UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

Maintaining cerebral blood flowFrom heart to brainBronzwaer, A.-S.G.T.

Publication date2017Document VersionFinal published versionLicenseOther

Link to publication

Citation for published version (APA):Bronzwaer, A-SGT. (2017). Maintaining cerebral blood flow: From heart to brain.

General rightsIt 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), other than for strictly personal, individual use, unless the work is under an opencontent license (like Creative Commons).

Disclaimer/Complaints regulationsIf you believe that digital publication of certain material infringes any of your rights or (privacy) interests, pleaselet the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the materialinaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letterto: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. Youwill be contacted as soon as possible.

Download date:10 Jun 2021

https://dare.uva.nl/personal/pure/en/publications/maintaining-cerebral-blood-flow(ec61289e-fe5f-4c31-8afe-3ab38de64e98).html

Maintaining cerebral blood flow

- From heart to brain -

Anne-Sophie G.T. Bronzwaer

ANNE-SOPHIE G.T. BRONZWAER

MAINTAINING CEREBRAL

BLOOD FLOWFrom heart

to brain

MA

INTA

ININ

G C

ER

EB

RA

L B

LOO

D F

LOW

Fro

m heart to

brain

AN

NE

-SO

PH

IE G

.T. BR

ON

ZW

AE

R

Maintaining cerebral blood flow

- From heart to brain -

Anne-Sophie G.T. Bronzwaer

Maintaining cerebral blood flow; – From heart to brain –Academic thesis, University of Amsterdam, the Netherlands

ISBN: 978-94-92683-42-7

© 2017 Anne-Sophie G.T. BronzwaerAll rights reserved. No part of this thesis may be reproduced, stored or transmitted in any form or by any means without prior permission of the author or the copyright own-ing journal.

Cover design, layout and printing: Optima Grafische Communicatie, Rotterdam, The Netherlands

Financial support by the Rembrandt Institute for Cardiovascular Science for the publica-tion of this thesis is gratefully acknowledged.

Maintaining cerebral blood flow

– From heart to brain –

ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad van doctoraan de Universiteit van Amsterdamop gezag van de Rector Magnificus

prof. dr. ir. K.I.J. Maexten overstaan van een door het College voor Promoties ingestelde commissie,

in het openbaar te verdedigen in de Agnietenkapelop dinsdag 4 juli 2017, te 12:00 uur

door

Anne-Sophie Gertrude Titine Bronzwaergeboren te Eindhoven

PROMOTIECOMMISSIE

Promotores:Prof. dr. M.J.A.P. Daemen AMC-UvAProf. dr. J.J. van Lieshout The University of Nottingham

Copromotor:Dr. ir. M.J.P. van Osch Universiteit Leiden

Overige leden:Prof. dr. E.T. van Bavel AMC-UvAProf. dr. Y.B.W.E.M. Roos AMC-UvAProf. dr. B.J.M. Mulder-van der Wall AMC-UvADr. ir. A.J. Nederveen AMC-UvAProf. dr. J. Hendrikse Universiteit UtrechtDr. J.A.H.R. Claassen Radboud Universiteit Nijmegen

Faculteit der Geneeskunde

TABlE OF COnTEnTS

Chapter 1 Introduction and outline of the thesis 9

Chapter 2 Physiological principles 17

Chapter 3 Methods 27

Chapter 4 Challenging systemic circulatory control 374.1 Cardiovascular response patterns to sympathetic stimulation by

central hypovolemia39

4.2 The cerebrovascular response to lower body negative pressure versus head-up tilt

57

4.3 Enabling lower body negative pressure in the MRI scanner: a newly designed compact box reduces head motion with preservation of the hemodynamic response

71

4.4 Arterial pressure variation as a biomarker of preload dependency in spontaneously breathing subjects – A proof of principle

85

4.5 Abnormal hemodynamic postural response in patients with chronic heart failure

99

Chapter 5 Quantification of cerebral blood flow 1155.1 Assessment of middle cerebral artery diameter during hypocapnia

and hypercapnia in humans using ultra-high-field MRI117

5.2 Middle cerebral artery diameter changes during rhythmic hand-grip exercise in humans

129

5.3 Transcranial Doppler determined middle cerebral artery blood flow velocity does not match MRI global or regional cerebral tissue perfusion during handgrip exercise

141

Chapter 6 Cardiac output and cerebral blood flow 1576.1 Arterial pressure variations as parameters of brain perfusion in

response to central blood volume depletion and repletion159

6.2 The influence of aging on the relationship between changes in cardiac output and middle cerebral artery blood flow velocity

177

Chapter 7 Summary | Samenvatting 191

Chapter 8 General discussion 201

References 207list of abbreviations 235Authors and affiliations 239PhD Portfolio 243Dankwoord 247

1Introduction and outline of the thesis

Introduction and outline of the thesis 11

11.1 InTRODuCTIOn

‘…Claude Bernard also repeatedly insists, and this deserves especial notice, that when the heart is affected it reacts on the brain; and the state of the brain again reacts through the pneumo-gastric (vagus) nerve on the heart; so that under any excitement there will be much mutual action and reaction between these, the two most important organs of the body...’ Darwin C. (1872)1,2

The brain is a highly metabolic active organ and although it accounts for a small frac-tion of total body weight (2%), it consumes in the resting condition about 20% of total body oxygen and 25% of total body glucose.3 Interestingly, the brain by itself has hardly no capillary recruitment and lacks almost any form of energy storage. This requires suf-ficient and uninterrupted supply of blood to the brain and, to this end, it receives ~15% of the cardiac output (amount of blood being pumped by the heart in one minute).4 Acute interruption of the cerebral blood flow (CBF) and with that oxygen delivery to brain tissue, even for a few seconds, has deleterious effects and results in loss of con-sciousness.5 On the other hand, an increase in CBF to well above normal values may elicit (brain tissue) hyperperfusion with edema development leading to migraine-like headaches, seizures and intracerebral hemorrhage.6 Therefore, safeguarding the brain from respectively hypo- and hyperperfusion is of major importance.

The regulation of CBF is complex and involves a powerful set of control mechanisms including the cerebral autoregulation, chemoregulation, neurogenic regulation and neurovascular coupling.7,8 These mechanisms are largely unique to the cerebral vascu-lature and act directly on the cerebral vascular resistance. CBF regulation at the level of the brain itself is often assumed to be so efficacious that it is treated as a separate entity.8 The perfusion of the brain is, however, also reliant on a sufficient supply of blood from the heart: acute interruption of CBF due to cardiac arrest9 or occlusion of a large cerebral artery10 results in loss of consciousness within seconds. Thus in addition to an effective set of cerebrovascular control mechanisms, the function of the heart and condition of the large brain-feeding arteries are also important in securing the blood supply to the brain. In other words, CBF also depends on factors beyond the brain.11-14 The idea of considering the cardiovascular system as an integral component of CBF control has laid the foundation of this thesis.

1.1.1 Heart-brain connection

The circulation of blood starts with the heart pumping blood through the blood vessels to the tissue under consideration, in this case the brain. To guarantee a sufficient CBF, all of these elements need to function properly. Evidence has emerged that cardiac dysfunction, impaired vessel patency or a critical reduction in central blood volume lead

12 Chapter 1

to insufficient blood supply to the brain. This affects CBF independently from cerebro-vascular control mechanisms whereas, in turn, insufficient brain perfusion may lead to a decline in cognitive performance.15

This followed, among others, from observations in patients with chronic heart failure who showed a lower CBF16-18 which is associated with a larger prevalence of cognitive dysfunction.19,20 More recently, the Framingham Heart Study revealed in a longitudinal cohort a similar association between a low cardiac output and an increased risk of in-cident dementia and Alzheimer’s disease.21 In addition, about half of the patients with impaired patency of large brain-feeding vessels (for instance in carotid artery occlusive disease), have signs and symptoms of mild cognitive impairment.22,23 In patients with chronic heart failure, improvement or reinstitution of cardiac function by, respectively, a left ventricular assist device, cardiac resynchronization therapy or a heart transplanta-tion, improved or restored CBF24-26 with beneficial effects on cognitive performance.27-29 There is, as yet, no conclusive evidence that extracranial-intracranial bypass surgery improves cognitive function in patients with a symptomatic internal carotid artery stenosis.30 A decrease in central blood volume too large to secure cardiac output may also affect CBF. An example for this phenomenon follows from the daily life physiological observation when healthy humans stand-up. Changing body position from the supine to the upright posture leads to a gravity induced shift in blood from the upper body towards the legs affecting the return of blood towards the right heart (venous return).31 The postural reduction in the amount of blood that is directly available to the heart (central blood volume) results in a decline in cardiac output with a drop in CBF and CBF velocity.3,32-35

Aforementioned observations suggest a functional link between the effective arterial blood volume, the heart and brain-feeding vessels on one hand, and brain perfusion and cognition on the other. Optimizing blood supply to the brain by improving car-diovascular function may become a new preventive or therapeutic target for cognitive disorders. As yet, the current monodisciplinary approach by clinicians and researchers alike, leaves the complete heart-brain connection as an incompletely explored territory. A synergy of different expertise fields may provide more insight into the interesting and clinically increasingly important relationships between the heart and the brain.

1.1.2 Aim

From current knowledge on heart-brain interactions, we hypothesize that maintain-ing CBF requires cardiovascular functional integrity in addition to cerebrovascular autoregulatory control. The challenge of disentangling the contributions of systemic versus local brain blood flow regulatory mechanisms requires expert contribution from various disciplines. At the onset of this research project we had the good fortune to start a multidisciplinary collaboration, working closely with a MRI perfusion expert, a

Introduction and outline of the thesis 13

1neuro-radiologist, a pathologist, and an internist-physiologist sharing an interest in the heart-brain connection. The primary aims of the studies presented in this thesis are the delineation of the human heart-brain connection by integrating physiological concepts into the MRI environment and quantification of CBF and its regulation mechanisms at the macrovascular level (using transcranial Doppler ultrasound) and at the tissue level (using arterial spin labeling MRI). This brought together expertise on the characteriza-tion of the systemic and cerebrovascular response to physiological challenges in health and disease as well as on brain perfusion at the tissue level using high resolution MRI. From observations in healthy volunteers and in patients with heart failure this thesis attempts to provide at least a partial answer to the following questions:

1. How to challenge the cardiovascular system in a controlled and reproducible man-ner (Chapter 4.1) and how to apply these interventions in a MRI setting (Chapters 4.2 and 4.3)?

2. Can central hypovolemia be detected non-invasively by arterial pressure wave analysis (Chapter 4.4) and, if so, is cerebral perfusion under those conditions pre-dominantly related to arterial pressure or systemic blood flow (Chapter 6.1)?

3. Is the postural cardiovascular response affected in patients with chronic heart failure and, if so, related to multidrug blockade treatment or the cardiac condition (Chapter 4.5)?

4. What is the effect of CO2 partial pressure (Chapter 5.1) and of sympathetic activa-tion (Chapter 5.2) on the diameter of the middle cerebral artery as one of the main brain-feeding arteries? Does transcranial Doppler determined CBF velocity in a large cerebral artery reflect CBF in brain tissue as measured with arterial spin labeling MRI (Chapter 5.3)?

5. Does aging influence the relationship between cardiac output and CBF velocity as measured with transcranial Doppler ultrasound (Chapter 6.2)?

1.2 OuTlInE OF THE THESIS

The studies described in this thesis have been performed in two different centers: the Laboratory for Clinical Cardiovascular Physiology of the Academic Medical Center (AMC; affiliated with the University of Amsterdam) and the C.J. Gorter Center for High-Field MRI, Department of Radiology of the Leiden University Medical Center (LUMC).

Chapter 2 provides a brief overview of the physiological principles involved in systemic and cerebral blood flow control. Chapter 3 describes the methods of continuous and non-invasive monitoring of systemic, cerebral and respiratory parameters, followed by

14 Chapter 1

a description of the physiological interventions that were used to assess autonomic cardio- and cerebrovascular control.

Chapter 4 focuses on the identification and development of cardiovascular challenges that will bring the circulatory system out of balance while being suitable for application in the physiology laboratory as well as inside the MRI scanner. To be able to study the effects of systemic blood flow on brain perfusion, such a challenge need to acutely and reversibly affect cardiac output while being patient friendly (e.g. safe, non-invasive and easily applicable). In Chapter 4.1 we investigated the reproducibility and robustness of the cardiovascular response to lower body negative pressure (LBNP). LBNP has the potential to serve as a MRI compatible surrogate of orthostatic stress as it also reduces the magnitude of the central blood volume while the subject remains in the supine position. In Chapter 4.2 we tested whether the cardio- and cerebrovascular response to LBNP performed in the supine body position is representative for the response in-duced by passive head-up tilt. In Chapter 4.3 we tested whether we could reduce head movement with a custom designed compact LBNP box (covering only the pelvic region and upper legs) while imposing a similar cardiovascular challenge compared to conven-tional LBNP. This would be essential to increase the reliability of MRI measurements of cerebrovascular hemodynamics during LBNP.

Whereas the desired effect of cardiovascular challenges applied in this thesis have in common that they modulate the central blood volume, quantifying the central blood volume itself remains difficult. In Chapter 4.4, we investigated the value of respiration-related variation in arterial pressure (e.g. pulse- and systolic pressure variation) as potentially useful biomarkers in the detection of central hypovolemia in spontaneously breathing subjects. In addition to studies in healthy volunteers, we determined in Chap-ter 4.5 the cardiovascular response to orthostatic stress in patients with a chronically compromised cardiac function.

Chapter 5 addresses the validity of CBF velocity measurements with transcranial Dop-pler ultrasound under a variety of circumstances. It has thus far been assumed that the diameter of the large intracranial arteries remains unaffected by changes in arterial CO2 partial pressure and cerebral perfusion pressure. Based on this assumption, changes in CBF velocity are considered directly proportional to changes in CBF. However, inconsis-tent findings in literature do not exclude the possibility of diameter changes in large intracranial arteries under the conditions of the studies presented in this thesis. Validity of this assumption is relevant for the interpretation of data on CBF velocity obtained by transcranial Doppler ultrasound. We therefore tested the effects of CO2 partial pressure (Chapter 5.1) and sympathetic activation by dynamic handgrip exercise (Chapter 5.2) on the diameter of the middle cerebral artery using ultra-high-field MRI at 7 Tesla.

Introduction and outline of the thesis 15

1Chapter 5.3 addressed whether CBF velocity changes in large brain feeding arteries reflect perfusion changes at the brain tissue level. We therefore compared the CBF veloc-ity response measured in the middle cerebral artery by transcranial Doppler with the CBF tissue response as assessed with arterial spin labeling MRI to sympathetic activation by dynamic handgrip exercise.

Chapter 6 focuses on the relationship between cardiac output and transcranial Dop-pler determined CBF velocity. In Chapter 6.1 we report on the relationship between CBF velocity and cardiac output on one hand and CBF velocity and arterial pressure on the other when manipulating the central blood volume from depletion to repletion by head-up tilt. Chapter 6.2 describes the influence of aging on the CO-CBF velocity relationship.

2Physiological principles

Physiological principles 19

2

2.1 REGulATIOn OF SySTEMIC BlOOD FlOw

The following sections describe the basic concepts of systemic hemodynamics, cardiac output and cardiovascular control as these concepts are part of the research that we performed in Chapters 4-6.

2.1.1 Systemic hemodynamics

Hemodynamics is defined as the dynamics of blood flow. Flow through a vessel follows the laws of mass-equality and inertia, but is simplified in steady-state by using the he-modynamic analog of Ohm’s law (generally used to describe electrical current). When applied to the cardiovascular system the blood flow is driven by the pressure difference between two points along the vessel together with the vascular resistance (flow = pres-sure difference / vascular resistance). Under steady state conditions arterial pressure is considered as the driving force behind flow and organ perfusion, and is therefore tightly regulated.36

Quantification of blood flow in living humans continues to be difficult whereas the measurement of blood pressure is much easier. The first non-invasive arterial blood pressure device was already invented in 1881 by Von Basch.37 Notwithstanding that the concept of measuring blood flow was already published in 1870,38 the measurement of cardiac output, either invasively or non-invasively was not possible before the mid of the 20th century. Determination of blood flow, and in particular the flow leaving the heart, is considered vital when managing critically ill patients (for instance those with severe cardiac or pulmonary disease and/or multi-organ failure). Nowadays, several non-invasive techniques (based on rebreathing Fick, Doppler, impedance and finger plethysmography) are available and provide real-time and continuous estimates of cardiac output and stroke volume.39 Also MRI can provide non-invasive quantifications of cardiac output.40 In this thesis we primarily focus on blood flow as the mechanism responsible for the transport of oxygen and nutrients, with blood pressure as the driving power behind flow.

2.1.2 Cardiac output

The amount of blood leaving the heart in one minute is designated as the cardiac output (CO) and equal to left ventricular stroke volume (the amount of blood ejected with each cardiac cycle) times heart rate. Stroke volume is dependent upon preload, contractility and afterload. Cardiac preload is defined as the amount of blood directly available to the left ventricle and relates to thoracic fluid content rather than to central vascular pressures.41 Therefore cardiac preload is often –also in this thesis– being referred to as ‘central blood volume’. Contractility refers to the intrinsic muscle strength of the (left) ventricle independent of its loading condition, and afterload refers to the force that

20 Chapter 2

opposes ejection of blood out of the ventricle (largely influenced by aortic impedance, arterial blood pressure and peripheral vascular resistance). Activities of daily living such as standing-up and physical exercise but also pathophysiological conditions including heart failure and dehydration modify CO by changing one or more of these (extra)cardiac factors.42

2.1.3 Cardiovascular control

To ensure a continuous and appropriate supply of oxygenated blood to the tissues, the systemic circulation is subject to precise cardiovascular control mechanisms. Although the cardiovascular system is under control of both neural and humoral components of the autonomic nervous system, in this thesis the focus will be on the neuro-cardiovascu-lar system with the arterial baroreflex as its best known example. The arterial baroreflex is the fastest control mechanism of blood pressure and generally acts within seconds.43 The reflex arc consist of peripheral (stretch) receptors embedded in the aortic arch and carotid arteries, afferent pathways (toward the central nervous system), pathways within the central nervous system itself, efferent nerves and finally the effector organs (e.g. cardiac conductive tissue, cardiac muscle and vascular wall muscle fibers),44 see Figure 2.1.1. Hence, a fall in arterial pressure leads to reflex adjustments by parasympa-thetic inhibition and sympathetic activation, with an increase in heart rate, in cardiac contractility, and in vascular resistance and venous return.45 Conversely, an increase in arterial pressure results in opposite reflex changes. This provides a real-time feedback control loop on the arterial blood pressure, involving the heart but also the peripheral vasculature.

2.1.4 Central hypovolemia

Severe central hypovolemia challenges cardiovascular control. An example of a critical reduction in central blood volume is uncontrolled hemorrhage or severe dehydration, e.g. by burn wounds or cholera. The central blood volume will also decrease in response to orthostatic stress (e.g. standing or passive head-up tilt) due to a gravitational shift of blood from the central circulation into the lower extremities leading to a decline in venous return. These (patho)physiological conditions are characterized by insufficient availability of intravascular volume, resulting in a progressive reduction in preload, stroke volume and finally CO. When reflex adjustments in heart rate and vascular resis-tance can no longer compensate for the progressive reduction in central blood volume, re-establishment of normovolemia by fluid administration is considered the cornerstone of treatment in hemodynamically unstable patients.47 The rationale behind volume expansion is to restore appropriate tissue perfusion and oxygenation by increasing the central blood volume.

Physiological principles 21

2

Arterial baroreceptor

activity

Parasympathetic activityArterial

pressure

Negative feedback

Sympathetic activity

Heart rate

Stroke volume

Vascular resistance

Cardiac output

+

−

+

+

+ +

+

−

+ +

Vessel

HeartAortic arch

Carotid sinus

Central nervous system

Figure 2.1.1. Schematic drawing of the afferent and efferent pathways of the baroreceptor reflex arc (modified from Wehrwein and Joyner 201346) and a block diagram of baroreflex mediated adjustments to a fall in arterial pressure.

Diagnosing a clinically relevant volume deficit is difficult because traditional clinical signs considered specific for hypovolemia including diminished skin turgor and high urine osmolarity, do regularly not accurately reflect a reduction in central blood volume.48 A meta-analysis of 12 clinical studies showed that with current clinical practice, between 40 and 70% of critically ill patients are so-called ‘fluid responders’.49 The substantial number of patients not responding to fluid administration by increasing stroke volume or CO, calls for physiological markers capable to detect fluid responsiveness. To this end, variations in arterial pressure, like systolic (SVP) and pulse pressure variation (PPV) have been proposed as potentially useful biomarkers to guide fluid administration.50 These dynamic indices are based on respiratory-induced changes in venous return51,52 and associated variations in left ventricular preload transferred to arterial pressure (Figure 2.1.2).Although in previous research these indices have been proven of clinical value, their application remains limited to patients who are mechanically ventilated with high tidal volumes.53-56 Chapters 4.4 and 6.1 focuses on arterial pressure variations as biomarkers for fluid responsiveness in spontaneously breathing subjects and their relation with brain perfusion as actual therapeutic endpoint.

22 Chapter 2

60

80

100

120

140

-10

0

10

20

30

40

50

Time

CO2 (

mm

Hg)

Art

eria

l pre

ssur

e (m

mH

g) PPmax

SPmaxSPminPPmin

Inspiration Expiration

Figure 2.1.2. Variation in pulse and systolic pressure (PP and SP, respectively) during the respiratory cycle.

2.2 REGulATIOn OF CEREBRAl BlOOD FlOw

CBF is defined as the blood supply to (a given part of ) the brain in a given time, which in rest equals to approximately 50-55 mL per 100 g brain tissue per minute in normoten-sive adults.4 The regulation of CBF is considered to act directly on the cerebrovascular resistance independently of systemic hemodynamics. The mechanisms involved in CBF control, i.e. cerebral autoregulation (also known as mechanoregulation), chemoregula-tion, neurogenic regulation and neurovascular coupling will be discussed in the follow-ing paragraphs.

2.2.1 Cerebral autoregulation

In the late fifties of the last century, Lassen proposed the ‘classic’ cerebral autoregulation (CA) curve relating CBF to cerebral perfusion pressure (CPP).4 The CPP is the difference be-tween mean arterial blood pressure (MAP) at the level of the circle of Willis and intracranial pressure, encompassing central venous pressure and the cerebral spinal fluid pressure. The traditional CA curve suggests more or less constant CBF for a wide range of perfusion pressures via adaptations in the cerebrovascular resistance (Figure 2.2.1). Recent data showed that the human brain is more effective at compensating for transient hyperten-sion than hypotension, providing evidence for hysteresis in the cerebral pressure-flow re-lationship.57 Beyond the so-called pressure limits of regulation (e.g. lower and upper limit), autoregulation is lost and CBF changes proportionally to the change in CPP. It has been

Physiological principles 23

2

suggested that the lower limit is not a fixed value and that it may shift towards a higher CPP in hypertensive subjects and vice versa to a lower CPP in patients with orthostatic hypotension related to sympathetic failure.12,13,58-61 When the CPP falls below the lower limit of autoregulation, CBF decreases and cerebral ischemia ensues.

Cerebral perfusion pressure (mmHg)

Cere

bral

blo

od �

ow (v

eloc

ity) Cerebral vascular resistance

60 150

Figure 2.2.1. The classical cerebral autoregulation curve describing the pressure-flow relationship for the brain. The CBF (velocity) is maintained more or less constant (the so-called autoregulatory plateau) via changes in the cerebrovascular resistance. Below and above the limits of autoregulation (150 mmHg), the brain becomes ‘pressure-passive’ as represented by the linear portion of the curve. Modified from Lucas et al. 2010.65

Under physiological conditions, CA adapts cerebrovascular tone in response to changes in CPP using a stretch sensing mechanism of the vascular smooth muscle cells (‘the Bayliss myogenic response’).62 Stretching of these cells is considered to induce the signals that provoke vasoconstriction while a reduction in transmural pressure leads to vasodilation. Although this phenomenon is assumed to be an inherent property of vascular smooth muscle cells,63 the exact underlying signaling pathways are incompletely defined.

To adapt to the brain’s metabolic demand on CBF in daily life, both fast and slower acting components of the CA are required. In the laboratory, static (hours to days) and dynamic (seconds to minutes) components of CA can be distinguished in either the time- or frequency domain. The dynamic component is assumed to reflect the capacity to counteract alterations in CBF in response to fast changes in blood pressure. This com-ponent operates within seconds and represents the latency of the cerebral vasoregula-tory system.64 The static component reflects the capacity to adapt the CBF in response to long term changes in blood pressure and rather represents the efficiency of the system.

2.2.2 Chemoregulation

The cerebral chemoregulation involves the strong cerebrovascular responsiveness to changes in arterial partial pressures of carbon dioxide (PaCO2) and oxygen (PaO 2).66-70

24 Chapter 2

Lowering versus elevating the PaCO2 (e.g. hypo- and hypercapnia) causes, respectively, vasoconstriction and vasodilatation of the capillaries, arterioles, and large extracranial vessels, leading to alterations in CBF. Conceptually, chemoregulation operates indepen-dently from the CA, but they may have common pathways and mechanisms.8,71-75 The exact mechanism of action has not been fully clarified. For long it has been thought that the PaCO2-driven changes in pH modify CBF by direct relaxation and contraction of the smooth muscle cells.76-78 There is, however, also data suggesting that PaCO2 regulates CBF both independently and/or in conjunction with altered pH.78

The sensitivity of CBF to changes in carbon dioxide is generally expressed as its percentage change per mmHg in PaCO2 (the CO2 reactivity of the brain), and is often quantified non-invasively by relating changes in CBF (velocity) to those in end-tidal CO 2. In the normocapnic range, transcranial Doppler determined CBF velocity in the middle cerebral artery changes approximately 3.5% per mmHg change in end-tidal CO 2.75,79-81

2.2.3 neurogenic regulation

Cerebral arteries are abundantly innervated by sympathetic nerves originating from the cervical ganglion but their role in CBF control is still under debate.13,82,83 It is assumed that under normal physiological conditions there is probably little influence of the central nervous system on CBF and its regulation.83 However, increased sympathetic activation during, for instance, exercise likely enhances cerebral vascular tone thus counteract-ing imminent cerebral hyperperfusion as a consequence of an excessive increase in BP beyond the cerebral autoregulatory range.84-87 Both sympathetic and cholinergic mechanisms are considered important for restricting the exercise-induced increase in cerebral perfusion on CBF without affecting the cerebral metabolic rate for oxygen.84,88

2.2.4 neurovascular coupling

Neuronal activation increases cerebral metabolic demand. Since the brain lacks exten-sive storage of energetic compounds, the required supply of oxygen and nutrients is rapidly adjusted by locally increasing the CBF.7,8,89-91 This interplay of supply and demand implies a connection between neurons and the local cerebral vasculature, to which the term neurovascular coupling was coined. This process is probably mediated through the astrocytes that surround the arterioles.92-95 In response to neuronal activation, elic-ited for example by a sensory stimulus, the CBF shows an overshoot with the supply transiently exceeding the demand, resulting in a locally increased oxygenation level. Imaging methods sensitive to either CBF or oxygen concentration, apply this principle in order to identify functional regions in the brain in relation to neuronal activation. The localized nature of the response is the main distinguishing property of neurovascular coupling compared to other regulatory mechanisms.

3Methods

Methods 29

3

3.1 PARAMETERS

Investigation of the cardio- and cerebrovascular response to physiological stress requires (simultaneous) monitoring of systemic, cerebral and respiratory parameters. This chapter provides an overview of the parameters that were measured in the studies described in this thesis.

3.1.1 Systemic

Blood pressureContinuous arterial blood pressure (BP) can be measured non-invasively by finger pleth-ysmography (Nexfin, Edwards Lifesciences BMEYE, Amsterdam, the Netherlands) using a volume-clamp technique.96 An optical plethysmograph in the finger cuff measures arte-rial volume continuously whereas the volume is clamped by applying variable pressures in an inflatable cuff around the finger countering the pulsatile arterial pressure. The cuff is placed around the midphalanx of the non-dominant hand and held at heart level. A height reference system is placed around a finger next to the cuff and at heart level to account for the hydrostatic pressure difference. To detect changes in BP correctly, an automatic built-in calibration system (Physiocal) tracks the unloaded diameter of the finger artery to establish and adjust the arterial unloaded volume.97 Finger arterial pressure is fundamentally different from brachial pressure in terms of wave shape and absolute levels such that waveform transformation and level corrections are applied in the Nexfin system to reconstruct brachial pressure.98

In case finger plethysmography is not available (for instance in the MRI environment), BP measurements are taken every 2–4 min using an inflatable arm-cuff (Magnitude, In-Vivo, Orlando, FL) while HR can continuously be monitored by means of an MRI compat-ible finger pulse-oximetry unit.

Heart rate, stroke volume and cardiac outputA pulse contour method (Nexfin CO-trek, Edwards Lifesciences BMEYE, Amsterdam, the Netherlands), adapted for age, sex, height and weight,99 provides left ventricular stroke volume (SV) and CO (equal to SV multiplied by instantaneous HR). This method has been thoroughly validated against invasive thermodilution measurements.99,100

In Chapter 6, CO was also measured by means of inert gas rebreathing (Innocor, In-novision A/S, Odense, Denmark).101,102 During the rebreathing procedure, blood-soluble N2O diffuses from the lung alveoli to the systemic circulation, and blood-insoluble NF6 remains in the pulmonary fields. The disappearance rate in the bag volume is propor-tional to the pulmonary blood flow which is assumed to be equal to the CO of the left ventricle.103

30 Chapter 3

3.1.2 Brain

In this thesis, we assessed with two non-invasive modalities the CBF response to a variety of physiological challenges. First, transcranial Doppler ultrasound (TCD) provides high temporal resolution assessments of CBF velocity (CBFv) in large brain-feeding arteries. This method is a relatively simple and low-cost bedside technique and is assumed to reflect mean CBF over a large area of the brain; that is, the flow territory perfused by the insonated artery. In addition, MRI techniques such as arterial spin labeling (ASL) and blood-oxygen-level dependent (BOLD) imaging enable the measurement of whole brain CBF and oxygenation at the microvascular tissue level. MRI is a complex, costly and time-consuming procedure that offers a non-invasive measure of brain perfusion and oxygenation at a high spatial resolution. A combination of TCD and MRI based quantifications of CBF has the potential to complement each other in obtaining a more complete understanding of brain perfusion at both the macro- and microvascular level. In the following paragraphs we will discuss both modalities into more detail.

Middle cerebral artery blood flow velocityMeasuring CBFv in the basal cerebral arteries by TCD was introduced in the early eighties of the twentieth century by Aaslid and coworkers,104 and has found wide acceptance in both clinical and research settings. The ultrasound probe emits a high-pitched sound wave through the intact scull, which is then reflected back from erythrocytes moving in its path. The CBFv is recorded from the Doppler shift spectrum of the reflected sound waves.105 Mean CBFv reports the velocity associated with the maximal frequency of the Doppler shift (‘the envelope’) rather than the intensity-weighted mean flow veloc-ity or the total signal power. These latter two variables are sensitive to small changes in insonation angle of the artery and therefore the maximum velocity is preferred as reported entity (Secher, Seifert et al. 2008). In the studies described in this thesis, a TCD system (DWL Multidop X4, Sipplingen, Germany) with a pulsed ultrasound frequency of 2 MHz was used in order to satisfactory penetrate the skull. As the bone of the of temporal region is thin and therefore the best promising area for penetration,104 CBFv was measured in the proximal segments of the left or right middle cerebral artery (MCA). The ultrasound probe was placed on the temporal region of the skull just above the zygomatic arch (Figure 3.1.1). Once the optimal signal-to-noise ratio was obtained at an insonation depth between 45 and 60 mm, the probe was secured in position by a head-band.

Methods 31

3A. B.

Figure 3.1.1. The ultrasound probe is placed on the temporal region of the skull (dotted line indicates the ‘temporal window’) just above the zygomatic arch (A). A frontal view of the ultrasound probe directed toward the MCA (B). The cylinder around the MCA indicates the observation region (sampling volume) and the distance from the middle of the cylinder to the probe corresponds to the depth setting. Reprinted with permission from J Neurosurg.104

The relation between calculated versus actual CBFv depends on angle of insonation.106 When the angle increases from 0° to 30°, its cosine will decrease from 1 to 0.86 result-ing in a maximum error up to 15%.105 By immobilizing the probe by a head-band, we minimized the influence of a potential change in angle as might occur during the experiments. A critical issue of TCD is to what extent blood flow velocity reflects actual blood flow. Changes in blood flow velocity reflect those in blood flow when the cross-sectional area of the insonated vessel remains constant (blood flow = blood flow veloc-ity x cross-sectional area). Direct observations made during craniotomy reveal that the vessel diameter does not change significantly during moderate variations in mean BP or CO2 partial pressure.107 Also orthostatic stress, as stimulated by lower body negative pressure (LBNP), does not alter the diameter of the MCA significantly as assessed with 3 Tesla MRI.108 These findings suggest that the MCA diameter remains constant and that changes in TCD determined CBFv will track those in CBF. In two studies presented in Chapters 5.1 and 5.2, we examined this assumption under different circumstances using high-resolution MRI at 7 Tesla.

Whole brain blood flowSince the proposal of ASL two decades ago,109 the non-invasive quantification of regional CBF with MRI has progressively developed into a well-accepted and clinically suitable technique. ASL-MRI is based on the detection of a tracer that is delivered to and cleared from the tissue by blood flow,110 and usually expressed in ml·min-1 per 100g tissue.111,112 With ASL, an endogenous tracer is created by inverting the proton spins of blood, mainly located in water molecules (H2O). Magnetic labeling of arterial blood water spins is done by a long series of radiofrequency pulses (in pseudo-continuous ASL) that are applied in a plane perpendicular to the neck. Subsequently, the labeled protons in arte-rial blood water act as freely diffusible tracers. From the labeling location the labeled

32 Chapter 3

protons migrate within 1-2 seconds via the arterial vessels and capillaries into the brain tissue where the label accumulates, thereby altering the local tissue magnetization. The change in tissue magnetization is measured by comparing multiple image slices covering the whole brain with identical control scans in which the inflowing blood was not labeled. A 3-dimensional perfusion map can be obtained by subtracting the labeled image volume from the control image volume (no label) (Figure 3.1.2).

Labeling Post labeling delay Acquisition

Figure 3.1.2. Principle of arterial spin labeling. The magnetic labeling of arterial protons is carried out up-stream from the volume of interest, at the neck vessels, by radiofrequency pulses. The labeled protons then migrate via the arterial vessels towards the brain tissue where they extravasate from the capillary compart-ment to the extravascular compartment. After the labeling, a delay time (the so-called post-labeling delay; PLD) of approximately 1.5-2.0 seconds allows the labeled protons to reach the tissue compartment, after which the images are acquired. The control acquisition is obtained in a highly similar manner, but without inverting the arterial magnetization.

The ASL-signal, i.e. the difference in signal intensity between label and control images, is small (~1%). To obtain a sufficient signal-to-noise ratio (SNR), many repetitions of the control and label pairs are acquired during 3-5 minutes. The ASL technique applied here is pseudo-continuous ASL, as the recommended standard for use in a clinical setting,112 and which has been recently compared with 15O H2O positron emission tomography (PET) CBF measurements.113 Background suppression RF pulses were used to enhance the SNR of the CBF signal. In addition, the imaging module was extended with an extra echo block to obtain the BOLD fMRI signal with minimal additional scan time.114 The BOLD signal is mainly sensitive to the concentration of deoxy-hemoglobin, and also de-pends on blood flow, blood volume and tissue properties, such as diffusion. This makes this method less specific than ASL. Changes in ASL or BOLD determined regional are of-ten used as proxy for neuronal activation.115 Chapter 5.3 describes a comparative study of the determination of CBF changes upon small muscle group exercise as measured by either ASL-MRI and TCD.

Methods 33

3

3.1.3 Respiration

Brain perfusion is highly sensitive to changes in PaCO2. To enable a correct interpretation of the CBF and CBFv responses, it is highly recommended to also monitor (changes in) PaCO2. The partial pressure of CO2 in exhaled air (designated as end-tidal CO2 partial pressure; PetCO2) is generally used as a non-invasive proxy for PaCO2 and therefore measured in the studies described in this thesis.

Partial end-tidal carbon dioxide pressurePetCO2 was continuously monitored, via a nasal cannula, by a sampling infrared cap-nograph (Tonocap, Datex-Ohmeda, Madison, USA or Datex Normocap 200, Helsinki, Finland). This technique is based upon the absorption of infrared radiation by CO2, with the amount of absorbed radiation having a nearly exponential relation to the CO2 con-centration. Detecting a change in infrared radiation levels by means of photo-detectors, allows for the calculation of the CO2 concentration in the gas sample.

3.2 PHySIOlOGICAl CHAllEnGES

The present thesis discusses various physiological methods (including orthostatic stress tests, small muscle group exercise and inhalation of a gas mixture containing CO2) that were applied to address autonomic cardio- and cerebrovascular control. This section summarizes these methods with respect to their known physiological mechanism and the way in which they were performed.

3.2.1 Orthostatic stress tests

Passive head-up tilt (HUT), lower body negative pressure (LBNP) and orthostasis (stand-ing up) all lead to a gradual translocation of blood from the intra-thoracic region into the lower parts of the body. As a result, CO decreases and a series of cardiovascular regulating mechanisms and reflexes come into action to maintain arterial blood pres-sure and cerebral perfusion.

Passive head-up tiltPassive HUT is performed with the subject lying supine and safely strapped on a tilt table (custom built by AMC Medical Technological Development / Dr. Kaiser Medizintechnik, Bad Hersfeld, Germany) and then either mechanically or manually tilted to a semi-supine (30°), semi-upright (45°) and/or almost completely upright (70°) position. Tilting back from 70° HUT to the supine position lead to central blood volume repletion and mimics a fluid challenge.

34 Chapter 3

Translocation of blood takes place because the hydrostatic indifference point for intravascular pressure (i.e. point in the vascular tree at which pressure remains constant independent of body position) is situated at the level of the diaphragm. Moreover, the indifference point for volume as monitored by electrical impedance is even lower posi-tioned between the navel and iliac crest.116 As a consequence, from supine to upright approximately 70% of the blood volume becomes positioned below the level of the heart, largely being contained in the (compliant) venous compartment and, therefore, not contributing to the effective arterial blood volume.117,118 This shift in blood volume distribution is estimated to be 300-800 ml of which 50% takes place within the first few seconds.42,119,120 The central blood volume is challenged further by an estimated 10% or ~500 ml reduction after 5 min and 15% or ~750 ml reduction after another 5 minutes in the HUT position.121

Lower body negative pressureDuring LBNP, sub-atmospheric pressure is applied to the lower limbs in a supine sub-ject such that blood redistributes from the upper parts of the body into the compliant compartment of the lower extremities. In preparation for LBNP, the lower body of the subject is positioned inside a LBNP box (Dr. Kaiser Medizintechnik, Bad Hersfeld, Ger-many / Dept. Instrumental Development, Leiden University Medical Center, Leiden, the Netherlands) and sealed at the level of the iliac crest.122 An advantage of this technique compared to passive HUT, is its usability within the static and horizontal setup of the MRI scanner. Chapter 4.2 provides a formal comparison of the cardio- and cerebrovascular response to LBNP and passive HUT. In Chapter 4.3, we introduce a custom designed compact LBNP box that meets MRI requirements more closely.

Standing upThe presumed mechanism behind the gravitational translocation of blood from the intra-thoracic region to the veins in the legs during passive HUT is similar to that when humans stand up from the supine position. However, the active change in posture dur-ing standing-up produces a hemodynamic response that is different from what happens with passive tilt during the first 30 s of upright posture.123

3.2.2 Small muscle group exercise

Rhythmic handgrip exercise is a form of small muscle group exercise that increases HR and CO, with modest changes in blood pressure.124 An advantage of rhythmic handgrip-ping is that it can be performed in the supine position while ensuring minimal (head) motion, which makes it a suitable exercise method during MRI monitoring. Moreover, a mild to moderate handgrip exercise level can be maintained for a longer period of time to achieve the steady state needed for acquisition of ASL measurements, which

Methods 35

3

typically take about 3-5 minutes. To standardize the workload between individuals, the subjects were first instructed to squeeze a handgrip dynamometer (Gripforce 500N, Current Designs Inc., Philadelphia PA, USA) to the maximum extent possible for 2-3 s without tensing the entire body. The so-measured maximum force was taken as 100%. The exercise experiments described in this thesis consisted of 0.5 Hz intermittent hand-grip contractions performed for the first minute at 80% of the maximum force followed by 4 minutes at 60%. The decreasing force protocol was used to achieve a steady-state in minutes 3 to 5.

3.2.3 Inhalation of a gas mixture containing CO2In order to quantify the cerebral vasomotor reactivity, a wide range of PetCO2 was es-tablished by inhalation of a gas mixture containing 5% CO 2 and 95% O 2 (Carbogen) for 2 minutes, followed by 2 minutes of breathing room air and, finally, hyperventilating for another 2 minutes.

4Challenging systemic circulatory control

4.1Cardiovascular response patterns to sympathetic stimulation by central hypovolemia

Anne-Sophie G.T. Bronzwaer* Jasper Verbree*Wim J. StokMark A. van BuchemMat J.A.P. DaemenMatthias J.P. van OschJohannes J. van Lieshout

*shared first authorship

Frontiers in Physiology 2016 Jun 20;7:235

40 Chapter 4.1

ABSTRACT

In healthy subjects, variation in cardiovascular responses to sympathetic stimulation evoked by submaximal lower body negative pressure (LBNP) is considerable. This study addressed the question whether inter-subject variation in cardiovascular responses co-incides with consistent and reproducible responses in an individual subject. In 10 healthy subjects (5 female, median age 22 years), continuous hemodynamic parameters (finger plethysmography; Nexfin, Edwards Lifesciences) and time-domain baroreflex sensitivity (BRS) were quantified during three consecutive 5-minute runs of LBNP at -50 mmHg. The protocol was repeated after one week to establish intra-subject reproducibility. In response to LBNP, 5 subjects (3 females) showed a prominent increase in heart rate (HR; 54±14%, p=0.001) with no change in total peripheral resistance (TPR; p=0.25) whereas the other 5 subjects (2 females) demonstrated a significant rise in TPR (7±3%, p=0.017) with a moderate increase in HR (21±9%, p=0.004). These different reflex responses coin-cided with differences in resting BRS (22±8 vs. 11±3 ms/mmHg, p=0.049) and resting HR (57±8 vs. 71±12 bpm, p=0.047) and were highly reproducible over time. In conclusion, we found distinct cardiovascular response patterns to sympathetic stimulation by LBNP in young healthy individuals. These patterns of preferential autonomic blood pressure control appeared related to resting cardiac BRS and HR and were consistent over time.

Cardiovascular response patterns to central hypovolemia 41

4

InTRODuCTIOn

Lower body negative pressure (LBNP) is used in research settings as a model to study the cardiovascular effects of central hypovolemia in humans.125 Application of sub-atmospheric pressure to the lower body redistributes fluid from the upper parts of the body into the compliant compartment of the lower extremities, leading to a decrease in venous return and central blood volume.126 Central blood volume is important for filling of the heart and directly affects stroke volume (SV) and cardiac output (CO). In response to a progressive reduction of central blood volume as elicited by LBNP, both SV and CO decrease modifying arterial pulse pressure and its pulsatility.127,128 This re-sults in a baroreceptor mediated reflex increase in heart rate (HR) and total peripheral (vascular) resistance (TPR).129,130 A reduction in central blood volume evoked by LBNP or posture changes (e.g. standing up) elicits a wide range of HR and blood pressure (varia-tion) responses among healthy individuals.131-134 Studies addressing the cardiovascular responses and specifically tolerance to a reduction in central blood volume evoked by LBNP reported that tolerance time and cardiovascular responses were reproducible in a test-retest condition at varying time intervals.135-139 From observations in our lab we found considerable differences in cardiovascular response patterns to LBNP between subjects. Only a few studies have specifically addressed individual response patterns from rest to maximal LBNP into some detail.122,140 We questioned whether the large variation in response patterns between subjects to submaximal LBNP coincides with consistent and reproducible responses in an individual subject. Therefore, the present study was designed to evaluate the individual cardiovascular reflex responses to sym-pathetic stimulation and their robustness. To that purpose, we determined intra-subject reproducibility of responses by central hypovolemia evoked by LBNP in young healthy subjects over short (5 minutes) and longer (1 week) time intervals.

METHODS

Subjects

10 healthy, non-smoking Caucasian subjects (5 females), with normal physical fitness and with a median (range) age of 22 (19-26) years, height of 174 (166-177) cm and weight 69 (55-77) kg participated in this study. Exclusion criteria included a medical history of cardio- and/or cerebrovascular disease, neurological disorders, diabetes mellitus, regu-lar fainting and the use of medication (either prescription or non-prescription). Subjects abstained from heavy exercise and caffeinated beverages 5h prior to the experiment. Phase of menstrual cycle in female subjects was not accounted for. Experiments were conducted in a temperature-controlled laboratory (20-22°C) at the same time of the day

42 Chapter 4.1

(12 to 4 pm) to avoid potential effects of circadian rhythm on the study outcomes. The institutional Medical Ethics Committee approved the protocol and written informed consent was obtained.

Experimental protocol

Measurements were performed in a quiet room with subjects in the supine position. After instrumentation, the lower body was positioned inside the LBNP box (Dr. Kaiser Medizintechnik, Bad Hersfeld, Germany) and sealed at the level of the iliac crest.122 The study protocol included five minutes of rest, followed by three 5 minute trials of LBNP at -50 mmHg separated by 5 minutes of rest. During the experiment, subjects were instructed to breathe normally and to avoid body movement. Reproducibility of cardio-vascular responses was evaluated by repeating the protocol 7 days later.

The LBNP box was equipped with a saddle to avoid leg muscle pump activation during the application of sub-atmospheric pressure. The pressure inside the box was manually controlled and established within 10-20 seconds. LBNP was terminated upon request or in case of (pre-)syncopal symptoms which were determined by one or more of the following criteria: systolic arterial pressure (SAP) below 80 mmHg, or rapid drop (SAP by ≥20 mmHg·min-1, diastolic (DAP) by ≥10 mmHg·min-1), drop in HR by ≥15 bpm, and/or sweating, light-headedness, nausea, blurred vision or skin pallor.

Measurements and Analysis

Continuous arterial pressure (AP) was measured non-invasively was measured non-inva-sively by a volume clamp method using finger plethysmography (Nexfin, Edwards Life-sciences BMEYE, the Netherlands). HR was expressed as the pulse pressure interval. Left ventricular stroke volume (SV) and cardiac output (CO; SV multiplied by instantaneous HR) were measured by a pulse contour method (Nexfin CO-trek, Edwards Lifesciences BMEYE, Amsterdam, the Netherlands) which is validated against thermodilution esti-mates of CO.39,99 TPR was defined as the ratio of mean arterial pressure (MAP) and CO. All recorded signals were visually inspected for artifacts and analyzed offline (Matlab R2007b, Mathworks Inc. MA, USA).

Time-domain cardiac BRS was analyzed using the cross-correlation method.141,142 First, beat-to-beat SAP and inter-beat interval (IBI) were fitted with cubic spline functions and resampled at 1s intervals. The cross-correlation between 10s series of resampled SAP and IBI signals were computed for various delays (τ) in IBI of 0-5s. The delay between SAP and IBI with the highest cross-correlation was selected if the correlation was significant at p

Cardiovascular response patterns to central hypovolemia 43

4

Statistical Analysis

Variables were presented as mean ± SD. One-way repeated measures ANOVA’s were used to compare the last minute of rest and the last minute of LBNP across three con-secutive trials on day 0 and day 7, followed by Holm-Sidak post hoc tests. The responses were grouped together per day when there were no differences in baseline and LBNP responses across trials. The effect of LBNP on the measured parameters was analyzed with a paired two-tailed Student’s T-test (Sigmaplot 11.0, Systat Software Inc., USA) comparing the last minute of rest with the last minute of LBNP (average of three trials). Intraclass correlation coefficients (ICC) and coefficients of variation (CV) were calculated (IBM SPSS statistics 20, IBM corporation, USA) to assess intra-subject reproducibility. Intra-subject reproducibility was evaluated across LBNP trials (5 minute periods; three trials on day 0) and sessions (one week period; average response of three trials between day 0 and day 7). One trial was defined as the last 3 minutes of rest, 5 minutes of LBNP and 2 minutes of recovery. ICC and CV values were calculated per trial time point and then averaged for all time points. No universal standard exists for classifying ICC and tests of statistical significance of reproducibility measures are of little practical utility.143 Therefore, reproducibility was defined as poor if ICC

44 Chapter 4.1

Tabl

e 4.

1.1.

Bas

elin

e ch

arac

teris

tics

and

hem

odyn

amic

resp

onse

to L

BNP

for t

hree

con

secu

tive

tria

ls a

t day

0 a

nd d

ay 7

.

Tria

l 1Tr

ial 2

 Tr

ial 3

 Tr

ial 1

vs.

2 v

s. 3

(p-v

alue

s)Re

stLB

NP

Rest

LBN

PRe

stLB

NP

Rest

LBN

P

SAP

day

012

0±

1510

9±

14*

121

±14

107

±13

*12

3±

1210

9±

11*

0.67

00.

237

(mm

Hg)

day

711

6±

1010

2±

5*11

5±

910

4±

4*11

8±

810

4±

4*0.

194

0.24

5

  

  

  

  

  

  

  

  

  

  

  

DA

Pda

y 0

69±

1271

±10

69±

1069

±9

70±

971

±7

0.77

20.

096

(mm

Hg)

day

767

±6

67±

267

±5

68±

468

±5

68±

30.

191

0.14

5

  

  

  

  

  

  

  

  

  

  

  

MA

Pda

y 0

88±

1485

±12

*89

±12

83±

10*

89±

1085

±8*

0.77

30.

124

(mm

Hg)

day

785

±7

80±

3*84

±6

81±

4*86

±6

81±

3*0.

124

0.12

4

  

  

  

  

  

  

  

  

  

  

  

HR

day

066

±13

90±

10*

64±

1285

±12

*62

±13

87±

10*

0.06

70.

432

(bea

ts·m

in-1

)da

y 7

62±

1383

±14

*60

±12

83±

15*

60±

1282

±16

*0.

137

0.49

6

  

  

  

  

  

  

  

  

  

  

  

SVda

y 0

112

±16

78±

16*

113

±17

78±

17*

114

±17

79±

16*

0.23

30.

810

(mL)

day

711

1±

1576

±13

*10

9±

1273

±12

*11

0±

1275

±11

*0.

476

0.32

6

  

  

  

  

  

  

  

  

  

  

  

COda

y 0

7.1

±1.

36.

8±

1.2*

7.0

±1

6.5

±1.

1*6.

9±

1.1

6.6

±1.

1*0.

435

0.09

8

(L·m

in-1

)da

y 7

6.4

±1.

16.

3±

0.9

6.5

±1.

16.

1±

0.7*

6.5

±1.

26.

1±

0.9*

0.97

60.

178

  

  

  

  

  

  

  

  

  

  

  

TPR

day

010

00±

152

1048

±20

410

23±

180

1033

±17

210

39±

164

1026

±17

20.

081

0.15

6

(dyn

·s·c

m-5

)da

y 7

1017

±17

610

42±

140

1046

±18

410

86±

142

1045

±18

010

78±

156

0.49

70.

510

  

  

  

  

  

  

  

  

  

  

  

BRS

day

017

±8

8±

4*19

±8

8±

2*17

±7

8±

3*0.

089

0.76

5

(ms·

mm

Hg-

1 )da

y 7

17±

610

±3*

20±

710

±3*

20±

1010

±4*

0.41

90.

877

  

  

  

  

  

  

  

  

  

  

  

Dat

a ar

e pr

esen

ted

as m

ean

± SD

. AP,

arte

rial p

ress

ure

(Sys

tolic

, Dia

stol

ic a

nd M

ean,

); H

R, h

eart

rate

; SV,

stro

ke v

olum

e; C

O, c

ardi

ac o

utpu

t; TP

R, to

tal p

erip

hera

l res

ista

nce;

BR

S, b

aror

eflex

sen

sitiv

ity. R

est a

nd L

BNP

resp

onse

s w

ere

com

pare

d be

twee

n Tr

ial 1

, Tria

l 2 a

nd T

rial 3

. *p<

0.05

vs.

rest

.

Cardiovascular response patterns to central hypovolemia 45

4

increased with a fall in systolic (SAP; -11±6%, p

46 Chapter 4.1

Individual responses

The cardiovascular compensatory response to LBNP differed between subjects (Figure 4.1.1, grey lines). Figure 4.1.2 gives the distribution of maximal changes in HR, SV and TPR . Subsequently, subjects were dichotomized into two equal-sized groups (A and B) based on the median change in HR (Figure 4.1.3 and Table 4.1.2). Group A showed a prominent increase in HR (54±14%, p=0.001) with no significant change in TPR (p=0.25) vs. group B demonstrating a moderate increase in HR (21±9%, p=0.004) and a rise in TPR (7±3%, p=0.017). The response pattern of group A vs. B coincided with a larger decrease in SV (-37±8 vs. -25±6%, p=0.026) and BRS (-54±14 vs. -34±15%, p=0.036) see Figure 4.1.4 and Figure 4.1.5. Group A vs. B subjects did not differ in sex, age or body mass index (BMI), however lower resting HR (57±8 vs. 71±12 bpm, p=0.047) and higher resting BRS (22±8 vs. 11±3 ms·mmHg-1, p=0.049) were found in group A. Figure 4.1.6 shows BRS results of one representative subject.

∆SV

(%)

∆HR

(%)

∆TPR

(%)

Day 0Day 7

subject #: 1 2 3 4 5 6 7 8 9 10

Figure 4.1.2. Distribution of maximal change in SV, HR and TPR in response to LBNP at day 0 (black bars) and day 7 (grey bars). For abbreviations see Figure 4.1.1

Cardiovascular response patterns to central hypovolemia 47

4

Δ T

PR (%

)Δ

TPR

(%)

Δ H

R (%

)Δ

HR

(%)

ΔSV

(%)

ΔSV

(%)

ΔM

AP

(%)

ΔM

AP

(%)

B

ALBNP LBNP LBNP LBNP LBNP

LBNP LBNP LBNP LBNP LBNP

#1 #4 #5 #7 #10

#2 #3 #6 #8 #9

Time (min) Time (min) Time (min) Time (min) Time (min)

Time (min) Time (min) Time (min) Time (min) Time (min)

Figure 4.1.3. Individual responses to LBNP for day 0 (solid line) and day 7 (dashed line). Data was normal-ized to the last two minutes of rest. Group A (upper panel) showed a predominant increase in HR whereas group B (lower panel) responded by a consistent increase in TPR with smaller change in HR. For abbrevia-tions see Figure 4.1.1.

48 Chapter 4.1

Table 4.1.2. Baseline characteristics and hemodynamic response to LBNP for group A and group B.

  Group A (n=5) Group B (n=5)

Rest LBNP Rest LBNP

Sex (M/F)   2 / 3       3 / 2      

Age (years)   21 ± 2       24 ± 2      

BMI (kg·m-2)   23 ± 3       23 ± 3      

                           

SAP (mmHg) day 0 123 ± 10 106 ± 14* 120 ± 17 110 ± 13*

  day 7 117 ± 8 102 ± 4* 116 ± 9 104 ± 4*

                           

DAP (mmHg) day 0 67 ± 7 68 ± 8 71 ± 14 71 ± 12

  day 7 68 ± 4 69 ± 2 67 ± 6 67 ± 6

                           

MAP (mmHg) day 0 87 ± 7 82 ± 10 88 ± 16 87 ± 14

  day 7 86 ± 5 81 ± 2 84 ± 7 81 ± 5

                           

HR (beats·min-1) day 0 57 ± 8 87 ± 13* 71 ± 12† 86 ± 12*

  day 7 55 ± 7 88 ± 14* 67 ± 13† 81 ± 17*

                           

SV (mL) day 0 121 ± 13 77 ± 17* 105 ± 17 79 ± 16*

  day 7 113 ± 16 70 ± 13* 108 ± 10 76 ± 13*

                           

CO (L·min-1) day 0 6.8 ± 1.1 6.6 ± 1.3 7.3 ± 1.2 6.6 ± 1.0*

  day 7 6.1 ± 1.1 6.0 ± 0.8 6.8 ± 1.1 6.0 ± 0.8*

                           

TPR (dyn·s·cm-5) day 0 1042 ± 80 1015 ± 89 994 ± 213 1062 ± 261*

  day 7 1149 ± 160 1111 ± 105 977 ± 187 1100 ± 179*

                           

BRS (ms·mmHg-1) day 0 22 ± 8 10 ± 4* 11 ± 4† 9 ± 4

  day 7 22 ± 7 10 ± 3* 13 ± 3† 10 ± 3

Data are presented as mean ± SD. BMI, body mass index; For other abbreviations see Table 4.1.1. *p

Cardiovascular response patterns to central hypovolemia 49

4

∆HR

(%)

∆MA

P (%

)∆C

O (%

)

∆BRS

(%)

∆SV

(%)

∆TPR

(%)

Day 0 Day 7 Day 0 Day 7

††

†

†

†

† †

*

*

*

*

*

*

*

*

*

*

* *

*

*

Figure 4.1.4. Percentage change from rest to LBNP for group A (white bars) and B (grey bars). BRS, barore-flex sensitivity; for other abbreviations see Figure 4.1.1. *p

50 Chapter 4.1

Rest LBNP Rest LBNP Rest LBNP Rest LBNP

Day 0 Day 7

p=0.01p=0.01

p=0.34 p=0.10

Group A Group B Group A Group BBR

S (m

s·m

mH

g-1 )

Figure 4.1.5. Individual (grey) and average (black) BRS responses to LBNP for group A (filled symbols) and group B (open symbols) at day 0 and day 7. No BRS values could be determined for subject 2 (group B) at day 0 due to signal artifacts. BRS, baroreflex sensitivity.

LBNP

2 5 10 15Time (min)

BRS

(ms·

mm

Hg-

1 )

Figure 4.1.6. A representative example (subject 5) of BRS computations in response to LBNP. Each dot rep-resents a BRS result, drawn horizontal dashed lines represent period averages and the solid line represents moving averages. BRS, baroreflex sensitivity.

Cardiovascular response patterns to central hypovolemia 51

4

Reproducibility

Figure 4.1.3 (dashed lines) visually demonstrated that the majority of individuals responded similarly one week later. Intra-subject reproducibility of the cardiovascular response to LBNP according to the ICC and the CV is given in Table 4.1.3. HR, SV, CO and TRP demonstrated good to excellent intra-subject reproducibility (ICC≥0.61 and CV

52 Chapter 4.1

sympathetic influence on both HR and TPR which occurs more slowly due to a longer time-constant.45,148 Recently, a more balanced model of sympatho-vagal control repre-senting a continuous interplay between vagal and sympathetic modulation of HR has been proposed without clear on/off thresholds.149 In humans, the relative contribution of arterial vs. cardiopulmonary baroreflex involvement cannot be ascertained; for instance, even mild LBNP reduces aortic dimensions contesting selective low pressure area receptor activation.150 The present study demonstrated two qualitatively differ-ent cardiovascular reflex patterns in response to a similar degree of exposure of LBNP varying from a predominant effect on HR to a consistent increase in TPR with a smaller change in HR. These differential responses between HR and TPR coincided with a larger decline in SV and CO. This conforms to data from Fu et al.151 who raised the hypothesis that decreases in pulse amplitude (a function of SV) may preferentially influence the vagal component of the baroreflex, whereas flow in baroreceptive arteries (a function of CO) dominates the sympathetic component. In addition, our data show that differ-ent reflex responses coincided with resting values of dynamic baroreflex control and HR. Together with an insubstantial increase in TPR in subjects with higher resting BRS, this alludes to differential cardiovascular reflex control in response to simulated central hypovolemia. The observed variance in responses was reproducible for the individual subjects suggesting an individually determined autonomic reflex response. We consider that humans may present with an identical sigmoidal baroreflex relationship as esti-mated by neck cuff suction-pressure plots but nevertheless may deliver different BRS values depending on the operating set point.152 Yielding multiple BRS values per minute rather than a single value reduces the risk of inaccurate reflections of baroreflex sensitiv-ity which we consider a strength of the cross-correlation method used in the present study (see Figure 4.1.6).

Generally, subjects with high tolerance to central hypovolemia display signs and symptoms of greater sympathetic activition, e.g. higher HR and peripheral vasoconstric-tion with elevated neurohormonal activation.153-156 Specifically, Convertino et al. demon-strated greater increases in muscle sympathetic nerve activity (MSNA) with an elevated total peripheral resistance and also an elevated HR, higher baseline cardiovagal BRS, and greater reductions in cardiovagal BRS in individuals with high tolerance to LBNP.155 The magnitude of the autonomic responses to LBNP has been defined as the HR and vasoconstrictor ‘reserve’ according to the concept that a greater physiological reserve capacity for tachycardia and vasoconstriction related to high tolerance to central hy-povolemia is associated with greater reserves for sympathoexcitation and cardiac vagal withdrawal.155,157,158 We consider that the present study addressed the cardiovascular responses to sub-maximal LBNP, i.e. beyond the ‘compensatory reserve’.159 It appears as though subjects in group B may have a lower “HR reserve” and “BRS reserve due to a higher resting HR and lower resting BRS. We did not determine the reproducibility

Cardiovascular response patterns to central hypovolemia 53

4

of responses to maximal LBNP leaving the question how these responses might affect tolerance to maximal LBNP.

Of interest, in a retrospective study in a seemingly homogenous population with high tolerance to central hypovolemia a higher HR, TPR, SNA, and BRS appeared not associ-ated with greater tolerance to a reduced central blood volume.153 This suggests that the autonomic make-up determines whether an individual relies on cardiac filling and vagal withdrawal to defend arterial pressure, or on sympathoexcitation to elevate HR and TPR 153. These findings were interpreted as to demonstrate the existence of subpopulations with analogous physiological abilities though diverse contributions of cardiovascular compensatory mechanisms to central blood volume depletion, and the present data conform to that concept. The recent observation that those subjects with high tolerance to central hypovolemia appear to be protected by maintained frontal lobe cortical tissue oxygen saturation links cerebral oxygen supply directly to brain function.160

Cardiovascular control is subjected to considerable environmental influences includ-ing level of deconditioning, hydration status and disease.42,161-163 Resting HR is a deter-minant of BRS,164 and physiological factors, particularly age and sex, have significant impact on BRS in healthy subjects.164,165

Age and sex

With aging the magnitude of the reflex increase in HR declines with BP maintained by a more substantial increase in forearm and total peripheral resistance.166 Regarding a gender effect on the response to sympathetic stimulation, the available data on vascular responsiveness as well as on changes in HR and TPR during passive head-up tilt and/or LBNP is not uniform.167,168 A gender effect on the vascular but not the HR response to LBNP has been reported with lesser increase in TPR in women.168 This conforms to recent data showing that a strong association between MSNA and TPR is expressed in young males only.169 In contrast, Shoemaker et al. demonstrated gender-related differ-ences in HR and MSNA burst frequency but not in TPR in response to head-up tilt.167 An influence of age does not apply to the participants in this study. We however cannot exclude that the sample size may have obscured a possible sex effect but we do consider that sex differences in baroreflex BP control during carotid hypotension have not been established.170

Physical fitness

An influence of differences in physical fitness level could be considered. We did not quantify maximal oxygen uptake but included subjects with normal physical fitness but without specific sports training as documented by a questionnaire. Exercise training does not affect vagal–cardiac control or cardiovagal BRS in young and middle-aged healthy subjects.171,172 Accordingly, the reduction in resting HR by exercise training in

54 Chapter 4.1

young and middle-aged adults is limited (~-5 bpm),172,173 and does not account for the more than twofold difference in resting HR between the two groups. The consistency of distinct cardiovascular response patterns as well as BRS with both autonomic response patterns maintaining blood pressure within the timeframe of the simulated central hypovolemia rather suggests an individually programmed strategy of reflex responses to sympathetic stimulation.

Genetic vs. environmental

MSNA is considered as the primary index of sympathetic activity in humans,174,175 and is characterized by large inter-individual differences but robust intra-individual repro-ducibility over many years which is in support of a genetic component.175 Kardos et al. showed that only half of the variance in BRS is attributable to simple anthropometric variables and common risk factors like smoking and alcohol consumption.164 They sug-gested that the remaining variability reflects the subjects’ different genetic background. In contrast, the significant degree of variance in cardiovascular responses to head-up tilt in identical twins suggests environmental respectively epigenetic factors as important contributing factors,176 which were not addressed in the present study.

Methodological Considerations

Several methodological considerations pertain to our data inherent to the study design. We do not directly measure individual differences in actual magnitude of the evoked fluid shift with LBNP. Inter-individual differences in SV are strongly correlated to central venous pressure,177 so may reflect these fluid shifts, with subsequent consequences for baroreceptor reflex responses.178 The inter-individual differences in central hypovolemia expressed as the LBNP-induced reduction in SV, seem a confounding factor related to the differential cardiovascular reflex responses between groups. The degree of central hypovolemia during LBNP is considered as the primary mover of cardiovascular re-sponses. Generally, differences in translocated volume between subjects for the same box pressure are an intrinsic limitation of LBNP. Surprisingly few studies have addressed study paradigms enabling individualization LBNP-induced central blood volume shifts. We propose leg (blood) volume179 as a controller of input to a sub-atmospheric pressure feedback loop. The finding that resting cardiovascular variables (including SV, HR and MAP) were similar between consecutive trials, suggest that 5 min of rest is sufficient to reverse LBNP-induced fluid shifts.

Intra-subject reproducibility has been assessed in the present study by two different methods of reliability testing. We found that the results of both tests were contradictory for arterial blood pressure responses: poor vs. good reproducibility. Using ICCs can lead to inaccurately low reliability measurements as it is highly sensitive for the spread of the data whereas typical error (e.g. CV) is not.180 Several researchers reported (highly)

Cardiovascular response patterns to central hypovolemia 55

4

reproducible blood pressure responses to repeated LBNP trials,135,138,139 suggesting that CV is a more robust and accurate measure of reliability.

A sudden blood pressure drop observed in two subjects (one from group A and one from group B) during the last LBNP trial of that session emphasizes that we do not know whether blood pressure control would have been maintained beyond the applied timeframe of simulated central hypovolemia. The finding that a decline in blood pres-sure is more likely to occur with accumulating exposure to LBNP conforms to previous research.181,182

Sympathetic stimulation is considered to enhance specifically the inotropic condition and lusitropic properties of the healthy heart.183 We acknowledge that in the physiologi-cal laboratory as well as in a clinical environment changes in intrinsic cardiac muscle properties during sympathetic activation usually go by unnoticed in part or in whole.

We did not evaluate MSNA respectively the sympathetic BRS response, which might provide further insight into the dichotomy.

This study does not answer the question whether or not the distinct cardiovascular response patterns would have remained consistent over months or years. Convertino re-ported that re-testing after one year delivered comparable cardiovascular responses to LBNP.139 With aging the interaction between neural and hemodynamic factors changes with different sex effects which is expected to modify the cardiovascular response to LBNP on the long run.169

Summary

In summary, the present study demonstrated distinct and reproducible cardiovascular response patterns to sympathetic stimulation by central hypovolemia in young healthy adults. Differences in resting HR and BRS between subjects suggest individually pro-grammed reflex strategies of autonomic blood pressure control which may contribute to the variance observed in cardiovascular reflex responses to central hypovolemia. The mechanisms responsible for this phenomenon and the extent to which they operate in other groups of subjects deserve attention.

4.2The cerebrovascular response to lower body negative pressure versus head-up tilt

Anne-Sophie G.T. Bronzwaer*Jasper Verbree*Wim J. StokMat J.A.P. DaemenMark A. van BuchemMatthias J.P. van OschJohannes J. van Lieshout

*shared first authorship

Journal of Applied Physiology 2017 Jan 12:jap.00797.2016 [Epub ahead of print]

58 Chapter 4.2

ABSTRACT

Lower body negative pressure (LBNP) has been proposed as a MRI compatible surrogate for orthostatic stress. Although the effects of LBNP on cerebral hemodynamic behavior have been considered to reflect those of orthostatic stress, a direct comparison with actual orthostasis is lacking. We assessed the effects of LBNP (-50 mmHg) versus head-up tilt (HUT; at 70°) in 10 healthy subjects (5 female) on transcranial Doppler determined cerebral blood flow velocity (CBFv) in the middle cerebral artery and cerebral perfusion pressure (CPP) as estimated from the blood pressure signal (finger plethysmography). CPP was maintained during LBNP but decreased after 2 min in response to HUT lead-ing to a ~15% difference in CPP between LBNP and HUT (p≤0.020). Mean CBFv initially decreased similarly in response to LBNP and for HUT but from minute 3 on, the decline became ~50% smaller (p≤0.029) during LBNP. The reduction in end-tidal PCO2 was comparable but with an earlier return towards baseline values in response to LBNP but not during HUT (p=0.008). We consider the larger decrease in CBFv during HUT versus LBNP attributable to the pronounced reduction in end-tidal PCO2 and to gravitational influences on CPP and this should be taken into account when applying LBNP as an MRI compatible orthostatic stress modality.

The cerebrovascular response to LBNP versus head-up tilt 59

4

InTRODuCTIOn

Orthostatic stress elicits a gravitational shift of blood away from the upper towards the lower parts of the body. The postural decline in arterial pressure at the level of the brain is subsequently being counteracted by a cerebrovascular autoregulatory adaptive response. Nevertheless, steady-state cerebral blood flow velocity (CBFv) as determined by transcranial Doppler (TCD) and cortical oxygenation decrease when changing from the supine to the upright body position.34,184 If the postural reduction in cerebral perfu-sion is substantial, the mismatch between cerebral oxygen demand vs. supply results in a transient loss of consciousness.12,185 Passive head-up tilt (HUT) is a commonly used orthostatic stress test, for example employed for the evaluation of patients with syn-cope of unknown origin.186,187 Cerebral hemodynamics are then often monitored by TCD which measures CBFv in a large brain-feeding artery as a proxy for cerebral blood flow (CBF).188,189

Advances in functional MRI techniques have enabled the measurement of not only cerebral blood flow but also volume and oxygenation at the brain tissue level which would provide a more complete insight into the cerebral hemodynamic response to orthostatic stress.112 Evaluating the cerebrovascular orthostatic response with MRI is, however, challenging because either active standing or passive HUT is incompatible with the horizontal and static setup of an MRI scanner. Lower body negative pressure (LBNP) has the potential to serve as an MRI compatible surrogate of orthostatic stress as it reduces central blood volume while the subject remains in the supine body posi-tion.190,191 Although the effects of LBNP on cerebral hemodynamic behavior have been considered to reflect those elicited by orthostatic stress,192 a direct comparison with actual orthostasis is not available. The present study aimed to provide a formal compari-son of the systemic and cerebrovascular responses to LBNP (-50 mmHg) versus HUT (70°) in young healthy volunteers.

METHODS

Subjects

Ten healthy, non-smoking subjects (5 female), with a median (range) age of 22 (19-26) years, height of 174 (166-177) cm and weight 69 (55-77) kg with no history of fainting and/or cardiac arrhythmia nor taking cardiovascular medication participated in this study. Phase of menstrual cycle was not accounted for. Subjects abstained from heavy exercise and caffeinated beverages for at least 5 h prior to the experiments, that were conducted in a temperature-controlled laboratory (20-22°C) at the same time of the day

60 Chapter 4.2

(12 to 4 pm). The institutional Medical Ethics Committee approved the study protocol and both oral and written informed consent was obtained from all participants.

Experimental protocol

Measurements were performed with subjects in the supine position on a manually operated tilt table (Dr. Kaiser Medizintechnik, Bad Hersfeld, Germany). Following instru-mentation, the lower body was positioned inside the LBNP box and sealed at the level of the iliac crest. The study protocol consisted of a 5 min baseline period followed by either a 5-min period of LBNP (-50 mmHg) or a 5 min period of HUT (70°). All partici-pants underwent both LBNP and HUT. As these experiments were part of a larger study program, LBNP and HUT were applied in a fixed order starting with LBNP. To prevent a downward shift of the body into the LBNP box disrupting the tight air sealing with loss of sub-atmospheric pressure, the LBNP box was supplied with a saddle. During HUT, subjects were instructed not to tense their legs while their body was restraint to the table by straps. The transition to -50 mmHg LBNP or 70° HUT was established within 10 s. The safety guidelines of the protocol determined that LBNP and HUT would both be terminated in case of (pre-)syncopal symptoms including sweating, light-headedness, nausea, blurred vision and/or signs meeting one or more of the following criteria: sys-tolic arterial pressure (SAP) below 80 mmHg, or rapid drop (SAP by ≥20 mmHg·min-1, diastolic arterial pressure (DAP) by ≥10 mmHg·min-1), and drop in HR by ≥15 bpm.

Measurements

Continuous blood pressure (BP) was measured non-invasively by finger plethysmogra-phy with the cuff placed around the middle phalanx of the non-dominant hand placed at heart level (Nexfin, Edwards Lifesciences BMEYE, the Netherlands). Left ventricular stroke volume (SV) was determined by a pulse contour method (Nexfin CO-trek, Edwards Lifesciences BMEYE, Amsterdam, the Netherlands) that has been validated against thermodilution measurements.99 Cardiac output (CO) was calculated as SV times HR and total peripheral resistance (TPR) was the ratio between mean arterial pressure (MAP) and CO. End-tidal CO2 partial pressure (PetCO2) was monitored through a nasal cannula connected to a capnograph (Datex Normocap 200, Helsinki, Finland). Changes in CBFv were followed in the proximal segment of the middle cerebral artery (MCA) by means of TCD (DWL Multidop X4, Sipplingen, Germany). The left MCA was insonated through the temporal window just above the zygomatic arch at a depth of 40–60 mm with a pulsed 2 MHz probe. After signal optimization, the probe was immobilized by a head-band.

Data Analysis

All signals were inspected and artefacts removed. Blood pressure at the level of the MCA (BPbrain ) was estimated by subtracting the hydrostatic column between the level of the

The cerebrovascular response to LBNP versus head-up tilt 61

4

left atrium and the insonation point of the TCD probe (present during HUT only), from BP at heart level (BPhea