A: Sympathetic predominance in essential hypertension: a study employing spectral analysis of heart rate variability. J.Hypertens. 6: , Singh JP, Larson MG, Tsuji H, Evans JC, O Donnell CJ, Levy D: Reduced heart rate variability and new onset hypertension: insights into pathogenesis of hypertension: the Framingham Heart Study. Hypertension 32: , Parmer RJ, Cervenka JH, Stone RA: Baroreflex sensitivity and heredity in essential hypertension. Circulation 85: ,

48 Autonomic Dysfunction in Cardiovascular Disease. A consequence of vascular damage 73. McCubbin JW, Green JH, Page IH: Baroceptor function in chronic renal hypertension. Circ.Res. 4: , Eckberg DL: Carotid baroreflex function in young men with borderline blood pressure elevation. Circulation 59: , Mancia G, Ferrari A, Gregorini L, Parati G, Pomidossi G, Bertinieri G, Grassi G, Di Rienzo M, Pedo i A, Zanche i A: Blood pressure and heart rate variabilities in normotensive and hypertensive human beings. Circ.Res. 53:96 104, Bristow JD, Honour AJ, Pickering GW, Sleight P, Smyth HS: Diminished baroreflex sensitivity in high blood pressure. Circulation 39:48 54, Norman RA, Jr., Coleman TG, Dent AC: Continuous monitoring of arterial pressure indicates sinoaortic denervated rats are not hypertensive. Hypertension 3: , Asmar RG, Julia PL, Mascarel VL, Fabiani JN, Benetos A, Safar ME: Ambulatory blood pressure profile after carotid endarterectomy in patients with ischaemic arterial disease. J.Hypertens. 12: , Zanche i A: Overview of cardiovascular reflexes in hypertension. Am.J.Cardiol. 44: , Bonyhay I, Jokkel G, Kollai M: Relation between baroreflex sensitivity and carotid artery elasticity in healthy humans. Am.J.Physiol 271:H1139 H1144, Van Merode T, Hick PJ, Hoeks AP, Rahn KH, Reneman RS: Carotid artery wall properties in normotensive and borderline hypertensive subjects of various ages. Ultrasound Med.Biol. 14: , Chen HI, Chapleau MW, McDowell TS, Abboud FM: Prostaglandins contribute to activation of baroreceptors in rabbits. Possible paracrine influence of endothelium. Circ.Res. 67: , Xie PL, Chapleau MW, McDowell TS, Hajduczok G, Abboud FM: Mechanism of decreased baroreceptor activity in chronic hypertensive rabbits. Role of endogenous prostanoids. J.Clin.Invest 86: , Li Z, Mao HZ, Abboud FM, Chapleau MW: Oxygen derived free radicals contribute to baroreceptor dysfunction in atherosclerotic rabbits. Circ.Res. 79: ,

49 Chapter Li Z, Abboud FM, Chapleau MW: Aggregating human platelets in carotid sinus of rabbits decrease sensitivity of baroreceptors. Circ.Res. 70: , Gillman MW, Kannel WB, Belanger A, D Agostino RB: Influence of heart rate on mortality among persons with hypertension: the Framingham Study. Am.Heart J. 125: , Head GA, Minami N: Importance of cardiac, but not vascular, hypertrophy in the cardiac baroreflex deficit in spontaneously hypertensive and stroke prone rats. Am.J.Med. 92:54S 59S, Muiesan ML, Rizzoni D, Zulli R, Castellano M, Be oni G, Porteri E, Agabiti Rosei E: Power spectral analysis of the heart rate in hypertensive patients with and without left ventricular hypertrophy: the effect of a left ventricular mass reduction. J.Hypertens. 16: , Lohmeier TE, Hildebrandt DA, Warren S, May PJ, Cunningham JT: Recent insights into the interactions between the baroreflex and the kidneys in hypertension. Am.J.Physiol Regul.Integr.Comp Physiol 288:R828 R836, Palatini P, Julius S: Heart rate and the cardiovascular risk. J.Hypertens. 15:3 17, Julius S: Corcoran Lecture. Sympathetic hyperactivity and coronary risk in hypertension. Hypertension 21: , Ewing DJ, Martyn CN, Young RJ, Clarke BF: The value of cardiovascular autonomic function tests: 10 years experience in diabetes. Diabetes Care 8: , Rossing P, Breum L, Major Pedersen A, Sato A, Winding H, Pietersen A, Kastrup J, Parving HH: Prolonged QTc interval predicts mortality in patients with Type 1 diabetes mellitus. Diabet. Med. 18: , O Brien IA, McFadden JP, Corrall RJ: The influence of autonomic neuropathy on mortality in insulin dependent diabetes. Q. J.Med. 79: , Rathmann W, Ziegler D, Jahnke M, Haastert B, Gries FA: Mortality in diabetic patients with cardiovascular autonomic neuropathy. Diabet.Med. 10: ,

50 Autonomic Dysfunction in Cardiovascular Disease. A consequence of vascular damage 96. Reichard P, Pihl M: Mortality and treatment side effects during long term intensified conventional insulin treatment in the Stockholm Diabetes Intervention Study. Diabetes 43: , Veglio M, Sivieri R, Chinaglia A, Scaglione L, Cavallo Perin P: QT interval prolongation and mortality in type 1 diabetic patients: a 5 year cohort prospective study. Neuropathy Study Group of the Italian Society of the Study of Diabetes, Piemonte Affiliate. Diabetes Care 23: , Veglio M, Borra M, Stevens LK, Fuller JH, Perin PC: The relation between QTc interval prolongation and diabetic complications. The EURODIAB IDDM Complication Study Group. Diabetologia 42:68 75, Christensen PK, Gall MA, Major Pedersen A, Sato A, Rossing P, Breum L, Pietersen A, Kastrup J, Parving HH: QTc interval length and QT dispersion as predictors of mortality in patients with non insulin dependent diabetes. Scand.J.Clin.Lab Invest 60: , Sawicki PT, Kiwi S, Bender R, Berger M: The value of QT interval dispersion for identification of total mortality risk in non insulin dependent diabetes mellitus. J.Intern.Med. 243:49 56, Clarke BF, Ewing DJ, Campbell IW: Diabetic autonomic neuropathy. Diabetologia 17: , Fra ola A, Parati G, Gamba P, Paleari F, Mauri G, Di Rienzo M, Castiglioni P, Mancia G: Time and frequency domain estimates of spontaneous baroreflex sensitivity provide early detection of autonomic dysfunction in diabetes mellitus. Diabetologia 40: , Weston PJ, James MA, Panerai R, McNally PG, Po er JF, Thurston H, Swales JD: Abnormal baroreceptor cardiac reflex sensitivity is not detected by conventional tests of autonomic function in patients with insulin dependent diabetes mellitus. Clin.Sci.(Lond) 91:59 64, Malpas SC, Maling TJ: Heart rate variability and cardiac autonomic function in diabetes. Diabetes 39: ,

51 Chapter Lefrandt JD, Hoogenberg K, van Roon AM, Dullaart RP, Gans RO, Smit AJ: Baroreflex sensitivity is depressed in microalbuminuric Type I diabetic patients at rest and during sympathetic manoeuvres. Diabetologia 42: , Rana BS, Lim PO, Naas AA, Ogston SA, Newton RW, Jung RT, Morris AD, Struthers AD: QT interval abnormalities are often present at diagnosis in diabetes and are be er predictors of cardiac death than ankle brachial pressure index and autonomic function tests. Heart 91:44 50, Ewing DJ, Campbell IW, Burt AA, Clarke BF: Vascular reflexes in diabetic autonomic neuropathy. Lancet 2: , Lluch I, Hernandez A, Real JT, Morillas C, Tenes S, Sanchez C, Ascaso JF: Cardiovascular autonomic neuropathy in type 1 diabetic patients with and without peripheral neuropathy. Diabetes Res.Clin.Pract. 42:35 40, Young RJ, Zhou YQ, Rodriguez E, Presco RJ, Ewing DJ, Clarke BF: Variable relationship between peripheral somatic and autonomic neuropathy in patients with different syndromes of diabetic polyneuropathy. Diabetes 35: , Toyry JP, Partanen JV, Niskanen LK, Lansimies EA, Uusitupa MI: Divergent development of autonomic and peripheral somatic neuropathies in NIDDM. Diabetologia 40: , Frost D, Beischer W: Determinants of carotid artery wall thickening in young patients with Type 1 diabetes mellitus. Diabet.Med. 15: , Lambert J, Smulders RA, Aarsen M, Donker AJ, Stehouwer CD: Carotid artery stiffness is increased in microalbuminuric IDDM patients. Diabetes Care 21:99 103, Molgaard H, Christensen PD, Hermansen K, Sorensen KE, Christensen CK, Mogensen CE: Early recognition of autonomic dysfunction in microalbuminuria: significance for cardiovascular mortality in diabetes mellitus? Diabetologia 37: , Ru er MK, McComb JM, Forster J, Brady S, Marshall SM: Increased left ventricular mass index and nocturnal systolic blood pressure in patients with Type 2 diabetes mellitus and microalbuminuria. Diabet.Med. 17: ,

52 Autonomic Dysfunction in Cardiovascular Disease. A consequence of vascular damage 115. Jensen T, Bjerre Knudsen J, Feldt Rasmussen B, Deckert T: Features of endothelial dysfunction in early diabetic nephropathy. Lancet 1: , Deckert T, Feldt Rasmussen B, Borch Johnsen K, Jensen T, Kofoed Enevoldsen A: Albuminuria reflects widespread vascular damage. The Steno hypothesis. Diabetologia 32: , Clarke CF, Eason M, Reilly A, Boyce D, Werther GA: Autonomic nerve function in adolescents with Type 1 diabetes mellitus: relationship to microalbuminuria. Diabet.Med. 16: , Smulders YM, Jager A, Gerritsen J, Dekker JM, Nijpels G, Heine RJ, Bouter LM, Stehouwer CD: Cardiovascular autonomic function is associated with (micro )albuminuria in elderly Caucasian sujects with impaired glucose tolerance or type 2 diabetes: the Hoorn Study. Diabetes Care 23: , Messent JW, Ellio TG, Hill RD, Jarre RJ, Keen H, Viberti GC: Prognostic significance of microalbuminuria in insulin dependent diabetes mellitus: a twenty three year follow up study. Kidney Int. 41: , Sundkvist G, Lilja B: Autonomic neuropathy predicts deterioration in glomerular filtration rate in patients with IDDM. Diabetes Care 16: , Weinrauch LA, Kennedy FP, Gleason RE, Keough J, D Elia JA: Relationship between autonomic function and progression of renal disease in diabetic proteinuria: clinical correlations and implications for blood pressure control. Am.J.Hypertens. 11: , Foss CH, Vestbo E, Froland A, Gjessing HJ, Mogensen CE, Damsgaard EM: Autonomic neuropathy in nondiabetic offspring of type 2 diabetic subjects is associated with urinary albumin excretion rate and 24 h ambulatory blood pressure: the Fredericia Study. Diabetes 50: , Po s JT, Hatanaka T, Shoukas AA: Effect of arterial compliance on carotid sinus baroreceptor reflex control of the circulation. Am.J.Physiol 270:H , Lage SG, Polak JF, O Leary DH, Creager MA: Relationship of arterial compliance to baroreflex function in hypertensive patients. Am.J.Physiol 265:H232 H237,

53 Chapter Bonyhay I, Jokkel G, Kollai M: Relation between baroreflex sensitivity and carotid artery elasticity in healthy humans. Am.J.Physiol 271:H1139 H1144, Brown AM, Saum WR, Tuley FH: A comparison of aortic baroreceptor discharge in normotensive and spontaneously hypertensive rats. Circ.Res. 39: , Brunner MJ: Carotid sinus compliance and baroreflex responses in hypertensive dogs. Am.J.Hypertens. 4: , Xie PL, Chapleau MW, McDowell TS, Hajduczok G, Abboud FM: Mechanism of decreased baroreceptor activity in chronic hypertensive rabbits. Role of endogenous prostanoids. J.Clin.Invest 86: , Li Z, Abboud FM, Chapleau MW: Aggregating human platelets in carotid sinus of rabbits decrease sensitivity of baroreceptors. Circ.Res. 70: , Li Z, Mao HZ, Abboud FM, Chapleau MW: Oxygen derived free radicals contribute to baroreceptor dysfunction in atherosclerotic rabbits. Circ.Res. 79: , Schwartz PJ, Zaza A, Pala M, Locati E, Beria G, Zanche i A: Baroreflex sensitivity and its evolution during the first year after myocardial infarction. J.Am.Coll.Cardiol. 12: , La Rovere MT, Pinna GD, Maestri R, Mortara A, Capomolla S, Febo O, Ferrari R, Franchini M, Gnemmi M, Opasich C, Riccardi PG, Traversi E, Cobelli F: Short term heart rate variability strongly predicts sudden cardiac death in chronic heart failure patients. Circulation 107: , La Rovere MT, Pinna GD, Hohnloser SH, Marcus FI, Mortara A, Nohara R, Bigger JT, Jr., Camm AJ, Schwartz PJ: Baroreflex sensitivity and heart rate variability in the identification of patients at risk for life threatening arrhythmias: implications for clinical trials. Circulation 103: , Eckberg DL, Sleight P. Congestive heart failure. Human Baroreflexes in Health and Disease In: Eckberg DL, Sleight P, eds. Oxford, UK: Clarendon Press. Ref Type: Generic 135. Wang W, Chen JS, Zucker IH: Carotid sinus baroreceptor sensitivity in experimental heart failure. Circulation 81: ,

54 Autonomic Dysfunction in Cardiovascular Disease. A consequence of vascular damage 136. Dibner Dunlap ME, Thames MD: Baroreflex control of renal sympathetic nerve activity is preserved in heart failure despite reduced arterial baroreceptor sensitivity. Circ.Res. 65: , Niebauer M, Zucker IH: Static and dynamic responses of carotid sinus baroreceptors in dogs with chronic volume overload. J.Physiol 369: , Porter TR, Eckberg DL, Fritsch JM, Rea RF, Beightol LA, Schmedtje JF, Jr., Mohanty PK: Autonomic pathophysiology in heart failure patients. Sympathetic cholinergic interrelations. J.Clin.Invest 85: , Sartori C, Scherrer U: Insulin, nitric oxide and the sympathetic nervous system: at the crossroads of metabolic and cardiovascular regulation. J.Hypertens. 17: , Sartori C, Lepori M, Scherrer U: Interaction between nitric oxide and the cholinergic and sympathetic nervous system in cardiovascular control in humans. Pharmacol.Ther. 106: , Lepori M, Sartori C, Trueb L, Owlya R, Nicod P, Scherrer U: Haemodynamic and sympathetic effects of inhibition of nitric oxide synthase by systemic infusion of N(G) monomethyl L arginine into humans are dose dependent. J.Hypertens. 16: , Sakima A, Teruya H, Yamazato M, Matayoshi R, Muratani H, Fukiyama K: Prolonged NOS inhibition in the brain elevates blood pressure in normotensive rats. Am.J.Physiol 275:R410 R417, Lembo G, Napoli R, Capaldo B, Rendina V, Iaccarino G, Volpe M, Trimarco B, Sacca L: Abnormal sympathetic overactivity evoked by insulin in the skeletal muscle of patients with essential hypertension. J.Clin.Invest 90:24 29, Vollenweider P, Tappy L, Randin D, Schneiter P, Jequier E, Nicod P, Scherrer U: Differential effects of hyperinsulinemia and carbohydrate metabolism on sympathetic nerve activity and muscle blood flow in humans. J.Clin.Invest 92: , Randin D, Vollenweider P, Tappy L, Jequier E, Nicod P, Scherrer U: Effects of adrenergic and cholinergic blockade on insulin induced stimulation of calf blood flow in humans. Am.J.Physiol 266:R809 R816,

55 Chapter Duplain H, Burcelin R, Sartori C, Cook S, Egli M, Lepori M, Vollenweider P, Pedrazzini T, Nicod P, Thorens B, Scherrer U: Insulin resistance, hyperlipidemia, and hypertension in mice lacking endothelial nitric oxide synthase. Circulation 104: , Steinberg HO, Brechtel G, Johnson A, Fineberg N, Baron AD: Insulin mediated skeletal muscle vasodilation is nitric oxide dependent. A novel action of insulin to increase nitric oxide release. J.Clin.Invest 94: , Sartori C, Trueb L, Nicod P, Scherrer U: Effects of sympathectomy and nitric oxide synthase inhibition on vascular actions of insulin in humans. Hypertension 34: , Lepori M, Sartori C, Duplain H, Nicod P, Scherrer U: Interaction between cholinergic and nitrergic vasodilation: a novel mechanism of blood pressure control. Cardiovasc.Res. 51: , Jumrussirikul P, Dinerman J, Dawson TM, Dawson VL, Ekelund U, Georgakopoulos D, Schramm LP, Calkins H, Snyder SH, Hare JM, Berger RD: Interaction between neuronal nitric oxide synthase and inhibitory G protein activity in heart rate regulation in conscious mice. J.Clin.Invest 102: , Choate JK, Danson EJ, Morris JF, Paterson DJ: Peripheral vagal control of heart rate is impaired in neuronal NOS knockout mice. Am.J.Physiol Heart Circ.Physiol 281:H2310 H2317, Danson EJ, Paterson DJ: Enhanced neuronal nitric oxide synthase expression is central to cardiac vagal phenotype in exercise trained mice. J.Physiol 546: , Brunner F, Andrew P, Wolkart G, Zechner R, Mayer B: Myocardial contractile function and heart rate in mice with myocyte specific overexpression of endothelial nitric oxide synthase. Circulation 104: , Chowdhary S, Vaile JC, Fletcher J, Ross HF, Coote JH, Townend JN: Nitric oxide and cardiac autonomic control in humans. Hypertension 36: , Chowdhary S, Ng GA, Nu all SL, Coote JH, Ross HF, Townend JN: Nitric oxide and cardiac parasympathetic control in human heart failure. Clin.Sci.(Lond) 102: , Scherrer U, Randin D, Tappy L, Vollenweider P, Jequier E, Nicod P: Body fat and sympathetic nerve activity in healthy subjects. Circulation 89: ,

56 Autonomic Dysfunction in Cardiovascular Disease. A consequence of vascular damage 157. Grassi G, Dell Oro R, Quarti Trevano F, Scopelliti F, Seravalle G, Paleari F, Gamba PL, Mancia G: Neuroadrenergic and reflex abnormalities in patients with metabolic syndrome. Diabetologia 48: , Singh JP, Larson MG, O Donnell CJ, Wilson PF, Tsuji H, Lloyd Jones DM, Levy D: Association of hyperglycemia with reduced heart rate variability (The Framingham Heart Study). Am. J.Cardiol. 86: , Laitinen T, Vauhkonen IK, Niskanen LK, Hartikainen JE, Lansimies EA, Uusitupa MI, Laakso M: Power spectral analysis of heart rate variability during hyperinsulinemia in nondiabetic offspring of type 2 diabetic patients: evidence for possible early autonomic dysfunction in insulin resistant subjects. Diabetes 48: , Pikkujamsa SM, Huikuri HV, Airaksinen KE, Rantala AO, Kauma H, Lilja M, Savolainen MJ, Kesaniemi YA: Heart rate variability and baroreflex sensitivity in hypertensive subjects with and without metabolic features of insulin resistance syndrome. Am.J.Hypertens. 11: ,

57 Chapter 2 48

58 Part II: Diabetes mellitus

59

60 Chapter 3 Baroreflex sensitivity is depressed in Microalbuminuric Type 1 diabetic patients at rest and during sympathetic manoeuvres Diabetologia. 1999;42: J.D. Lefrandt 1, K.H. Hoogenberg 2, A.M. van Roon 1, R.P.F. Dullaart 2, R.O.B. Gans 1, A.J. Smit 1 Department of Internal Medicine, Divisions of Angiology and General Medicine 1 and Endocrinology 2, State University Hospital, Groningen, The Netherlands.

61 Chapter 3 Summary Aims/hypothesis. To evaluate baroreflex sensitivity (BRS) in micro and normoalbuminuric Type 1 (insulin dependent) diabetic patients without autonomic neuropathy and in healthy controls. Methods. 15 Microalbuminuric Type 1 diabetic patients were matched for age, gender, body mass index (BMI) and smoking habits with 15 normoalbuminuric patients and with 15 healthy controls. All studied subjects had a blood pressure <160/95 mmhg, a BMI <30 kg/m 2 and normal autonomic function on standard tests. Blood pressure and heart rate were measured non invasively (Finapres) at rest and during sympathetic activation (handgrip, mental stress, standing). The BRS was defined as the mean gain between blood pressure variability and heart rate variability in the Hz frequency band. Results. Resting BRS was decreased in the microalbuminuric patients (3.5 ± 0.4 msec/mmhg) compared with the normoalbuminuric patients and the healthy subjects (7.6 ± 1.6 and 9.5 ± 1.1 msec/mmhg, respectively, p<0.001). The sympathetic tests reduced BRS similarly in the groups without changing the between group differences. Conclusion/interpretation. BRS is reduced in Type 1 diabetic patients with microalbuminuria but without autonomic neuropathy. A prospective study should indicate whether this early abnormality in cardiovascular reflex function is a risk factor of cardiovascular mortality in these patients. [Diabetologia (1999) 42: ] Introduction The presence of microalbuminuria indicates an increased risk of cardiovascular disease in patients with Type 1 (insulin dependent) diabetes mellitus 1. Clinical autonomic neuropathy adds to a particular poor prognosis in these patients 2 5. Because autonomic neuropathy is highly prevalent among proteinuric patients, it could in part account for the increased mortality rate associated with diabetic nephropathy 6,7. In addition, autonomic neuropathy may play a role in the development and the progression of proteinuria 8. Abnormal cardiovascular reflexes predict a lower than expected glomerular filtration rate 9,10, and the presence of autonomic neuropathy blunts the nocturnal fall in blood 52

62 Baroreflex sensitivity is depressed in micro albuminuric Type 1 diabetic patients at rest and during sympathetic manoeuvres pressure and albumin excretion rate 11. Baroreflex sensitivity (BRS) measurements are highly sensitive to detect early abnormalities in cardiovascular autonomic function 12,13. BRS has been reported to be impaired in uncomplicated Type 1 diabetic patients 13 15, but the clinical significance of a decline in BRS in these patients is not yet clear. It could be a risk indicator of cardiovascular mortality, in analogy to observations in post myocardial patients 16,17. An increase in the carotic wall thickness 18 and a reduced distensibility 19 have both been reported in microalbuminuric Type 1 diabetic patients and could affect the function of baroreceptors in the carotic wall. An impaired cardiac vagal function 6 and a diminished endothelial function 20 are other abnormalities reported in microalbuminuric Type 1 diabetic patients that could also contribute to an abnormal baroreflex function in these patients. In the present study, we evaluated the possibility of an abnormal baroreflex function in microalbuminuric Type 1 diabetic patients. Therefore, we assessed BRS in matched micro and normoalbuminuric patients without autonomic neuropathy and healthy subjects at rest and during sympathetic manoeuvres. Subjects and methods Subjects. All subjects consented to participate in the study, which was approved by the local medical ethics commi ee. Eligible subjects had an age between 30 and 65 years, a serum creatinine <120 µmol/l, a body mass index <30 kg/m 2, a systolic blood pressure <160 mmhg and a diastolic blood pressure <95 mmhg, and no signs of autonomic neuropathy as assessed by standard tests 21. Autonomic function was evaluated by the Valsalva maneuver (15 sec breathing with 40 mmhg counter pressure), beat to beat variation during 6 cycles of maximal in and expiration, the drop in systolic blood pressure and the 30:15 ratio of the change in heart rate after standing 21. Results were compared to age adjusted normalised values 22 and autonomic neuropathy was considered to be present when 2 or more tests were abnormal 21. The diabetic patients were considered insulin depend ent because of ketosis prone diabetes and a disease onset before 35 yr. In case of doubt, insulin deficiency was confirmed by a glucagon stimulated plasma 53

63 Chapter 3 Table 1. Clinical characteristics of the study groups. Microalbuminuric Normoalbuminuric Control Diabetic patients Diabetic patients subjects Number Age (yr) 50 ± 3 49 ± 3 47 ± 3 Duration of disease (yr) 28 ± 2 28 ± 3 Gender (males/females) 11/4 11/4 11/4 Cigare e smokers (n) Body mass index (kg/m 2 ) 26.3 ± ± ± 0.8 Retinopathy (A/B/P) 1/8/6 a 8/5/2 Systolic blood pressure (mmhg) 142 ± ± ± 4 Diastolic blood pressure (mmhg) 75 ± 2 73 ± 3 79 ± 2 Pulse rate (beats/min) 78 ± 3 b 67 ± 3 67 ± 3 Serum creatinine (µmol/l) 88 ± 6 84 ± 3 85 ± 3 Urinary albumin excretion (µg/min) 54.9 ± 9.6 c 4.3 ± ± 0.6 Glycated haemoglobin (%) 8.2 ± ± ± 0.1 d Insulin dose (U. kg 1. day 1 ) 0.8 ± ± 0.1 Valsalva ratio 1.38 ± ± ± 0.07 Beat to beat variation during 12 ± 1 14 ± 2 15 ± 1 deep breathing Drop in systolic blood pressure 14 ± 2 12 ± 2 12 ± 2 to standing (mmhg) 30:15 Ratio after standing 1.16 ± ± ± 0.06 Values are given in mean ± SEM, except urinary albumin excretion rate which is given in geometric mean ± antilog SEM (not normal distributed). Retinopathy A: absent, B: background, P: proliferative a p<0.05 vs normoalbuminuric diabetic patients; b p<0.02 and c p<0.001 vs normoalbuminuric diabetic patients and control subjects; d p<0.001 vs normo and microalbuminuric diabetic patients. C peptide level <0.2 nmol/l. They had a duration of disease of at least 10 years and were categorised for the presence of microalbuminuria. Microalbuminuria was defined as an urinary albumin excretion rate between 20 and 200 µg/min in at least 2 of the 3 overnight urine collections in the preceding year. Patients with a history of coronary heart disease, peripheral vascular disease, and the use of beta or calcium channel blockers were excluded. Fifteen microalbuminuric, 15 normoalbuminuric Type 1 diabetic patients and 15 healthy control subjects were included in the study. They were matched for gender, age (within 5 years), body mass index (within 2.5 kg/m 2 ) and smoking habits (Table 1). Metabolic control and daily insulin requirement was similar in the normo and microalbumi nuric diabetic groups. Systolic and diastolic blood pressure were not essentially different between the groups, while pulse rate was higher in the microalbuminuric compared to normoalbuminuric patients and healthy control subjects (p<0.02, Table 1). W studied five microalbumi nuric and two normoalbuminuric patients while using an angiotensin converting enzyme (ACE) inhibitor. Of these, three microalbuminuric and one normoalbuminuric patients were additionally treated with diuretics. Because of their participation in a study requiring discontinuation of medication ten microalbuminuric patients stopped their ACE inhibition and/or diuretic mediations 4 to 6 weeks before the measurements 54

64 Baroreflex sensitivity is depressed in micro albuminuric Type 1 diabetic patients at rest and during sympathetic manoeuvres Baroreflex sensitivity measurements during rest and sympathetic activation. The studies were performed 2 h after breakfast at which the patients had taken their usual morning insulin dose. The measurements were carried out in a quiet temperature controlled (22 o C) room with the participants in the supine position. Smoking and caffeine bearing drinks were not allowed during the study day. After a 30 min rest, a Finapres cuff (Finapres, Ohmeda 2300, Englewood, Colo., USA) was applied to the midphalanx of the third finger for continuous blood pressure and heart rate monitoring. After 20 min of baseline measurements, the subjects underwent a further ba ery of autonomic function tests. These included isometric handgrip (3 min isometric contraction at 30% of maximum strength), mental stress (3 min of arithmetic calculations) and standing (5 min standing in upright position). Baroreflex sensitivity was determined by the transfer function technique using the CARSPAN program, as described 23, 24. This program allows discrete Fourier transformation of non equidistant samples of blood pressure and RR interval series. The analysed time series are corrected for artefacts and checked for stationarity. Baroreflex sensitivity is defined as the mean modulus between spectral values of systolic blood pressure variability and heart rate variability in the Hz frequency band with a coherence of >0.3. Baroreflex sensitivity is expressed in ms/mmhg. A BRS of 10 ms/mmhg indicates that a rise of 1 mmhg in systolic blood pressure will induce 10 ms of RR interval lengthening. The coefficient of variation of this method is 13%. In the present study, resting BRS was determined from three periods of 100 to 300 sec. Modulation of BRS in response to the sympathetic tests was assessed from a stationary segment of at least 100 sec. Statistical analysis. Power spectral values have a skewed distribution that is normalised after logarithmic transformation. Therefore, the natural logarithm of BRS is used in the analyses. Within and between group differences were analysed with ANOVA and post hoc analysis with Bonferroni adjustment for multiple comparisons. Correlation were sought with Pearson tests, the independent contribution of various parameters was evaluated with multiple regression analysis. Data are given as means ± SEM, unless stated otherwise. The findings of Valsalva manoeuvre, beat to beat variation on deep breathing, 30:15 ratio and drop in systolic blood pressure after standing were ranked 55

65 Chapter 3 separately and the average rank was used as an arbitrary measure of overall autonomic function. A p value of <0.05 was considered significant. Results Resting supine BRS was reduced in microalbuminuric patients compared with normoalbuminuric patients and healthy controls (p<0.001, Figure 1). There was no numerical difference in BRS between diabetic patients studied with and without ACE inhibition therapy (Figure 1). Isometric handgrip, mental stress and standing induced similar changes in systolic pressure in the three groups (change in systolic blood pressure during handgrip: 35.4 ± 5.7, 31.6 ± 4.1, 28.8 ± 3.4 mmhg; during mental stress: 28.5 ± 3.3, 25.8 ± 2.6, 25.3 ± 3.2 mmhg in the micro, and normoalbuminuric diabetic patients and healthy control, respectively, p<0.001 for all; during standing: see Table 1). Baroreflex sensitivity decreased significantly during isometric handgrip and to standing, increased post standing (p<0.01 in all groups, Figure 2), but was not significantly altered by mental stress. The changes in BRS during handgrip, standing and post standing did not differ between the groups (Figure 2). Baseline BRS correlated significantly with age (r= 0.46), diabetes duration (r= 0.54), HbA 1c (r= 0.42), BMI (r= 0.40), overall autonomic function test score (r= 0.52), grade of retinopathy (defined as absent, background or proliferative) (r=0.57), heart rate (r= 0.58), systolic blood pressure (r= 0.48), and urinary albumin excretion rate (r=0.57, p<0.01 for all). Multiple regression analysis showed that resting heart rate contributed to 15% (p<0.001), urinary albumin excretion rate to 14% (p<0.001), age to 13% (p<0.001) and overall autonomic function to 7% (p<0.05) to the variance in BRS (multiple r=0.70, p<0.001), without statistically significant contributions of diabetes duration, HbA 1c, grade of retinopathy and systolic blood pressure (all p>0.20). Furthermore, multiple regression analysis demonstrate that the Valsalva ratio (r=0.45, p<0.001), beat to beat variation during deep breathing (r=0.46, p<0.001) and 30:15 ratio after standing (r=0.55, p<0.001) only were explained by the baseline BRS measurements without significant contributions of other clinical variables. Baseline BRS was not related 56

66 Baroreflex sensitivity is depressed in micro albuminuric Type 1 diabetic patients at rest and during sympathetic manoeuvres BRS (mm/mmhg zz 00 Microalbuminuric Normoalbuminuric Controls Type 1 diabetic Type 1 diabetic subjects patients patients Figure 1. Individual resting supine BRS measurements in micro, normoalbuminuric type 1 diabetic patients and healthy control subjects. measurements of patients using an angiotensin converting enzyme inhibitor. Horizontal lines indicate geometric means. * p<0.001 vs normoalbuminuric patients and healthy controls BRS (mm/mmhg Basal * handgrip * * mental stress standing post standing Figure 2. BRS measurements at rest, isometric handgrip, mental stress, standing and after standing in the microalbuminuric ( ), the normoalbuminuric ( ) type 1 diabetic patients and the healthy control subjects ( ). Data are represented as geometric means ± antilog SEM. * p<0.01 from resting supine BRS and ** p<0.01 from standing BRS. * * * ** ** ** to the blood pressure responses at handgrip, mental stress and standing testing. 57

67 Chapter 3 Discussion In the present study, baroreflex control of blood pressure was reduced in Type 1 diabetes mellitus complicated by microalbuminuria, but still showed the ability to modulate in response to sympathetic stress tests. The impaired BRS indicates that autonomic dysfunction is already present at an early stage of diabetic renal involvement even when there is no clinical autonomic neuropathy as defined by accepted criteria 21. In a multiple regression analysis, we found that microalbuminuria independently from well known other factors that influence BRS, like age and autonomic function 13 15, contributed to the reduction in BRS. Thus, the presence of microalbuminuria seems to be an important factor to take into consideration in BRS abnormalities of Type 1 diabetic patients. We did not find a numerical difference in baseline BRS between diabetic patients with and without ACE inhibition therapy. We can only speculate on the clinical relevance on the association between microalbuminuria and reduced BRS in the Type 1 diabetic patients. The aim of the baroreflex is short term blood pressure control by modulation of vagal and sympathetic influence on heart rate, cardiac contractility and peripheral vascular resistance. A sensitive system effectively responds to small rises in blood pressure with an increase in vagal activity and a decrease in sympathetic tone, resulting in a deceleration of heart rate, a diminished contractility and a reduction of peripheral resistance. A drop in blood pressure causes opposite responses. BRS is marker of the capability of the system to buffer blood pressure variations and a reduction in BRS has been found to be highly predictive of mortality in post myocardial infarct patients 16,17. The increased mortality rate of patients with a low BRS is a ributed to an increased vulnerability to develop lethal arrhythmias 16,17. Based on the observations in post myocardial infarction patients, we suggest our finding of a reduced BRS in conjunction with the presence of microalbuminuria in Type 1 diabetic patients could similarly indicate an increased cardiovascular risk. Indeed, the occurrence of microalbuminuria is associated with an increased risk of cardiovascular disease 1. Moreover, a higher incidence of sudden cardiac deaths has been reported in diabetic patients 25. A diminished BRS indicates an impaired responsiveness to blood pressure variations of the baroreflex system. Therefore, another consequence of the diminished BRS of the microal- 58

68 Baroreflex sensitivity is depressed in micro albuminuric Type 1 diabetic patients at rest and during sympathetic manoeuvres buminuric Type 1 diabetic patients could be an impairment of short term blood pressure control. However, we were unable to demonstrate an altered blood pressure responsiveness or variability despite the use of sympathetic stimulation tests. This does not exclude the possibility that microalbuminuric Type 1 diabetic patients are prone to larger blood pressure excursions during daily life activities when BRS is impaired. The suggestion warrants further investigation as a decreased systemic blood pressure buffering capacity could be of importance in patients with impaired renal haemodynamics 26. Baroreflex sensitivity was not essentially lower in the normoalbuminuric Type 1 diabetic patients compared to the healthy control subjects, despite some of these patients had distinct abnormal values. This finding is at variance with recent observations that showed a decreased BRS in uncomplicated Type 1 diabetic patients However, the methods to calculate BRS differed among these studies, as the α method 13, the sequential technique and the transfer function method [present study] were used. Despite these techniques having all been validated against classical pharmacological BRS assessments 24,27, the impact of the differences in BRS outcome in diabetic patients awaits further evaluation. The conclusion that BRS measurements unequivocally detect early abnormalities in cardiovascular autonomic function is illustrated by the findings of multiple regression analysis in which only the BRS, independent from other factors, contributed to the variance of the standard autonomic function tests. It is generally thought that standard autonomic function tests are hampered by a low discriminative power to detect abnormalities in cardiovascular autonomic function 6,8,16. Despite the accuracy to measure early abnormalities, modern analyses of cardiovascular rhythms and reflexes, like BRS assessments, have not yet been documented to be a risk factor of mortality in diabetic patients. This contrasts with the well documented power of autonomic function tests to predict mortality in diabetic patients 2 5. In conclusion, BRS is reduced in Type 1 diabetic patients complicated by microalbuminuria. Further investigations are warranted to assess the possible predictive power of a reduced BRS on mortality and progression of renal disease in these patients. 59

69 Chapter 3 Acknowledgements We are indebted to Marianne Bruin, Berta Buist, Anne van Gessel, Wietze Kuipers and Margreet Teune for their skilful laboratory assistance. 60

70 Baroreflex sensitivity is depressed in micro albuminuric Type 1 diabetic patients at rest and during sympathetic manoeuvres References 1. Messent JW, Ellio TG, Hill RD, Jarre RJ, Keen H, Viberti GC (1992) Prognostic significance of microalbuminuria in insulin dependent diabetes mellitus: a twenty three year follow up study. Kidney Int 41: Ewing DJ, Campbell IW, Clarke BF (1976) Mortality in diabetic autonomic neuropathy. Lancet i: O Brien IA, McFadden JP, Corrall RJM (1991) The influence of autonomic neuropathy on mortality in insulin dependent diabetes. Q J Med 79: Rathmann W, Ziegler D, Jahnke M, Haastert B, Gries FA (1993) Mortality in diabetic patients with cardiovascular autonomic neuropathy. Diabet Med 10: Reichard P, Pihl M (1994) Mortality and treatment side effects during long term intensified conventional insulin treatment in the Stockholm Diabetes Intervention Study. Diabetes 43: Mølgaard H, Christensen PD, Hermansen K, Serensen KE, Christensen KE, Mogensen CE (1994) Early recognition dysfunction in microalbuminuria: significance for cardiovascular mortality in diabetes mellitus? Diabetologia 37: Sawicki PT, Dahne R, Bender R, Berger M (1996) Prolonged QT interval as a predictor of mortality in diabetic nephropathy. Diabetologia 39: Spallone V, Menzinger G (1997) Diagnosis of cardiovascular autonomic neuropathy in diabetes. Diabetes 46: [Suppl 2] S67 S Sundkvist G, Lilja B (1993) Autonomic neuropathy predicts deterioration in glomerular filtration rate in patients with IDDM. Diabetes Care 16: Weinrauch LA, Kennedy FP, Gleason RE, Keough J, D Elia JA (1998) Relationship between autonomic function and progression of renal disease in diabetic proteinuria: clinical correlations and implications for blood pressure control. Am J Hypertens 11: Spallone V, Gambardella S, Maiello MR, Barini A, Frontoni S, Menzinger G (1994) Relationship between autonomic neuropathy, 24 h blood pressure profile and nephropathy in normotensive IDDM patients. Diabetes Care 17:

71 Chapter De Ferrari GM, Landolina M, Mantica M, Manfredini R, Schwartz PJ, Lo o A (1995) Baroreflex sensitivity, but not heart rate variability, is reduced in patients with life threatening ventricular arrhythmias long after myocardial infarction. Am Heart J 130: Fra ola A, Parati 6, Gamba P., Paleari F, Mauri 6, DiRienzo M, Castiglioni P, Mancia G (1997) Time and frequency domain estimates of spontaneous baroreflex sensitivity provide early detection of autonomic dysfunction in diabetes mellitus. Diabetologia 40: Weston P J, James MA, Panerai R, McNally PG, Po er JF, Thurston H, Swales JD (1996) Abnormal baroreceptor cardiac reflex sensitivity is not detected by conventional tests of autonomic function in patients with insulin dependent diabetes mellitus. Clin Sci 91: Weston P J, Panerai RB, McCullough A, McNally PG, James MA, Po er JF, Thurston H, Swales JD (1996) Assessment of baroreceptor cardiac reflex sensitivity using time domain analysis in patients with IDDM and the relation to left ventricular mass index. Diabetologia 39: Farrell TG, Odemuyiwa O, Bashir Y, Cripps TR, Malik M, Ward DE, Camm AJ (1992) Prognostic value of baroreflex sensitivity testing after acute myocardial infarction. Br Heart J 67: LaRovere, MT, Bigger JT, Marcus FI, Mortara A, Schwartz PJ (1998) Baroreflex sensitivity and heart rate variability in prediction of total cardiac mortality after myocardial infarction. Lancet 351: Frost D, Beischer W (1998) Determinants of carotic artery wall thickening in young patients with Type 1 diabetes mellitus. Diabet Med 15: Lambert J, Smulders RA, Aarsen RVT, Donker AJM, Stehouwer CDA (1998) Carotic artery stiffness is increased in microalbuminuric IDDM patients. Diabetes Care 21: Jensen T, Bjerre Knudsen J, Feldt Rasmussen B, Deckert T (1989) Features of endothelial dysfunc tion in early diabetic nephropathy. Lancet i: Ewing DJ, Clarke BF (1982) Diagnosis and management of diabetic autonomic neuropathy. Br Med J 285:

72 Baroreflex sensitivity is depressed in micro albuminuric Type 1 diabetic patients at rest and during sympathetic manoeuvres 22. Piha SJ (1991) Cardiovascular autonomic reflex tests: normal responses and age related reference values. Clin Physiol 11: Robbe HW, Mulder LJ, Ruddel H, Langewitz WA, Veldman JB, Mulder G (1997) Assessment of baroreceptor reflex sensitivity by means of spectral analysis. Hypertension. 10: Saul JP, Berger RD, Albrecht P, Stein SP, Chen MH, Cohen RJ (1991) Transfer function analysis of the circulation: unique insights into cardiovascular regulation. Am J Physiol. 261: H1231 H Ta ersall RB, Gill GV (1991) Unexplained death of type 1 diabetic patients. Diabet Med 8: Mogensen CE (1995) Diabetic renal disease: The quest for normotension and beyond. Diab Med 12: Watkins LL, Grossman P, Sherwood A (1996) Noninvasive assessment of baroreflex control in borderline hypertension. Hypertension 28:

73 Chapter 3 64

74 Chapter 4 Heart rate variability and baroreflex sensitivity in diabetic patients with peripheral neuropathy are indices of both neural and vascular disease. Submi ed. J.D. Lefrandt 2, J.H. van der Hoeven 4, A.M. van Roon 2, R.P.F Dullaart 3, R.O.B. Gans 2, A.J. Smit 2, K.H. Hoogenberg 1, Martini Hospital Groningen, Department of Internal Medicine 1 and University Hospital Groningen, Department of Internal Medicine, Divisions of General Medicine 2 and Endocrinology 3, Department of Neurology 4, Groningen, The Netherlands.

75 Chapter 4 Summary Aims: Cardiovascular autonomic neuropathy (CAN) assessed by conventional tests, is variably related to the presence of diabetic peripheral neuropathy (DPN). Power spectral analysis of heart rate variability (HRV) and noninvasive measurement of baroreflex sensitivity (BRS) are advanced measures of CAN. It is unknown whether HRV and BRS measurements are impaired in the presence of clinical DPN, and may be useful as indicators of an increased cardiovascular risk in these patients. Therefore, we studied HRV and BRS in carefully selected diabetic patients with and without DPN, and healthy controls. Methods: Eighteen diabetic patients with DPN (DN) were individually matched with 18 diabetic patients without DPN (DC) and healthy controls (C). Patients with cardiac and peripheral arterial occlusive disease were excluded. HRV and BRS were calculated by spectral analysis of continuous heart rate and blood pressure recordings by non invasive Finapres measurements. The HRV was analyzed among the different frequency bands, the BRS was defined as the mean gain between blood pressure variability and heart rate variability in the Hz frequency band. Results: Total power in HRV was lower in DN (7.1 ± 0.3 ln(ms 2 ) [mean ± SEM], P<0.001 for both) compared to DC (8.9 ± 0.3 ln(ms 2 ) and C (9.3± 0.3 ln(ms 2 )). BRS was similarly impaired in DN (3.0 ± 0.7 ms/mmhg, P< 0.01 for both) compared to DC (6.0 ± 0.6 ms/mmhg) and C (6.7 ± 1.1 ms/mmhg). In a multiple regression analysis, peripheral neuropathy and an elevated urinary albumin excretion rate (UAE) were the main determinants of a lower HRV, whereas the presence of an elevated UAE almost uniquely explained the variance BRS. It was further discovered that the lower HRV and BRS in patients with DN appeared to be limited to those patients who had concomitant elevated urinary albumin excretion (UAE) (HRV, 6.2 ± 0.4 ln(ms 2 ) and BRS, 1.7 ± 0.6 ms/mmhg, n=9) vs. normal UAE (HRV 8.0 ± 0.4 ln(ms 2 ) and BRS 5.2 ± 1.1 ms/mmhg, n=9, p<0.01 for both). Conclusions: Elevations in UAE are important independent determinants of a lower HRV and BRS in diabetic patients with peripheral neuropathy. We suggest that the presence of vascular abnormalities should to be taken into account while interpreting these measurements of cardiovascular reflex function. 66

76 Heart rate variability and baroreflex sensitivity in diabetic patients with peripheral neuropathy are indices of both neural and vascular disease. Introduction Foot ulceration in patients with diabetes mellitus is associated with a high morbidity and mortality (1). Distal peripheral neuropathy (DPN) and peripheral vascular disease play a dominant role in the multifactorial etiology of this complication (2). The coexistence of cardiovascular autonomic neuropathy (CAN) has been proposed to contribute to the high mortality rate in these patients, as the presence of CAN is a determinant of a poor prognosis in diabetic patients (3). There is, however, a variable relation between CAN and DPN, and CAN is frequently observed in the absence of DPN (4 6). This contrasts with the possibility that the sensory and motor nerves of the lower extremities and the unmyelinated post ganglionic sympathetic and parasympathetic vagal nerves to the heart are equally vulnerable to neuronal damage by the diabetic milieu. It has even been suggested that CAN and DPN are distinct entities with differences in pathogenesis as different risk factors have been described for the development of CAN and DPN (6,7). The lower than expected concordance between CAN and DPN may be also a ributable to the methodology employed to diagnose CAN (3). More advanced evaluation of CAN by power spectral analysis of heart rate variability (HRV) and non invasive measurement of baroreflex sensitivity (BRS) have shown to be highly sensitive to detect CAN at an early stage in diabetic patients (8 10). While HRV evaluates the efferent part of the baroreflex arc, BRS measures both its afferent and efferent function by cross spectral analysis of heart rate and blood pressure variations (9 12). Interestingly, HRV was evidently lowered patients with DPN and foot ulceration, and it was suggested that this parameter could be of additional value for cardiovascular risk stratification (13). In the present study, we compared HRV and BRS measurements in diabetic patients, carefully selected for the presence or absence of clinical DPN, and healthy control subjects. We evaluated the possibility that these measures of CAN are more consistently abnormal in diabetic patients in whom neuropathy is manifested at the feet than what has been reported for conventional CAN testing. 67

77 Chapter 4 Materials and Methods Subjects. The study consisted of three groups: 18 diabetic patients with peripheral neuropathy (DN), 18 diabetic patients without peripheral neuropathy (D), and 18 control subjects with normal glucose tolerance (C). Patients were recruited from the Diabetes Outpatient Clinic (University Hospital Groningen) and the Rehabilitation Center Beatrixoord Haren on the basis of hospital records were checked for previous foot ulceration of neuropathic in origin and. in whom peripheral vascular disease was not considered to have contributed to the foot ulcers. After this screening, they were recruited in a randomized order. The first group consisted of 24 diabetic patients known to have had neuropathic foot ulcers (DU group). These ulcers were purely neuropathic by origin, as confirmed by their localization (plantar surface of the foot at high pressure points) and the absence of peripheral arterial disease, as described below. In the second group, 24 diabetic patients without clinical neuropathy or foot ulcers (DC group) were included. To confirm this, the 10 g Semmes Weinstein monofilament was tested on the plantar surface of the hallux and central at the heel. The ability to correctly sense the monofilament in six trials on both locations was defined as normal, whereas the inability to sense the monofilament correctly in one or more trials was defined as disturbed. The third group consisted of 21 control subjects with normal glucose tolerance (C group). All groups were matched for sex and age (within 5 years), and the diabetic groups were also matched for duration and type of diabetes (type 1/type 2 diabetes; type 1 diabetes was considered on clinical grounds when the onset of the disease was a ketoacidosis or before the age of 40 years). Subjects with a history of or clinically apparent cardiac disease, with electrocardiographic abnormalities, or who used ß blockers or calcium antagonists were excluded. Peripheral arterial disease was excluded by normal ankle arm indexes (>0.90), toe arm indexes (>0.70), and normal plethysmography (crest time 0.22 s) in all groups. Normal glucose tolerance of the control subjects was demonstrated by a fasting capillary blood glucose <6.1 mmol/l and a blood glucose <7.8 mmol/l 2 h after a 75 g oral glucose tolerance test. Details of the clinical characteristics of each group are given in 68

78 Heart rate variability and baroreflex sensitivity in diabetic patients with peripheral neuropathy are indices of both neural and vascular disease. The study was approved by the local medical ethics commi ee and wri en informed consent was obtained from all participants after explanation of the purpose of the study. The patients were recruited from the outpatient clinic of the University Hospital Groningen and the Rehabilitation Centre Beatrixoord. Healthy control subjects were recruited by an advertisement in a local newspaper. The study consisted of three groups: 18 diabetic patients with peripheral neuropathy (DN), 18 diabetic patients without peripheral neuropathy (D), and 18 control subjects with normal glucose tolerance (C). All participants were recruited from the Diabetes Outpatient Clinic (University Hospital of Groningen) and from the Beatrixoord Rehabilitation Center in Haren, The Netherlands. Three groups of subjects were studied. The first group consisted of 18 diabetic neuropathic patients with a previous history of neuropathic foot ulceration (group DN); the second group of 18 diabetic patients without neuropathy (group DC); and the third group of 18 control subjects with normal glucose tolerance (group C). All groups were individually matched for sex and age (within 5 yr), and the diabetic groups for type of diabetes (type 1/ type 2) as well. Eligible subjects were 35 to 70 years of age. Subjects with a history of or clinically apparent cardiac disease, electrocardiographic abnormalities or using betablockers or calcium antagonists were excluded. Peripheral arterial disease was excluded by normal ankle arm indices (>0.90), toe arm indices (>0.70) and normal plethysmography (crest time 0.22 sec). Normal glucose tolerance of the control subjects was demonstrated by a fasting capillary blood glucose < 6.1 mmol/l and a blood glucose < 7.8 mmol/l 2h after a 75 gr oral glucose tolerance test. The protocol was approved by the local Ethical Commi ee and all participants gave wri en informed consent. Blood glucose was measured on APEC glucose analyzer (APEC Inc., Danvers, MA, USA), HbA1c was measured by HPLC (Bio Rad, Veenendaal, The Netherlands, normal range 4.6 to 6.1%), serum creatinine and cholesterol by a SMA(C) autoanalyzer (Technicon Instruments Inc. Tarrytown, N.Y., USA), and urinary albumin by radioimmuno assay (Diag nostic Products Corporation, Apeldoorn, The Netherlands). A urinary albumin creatinine ratio lower than 2.5 mg/µmol in fresh voided morning urine excluded 69

79 Chapter 4 Table 1 Clinical characteristics of study subjects. Neuropathic Diabetic Healthy diabetic control controls subjects subjects (DN) (DC) (C) N Male/Female 9/9 9/9 9/9 Age (years) 57 ± 3 56 ± 2 57 ± 2 Systolic blood pressure (mmhg) 146 ± ± ± 3 Diastolic blood pressure (mmhg) 80 ± 2 81 ± 2 86 ± 2 Number of antihypertensives (0/1/2) 7 / 7 / 4* 9 / 7 / 2H 17 / 1 / 0 Body mass index (kg/m 2 ) 30.4 ± ± ± 1.3 Smokers (yes/no) 10/8 6/12 5/13 Type of diabetes (1/2) 4/14 4/14 / Diabetes duration (years) 17 ± 3 12 ± 2 / HbA 1c (%) 8.4 ± 0.3I 7.6 ± ± 0.1 Blood glucose (mmol/l) 9.8 ± ± 0.6H 5.8 ± 0.2 Serum creatinine (mmol/l) 96 ± 5 89 ± 3 89 ± 2 Normo / micro / albuminuria 9/7/2 * 17/1/0 18/0/0 Urinary albumin excretion rate # 164 (54 616) 102 (mg/day) in albuminuric patients (n=9) (n=1) Retinopathy (absent/background/proliferative) 4/10/4 13/5/0H 18/0/0 Serum cholesterol (mmol/l) 4.9 ± 0.2H 5.0 ± 0.2H 6.2 ± 0.4 Patients on cholesterol lowering agents DNE 8.9 ± 0.5& 1.2 ± ± 0.2 NSS 5.6 ± 0.7& 0.8 ± ± 0.4 Semmes Weinstein monofilaments 6.6 ± 0.1& 5.1 ± ± 0.1 Hand: motor conduction n. medianus Measurable 17/18 16/18 18/18 Velocity (ms) 47.8 ± 1.3& 54.5 ± ± 1.1 Amplitude (mv) 7.2 ± 0.6 I 11.1 ± ± 0.8 Hand: sensory conduction n. medianus Measurable 17/18 15/18 18/18 Velocity (ms) 32.4 ± 1.5& 40.6 ± ± 1.6 Amplitude (mmv) 8.5 ± 1.7& 30.4 ± ± 3.7 Foot: motor conduction n. peroneus Measurable 18/18 18/18 18/18 Velocity (ms) 44.2 ± 1.6& 56.0 ± ± 2.0 Amplitude (mv) 6.3 ± 0.8*& 10.5 ± ± 0.5 Foot: sensory conduction n. suralis Measurable 9/18 17/18 18/18 Velocity (ms) 37.8 ± ± ± 0.8 Amplitude (mmv) 1.8 ± ± ± 1.3 Peripheral neuropathy score 3.0 ± 0.2& 1.3 ± ± 0.2 Data are number or means ± SE, # urinary albumin excretion is given as median (range). Microalbuminuria denotes an urinary albumin excretion of mg/day, albuminuria was present in 2 patients but did not exceed 1000 mg/day. DNE: diabetic neuropathy examination score, NSS: neurological symptom score. Only of measurable nerves, the mean velocities and amplitudes are given. If there was no detectable nerve function, the nerve was classified as abnormal. Peripheral neuropathy score is the number of abnormal nerves (conduction velocity or amplitude) of the n. medianus (motor), n. medianus (sensory), n. peroneus and n. suralis. *p<0.01 vs. C, Hp<0.05 vs. C, Ip<0.05 vs. DC, p<0.001 vs. C, p<0.01 vs. DC, &p<0.001 vs. DC. microalbuminuria. In case of a positive test, a 24h urine collection was made to quantify albumin excretion (elevated when >30 mg/day). Details of the clinical characteristics of each group are given in Table 1. All groups were comparable for sex, age, systolic and diastolic blood pressure, body mass index and number of smokers. The diabetic 70

80 Heart rate variability and baroreflex sensitivity in diabetic patients with peripheral neuropathy are indices of both neural and vascular disease. groups were comparable for type and duration of diabetes. HbA 1c was higher and retinopathy was more severe in the DN compared to the DC group. Elevated UAE was highly prevalent in DN; 7 patients had microalbuminuria ( mg/day) and 2 had albuminuria between mg/day (Table 1). Blood glucose was higher in the diabetic groups. The use of antihypertensive drugs was more frequent in the diabetic patients. Cholesterol lowering agents were also more frequently used in both groups of diabetic patients. As a result, serum total cholesterol was lower in both groups of diabetic patients compared to healthy subjects. Assessment of neuropathy. Diabetic neuropathy was diagnosed according to the San Antonio Consensus Statement criteria (14). Clinical neuropathy signs were scored by means of the Diabetic Neuropathy Examination (DNE) score (15) that has a maximum score of 16 and is a modification of Neuropathy Disability Score. Neurological symptoms were scored by a modification of the Neurological Symptom Score (NSS) (16) with a maximum score of 12. Quantitative sensory testing was performed with Semmes Weinstein monofilaments. Motor nerve conduction velocities and amplitudes were measured in the median and peroneal nerves (tibialis anterior), and sensory nerve conduction velocities and amplitudes in the median (first digit) and sural nerves. An overall peripheral neuropathy score (PNS) was defined as the number of these four nerves that had an abnormal conduction velocity and amplitude, ranging from 0 (all normal) to 4 (all abnormal). Since it was the goal of the present study to evaluate cardiovascular autonomic function, autonomic neuropathy was not assessed before selection of the patients. The neuropathy scores are given in Table 1. As expected from the selection procedure, severe neuropathy was present in the DN group, while neuropathy scores did not differ between the DC and C groups. Evaluation of cardiovascular autonomic function. Cardiovascular autonomic function was assessed by analysis of heart rate variability (HRV) and baroreflex sensitivity (BRS). All participants were studied in the late morning, 2 hours after they had used a light breakfast and the diabetic patients taken their oral hypoglycemic agents or insulin dose. All measurements took place in a quiet room with the temperature kept constant at 22 o C. Blood pressure was monitored by a Finapres (Ohmeda 2300, Inglewood, Col., USA) and heart rate by an ECG monitor 71

81 Chapter 4 (Hewle Packard 78351T, Palo Alto, Ca., USA). After 30 min of supine rest, the Finapres and ECG signal were sampled at 100 Hz and stored on a personal computer during 15 min. Offline, 300 seconds of each recording was analyzed by the CARSPAN program (IEC ProGamma, Groningen, the Netherlands), as described previously (12,17). After artifact correction and stationarity check, discrete Fourier transformation of systolic blood pressure and RR interval length was performed. HRV was analysed in accordance with the guidelines of the Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology (18). The power spectrum of HRV was divided into very low frequency (VLF) ( Hz), low frequency (LF) ( Hz) and high frequency (HF) ( Hz) bands and expressed in ln(ms 2 ) (12). The total power frequency band (TP) was defined as Hz. BRS was determined by the transfer function method and defined as the mean modulus between systolic blood pressure and heart rate variability in the Hz frequency band with at least 0.5 coherence, expressed in ms/mmhg (10,12,17) Heart Rate Variability ln(ms 2 ) * * * * 4 DN DC C Figure 1. Heart rate variability in diabetic patients with peripheral neuropathy (DN), diabetic patients without peripheral neuropathy (DC) and healthy control subjects (C). Data are shown as mean (SEM). Very Low Frequency Power (VLF, Hz), Low Frequency Power (LF, Hz) High Frequency Power (HF, Hz), Total Power (TP, Hz). *p<0.001 versus DC and versus C. 72

82 Heart rate variability and baroreflex sensitivity in diabetic patients with peripheral neuropathy are indices of both neural and vascular disease. Statistical analysis. The Statistical Package for Social Sciences for Windows version 9.0 (SPSS, Chicago) was used for the statistical analysis. Group differences were sought with analysis of variance and post hoc comparisons with Bonferroni correction. The spectral parameters that have a χ 2 distribution and BRS that has an F distribution were normalized by log transformation. For nonparametrically distributed data, the Kruskal Wallis test was used. The χ 2 test was used for comparison of dichotomous variables. Multiple linear regression was used to evaluate the independent contribution of clinical variables to the variance in HRV and BRS. A P value 0.05 was considered significant. Results The total power, an overall index of HRV, was lower in the DN group compared to the DC and C groups (DN 7.1 ± 0.3 vs DC 8.9 ± 0.3 and C 9.3± 0.3 ln(ms 2 ), P<0.001 for both, Figure 1). These differences were present in all frequency ranges, VLF, LF and HF (DN < D and C, P<0.001 for all frequency ranges, Figure 1). Similar, BRS was depressed in the DN group (3.0 ± 0.7) compared to the DC group (6.0 ± 0.6, P=0.003) and the C group (6.7 ± 1.1 ms/mmhg, P=0.001). Multiple regression analysis in the combined groups (n=54), disclosed that an elevated UAE (P<0.001), peripheral neuropathy score (P<0.02), diastolic blood pressure (P<0.05) and blood glucose level (P<0.05) independently contributed to 22%, 8%, 4% and 4%, respectively, to the variance in HRV (r 2 =0.89, P<0.001). The analysis showed that an elevated UAE (P<0.001) and blood glucose level (P<0.02) contributed to 27% and 8%, respectively, to the variance in BRS (r 2 =0.62, P<0.001), while diastolic blood pressure (P=0.07) and peripheral neuropathy score (P=0.20) did not significantly contribute to the model. The large effects of the presence of an elevated UAE on the HRV and BRS measurements is further illustrated in Figure 2A, B. The HRV (total power) in the patients of the DN group with an elevated UAE (6.2 ± 0.4 ln(ms 2 ), n=9) differed from the patients in the DN group with a normal UAE (8.0 ± 0.4 ln(ms 2 ), n=9, P<0.01, Figure 2A). Similarly, BRS in the patients of the DN group with an elevated UAE (1.7 ± 0.6, n=9) differed from the patients in the DN group with a normal UAE (5.2 ± 1.1 ms/mmhg, n=9, P<0.01, Figure 2B). 73

83 Chapter 4 15 (A) Heart Rate Variability 10 ln(ms 2 ) DN DC C (B) Baroreflex Sensitivity ms/mmhg DN DC C Figure 2. Heart rate variability (total power) and baroreflex sensitivity in diabetic patients with peripheral neuropathy (DN), diabetic patients without peripheral neuropathy (DC) and healthy control subjects (C), patient with normal and an elevated urinary albumin excretion. Discussion The present study shows that cardiovascular autonomic function, assessed by HRV and BRS, is impaired in diabetic patients with peripheral neuropathy, and is in agreement with a previous report on HRV in such a patient group (13). Interestingly, multiple regression analysis indicated 74

84 Heart rate variability and baroreflex sensitivity in diabetic patients with peripheral neuropathy are indices of both neural and vascular disease. that HRV and BRS were particular abnormal in the patients in whom peripheral neuropathy was complicated by an elevated UAE. Thus independent from compounding variables, like duration of diabetes, blood glucose, HbA1c, blood pressure and need for antihypertensive medication and degree of retinopathy, which all directed towards a more severely complicated diabetes in the neuropathic patients, the presence of an elevated UAE allocated for abnormalities in HRV and BRS. Although not anticipated on beforehand, elevations in UAE were highly prevalent in the diabetic patient with neuropathy, and this notion is in line with recent observations showing that a large portion of patient with abnormal CAN tests have elevated UAE levels (7,10,19). Also, in the study Aso et al. (13), the DPN patients had an elevated UAE though this but was not further considered. It is unknown, how abnormal CAN tests either by conventional means (7,19) or by the presently used HRV and BRS (10) pathogenetically relate to elevations in UAE. Elevations in UAE are associated with cardiovascular risk factors like dyslipidemia (20), hypertension (21), insulin resistance (22), increased intima media thickness and stiffness of the carotid arteries (23) and endothelial dysfunction (24), and are conceptually seen to reflect generalized vascular disease (25,26). The presence of vascular disease as indicated an elevated UAE may have affected the HRV and BRS. The baroreflex input is derived from blood pressure changes sensed by baroreceptors in the carotid arteries and aorta, and its output is modulation of heart rate, myocardial contractility and peripheral arterial resistance (17,27). Thus measurement of HRV and BRS do to certain extent depend on vessel wall properties. Subclinical atherosclerosis at the site of the baroreceptors, a diminished arterial compliance (23), subclinical cardiac contraction abnormalities (28,29) and an impaired vascular endothelial function (24) may all have contributed to the impairments in HRV and BRS, as these abnormalities have all been documented in the presence of UAE elevations in diabetic patients. Conversely, it is also possible that CAN precedes elevations in UAE and increases the likelyhood of the development of microalbuminuria, as this has previously been suggested (10). Endothelial dysfunction could be an important denominator (30,31). The fact that neuropathy score related to HRV and not to BRS in the multiple regression analysis may be due to the fact that HRV is more close to autonomic nerve function than BRS, as HRV assesses heart rate modulations by the efferent vagal and the sympathetic nerves. It further 75

85 Chapter 4 suggests that BRS, that evaluates both blood pressure and heart rate variations more than HRV depends on vascular function. However, elevations in UAE were present in a small group of patients, in the DPN group 2 had albuminuria, 7 had microalbuminuria and 9 had normo albuminuria, and in the diabetic group without DPN, only 1 was found to have microalbuminuria. Therefore, we cannot exclude the possibility that the lack of independent association between neuropathy score and BRS was due to the rather restricted sample size of this post hoc study observation. Finally the relation between a lower HRV and BRS and actual blood glucose level has recently been given a ention in several reports (32,33), and this intriguing inverse relationship is still awaiting for an explanation (34). In conclusion, elevations in UAE are important independent determinants of a lower HRV and BRS in diabetic patients with peripheral neuropathy. As elevations in UAE are associated with many vascular abnormalities, we raise the hypothesis that impairments of these cardiovascular reflex measurements in diabetic patients reflect both vascular and neural disease. The role of vascular function in autonomic neuropathy testing deserves further investigation. Acknowledgments We are indebted to Marianne Bruin, Anne van Gessel, Wietze Kuipers and Margreet Teune for their skillful technical assistance at the vascular laboratory. Ymie Talsma is acknowledged for her help with the neurography measurements. Dr. E. Blaauwwiekel, Dr. T. Links and Dr. J.W. Meijer from the Beatrixoord Rehabilitation Center in Haren recruited many patients for the study. 76

86 Heart rate variability and baroreflex sensitivity in diabetic patients with peripheral neuropathy are indices of both neural and vascular disease. References 1. Boyko EJ, Ahroni JH, Smith DG, Davignon D: Increased mortality associated with diabetic foot ulcer. Diabet.Med. 13: , Young MJ, Veves A, Boulton AJ: The diabetic foot: aetiopathogenesis and management. Diabetes Metab Rev. 9: , Vinik AI, Maser RE, Mitchell BD, Freeman R. Diabetic autonomic neuropathy. Diabetes Care. 26: , Lluch I, Hernandez A, Real JT, Morillas C, Tenes S, Sanchez C, Ascaso JF: Cardiovascular autonomic neuropathy in type 1 diabetic patients with and without peripheral neuropathy. Diabetes Res.Clin.Pract. 42:35 40, Young RJ, Zhou YQ, Rodriguez E, Presco RJ, Ewing DJ, Clarke BF: Variable relationship between peripheral somatic and autonomic neuropathy in patients with different syndromes of diabetic polyneuropathy. Diabetes 35: , Töyry JP, Partanen JV, Niskanen LK, Lansimies EA, Uusitupa MI: Divergent development of autonomic and peripheral somatic neuropathies in NIDDM. Diabetologia 40: , Cohen JA, Jeffers BW, Faldut D, Marcoux M, Schrier RW: Risks for sensorimotor peripheral neuropathy and autonomic neuropathy in non insulin dependent diabetes mellitus (NIDDM). Muscle Nerve 21:72 80, Bellavere F, Balzani I, De Masi G, Carraro M, Carenza P, Cobelli C, Thomaseth K: Power spectral analysis of heart rate variations improves assessment of diabetic cardiac autonomic neuropathy. Diabetes 41: , Fra ola A, Parati G, Gamba P, Paleari F, Mauri G, Di Rienzo M, Castiglioni P, Mancia G: Time and frequency domain estimates of spontaneous baroreflex sensitivity provide early detection of autonomic dysfunction in diabetes mellitus. Diabetologia 40: , Lefrandt JD, Hoogenberg K, van Roon AM, Dullaart RP, Gans RO, Smit AJ: Baroreflex sensitivity is depressed in microalbuminuric Type I diabetic patients at rest and during sympathetic manoeuvres. Diabetologia 42: ,

87 Chapter Akselrod S, Gordon D, Ubel FA, Shannon DC, Berger AC, Cohen RJ: Power spectrum analysis of heart rate fluctuation: a quantitative probe of beat to beat cardiovascular control. Science 213: , Robbe HW, Mulder LJ, Ruddel H, Langewitz WA, Veldman JB, Mulder G: Assessment of baroreceptor reflex sensitivity by means of spectral analysis. Hypertension 10: , Aso Y, Fujiwara Y, Inukai T, Takemura Y: Power spectral analysis of heart rate variation in diabetic patients with neuropathic foot ulceration. Diabetes Care 21: , Report and recommendations of the San Antonio conference on diabetic neuropathy. Consensus statement. Diabetes 37: , Meijer JW, van Sonderen E, Blaauwwiekel EE, Smit AJ, Groothoff JW, Eisma WH, Links TP: Diabetic neuropathy examination: a hierarchical scoring system to diagnose distal polyneuropathy in diabetes. Diabetes Care 23: , Dyck PJ: Detection, characterization, and staging of polyneuropathy: assessed in diabetics. Muscle Nerve 11:21 32, de Boer RW, Karemaker JM, Strackee J: Hemodynamic fluctuations and baroreflex sensitivity in humans: a beat to beat model. Am.J.Physiol 253:H680 H689, Heart rate variability: standards of measurement, physiological interpretation and clinical use. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Circulation 93: , Clarke CF, Eason M, Reilly A, Boyce D, Werther GA: Autonomic nerve function in adolescents with Type 1 diabetes mellitus: relationship to microalbuminuria. Diabet.Med. 16: , Dullaart RP, Dikkeschei LD, Doorenbos H: Alterations in serum lipids and apolipoproteins in male type 1 (insulin dependent) diabetic patients with microalbuminuria. Diabetologia 32: , Mangili R, Deferrari G, Di Mario U, Giampietro O, Navalesi R, Nosadini R, Rigamonti G, Spezia R, Crepaldi G: Arterial hypertension and microalbuminuria in IDDM: the Italian Microalbuminuria Study. Diabetologia 37: ,

88 Heart rate variability and baroreflex sensitivity in diabetic patients with peripheral neuropathy are indices of both neural and vascular disease. 22. Mykkanen L, Haffner SM, Kuusisto J, Pyorala K, Laakso M: Microalbuminuria precedes the development of NIDDM. Diabetes 43: , Lambert J, Smulders RA, Aarsen M, Donker AJ, Stehouwer CD: Carotid artery stiffness is increased in microalbuminuric IDDM patients. Diabetes Care 21:99 103, Jensen T, Bjerre Knudsen J, Feldt Rasmussen B, Deckert T: Features of endothelial dysfunction in early diabetic nephropathy. Lancet 1: , Yudkin JS, Forrest RD, Jackson CA: Microalbuminuria as predictor of vascular disease in non diabetic subjects. Islington Diabetes Survey. Lancet 2: , Deckert T, Feldt Rasmussen B, Borch Johnsen K, Jensen T, Kofoed Enevoldsen A: Albuminuria reflects widespread vascular damage. The Steno hypothesis. Diabetologia 32: , Saul JP, Berger RD, Albrecht P, Stein SP, Chen MH, Cohen RJ: Transfer function analysis of the circulation: unique insights into cardiovascular regulation. Am.J.Physiol 261:H1231 H1245, Kelbaek H, Jensen T, Feldt Rasmussen B, Christensen NJ, Richter EA, Deckert T, Nielsen SL: Impaired left ventricular function in insulin dependent diabetic patients with increased urinary albumin excretion. Scand.J.Clin.Lab Invest 51: , Ru er MK, McComb JM, Brady S, Marshall SM: Silent myocardial ischemia and microalbuminuria in asymptomatic subjects with non insulin dependent diabetes mellitus. Am.J.Cardiol. 83:27 31, Chowdhary S, Vaile JC, Fletcher J, Ross HF, Coote JH, Townend JN: Nitric oxide and cardiac autonomic control in humans. Hypertension 36: , Veves A, Akbari CM, Primavera J, Donaghue VM, Zacharoulis D, Chrzan JS, DeGirolami U, LoGerfo FW, Freeman R: Endothelial dysfunction and the expression of endothelial nitric oxide synthetase in diabetic neuropathy, vascular disease, and foot ulceration. Diabetes 47: , Watkins LL, Surwit RS, Grossman P, Sherwood A: Is there a glycemic threshold for impaired autonomic control? Diabetes Care 23: ,

89 Chapter Lefrandt JD, Mulder MC, Bosma E, Smit AJ, Hoogenberg KH: Inverse relationship between blood glucose and autonomic function in healthy subjects. Diabetes Care 23: , Gerritsen J, Dekker JM, Ten Voorde BJ, Bertelsmann FW, Kostense PJ, Stehouwer CDA, Heine RJ, Nijpels G, Heethaar RM, Bouter LM: Glucose tolerance and other determinants of cardiovascular autonomic function: the Hoorn study. Diabetologia 43: ,

90 Chapter 5 Sympathetic mediated vasomotion and skin capillary permeability in diabetic patients with peripheral neuropathy Diabetologia. 2003;46:40 47 J. D. Lefrandt 1, E. Bosma 1, P. H. N. Oomen 1, J. H. van der Hoeven 2, A. M. van Roon 1, A. J. Smit 1, K. H. Hoogenberg 3 1 Divisions of Angiology, Department of Internal Medicine, University Hospital, Groningen, The Netherlands 2 Department of Neurology, University Hospital, Groningen, The Netherlands 3 Department of Internal Medicine, Endocrinology and Diabetes, Martini Hospital, Groningen, The Netherlands

91 Chapter 5 Summary Aims/hypothesis. A loss of sympathetic function could lead to changes in capillary fluid filtration in diabetic patients. We investigated whether a decreased sympathetically mediated vasomotion in the skin in diabetic patients with peripheral neuropathy is associated with an abnormal capillary leakage. Methods. Three matched groups were studied: 18 diabetic patients with documented peripheral neuropathy (DN), 18 diabetic patients without peripheral neuropathy (D), and 18 healthy control subjects (C). Sensory and motor nerve function of the distal extremities were assessed by standard neurography, and expressed in a sensory motor nerve function score. Sympathetic vasomotion of the skin microcirculation was assessed by determining the power of blood flow variability in the low frequency ( Hz) band by spectral analysis of laser Doppler flowmetry at the median ankle. Skin capillary leakage was evaluated by sodium fluorescein videodensitometry at the same site of the foot. Results. Sympathetically mediated vasomotion of the foot skin microcirculation was lower in diabetic patients with documented peripheral neuropathy compared with diabetic patients without peripheral neuropathy and control subjects (p<0.001). Capillary sodium fluorescein leakage was larger in 18 diabetic patients with documented peripheral neuropathy than in diabetic patients without peripheral neuropathy (p<0.02) and C (p<0.005). Multiple regression analysis disclosed that a reduced sympathetically mediated vasomotion, together with a lower sensory motor nerve function score, independently contributed to the variance in sodium fluorescein leakage, for 30% (p<0.001) and 17% (p<0.01), respectively. Conclusions. A loss of sympathetic tone, apart from sensory motor nerve dysfunction, seems to be a major determinant of an increased capillary permeability in diabetic patients with neuropathy. [Diabetologia (2003) 46:40 47] 82

92 Sympathetic mediated vasomotion and skin capillary permeability in diabetic patients with peripheral neuropathy Introduction The skin microcirculation is abundantly innervated by sympathetic nerves that regulate blood flow by opening and closing of arteriovenous anastomoses and precapillary arterioles. Blood flow through the skin changes in a cyclical manner, known as vasomotion 1. These oscillations are under sympathetic control, either directly through periodic nerve discharges or through facilitation of an endogenous vascular pacemaker 2,3. The rhythmicity is considered to be a fundamental part of proper tissue perfusion 2-4, and a loss of sympathetically mediated vasomotion could derange fluid filtration at the capillary level. In diabetic patients, sympathetic denervation as well as microcirculatory alterations have been described. A decrease in sympathetic tone 5 8, an increase of capillary flow and pressure 9 11 and an increase of capillary permeability have been documented at various stages of the disease. The role of the sympathetic nervous system in the development of diabetes associated microcirculatory alterations is not fully understood. Spectral analysis of skin blood flow oscillations, recorded continuously by laser Doppler flowmetry, can be used to study sympathetically mediated vasomotion 3,17,18. Sympathetic stimulation by postural change 17,18 and sympathetic blockade by anesthetics 17, have the largest effect on the low fre quency (LF; 0.02 to 0.14 Hz) power component of skin blood flow variability (BFV). The skin LF-BFV power has been shown to be lowered in diabetic patients 3. Sodium fluorescein (NaF) videodensitometry is used to study capillary permeability. With different visualizing techniques, an increased NaF leakage has been found in the capillary beds of the retina 19, the nailfold 12 and the foot skin 20 of diabetic patients. With an improved technique 21, we found that angiotensin converting enzyme inhibitors, independent from its blood pressure lowering effect, reduced NaF leakage in microalbuminuric diabetic patients, suggesting that changes that at the microcirculatory level influence capillary NaF permeability 15. In this study, we investigated the role of sympathetically mediated vasomotion on capillary permeability in diabetic patients with and without overt peripheral neuropathy and healthy control subjects by combining the laser Doppler based assessments of skin LF-BFV and the skin capillary NaF videodensitometry. 83

93 Chapter 5 Table 1. Clinical characteristic of the study groups Study groups* DN D C Number (n) Age (years) 57±3 54±2 58±2 Sex (male/female) 8/10 8/10 8/10 BMI (kg/m 2 ) 29.3± ± ±1.3 Type of diabetes (Type1/Type 2) 4/14 4/14 Diabetes duration (years) 18±3 12±2 Patients using insulin (n) Insulin dose (IU/day) 69±15 a 42±7 Smokers (n) Ankle arm index 1.16± ± ±0.03 Toe pressure (mmhg) 129±7 118±5 123±4 Plethysmographic crest time (s) 0.19± ± ±0.02 Glycated haemoglobin (%) 8.2±0.2 b 7.6±0.2 b 5.7±0.1 Retinopathy (Absent/Background/Proliferative) 3/11/4a 13/5/0 Microalbuminuria (n) 9c 1 0 Albumin/creatinine ratio (mg/µmol) 17.6±10.9c 1.3± ±0.1 Serum creatinine (µmol/l) 93.4± ± ±2.1 Systolic/Diastolic Blood Pressure (mmhg) 148±5/80±3 136±3/79±1 136±4/84±2 Number of antihypertensives (0/1/2) a 6/6/6 d 11/5/2 16/2/0 ACE inhibition therapy (n (%)) 10/18(55%)d 6/18(30%) 2/18(11%) Diabetic Neuropathy Examination score 8.9±2.1 c 1.2± ±1.6 Semmes Weinstein monofilaments 6.60±0.12 c 5.12± ±0.08 Foot: Sensory conduction n. suralis measurable 9/18 17/18 18/18 velocity (m/s) 37.8±1.3 c 45.4± ±0.7 amplitude (mv) 0.9±0.2 c 7.9± ±1.8 ranked scores 10.5±1.4 c 33.9± ±1.9 Foot: Motor conduction n. peroneus (tibialis anterior) Measurable 18/18 18/18 18/18 velocity (m/s) 43.8±1.7 c 55.9± ±1.5 amplitude (mv) 5.7±0.8 c 10.2± ±0.8 ranked scores 14.6±2.1 c 35.6± ±1.9 Combined ranked sensory motor responses 12.6±1.4 c 34.7± ±1.7 Data are given in means ± SEM. * Groups DN, D and C, see text. a p<0.05 from D, b p<0.001 from C, c p<0.001 from D and C, d p<0.005 from D and C. Subjects, Materials and methods Characterization of patients and healthy subjects. The study was approved by the local medical ethics commi ee and wri en informed consent was obtained from all participants after explanation of the purpose of the study. Patients were recruited from the outpatient clinic of the University Hospital Groningen and the Rehabilitation Centre Beatrixoord. Healthy control subjects were recruited by an advertisement in a local newspaper. The study consisted of three groups: 18 diabetic patients with peripheral neuropathy (DN), 18 diabetic patients without peripheral neuropathy (D), and 18 control subjects with normal glucose tolerance (C). The groups were individually matched for age (within 5 yr), sex and body mass index (BMI, within 5 kg/m2). The diabetic patients were matched for type of diabetes (Type 1 and Type 2). The results of the matching procedure are given in Table 1. The dia- 84

94 Sympathetic mediated vasomotion and skin capillary permeability in diabetic patients with peripheral neuropathy betes type was assessed on clinical grounds, and in the case of doubt by glucagon stimulated C peptide concentrations. The proportion of diabetic patients using insulin was similar in DN and D, although DN used higher doses than D (p<0.05, Table 1). Normal glucose tolerance in the control subjects was assessed by a blood glucose concentration less than 7.8 mmol/l 120 min after ingestion of 75 grams of glucose. The number of smokers was not different among the groups. Peripheral vascular disease was excluded in all participants by normal ankle arm indexs (>0.90), toe pressures (>100 mmhg) and plethysmography (crest time <0.22 s, Table 1). Clinically apparent cardiac disease was excluded by history, and all had a normal resting electrocardiography, confirming sinus rhythm. Glycated haemoglobin concentrations indicated that the diabetic patients were in moderate metabolic control (Table 1). Neuropathy coincided with other diabetes associated complications since retinopathy was more severe (p<0.05 from D, Table 1) and the prevalence of microalbuminuria higher in DN (p<0.001 from D and C, Table 1). Patients using α, β adrenergic inhibitors or calcium antagonists were excluded. The use of angiotensin converting enzyme inhibitors, angiotensin II receptor antagonists and diuretics was allowed. Although systolic and diastolic blood pressure did not differ, a higher proportion of patients in DN used antihypertensive medication (p<0.005, Table 1). Patients with alcohol consumption of more than four units per day, hepatic and renal insufficiency, folic acid and hydroxycobolamin deficiencies were excluded. Assessment of neuropathy. Diabetic neuropathy was diagnosed according to the San Antonio Consensus Statement criteria 22. Clinical neuropathy signs were scored by means of the Diabetic Neuropathy Examination (DNE) score 23 that has a maximum score of 16 and is a modification of Neuropathy Disability Score 24. Quantitative sensory testing was carried out with Semmes Weinstein monofilaments. Motor nerve conduction velocities and amplitudes were measured in the peroneal nerve (tibialis anterior), and sensory nerve conduction velocities and amplitudes in the sural nerve 25. Nerve conduction velocities and amplitudes of the peroneal and the sural nerve were ranked separately. The mean of the rank of the conduction velocity and the amplitude of these nerves, as well as the overall mean in ranks of both nerves, was calculated as indexes of motor nerve function, sensory nerve func- 85

95 Chapter 5 tion and combined sensory motor nerve function, respectively. The neuropathy scores are given in Table 1. As expected from the selection criteria, severe neuropathy was present in the DN group, and no differences in neuropathy scores existed between the DC and C groups (Table 1). Study protocol. The participants were studied in the morning, 1 1/2 h after breakfast during which the diabetic patients had taken their usual oral hypoglycaemic drugs or regular insulin injections. Blood glucose was measured at arrival, and at hourly intervals thereafter. Blood glucose concentrations ranging from 3.5 to 15.0 mmol/l were accepted during the studies. Subjects were instructed not to smoke or consume drinks containing caffeine 24 h before the study. They were Figure 1A, B. (A) 5 minutes of continuous skin blood flow recording by the laser Doppler flowmetry. (B) Power density spectrum of the changes in skin blood flow 86

96 Sympathetic mediated vasomotion and skin capillary permeability in diabetic patients with peripheral neuropathy studied in a supine position and allowed to acclimatize for 30 min to the environmental conditions of a quiet, temperature controlled room (24 C). Ankle skin temperature was measured with a thermocouple (Ellab DU3 s, Copenhagen, Denmark), a skin temperature above 28 C was considered appropriate to start the measurements 10. Blankets, covering the legs, were used to maintain the temperature at the desired level, while exogenous heating was avoided as this could influence skin blood flow oscillations. Skin blood flow variability. Skin blood flow was measured at the median ankle with laser Doppler flowmetry (Diodopp, Applied laser technology, Asten, The Netherlands), which has been validated previously 26. The probe was a ached with double sided adhesive tape at the medial malleolus. The laser Doppler flux signal (Figure 1A), expressed in arbitrary units (AU), was recorded online for 10 min by a personal computer program (Poly, version 4.9, Inspektor Research Systems, Amsterdam, The Netherlands) with a sampling frequency of 100 Hz, and stored on disk for analysis afterwards. Skin BFV was assessed by spectral analysis using the CARSPAN program (version 2.0, Progamma, Groningen, the Netherlands) 27,28. The systolic component of the laser Doppler signal was taken for discrete Fourier transformation (Figure 1b). The following spectral characteristics were obtained: the total BFV was the power in the 0.02 to 0.40 Hz frequency band (AU2), the low frequency (LF) BFV was the power in the Hz frequency band (AU2), and the high frequency (HF) BFV was the power in the Hz frequency band (AU 2 ). The LF-BFV was also expressed in normalized units (NU), which represents the ratio between the LF-BFV power and total BFV power 3,18. Sodium fluorescence videodensitometry. Large window NaF videodensito metry was carried out to measure skin capillary permeability. The system consists of an epiilumination microscope (Olympus BHMJ, Tokyo, Japan) to which a 75 wa xenon lamp is mounted (Osram XBO, Berlin, Germany). Emi ed light is filtered using a fluorescence filter set (Olympus BH2 UDMB, excitation nm, barrier 515 nm, Tokyo, Japan). A 2 3 mm section of the skin of the medial malleolus of the ankle was visualized (magnification 100), adjacent to the place where the LDF probe was situated. Immersion oil (Leitz, din 58884, Wetzlar, 87

97 Chapter 5 Figure 2A D. Sodium fluorescein (NaF) videodensitometry images at the time of dye arrival in the skin capillaries in a healthy subject (A) and in a diabetic patient with peripheral neuropathy (B). Similar images after 60 s of interstitial NaF diffusion in a healthy subject (C) and a diabetic patient with peripheral neuropathy (D) Germany) was applied to the skin to increase skin transparency. A bolus of NaF solution [0.3 ml of a 15% NaF (MW 376) solution per litre of estimated blood volume] was injected intravenously. The epiillumination microscope visualizes the rapid capillary appearance (Figure 2A) and the subsequent interstitial leakage of NaF (Figure 2B). Images were continuously recorded for 20 min using a video camera (Grundig FA 85, Fürth, Germany), from which the automatic gain function had been removed, and a S VHS video recorder (JVC HR S7500E/EH, Japan). Every second, video images were digitized for analysis (Figure 2). Total fluorescence light intensity of each image was expressed in arbitrary units (AU). Background fluorescence intensity was obtained from one baseline image and was subtracted from all subsequent intensities. To avoid the effects of differences in skin transparency and body composition between individuals, fluorescent light intensities were expressed as a percentage of individual maximal light intensity. The relative fluorescence values provided a semiquantitative way to describe NaF leakage curves. The intra individual day to day reproducibility has a coefficient of variation of 10% 21. The time interval between the NaF injection 88

98 Sympathetic mediated vasomotion and skin capillary permeability in diabetic patients with peripheral neuropathy until appearance in the skin capillaries was defined as dye arrival time. Capillary density was measured by counting the visualized capillaries from the tape recordings after appearance of NaF in the skin capillaries (Figure 2a). Statistical analysis. Data are given as means ± SEM. The natural log of the spectral powers of skin BFV were used for analyses since spectral parameters have a χ 2 distribution that is normalized after log transformation 28. Both the NaF leakage curves from 0 to 7 min, as well as the area under the curve of the relative fluorescence intensity (denoted as AUC and the time after dye arrival in seconds) were analysed. One way ANOVA with Bonferroni correction for multiple comparisons was used to analyse differences between the groups. Chi square analysis was used to show differences in prevalence between the groups. Repeated ANOVA measurements, with Student Newmann Keuls correction for multiple comparisons, evaluated between group differences of the 0 to 7 min NaF leakage curves. Univariate correlations were expressed as Pearson s correlation coefficients. The independent contribution of variables was evaluated by stepwise multiple regression analysis. A double sided p value of less than 0.05 was considered statistically significant. Table 2. Spectral analysis parameters of skin blood flow variability (BFV) and skin capillary sodium fluorescein leakage Study groups* DN D C Blood glucose during study (mmol/l) 8.9±0.8 a 8.7±0.6 a 5.9±0.2 Mean skin temperature ( C) 31.4±0.2 b 30.4± ±0.2 Total power skin BFV ln(au 2 ) 8.9±0.2 c 9.4± ±0.2 Low frequency power skin BFV ln(au 2 ) 8.4±0.2 a,c 9.1± ±0.2 High frequency power skin BFV ln(au 2 ) 8.0± ± ±0.1 Low frequency BFV (NU) 0.56±0.04 b 0.77± ±0.03 Number of Capillaries (per mm 2 ) 29±3 d 21±2 22±2 Dye Arrival Time (s) 30.6±1.6 b 41.2± ±3.6 Maximum fluorescence (AU) 19601± ± ±2705 AUC 30 s fluorescence curve after 30 s 30.3±5.0 a,e 13.8± ±1.1 AUC 60 s fluorescence curve after 60 s 37.0±5.1 a,e 20.6± ±1.6 AUC 120 s fluorescence curve after 120 s 45.5±4.9 a,f 30.5± ±2.0 AUC 420 s fluorescence curve after 420 s 68.0±3.4 g 60.3± ±2.0 Data in means ± SEM. * Groups DN, D and C, see text. Spectral power are natural logs. AUC, area under the curve at given time point. a p<0.001 from C; b p<0.001 from D and C; c p<0.01 from C; d p<0.05 from D and C, e p<0.01 from D, f p <0.05 from D, g p<0.01 from C 89

99 Chapter 5 Figure 3. (A) Fluorescence light intensity after sodium fluorescein arrival in the skin of diabetic patients with peripheral neuropathy ( ), without peripheral neuropathy ( ) and healthy controls ( ). vs., p<0.02; vs., p<0.01. (B) Relationship between the low frequency power of skin blood flow variability (LF BFV) in normalized units and the area under the curve 60 s after sodium fluorescein arrival., Type 1 ( ) and Type 2 ( ) diabetic patients with peripheral neuropathy, Type 1 ( ) and Type 2 ( ) diabetic patients without peripheral neuropathy and healthy controls ( ), r= 0.64, p<

100 Sympathetic mediated vasomotion and skin capillary permeability in diabetic patients with peripheral neuropathy Results The mean blood glucose concentration was higher in D and DN than in C (p<0.001, Table 2). Skin temperature was higher in DN than in D and C (p<0.001, Table 2). The total power BFV was lower in DN compared to C (p<0.01, Table 2). The LF-BFV power was lower in DN (p<0.05 from D and p<0.005 from C), while the HF BFV power was not different among the groups (Table 2). The differences in LF-BFV, the band of interest, were even more evident when expressed in normalized units, i.e. the ratio of LF-BFV/total BFV (p<0.001 from D and C, Table 2). The normalized LF-BFV was further evaluated in statistical analysis. The number of visible skin capillaries was larger in DN than in D and C (p<0.05, Table 2). The NaF arrival time in the skin capillaries was shorter in DN than in D and C (p<0.001 for both, Table 2). Maximal fluorescence was similar between the groups (Table 2). The NaF leakage curve of relative fluorescence from 0 to 420 s, was higher in DN (p<0.02 from D and p<0.005 from C, Figure 3A). Expressed as AUC s from 0 until 30, 60, 120 and 420 s, the differences in the NaF leakage were more pronounced in the first 60 s (Table 2). For practical purposes, therefore, the NaF leakage at 60 s (NaF AUC 60 s ) was taken for further analysis. The normalized LF-BFV correlated with NaFAUC 60 s (r= 0.64, p<0.001, n=54, Figure 3B), diabetes duration (r= 0.51, n=54, P<0.001), grade of retinopathy (r= 0.51, p<0.001, n=54), combined sensory motor nerve responses (r=0.48, p<0.001, n=54), NaF capillary arrival time (r= 0.46, p<0.01, n=54), HbA1c concentration (r= 0.44, p<0.01, n=54), skin temperature (r= 0.43, p<0.01, n=54) and number of capillaries (r= 0.33, p<0.02, n=54). The NaF AUC 60 s correlated with the normalized LF-BPV (r= 0.64, p<0.001, n=54), combined sensory motor nerve responses (r= 0.53, p<0.001, n=54), diabetes duration (r=0.50, p<0.001, n=54), HbA1 c concentration (r=0.41, p<0.01, n=54), NaF arrival time (r= 0.40, p<0.01, n=54), grade of retinopathy (r=0.38, p<0.01, n=54), the number of capillaries (r= 0.36, p<0.01, n=54) and systolic blood pressure (r=0.27, p<0.05, n=54). Stepwise multiple regression analysis disclosed that when normalized LF BPV, sensory motor nerve responses, diabetes duration, HbA 1c concentration, grade of retinopathy, systolic blood pressure, NaF arrival time and number of capillaries were included as the determinants 91

101 Chapter 5 of NaF AUC 60 s, only normalized LF-BFV and combined sensory motor nerve responses independently contributed to 30% (p<0.001) and 17% (p<0.01), respectively, to the variance in NaFAUC 60 s. In a second analysis, sensory motor nerve function, diabetes duration, grade of retinopathy and HbA 1c concentration were included as the determinants of normalized LF-BFV. Diabetes duration and combined sensory motor nerve responses explained 16% (p<0.01 ) and 9% (p<0.05), respectively, of the variance in normalized LF-BFV. The type of diabetes (Type 1/Type 2) and the use of an ACE inhibitor (yes/no), included as categorical variables, did not change the multiple regression models of NaFAUC 60 s and LF-BFV. Discussion In this study a loss of sympathetically mediated vasomotion in the skin of diabetic patients with neuropathy was associated with an increased NaF leakage. This finding suggests that sympathetic nerves influence microvascular fluid homeostasis in a way that a loss of sympathetic function relates to an increased capillary permeability. The power of the low frequency component of blood flow variability, resembling sympathetic modulation, was decreased in the skin of the diabetic patients with peripheral neuropathy. Previous studies, using the laser Doppler flowmetry to record changes in skin blood flow documented blunted responses to sympathetic maneuvers in patients with diabetic neuropathy 10, Another study 3 used spectral analysis on the signals of the laser Doppler, applied at the forearm and finger region, to demonstrate abnormalities in sympathetic modulation in diabetic patients with neuropathy. We extended these observations by measuring skin LF-BFV at the feet, where the long and thin unmyelinated sympathetic fibres are particularly vulnerable to neuronal damage by the diabetic state. Our finding that a reduction in skin LF-BFV power related to the grade of sensory motor dysfunction as well as to the duration of diabetes, indicates that sympathetically mediated vasomotion decreases in conjunction with sensory motor neuropathy and prolonged glycaemic exposure. Capillary NaF leakage, taken as an index of capillary permeability, was increased in the diabetic patients with peripheral neuropathy. The 92

102 Sympathetic mediated vasomotion and skin capillary permeability in diabetic patients with peripheral neuropathy amount of interstitial NaF is proportional to the rate of NaF diffusion through the skin capillaries 15,21, since NaF delivery to the interstitial space is relatively fast, reaching a maximal interstitial concentration after 5 to 10 min, while NaF clearance from the interstitium takes at least several hours. Estimation of the trans capillary escape rate of albumin 13,16, the transcapillary diffusion gradient 14 and NaF leakage 15 showed an increase in capillary permeability in Type 1 diabetic patients, particularly when complicated by microalbuminuria. No previous studies have evaluated the role of neuropathy in the increased capillary permeability. Multiple regression analysis disclosed that a reduced LF-BFV and to lesser extent a loss of sensory motor nerve function contributed to an increase in NaF, independently from other covariates. Although microalbuminuria was highly prevalent among the patients with neuropathy, it was not related to a higher NaF leakage. A previously reported association between microalbuminuria and NaF leakage 15 could have had neuropathy as an important intermediary factor. In that study, neurography and blood flow variability were not measured, as ACE inhibition therapy reduces NaF leakage 15, the true NaF leakage could have even been higher in the neuropathy group as more of these patients were on ACE inhibitors. The finding that the use of an ACE inhibitor did not contribute to the NaF leakage, does not exclude this possibility, but is probably beyond the scope of this study. The following pathophysiological mechanisms could be involved in the relation between the decline in LF-BFV and the increased NaF leakage in the patients with neuropathy. According to Starlings forces of fluid filtration, the transcapillary passage of fluids depends on the transcapillary pressure gradient, which is a result of the net hydrostatic and the net osmotic pressures, the capillary filtration area and the diffusion coefficient 33. The augmented NaF leakage can, therefore, be explained in terms of hydrostatic capillary pressure, the filtration surface area and barrier properties of the capillary wall 15. The lower LF-BFV could indicate a lower state of vasoconstriction of those skin blood vessels that are normally innervated by sympathetic nerves i.e. arterio venous anastomoses and precapillary arterioles. Consequently, a reduced vasomotor tone might enhance skin blood flow as observed in this study by the higher skin temperature and the faster arrival and initial appearance of NaF in the skin capillaries of the patients with neuropathy. Indeed, an increased arterio venous shunting and at the 93

103 Chapter 5 same time increased capillary perfusion, have been well documented in the foot skin of diabetic patients complicated by neuropathy 9,10. The proposed precapillary arteriolar vasodilation could increase capillary hydrostatic pressure, which could have promoted NaF leakage. The increased skin capillary density of the patients with neuropathy in this study 10 suggests that the reduction in sympathetically mediated vasomotion is associated with capillary recruitment and, thereby, with a larger capillary filtration area for NaF. Likewise, at the single capillary level, a decrease in the cyclic modulations of skin blood vessel constriction could, persistently or intermi ently, increase the available capillary diffusion area. The diffusion coefficient could also have been altered in the patients with neuropathy. A decrease in cyclic modulations could have altered the properties of the capillary wall itself, by increasing the size of the intercellular cell junctions, through which NaF may pass. Taken together, probably an increased capillary filtration pressure, but more likely a larger capillary filtration area and a decreased capillary filtration barrier are proposed mechanisms, by which a reduced sympathetic tone promotes NaF leakage in the diabetic patient with peripheral neuropathy. Early sympathetic dysfunction has been suggested to play a role in diabetes associated haemodynamic alterations and has been implicated in the development of diabetic complications such as diabetic foot ulceration and diabetic nephropathy in patients with Type 1 diabetes mellitus and, although more controversial, in patients with Type 2 diabetes mellitus 33. The fact that the multiple regression models of LF-BFV and NaF leakage parameters of microcirculatory function were not influenced by the type of diabetes, challenges the thought that an increased capillary permeability in the skin is a phenomenon of autonomic neuropathy. Endothelial dysfunction could be common denominator of the observed association between NaF leakage and LF-BFV in these patients groups. Endothelial dysfunction, as demonstrated by a reduced vasodilation on the dorsal foot in response to heating and iontophoresis of acetylcholine, was found in neuropathic Type 1 and Type 2 diabetic patients both with and without vascular disease, but not in diabetic patients without neuropathy and control subjects 34. Further, the vasodilatory response to acetylcholine at the foot was reduced in diabetic patient with neuropathy compared to diabetic patients without neuropathy and control subjects

104 Sympathetic mediated vasomotion and skin capillary permeability in diabetic patients with peripheral neuropathy In this cross sectional and observational study, we found a statistically independent contribution of skin LF-BFV on NaF leakage in diabetic patients complicated by neuropathy. Thus, these data add further insights in the factors contributing to the microcirculatory changes as described by the haemodynamic hypothesis 11,33. Further studies are needed to elaborate the putative role of sympathetic nerve dysfunction on skin microcirculatory changes in patients with diabetes mellitus, and to evaluate whether these findings are also present in other organ tissues than the foot skin. In conclusion, this study documents that peripheral autonomic neuropathy is related to increased capillary permeability in diabetic patients. A loss of sympathetic tone seems to be a major determinant of an increased capillary fluid filtration, which could be deleterious to the skin nutritive capillary function in diabetic patients with neuropathy. Acknowledgements We are indebted to Y. Talsma for her support during the neurography measurements. E. Blaauwwiekel, T. Links and J. W. Meijer of the Beatrixoord Rehabilitation Centre of Haren which recruited many patients for the study. R. Henning carried out the statistical analysis of the sodium fluorescein curves. We thank M. Bruin, B. Buist, A. van Gessel, W. Kuipers and M. Teune for their skilful assistance at the vascular laboratory. Finally, the Groningen Endocrinology Foundation and the Keyzer en de Houtman Foundation provided financial support. 95

105 Chapter 5 References 1. Christensen NJ (1969) Spontaneous variations in resting blood flow, postischaemic peak flow and vibratory perception in the feet of diabetics. Diabetologia 5: Benbow SJ, Pryce DW, Noble K, MacFarlane IA, Friedmann PS, Williams G (1995) Flow motion in peripheral diabetic neuropathy. Clin Sci (Colch) 88: Bernardi L, Rossi M, Leuzzi S et al. (1997) Reduction of 0.1 Hz microcirculatory changes as evidence of sympathetic dysfunction in insulin dependent diabetes. Cardiovasc Res 34: Wiensperger N (2000) Defects in microvascular haemodynamics during prediabetes: contributor or epiphenomenon? Diabetologia 43: Wiernsperger NF (2001) In defense of microvascular constriction in diabetes. Clin Hemorheol Microcirc 25: Hoffman RP, Sinkey CA, Kienzle MG, Anderson EA (1993) Muscle sympathetic nerve activity is reduced in IDDM before overt autonomic neuropathy. Diabetes 42: Schnell O, Kirsch CM, Stemplinger J, Haslbeck M, Standl E (1995) Scintigraphic evidence for cardiac sympathetic dysinnervation in long term IDDM patients with and without ECG based autonomic neuropathy. Diabetologia 38: Ziegler D, Weise F, Langen KJ et al. (1998) Effect of glycaemic control on myocardial sympathetic innervation assessed by [123I]me taiodobenzylguanidine scintigraphy: a 4 year prospective study in IDDM patients. Diabetologia 41: Flynn MD, Edmonds ME, Tooke JE, Watkins PJ (1988) Direct measurement of capillary blood flow in the diabetic neuropathic foot. Diabetologia 31: Ne en PM, Wollersheim H, Thien T, Lu erman JA (1996) Skin microcirculation of the foot in diabetic neuropathy. Clin Sci (Colch) 91: Sandeman DD, Shore AC, Tooke JE 1992) Relation of skin capillary pressure in patients with insulin dependent diabetes mellitus to complications and metabolic control. N Engl J Med 327:

106 Sympathetic mediated vasomotion and skin capillary permeability in diabetic patients with peripheral neuropathy 12. Bollinger A, Frey J, Jäger K, Furrer J, Seglias J, Siegenthaler W (1982) Pa erns of diffusion through skin capillaries in patients with long term diabetes. N Engl J Med 307: Feldt Rasmussen B (1986) Increased transcapillary escape rate of albumin in Type 1 (insulin dependent) diabetic patients with microalbuminuria. Diabetologia 29: Jaap AJ, Shore AC, Gartside IB, Gamble J, Tooke JE (1993) Increased microvascular fluid permeability in young Type 1 (insulin dependent) diabetic patients. Diabetologia 36: Oomen PHN, Jager J, Hoogenberg K, Dullaart RPF, Reitsma WD, Smit AJ (1999) Capillary permeability is increased in normo and microalbuminuric Type 1 diabetic patients: amelioration by ACE inhibition. Eur J Clin Invest 29: Vervoort G, Lu erman JA, Smits P, Berden JH, Wetzels JF (1999) Transcapillary escape rate of albumin is increased and related to haemodynamic changes in normo albumin uric type 1 diabetic patients. J Hypertens 17: Bernardi L, Rossi M, Fratino P, Finardi G, Mevio E, Orlandi C (1989) Relationship between phasic changes in human skin blood flow and autonomic tone. Microvasc Res 37: Bernardi L, Radaelli A, Solda PL et al. (1996) Autonomic control of skin microvessels: assessment by power spectrum of photoplethysmographic waves. Clin Sci (Colch) 90: Cunha Vaz J, Faria de Abreu JR, Campos AJ (1975) Early breakdown of the blood retinal barrier in diabetes. Br J Ophthalmol 59: Frey J, Furrer J, Bollinger A (1983) Transcapillary diffusion of Na fluorescein in skin areas of the dorsum of the foot in juvenile diabetics. Schweiz Med Wochenschr 113: Jager J, Oomen PHN, Sluiter WJ, Reitsma WD, Smit AJ (1997) Improved reproducibility of the large window method of assessing transcapillary and interstitial fluorescein diffusion in the skin in healthy subjects and in subjects with insulin dependent diabetes mellitus. Int J Microcirc Clin Exp 17: Report and recommendations of the San Antonio Conference on diabetic neuropathy (1988) Diabetes 37: Dyck PJ (1998) Detection, characterization and staging of polyneuropathy: assesses in diabetics. Muscle Nerve 11:

107 Chapter Meijer JW, Van Sonderen E, Blaauwwiekel EE et al. (2000) Diabetic neuropathy examination: a hierarchical scoring system to diagnose distal polyneuropathy in diabetes. Diabetes Care 23: Standardized measures of diabetic neuropathy (1995) Diabetes Care 18 [Suppl 1]: Ne en PM, Keeris LM, De Boo T, Wollersheim H, Thien T (1993) A clinical comparison of two laser Doppler instuments. Int J Microcirc Clin Exp 12: Robbe HW, Mulder LJ, Ruddel H, Langewitz WA, Veldman JB, Mulder G (1997) Assessment of baroreceptor reflex sensitivity by means of spectral analysis. Hypertension 10: Lefrandt JD, Hoogenberg K, Van Roon AM, Dullaart RPF, Gans ROB, Smit AJ (1999) Baroreflex sensitivity is depressed in microalbuminuric Type I diabetic patients at rest and during sympathetic manoeuvres. Diabetologia 42: Aso Y, Inukai T, Takemura Y (1997) Evaluation of skin vasomotor reflexes in response to deep inspiration in diabetic patients by laser Doppler flowmetry. Diabetes Care 20: Stansberry KB, Hill MA, Shapiro SA, McNi PM, Bha BA, Vinik AI (1997) Impairment of peripheral blood flow responses in diabetes resembles an enhanced aging effect. Diabetes Care 20: Bornmyr S, Castenfors J, Svensson H et al. (1999) Detection of autonomic sympathetic dysfunction in diabetic patients. A study using laser Doppler imaging. Diabetes Care 22: Bornmyr S, Castenfors J, Svensson H, Wroblewski M, Sundkvist G, Wollmer P (2000) Abnormal vasoreaction to arousal stimuli an early sign of diabetic sympathetic neuropathy demonstrated by laser Doppler flowmetry. J Clin Neurophysiol 17: Tooke JE (1995) Microvascular function in human diabetes. A physiological perspective. Diabetes 44: Veves A, Akbari CM, Primavera J et al. (1998) Endothelial dysfunction and the expression of endothelial nitric oxide synthetase in diabetic neuropathy, vascular disease, and foot ulceration. Diabetes 7: Arora S, Smakowski P, Frykberg RG et al. (1998) Differences in foot and forearm skin microcirculation in a group of Type 1 and Type 2 diabetic patients with and without neuropathy. Diabetes Care 8:

108 Part III: Hypertension

109

110 Chapter 6 Autonomic Function in Hypertensive and Normotensive Subjects The Importance of Gender Hypertension June;37: K. Sevre 1, J. D. Lefrandt 4, G. Nordby 2, I. Os 2, M. Mulder 4, R. O.B. Gans 4, M. Rostrup 2,3, A. J. Smit 4. Departments of Cardiology 1, Internal Medicine 2, and Research Forum 3, Ullevål University Hospital, Oslo, Norway; Department of Internal Medicine 4, University Hospital of Groningen, The Netherlands.

111 Chapter 6 Summary Baroreceptor reflex sensitivity (BRS) has been found lower and heart rate variability (HRV) parasympathetic markers have been found higher in healthy women than in healthy men. Thus, in the present study we hypothesized gender differences in the autonomic function among hypertensive subjects. Forty one hypertensive patients and 34 normotensive subjects, age 53 ± 1 years, were examined. Four weeks after cessation of antihypertensive therapy, HRV was assessed in 24 hour Holter ECGs, and BRS was calculated with the transfer technique. A t test was performed after log transformation of spectral values. Resting blood pressure and heart rate in the hypertensive and the normotensive groups were 150 ± 2/100 ± 1 (mean ± SEM) and 121 ± 2/81 ± 1 mmhg, respectively, and 68 ± 1 and 60 ± 1 bpm, respectively (P<0.0005). Compared with normotensive controls, hypertensive patients had lower total power (1224 ± 116 versus 1797 ± 241 ms 2 ; P=0.03), lower low frequency power (550 ± 57 versus 813 ± 115 ms 2 ; P=0.04), lower high frequency power (141 ± 23 versus 215 ± 38 ms 2 ; P=0.06), lower root mean square successive difference (28.7 ± 2.7 versus 35.7 ± 3.0 ms; P=0.03), and PNN50 (4.9 ± 0.6% versus 9.8 ± 1.5%; P=0.003). BRS was also lower in the hypertensive subjects (7.6 ± 0.6 versus 10.4 ± 0.8 ms/ mmhg; P=0.005). When comparing the same parameters between normotensive subjects and hypertensive subjects within the same gender group, we found significant reduction (P<0.05) only within the female group. The difference in BRS within the female group was twice that within the male group. Stepwise multiple regression analysis revealed gender, age, HDL cholesterol, and blood pressure as independent explanatory variables of BRS and HRV. Our results suggest that gender is an important determinant of BRS and HRV. Autonomic function parameters were especially impaired in hypertensive women compared with hypertensive men. 102

112 Autonomic Function in Hypertensive and Normotensive Subjects: The Importance of Gender Introduction The autonomic nervous system plays a crucial role in blood pressure (BP) and heart rate (HR) control and may thus be an important pathophysiological factor in the devel opment of hypertension. There have been numerous studies on plasma catecholamines in essential hypertension, 1 most of which have shown increased levels in hypertensive subjects. Moreover, disturbed autonomic HR and BP control has been demonstrated in several studies by means of HR variability (HRV) and baroreceptor reflex sensitivity (BRS) HRV, which estimates the tonic HR control, is gener ally reduced (standard deviation of all R R intervals [SDNN] and total power [energy in the heart period spectrum between and 0.40 Hz] [TP]) in hypertensive patients. 2,4 6 Markers of sympathetic predominance are increased in some 3 but not all studies. 4 6 BRS, which estimates the reflex vagal HR control, is reduced in hypertensive subjects Both BRS and HRV parameters (except low frequency power [energy in the heart period spectrum between 0.04 and 0.15 Hz] [LF]/high fre quency power [energy in the heart period spectrum between 0.15 and 0.40 Hz] [HF]) decrease with increasing age in healthy 10,14 and hypertensive subjects. 10 It is also proposed that BRS stabilizes after middle age. 8 There may be gender differences in the pathophysiology of essential hypertension. We have previously observed that hypertensive women have low renin hypertension 15 and less cardiovascular reactivity to stress compared with hyperten sive men. 16 However, thus far there has been only 1 specific study on possible gender differences in HRV in hypertension, 4 and there have been none on BRS. Singh et al 4 found reduced LF/HF ratio but not TP, percentage of adjacent R R intervals differing >50 ms (PNN50), square root of the mean of the sum of the squares of differences between adjacent R R intervals (RMSSD), LF, and HF in hypertensive women compared with hypertensive men. However, there are more studies on gender differences in healthy subjects. SDNN, 4,17,18 LF/HF ratio, and LF normal ized units [LF/ (LF+HF) 100] 4,17,19 have been found lower and HF normalized units [HF/(LF+HF) 100] 17 and HF 4,17,19 have been found higher in women than in men. Moreover, BRS is reduced in women compared with men, 14,17,18,20 but the difference is not significant in those age 60 years

113 Chapter 6 In both healthy and hypertensive subjects, previous studies suggest a higher tonic parasympathetic activity in women than in men. Surprisingly, these studies also propose decreased reflex vagal responses in healthy women compared with healthy men. However, the correlation between HRV and BRS is weak. 12 While gender differences in HRV and BRS have been studied in healthy subjects, no one has addressed specifically gender differences in BRS in hypertensive patients compared with normotensive controls. This article presents some new findings suggesting that autonomic dysfunction may play a more prom inent role in female than in male hypertension. Methods Subjects. Patients were eligible if they were age >18 years and suffered from mild to moderate hypertension (systolic BP 140 and 180 mmhg, diastolic BP 90 mmhg). Patients were excluded if secondary hypertension was suspected, if they had recently suffered from a cardiovascular event, if organ failure was present, or if they had diabetes mellitus, autoimmune disease, or Parkinson s disease. Use of neuroleptics, antidepressants, lithium, antiarrhythmics, and cime tidine was not allowed. These criteria were similar for the normo tensive controls except for their BP level, which had to be <140/90 mmhg. Patients were recruited from the outpatient clinic for hypertensive patients, Ullevål University Hospital, Oslo, Norway, and the University Hospital of Groningen, Netherlands. Normotensive controls were partly former participants in a screening program for cardiovascular Table 1. Basal Characteristics in Normotensive vs Hypertensive Subjects Basal Parameters Normotensive Hypertensive P n Age, y 52.7 ± ± 1.4 NS BMI, kg/m ± ± 0.5 NS Smoker, % NS Hematocrit, fraction 0.40 ± ± NS Creatinine, µmol/l 84.0 ± ± 2.1 NS Total cholesterol, mmol/l 5.8 ± ± 0.3 NS HDL cholesterol, mmol/l 1.2 ± ± Triglyceride, mmol/l 1.2 ± ± 0.2 NS Systolic BP, mmhg 121 ± ± 2 < Diastolic BP, mmhg 81 ± ± 1 < HR, bpm 60 ± 1 68 ± 1 < Data are mean ± SEM. 104

114 Autonomic Function in Hypertensive and Normotensive Subjects: The Importance of Gender Table 2. Basal Characteristics in Normotensive Women vs Hypertensive Women and Normotensive Men vs Hypertensive Men Female Male Basal Parameters Normotensive Hypertensive P Normotensive Hypertensive P n Age, y 51.4 ± ± 1.5 NS 53.6 ± ± 2.1 NS BMI, kg/m ± ± 0.9 NS 26.4 ± ± 0.6 NS Smoker, % NS NS Hematocrit, fraction 0.38 ± ± NS 0.42 ± ± NS Creatinine, µmol/l 74.1 ± ± 2.2 NS 91.9 ± ± 2.2 NS Total cholesterol, mmol/l 5.7 ± ± 0.5 NS 5.8 ± ± 0.3 NS HDL cholesterol, mmol/l 1.5 ± ± 0.2 NS 1.1 ± ± Triglyceride, µmol/l 1.1 ± ± 0.3 NS 1.2 ± ± 0.3 NS Systolic BP, mmhg 121 ± ± 2 < ± ± 3 < Diastolic BP, mmhg 81 ± 1 99 ± 1 < ± ± 1 < HR, bpm 60 ± 1 69 ± ± 1 67 ± 2 < Data are mean ± SEM. Sevre et al risk factors. The subjects entered the study between September 1996 and February All patients gave wri en informed consent. The regional ethical research commi ees in both countries approved the protocol. Baseline characteristics at inclusion are summarized in Tables 1 and 2. Study Procedure. All subjects were examined at 2 visits. At the first visit, patients were advised to stop taking any antihypertensive drugs. The patients who terminated antihypertensive medication were scheduled for BP control once a week after the first visit. At 4 weeks after the first visit, the second visit, the final assessment of eligibility, was performed. All examinations were performed in the morning in a quiet room with temperature 22 C to 24 C. The subjects were examined after an overnight fast and had refrained from alcohol and tobacco for at least the last 24 hours. HR and si ing sphygmomanometric BP were measured after 10 minutes of rest. Beat to beat BP and HR were recorded with the patient in the supine position with a Finapres (Ohmeda 2300) noninva sive BP monitor with the appropriate cuff applied to the third finger of the left hand. This instrument has been validated, and the accuracy and precision have been found sufficient for tracking of changes in BP and HR. 21 A 24 hour Holter ECG was applied to the chest (Marque e series 8500). Because mental stress can influence autonomic functions, the purpose of visit 1 was to familiarize the patient with the study procedure. Thus, only data from the second visit are presented. 105

115 Chapter 6 HRV Analysis. Twenty four hour ambulatory ECG recordings were analyzed on a Marque e laser Holter system (series 8000XP). HRV was analyzed as described previously 22 and in accordance with international guidelines. 23 Three ECG leads (modified leads V1,V5, and avf) and a time signal to correct for tape speed irregularities were recorded. The 24 hour recordings were divided into 288 segments of 5 minutes. Twelve 5 minute segments were averaged to obtain hourly mean values of the HRV parameters. All ectopic beats were classified, and only segments with 15% ectopy were used. Each nonnormal R R interval was substituted by the subsequent R R interval. Two experienced Holter analysts, with supervision of a cardiologist, analyzed all recordings. Normalized units, TP, LF, and HF have been defined previously in this report. BRS Measurement. Finapres recordings of 8 segments of 300 seconds of beat to beat BP and HR during rest in the supine position were used for determina tion of the BRS with the CARSPAN program (Pro- GAMMA bv), as described previously. 7,24,25 This program allows discrete Fourier transformation of nonequidistant samples of BP and R R interval series. The signals were tested for stationarity, and artifacts were corrected. Nonstationary signals or periods with >10% correction were excluded. Segments that lasted <100 seconds after this procedure were excluded. Subsequently, spectral analysis of systolic BP and R R interval length was performed, and BRS was calculated by the transfer function method. This method defines the BRS as the mean modulus between systolic BP and R R interval length spectra in the midfrequency band (0.07 to 0.14 Hz) with a coherence of >0.5. BRS is expressed in ms/mmhg. A BRS of 10 ms/mmhg indicates that a rise of 1 mmhg in systolic BP will induce 10 ms of R R interval lengthening. Statistical Analysis. On the basis of previous studies, 7,9 we expected a possible difference in BRS of 3 ms/mmhg between normotensive and hypertensive subjects. With a possible SD of 3 ms/mmhg, at least 15 patients and 15 controls should be examined with a power of 80%. Because we also planned a subgroup analysis based on gender differences, however, we included more than twice as many subjects in both groups. The data were analyzed with the use of SPSS statistical package (SPSS Inc). Nonnormal distributed data were natural log trans- 106

116 Autonomic Function in Hypertensive and Normotensive Subjects: The Importance of Gender formed. Two tailed statistical analyses of data were performed with Student s t test (P) and Pearson s correlation coefficient (r). Stepwise multiple regression analysis was performed with gender, BP group (hypertensive or normotensive), age, body mass index (BMI), smoking status, triglyceride, and total and HDL cholesterol as predictors and HRV parameters and BRS as dependent variables. Data are presented as mean ± SEM. The level of statistical signifi cance was set at P=0.05. Results Baroreceptor Reflex Sensitivity. All subjects had at least 1 available 300 second period for BRS measurement. Differences Between Hypertensive and Normotensive Subjects. BRS was reduced in the hypertensive patients compared with the normotensive controls (7.6 ± 0.6 versus 10.4 ± 0.8 ms/mmhg, respectively; P=0.005). Gender Differences. Hypertensive women had lower BRS than normotensive women. The difference in BRS between male hypertensive and normotensive subjects did not reach statistical signifi cance. Female hypertensive subjects had lower BRS than male hypertensives. BRS did not differ significantly between the 2 normotensive groups (Figure 1). BRS correlated with age in the hypertensive and normotensive male 15 BRS (ms/mmhg) NT females HT females NT males HT males Figure 1. BRS in 41 hypertensive (HT) and 34 normotensive (NT) subjects. BRS was significantly lower in the female hypertensive group than in the female normotensive group but did not differ sig nificantly between the 2 male groups. BRS was significantly lower in the female hypertensive group than in the male hypertensive group but did not differ significantly between the 2 normotensive groups. ***P<0.0005; **P<

117 Chapter 6 Table 3. Stepwise Multiple Regression With Gender, BP Group, Age, HDL Cholesterol, BMI, Total Cholesterol, Triglyceride, and Smoking Status as Predictors and BRS, Mean R R, SDNN, PNN50, RMSSD, TP, LF, HF, LF/HF Ratio, LF Normalized Units, and HF Normalized Units as Dependent Variables Gender* HT or NT Age HDL Cholesterol Dependent Variables β P β P β P β P R 2 BRS, ms/mmhg < NS 0.42 Mean R R, ms NS < SDNN, ms NS NS NS PNN50, % < < RMSSD, ms NS NS 22.5 < TP, ms < < LF, ms < NS 0.35 HF, ms NS LF/HF ratio NS NS < LF normalized units NS NS NS 11.0 < HF normalized units NS NS NS 11.0 < BMI, total cholesterol, and triglyceride were not independent explanatory variables. Smoking status was significantly related to TP and RMSSD (P=0.03, both). Only gender, BP group, age, and HDL cholesterol were included in R 2. * Male=1, female=2. Normotensivecontrols (NT)=0; hypertensive patients (HT)=1. groups (r= 0.65 and r= 0.61, respectively; P<0.01) but not in the female groups. In the female hypertensive group only, we found a significant correlation between systolic BP and BRS (r= 0.51; P<0.04). Multiple Regression Analysis. Gender, age, and presence of hypertension were the only significant independent explanatory variables of BRS. BRS decreased with increasing age and was lower in female and hypertensive subjects than in men and normotensive subjects (Table 3). Heart Rate Variability. Seventy five 24 hour ECG recordings were analyzed. All had at least 18 hour recordings suitable for HRV analysis. Differences Between Hypertensive and Normotensive Subjects. PNN50, RMSSD, TP, LF, and HF were lower in the hyper tensive group than in the normotensive group. We did not find any significant differences in normalized units or LF/HF ratio between the hypertensive and normotensive subjects, although there was a nonsignificant tendency of higher LF/HF ratio in the hypertensive group (Table 4). Gender Differences. Hypertensive women had higher HF normalized units and lower LF normalized units and LF/HF ratio than hypertensive men. Similar gender differences were found in the normotensive groups. The hypertensive women had significantly lower TP and LF 108

118 Autonomic Function in Hypertensive and Normotensive Subjects: The Importance of Gender than the hypertensive men (Figures 2 and 3). There were significant differences between hypertensive and normotensive women in the same parameters as seen between the entire hypertensive and entire normotensive group (Figures 2 and 3). We did not find any significant differences between hyperten sive and normotensive men in any of the HRV parameters. Multiple Regression Analysis. Age, gender, HDL cholesterol, and presence of hypertension were the only significant independent explanatory variables of HRV numerical values, ie, the variability decreased with increasing age and was blunted in hypertensive subjects and women as opposed to normotensive subjects and men. HRV increased with increasing HDL concentrations. The normal ized units of HF and LF were only related to HDL choles terol, ie, the higher the HDL concentrations were, the lower were LF normalized units and the higher were HF normalized units. LF/HF ratio was related to both HDL and age, ie, LF/HF ratio decreased with increasing HDL and age (Table 3). Hemoglobin, Hematocrit, Creatinine, and Blood Lipids. Table 4. HRV in Normotensive vs Hypertensive Subjects HRV Parameters Normotensive Hypertensive P n Mean R R, ms 836 ± ± 12 NS SDNN, ms 151 ± ± 5 NS PNN50, % 9.8 ± ± RMSSD, ms 35.7 ± ± TP, ms ± ± LF, ms ± ± HF, ms ± ± 23 NS (0.055) LF/HF ratio 4.75 ± ± 0.48 NS LF normalized units 80.2 ± ± 1.4 NS HF normalized units 19.8 ± ± 1.4 NS Data are mean ± SEM. Differences Between Hypertensive and Normotensive Subjects. HDL cholesterol was higher in the hypertensive than in the normotensive group. We did not find any statistically significant differences in hematocrit, creatinine, total cholesterol, or triglycerides (Table 1). Gender Differences. Hypertensive men had higher HDL than normotensive men. There were no statistically significant HDL differences be tween the 2 female groups. Female normotensive subjects had higher HDL (P<0.0005) than male normotensive subjects. Creatinine 109

119 Chapter TP, LF, HF (ms 2 ), HF nu, LF nu(%), LF/HF TP LF HF LF nu HF nu LF/HF Figure 2. HRV frequency domain parameters in 41 hypertensive (HT) and 34 normotensive (NT) subjects. TP and LF were signifi cant lower in hypertensive women than in normotensive women. HF was lower as well, but not at the level of significance. TP, LF, HF normalized units (nu), LF normalized units, and LF/HF ratio differed significantly between the male and female hyper tensive groups. LF normalized units, HF normalized units, and LF/HF ratio differed significantly between the male and female normotensive groups. *P0.05; **P0.01. Mean RR, SDNN, RMSSD (ms), PNN50 (%) Mean RR SDNN RMSSD PNN50 Figure 3. HRV time domain parameters in 41 hypertensive (HT) and 34 normotensive (NT) subjects. PNN50 and RMSSD were significantly lower in hypertensive women than in normotensive females. *P<0.05; **P<

120 Autonomic Function in Hypertensive and Normotensive Subjects: The Importance of Gender and hematocrit was higher in the male group than in the female group (P<0.001). We did not find any gender differences in cholesterol and triglyceride (Table 2). Discussion The present study demonstrated significantly lower BRS in hypertensive subjects than in normotensive controls. More over, the reduction in BRS was most pronounced in the female group and did not reach statistical significance in men. BRS was also significantly reduced in hypertensive women compared with hypertensive men. Systolic BP correlated significantly with BRS in the female hypertensive group only. In addition, we found significantly reduced TP, LF, PNN50, and RMSSD in hypertensive patients compared with normoten sive controls. In our subgroup analysis these differences were only observed in the women and not in the men. LF normalized units and LF/HF ratio were higher in men than in women in both BP groups. By multiple regression analysis, gender, age, and BP were independent determinants of BRS and HRV, TP, LF, HF, PNN50, and RMSSD. High HDL was associated with high values of HRV with the exception of LF normalized units and LF/HF ratio, which were negatively correlated. Thus, in the present study we have demonstrated substantial changes in the autonomic function in hypertensive patients. Moreover, autonomic dysfunction seems to play a more prominent role in female than in male hypertension. The 4 groups were well matched with respect to gender, smoking, BMI, and age, making our subgroup analyses feasible without any confounding factors. BRS was calculated with the transfer technique, which utilizes the spontaneous fluctuations in systolic BP and HR mainly induced by the respiration to estimate the BRS. It does not interfere with any cardiovascular control mechanisms 26 and correlates well with the phenylephrine ramp method in healthy 24 and hypertensive subjects. 7 Our BRS results confirm the findings of reduced BRS in hypertensive patients BRS did not differ between normo tensive men and women. This is in accordance with Laitinen et al, 14 who did not find any differ- 111

121 Chapter 6 ence between healthy postmeno pausal female subjects aged 60 to 77 years and male age matched subjects. Our finding of a substantial reduction of BRS, especially in female hypertension, warrants some comments. Rapid changes in HR following alterations in BP are mediated by the barore ceptor reflex arc, which is an important part of hemodynamic homeostasis. It has been suggested that hypertension is associ ated with a rese ing of the reflex arc at a higher set point. 27 Female sex hormones may also be important in modeling female arteries. The compliance of the brachial artery is higher in women than in men. 28 Pregnant women have thinner arterial intima layer and thicker media layer than controls. 29 Estrogen therapy alone or in combination with simvastatin improves flow mediated dilation of the brachial artery. 30 Interestingly, estrogen alone also increases HDL cholesterol concentrations. 30 These studies propose that female sex hormones model the arterial layers, which are crucial to the arterial BP buffer capacity and hence blood pressure variability and BRS. Furthermore, estrogen replacement therapy increases BRS in postmenopausal healthy women. 17 In our study all but 1 woman were postmeno pausal, and BRS did not differ significantly between the male and the female normotensive groups. Arteriosclerotic plaques may mechanically cause arterial rigidity and consequently de crease BRS. 8 It is unlikely, however, that the women in our hypertensive group have more arteriosclerotic changes than their male counterparts. BRS is also lower in hypertensive subjects with insulin resistance than in hypertensive subjects with normal insulin tolerance. 6 Our study lacks data regarding insulin resis tance, but there were no significant differences in BMI between the 4 groups, nor did any subjects suffer from diabetes. For these reasons we find it unlikely that different sex hormone concentrations, arteriosclerotic changes, or insulin resistance can ex plain our results. Moreover, the close relationship between BRS and BP only in hypertensive women supports the assumption that autonomic dysfunction may play a more important role in female hypertension. The HRV differences between hypertensive and normoten sive subjects demonstrated in the present study are in accordance with previous studies on HRV regarding the reduction in TP, HF, LF, 2,4 6 and PNN50. 2 We found a small and nonsignificant increase in the LF/HF ratio and LF normalized units in the hypertensive group compared with the normotensive group. This is in accordance with Pikkujamsa 112

122 Autonomic Function in Hypertensive and Normotensive Subjects: The Importance of Gender et al 6 but not with Guzze i et al, 3 who demonstrated significantly increased LF normalized units in the hypertensive group, and Huikuri et al, 5 who dis played decreased LF/HF ratio in the hypertensive group. Our subjects were examined twice, whereas Guzze i et al only examined their subjects once. This may explain the diverging results. Most probably, the responses to the laboratory exami nation per se will differ between the first and second examinations. 31 The subjects in the study by Huikuri et al 5 were using vasoactive drugs at the time of the examination, which may have influenced the results. As demonstrated by other investigators 4,17,19 we found higher PNN50, LF/HF ratio, and LF nor malized units in male normotensives than in female normoten sives. We also confirmed the results of Singh et al 4 regarding PNN50, RMSSD, and LF/HF ratio in hypertensive men com pared with hypertensive women, but, in contrast, we did not find any significant HRV discrepancies between the male hyperten sive and normotensive groups. The probability of type II errors should be considered. The reduction of overall HRV in hypertensive patients is more pronounced in women than in men. On the other hand, the relative HRV parameters, ie, normalized units and the LF/HF ratio, did not reveal any significant differences between the normotensive and hypertensive subjects in either the male or the female group. We can only speculate as to possible explanations of these results. One interpretation could be a generally more pronounced withdrawal of autonomic HR control in hyperten sive women than in hypertensive men, even though the balance between the sympathetic and parasympathetic nervous system is similar. This assumption is further supported by the BRS results, which suggest less HR responsiveness to changes in systolic BP in hypertensive women compared with hypertensive men and normotensive subjects. During the past decade, BRS has proven to be a powerful independent marker of increased risk for malignant cardiac arrhythmias and sudden death after myocardial infarction. 12,32 HRV measurements have also been investigated as predictors of cardiovascular morbidity. However the results have been diverg ing. While TP, 33 ultra low frequency power, 33 very low fre quency power, 33 and SDNN 32 have proven to be independent markers of cardiovascular morbidity, the results regarding HF and LF have been less convincing. 33 We still lack a physiolog ical understanding of the former HRV parameters, beyond that they reflect 113

123 Chapter 6 a general variability. 23,33 Conversely, the physiolog ical basis for the latter is be er understood. 23 On the basis of reports in patients with other kinds of cardiovascular diseases, 12,32,33 however, we might anticipate an association between low BRS and reduced HRV with cardio vascular morbidity in hypertensive subjects as well. On the basis of these considerations, our findings may imply that hyperten sive women are more susceptible to cardiac events and arrhyth mias than hypertensive men. No long term studies, however, have been performed to investigate this possibility. Acknowledgment This study was supported by KNOLL, BASF Pharma, The Netherlands. 114

124 Autonomic Function in Hypertensive and Normotensive Subjects: The Importance of Gender References 1. Goldstein DS. The fact of organization. In: Goldstein DS, ed. Stress, Catecholamines, and Cardiovascular Disease. New York, NY: Oxford University Press; 1995: Chakko S, Mulingtapang RF, Huikuri HV, Kessler KM, Materson BJ, Myerburg RJ. Alterations in heart rate variability and its circadian rhythm in hypertensive patients with left ventricular hypertrophy free of coronary artery disease. Am Heart J. 1993; 126 : Guzze i S, Piccaluga E, Casati R, Ceru i S, Lombardi F, Pagani M, Malliani A. Sympathetic predominance in essential hypertension: a study employing spectral analysis of heart rate variability. J Hypertens. 1988;6: Singh JP, Larson MG, Tsuji H, Evans JC, O Donnell CJ, Levy D. Reduced heart rate variability and new onset hypertension: insights into pathogenesis of hyper tension: the Framingham Heart Study. Hypertension. 1998;32: Huikuri HV, Ylitalo A, Pikkujamsa SM, Ikaheimo MJ, Airaksinen KE, Rantala AO, Lilja M. Heart rate variability in systemic hypertension. Am J Cardiol. 1996;77: Pikkujamsa SM, Huikuri HV, Airaksinen KE, Rantala AO, Kauma H, Lilja M, Savolainen MJ, Kesaniemi YA. Heart rate variability and baroreflex sensitivity in hypertensive subjects with and without metabolic features of insulin resistance syndrome. Am J Hypertens. 1998;11: Watkins LL, Grossman P, Sherwood A. Noninvasive assessment of baroreflex control in borderline hypertension: comparison with the phenyl ephrine method. Hypertension. 1996;28: James MA, Robinson TG, Panerai RB, Po er JF. Arterial baroreceptor cardiac reflex sensitivity in the elderly. Hypertension. 1996;28: Grassi G, Ca aneo BM, Seravalle G, Lanfranchi A, Mancia G. Baroreflex control of sympathetic nerve activity in essential and secondary hypertension. Hypertension. 1998;31: Korner PI, West MJ, Shaw J, Uther JB. Steady state properties of the baroreceptor heart rate reflex in essential hypertension in man. Clin Exp Pharmacol Physiol. 1974;1:

125 Chapter La Rovere MT. Autonomic markers of prognosis after myocardial infarction. Clin Sci (Colch). 1996;91(suppl): Hohnloser SH, Klingenheben T, van de Loo A, Hablawetz E, Just H, Schwartz PJ. Reflex versus tonic vagal activity as a prognostic parameter in patients with sustained ventricular tachycardia or ventricular fibrillation. Circulation. 1994;89: Schwartz PJ, La Rovere MT, Vanoli E. Autonomic nervous system and sudden cardiac death. Circulation. 1992;85(suppl I):I 77 I Laitinen T, Hartikainen J, Vanninen E, Niskanen L, Geelen G, Lansimies E. Age and gender dependency of baroreflex sensitivity in healthy subjects. J Appl Physiol. 1998;84: Nordby G, Os I, Kjeldsen SE, Eide I. Mild essential hypertension in nonobese premenopausal women is characterized by low renin. Am J Hypertens. 1992; 5: Mundal HH, Nordby G, Lande K, Gjesdal K, Kjeldsen SE, Os I. Effect of cold pressor test and awareness of hypertension on platelet function in normotensive and hypertensive women. Scand J Clin Lab Invest. 1993;53: Huikuri HV, Pikkujamsa SM, Airaksinen KEJ, Ikaheimo MJ, Rantala AO, Kauma H, Lilja M, Kesaniemi YA. Sex related differences in autonomic modulation of heart rate in middle aged subjects. Circulation. 1996;94: Convertino VA. Gender differences in autonomic functions associated with blood pressure regulation. Am J Physiol. 1998;275(pt 2): R1909 R Ryan SM, Goldberger AL, Pincus SM, Mietus J, Lipsitz LA. Gender and age related differences in heart rate dynamics: are women more complex than men? J Am Coll Cardiol. 1994;24: Abdel Rahman AR, Merrill RH, Wooles WR. Gender related differences in the baroreceptor reflex control of heart rate in normotensive humans. J Appl Physiol. 1994;77: Imholz BP, Wieling W, van Montfrans GA, Wesseling KH. Fifteen years experience with finger arterial pressure monitoring: assessment of the tech nology. Cardiovasc Res. 1998;38:

126 Autonomic Function in Hypertensive and Normotensive Subjects: The Importance of Gender 22. Tuininga YS, Crijns HJGM, Brouwer J, van den Berg MP, in t Veld AJM, Mulder G, Lie KI. Evaluation of importance of central effects of atenolol and metoprolol measured by heart rate variability during mental performance tasks, physical exercise, and daily life in stable postinfarct patients. Circu lation. 1995;92: Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Heart rate variability: standards of measurement, physiological interpretation, and clinical use. Eur Heart J. 1996;17: Robbe HW, Mulder LJ, Ruddel H, Langewitz WA, Veldman JB, Mulder G. Assessment of baroreceptor reflex sensitivity by means of spectral analysis. Hypertension. 1987;10: Lefrandt JD, Hoogenberg K, van Roon AM, Dullaart RP, Gans RO, Smit AJ. Baroreflex sensitivity is depressed in microalbuminuric type I diabetic patients at rest and during sympathetic manoeuvres. Diabetologia. 1999;42: Maestri R, Pinna GD, Mortara A, La Rovere MT, Tavazzi L. Assessing baroreflex sensitivity in post myocardial infarction patients: comparison of spectral and phenylephrine techniques. J Am Coll Cardiol. 1998;31: Parati G, Fra ola A, Omboni S, Mancia G, Di Rienzo M. Analysis of heart rate and blood pressure variability in the assessment of autonomic regulation in arterial hypertension. Clin Sci (Colch). 1996;91(suppl): Heijden Spek JJ, Staessen JA, Fagard RH, Hoeks AP, Boudier HA, van Bortel LM. Effect of age on brachial artery wall properties differs from the aorta and is gender dependent: a population study. Hypertension. 2000;35: Sator MO, Joura EA, Gruber DM, Obruca A, Zeisler H, Egarter C, Huber JC. Non invasive detection of alterations of the carotid artery in pregnant women with high frequency ultrasound. Ultrasound Obstet Gynecol. 1999;13: Koh KK, Cardillo C, Bui MN, Hathaway L, Csako G, Waclawiw MA, Panza JA, Cannon RO III. Vascular effects of estrogen and cholesterol lowering therapies in hypercholesterolemic postmenopausal women. Circulation. 1999;99:

127 Chapter Benetos A, Safar ME. Response to the cold pressor test in normotensive and hypertensive patients. Am J Hypertens. 1991; 4(pt 1): La Rovere MT, Bigger JT Jr, Marcus FI, Mortara A, Schwartz PJ. Baroreflex sensitivity and heart rate variability in prediction of total cardiac mortality after myocardial infarction. Lancet. 1998;351: Bigger JT Jr, Fleiss JL, Steinmann RC, Rolnitzky LM, Kleiger RE, Ro man JN. Frequency domain measures of heart period variability and mortality after myocardial infarction. Circulation. 1992;85:

128 Chapter 7 The Effects of Dihydropyridine and Phenylalkylamine Calcium Antagonist Classes on Autonomic Function in Hypertension: The VAMPHYRE Study American Journal of Hypertension Nov;14: J.D. Lefrandt 1, J. Heitmann 2, K. Sevre 3, M. Castellano 4, M. Hausberg 5, M. Fallon 6, L. Fluckiger 7, A. Urbigkeit 2, M. Rostrup 3, E. Agabiti Rosei 4, K.H. Rahn 5, M. Murphy 6, F. Zannad 7, PJ. de Kam 1, A.M. van Roon 1, and A.J. Smit 1 1 University Hospital of Groningen, Groningen, The Netherlands; 2 Klinikum der Philipps Universität, Marburg, Germany; 3 Ullevål Hospital, Oslo, Norway; 4 Università Degli Studi di Brescia, Brescia, Italy; 5 Klinik der Westfa lischen Wilhelms Universität, Münster, Germany; 6 Cork University Hospital, Cork, Ireland; and 7 Centre d Investigation Clinique Hoˆpital Jeanne d Arc (LF, FZ), Nancy, France.

129 Chapter 7 Summary The aim of the present study was to compare the effects of a long acting dihydropyridine (amlodipine) and a nondihydropyridine (verapamil) on autonomic function in patients with mild to moderate hypertension. A total of 145 patients with a diastolic blood pressure (BP) between 95 and 110 mmhg received 8 weeks of verapamil sustained release (240 mg) and amlodipine (5 mg) in a prospective randomized, double blind, cross over study, both after 4 weeks of placebo. The 24 h autonomic balance was measured by analysis of 24 h heart rate variability and short term autonomic control of BP by baroreflex sensitivity measurements. Plasma norepinephrine was sampled at rest. Blood pressure was equally reduced from 153/100 mmhg to 139/91 mmhg with verapamil and 138/91 mmhg with amlodipine, P=0.50/.59. The low to high frequency ratio (LF/HF), reflecting sympathovagal balance, was higher with amlodipine than with verapamil (4.66 v 4.10; P=0.001). Baroreflex function was improved by both treatments; however, baroreflex sensitivity (BRS) was significantly higher with verapamil than with amlodipine (8.47 v 8.06 msec/mmhg; P=0.01). Plasma norepinephrine (NE) level was higher with amlodipine than with verapamil (1.59 v 1.32 nmol/l; P<0.0001). Amlodipine induces a shift in sympathovagal balance, as measured by heart rate variability indices and plasma NE, toward sympathetic predominance compared with vagal predominance with verapamil. Short term autonomic control of BP, as assessed by BRS, is more effectively improved by verapamil than by amlodipine. These contrasting effects on autonomic function suggest that the nondihydropyridine calcium antagonist verapamil may have additional beneficial effects beyond lowering BP compared with the dihydropyridine amlodipine. Introduction Increased heart rate has long been identified as a marker of sympathetic predominance in hypertension and is related to a higher mortality. 1,2 Spectral analysis of heart rate variability (HRV) and recordings of muscle sympathetic nerve activity by microneurography have shown increased sympathetic activity and decreased vagal activity in hyper- 120

130 The Effects of Dihydropyridine and Phenylalkylamine Calcium Antagonist Classes on Autonomic Function in Hypertension: The VAMPHYRE Study tension. 3,4 Measurement of baroreflex sensitivity (BRS) by cross spectral analysis and phenylephrine infusions has revealed a diminished baroreflex function in hypertension. 5 Although it is still unclear whether these changes to the autonomic nervous system are pathogenetically involved or are secondary to consequences of chronically elevated blood pressure (BP), they may contribute to the increased cardiovascular risk in established hypertension. 2,6 Reducing sympathetic tone and restoration of baroreflex function by antihypertensive drugs could therefore be beneficial; or, at least, antihypertensive drugs should not deteriorate them. Calcium antagonists are widely used to treat hypertension. However, the influences on the autonomic nervous system may differ between the various classes The use of short acting calcium antagonists has been associated with increased morbidity and mortality in patients with hypertension and coronary artery disease. 12 It has been suggested that this could at least partly be due to reflex sympathetic activation in response to vasodilation and a reduction of peripheral resistance. 7 In contrast to short acting dihydropyridines (DHP), long acting DHP and non DHP would have similar effects and would not increase sympathetic activity. However, very few data are available from direct comparisons of the effects of the la er two classes on autonomic function in hypertension. Therefore, the aim of the present study was to compare the effects of a long acting dihydropyridine (amlodipine) and a nondihydropyridine (verapamil) on autonomic function in patients with mild to moderate hypertension in a multinational prospective, double blind, randomized, cross over comparison. The low to high frequency (LF/HF) ratio and BRS were chosen as primary efficacy parameters to evaluate the effects of the drugs on 24 h sympathovagal balance 13,14 and on short term reflex autonomic control of BP. 15 Methods Study Design and Subjects. The effects of Verapamil and Amlodipine on autonomic function in Patients with Hypertension at Rest and during Exercise (VAMPHYRE) study was a comparison of verapamil sustained release (SR, 240 mg) and amlodipine (5 mg) in hypertensive patients >18 years of age. Hypertension was defined as a diastolic BP 121

131 Chapter 7 (DBP) 95 mmhg on at least three occasions. In these patients, either hypertension was newly diagnosed or the current antihypertensive treatment did not meet therapeutic goals (systolic BP [SBP] <140 mmhg and DBP 90 <mmhg). Patients were excluded if secondary hypertension was suspected or if they had had a recent cardiovascular event, had diabetes mellitus, or used drugs known to influence the autonomic nervous system. Seven centers in six European Community countries participated. Informed consent was obtained from all patients. The local ethics commi ees at each participating hospital approved the VAMPHYRE protocol. At the first patient visit, any antihypertensive treatment was stopped and 4 weeks of placebo treatment was started. Si ing BP was measured three times after 10 min of rest with 2 min intervals. Randomization to sequence 1 (8 weeks of verapamil, 4 weeks of placebo, 8 weeks of amlodipine) or sequence 2 (8 weeks of amlodipine, 4 weeks of placebo, 8 weeks of verapamil) followed at the second visit if DBP was 95 and 110 mmhg and SBP was 180 mmhg. The first placebo period was open; both active drug periods were double blind; and the second placebo period was single blind. Blood pressure measurement and all assessments of autonomic function were performed at the start of the study and after each placebo and each treatment period. Study Procedure. All tests were performed in the morning in a warm, quiet room. Patients refrained from eating and smoking and from drinking alcohol, coffee, and tea that day. A Finapres (Ohmeda 2300; Ohmeda, Liberty Corner, NJ) blood pressure monitor was a ached to the third finger of the patient s right hand. Ten min after insertion of an intravenous catheter, heart rate and si ing and supine BP were measured. Venous blood, 6 ml, was sampled in a prechilled EGTA gluthation tube for determination of plasma norepinephrine (NE) level by electrochemical detection after high pressure liquid chromatography in a central laboratory (Analytico, Breda, the Netherlands) for all centers. After 30 min of rest, BRS was measured; then a 24 h Holter electrocardiographic (ECG) recorder was applied. Because the autonomic nervous system is very much influenced by mental stress, the purpose of the first visit was only to get the patient acquainted with the study procedures. No ECG recorder was applied on the first visit. 122

132 The Effects of Dihydropyridine and Phenylalkylamine Calcium Antagonist Classes on Autonomic Function in Hypertension: The VAMPHYRE Study Holter ECG: Heart Rate Variability Analysis. Twenty four hour ambulatory ECG recordings were acquired by a Marque e Holter recorder (Series 8500) (Marque e Electronics Inc., Milwaukee, WI) and analyzed by an experienced analyst. Three ECG leads (modified leads V 1,V 5, and a VF) and a time signal to correct for tape speed irregularities were recorded. The HRV was analyzed with the COHORT program, as described previously 16 and in accordance with the recommendations of the Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. 17 Low frequency power (LF, 0.04 to 0.15 Hz) reflects sympathetic and partly vagal activity in contrast to high frequency power (HF, 0.15 to 0.40 Hz), which primarily reflects vagal activity. 18 The LF/HF ratio is the ratio of LF and HF powers and is considered to reflect sympathovagal balance. 13,14 Baroreflex Sensitivity Measurement. Three hundred second segments of Finapres recordings during rest were used to determine the BRS with the CARSPAN program (IEC ProGamma, the Netherlands), as described previously. 15 After detection and correction of artifacts, BRS was calculated by the transfer function method, defining BRS as the mean modulus of the cross spectrum of SBP and R peak to R peak (RR) interval variability in the 0.07 to 0.15 Hz frequency band with at least 0.5 coherence. The phase between SBP and RR interval length spectra indicates the delay between a BP variation and the subsequent adaptation of RR interval length. Because RR interval modulation usually follows a change in BP, the phase is mostly a negative value. As vagal reflexes act more rapidly than sympathetic reflexes, a change in phase toward a less negative value (ie, a shorter delay from change in BP to RR interval response) indicates a shift toward vagal predominance. In contrast, a change in phase toward a more negative value (ie, a longer delay) indicates a shift toward sympathetic predominance. 19 Sample Size and Statistical Analysis. Monitoring and statistical analysis (SAS software 6.12; SAS Institute, Cary, NC) were performed by an independent monitoring agency (TCC, Groningen, the Netherlands). Sample size calculation was based on the 24 h LF/HF ratio of a recent trial performed in our hospital. 16 Anticipating a standard deviation of 2.5, a total of 130 evaluable patients (65 in each sequence) were needed to detect a difference of in a two sided t test for α=0.05 with 80% power. 123

133 Chapter 7 Table 1. Patient characteristics at randomization. Mean (Range) Characteristic or Number (%) Male/female, N (%) 95/50 (65%/35%) Age (y) 51 (20 72) Systolic blood pressure (mmhg) 153 ( ) Diastolic blood pressure (mmhg) 100 (95 110) Heart rate (beats/min) 67.8 (51 102) Body mass index (kg/m 2 ) 28.6 ( ) Smokers, current, N (%) 33 (23%) Dyslipidemia, N (%)* 57 (40%) Family history of premature atherosclerotic disease, N (%) 13 (9%) Serum creatinine (µmol/l) 85 (59 141) * Dyslipidemia is defined as total cholesterol >6.5 mmol/l, low density lipoprotein >5 mmol/l, high density lipoprotein <0.9 mmol/l, or triglycerides >2.5 mmol/l. Primary statistical analysis was performed on the intention to treat population, which consisted of all patients randomized with at least one dose of study medication taken. The per protocol population was defined as all patients with all four fully evaluable Holter ECG and no major protocol violations. A mixed analysis of covariance model was defined to test the null hypothesis of no difference between both treatments. Period, treatment, and center were defined as fixed factors, the placebo period preceding the active drug as covariate, and patient within center as random effect to account for the intrapatient correlation in this cross over se ing. A P value <0.05 in the two sided test was considered statistically significant. The student t test was performed for all endpoints to check for any differences between the first and second placebo period. All parameters that were nonnormally distributed and could not be transformed to a normal distribution were tested with a Wilcoxon test. Secondary models were defined to exclude the influence of a carry over effect. Results Patient Recruitment,Follow Up, and Adverse Events. Of 168 patients who entered the placebo run in phase, 23 patients left the study because they did not meet randomization criteria, and 145 were randomized at the second visit. Patient characteristics at randomization are summarized in Table 1. Because no significant difference between both placebo periods existed for BP or any autonomic function parameter, the mean 124

134 The Effects of Dihydropyridine and Phenylalkylamine Calcium Antagonist Classes on Autonomic Function in Hypertension: The VAMPHYRE Study Table 2. Effects of verapamil and amlodipine on short term autonomic control of blood pressure. Treatment P Value Verapamil Amlodipine Difference Treatment Placebo (V) (A) (V A) Difference BRS (msec/mmhg) (6.42, 6.90) (7.11, 10.1) (6.76, 9.60) (0.09, 0.68) BPV <.0001 (mmhg 2 ) (5.01, 5.49) (2.30, 3.57) (2.96, 4.58) ( 0.75, 0.44) Phase <.0001 (radians) ( 1.33, 1.41) ( 1.48, 1.18) ( 1.65, 1.34) (0.11, 0.22) BRS = baroreflex sensitivity; BPV = blood pressure variability. Data given as means (95% confidence intervals). of both is shown. A total of 22 patients later dropped out of the study: 13 during verapamil use, four during the second placebo period, and five during amlodipine use. The mean compliance, assessed by tablet count, was >95% for all treatments. The main reasons for premature study discontinuation were adverse events and lack of compliance. The design of the study does not allow an unbiased assessment of the relation between adverse events and study drugs. The adverse events with a probable relation with the drugs were most frequently common side effects: headache, dizziness, palpitations, constipation, and peripheral edema. However, three generalized allergic responses to verapamil that faded after withdrawal of the drug were observed. Blood Pressure and Heart Rate Response. After placebo, the mean SBP and DBP were 153 and 100 mmhg, respectively. Both were equal after verapamil and amlodipine treatment (139/91 v 138/91 mmhg; P=.50/.59). After placebo, mean heart rate was 68 beats/min. Heart rate was lower with verapamil than with amlodipine (65 v 69 beats/min; P<.0001). Twenty Four Hour Heart Rate Variability. Of the 528 Holter ECG that were applied, 453 had at least 18 h suitable for HRV analysis after removal of artifacts, nonstationary segments, and segments with >15% ectopy (Figure 1). After placebo, the primary endpoint LF/HF ratio was The LF/HF ratio was higher after amlodipine than after verapamil (4.66 v 4.10). This difference was highly significant in both in the intention to treat (P=0.001) and per protocol (P=0.01) analyses. No change in overall HRV (total power) were observed after treatment. Baroreflex Sensitivity. Both verapamil (+27%) and amlodipine (+21%) 125

135 Chapter 7 Figure 1. Heart rate variability before and after treatment with verapamil and amlodipine. total power ( to 0.40 Hz); low frequency power (0.40 to 0.15 Hz); high frequency power (0.15 to 0.40 Hz); LF/HF ratio (low to high frequency power, reflecting sympathovagal balance. Data given as mean (SEM). *P=0.05, **P=0.001 verapamil v amlodipine. increased BRS (Table 2), but BRS was significantly higher after verapamil (P=0.01). Short term BPV was reduced by both verapamil ( 45%) and amlodipine ( 30%), but was significantly lower with verapamil (P<.0001). Phase was significantly shorter during verapamil treatment (P<.0001). Plasma Norepinephrine. Of the 477 samples drawn, two were below detection limit (<0.10 nmol/l). After placebo, plasma NE level was 1.35 nmol/l. Verapamil did not change plasma NE, but amlodipine increased plasma NE by 22%, resulting in significantly higher NE levels compared with those achieved with verapamil (1.59 v 1.32 nmol/l; P<.0001, Figure 2). 126

136 The Effects of Dihydropyridine and Phenylalkylamine Calcium Antagonist Classes on Autonomic Function in Hypertension: The VAMPHYRE Study Figure 2. Plasma norepinephrine before and after treatment with verapamil and amlodipine. Data given as mean (SEM). *P<0.001 verapamil v amlodipine. Discussion This study demonstrates that the effects of the long acting dihydropyridine amlodipine and the phenylalkylamine verapamil on the autonomic nervous system are markedly different, whereas the reduction in BP by the two drugs is similar. Effects on Sympathovagal Balance. The primary endpoint, LF/HF ratio, was significantly higher with amlodipine compared with verapamil, suggesting sympathetic predominance during amlodipine treatment compared with vagal predominance during verapamil treatment. The other indices of sympathovagal interaction, heart rate, and phase between short term BP and HRV showed the same contrasting effects. One might object that the contrasting effect on LF/HF ratio is due to the heart rate lowering effect of verapamil (caused by the direct effect of verapamil on the sinus node), rather than to modulation of the sympathovagal balance. However, when we added the mean 24 h RR interval length to the model as a covariate to study the influence of the effects of heart rate on the primary outcome, the difference in LF/HF ratio between verapamil and amlodipine remained significant (4.20 v 4.55, P=0.03). These contrasting effects on sympathovagal balance, as measured by LF/HF ratio, could result from modulation of sympathetic or vagal activity, or both. Compared with baseline level, amlodipine 127

137 Chapter 7 increased plasma NE, whereas verapamil had no effect. The HF power (reflecting vagal modulation) increased with verapamil, in contrast to the absence of effect with amlodipine. These findings imply that amlodipine increased sympathetic activity, whereas verapamil increased vagal tone. There is no consensus whether amlodipine raises sympathovagal balance in hypertension when given long term. De Champlain et al found that amlodipine raised heart rate and increased plasma NE by 50% after 6 weeks therapy. 8 In contrast, Hamada et al reported no changes in heart rate or urinary NE and decreases in plasma NE and LF/HF ratio after 4 weeks of amlodipine in a comparison with nifedipine and slow release nifedipine. 20 However, large differences in baseline values of urinary and plasma NE and of heart rate existed between the small groups (n=16), which often exceeded the treatment effect. Sakata et al found that amlodipine did not significantly change plasma NE in 24 patients after 2 weeks to 3 months, although the absolute difference between mean plasma NE before and after therapy was about as large (increase of 0.24 nmol/l) as in our study. 9 However, all of these studies had a relatively small sample size (n=16 to 24 in the amlodipine group) and a parallel design, therefore increasing the chance of a type II error. In contrast, the 145 patients and the cross over design of the present study result in a considerably greater statistical power. The heart rate moderating effect of verapamil is well documented. 10,11,21 Furthermore, verapamil is thought either to not affect plasma NE 21 or to decrease it. 11 In one of the few studies with a cross over design, 4 weeks of felodipine had no effect on heart rate and plasma NE, whereas 4 weeks of verapamil lowered both. 10 In a review of 20 clinical studies reporting the long term effects of long acting calcium antagonists in patients with hypertension, heart rate was unchanged; plasma NE levels increased by 14.5% ± 5% after long acting DHP calcium antagonists; and heart rate and plasma NE levels decreased after long acting non DHP calcium antagonists by 7.1% ± 1.8% and 20.7% ± 9.8%, respectively (P<0.001 between groups for both comparisons). 7 Our study supports the notion of opposite effects of verapamil and amlodipine on sympathovagal balance, as assessed by HRV indices and plasma NE. Vasodilation induced by amlodipine probably accounts for the reflex increase of sympathetic activity. Nazzaro et al showed, besides a decrease in heart rate with verapamil and an increase in heart 128

138 The Effects of Dihydropyridine and Phenylalkylamine Calcium Antagonist Classes on Autonomic Function in Hypertension: The VAMPHYRE Study rate with amlodipine, that both drugs reduced BP and total vascular resistance equally in 23 hypertensive patients. 11 Possibly, reflex increase of sympathetic activity is counteracted by a direct inhibition of presynaptic release of NE by verapamil. 22,23 Alternatively, verapamil might directly decrease sympathetic activity or increase vagal activity by a central effect on the cardiovascular control center, inasmuch as verapamil can pass the blood brain barrier. 24 The changes in heart rate, LF/HF ratio, and plasma NE were negatively associated with their baseline levels for both amlodipine (Spearman s r= 0.34, P=0.004; r= 0.38, P=0.004; r= 0.23, P=0.07, respectively) and verapamil (r= 0.58, P=0.001; r= 0.25, P=0.10; r= 0.47, P=0.0006, respectively). As all these parameters tended to increase with amlodipine and to decrease with verapamil, one might suggest that subjects with a relatively low sympathetic activity before treatment show the largest increase of sympathetic activity with amlodipine. In contrast, subjects with a relatively high sympathetic activity before treatment might benefit most from treatment with verapamil in terms of a reduction of sympathetic activity, as assessed by heart rate and plasma NE. Effects on Short Term Autonomic Control of Blood Pressure. Although both drugs improved BRS, this effect was considerably greater with verapamil than with amlodipine. Increased autonomic control of heart rate, as a consequence of increased vagal tone during verapamil treatment, might explain this. Kailasam et al reported no significant differences among 2 weeks of placebo, 4 weeks of felodipine, and 4 weeks of verapamil on phenylephrine BRS, although BRS was higher during verapamil treatment (9.3 msec/mmhg) compared with felodi pine (5.9 msec/mmhg). 10 However, sample size was small (n=15). Second, the different BRS technique could explain the discrepancy with our results. Although the transfer function technique (present study) and the phenylephrine technique (Kailasam et al) are highly correlated, 5,15 the phenylephrine technique has been criticized because the drug influences the baroreflex itself, 25 whereas the transfer function technique analyzes the baroreflex response to spontaneous variations in BP. 15 Clinical Implications and Limitations of the Study. Our study demonstrates that the effects of different classes of calcium antagonists on sympathovagal balance and baroreflex function, as measured by HRV, 129

139 Chapter 7 BRS, and plasma NE, are markedly different in hypertension. However, a number of issues have to be considered before clinical implications can be drawn. First, there is debate about how to measure sympathovagal balance and whether noninvasive HRV methods are suitable. 26 However, an increase in LF/HF ratio observed during tilt 13,14 that can be a enuated by β blockade 13 suggests that an increase in LF/HF ratio reflects a shift toward sympathetic predominance. Furthermore, the contrasting effects of verapamil and amlodipine on LF/HF are endorsed by their similar effects on heart rate, phase between BP and heart rate variability, and plasma NE. Second, the present study is a short term study with a treatment duration of two times 8 weeks. Although this period is probably long enough to overcome short term effects and to have a stable autonomic balance, the long term effects have to be established in studies with longer duration. Still, the present study may have clinical implications. Because impaired baroreflex function and increased sympathovagal balance often accompany hypertension, recuperation of baroreflex function and autonomic balance may be valuable additional goals in the treatment of hypertension. Unfortunately, no trial has yet been designed to show that reduction of sympathovagal balance reduces cardiovascular mortality in hypertension. However, there is considerable circumstantial evidence. First, increased heart rate is a marker of increased risk for death from coronary heart disease, cardiovascular disease, and even for all causes of mortality in hypertension. 2 Although the predictive value of BRS and HRV has not yet been proved in hypertension, these parameters are markers of increased risk after myocardial infarction and in the general population. 27,28 Second, several pathophysiologic mechanisms may account for this increased risk. Chronically increased sympathovagal balance is not only associated with a higher BP and heart rate, 1 but also with end organ damage (left ventricular hypertrophy), 29 increased platelet coagulation, increased plasma cholesterol and triglycerides, increased plasma glucose, increased insulin, and increased hematocrit 30 and arrhythmias. 31 Third, a reduction in end organ damage (assessed by left ventricular hypertrophy), which was confined to hypertensive patients with a reduction in sympathetic tone, supports their close association

140 The Effects of Dihydropyridine and Phenylalkylamine Calcium Antagonist Classes on Autonomic Function in Hypertension: The VAMPHYRE Study Conclusion We conclude that 1) the long acting dihydropyridine amlodipine induces a shift in sympathovagal balance, as assessed by HRV and plasma NE, toward sympathetic predominance compared with vagal predominance with the non dihydropyridine verapamil; and 2) short term autonomic control of BP, as measured by BRS, is more effectively restored by verapamil than by amlodipine. These contrasting effects of calcium antagonists of different classes may have therapeutic implications beyond lowering BP. However, further long term studies are needed to determine whether reduction of sympathetic tone and restoration of baroreflex function can reduce the risk of complications and mortality in hypertension. 131

141 Chapter 7 References 1. Palatini P, Casiglia E, Paule o P, Staessen J, Kaciroti N, Julius S: Relationship of tachycardia with high blood pressure and metabolic abnormalities: a study with mixture analysis in three populations. Hypertension 1997;30: Gillman MW, Kannel WB, Belanger A, D Agostino RB: Influence of heart rate on mortality among persons with hypertension: the Framingham Study. Am Heart J 1993;125: Chakko S, Mulingtapang RF, Huikuri HV, Kessler KM, Materson BJ, Myerburg RJ: Alterations in heart rate variability and its circadian rhythm in hypertensive patients with left ventricular hypertrophy free of coronary artery disease. Am Heart J 1993;126: Grassi G, Ca aneo BM, Seravalle G, Lanfranchi A, Mancia G: Baroreflex control of sympathetic nerve activity in essential and secondary hypertension. Hypertension 1998;31: Watkins LL, Grossman P, Sherwood A: Noninvasive assessment of baroreflex control in borderline hypertension: Comparison with the phenylephrine method. Hypertension 1996;28: Julius S, Jamerson K: Sympathetics, insulin resistance and coronary risk in hypertension: the chicken and egg question. J Hypertens 1994;12: Grossman E, Messerli FH: Effect of calcium antagonists on plasma norepinephrine levels, heart rate, and blood pressure. Am J Cardiol 1997;80: de Champlain J, Karas M, Nguyen P, Cartier P, Wistaff R, Toal CB, Nadeau R, Larochelle P: Different effects of nifedipine and amlodipine on circulating catecholamine levels in essential hypertensive patients. J Hypertens 1998;16: Sakata K, Shirotani M, Yoshida H, Nawada R, Obayashi K, Togi K, Miho N: Effects of amlodipine and cilnidipine on cardiac sympathetic nervous system and neurohormonal status in essential hypertension. Hypertension 1999;33: Kailasam MT, Parmer RJ, Cervenka JH, Wu RA, Ziegler MG, Kennedy BP, Adegbile IA, O Connor DT: Divergent effects of dihydropyridine and phenylalkylamine calcium channel antagonist classes on autonomic function in human hypertension. Hypertension 1995;26:

142 The Effects of Dihydropyridine and Phenylalkylamine Calcium Antagonist Classes on Autonomic Function in Hypertension: The VAMPHYRE Study 11. Nazzaro P, Manzari M, Merlo M, Triggiani R, Scarano AM, Lasciarrea A, Pirrelli A: Antihypertensive treatment with verapamil and amlodipine: Their effect on the functional autonomic and cardiovascular stress responses. Eur Heart J 1995;16: Psaty BM, Heckbert SR, Koepsell TD, Siscovick DS, Raghunathan TE, Weiss NS, Rosendaal FR, Lemaitre RN, Smith NL, Wahl PW: The risk of myocardial infarction associated with antihypertensive drug therapies. JAMA 1995;274: Pagani M, Lombardi F, Guzze i S, Rimoldi O, Furlan R, Pizzinelli P, Sandrone G, Malfa o G, Dell Orto S, Piccaluga E: Power spectral analysis of heart rate and arterial pressure variabilities as a marker of sympatho vagal interaction in man and conscious dog. Circ Res 1986;59: Montano N, Ruscone TG, Porta A, Lombardi F, Pagani M, Malliani A: Power spectrum analysis of heart rate variability to assess the changes in sympathovagal balance during graded orthostatic tilt. Circulation 1994;90: Robbe HW, Mulder LJ, Ruddel H, Langewitz WA, Veldman JB, Mulder G: Assessment of baroreceptor reflex sensitivity by means of spectral analysis. Hypertension 1987;10: Tuininga YS, Crijns HJ, Brouwer J, van den Berg MP, Man i, V, Mulder G, Lie KI: Evaluation of importance of central effects of atenolol and metoprolol measured by heart rate variability during mental performance tasks, physical exercise, and daily life in stable postinfarct patients. Circulation 1995;92: Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology: Heart rate variability: standards of measurement, physiological interpretation and clinical use. Circulation 1996;93: Appel ML, Berger RD, Saul JP, Smith JM, Cohen RJ: Beat to beat variability in cardiovascular variables: noise or music? J Am Coll Cardiol 1989;14: Saul JP, Berger RD, Albrecht P, Stein SP, Chen MH, Cohen RJ: Transfer function analysis of the circulation: unique insights into cardiovascular regulation. Am J Physiol 1991;261:H1231 H

143 Chapter Hamada T, Watanabe M, Kaneda T, Ohtahara A, Kinugawa T, Hisatome I, Fujimoto Y, Yoshida A, Shigemasa C: Evaluation of changes in sympathetic nerve activity and heart rate in essential hypertensive patients induced by amlodipine and nifedipine. J Hypertens 1998;16: Schmieder RE, Messerli FH, Garavaglia GE, Nunez BD: Cardiovascular effects of verapamil in patients with essential hypertension. Circulation 1987;75: Meyer EC, Sommers DK, Avenant JC: The effect of verapamil on cardiac sympathetic function. Eur J Clin Pharmacol 1991;41: Terland O, Gronberg M, Flatmark T: The effect of calcium channel blockers on the H() ATPase and bioenergetics of catecholamine storage vesicles. Eur J Pharmacol 1991;207: Doran AR, Narang PK, Meigs CY, Wolkowitz OM, Roy A, Breier A, Pickar D: Verapamil concentrations in cerebrospinal fluid after oral administration (le er). N Engl J Med 1985;312: Chowdhary S, Vaile JC, Fletcher J, Ross HF, Coote JH, Townend JN: Nitric oxide and cardiac autonomic control in humans. Hypertension 2000;36: Eckberg DL: Sympathovagal balance: a critical appraisal. Circulation 1997;96: La Rovere MT, Bigger JT J, Marcus FI, Mortara A, Schwartz PJ: Baroreflex sensitivity and heart rate variability in prediction of total cardiac mortality after myocardial infarction: ATRAMI. Lancet 1998;351: Dekker JM, Crow RS, Folsom AR, Hannan PJ, Liao D, Swenne CA, Schouten EG: Low heart rate variability in a 2 minute rhythm strip predicts risk of coronary heart disease and mortality from several causes: the ARIC Study. Atherosclerosis Risk In Communities. Circulation 2000;102: Marcus R, Krause L, Weder AB, Dominguez MA, Schork NJ, Julius S: Sex specific determinants of increased left ventricular mass in the Tecumseh Blood Pressure Study. Circulation 1994;90: Palatini P, Julius S: Association of tachycardia with morbidity and mortality: pathophysiological considerations. J Hum Hypertens 1997;11(suppl 1):S19 S

144 The Effects of Dihydropyridine and Phenylalkylamine Calcium Antagonist Classes on Autonomic Function in Hypertension: The VAMPHYRE Study 31. Schwartz PJ, La Rovere MT, Vanoli E: Autonomic nervous system and sudden cardiac death: Experimental basis and clinical observations for post myocardial infarction risk stratification. Circula tion 1992;85:I77 I Muiesan ML, Rizzoni D, Zulli R, Castellano M, Be oni G, Porteri E, Agabiti Rosei E: Power spectral analysis of the heart rate in hypertensive patients with and without left ventricular hypertrophy: the effect of a left ventricular mass reduction. J Hypertens 1998;16:

145 Chapter 7 136

146 Chapter 8 Contrasting effects of verapamil and amlodipine on cardiovascular stress responses in hypertension Britisch Journal of Clinical Pharmacology Dec;52: J.D. Lefrandt 1, J. Heitmann 2, K. Sevre 3, M. Castellano 4, M. Hausberg 5, M. Fallon 6, A. Urbigkeit 2, M. Rostrup 3, E. Agabiti Rosei 4, K.H. Rahn 5, M. Murphy 6, F. Zannad 7, P.J. de Kam 1 & A.J. Smit 1 1 University Hospital of Gronigen, Groningen, The Netherlands, 2 Klinikum der Philipps Universität, Marburg, Germany, 3 Ullevål Hospital, Oslo, Norway, 4 Università Degli Studi di Brescia, Brescia, Italy, 5 Klinik der Westfälischen Wilhelms Universität, Münster, Germany, 6 Cork University Hospital, Cork, Ireland and 7 Centre d Investigation Clinique Hôpital Jeanne d Arc, Nancy, France.

147 Chapter 8 Summary Aims To compare the effects of two long acting calcium antagonists of different types on cardiovascular stress responses in hypertension. Methods One hundred and forty five patients with mild to moderate hypertension and a mean (± s.e.mean) age of 51 ± 0.9 years received for 8 weeks the phenylalkylamine verapamil sustained release (240 mg) and the dihydropyridine amlodipine (5 mg) in a double blind cross over design, both after 4 weeks of placebo. Blood pressure, heart rate and plasma noradrenaline were monitored during 3 min of sustained isometric handgrip and 2 min of cold pressor. Results Blood pressure was equally reduced by both drugs. After 3 min handgrip, systolic blood pressure, heart rate and rate pressure product were lower with verapamil compared with amlodipine. Verapamil a enuated the increases in systolic blood pressure (25 ± 2 vs 30 ± 2 mmhg, difference 4.6, 95% CI (1.0, 8.1), P<0.01) and rate pressure product (3.1 ± 0.2 vs 3.6 ± mmhg beats min 1, difference 0.5, 95% CI (0.1, 0.9), P<0.01) during handgrip compared with amlodipine. Similar results were observed during cold pressor. Plasma noradrenaline levels were lower with verapamil compared with amlodipine at rest and after both tests, but the increases in plasma noradrenaline were not significantly different. Conclusions Verapamil is more effective in reducing blood pressure and rate pressure product responses to stress compared with amlodipine. Although plasma noradrenaline is lower with verapamil at rest and after stress, the increase during stress is not different. 138

148 Contrasting effects of verapamil and amlodipine on cardiovascular stress responses in hypertension Introduction Increased cardiac output and heart rate at rest and adrenergic hyperreactivity to stress in patients with borderline hypertension 1 3, as well as normotensive offspring of hypertensive patients 4, have characterized the pathophysiological haemodynamic changes in the development of hypertension. In established hypertension, increased peripheral resistance rather than increased cardiac output is characteristic 5. The increased pressure workload induces adaptive changes in the endothelium, the vascular smooth muscle and the extracellular matrix of vessels and the heart 6,7. Development of left ventricular hypertrophy adds to the increased cardiovascular risk in hypertensive patients 8. Furthermore, increased blood pressure and heart rate responsiveness during stress could trigger myocardial ischaemia and infarction due to inability to supply oxygen during this high demand at increased workload 9,10. The incidence of ischaemic events follows a pa ern similar to rises in heart rate and blood pressure 11,12. Thus, drugs that would reduce cardiac load both at rest and during exertion could be beneficial in hypertension. Calcium antagonists are widely used to treat hypertension. Although the blood pressure reducing efficacy of different types of calcium antagonists is comparable, their mechanisms of action are not the same. While dihydropyridines induce peripheral vasodilatation and have li le or no effect on myocardial contractility and sinus node automaticity, nondihydropyridines have less influence on peripheral vascular smooth musculature and direct effects on cardiac inotropy and chronotropy 13,14. The aim of the present study was to compare the effects of a long acting dihydropyridine (amlodipine) and a phenylalkylamine (verapamil) on cardiovascular stress responses by noninvasive methods during isometric handgrip and cold pressor stress tests in patients with mild to moderate hypertension. This paper was presented as an abstract at the Eighteenth Scientific Meeting of the International Society of Hypertension in Chicago, August

149 Chapter 8 Methods Study design and subjects. The study was a double blind randomized cross over comparison of verapamil Sustained Release (SR, 240 mg) and amlodipine (5 mg), corresponding to the registered and recommended starting dosage of these drugs in The Netherlands. All patients were at least 18 years of age and had mild to moderate hypertension, defined as a diastolic blood pressure between 95 and 110 mmhg on at least three occasions. In these patients, either hypertension was newly diagnosed, or current antihypertensive treatment did not meet therapeutic goals. Patients were excluded if secondary hypertension was suspected or if they had had a recent cardiovascular event. Seven centres in six European Community countries participated. Informed consent was obtained from all patients. The local ethics commi ees at each participating hospital approved the protocol. At the first patient visit, any antihypertensive treatment was stopped and 4 weeks of placebo treatment was started. Si ing blood pressure was measured three times after 10 min of rest with 2 min intervals. Randomization to sequence 1 (8 weeks of verapamil 4 weeks of placebo 8 weeks of amlodipine) or 2 (8 weeks of amlodipine 4 weeks of placebo 8 weeks of verapamil) followed at the second visit if diastolic blood pressure was 95 and 110 mmhg and systolic blood pressure 180 mmhg. The first placebo period was open, both active drug periods were double blind and the second placebo period was single blind. All tablets were taken in the evening. Study procedures. All tests were performed in the morning in a warm, quiet room. Patients refrained from eating, smoking and drinking alcohol, coffee and tea that day. A Finapres (Ohmeda 2300, Englewood, Colo., USA) blood pressure monitor was a ached to the third finger of the patient s right hand. Ten minutes after insertion of a catheter in an antecubital vein of the left arm, heart rate and si ing sphygmomanometric blood pressure were measured and the Finapres blood pressure monitoring was started. After another 10 min of supine rest, the isometric handgrip test was performed; the patient had to squeeze a dynamometer at 30% of the predetermined maximum strength, during 3 min, with his left hand. After a recovery period of 10 min, the cold pressor test was performed; the left hand was immersed in 140

150 Contrasting effects of verapamil and amlodipine on cardiovascular stress responses in hypertension ice water (0 4 C) for 2 min. Then, a second recovery period of 10 min followed. Since the autonomic nervous system is very much influenced by mental stress, the sole purpose of the first visit was to get the patient acquainted with the study procedures. Venous blood (6 ml) was sampled in a prechilled EGTA glutathion tube 1 min before and 1 min after both handgrip and cold pressor. Plasma noradrenaline concentrations were determined by electrochemical detection after high pressure liquid chromatography in a central laboratory (Analytico, Breda, the Netherlands) for all centres. Beat to beat systolic (SBP) and diastolic blood pressure (DBP) and heart rate (HR) were obtained from the RS232 interface of the Finapres and stored on a personal computer. Rate pressure product was calculated by multiplying HR by SBP. Mean SBP, DBP, HR and rate pressure product were determined for consecutive periods of 20 s. The changes in SBP, DBP, HR and rate pressure product were calculated for every minute. Statistical analysis. The primary endpoint of this analysis was defined as the difference in increase of the rate pressure product with verapamil and amlodipine during the stress tests. Sample size calculation was based on the rate pressure product responses to different types of stress of a recent trial comparing verapamil and amlodipine in hypertensive subjects 15. Anticipating a standard deviation of 1100, 134 evaluable patients were needed to detect a difference of 500 mmhg beats min 1 in a 2 sided t test for α = 0.05 with 95% power. A mixed ancova model was defined to test the null hypothesis of no difference between both treatments. Period, treatment and centre were defined as fixed factors, the placebo period preceding the active drug as covariate and patient within centre as random effect. Bonferroni s correction for multiple comparisons was used. A P value < 0.05 in the 2 sided test was considered statistically significant. A Wilcoxon test was used for all parameters that had a skewed distribution and could not be normalized by log transformation. Student s t test was performed for all parameters, to check for any differences between the first and second placebo period. A secondary model tested the possibility of a carry over effect. All data are shown as mean±s.e.mean. Differences between both drugs are shown as mean with 95% CI or natural log 141

151 Chapter 8 (LN) mean with 95% CI for parameters that were log transformed before analysis. Table 1. Patient characteristics at randomization. Dyslipidaemia is as defined as total cholesterol >6.5 mmol l 1 or low density lipoprotein >5 mmol l 1 or high density lipoprotein <9 mmol l 1 triglycerides >2.5 mmol l 1. Mean (range) or Number (%) Male/female n (%) 95/50 (65%/35%) Age (years) 51 (20 72) Systolic blood pressure (mmhg) 153 ( ) Diastolic blood pressure (mmhg) 100 (95 110) Heart rate (beats min 1 ) 67.8 (51 102) Body mass index (kg/m 2 ) 28.6 ( ) Smokers current n (%) 33 (23%) Dyslipidaemia n (%) 57 (40%) Family history of premature 13 (9%) atherosclerotic disease n (%) Serum creatinine (µmol l 1 ) 85 (59 141) Results One hundred 45 patients were randomized at the second visit. Patient characteristics at randomization are summarized in Table 1. Twenty two patients later dropped out of the study due to adverse events, lack of compliance and withdrawal of their consent. The data were analysed according to the intention to treat (ITT) principle: all patients that had used at least one tablet of active treatment entered the ITT population. The design of the study does not allow an unbiased assessment of the relation between adverse events and study drugs. The adverse events with a probable relation with the drugs (verapamil/amlodipine) were most frequently common side effects: headache (2/1), dizziness (3/1), palpitations (0/1), constipation (1/1) and peripheral oedema (2/1). In addition, three generalized allergic responses to verapamil that faded after withdrawal were observed. The mean compliance, assessed by tablet count, was > 95% for all treatments. Because no significant difference between both placebo periods existed for any parameter, the mean of both is presented. Sphygmomanometric blood pressure and heart rate. During placebo, mean systolic and diastolic blood pressure were 153 ± 1 and 100 ± 1 mmhg, respectively. Both were equal after verapamil and amlodipine treatment: SBP 139 ± 2 vs 138 ± 2 mmhg, difference 0.86, 95% CI ( 1.65, 3.38), P=0.50 and DBP 91 ± 1 vs 91 ± 1 mmhg, difference 0.43, 95% CI 142

152 Delta systolic blood pressure (mmhg) Delta rate pressure product ( mmhg*beats i n Blood pressure (mmhg) Rate pressure product ( mmhg*beats i n Delta systolic blood pressure (mmhg) Delta rate pressure product (mmhg * beats min -1 ) Blood pressure (mmhg) Rate pressure product (mmhg * beats min -1 ) Contrasting effects of verapamil and amlodipine on cardiovascular stress responses in hypertension SBP Heartrate (beats min HG HG DBP HG Time (min) Time (min) Time (min) Figure 1. Mean ± s.e. mean for blood pressure, heart rate and rate pressure product during and after 3 min of handgrip (HG), for segments of 20 s. For the delta systolic blood pressure and delta rate pressure product the change from before the test are shown for each minute. *P<0.01, **P< , for verapamil vs amlodipine. *, placebo;, verapimil;, amlodipine SBP Heartrate (beats min CP ) m CP CP 0 1 DBP Time (min) Time (min) ) m Time (min) Figure 2. Mean ± s.e. mean for blood pressure, heart rate and rate pressure product during and after 2 min of cold pressor (CP), for segments of 20 s. For the delta systolic blood pressure and delta rate pressure product the change from before the test are shown for each minute. *P<0.01, **P< , for verapamil vs amlodipine. *, placebo;, verapimil;, amlodipine. ( 1.36, 2.22), P=0.59. During placebo, mean heart rate was 68 ± 1 beats min 1. Heart rate was lower with verapamil than with amlodipine: 65 ± 1 vs 69 ± 1 beats min 1, difference 4.0, 95% CI ( 6.6, 2.7), P< Haemodynamic responses to handgrip and cold pressor. SBP, DBP, HR and rate pressure product responses to handgrip and cold pressure are depicted in Figures 1 and 2. Before both tests, SBP and DBP were not different with verapamil and amlodipine; HR and rate pressure 143

153 Chapter 8 product were significantly lower with verapamil compared to amlodipine (P< and P=0.002, respectively). Handgrip. During 3 min handgrip, SBP, DBP, HR and rate pressure product increased significantly with both drugs (P< for all). SBP was significantly lower with verapamil compared with amlodipine during the last 40 s of the test (after 3 min handgrip: 169 ± 3 vs 177 ± 4 mmhg, natural log difference 0.048, 95% CI (0.013, 0.083), P<0.01). The difference in DBP did not reach statistical significance (after 3 min handgrip: 100 ± 2 vs 103 ± 3 mmhg, natural log difference 0.037, 95% CI ( 0.006, 0.079), P=0.09). HR and rate pressure product remained significantly lower during and after the test with verapamil compared with amlodipine (after 3 min handgrip: HR 71 ± 1 vs 75 ± 1 beats min 1, natural log difference 0.057, 95% CI (0.029, 0.085), P<0.0001; rate pressure product 11.8 ± 0.3 vs 13.2 ± mmhg beats min 1, natural log difference 0.107, 95% CI (0.060, 0.153), P<0.0001). The increase in SBP was significantly smaller during the third minute of handgrip and the increase in rate pressure product was significantly smaller during all 3 min of handgrip with verapamil compared with amlodipine (third minute of handgrip: increase in SBP 25 ± 2 vs 30 ± 2 mmhg, difference 4.6, 95% CI (1.0, 8.1), P<0.01; increase in rate pressure product 3.1 ± 0.2 vs 3.6 ± mmhg beats min 1, difference 0.5, 95% CI (0.1, 0.9), P<0.01). Cold pressor. During 2 min cold pressor, SBP, DBP, HR and rate pressure product increased significantly with both drugs (P< for all). With verapamil, SBP was significantly lower from 40 s after the start of cold pressor until 3 min and 20 s after the test, compared to amlodipine (after 2 min cold pressor: 174 ± 4 vs 184 ± 4 mmhg, natural log difference 0.059, 95% CI (0.020, 0.099), P<0.01). DBP was significantly higher with amlodipine compared with verapamil during the recovery period, during the third minute after the test. HR and rate pressure product remained significantly lower during and after the test with verapamil compared with amlodipine (after 2 min cold pressor: HR 67 ± 1 vs 72 ± 1 beats min 1, natural log difference 0.069, 95% CI (0.043, 0.095), P<0.0001; rate pressure product 11.6 ± 0.3 vs 13.2 ± mmhg beats min 1, natural log difference 0.13, 95% CI (0.08, 0.19), P<0.0001). The increase in rate pressure product was significantly smaller with verapamil compared with amlodipine during the second minute of cold pressor (after 2 min cold pressor: 3.2 ± 0.2 vs 3.9 ± mmhg beats min 1, difference 0.7, 95% CI (0.2, 1.1), P<0.01). 144

154 Contrasting effects of verapamil and amlodipine on cardiovascular stress responses in hypertension Plasma noradrenaline before and after handgrip and cold pressor (Table 2). Before and after the handgrip as well as before and after the cold pressor test, plasma noradrenaline concentrations were higher with amlodipine compared with verapamil (P<0.001, P<0.001, P<0.001 and P=0.01). The increases in plasma noradrenaline concentration after both tests were not different between the drugs. Table 2. Mean ± s.e. mean for plasma noradrenaline before and after isometric handgrip and cold pressor. Median (Q1, Q3) for changes in plasma noradrenaline. Mean (95% CI) for the difference between verapamil and amlodipine, except for the difference in changes (*median, Q1, Q3). Noradrenaline(nmol l 1 ) Placebo Verapamil Amlodipine Difference P value Handgrip 1 min before test 1.30 ± ± ± ( 0.21, 0.06) < min after test 1.38 ± ± ± ( 0.28, 0.09) <0.001 Change ( 0.13, 0.26) ( 0.11, 0.34) ( 0.20, 0.27) 0.07 ( 0.37, 0.41)* 0.94 Cold pressor 1 min before test 1.35 ± ± ± ( 0.23, 0.10) < min after test 1.81 ± ± ± ( 0.51, 0.07) 0.01 Change ( 0.19, 0.72) ( 0.11, 0.34) ( 0.28, 0.94) 0.13 ( 0.42, 0.45)* 0.34 Discussion This study demonstrates that the effects of the long acting dihydropyridine amlodipine and the phenylalkylamine verapamil on cardiovascular stress responses to exertion are markedly different, whereas the reduction in resting blood pressure by the two drugs is similar. Calcium antagonists reduce blood pressure mainly by decreasing peripheral resistance. We found a similar reduction in blood pressure by verapamil and amlodipine at rest. This is in agreement with a study of Nazarro et al. 15 in 23 hypertensive patients that reported an equal reduction of blood pressure with verapamil and amlodipine after 4 weeks treatment. Furthermore, they found both drugs equally reduced peripheral resistance. The blood pressure reducing efficacy at rest therefore seems to be the same for both drugs. However, the effects on resting rate pressure product and heart rate are different. We found that verapamil reduced resting rate pressure product while amlodipine did not. This resulted from reduction in resting heart rate by verapamil, which directly effects sinus node automaticity 16. During stress, the effects of verapamil and amlodipine on haemodynamics showed even more contrasts. Blood pressure lowering efficacy during handgrip in hypertensive patients has been reported with both verapamil 17 and amlodipine 18. However, the present study also shows diminished systolic blood pressure and rate pressure product 145

155 Chapter 8 responsiveness during handgrip with verapamil compared with amlodipine. Diminished blood pressure responsiveness during handgrip has been reported with verapamil only 17. Isometric handgrip is a potent α adrenergic stimulus. Verapamil has been shown to bind to the α adrenergic receptor 19. Furthermore, verapamil might directly inhibit presynaptic release of noradrenaline 20,21. However, although plasma noradrenaline concentrations before and after isometric handgrip were lower with verapamil compared with amlodipine, the increase in plasma noradrenaline concentration during handgrip was not different in our study. The cold pressor test is less frequently used to study the effects of calcium antagonists on cardiovascular stress responses in hypertension. Both handgrip and cold pressor test increase sympathetic outflow and peripheral resistance that results in an increase in blood pressure. The blood pressure and heart rate responses to handgrip showed a linear increase through time in our study (Figure 1). In contrast, the cold pressor test induced changes which exhibited a hyperbolic time course. However, the effects of verapamil and amlodipine showed interesting contrasts. Verapamil blunted the systolic blood pressure, heart rate and rate pressure product response during cold pressor compared to amlodipine. The change in rate pressure product was lower with verapamil although the difference in systolic blood pressure response did not reach statistical significance. In an earlier study of 13 hypertensive patients, verapamil reduced systolic blood pressure but not the increase in systolic blood pressure during cold pressor compared with placebo 22. Although the present study showed a clear reduction in blood pressure and rate pressure product responses to stress and a reduction of plasma noradrenaline at rest and after stress with verapamil compared with amlodipine, the potential advantage of these observations in the treatment of hypertension remains to be confirmed. Still, the present study may have clinical implications. Firstly, the higher efficacy of verapamil in reducing blood pressure during exertion suggests that its antihypertensive effect is more preserved during normal daily activities that often contain isometric handgrip. Secondly, the reduction in rate pressure product at rest and during exertion and the smaller increase in rate pressure product during exertion with verapamil should decrease myocardial oxygen consumption 23 at rest and during stress. 146

156 Contrasting effects of verapamil and amlodipine on cardiovascular stress responses in hypertension This cardioprotective effect of verapamil could be especially useful in patients with associated coronary artery disease, as supported by a study of 551 patients with chronic stable angina, in whom verapamil reduced the total duration of ischaemic episodes, in contrast to amlodipine which increased the total duration of ischaemic episodes 24. The DAVIT II study showed a reduction in major cardiovascular events in post myocardial patients without heart failure by verapamil compared with placebo 25. However, the use of a rate limiting drug as primary prevention of ischaemic events remains speculative. To determine whether the favourable effects of verapamil on blood pressure reactivity and heart rate would translate into improved prognosis, a large outcome study in hypertensive patients is needed. We conclude that verapamil is more effective in reducing blood pressure and rate pressure product responses to stress compared with amlodipine, while resting blood pressure is equally reduced by both. Although plasma noradrenaline is lower with verapamil at rest and after stress, the increase during stress is not different compared with amlodipine. Acknowledgments This study was financially supported by KNOLL AG, Germany. 147

157 Chapter 8 References 1. Julius S, Krause L, Schork NJ, et al. Hyperkinetic borderline hypertension in Tecumseh, Michigan. J Hypertens 1991; 9: Marraccini P, Palombo C, Giaconi S, et al. Reduced cardiovascular efficiency and increased reactivity during exercise in borderline and established hypertension. Am J Hypertens 1989; 2: Borghi C, Costa FV, Boschi S, Mussi A, Ambrosioni E. Predictors of stable hypertension in young borderline subjects: a five year follow up study. J Cardiovasc Pharmacol 1986; 8(Suppl 5): S138 S Radice M, Alli C, Avanzini F, et al. Role of blood pressure response to provocative tests in the prediction of hypertension in adolescents. Eur Heart J 1985; 6: Julius S. Transition from high cardiac output to elevated vascular resistance in hypertension. Am Heart J 1988; 116: Frohlich ED, Apstein C, Chobanian AV, et al. The heart in hypertension. N Engl J Med 1992; 327: Schwartzkopff B, Motz W, Frenzel H, Vogt M, Knauer S, Strauer BE. Structural and functional alterations of the intramyocardial coronary arterioles in patients with arterial hypertension. Circulation 1993; 88: Levy D, Garrison RJ, Savage DD, Kannel WB, Castelli WP. Prognostic implications of echocardiographically determined left ventricular mass in the Framingham Heart Study. N Engl J Med 1990; 322: Nitenberg A, Antony I, Aptecar E, Arnoult F, Lerebours G. Impairment of flow dependent coronary dilation in hypertensive patients. Demonstration by cold pressor test induced flow velocity increase. Am J Hypertens 1995; 8: 13S 18S. 10. Antony I, Aptecar E, Lerebours G, Nitenberg A. Coronary artery constriction caused by the cold pressor test in human hypertension. Hypertension 1994; 24: Deedwania PC & Nelson JR. Pathophysiology of silent myocardial ischemia during daily life. Hemodynamic evaluation by simultaneous electrocardiographic and blood pressure monitoring. Circulation 1990; 82:

158 Contrasting effects of verapamil and amlodipine on cardiovascular stress responses in hypertension 12. Muller JE, Abela GS, Nesto RW, Tofler GH. Triggers, acute risk factors and vulnerable plaques: the lexicon of a new frontier. J Am Coll Cardiol 1994; 23: Wood AJ. Calcium antagonists. Pharmacologic differences and similarities. Circulation 1989; 80: IV184 IV Freedman DD & Waters DD. Second generation dihydropyridine calcium antagonists. Greater vascular selectivity and some unique applications. Drugs 1987; 34: Nazzaro P, Manzari M, Merlo M, et al. Antihypertensive treatment with verapamil and amlodipine. Their effect on the functional autonomic and cardiovascular stress responses. Eur Heart J 1995; 16: Alvarez L, Escudero C, Torralba A, Millan I. Electrophysiologic assessment of calcium channel blockers in transplanted hearts: an experimental study. J Electrocardiol 1998; 31: Cleroux J, Beaulieu M, Kouame N, Lacourcierce Y. Comparative effects of quinapril, atenolol, and verapamil on blood pressure and forearm hemodynamics during handgrip exercise. Am J Hypertens 1994; 7: Lopez LM, Thorman AD, Mehta JL. Effects of amlodipine on blood pressure, heart rate, catecholamines, lipids and responses to adrenergic stimulus. Am J Cardiol 1990; 66: van Zwieten PA, van Meel JC, Timmermans PB. Pharmacology of calcium entry blockers: interaction with vascular alpha adrenoceptors. Hypertension 1983; 5: II8 II Meyer EC, Sommers DK, Avenant JC. The effect of verapamil on cardiac sympathetic function. Eur J Clin Pharmacol 1991; 41: Terland O, Gronberg M, Flatmark T. The effect of calcium channel blockers on the H (+) ATPase and bioenergetics of catecholamine storage vesicles. Eur J Pharmacol 1991; 207: Gebara OC, Jimenez AH, McKenna C, et al. Stress induced hemodynamic and hemostatic changes in patients with systemic hypertension: effect of verapamil. Clin Cardiol 1996; 19: Hoeft A, Sonntag H, Stephan H, Ke ler D. Validation of myocardial oxygen demand indices in patients awake and during anesthesia. Anesthesiology 1991; 75:

159 Chapter Frishman WH, Glasser S, Stone P, Deedwania PC, Johnson M, Fakouhi TD. Comparison of controlled onset, extended release verapamil with amlodipine and amlodipine plus atenolol on exercise performance and ambulatory ischemia in patients with chronic stable angina pectoris. Am J Cardiol 1999; 83: Hansen JF, for the Danish study group on verapamil in myocardial infarction. Effect of verapamil on mortality and major events after acute myocardial infarction. DAVIT II. Am J Cardiol 1990; 66:

160 Chapter 9 Less Adrenergic Response to Mental Task During Verapamil Compared to Amlodipine Treatment in Hypertensive Subjects Blood Pressure Oct;10:111 5 K. Sevre 1, J.D. Lefrandt 2, I. Eide 3, A.J. Smit 2, M. Rostrup 3,4 1 Department of Cardiology, Ullevål University Hospital, Oslo, Norway, 2 Department of Internal Medicine, University Hospital of Groningen, The Netherlands, 3 Department of Internal Medicine and 4 Research Forum, Ullevål University Hospital, Oslo, Norway

161 Chapter 9 Summary We compared the effects of amlodipine and verapamil slow release on autonomic responses to a 5 min mental arithmetic test (MST) in patients with mild to moderate hypertension. Twenty subjects received 8 weeks of verapamil slow release 240 mg or amlodipine 10 mg in a double blind crossover design, both after 4 weeks placebo. Heart rate (HR) and blood pressure (BP) were continuously monitored. Venous plasma catecholamines were analysed by a radioenzymatic assay. Baroreflex sensitivity (BRS) was estimated with the transfer function technique. Calculations of the area under the curve (AUC) were used to estimate average HR, BP and catecholamine concentrations. The reactivity to MST was estimated as percent change from the basal AUC. A paired t test was performed. Data are means ± SEM. Compared to verapamil, amlodipine increased average noradrenaline (NA) concentrations (245 ± 23 vs 191 ± 17 pg/l, respectively, p=0.005), NA reactivity (14.0 ± 5.5% vs 2.9 ± 3.3, p=0.004), average HR (65 ± 2 vs 61 ± 2 beats/min, p<0.001) and HR reactivity (2.5 ± 1.0 vs 0.1 ± 0.9%, p=0.056). BP did not differ significantly. BRS correlated with average and baseline HR on both medications (r= 0.53 and 0.63, p 0.03). We conclude that adrenergic responses to MST are blunted on treatment with verapamil compared to amlodipine in hypertensive patients. Introduction Surges of adrenergic activity as seen during mental stress trigger cardiac events and ventricular arrhythmias 1 3. A major goal in treatment of patients with coronary artery disease (CAD) is to prevent such events 4. Calcium channel blockers are used in treatment of CAD, but preferably in those who are not tolerating adrenergic betablockers 4,5. Both drugs are widely used in treatment of hypertension 5. Calcium channel blockers are a heterogeneous group with diverging effects on the autonomic nervous system. Grossman et al. 6 meta analysed studies which involved 1252 hypertensive patients treated with various calcium antagonists. After more than 1 week s medication, heart rate (HR) was unchanged and noradrenaline (NA) increased (15%) on treatment with long acting dihydropyridines (e.g. amlodipine, 152

162 Less Adrenergic Response to Mental Task During Verapamil Compared to Amlodipine Treatment in Hypertensive Subjects nisoldipine), while the nondihydropyridines (e.g. verapamil slow release (SR), dilitazem) reduced HR (7%) and NA (21%). Gebara et al. 7 showed lower overall systolic blood pressure (SBP) and blood platelet aggregability during mental stress (MST) and cold pressor test (CPT) on verapamil SR than placebo medication in 13 moderately hypertensive patients. Ludwig et al. 8 demonstrated higher concentrations of NA, but no HR differences and lower blood pressure (BP) during MST on treatment with nisoldipine compared to placebo in moderately hypertensive subjects. Nazzaro et al. 9 demonstrated reduced overall HR but similar BP response to the total load of a battery of stress tests (two different MSTs, CPT and isometric handgrip) on verapamil SR treatment compared to amlodipine in 23 moderately hypertensive patients. However, these tests trigger different stress responses Thus, by not differentiating between the different tests, important information might have been lost. To our knowledge, nobody has addressed changes in cardiovascular and plasma catecholamine reactivity during specific mental stress on dihydropyridine compared to verapamil treatment. Thus, in the present study we wanted to compare the effects of amlodipine and verapamil SR on autonomic responses to MST in patients with mild to moderate hypertension. Both drugs are widely used in treatment of CAD and hypertension. MST was preferred, as this test evokes the fight and flight response 10,13,14, which closely resembles an everyday situation that may trigger cardiac events. Furthermore, as baroreceptor reflex sensitivity (BRS) is reduced in hypertensive patients 15, and decreases during mental stress 16 we wanted to relate BRS to the autonomic responses during MST. Materials and methods The present study was a substudy to the Vamphyre study, The effects on autonomic function of Verapamil versus Amlodipine in Patients with mild to moderate Hypertension during Rest and Exercise which was carried out in six European countries. Patients entry were eligible if they were >18 years old, and suffered from mild to moderate hypertension [140 systolic blood pressure (SBP) 180, diastolic blood pressure (DBP) 90 mmhg]. Patients were excluded if secondary hypertension 153

163 Chapter 9 was suspected, if they had recently suffered a cardiovascular event, if organ failure was present or they had diabetes mellitus, autoimmune disease or Parkinson s Disease. At the first visit, hypertensive treatment, if any, was stopped and 4 weeks placebo treatment was started. Randomization to sequence 1 (8 weeks verapamil, 4 weeks placebo, 8 weeks amlodipine) or 2 (8 weeks amlodipine, 4 weeks placebo, 8 weeks verapamil) followed at the second visit if DBP was 95 and µ110 mmhg and SBP µ180 mmhg. The patients were scheduled for five visits. Visit 1 took place the first day of the study, the remaining visits were performed at the last day of each medication period. The first placebo period was open, the second placebo period was single blind and both active drug periods were double blind. The purpose of visit one was to make the patient acquainted to the study procedures. The patients who terminated antihypertensive medication at visit 1 were scheduled for a visit to check BP 1 week later. Patients were recruited from the outpatient clinic for hypertensive patients, Ullevål University Hospital, Oslo, Norway. The Regional Board of Research Ethics approved the study, and informed wri en consent was obtained from each participant. Subjects. Twenty three patients entered the study between March 1997 and February Two patients did not want to perform the MST. One patient was withdrawn from the study due to an eye vein thrombosis after the first medication period. The remaining 20 patients stayed throughout the study, and consisted of 10 male and 10 female participants. Before visit 1, 11 patients did not use any vasoactive drug, six used a calcium antagonist, two used an ACE inhibitor and one used a beta blocker. Three participants were smokers. Mean age was 55 years (range years). Mean body mass index was 26.2 ± 0.6 kg/m2. Mean si ing BP and HR were 151 ± 3/99 ± 1 mmhg and 71 ± 2 beats/min, respectively, at the second visit. Data are means ± SEM. The medication compliance, assessed by tablet count, was be er than 80% for all drugs. Study procedure. All examinations were performed in the morning in a quiet room with temperature of C, after an overnight fast and a refrain from alcohol and tobacco the last 24 h. HR and si ing sphygomomanometric SBP and DBP (Korotkoff sounds I and V) were 154

164 Less Adrenergic Response to Mental Task During Verapamil Compared to Amlodipine Treatment in Hypertensive Subjects measured three times after a 10 min rest with a 2 min interval. The final BP was averaged from these three measurements. Beat to beat BP and HR were recorded in the supine position with a Finapres (Finapres, Ohmeda 2300, Englewood, Colorado, USA) non invasive blood pressure monitor with the appropriate cuff applied to the third finger of the left hand. This instrument has been validated, and the accuracy and precision found sufficient for tracking of changes in BP and HR 17. Mental stress test. The subjects were asked to subtract a two digit number, starting with a four digit number for 5 min as fast as possible. All wrong calculations were corrected. A metronome making noise at a frequency of 2 Hz was used to distract the subjects. The numbers used for subtractions were changed within the different examinations. The subjects rested for 10 min in the supine position before and after finishing MST. MST was carried out h after the last dose of study medication. The Finapres software automatically registered HR, SBP and DBP 1 min before starting the subtractions, and after 1, 3 and 5 min calculation and 2, 5, 7 and 10 min after finishing the subtractions. Mean arterial pressure (MAP) was calculated with the formula: MAP=DBP + 1 / 3 (SBP DBP). Baroreceptor reflex sensitivity. Finapres recordings of 300 s segments beat to beat BP and HR during rest in the supine position and during the 5 min stress were used for determination of the BRS with the CARS- PAN program (ProGAMMA bv, Groningen, Netherlands), as described previously 15,16. This program allows discrete Fourier transformation of nonequidistant samples of BP and RR interval series. Nonstationary signals or periods with less than 90% normal to normal beats were excluded. Subsequently, spectral analysis of SBP and RR interval length was performed, and BRS calculated with the Transfer Function method. This method defines the BRS as the mean modulus between SBP and RR interval length spectra in the midfrequency band ( Hz) with a coherence of more than 0.5. BRS is expressed in ms/mmhg. Catecholamines. Venous blood was sampled through an 18 gauge Venflon (BOC, Ohmeda AB, Helsingborg, Sweden) in a cubital vein 1 min before starting the subtractions, after 1 and 5 min calculations and 10 min after finishing subtractions. Blood was sampled on prechilled 155

165 Chapter 9 EGTA gluthation tubes, then centrifuged and the plasmas frozen at 70 C.Plasma catecholamines were analysed by a sensitive radioenzymatic technique 18. Data analysis and statistics. The data were analysed using SPSS statistical package (SPSS Inc, Chicago, IL, USA). Non normally distributed data were natural log transformed. Two tailed statistical analyses of data were carried out using Student s t test for paired samples (p), Spearman s correlation coefficient was also used. The HR, MAP and catecholamine reactivity to MST were estimated as percentage change in the area under the curve (AUC) from baseline AUC as described previously 9,19. Baseline AUC was defined as baseline value multiplied with the total duration of the monitored period (16 min). Average levels were defined as the AUC for the measured parameters divided by the length of the monitored period. The level of statistical significance was set at p=0.05. Data are presented as means ± SEM. The study was designed to compare results from the examinations after the two periods with amlodipine and verapamil. Results Catecholamines. Average plasma NA was lower on verapamil than amlodipine treatment, 191 ± 17 pg/l vs 245 ± 23 pg/l, respectively, p= NA reactivity was abolished on verapamil while it was present on amlodipine medication, 2.9 ± 3.3% vs 14.0 ± 5.5%, respectively, p=0.004 (Figure 1). Average plasma A on verapamil was 26 ± 4 pg/l vs 32 ± 5 pg/l on amlodipine, p=0.5. The A reactivity on verapamil was 25.7 ± 11.2% vs 34.3 ± 13.5% on amlodipine, p=0.9 (Figure 1). Heart rate and blood pressure. Average HR was lower on verapamil than amlodipine treatment, 61 ± 2 beats/min vs 65 ± 2 beats/min, p< HR reactivity was lower on verapamil compared to amlodipine medication, 0.1 ± 0.9% vs 2.5 ± 1.0%, respectively, p=0.056 (not significant; Figure 1). BP reactivity, overall and baseline BP did not differ significantly between the two medication periods (Figure 1). Baseline BP on verapamil and amlodipine treatment were 146 ± 3/93 ± 2 mmhg and 144 ± 2/91 ± 156

166 Less Adrenergic Response to Mental Task During Verapamil Compared to Amlodipine Treatment in Hypertensive Subjects Figure 1. Plasma noradrenaline (NA), mean arterial pressure (MAP), heart rate (HR) and plasma adrenaline (A) changes during mental stress test (MST) in 20 hypertensive subjects. Data are means ± SEM. Plasma NA was lower on verapamil treatment compared to amlodipine treatment after 1 and 5 min MST, and 10 min after finishing the calculations, p The NA reactivity was blunted on verapamil compared to amlodipine treatment, p= There was no significant difference in MAP between the two medication periods. Overall HR was lower and the HR reactivity was lower on verapamil compared to amlodipine treatment. There was no significant difference in A between the two medication periods. Amlodipine, Verapamil. 2 mmhg, respectively. Baroreceptor reflex sensitivity. BRS was calculated by two independent investigators. The two BRS estimations correlated significantly (r=0.96, p<0.001). BRS estimated during the 5 min mental stress correlated with average HR, i.e. the total number of heartbeats spent during the monitored period on verapamil and amlodipine medication, r= 0.63 and 0.53, respectively, the regression coefficient (β) was 18.9 and 22.2, respectively, p 0.05 (Figure 2). Baseline HR also correlated, r= 0.53 and 0.49, respectively; β was 1.0 and 1.4, respectively, p We did not find any significant BRS differences between the two medication periods, although there was a tendency of higher BRS both at rest and during MST on verapamil compared to amlodipine (p>0.2). Discussion The present study demonstrated that verapamil, when compared to amlodipine treatment, reduced HR and NA reactivity to MST, as 157