Role of the autonomic nervous system in vasoconstriction responses to mental stress in sickle cell disease: a bioengineering perspective

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Request accessible transcript

Transcript (if available)

Content ROLE OF THE AUTONOMIC NERVOUS SYSTEM IN
VASOCONSTRICTION RESPONSES TO MENTAL STRESS
IN SICKLE CELL DISEASE: A BIOENGINEERING
PERSPECTIVE

by
Wanwara Thuptimdang

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
BIOMEDICAL ENGINEERING

August 2021

Copyright 2021 Wanwara Thuptimdang

ii
Acknowledgements
The PhD journey has finally come to an end. I would like to express sincere
gratitude to all the people who provided a great deal of support and help throughout this
journey. I would first like to thank my supervisor, Professor Michael Khoo, who gave me
the opportunity to be part of this journey. Your insightful feedback encouraged me to
sharpen my thinking and perspectives. Your patience and understandings allowed me to
walk at my own pace. Next, I would like to thank Dr. Thomas Coates who gave me the
opportunity to be part of our research. I am deeply appreciated that you let me took a
major role in developing the software for our group. Your passion has always been a
great inspiration to me. I also want to thank the committee members - Dr.John Wood, Dr.
Jon Detterich and Dr.Brent Liu for generously offering your time and support.
This work would not have been possible without dedicated people who took part
in this research. I would like to thank Payal Shah who took a primary role on collecting
the experimental data used in this work. I would like to thank John Sunwoo and Roberta
Kato for troubleshooting problems with our devices. I also want to thank Saranya
Veluswamy, Christopher Denton and Sarah Martin for providing valuable feedbacks on
this work. Lastly, I would like to thank Nathan Smith for helping me with all the paper
works I needed during my time at CHLA hospital.
I am fortunate to have supportive friends and family. I would like to thank my fellow
K-lab members: Alison, John, Kelby, Leo, Yunhua, Winston and Sang for your support
and valuable feedbacks. Thank you for making my life at the lab enjoyable. I also
appreciated the lunch time I spent with Sahar, Gunce and Gene. Thank you for sharing

iii
memories of being a graduate student together. I would like to thank Silvie and Iris for
making my time at CHLA less lonely. I used to be in the lab alone, coding. Thank you
for spending your coffee breaks with me. I am also grateful to know these amazing Thai
friends: Jelly, Bug, Wa, Nam, Beer, Toy, Gor, Title. All of you have inspired me to
overcome my negativity, be strong and be a better person.
I would like to thank my girlfriend and my family. I want to thank Phon for taking
care of me and cheering me up through the ups and downs of PhD. You always believe
in me more than I do. This always gives me strength to strive. I would like to thank my
brother for supporting me, giving good advice and encouraging me to get through tough
times. I used to have you proof-read my essays, corrected my grammars. I really
appreciated all your help. I could not have asked for a better brother. Lastly, I would like
to thank my parents – Dad and Mom for believing in me and supporting me in every
decision I made. Thank you for always listening to me, and trying your best to understand
what it is meant to be a PhD student. Without your support, I would not have come this
far.
Sincerely,
To-Ey

iv
Table of Contents
Acknowledgements ............................................................................................................. ii
List of Tables ...................................................................................................................... vi
List of Figures .................................................................................................................... vii
Abstract ................................................................................................................................ x
Chapter 1 Introduction ........................................................................................................ 1
Chapter 2 Background ....................................................................................................... 4
The Autonomic Nervous System .................................................................................... 4
The sympathetic pathway ........................................................................................... 5
The parasympathetic pathway .................................................................................... 6
Autonomic control of the heart ....................................................................................... 7
Sympathetic control of blood vessels ............................................................................. 9
Baroreceptor Reflex ...................................................................................................... 12
Anatomy and activation ............................................................................................. 13
Sickle cell disease......................................................................................................... 14
Pathology and pathophysiology ................................................................................ 15
Vaso-occlusive crisis (VOC) ..................................................................................... 17
Autonomic nervous system involvement in VOC ..................................................... 18
Chapter 3 Experimental methods and data processing .................................................. 22
Measurements and Data Processing ........................................................................... 22
Photoplethysmography ............................................................................................. 23
Continuous noninvasive blood pressure monitoring (Nexfin) .................................. 24
Pulse amplitude of PPG and blood pressure are not necessarily correlated .......... 27
Experimental protocols ................................................................................................. 31
Chapter 4 The role of baseline baroreflex on vasoconstriction to mental stress in SCD 33
Method .......................................................................................................................... 34
Quantification of vasoconstriction................................................................................. 34
Characterization of baseline autonomic function ......................................................... 34
(a) Baroreflex sensitivities using the “sequence” technique: ................................. 35
(b) Baroreflex sensitivities using the “spectral” technique: .................................... 36
Statistical Analysis ........................................................................................................ 38
Results .......................................................................................................................... 40
Subject characteristics .............................................................................................. 40
Baseline cardiovascular and autonomic indices ...................................................... 42
Correlation analysis between baseline autonomic indices and the magnitude of
vasoconstriction......................................................................................................... 45
Effect of mental stress on vasoconstriction after adjusting for baseline vascular
baroreflex sensitivity.................................................................................................. 48
Discussion ..................................................................................................................... 50

v
Role of the baroreflexes in modulating the cardiovascular responses to mental
stress ......................................................................................................................... 50
Reduced baseline vascular baroreflex sensitivity in SCD ....................................... 51
Variability of vasoconstriction responses resulting from differences in source of
autonomic stimulation ............................................................................................... 52
Estimation of vascular baroreflex sensitivity ............................................................ 53
Conclusion .................................................................................................................... 55
Chapter 5 The relationships between autonomic indices and patterns of
vasoconstriction during mental stress .............................................................................. 56
Method .......................................................................................................................... 59
Measuring sustained vasoconstriction ..................................................................... 59
Assessment of autonomic responses ....................................................................... 63
Statistical Analysis .................................................................................................... 66
Results .......................................................................................................................... 67
Block entropy of PPGa time-series........................................................................... 67
Relationships between block entropy and mental stress ......................................... 71
Effects of mental stress on the autonomic nervous system .................................... 74
Relationships between sustained vasoconstriction and the ANS measures .......... 79
Discussion ..................................................................................................................... 85
Conclusion .................................................................................................................... 87
Chapter 6 Summary, limitations and future directions .................................................... 88
Summary and significant new findings ......................................................................... 88
Limitations ..................................................................................................................... 88
Future directions ........................................................................................................... 91
References........................................................................................................................ 92

vi
List of Tables
Table 4-1 Subject characteristics and hematological parameters .................................. 41
Table 5-1 Mean cardiovascular variables during cognitive tasks ................................... 76
Table 5-2 Comparison of baroreflex indices between controls and SCD subjects during
cognitive tasks .................................................................................................................. 77
Table 5-3 Comparison of spectral indices between controls and SCD subjects during
cognitive tasks .................................................................................................................. 78

vii
List of Figures
Figure 2-1: Simplified sketch of the autonomic pathways to the heart and blood vessels.
The black lines represent the preganglionic sympathetic fibers. The black dashed lines
represent the postganglionic sympathetic fibers. Sympathetic stimulation increases
heart rate and causes vasoconstriction. Parasympathetic stimulation slows heart rate.
Adopted from (Klabunde, 2011). ........................................................................................ 6
Figure 2-2: Anatomy of the heart. The heart consists of four chambers: right, left atrium
and right, left ventricle. Blood flows from the right atrium into the right ventricle and
enters the pulmonary circulation. Blood returning from the lungs enters the left atrium,
flows into the left ventricle which ejects blood out of the heart. SA, the sinoatrial node;
AV, the atrioventricular node. ............................................................................................. 8
Figure 2-3: Schematic diagram of hypothetical cases of parasympathetic tone. From A)
to B), increase in parasympathetic tone prolongs the start of cardiac cycle; thus,
decreasing heart rate. ........................................................................................................ 9
Figure 2-4: Schematic diagram of cardiac and vascular baroreflexes........................... 12
Figure 2-5: Hemoglobin S polymerizes after deoxygenation ......................................... 14
Figure 2-6: Sickle erythrocytes. Peripheral blood smear from a patient with SCD
obtained during a routine clinic visit. The smear shows classical sickle-shaped (arrows)
and various other misshaped erythrocytes (arrowheads). Republished with permission
of American Society for Clinical Investigation, from (Frenette and Atweh, 2007) .......... 16
Figure 3-1: Example of PPG and PPGa signals. The PPG only contains the AC
component as the DC component is filtered out by the measuring device. ................... 24
Figure 3-2: Volume clamp principle observing the dynamic unloading of the arterial wall
by using an infrared optical plethysmograph built in the cuff. Republished with
permission of Springer Nature from (Settels, 2014). ....................................................... 25
Figure 3-3: The volume of arteries under the cuff is held constant by dynamically
varying the counter pressure, guided by the volume signal from the optical
plethysmograph. The amplifier senses the deviation from the set-point and force the
control valve to increase the cuff pressure. Republished with permission of Springer
Nature from (Settels, 2014). ............................................................................................. 26
Figure 3-4: Nexfin start-up and Physiocal adjustments. Initially, the set-point is
calibrated every 10 beats. The data is recorded during the baseline period of the
mental stress protocol. ..................................................................................................... 27
Figure 3-5: A typical wave form of PPG and arterial pulse. ........................................... 28

viii
Figure 3-6: The relationship between PPGa and DBP signals. Red rectangle indicates
negative correlation between two signals. Blue rectangle indicates positive correlation
between two signals. ........................................................................................................ 29
Figure 3-7: Experimental measurements. The measurements are exported into a .MAT
file. ..................................................................................................................................... 32
Figure 4-1: An example of power spectra of cardiovascular variables during baseline of
a presentative subject. All signals are resampled at 2Hz. Oscillations below 0.03 Hz
were removed. .................................................................................................................. 38
Figure 4-2: (A) Correlation analyses between baseline BRSv and ∆PPGa: the
relationship between baseline BRSv and ∆PPGa is stronger in SCD subjects
(p=0.0005). (B) Correlation analyses between baseline BRSc and ∆RRI. No interaction
was observed between baseline BRSc ........................................................................... 45
Figure 4-3: An example of cardiovascular variables during the mental stress protocol of
a single subject. “Tasks” (top panel) displays the output of the E-prime software where
the height of the bars represents the difficulty of the task. .............................................. 46
Figure 4-4: Correlation analyses between baseline BRSv and ∆PPGa in SCD and non-
SCD during N-back (A), Stroop (B) and pain anticipation (C). ........................................ 47
Figure 4-5: The bar graphs show magnitude of vasoconstriction (mean SE) in SCD
and non-SCD during MTS before (A) and after (B) adjusting for baseline BRSv........... 48
Figure 5-1: PPGa time-series of three subjects during baseline and Stroop. Subject 3
had sustained vasoconstriction during Stroop. Subject 16 exhibited large fluctuations in
the PPGa response during Stroop. .................................................................................. 57
Figure 5-2: The block entropies calculated from PPGa responses during Stroop. m
indicates the block size. Each line represents each subject. .......................................... 62
Figure 5-3: (Top) Example of a sustained vasoconstriction to mental stress. (Bottom)
Example of the discretized PPGa time-series using 6 quantization levels. .................... 68
Figure 5-4: Application of block entropy on non-sustained signals during Stroop. Both
signals were from different SCD subjects. These signals had the same variability as
quantified by the standard deviation of 0.08. However, the values of block entropy were
different. ............................................................................................................................ 69
Figure 5-5: Association between the number of unique blocks and block entropy during
Stroop. The yellow rectangle indicates the subject with sustained vasoconstriction. The
blue box indicates the subject with non-sustained vasoconstriction............................... 70
Figure 5-6: Mean  SE of block entropy during baseline, N-back and Stroop in controls
(n=15) and SCD subjects (n=15). .................................................................................... 72

ix
Figure 5-7: (Top) Changes in the block entropy from baseline to N-back. (Bottom)
Changes in the block entropy from baseline to Stroop. The error bars indicate mean 
SE of block entropy. ......................................................................................................... 73
Figure 5-8: Pearson’s correlation revealed positive relationships between block
entropies and mean PPGa. .............................................................................................. 81
Figure 5-9: Pearson's correlation revealed positive relationships between block
entropies and LFPPPG. ...................................................................................................... 81
Figure 5-10: Pearson's correlation between block entropies and BRSv in non-SCD and
SCD subjects. In the SCD group, “smibr” was excluded from N-back and “wasjy” was
excluded from Stroop. ...................................................................................................... 82
Figure 5-11: Pearson’s correlation showed that there were no relationships between
SEQ% BRSv and block entropies. ...................................................................................... 83
Figure 5-12: Pearson’s correlation showed no relationships between BRSv and SEQ%
BRSv. ................................................................................................................................... 83
Figure 5-13: Pearson’s correlation showed that there were no relationships between
LFPnu PPG and block entropies. ....................................................................................... 84

x
Abstract
In sickle cell disease (SCD), flexible red blood cells, after releasing oxygen to the
tissues, become rigid structures, producing regional obstructions of capillary flow that can
potentially cascade into full-blown painful vaso-occlusive crises (VOC). Anecdotal
evidence suggests that VOC is often precipitated by stress, cold exposure or pain.
Recent studies have shown that individuals with sickle cell disease (SCD) exhibit
greater vasoconstriction responses to physical autonomic stressors, such as heat pain
and cold pain than normal individuals, but this is not the case for mental stress.
Mental stress caused neural-mediated vasoconstriction in SCD and control
subjects; however, there was substantial variability in the magnitude and the patterns of
vasoconstriction responses across groups and individuals. In this thesis, we sought to
determine whether the variability of magnitude and the patterns of vasoconstriction
responses are related to inter-individual differences in autonomic function.
In our experimental protocol, fifteen subjects with SCD and 15 healthy volunteers
participated in 3 mental stress tasks: N-back, Stroop, and pain anticipation. R-R interval,
arterial blood pressure and finger photoplethysmogram (PPG) were continuously
monitored before and during these mental stress tasks. The magnitude of
vasoconstriction was quantified using changes in PPG amplitude (PPGa) from the
baseline period. The patterns of vasoconstriction were characterized from the patterns
of PPGa signals during N-back and Stroop tasks. “Block Entropy” was applied to PPGa
signals to quantify how close the pattern of vasoconstriction is to sustained
vasoconstriction.

xi
To assess autonomic function, we estimated cardiac and vascular baroreflex
sensitivities. Cardiac baroreflex sensitivity (BRSc) was estimated by applying both the
“sequence” and “spectral” techniques to beat-to-beat measurements of systolic blood
pressure and R-R intervals. The vascular baroreflex sensitivity (BRSv) was quantified
using the same approaches, modified for application to beat-to-beat diastolic blood
pressure and PPG amplitude (PPGa) measurements. For each subject, BRSc and BRSv
were assessed during baseline and stress periods.
In the first part of this work (Chapter 4), we determined whether the magnitude of
vasoconstriction is related to inter-group difference in baseline BRSc and BRSv. We
found that baseline BRSc was not different between SCD and non-SCD subjects, was
not correlated with BRSv, and was not associated with the magnitude of vasoconstriction
responses to mental stress tasks. BRSv in both groups was correlated with mean PPGa,
and since both baseline PPGa and BRSv were lower in SCD, these findings suggested
that the SCD subjects were in a basal state of higher sympathetically-mediated vascular
tone. In both groups, baseline BRSv was positively correlated with the magnitude of
vasoconstriction responses to N-back, Stroop and pain anticipation. After adjusting for
differences in BRSv within and between groups, we found no difference in the
vasoconstriction responses to all 3 mental tasks between SCD and non-SCD subjects.
In the second part (Chapter 5), we examined whether sustained vasoconstriction
was associated with autonomic function assessed during N-back and Stroop tasks. We
found that sustained vasoconstriction was not particularly associated with SCD. In
addition, the subjects who had sustained vasoconstriction also had low mean PPGa and
BRSv during stress tasks. There was a strong correlation between BRSv and the degree

xii
of sustainability as indicated by block entropy. These findings suggested that sustained
vasoconstriction during mental stress was primarily driven by increase in sympathetic
tone.
Overall, this work demonstrates that inter-individual difference in sympathetic
function affects the variability of vasoconstriction responses to mental stress in sickle cell
disease.

1
CHAPTER 1 INTRODUCTION
Sickle cell disease (SCD) is caused by a single mutation in the ß-globin gene,
resulting in the production of sickle hemoglobin which polymerizes after deoxygenation
(Rees et al., 2010). The polymerization changes liquid hemoglobin into a solid,
transforming the flexible red blood cell (RBC) into sickled-shaped RBC that tend to
obstruct microvascular flow. The clinical manifestation of extensive obstruction of
microvascular flow is episodic painful vaso-occlusive crisis (VOC), the hallmark symptom
of SCD. VOCs account for most hospitalizations (Yang et al., 1995; Yusuf et al., 2010)
and are also associated with increased mortality rate (Platt et al., 1991). Crisis frequency
worsens with age (Darbari et al., 2012). Lower fetal hemoglobin concentration, higher
hemoglobin or higher blood viscosity have been associated with higher frequent pain
episodes (Nebor et al., 2011; Darbari et al., 2012; Darbari et al., 2013). Nonetheless,
these factors do not account for immediate transition from steady state to VOC, and the
mechanisms responsible for initiating VOC remain elusive.
In anecdotal reports, SCD patients clearly state that VOC is often precipitated by
stress, cold exposure or pain, the conditions known to activate autonomic nervous system
(ANS). Since ANS is a major regulator of precapillary blood flow, it has been suggested
that ANS dysregulation of peripheral blood flow can promote vaso-occlusion (Coates et
al., 2018; Veluswamy et al., 2019). ANS could be modulated by pro-inflammatory states,
and in turn, could exacerbate inflammatory responses creating a vicious cycle of pain
(Ballas et al., 2012). In the past recent years, our group has employed multiple laboratory
stressors, such as. hypoxia (Sangkatumvong et al., 2011), cold face stimulation

2
(Chalacheva et al., 2015), heat pain (Khaleel et al., 2017), cold pain (Veluswamy et al.,
2020) and orthostatic stress (Chalacheva et al., 2019), to determine the resulting
autonomic nervous system (ANS) responses in SCD and control subjects. These studies
demonstrated the presence of significant abnormalities in the ANS responses of the
subjects with SCD. Importantly, we found that the peripheral vasoconstriction in response
to heat and cold stimuli was greater in SCD.
Psychological factors play an important role in SCD patients’ quality of life. Higher stress,
along with mood, has been associated with increases in self-reported SCD pain frequency
and pain intensity (Porter et al., 2000; Gil et al., 2004). In turn, increases in SCD pain led
to increases in stress (Gil et al., 2003). Such observations suggest the possibility that
mental stress (MTS) may also play a role in the onset of VOC by exerting its impact
through the ANS. In a previous study (Shah et al., 2020), we measured the cardiovascular
responses of SCD and healthy subjects to MTS tasks that included N-back, Stroop and
anticipated pain tests. However, in contrast to physical autonomic stressors, such as heat
pain and cold pain, where SCD subjects showed exaggerated vasoconstriction responses
(Khaleel et al., 2017; Veluswamy et al., 2020), the responses of the SCD subjects to MTS
were not different from their control counterparts. We also observed substantial variability
in terms of the magnitude and patterns of the responses within the SCD and control
subjects. We sought to address this variability by analyzing the vasoconstriction
responses of SCD and controls to mental stress in the context of a more detailed and
quantitative characterization of the underlying autonomic physiology. The objectives of
this dissertation are as follows.

3
Objective 1: To investigate how resting autonomic function influences the magnitude of
vasoconstriction responses to mental stress.

Objective 2: To comprehensively examine the relationships between the patterns of
vasoconstriction and autonomic function during mental stress.

4
CHAPTER 2 BACKGROUND
The Autonomic Nervous System
The autonomic nervous system (ANS) is part of the central nervous system that
controls most visceral functions. The system is responsible for transmitting the
information from the central nervous system (CNS) to target organs. The ANS operates
mostly via visceral reflexes. In brief, sensory signals from a visceral organ trigger complex
interactions among higher brain regions such as the brainstem which in turn modulate
autonomic activity traveling back to control the organ. The autonomic signals are
transmitted to target organs via two autonomic divisions, the sympathetic nervous system
and the parasympathetic nervous system. The sympathetic nervous system is “well-
known” for its “fight-or-flight” response as the system is responsible for accelerating heart
rate, increasing blood pressure when the body is under stress or pain. The
parasympathetic nervous system, however, dominates during rest periods such as
digestion or sleep. In absence of external stimuli, a resting heart rate of a person is mostly
determined by the activity of the parasympathetic nervous system.

5

The sympathetic pathway
The efferent sympathetic pathway is composed of two neurons, a preganglionic
neuron and a postganglionic neuron. The preganglionic sympathetic fibers leave the
spinal cord and carry the signals from the central nervous system to one of the
sympathetic ganglia where they synapse with the postganglionic sympathetic neurons
(Figure 2-1). The postganglionic sympathetic fibers travel to target organs and terminate
near the locations of the effector cells. Here, these fibers release sympathetic
neurotransmitters, norepinephrine or epinephrine, to stimulate or inhibit the target cells
(Hall and Guyton, 2011; Klabunde, 2011). In special cases, the preganglionic
sympathetic fibers may bypass the sympathetic chain ganglia and synapse with the
postganglionic neurons at the peripheral chain ganglia near the tissue of target organs.
There is difference in the anatomical distribution between the efferent sympathetic
and the parasympathetic pathways. Some target organs such as blood vessels or
skeletal muscles are solely affected by sympathetic activity. Other organs such as the
heart, are innervated by both divisions.

6

Figure 2-1: Simplified sketch of the autonomic pathways to the heart and blood vessels.
The black lines represent the preganglionic sympathetic fibers. The black dashed lines
represent the postganglionic sympathetic fibers. Sympathetic stimulation increases
heart rate and causes vasoconstriction. Parasympathetic stimulation slows heart rate.
Adopted from (Klabunde, 2011).
The parasympathetic pathway
The efferent parasympathetic pathway is also a sequence of a preganglionic and
a postganglionic neuron. The majority of preganglionic parasympathetic neurons arise in
the brainstem’s gray matter. Their fibers leave the brainstem via cranial nerves III, VII,
IX, and X, and pass uninterrupted to parasympathetic ganglia locating in the wall of target
organs (Hall and Guyton, 2011). In the ganglia, they make synaptic connection with

7
extremely short postganglionic neurons. The postganglionic parasympathetic neurons
excite or inhibit the effector cells by secreting acetylcholine.
It is worth mentioning that the tenth cranial nerve, known as “vagus nerves”, is the
major nerve of the parasympathetic division. The vagus nerves carry about 75 percent
of preganglionic parasympathetic fibers that supply the heart, lungs, thoracic and
abdominal regions. Therefore, parasympathetic activity is often referred to as vagal
activity. Figure 2-1shows the simplified drawing of the parasympathetic nervous system
pathways to the heart.
Autonomic control of the heart
The heart serves as a primary pump that generates enough force (pressure) to
circulate blood within the body circulation (Figure 2-2). The heart contraction is initiated
by a group of pacemaker cells, known as the sinoatrial node (SA node), positioned on the
wall of the right atrium. The SA node spontaneously generates electrical impulses that
trigger action potentials of adjacent cardiomyocytes. The local action potentials quickly
spread over the atria and propagate to the atrioventricular node (AV node), another group
of pacemaker cells, which slows the propagation time, setting a coordinated contraction
between the atria and the ventricles. Functionally, the AV node delay permits the atria to
fully contract before the contraction of the ventricles is initiated. In an electrocardiogram,
the onset of P wave reflects the onset of atrial depolarization and the onset of QRS
complex reflects the onset of ventricular depolarization.

8

Figure 2-2: Anatomy of the heart. The heart consists of four chambers: right, left atrium
and right, left ventricle. Blood flows from the right atrium into the right ventricle and
enters the pulmonary circulation. Blood returning from the lungs enters the left atrium,
flows into the left ventricle which ejects blood out of the heart. SA, the sinoatrial node;
AV, the atrioventricular node.
Normally, the rate at which the SA node triggers heart beat is approximately 100
beats per minute (bpm) (Pappano and Wier, 2012). For this rate to change, it involves
the autonomic regulation. The SA and AV node are innervated by both autonomic
branches. Increase parasympathetic activity inhibits the SA node firing and impedes the
AV node conductance; thus, delaying the start of subsequent cardiac cycle (Robertson et
al., 2011; Pappano and Wier, 2012). Stimulation of the sympathetic nervous system not
only accelerates heart rate but also increases the heart contractility. The parasympathetic
and sympathetic activity change reciprocally to increase or decrease heart rate.
It should be noted that the autonomic nervous system is always active.
Deactivation on one of these branches merely decreases the frequency of nerve impulses
coming to the SA node, but never completely eliminates its presence. This continuous
nerve activity coming from the parasympathetic and the sympathetic nervous system is

9
known as “autonomic tone” (Figure 2-3). During rest, heart rate for a healthy person is
lower than the intrinsic rate of the SA node because of parasympathetic predominance
(Katona et al., 1982). External factors such as stress, pain or exercise can change the
balance toward sympathetic dominance.

Figure 2-3: Schematic diagram of hypothetical cases of parasympathetic tone. From A)
to B), increase in parasympathetic tone prolongs the start of cardiac cycle; thus,
decreasing heart rate.
Sympathetic control of blood vessels
Blood vessels are elastic conduits of the circulatory system. They come in different
types and sizes. Functionally, there are three types of blood vessels for 1) carrying blood
away from the heart 2) carrying blood toward the heart 3) delivering nutrients and oxygen
to surrounding cells and tissues. In order to understand how the sympathetic control of
blood vessels plays important role in the circulatory system, we must understand how
blood is transported from and to the heart.

10
When the heart contracts, oxygenated-blood is forced out of the heart through the
largest artery, known as the aorta. From there, the aorta branches into large arteries
which transport blood to specific organs. Large arteries are highly elastic; thus, capable
of reserving a potential energy necessary enough to maintain blood flow during diastole
phase. After large arteries reach organs, they branch into small arteries which distribute
blood within the organs. These small arteries keep branching into smaller vessels until
they reach the smallest arteries called arterioles. Eventually, arterioles further branch
into capillaries where nutrients and waste are exchanged. In the capillary bed, the blood
has a higher concentration of oxygen than the cells causing oxygen to diffuse to the cells.
This process deoxygenates the blood as the concentration of oxygen decreases.
Deoxygenated blood is transported from capillaries to bigger vessels called venules which
carry waste out of the capillary bed. Venules further join into even bigger vessels known
as veins and finally these vessels return to the inferior or superior vena cava which return
blood back to the heart.
The structure of blood vessels varies slightly among different types. For example,
capillaries only have one or two single layers and are not innervated by the sympathetic
nervous system (Klabunde, 2011). The others consist of three layers – intima, media and
adventitia. The media of the middle layer contains smooth muscle cells innervated
primarily by the sympathetic nervous system through adrenergic receptors. The
contraction or relaxation of smooth muscles cells is what causes vasoconstriction and
vasodilation in blood vessels. Depending on the type of adrenergic receptors (alpha or
beta receptors), sympathetic stimulation can cause excitatory effects in some organs but
inhibitory effects in others (Hall and Guyton, 2011). In the finger, where the walls of the

11
cutaneous vessels contain mostly the alpha receptors, sympathetic stimulation during
cold immersion causes vasoconstriction (Awad et al., 2001; Robertson et al., 2011;
Lombard and Cowley Jr, 2012). The beta-2 receptors are located in the arteries of the
skeletal muscles. When these receptors are stimulated, the vessels dilate so that skeletal
muscles can receive more blood flow pumped out of the heart.
Normally, blood vessels are partially constricted at all times. This constricted state
is known as intrinsic tone of the vessels and determined by the centrally driven basal
sympathetic activity as well as endothelial functions. In order to have rapid control of
blood pressure and blood flow, sympathetic control on blood vessels are generally
superimposed on the intrinsic tone. This rapid controls are brought about by reflex
mechanisms such as baroreceptor reflex. In summary, the effects of sympathetic nervous
system on blood vessels are a combination of basal sympathetic activity and autonomic
reflex mechanisms.

12
Baroreceptor Reflex
By far, the baroreflex function is the best-known negative feedback for
cardiovascular system. The reflexes are responsible for maintaining arterial blood
pressure and preventing large, short-term blood pressure fluctuations (Hall and Guyton,
2011; Pappano and Wier, 2012). Baroreceptor reflex modulates the parasympathetic and
sympathetic outflow to the heart and blood vessels, mainly to control CO and SVR in
response to acute changes in arterial blood pressure.

Figure 2-4: Schematic diagram of cardiac and vascular baroreflexes

13
Anatomy and activation
The arterial baroreflex is initiated by stretch receptors known as baroreceptors.
These receptors are presented in multiple areas of the body (e.g. in the atria, in the
ventricles) but the most sensitive ones located in the carotid sinus (Figure 2-4).
Baroreceptors are activated by the stretch of vessel walls. When blood pressure rises,
there is more force exerting on the walls of blood vessels. This makes the walls expand,
and at the same time, baroreceptors are being activated. Excited baroreceptors send
signals to the brain, primarily to the nuclear tractus solitarius (NTS) in the medulla
(Dampney, 2017). NTS has controls over the parasympathetic and sympathetic centers;
thus, can adjust the neural outflow according to the received inputs. Subsequent
reduction in SVR will attenuate the blood pressure rises. When arterial blood pressure
falls, the reflex responds by causing subsequent increase in heart rate and
vasoconstriction. This negative feedback allows the ANS to maintain blood pressure
within a normal range.

14
Sickle cell disease
Sickle cell disease (SCD) is a life-threatening blood disorder affecting nearly 90000
people in United States (Hassell, 2010). The average survival rate of SCD patients was
less 45 years in U.S. (Hassell, 2010) and less than 5 years in African countries (Makani
et al., 2011). Unexpected occurrences of extremely painful affect the mental and physical
health of the patients (Mahdi et al., 2010). The lack of comprehensive care for adults with
SCD may be associated with the overall increased mortality rate over the past 10 years
(Lanzkron et al., 2013).

Figure 2-5: Hemoglobin S polymerizes after deoxygenation

15
Pathology and pathophysiology
Sickle cell disease is a genetic blood disorder that distorts the shape of red blood
cells. The transformation process, known as sickling, is caused by the polymerization of
deoxygenated sickle hemoglobin (HbS) inside a red blood cell (RBC). Normally, the most
common form of adult hemoglobin (HbA) is composed of two alpha-globin and two beta-
globin subunits. In SCD patients, the sixth amino acid of the beta-globin chain is
substituted by valine due to the allele coding of the mutated beta-globin gene. The
absence of a single amino acid promotes aggregation of HbS molecules that grows and
disrupts the membrane of a red blood cell, turning a normal-shaped red blood cell into a
sickled shape red blood cell (Figure 2-5,2-6). Although the sickling process is reversible,
repeated disruption of the cell’s membrane leads to premature destruction of red blood
cells, known as hemolysis. As a result, most patients with SCD are anemic.

16

Figure 2-6: Sickle erythrocytes. Peripheral blood smear from a patient with SCD
obtained during a routine clinic visit. The smear shows classical sickle-shaped (arrows)
and various other misshaped erythrocytes (arrowheads). Republished with permission
of American Society for Clinical Investigation, from (Frenette and Atweh, 2007)

In sickle form, the red blood cells are less flexible and tend to obstruct blood flow
in microvasculature potentially causing vaso-occlusions. Although the exact mechanisms
of vaso-occlusions have not been fully understood, there are speculations that vaso-
occlusions are contributed by endothelial activation, interactions between sickled RBCs
and leukocytes or increased transit time via dysregulation of blood flow (Frenette, 2002;
Manwani and Frenette, 2013). Interactions with leukocytes were also suspected to be
the cause of the vaso-occlusion as leukocytosis was associated with clinical severity of
SCD (Frenette, 2002). The major consequences of the occlusions are inflammation,
acute and chronic pain. In severe cases, depending on the sites, vaso-occlusions can
lead to stroke, acute chest syndrome or multiorgan failure syndrome in early childhood
(Serjeant, 2013).

17
The severity of SCD is affected by the genotypes of the globin genes. The disease
is autosomal recessive, meaning that two alleles must be inherited to demonstrate clinical
syndromes. The most common and severe form is the homozygous state in which two
copies of the βs allele are inherited, producing HbSS. Other forms include the inheritance
of one βs allele along with βc allele, β-thalassemia or β
0
thalassemia producing HbSC,
HbSβ
+
and HbSβ
0
respectively. It was shown that the severity of the disease was milder
in HbSC, HbSβ
+
but was stronger in HbSβ
0
and HbSS genotype (Platt et al., 1991).
Vaso-occlusive crisis (VOC)
Pain in SCD patients is related to sickling process that affects microcirculation
multidimensionally. Sickled red blood cells cause obstruction in the microcirculation and
cause damages to the walls of blood vessels, triggering inflammatory responses that
cause pain. There are two types of sickling-related pain – chronic pain and vaso-
occlusive crisis (VOC).
Chronic pain is associated with chronic disorders such as avascular necrosis, back
pain, knee pain that may last more than 3 months (Taylor et al., 2010). Chronic pain can
be treated at home using analgesics, hot baths, massage or other relaxation activities.
The location of the pain is variable but the hips and the back were the most common sites
(Taylor et al., 2010).
Vaso-occlusive crisis (VOC) is characterized by acute episodes of pain (Darbari et
al., 2020). The pain associated with VOC is often described as sudden in onset and high
intensity. VOC is often treated in hospital emergency departments or inpatient units. It
is the primary cause of hospitalizations for SCD patients with an average hospital stay of
9-10 days in adults (Ballas and Lusardi, 2005; Panepinto et al., 2005).

18
There are a number of clinical consequences associated with VOC. For example, acute
chest syndrome occurs in 10-20% of hospitalized patients with VOC, usually manifest 1-
3 day after admission for severe pain (Hassell et al., 1994; Vichinsky et al., 2000). In
some cases, it progresses to acute respiratory failure and results in death (Gladwin and
Vichinsky, 2008).
Typically, VOC is treated with vigorous intravenous hydration and analgesics.
Although an opioid dose used to treat VOC is much higher than home treatments, the
treatment is not always effective. In many cases, the pain intensity was only improved in
the beginning of hospitalization (Ballas and Lusardi, 2005). Some patients were
discharged with severe pain. Re-admissions were also common.
Identification, treatment and prevention of VOC could significantly improve the quality of
life of SCD patients.
Autonomic nervous system involvement in VOC
Vaso-occlusive crisis is very much unpredictable. Although it has been widely
accepted that pain crisis stems from regional vaso-occlusions, it is not yet clear what
factors cause the immediate transition from a non-painful steady state to the rapid onset
of perceptible pain. In fact, sickling process may continually occur in the microcirculation
but is not extensive enough to cause noticeable pain. This leads to the speculation that
there are other factors on top of regional vaso-occlusions that reinforce the occlusion
process, and trigger the full-blown crisis. Our group strongly believes that the autonomic
nervous system may have come into play (Coates et al., 2018; Veluswamy et al., 2019).
Our hypothesis is influenced by the discovery in 1974 (Hofrichter et al., 1974). In
that experiment, it was found that upon deoxygenation, there was a delay time prior to

19
the HbS polymerization. This discovery led to the hypothesis that any factors that
decrease the delay time or increase the transit time through the microcirculation can
promote vaso-occlusions and trigger pain crisis (Hofrichter et al., 1974; Eaton et al., 1976;
Mozzarelli et al., 1987; Eaton, 2020). Because the ANS is a major regulator of
precapillary blood flow in the body, it was speculated that the ANS can affect the transit
time and promote vaso-occlusions which may lead to pain crisis (Coates et al., 2018;
Veluswamy et al., 2019).
Early studies on the ANS in SCD patients primarily focused on the cardiac ANS
(Mestre et al., 1997; Pearson et al., 2005; Treadwell et al., 2011; Martins et al., 2012). In
most of these studies, the ANS was assessed by the statistical measures or the frequency
power of heart rate variability. The major findings were that SCD patients had cardiac
autonomic dysfunction. However, the degree of abnormality was variable and the
connection between cardiac autonomic dysfunction and vaso-occlusive was still unclear.
The vascular autonomic function in SCD patients had been thoroughly investigated
by our group and a few others in the past ten years. Mainly, we were interested to see
what role the autonomic regulation of blood flow may have on the transition from a steady
state to pain crisis. Our pioneer work was conducted in 2008 (Sangkatumvong et al.,
2008; Sangkatumvong et al., 2011). In this study, night time hypoxia was mimic by having
SCD subjects breathed in 5 breaths of 100% nitrogen. Although we found that SCD
patients had more parasympathetic withdrawal than controls during transient hypoxia,
decreases in microvascular blood flow were not different between two groups.
In more recent years, we conducted the experiments where the autonomic-
mediated blood flow responses were measured during the exposure to stress or pain, the

20
known risk factors for VOC. In two of our experiments, pain stimulation in the form of
“heat” or “cold” was induced on the right arm of SCD subjects (Khaleel et al., 2017;
Veluswamy et al., 2020). We found that the pain stimulation produced vasoconstriction
responses in SCD patients. In addition, SCD patients had stronger vasoconstriction to
pain pulses. Further investigation revealed that exaggerated responses in SCD patients
were the results of abnormal interactions between the ANS and vascular interface. It was
also suggested these were the consequence of endothelial dysfunction (Chalacheva et
al., 2017).
Similar findings were observed in another experiment where experimentally
induced stress was used as stimuli. In this study, 85% of SCD subjects showed
vasoconstriction responses to mental stress. However, the responses were comparable
to healthy subjects. It was speculated that such hyperresponsiveness was not detected
because some patients were already vaso-constricted during the baseline period (Shah
et al., 2020).
In 2019, we conducted a tilt test where a subject was tilted from a supine to near-
upright position (Chalacheva et al., 2019). The test induced a transient drop in blood
pressure followed by tachycardia and peripheral vasoconstriction. While peripheral
vasoconstriction was observed in SCD patients and controls, we did not detect the
significant difference of the vasoconstriction responses between two groups. The only
difference seen between two groups was that the majority of controls showed both cardiac
and vascular responses while the SCD group was more prone to show peripheral
vasoconstriction.

21
According to these findings, it is still inconclusive whether SCD patients have
autonomic dysregulation of blood flow because the degree of abnormality varied across
experiments and SCD patients. Exaggerated vasoconstriction responses were only
observed during thermal pain stimulation but not during transient hypoxia, mental stress
or tilt test. In most of our experiments, we did not see the significant difference in the
cardiac ANS between controls and SCD patients.
Nevertheless, this does not negate the fact that the autonomic nervous system can
reduce microvascular blood flow during the exposure to known risk factors for VOC.
Importantly, because the reduction of microvascular blood flow can increase the transit
time of red blood cells through the microcirculation, we would predict that the SCD
patients who showed exaggerated vasoconstriction responses would experience more
crises. Certainly, future studies are still necessary to confirm this hypothesis.

22
CHAPTER 3 EXPERIMENTAL METHODS AND DATA
PROCESSING
The study was conducted under Institutional Review Board approved protocol at
Children’s Hospital Los Angeles (CHLA). All participants were patients or patients’
families who received care at CHLA. Age and ethnicity matched between non-SCD and
SCD groups. Inclusion criteria are participants who are at least 12 years old, free of vaso-
occlusive crisis in the past 10 days. Subjects were excluded if they were having active
anxiety disorder. Informed consent and assent were obtained prior to the study.
Measurements and Data Processing
Measurements were carried out in a quiet, dimmed light, temperature-controlled
room (27  0.5  C). The subjects rested comfortably at an approximately 45  angle on a
cushioned chair with arm and leg supports. The photoplethysmogram (PPG) was
obtained from the thumb of the subjects’ left hand (Nonin Medical Inc., USA). Continuous
blood pressure was measured from the left middle finger using a noninvasive finger cuff
(Nexfin; BMEYE, Amsterdam, The Netherlands). Other physical measurements include
laser Doppler flowmetry (Perimed, Sweden), electrocardiogram (ECG), respiration (zRip
DuraBelt, Philips) were also recorded. All measurements were acquired synchronously
through Biopac MP150 data acquisition system (Biopac, USA) at 250 Hz.
Beat-to-beat values were detected with respect to the R peaks on the ECG. R-to-
R interval (RRI) was defined as the time between two consecutive R-waves. Systolic and
diastolic blood pressure (SBP and DBP) were the peak and nadir of the blood pressure

23
pulse during the cardiac cycle. The pulse amplitude of the PPG waveform (PPGa), which
reflects changes in pulsatile finger blood volume, was measured from the difference
between the peak and nadir of the PPG waveform within the cardiac cycle. We used
PPGa as a surrogate measure of peripheral vascular conductance, thus assuming that
decreases in PPGa reflect vasoconstriction and increases reflect vasodilation (Alian and
Shelley, 2014). It has been shown that increases and decreases in PPGa generally
coincide with cutaneous blood flow, measured using laser doppler flowmetry (Rauh et al.,
2003; Khoo and Chalacheva, 2019), and that these changes take place primarily as a
result of peripheral vasodilation or vasoconstriction. However, it is also known that a
fraction of these changes in PPGa may be due to changes in pulse pressure and/or local
arterial distensibility (Khoo and Chalacheva, 2019).Since the PPG is a relative
measurement, PPGa was normalized to its own 95th percentile value of its full study
recording and expressed in normalized units (nu).
Photoplethysmography
Photoplethysmography is a low-cost and non-invasive technique used to detect
changes in blood volume in the microvascular bed of tissue. The basic form of
photoplethysmography requires a light emitting diode to irradiate tissue and a
photodetector to detect the intensity of transmitted or reflected light (Allen, 2007; Reisner
et al., 2008; Alian and Shelley, 2014). When light is directed through tissue at a
measurement site (e.g. finger, earlobe), part of the light is absorbed by bones, skin
pigments and blood. The amount of light not absorbed is measured and output as a raw
PPG. The raw PPG reflects blood volume variations as the detected light intensity varies
with the amount of blood presented in the circulation. Generally, the raw PPG has two

24
components - a pulsatile and non-pulsatile component. The pulsatile component (AC
component) reflects arterial blood volume variations caused by heartbeats (Allen, 2007;
Reisner et al., 2008; Korhonen and Yli‐Hankala, 2009; Elgendi, 2012; Alian and Shelley,
2014). The non-pulsatile component (DC component) is more static and related to
venous blood volume variations (Allen, 2007; Reisner et al., 2008; Korhonen and Yli‐
Hankala, 2009; Elgendi, 2012; Alian and Shelley, 2014). It is filtered out by most
commercial monitors, and only the AC component is displayed and analyzed further.

Figure 3-1: Example of PPG and PPGa signals. The PPG only contains the AC
component as the DC component is filtered out by the measuring device.
Continuous noninvasive blood pressure monitoring (Nexfin)
Arterial blood pressure is defined as the force per unit of surface that blood exerts
on the arterial wall when it is stressed. Most of this pressure results from the heart
pumping blood through the circulation. The Nexfin monitor measures arterial blood
pressure continuously and noninvasively by means of an inflatable finger cuff with a built-
in photo-electric plethysmograph (Eeftinck Schattenkerk et al., 2009; Martina et al., 2012;
Ameloot et al., 2013).

25
The finger cuff encloses a finger and is dynamically inflated to exert a counter
pressure to keep arterial blood volume unloaded or constant at a set-point volume. This
method is known as the Volume-Clamp technique (Penaz, 1973). When arterial blood
volume increases, the volume signal from the plethysmograph deviates from the set-point.
The deviation triggers a fast servo feedback system to increase the finger cuff pressure
to counter these changes so that the arterial wall is kept unloaded. The counter pressure
is measured and re-constructed into a brachial artery waveform (Gizdulich et al., 1996;
Gizdulich et al., 1997).

Figure 3-2: Volume clamp principle observing the dynamic unloading of the arterial wall
by using an infrared optical plethysmograph built in the cuff. Republished with
permission of Springer Nature from (Settels, 2014).

26

Figure 3-3: The volume of arteries under the cuff is held constant by dynamically
varying the counter pressure, guided by the volume signal from the optical
plethysmograph. The amplifier senses the deviation from the set-point and force the
control valve to increase the cuff pressure. Republished with permission of Springer
Nature from (Settels, 2014).
The Nexfin monitor uses the Physiocal method (Wesseling, 1990; 1995) to
determine the set-point or the unloaded state of arteries. This method gradually increases
a counter pressure from zero to supra-systolic while analyzing the amplitude and shape
of the recorded PPG. When a cuff pressure is held constant at each pressure level,
changes in transmural pressure can be extracted. The volume at which the transmural
pressure is zero is the set-point volume (Boehmer, 1987; Eeftinck Schattenkerk et al.,
2009; Settels, 2014). After the Nexfin start-up, the device adjusts the set-point every 10
beats, and increases the interval to 70 beats as stability increases. The example of Nexfin
Physiocal adjustments is shown in Figure 3-4. More details about the Physiocal
adjustments have been described in previous articles (Boehmer, 1987; Wesseling, 1990;
Eeftinck Schattenkerk et al., 2009; Martina et al., 2012; Settels, 2014).

27

Figure 3-4: Nexfin start-up and Physiocal adjustments. Initially, the set-point is
calibrated every 10 beats. The data is recorded during the baseline period of the
mental stress protocol.
Pulse amplitude of PPG and blood pressure are not necessarily
correlated
During a systolic phase, the volume of blood is pumped into the arterial system.
Therefore, arterial pressure and blood volume rise as indicated by the rising age of a PPG
and arterial pulse (see Figure 3-5). In contrast, arterial pressure and blood volume falls
during a diastolic phase. This is indicated by the falling edge of a PPG and arterial pulse.

28

Figure 3-5: A typical wave form of PPG and arterial pulse.
Although the morphology of the PPG signal is similar to the arterial blood pressure
waveform within each beat, the beat-to-beat parameters extracted from these two signals
are generally not correlated. The PPGa reflects a change in finger blood volume within
each cardiac cycle. This change is determined in part by pulse pressure and vascular
distensibility (compliance) of the blood vessels in the finger. On the other hand, arterial
pressure varies depending on location of measurement within the circulation, the overall
compliance properties of the vasculature, cardiac output, blood volume, total peripheral
resistance, and other phenomena like wave reflection from the periphery of the arterial
tree.

29
On a beat-to-beat basis, a PPGa signal is not necessarily correlated with SBP and
DBP signals. Correlations may appear transiently and can either be positive or negative,
depending on the underlying mechanisms that drive them. In addition, the relationships
may vary over time. Figure 3-6 shows examples of these “transient” correlations that can
be opposite in sign. In the first part of figure (red box), PPGa shows a drop between 2490
and 2496 s, and this is accompanied by a rise in DBP (between 2492 and 2496 s). This
likely reflects an autonomically mediated peripheral vasoconstriction that leads to an

Figure 3-6: The relationship between PPGa and DBP signals. Red rectangle indicates
negative correlation between two signals. Blue rectangle indicates positive correlation
between two signals.

30
increase in blood pressure in accordance with Ohm’s law (ie. increase in resistance
leads to increase in voltage when current remains relatively unchanged). Subsequently,
the increase in DBP triggers a reduction in peripheral resistance (displayed as an
increase in PPGa) via the vascular baroreflex. In the latter part of the figure (blue box),
there is a simultaneous reduction in both PPGa and DBP (between 2515 and 2519 s).

31
Experimental protocols
Measurements were performed in the autonomic laboratory where the participant
sat comfortably on a cushioned chair with arm and leg supports. After all measurements
stabilized, the subjects answered the State-Trait Anxiety Inventory (STAI) questionnaire.
Then, the experimenter started collecting baseline data for 5 minutes. Following the
baseline recording, the mental stress protocol was induced through a psychological
software (E-prime 2.0, Psychology software Tools, USA). Two trials of mental stress (N-
back and Stroop) were presented in a random order on a computer screen. The STAI
questionnaire was measured once more between two trials. Subjects responded to N-
back task using an ‘enter key’ and Stroop using a mouse. At the end of mental stress, a
pain anticipation task was presented. Subjects read the following sentence on their
computer screen: “You will receive a maximum pain stimulus in one minute. When you
cannot tolerate the pain any longer, say STOP and the device will cool down to normal
level immediately.” However, no actual pain stimulus applied. Finally, the post baseline
measurements were recorded for 3 minutes.
During N-back, a sequence of alphabetic letters appeared on a computer screen.
The subjects responded when a letter was repeated from n steps back earlier in a
sequence (n= 0, 1, 2, 3). N-back trial consisted of 12 sessions of 0, 1, 2 and 3-back
presented in a randomized order.
During Stroop, a sequence of color words appeared on the screen. The subjects
were asked to identify the ink color of the word, not the written name of the word. Stroop
trial consisted of 12 sessions of three difficulty levels of Stroop presented in a randomized

32
order. One session lasted about 1 minute and the time gap between sessions was 25
seconds.

Figure 3-7: Experimental measurements. The measurements are exported into a .MAT
file.

33
CHAPTER 4 THE ROLE OF BASELINE
BAROREFLEX ON VASOCONSTRICTION TO
MENTAL STRESS IN SCD
Previously, (Shah et al., 2020) addressed the role of mental stress on
microvascular blood flow in SCD and control subjects. They found that mental stress
consistently decreased mean PPGa in 85% of SCD subjects. In addition, there were
concurrent decreases in mean RRI and high frequency power of RRI (HFPRRI) during the
tasks. Their results strongly indicated that mental stress triggered sympathetic activation
in SCD subjects. However, similar responses were observed among the controls. There
was no significant difference in the magnitude of vasoconstriction responses between two
groups. This contradicts our previous evidences which found that SCD patients had
exaggerated vasoconstriction to heat and cold stimuli.
In healthy subjects, the effects of mental stress on muscle sympathetic nerve
activity (MSNA) have always been inconsistent. Mental stress has been reported to
increase and decrease MSNA. Several groups tried to evaluate factors that differentiate
between the non-responders and responders (Carter and Ray 2009; Donadio et al. 2012;
El Sayed et al. 2016). In 2016, El Sayed et al demonstrated that the rise of blood pressure
at the onset of mental stress was associated with the direction of change of MSNA
responses. They suggested that in non-responders, the baroreflex function may initially
respond to the rise in blood pressure by triggering compensatory vasodilation (decrease

34
in MSNA). While in responders, the baroreflex function was “reset” or “adapt” to operate
at a higher level of blood pressure; thus, allowing an increase in MSNA.
Although we did not measure actual MSNA in this work, we cannot neglect the
possibility that the variability in the magnitude of vasoconstrictions in our study may have
been influenced by factors such as resting baroreflex function. In this chapter, we sought
to investigate the variability in the magnitude of vasoconstriction responses within the
SCD and control subjects. In particular, we estimated baseline cardiac and vascular
baroreflex function, and determined how these indices may have influenced the
subsequent vasoconstriction responses to mental stress.
Method
Quantification of vasoconstriction
Vasoconstriction response, ∆PPGa, to each of the tasks was defined as the
difference between mean PPGa during baseline and the mean PPGa during each task.
Positive values of ∆PPGa therefore represented vasoconstriction, while negative values
reflected vasodilation.
Characterization of baseline autonomic function
Baseline autonomic function in each subject was assessed using the averages of the
beat-to-beat values of RRI, SBP, DBP and PPGa, along with estimates of the cardiac and
vascular baroreflex sensitivities. The latter autonomic indices were estimated based on
the spontaneous fluctuations in the cardiovascular variables during the baseline period
using two approaches:

35
(a) Baroreflex sensitivities using the “sequence” technique:
We used the “sequence” method (Parati et al., 2000) to estimate the cardiac baroreflex
sensitivity (BRSc), which reflects predominantly vagal control of the heart. In this approach,
we first identified the sequences of beat-to-beat SBP and RRI that changed together in
the same direction, either increasing or decreasing, for at least 3 consecutive beats. The
slope of the regression line between SBP and RRI was taken to represent an estimate of
BRSc over the duration of that sequence. The change in SBP or RRI from the current beat
to the next beat was required to be greater than a minimum threshold in order to qualify
as an increasing or decreasing sequence (Persson et al., 2001). RRI was allowed to lag
SBP by 0 to 5 beats (Porta et al., 2018). Only the sequences with the goodness of fit (R
2
)
of the regression line equal to or greater than 0.9 were considered in the analysis. The
average BRSc of all sequences identified in the baseline period was taken to represent
cardiac baroreflex sensitivity for that subject.
The vascular baroreflex sensitivity (BRSv) characterizes the factor by which
increases(decreases) in DBP induce vasodilation/vasoconstriction of the peripheral
vasculature. Previous studies have employed various measurements to quantify the
effector variable for this reflex, including muscle sympathetic nerve activity (Durocher et
al., 2011; Marchi et al., 2016), systemic vascular resistance (Borgers et al., 2014), and
skin blood flow (Porta et al., 2018). In our case, we used the changes in PPGa as the
surrogate measure of changes in peripheral vascular conductance. To estimate BRSv,
we used the approach proposed by (Porta et al., 2018) which was based on the sequence
technique used for estimating BRSc. In this case, sequences of consecutive increases

36
(decreases) in beat-to-beat DBP values were related to corresponding subsequent
sequences of parallel increases (decreases) in beat-to-beat PPGa. For each sequence,
the BRSv was defined as the slope of the linear regression of PPGa over DBP. The
sequences of DBP and PPGa values were composed of three or more consecutive beats.
The total change in DBP of each identified sequence had to be least 1 mmHg. The delay
for a sequence of DBP-PPGa values was allowed to range from 0 to 5 beats (Porta et al.,
2018). The delay was estimated via cross-correlation before the sequences were
identified, with the “optimal” delay corresponding to the maximum positive correlation
between DBP and PPGa beat-to-beat values. The average BRSv of all sequences
identified in the baseline period was taken to represent vascular baroreflex sensitivity for
that subject.
(b) Baroreflex sensitivities using the “spectral” technique:
As consistency checks on the aforementioned BRSc and BRSv estimates, we also
computed the corresponding baroreflex sensitivities using the “spectral” technique, which
has been used extensively for quantifying cardiac baroreflex sensitivity (Parati et al.,
2000). The beat-to-beat signals of RRI, SBP, DBP and PPGa were first converted into
uniformly sampled time series, with 0.5 s as the interval between samples, using an
interpolation and resampling algorithm (Berger et al., 1986). Subsequently, the mean and
very low frequency trend (0 to 0.01 Hz) were subtracted, following which the power
spectral density of each signal was computed using autoregressive modeling (Figure 4-
1). From each power spectra, we computed the area within the low-frequency (LF: 0.04
to 0.15 Hz) band in order to yield the following spectral indices: (a) LFPSBP (LF power of
SBP variability), (b) LFPRRI (LF power of RRI variability), (c) LFPDBP (LF power of DBP

37
variability), and (d) LFPPPGa (LF power of PPGa). Then, from the aforementioned spectral
indices, we computed the corresponding sensitivities (conventionally referred to as “ 
coefficients”) for both baroreflex arms. As in previous reports (Parati et al., 2000), for the
cardiac baroreflex, c was defined as:
∝
c
= √(
LFP
RRI
LFP
SBP
)
Along the same lines, we defined v for the sympathetic vascular baroreflex as:
∝
v
= √(
LFP
PPGa
LFP
DBP
)
Both spectral indices ( c, v) were subsequently compared against their corresponding
sequence-derived measures (BRSc, BRSv). The area of the power spectrum for RRI
variability within the high-frequency band (HF: 0.15 to 0.4 Hz) was also computed in order
to obtain HFPRRI (HF power of RRI variability), a well-accepted metric of vagal modulation
of heart rate.

38

Figure 4-1: An example of power spectra of cardiovascular variables during baseline of
a presentative subject. All signals are resampled at 2Hz. Oscillations below 0.03 Hz
were removed.
Statistical Analysis
The comparisons of continuous parameters of non-SCD and SCD were tested
using the Student’s t test or Wilcoxon rank sum test. The comparisons of dichotomous
variables were tested using Pearson’s Chi-squared test. Since each subject participated
in all MTS tasks, linear mixed model analysis was used to test for the significance of
stress tasks and diagnosis on the changes in cardiovascular variables. This included
regression analysis to determine the nature of the associations between the baseline
baroreflex indices are cardiovascular responses to the three different mental tasks, after
0 0.2 0.4
0
0.05
HRV (sec
2
)
0 0.2 0.4
0
100
200
BP (mmHg
2
)
DBP
SBP
0 0.2 0.4
0
0.1
0.2
PPGa(au
2
)
0 0.2 0.4
Freq(Hz)
0
0.05
Respiration(L
2
)
0 100 200 300
-0.4
-0.2
0
0.2
RRI (sec)
0 100 200 300
-20
0
20
BP (mmHg)
0 100 200 300
-0.2
0
0.2
PPGa (au)
0 100 200 300
Time(sec)
-0.2
0
0.2
0.4
Respiration (L)

39
accounting for other possible covariates (e.g. sex, age, hemoglobin). For all statistical
tests, multivariate regressions were accomplished using JMP® Pro 14.0.0

40

Results
Subject characteristics
Subject demographics are summarized in Table 4-1 Subject characteristics and
hematological parameters. Thirty-six subjects participated in the study. However, in
contrast with the previously published study (Shah et al.,2020), we selected only 30
subjects (15 non-SCD and 15 SCD) for analysis, excluding those datasets that contained
low-quality or artefactual blood pressure recordings. Two of the 30 subjects had unusable
blood pressure recordings during the pain anticipation task. There was no difference in
age between the two groups; although, the median age of the SCD group was slightly
higher than the non-SCD group. There was no difference in state anxiety and trait anxiety
scores between groups. The SCD subjects were predominantly SS patients and only two
subjects were SC and S 
+
. The difference in hemoglobin between groups was expected.
The percentage of hemoglobin S (HbS%) was considered to be zero in control subjects
with sickle trait as HbS% does not contribute to sickling process in sickle trait. Seven
(50%) SCD subjects were on chronic transfusion, six (36%) were being treated with
hydroxyurea and two (14%) were not receiving any treatment.

41

Table 4-1 Subject characteristics and hematological parameters
Non-SCD (N=15) SCD (N=15) P-value
Diagnosis
Healthy 11
Sickle cell trait 4
Homozygous SS 13
SC 1
S + thalassemia 1
-
Sex (M/F) 9/6 7/8 0.43
Age (years)* 15.93 (8.29) 21.61(13.05) 0.23
Hemoglobin (g/dL)* 12.8 (2.4) 9.6 (1.95) <0.0001
Hemoglobin S(%) - 53.1 (7.42) -
Normally distributed data are shown as mean(SE) with p-value from Student t-test. Non-
normally distributed data, indicated by *, are shown as median (IQR) with p-value from
Wilcoxon Rank Sum test. Bolded p-values indicate significance (p<0.05)

42
Baseline cardiovascular and autonomic indices
The baseline cardiovascular and autonomic descriptors for all subjects studied are
displayed in Table 4-2. The baseline mean PPGa in the SCD group was approximately
20% lower than that in the non-SCD group (p=0.006). BRSv in SCD subjects was on
average about 27% lower compared to non-SCD (p=0.03). v in SCD also tended to be
lower, but the group difference did not attain significance. On the other hand, BRSv was
strongly correlated with v (R = 0.83, p<0.0001). Similarly, BRSc was significantly
correlated with c (r=0.64, p=0.0001). There was no significant group difference in BRSc,
although there was a tendency for it to be slightly lower in SCD subjects. There was no
correlation between BRSv and BRSc.

43

Table 4-2. Baseline cardiovascular and autonomic descriptors
Cardiovascular/Autono
mic Descriptor
Non-SCD (N=15) SCD (N=15) P-value
RRI (sec) 0.91 (0.044) 0.84 (0.038) 0.11
SBP (mmHg) 117.16 (4.03) 115.84 (5.00) 0.42
PPGa (nu) 0.72 (0.033) 0.57(0.047) 0.006
DBP (mmHg) 71.68 (2.73) 70.58 (2.85) 0.39
HFPRRI * 0.00093 (0.002) 0.00089 (0.001) 0.44
BRSc (s mmHg
-1
) 0.0246 (0.0031) 0.0200 (0.0026) 0.13
BRSv (nu mmHg
-1
) 0.0325 (0.0038) 0.0236 (0.0022) 0.03
c (s mmHg
-1
)
0.0119 (0.0015) 0.0096 (0.0010) 0.11
v (au mmHg
-1
)
0.0328 (0.0032) 0.0273(0.0021) 0.08
Normally distributed data are shown as mean(SE) with p-value from Student t-test. Non-
normally distributed data, indicated by *, are shown as median (IQR) with p-value from
Wilcoxon Rank Sum test. Bolded p-values indicate significance (p<0.05).

44
As displayed in Figure 4-2A, we found moderate-to-strong correlations between
baseline mean PPGa and BRSv in non-SCD subjects (R=0.59, p=0.02) and SCD subjects
(R=0.89, p<0.0001). In the SCD group, the slope between baseline mean PPGa and
BRSv was 3.7 times higher than that of the non-SCD group (p=0.0005). Given that a
smaller mean PPGa represents a more vasoconstricted baseline state, the lower BRSv
would indicate a state of higher sympathetically-mediated vascular tone.
Parallel to this finding, baseline RRI was strongly correlated to BRSc in SCD
subjects (R=0.82, p=0.0002) but only moderately correlated in non-SCD subjects (R=0.60,
p=0.02) (Fig.1B). However, the slopes between baseline mean RRI and BRSc were not
different between two groups. BRSc was also strongly correlated with HFPRRI in both
controls (R=0.94, p<0.0001) and SCD subjects (R=0.82, p=0.0003).

45
The association of cardiac baroreflex sensitivity with mean RRI and HFPRRI suggests
that BRSc provides a representative measure of vagal tone.

Correlation analysis between baseline autonomic indices and the
magnitude of vasoconstriction
Figure 4-3 shows an example of vasoconstriction and changes in other
cardiovascular variables to MTS in one SCD subject. Once the first task began, there was
a significant drop in the mean PPGa, indicating vasoconstriction. There were small
increases in blood pressure and heart rate (or equivalently decreases in RRI).
(B) (A)
Figure 4-2: (A) Correlation analyses between baseline BRSv and ∆PPGa: the
relationship between baseline BRSv and ∆PPGa is stronger in SCD subjects (p=0.0005).
(B) Correlation analyses between baseline BRSc and ∆RRI. No interaction was observed
between baseline BRSc

46

Figure 4-3: An example of cardiovascular variables during the mental stress protocol of
a single subject. “Tasks” (top panel) displays the output of the E-prime software where
the height of the bars represents the difficulty of the task.

47
Figure 4-4 displays the relations between baseline BRSv and ∆PPGa stratified by
groups. In SCD, there was a significant correlation between baseline BRSv and ∆PPGa
in N-back (R=0.76, p=0.0011), Stroop (R=0.77, p=0.0008) and the PA (R=0.65,
p=0.0089). In non-SCD, the strongest correlation was found during Stroop (R=0.71,
p=0.0031). The correlations between baseline BRSv and ∆PPGa were moderate during
N-back (R=0.51, p=0.05) and PA (R=0.51, p=0.05). The effect of baseline BRSv on
∆PPGa was similar between SCD and non-SCD subjects.

(A) (B) (C)
Figure 4-4: Correlation analyses between baseline BRSv and ∆PPGa in SCD and non-
SCD during N-back (A), Stroop (B) and pain anticipation (C).

48
Effect of mental stress on vasoconstriction after adjusting for
baseline vascular baroreflex sensitivity

Figure 4-5: The bar graphs show magnitude of vasoconstriction (mean SE) in SCD and
non-SCD during MTS before (A) and after (B) adjusting for baseline BRSv.
Figure 4-5A. shows the significance of stress tasks and diagnosis on the
magnitude of vasoconstriction before adjusting for baseline BRSv. Linear mixed model
analysis showed that the PA task induced greater vasoconstriction than N-back
(p<0.0001) and Stroop (p=0.007) regardless of diagnosis. In cognitive tasks (N-back,
Stroop), the absolute task difficulty did not affect the magnitude of vasoconstriction. When
considering all the three tasks, the magnitude of ∆PPGa tended to be lower in the SCD
subjects but the diagnosis effect was not significant in the mixed model analysis (p=0.05).
However, when only the PA task was considered, the SCD subjects had lower responses
than the controls (p=0.01). Age, gender and hemoglobin were not related to the
magnitude of the vasoconstriction.
Figure 4-5B displays the magnitudes of vasoconstriction in SCD subjects and
controls, in response to the 3 mental tasks, after adjusting for the effect of baseline BRSv.
N-back Stroop Pain Anticipation
0
0.1
0.2
0.3
0.4
0.5
Non-SCD
SCD
N-back Stroop Pain Anticipation
0
0.1
0.2
0.3
0.4
0.5
Non-SCD
SCD
(A) (B)
Tasks: p < .0001
Diagnosis: p=0.05
Tasks*Diagnosis: p=0.22
Tasks: p < .0001
Diagnosis: p=0.39
Tasks*Diagnosis: p=0.23

49
When adjusted for baseline BRSv, there was no significant difference in ∆PPGa between
SCD and non-SCD groups over the three tasks. Eliminating differences in BRSv across
the subjects led on average to a reduction in the vasoconstriction response to each task
in the non-SCD controls, but an increase in vasoconstriction response in the
corresponding task in SCD subjects. Thus, a higher BRSv was associated with a stronger
vasoconstriction response, while subjects with lower BRSv had weaker vasoconstriction
responses.

50

Discussion
Role of the baroreflexes in modulating the cardiovascular responses
to mental stress
During mental stress, sympathetic outflow from the higher centers to the heart and
vasculature leads to increased heart rate, peripheral vasoconstriction and consequently
elevation in arterial blood pressure. These responses have been well documented in the
literature (Callister et al., 1992; Hoshikawa and Yamamoto, 1997). From a theoretical
perspective, the elevated blood pressure would stimulate the baroreceptors and produce
compensatory changes in heart rate and cardiac contractility, as well as vascular
resistance, which in turn lower blood pressure to a level consistent with the negative
feedback. If this was the only factor involved, we would have expected a negative
relationship between baroreflex sensitivity and the magnitude of vasoconstriction
resulting from MTS, since a higher BRSv would have triggered a compensatory
vasodilation to minimize the increase in blood pressure. Instead, the primary finding in
this study was that the individuals with higher baseline BRSv elicited a larger
vasoconstriction response to MTS in both subject groups. Thus, secondary baroreflex
adjustments brought about by the primary effects of sympathetic outflow elicited by MTS
were most likely not the predominant physiological mechanism responsible for the
observed vasoconstriction responses to MTS.
An alternative hypothesis is that the strong association we observed between
BRSv and the magnitude of vasoconstriction reflected an indirect, rather than direct, role

51
of the sympathetic baroreflex. There are several pieces of evidence to support this
argument. Most importantly, BRSv was strongly and positively correlated with baseline
PPGa, which we have taken to represent peripheral vascular conductance. For example,
subjects with low baseline BRSv tended to have low baseline PPGa, meaning that these
subjects were already vasoconstricted at baseline prior to the application of MTS. As such,
the magnitude of vasoconstriction (measured by the reduction in PPGa from baseline) for
the same increase in sympathetic neural drive would likely be smaller due to a smaller
reserve for further vasoconstriction, compared to those cases where the baseline PPGa
is larger. Therefore, baseline BRSv served as a marker for the pre-MTS level of
sympathetic vascular tone: the lower the BRSv, the higher the baseline vascular tone.
Reduced baseline vascular baroreflex sensitivity in SCD
The second important finding in this study was that the SCD subjects, on average,
had lower baseline BRSv than controls. As well, the SCD subjects had lower baseline
PPGa (Table 2 and Fig.1A). Applying the same reasoning as had been discussed in the
aforementioned section, we can only conclude that the SCD subjects had higher
sympathetic vascular tone at baseline. In humans, depressed cardiac baroreflex
sensitivity has been reported in other cardiovascular diseases with high sympathetic tone,
such as congestive heart failure (Osterziel et al., 1995). It is also well-documented that
sympathetic baroreflex sensitivity decreases in response to acute cardiovascular
stressors such as exercise or mental stress (Durocher et al., 2011; Dampney, 2017).
The possible causes for higher sympathetic tone in SCD subjects remain unclear.
Our previous work suggested that anxiety could be a factor (Shah et al., 2020); however,

52
we did not find a significant relation between anxiety score and baseline BRSv in this
work. The alteration in baseline BRSv could be due to chronic anemia (Martins et al.,
2012), but we found no significant correlation between hemoglobin and baseline BRSv.
Consistent with the literature, the majority of SCD subjects tend to have higher
sympathetic tone, making them more vasoconstricted at baseline than the non-SCD
controls. Despite a lower baseline BRSv in the SCD group, the slope between baseline
BRSv and the magnitude of vasoconstriction to mental stress was not different from non-
SCD subjects. Therefore, the association between baseline BRSv and vasoconstriction
was likely not altered by having SCD.
Variability of vasoconstriction responses resulting from differences in
source of autonomic stimulation
While we found that interindividual variability of BRSv contributes to the variability
in vasoconstriction responses across individuals, adjusting for differences in BRSv
actually further reduced the inter-group differences in MTS-induced vasoconstriction
responses (Fig.4B). This stands in contrast with the enhanced vasoconstriction
responses found in SCD subjects during thermal pain stimuli (Khaleel et al., 2017;
Veluswamy et al., 2020). Since there is increasing evidence that dysautonomia in SCD is
associated with autonomic hyperresponsiveness, one would have expected the SCD
subjects with high sympathetic tone to have displayed stronger vasoconstriction
responses to MTS. This contradiction may be related to the different stressors used to
elicit autonomic-mediated responses. The autonomic responses to MTS are driven
predominantly by sources that have a central origin, unlike the corresponding responses
to other types of stimulation that are applied peripherally (Carter and Goldstein, 2011;

53
Dampney, 2015). Thermal pain elicits changes in sympathetic outflow via peripheral
receptors (Morrison, 2001) while MTS triggers sympathetic outflow via central commands
(Dampney, 2015). It is unknown whether autonomic hyperresponsiveness, if it does occur,
occurs at the efferent pathways or afferent pathways of the ANS. In our previous study
where the autonomic stimulus was head-up tilt, we also found no differences in
vasoconstriction magnitude between SCD subjects and controls (Chalacheva et al.,2019).
In the cardiovascular response to orthostatic stress, it is known that, aside from
baroreflex-mediated adjustments, a strong determinant of the peripheral vasoconstriction
is increased central sympathetic drive from the vestibular system (Carter and Ray, 2008).
Estimation of vascular baroreflex sensitivity
In this chapter, we employed a method for assessing vascular baroreflex sensitivity
using the spontaneous fluctuations in diastolic blood pressure and PPGa that is
analogous to the sequence technique commonly used to estimate cardiac baroreflex
sensitivity. Porta et al. (2018) were the first to extend the sequence method for application
to the vascular baroreflex; however, they used skin blood flow, measured through laser-
doppler flowmetry, in conjunction with noninvasive arterial pressure as the surrogate
index of peripheral vascular conductance. Previous work (Rauh et al., 2003; Khoo and
Chalacheva, 2019) has shown that changes in PPGa are correlated with the
corresponding changes derived from laser-doppler flow. Just as the cardiac baroreflex
sensitivity estimates derived from the sequence technique have been found to correlate
closely with the corresponding estimates deduced from the “spectral technique” (Parati
et al., 2000; Perrson et al., 2001), we found our estimates of BRSv from both methods to
also be strongly correlated. However, it should be emphasized that these estimates of

54
BRSv differ significantly from previous studies that have used muscle sympathetic nerve
activity (MSNA) to quantify sympathetic baroreflex output, since those studies do not take
the vascular aspect of the entire baroreflex arc into account. The reflex response of MSNA
to changes in arterial pressure may not necessarily induce a vascular response in a linear
fashion. For instance, there was no correlation between MSNA burst incidence and the
percentage change in leg vascular conductance in young women (Hissen et al., 2019;
Robinson et al., 2019). Accordingly, a high sympathetic baroreflex gain derived using
neural drive as the output may not linearly translate into a high value of BRSv, since
mechanical and physical constraints in the blood vessels could result in a saturation of
the transfer relation between the neural and vascular interfaces. SCD subjects have other
factors like decreased nitric oxide availability and high levels of endothelin-1 making the
vessels constrict more given the same neural input (Ergul et al., 2004; Akinsheye and
Klings, 2010). Thus, at the highest levels of vascular tone, increasing sympathetic drive
further may lead to further increases in MSNA but no additional increases in
vasoconstriction, hence resulting in the low BRSv values under these conditions. To the
best of our knowledge, this is the first study ever to report measurements of vascular
baroreflex sensitivity in subjects with SCD.

55
Conclusion
In this chapter, we examined in greater detail the source of variability in the
vasoconstriction responses of SCD and controls to mental stress. We assessed both
cardiac and vascular arms of the baroreflex in order to quantify autonomic function during
the baseline period prior to application of mental stress. Baseline BRSc was not different
between SCD and non-SCD subjects, was not correlated with BRSv, and was not
associated with the vasoconstriction responses to mental stress tasks. BRSv in both
groups was correlated with mean PPGa, and since both baseline PPGa and BRSv were
lower in SCD, these findings suggested that the SCD subjects were in a basal state of
higher sympathetically-mediated vascular tone.
After adjusting for differences in BRSv within and between groups, we found no
difference in the vasoconstriction responses to all 3 mental tasks between SCD and non-
SCD subjects. The implications of these findings are significant in SCD since
vasoconstriction reduces microvascular flow and prolongs capillary transit time, thus
increasing the likelihood for VOC to be triggered following exposure to stressful events.

56
CHAPTER 5 THE RELATIONSHIPS BETWEEN
AUTONOMIC INDICES AND PATTERNS OF
VASOCONSTRICTION DURING MENTAL STRESS
During mental stress, we observed great variability in the patterns of the
vasoconstriction responses across individuals and the subject groups. For example,
Figure 5-1 shows that subject 3 had prolonged vasoconstriction throughout the Stroop
task whereas subject 27 had short recoveries during that period. In contrast, subject 16’s
blood flow fluctuated drastically. If these subjects were SCD patients, we would predict
that subject 3 would have higher chance of getting VOC because blood flow was reduced
for the longest period of time.

57

Figure 5-1: PPGa time-series of three subjects during baseline and Stroop. Subject 3
had sustained vasoconstriction during Stroop. Subject 16 exhibited large fluctuations in
the PPGa response during Stroop.
These patterns of vasoconstriction were also apparent in other studies.
Veluswamy et al. 2020 showed that the exposure to “cold stimuli” caused progressive
vasoconstriction in some SCD subjects. In particular, microvascular blood flow did not
recover after the stimuli were removed. In some other patients, blood flow nearly
recovered to the baseline level at the end of the cold application.
Tilting a subject from a supine to near-upright position induces a transient drop in
blood pressure that is normally followed by tachycardia and peripheral vasoconstriction
(Chalacheva et al., 2019). In this study, while 29.6% of SCD subjects displayed normal
responses, 37% responded with peripheral vasoconstriction only. Further analyses
confirmed that, this was specifically due to the loss of baseline parasympathetic activity.
0 100 200 300
0
0.5
1
1.5
2
PPGa[nu]
Baseline
0 100 200 300
0
0.5
1
1.5
2
PPGa[nu]
0 100 200 300
Time[sec]
0
0.5
1
1.5
2
PPGa[nu]
0 100 200 300 400 500 600 700 800
0
0.5
1
1.5
2
Stroop
0 100 200 300 400 500 600 700 800
0
0.5
1
1.5
2
0 100 200 300 400 500 600 700 800
Time[sec]
0
0.5
1
1.5
2
Subject 3
Subject 27
Subject 16

58
These evidences lead to the hypothesis that the patterns of vasoconstriction may
be associated to underlying autonomic function. In this chapter, we comprehensively
examined plausible connections between sustained vasoconstriction and autonomic
function during mental stress. Two following questions will be answered in this chapter.
1) How do we quantitatively measure sustained vasoconstriction?
2) Is sustained vasoconstriction related to autonomic measures during N-back
and Stroop?

59
Method
Measuring sustained vasoconstriction
As depicted in Figure 5-1, sustained vasoconstriction is indicated by a prolonged
vasoconstriction that lasts for the entire duration of the tasks. This response can be easily
identified. The challenge arises when the responses are deviated from being a sustained
one. In those cases, we would like to measure how close these responses are to
sustained vasoconstriction. To measure the degree of sustainability, we employ the
statistics called “block entropy” to quantify “randomness” or “uncertainty” in the data.
Basically, as the response becomes closer to sustained vasoconstriction, the block
entropy gets closer to zero. As a reference, sustained vasoconstriction has the block
entropy of zero.
Block Entropy
Given a symbol sequence X = x
1
, x
2
, … , x
N
, a block of size m is a segment of m
consecutive elements of the symbol sequence X, e.g. (x
i
, x
i+1
, … , x
i+m−1
). The block
entropy of X for blocks of size m is defined by the following:
BlockEn (X, m) = − ∑ p(x
i
, x
i+1
, … , x
i+m−1
) log
2
p(x
i
, x
i+1
, … , x
i+m−1
)
N−m+1
i=1
(5-1)
where p(x
i
, x
i+1
, … , x
i+m−1
) is the probability of occurrence of (x
i
, x
i+1
, … , x
i+m−1
).
In this context, p(x
i
, x
i+1
, … , x
i+m−1
) is calculated as follows:

60

p(x
i
, x
i+1
, … , x
i+m−1
) =
No. of blocks of the form (x
i
, x
i+1
, … , x
i+m−1
)
N − m + 1
(5-2)

61
For example, if X = 1 2 3 4 5 6 5 4 3 2 1 2 3 4 5 6 5 4 3 2 1, the blocks of size 3 in this
sequence are
Block (1,2,3) (2,3,4) (3,4,5) (4,5,6) (5,6,5) (5,4,3) (4,3,2) (3,2,1) (2,1,2) (6,5,4)
Occurrence 2 2 2 2 2 2 2 2 1 2
𝐩 (𝐱 𝐢 , . . , 𝐱 𝐢 +𝐦 −𝟏 ) 0.105 0.105 0.105 0.105 0.105 0.105 0.105 0.105 0.053 0.0125
Hence, BlockEn (X, 3) = 8 ∗ 0.105log
2
0.105 + 0.053log
2
0.053 + 0.0125log
2
0.0125 =
2.2878.
Symbolization process
The whole recording of a PPGa time series is transformed into a symbolic
sequence by following the procedure detailed below:
Step 1: A PPGa signal is uniformly sampled at 1 Hz. After, the signal artifacts are
removed using a moving median filter. Then, the signal is normalized to its 95
th

percentile.
Step 2: The difference between the minimum and the maximum of the PPGa signal is
divided into 6 bins of equal size, l. The PPGa values are transformed according to their
respective bin numbers.

62
S
i
={
1: min (x
i
) ≤ x
i
< 1 ∙ l
2: 1 ∙ l ≤ x
i
< 2 ∙ l
⋮
6: 5 ∙ l ≤ x
i
< 6 ∙ l
(5-3)
The bin size, l is unique for each subject because it is calculated based on the minimum
and the maximum of each subject’s PPGa signal.
Block size and the number of bins

Figure 5-2: The block entropies calculated from PPGa responses during Stroop. m
indicates the block size. Each line represents each subject.
The choice of m and the number of bins is dependent on the data. For a short
time-series, it is common to use a block of size 2 or 3. After comparing multiple values
of m and the number of bins, we found that the block entropy monotonically increased
with m and the number of bins (Figure 5-2). In this analysis, the block size was 3 and
the number of bins was 6.
4 6 8
0
1
2
3
4
5
BlockEn
m =1
4 6 8
0
1
2
3
4
5
BlockEn
m =2
4 6 8
No. of bins
0
1
2
3
4
5
BlockEn
m =3
4 6 8
No. of bins
0
1
2
3
4
5
BlockEn
m =4

63
Assessment of autonomic responses
In each subject, autonomic functions were evaluated using the estimates of the
baroreflex sensitivities, the degree of baroreflex involvement and the power spectra of
RRI, SBP, DBP and PPGa signals. All autonomic indices were assessed during the
period of baseline, N-back and Stroop period.
Baroreflex sensitivities
Both the sequence and spectral technique were used to determine the cardiac and
vascular baroreflex sensitivities. In the sequence method, the average slope of the
regression line between RRI and SBP was taken as a measure of cardiac baroreflex
sensitivity (BRSc). In the same way, the average slope of the regression line between
PPGa and DBP was taken as a measure of vascular baroreflex sensitivity (BRSv). In the
spectral method, the cardiac baroreflex sensitivity ( c) was calculated as the square root
of the ratio between the low frequency power of RRI and SBP, while the vascular
baroreflex sensitivity was calculated as the square root of the ratio between the low
frequency power of PPGa and DBP (for more detail, see Chapter 4).
Baroreflex involvement
To evaluate the degree of baroreflex involvement, we used the approach proposed
by (Porta et al., 2018). The degree of cardiac baroreflex involvement was represented
by the percentage of RRI-SBP sequences identified during each task. In particular, this
index , SEQ%BRSc, was defined as

SEQ%
BRSc
=
No. of RRI− SBP sequences
N − L + 1
× 100
(5-4)

64
In a similar fashion, the degree of vascular baroreflex involvement , SEQ%BRSv, was
defined as

SEQ%
BRSv
=
No. of PPGa− DBP sequences
N − L + 1
× 100
(5-5)
In both equations, N represents the signal length, and L represents the sequence
length. Both SEQ%BRSc and SEQ%BRSc range between 0-100, where 0 indicates the
absence of the RRI-SBP sequence that changes in the same direction, while 100
indicates that all possible sequences are of the baroreflex origin.

65
Power spectral indices
The beat-to-beat values of RRI, SBP, DBP and PPGa were first converted into
uniformly sampled time series, with a sampling frequency of 1 Hz. Then, these signals
were divided into three segments according to three experimental phrases – baseline, N-
back and Stroop. In each segment, the mean value was subtracted, and the power
spectrum was subsequently computed using the Welch method (Stoica and Moses, 2005).
This involved dividing the segment into smaller 5-minute segments with 50% overlap, and
taking the average of power spectra among them. The areas under the computed power
spectra reflect the modulations by different mechanisms. The high frequency band (HF:
0.15 to 0.4 Hz) of RRI variability corresponds to the modulations of vagal tone, and is
strongly linked to respiration (Saul 1990; Malliani et al. 1991). The low-frequency band
(LF: 0.04 to 0.15 Hz) of SBP variability reflects the combination of intrinsic vasomotor
rhythmicity and sympathetic modulation of blood vessels (Ryan et al. 2011; Pagani et al.
2012). The low-frequency band (LF: 0.04 to 0.15 Hz) of PPGa variability reflects local
and sympathetic control over peripheral circulation (Chan et al. 2012; Bernardi et al. 1996).
In summary, we extracted HFPRRI , LFPSBP, and LFPPPGa during baseline, N-back and
Stroop in each subject.
The frequency powers of SBP and PPGa variability were also expressed in
normalized units after divided by the total power (0-0.5Hz), then multiplied by 100. The
normalized low frequency powers of SBP and PPGa variability were denoted as LFPnu
SBP and LFPnu PPGa, respectively.

66
Statistical Analysis
A nonparametric Friedman’s ANOVA test for repeated measures was used for the
comparison of block entropies during baseline, N-back and Stroop. When a significant
change was detected, a post hoc Wilcoxon sign rank test was performed between the
baseline period and the task period. Linear mixed model analysis was used to test for
the significance of stress tasks and diagnosis on the changes in BRSc, BRSv, c, v,
SEQ%BRSc, SEQ%BRSc, HFPRRI , LFPSBP, LFPPPGa , LFPnu SBP, LFPnu PPGa and mean
cardiovascular variables. Any non-normal continuous variables were square-root
transformed before used in the mixed model analysis. The comparisons of dichotomous
variables were tested using Pearson’s Chi-squared test. All statistical tests were
accomplished using JMP® Pro 14.0.0.

67
Results
Block entropy of PPGa time-series
Figure 5-3 (the upper row) shows an example of sustained vasoconstriction during
mental stress of one healthy subject. It can be seen that despite some variations, the
PPGa values stayed below 0.3 for the whole tasks.
Figure 5-3 (bottom row) shows the quantization of this signal. This process
converted the original values into their corresponding bin numbers. During Stroop, we
can see that the new sequence was mostly composed of 1. The symbols emphasized
that most of its original values were constrained within the first interval. During the tasks,
the block entropy values were 18 times and 55 timers lower than the values during
baseline. This shows that the block entropy became lower as the response plateaued.

68

500 1000 1500 2000 2500 3000
0.2
0.4
0.6
0.8
1
1.2
1.4
PPGa[nu]
500 1000 1500 2000 2500 3000
Time(s)
1
2
3
4
5
6
Symbol
BlockEn
= 2.78
BlockEn = 0.05 BlockEn = 0.15
Stroop N-back Baseline
1
3
4
5
6
Figure 5-3: (Top) Example of a sustained vasoconstriction to mental stress. (Bottom)
Example of the discretized PPGa time-series using 6 quantization levels.
1
2
3
4
5
6

69

Figure 5-4: Application of block entropy on non-sustained signals during Stroop. Both
signals were from different SCD subjects. These signals had the same variability as
quantified by the standard deviation of 0.08. However, the values of block entropy were
different.
Figure 5-4 shows the application of block entropy on non-sustained signals during Stroop.
Both signals had a standard deviation of 0.08. However, the left sequence had lower
block entropy. This was because there was less variety in the symbols. We can see that
its symbols were mostly confined to the first bin.
In these presented data, there were strong associations between the values of
block entropy and the number of blocks during N-back (R=0.90; p<.0001) and Stroop
(R=0.93; p<.0001). Figure 5-5 demonstrates that the block entropy depended upon the
number of unique blocks. When the block entropy was zero, the response was completely
sustained (subject 9).
1200 1400 1600 1800 2000 2200 2400
0.2
0.4
0.6
0.8
1
1.2
PPGa[nu]
1200 1400 1600 1800 2000 2200 2400
Time(s)
1
2
3
4
5
6
Symbol
1200 1400 1600 1800 2000 2200 2400
0.2
0.4
0.6
0.8
1
1.2
1.4
PPGa[nu]
1200 1400 1600 1800 2000 2200 2400
Time(s)
1
2
3
4
5
6
Symbol
std = 0.10 std = 0.10
BlockEn = 0.80 BlockEn = 1.81
Stroop Stroop

70

Figure 5-5: Association between the number of unique blocks and block entropy during
Stroop. The yellow rectangle indicates the subject with sustained vasoconstriction. The
blue box indicates the subject with non-sustained vasoconstriction.
Task
Stroop
BlockEn
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
R ²: 0.873
F(1,29)=200.25, PValue=<.0001
0 10 20 30 40
No. Unique Blocks
Non-SCD
SCD
BlockEn
1300 1400 1500 1600 1700 1800 1900 2000
Time(s)
1
2
3
4
5
6
Symbol
1400 1600 1800 2000
Time(s)
1
2
3
4
5
6
Symbol
subject 9
subject 16

71
Relationships between block entropy and mental stress
As shown in Figure 5-6, there was a significant decrease in block entropy during
N-back (p=0.0003) and Stroop (p=0.0059). While the baseline block entropy was
significantly lower in SCD subjects, there was no difference in the block entropy between
the two groups during two cognitive tasks. Plotting individual changes revealed
substantial inter-individual variability within the SCD and control subjects (Figure 5-7). In
the control subjects, the SE of block entropy was 0.12 at baseline but increased more
than twice during N-back and Stroop. There were a few subjects who showed extremely
low values (BlockEn<0.5) during the cognitive tasks. Similarly, there was one SCD
subject who had zero entropy during Stroop. During the tasks, the standard errors of
block entropy were similar between the two groups. This suggests that the variability in
block entropies was unlikely related to the disease. The values of block entropy were not
associated with age, sex, hematocrit or percentage of HbS.

72

Figure 5-6: Mean  SE of block entropy during baseline, N-back and Stroop in controls
(n=15) and SCD subjects (n=15).

Non-SCD SCD
1.4
1.6
1.8
2
2.2
2.4
2.6
2.8
BlockEn
Baseline
Non-SCD SCD
N-back
Non-SCD SCD
1.4
1.6
1.8
2
2.2
2.4
2.6
2.8
Stroop
p=0.006

73

Figure 5-7: (Top) Changes in the block entropy from baseline to N-back. (Bottom)
Changes in the block entropy from baseline to Stroop. The error bars indicate mean 
SE of block entropy.
Baseline N-back
0
1
2
3
4
BlockEn
Non-SCD
Baseline N-back
0
1
2
3
4
SCD
Baseline Stroop
0
1
2
3
4
BlockEn
Baseline Stroop
0
1
2
3
4

74
Effects of mental stress on the autonomic nervous system
During mental stress, mean PPGa and RRI decreased significantly from the
baseline period (Table 5-1). Similarly, mean SBP and DBP increased significantly from
the baseline period.
Table 5-2 compared baroreflex sensitivities among three conditions – during
baseline, N-back and Stroop. We found that both BRSv and v significantly decreased
during the cognitive tasks. However, there was no difference in the changes between
thee controls and SCD subjects. BRSc and c did not change in response to mental
stress. There was no difference in cardiac baroreflex sensitivities between two groups.
The results show that mental stress only affected the vascular arm of baroreflex. In
addition, its effects on the vascular baroreflex were the same for both groups.
In Table 5-3, power spectra of RRI, SBP and PPGa variability were compared
under three conditions. All absolute powers exhibited significant changes from the
baseline period. HFPRRI dropped significantly during N-back (p=0.0074) and more during
Stroop (p<.0001). There was a significant reduction in LFPPPGa during N-back (p=0.0006)
and Stroop (p<.0001). Similarly, LFPSBP was significantly decreased during Stroop
(p=0.02). The group difference was not detected in HFPRRI, LFPSBP, LFPPPGa.
In both groups, the degree of baroreflex involvement, as indicated by SEQ% BRSc
and SEQ% BRSv did not change under the stress conditions. The degree of baroreflex
involvement was around 4% for the vascular arm and 10% for the cardiac arm (Table 5-
2). The normalized low-frequency powers, LFPnu SBP, LFPnu PPGa, were similar across

75
three conditions. These findings suggest that the involvement of underlying mechanisms
did not change during mental stress.

76

Table 5-1 Mean cardiovascular variables during cognitive tasks
Baseline N-back Stroop Significance
of Tasks
Non-SCD SCD Non-SCD SCD Non-SCD SCD
PPGa
(nu)
0.68(0.04) 0.57(0.05)
¶
0.46(0.06) 0.47(0.03) 0.45(0.06) 0.39(0.03) p<.0001
SBP
(mmHg)
120(4.7) 116(5) 121(4.9) 115(4.7) 121(5) 118(3.9) p=0.01
DBP
(mmHg)
73(2.8) 71(3.1) 74(3.2) 71(2.6) 75(3.6) 73(2.2) p=0.0089
RRI (s) 0.91(0.04) 0.84(0.04) 0.86(0.04) 0.79(0.03) 0.84(0.04) 0.79(0.03) p<.0001
Mean(SE).
¶
indicates p<0.05 for non-SCD vs SCD. The last column indicates the
significance of the fixed effect “task” in the mixed model analyses.

77

Table 5-2 Comparison of baroreflex indices between controls and SCD subjects during cognitive tasks
Baroreflex
indices
Baseline N-back Stroop Significance
of Tasks
Non-SCD SCD Non-SCD SCD Non-SCD SCD
BRSv (nu mmHg
-1
) 0.021(0.0025) 0.017(0.0018) 0.014(0.0016) 0.014(0.0013) 0.013(0.0016) 0.013(0.0015) p=0.0041
BRSc (s mmHg
-1
) 0.019(0.0022) 0.015(0.0023) 0.019(0.0027) 0.015(0.0019) 0.018(0.0027) 0.014(0.0023) p=0.20
v (nu mmHg
-1
)
0.027
(0.0037)
0.023
(0.0022)
0.02 (0.0032) 0.02 (0.0017) 0.018(0.0024) 0.018(0.0021) p=0.0024
c (s mmHg
-1
)
0.012(0.0015) 0.009(0.0011) 0.012(0.0015) 0.01(0.0012) 0.012(0.0015) 0.01 (0.001) p=0.59
SEQ% BRSv 4.26(0.58) 4.03(0.65) 4.19(0.79) 4.11(0.85) 3.10(0.44) 3.93(0.74) p=0.14
SEQ% BRSc 10.32(1.94) 10.46(1.88) 9.85(1.50) 10.27(1.42) 10.79(1.53) 8.84(1.17) p=0.97
Mean(SE). None of these parameters were different between controls and SCD. The last column indicates the
significance of the fixed effect “task” in the mixed model analyses.

78

Table 5-3 Comparison of spectral indices between controls and SCD subjects during cognitive tasks
Spectral indices Baseline N-back Stroop Significance of
Tasks
Non-SCD SCD Non-SCD SCD Non-SCD SCD
HFPRRI *(s
2
) 0.00054
(0.0016)
0.00042
(0.0009)
0.00038
(0.001)
0.00035
(0.0005)
0.00035
(0.0011)
0.00025
(0.0003)
p<.0001
LFPSBP * (mmHg
2
) 4.43(5.03) 3.86(3.46) 4.39(3.30) 4.20(3.32) 3.91(2.87) 3.93(2.99) p=0.03
LFPPPG * (nu
2
) 0.0028
(0.0029)
0.0025
(0.00216)
0.001
(0.0014)
0.0014
(0.0008)
0.0008
(0.0013)
0.001
(0.001)
p<.0001
LFPnu SBP * (%) 30(21.3) 30.8(22.3) 28.2(17.2) 25.4(19.3) 26(16.3) 23.9(14.5) p=0.55
LFPnu PPG * (%) 10.3(13.7) 21.4(10.2)
¶
13.2(5.8) 15.3(8.6) 15.7(6.2) 15(10.7) p=0.99
Median(IQR).
¶
indicates p<0.05 for non-SCD vs SCD. The last column indicates the significance of the fixed effect “task”
in the mixed model analyses.

79

Relationships between sustained vasoconstriction and the ANS
measures
Because the values of block entropy indicate the sustainability of the PPGa signals,
we tested for the linear relationships between block entropies and the ANS parameters
during the cognitive tasks. Pearson’s correlation analyses revealed that block entropies
were positively correlated with mean PPGa and LFP-PPGa in N-back and Stroop (Figure
5-8, 5-9). There was no interaction between the subject group and mean PPGa. Similar
relationships were observed for BRSv; however, the relationships appeared to be
stronger in the controls (Figure 5-10, R=0.81, p=0.0002) than in the SCD group (Figure
5-10, R=0.52, p=0.06). Further analysis showed that the interaction between the subject
group and BRSv was significant in N-back (p=0.02). However, when the outlier was
removed (subject 25), the interaction did not reach significance. Significant interaction
was not detected in Stroop. As a consistency check on whether the relationships between
block entropies and the vascular baroreflex were comparable between groups, we tested
for the effect v on the block entropy and found a strong correlation between block
entropies and v in N-back (R=0.56, p=0.0012) and Stroop (R=0.60, p=0.0005). We did
not find the significant interaction between v and the subject group. None of the positive
relationships was present for the parasympathetic indices i.e. BRSc, c and HFP-RRI.
These findings show that low values of block entropies were associated with low values
of mean PPGa and the vascular baroreflex sensitivity.

80
The significant relationships between block entropies and the vascular baroreflex
sensitivity prompted us to investigate further if the degree of vascular baroreflex
involvement was related with block entropies. We found that SEQ%BRSv was not
associated with block entropies in N-back and Stroop (Figure 5-11). In particular, the
subjects with low block entropies were accompanied by low BRSv but not low SEQ%BRSv
(Figure 5-12). The values of SEQ%BRSv were about 4% in most subjects. Similarly, the
correlation between LFPnu PPG and block entropies was weak in N-back (R=0.28, p=0.033)
and not significant in Stroop (R=0.18, p=0.33) (Figure 5-13).

81
Figure 5-8: Pearson’s correlation revealed positive relationships between block
entropies and mean PPGa.

Figure 5-9: Pearson's correlation revealed positive relationships between block
entropies and LFPPPG.

82
Figure 5-10: Pearson's correlation between block entropies and BRSv in non-SCD and
SCD subjects. In the SCD group, “smibr” was excluded from N-back and “wasjy” was
excluded from Stroop.

83
Figure 5-11: Pearson’s correlation showed that there were no relationships between
SEQ% BRSv and block entropies.

Figure 5-12: Pearson’s correlation showed no relationships between BRSv and SEQ%
BRSv.

84
Figure 5-13: Pearson’s correlation showed that there were no relationships between
LFPnu PPG and block entropies.

85
Discussion
The basic mechanism of vaso-occlusions in SCD stems from the polymerization
of deoxy-HbS that rigidifies red blood cells. In sickled form, these cells can become
lodged in microvasculature, blocking blood flow and potentially lead to acute vaso-
occlusive pain episodes. Therefore, any factors that increase the microvasculature transit
time of RBCs increase the chance for vaso-occlusions.
During mental stress, the patterns of vasoconstriction across SCD and control
individuals were apparent. Among these patterns, we speculate that sustained
vasoconstriction is the riskiest condition for SCD patients to be in because peripheral
blood flow is reduced for the longest period of time compared to other types of the
responses. Therefore, in the present study, we specifically examined if this type of the
responses can be uniquely identified by any of underlying autonomic mechanisms.
We have applied the “block entropy” method to quantify the sustainability of the
responses during mental stress. The block entropies calculated from the task periods
were significantly lower than that of the baseline period. Also, SCD individuals did not
necessarily have lower block entropies than the controls. From a methodological point of
view, low block entropies (e.g. BlockEn < 1) indicate “regularity” or “predictability” in the
data. In other words, there is not much variety in the symbolic sequence, making the
constructed blocks repetitive. This also indicates that the fluctuations in peripheral blood
flow (as indicated by the PPGa signals) are more restricted during mental stress; thus,
the sustainability increases.

86
It is very likely that a sustained vasoconstriction during mental stress is influenced
by increase in sympathetic tone rather than absence of underlying autonomic
contributions. There are major findings to support this statement. First, block entropies
were strongly correlated with the sympathetic measures such as mean PPGa, BRSv, and
LFPPPG as opposed to the vagal measures such as BRSc or HFPRRI. For example,
subjects with low BRSv tended to have low block entropies. This also indicates that for
the same change in blood pressure, the subsequent change in PPGa was smaller in the
subjects with lower block entropies.
When the degree of vascular baroreflex involvement was considered, we did not
find the difference in this parameter between the subjects with low entropies and high
entropies. Furthermore, when we compared the normalized low frequency power of
PPGa signals across the subjects, the percent of low-frequency oscillations in PPGa
(indicated by LFPnu PPGa) was similar between the subjects with low and high block
entropies. The lack of associations between block entropies, SEQ%BRSv and LFPnu PPGa
confirms that low block entropies are driven mainly by an increase in sympathetic tone
rather than the degree involvement of neural or local mechanisms.
While we thought that the SCD subjects with low block entropies would have worse
clinical outcomes, we did not find the relationships among block entropies, HbS, treatment
groups and the number of pain crises. We think our sample size is too small to detect
such relationships and future studies are needed to confirm this finding.

87
Conclusion
In the present work, we thoroughly determined the role of autonomic functions on
sustained vaso-constriction responses. Using the block entropy, we measured how close
the vasoconstriction is to a sustained vasoconstriction. We found that there was
substantial variability in the values of block entropy across individuals and subject groups.
Complete sustained vasoconstriction was elicited in few individuals and the only
parameters associated with this type of response were mean PPGa and vascular
baroreflex sensitivities. This suggests that sustained vasoconstriction is driven by an
increase in sympathetic tone.

88
CHAPTER 6 SUMMARY, LIMITATIONS AND FUTURE
DIRECTIONS
Summary and significant new findings
The primary focus of this work was to investigate whether autonomic nervous
system activity was an important determinant of the variability of vasoconstriction
responses to mental stress in SCD patients. In the majority of subjects, mental stress
caused reduction in the mean level of finger PPGa from the baseline period, indicating
peripheral vasoconstriction (Shah et al., 2020). The amount of reduction varied across
individuals, characterizing the inter-individual differences in the magnitude of the
responses. In the first part of this work, we addressed if these differences were
determined by baseline autonomic function. For each subject, baseline autonomic
function was assessed using the estimated cardiac and vascular baroreflex sensitivities.
We found that baseline BRSc was not related to the magnitude of the responses.
However, baseline BRSv was positively correlated with the magnitude of the responses.
In particular, the subjects who had lower baseline BRSv had weaker vasoconstriction.
Because lower baseline BRSv reflected higher sympathetic tone, this suggested that the
subjects who had higher sympathetic tone at rest had smaller room to vasoconstrict
during mental stress. This also explained why the magnitude of the responses tended to
be lower in the SCD group. Adjusting for the baseline BRSv reduced the difference in
magnitude of the responses between the two groups.

89
In the second part of this work, we focused on one unique pattern of the responses
characterized by prolonged vasoconstriction during tasks. For SCD patients, this pattern
may be of critical importance because prolonged vasoconstriction would mean prolonged
transit time, and consequently, an increase in the likelihood of generating vaso-occlusions.
To quantify sustained vasoconstriction, we used “Block Entropy” to measure how close
the response was to sustained response. For a complete sustained response, the block
entropy was zero. For each subject, block entropy values were extracted from the PPGa
signal during N-back and Stroop. We found that only 3-4 subjects had zero block entropy
indicating complete sustained vasoconstriction.
During mental stress, we also extracted cardiovascular and autonomic parameters
such as mean RRI, mean PPGa, BRSc, BRSv , c, v, HFPRRI, LFPPPGa. None of these
parameters except mean PPGa, BRSv and LFPPPGa were correlated with block entropy
values. Mean PPGa, BRSv and LFPPPGa had strong positive correlations with block
entropy values. The subjects who had zero block entropy had the lowest mean PPGa,
BRSv and LFPPPGa during the tasks. Since lower mean PPGa and BRSv reflected higher
sympathetic tone, these results suggested that sustained vasoconstriction was driven by
increase in sympathetic tone.
In summary, this work revealed that 1) baseline sympathetic tone affected the
variability in the magnitude of the vasoconstriction responses to mental stress 2)
sustained vasoconstriction during mental stress was driven by increase in sympathetic
tone and 3) none of these effects on the variability of the vasoconstriction responses were
different between SCD and control subjects.

90
Limitations
Although these findings may have shed some light on the fact that inter-individual
difference in sympathetic tone has a significant role on the vasoreactivity phenotypes
during mental stress, they were derived from a study with several limitations. First, our
sample size was small and the subjects studied were SCD patients who received regular
treatment and were fairly healthy at the time of the study. According to our records, only
two SCD patients had frequent painful crisis the year before the study. This prevented
us from detecting any clear significant associations between individual sympathetic tone
and VOC.
Second, the stressors used in this study may not be translated to a real-life stress.
Stress responses were reported to be dependent on individual perception and experience
(Callister et al., 1992). SCD patients experience more stressful events associated with
sickness and pain compared to healthy subjects (Porter et al., 2000; Gil et al., 2003; Gil
et al., 2004). Their perception of stress may be altered and thus is different from healthy
participants. In our study, the perception of stress was not measured. Therefore, the
vaso-reactivity phenotypes found in this study should not be generalized to vaso-reactivity
during sickle-related stressful events.
Lastly, the transfer relation between sympathetic activity and PPGa was not
considered in the assessment of BRSv. Low BRSv can be driven by increased
sympathetic outflow or a saturated end organ response (i.e. inability to constrict further
due to mechanical constraints). In SCD patients, high levels of endothelin-1 and NO
depletion can alter this transfer relation, making blood vessels more constricted to
sympathetic input (Ergul et al., 2004; Akinsheye and Klings, 2010).

91
Future directions
While our findings provide the comprehensive examination of the relations
between the sympathetic nervous system and vaso-reactivity phenotypes, it has not yet
been established how differences in vaso-reactivity phenotypes affect the variability of
crisis frequency. Therefore, the notion that SCD patients who had exaggerated or
sustained vasoconstriction would experience more VOC remains somewhat speculative.
Future work would benefit from a greater focus on examining the relationships
between vaso-reactivity phenotypes and the clinical outcomes such as pain crisis
frequency or HbS concentration. It is also worthwhile examining vaso-reactivity
phenotypes under different experimental conditions. Once we establish these
connections, it would be interesting to investigate how the sympathetic nervous system
has influenced these phenotypes.
To assess sympathetic function, we suggest incorporating other measures of the
sympathetic nervous system such as peroneal nerve activity (MSNA) or catecholamine
levels in addition to non-invasive measures used in this study. This will allow us to isolate
the effects of non-neural confounders, such as ET-1 and NO depletion, from effects
directly related to sympathetic activity in our analyses.

92
REFERENCES
Akinsheye, I., and Klings, E.S. (2010). Sickle cell anemia and vascular dysfunction: the
nitric oxide connection. Journal of cellular physiology 224(3), 620-625.
Alian, A.A., and Shelley, K.H. (2014). Photoplethysmography. Best Practice & Research
Clinical Anaesthesiology 28(4), 395-406.
Allen, J. (2007). Photoplethysmography and its application in clinical physiological
measurement. Physiological measurement 28(3), R1.
Ameloot, K., Van De Vijver, K., Broch, O., Van Regenmortel, N., Schoonheydt, K., Dits,
H., et al. (2013). Nexfin noninvasive continuous hemodynamic monitoring:
validation against continuous pulse contour and intermittent transpulmonary
thermodilution derived cardiac output in critically ill patients. The Scientific World
Journal 2013.
Awad, A.A., Ghobashy, M.A.M., Ouda, W., Stout, R.G., Silverman, D.G., and Shelley,
K.H. (2001). Different responses of ear and finger pulse oximeter wave form to
cold pressor test. Anesthesia & Analgesia 92(6), 1483-1486.
Ballas, S.K., Gupta, K., and Adams-Graves, P. (2012). Sickle cell pain: a critical
reappraisal. Blood, The Journal of the American Society of Hematology 120(18),
3647-3656.
Ballas, S.K., and Lusardi, M. (2005). Hospital readmission for adult acute sickle cell
painful episodes: frequency, etiology, and prognostic significance. American
journal of hematology 79(1), 17-25.
Berger, R.D., Akselrod, S., Gordon, D., and Cohen, R.J. (1986). An efficient algorithm
for spectral analysis of heart rate variability. IEEE Transactions on biomedical
engineering (9), 900-904.
Boehmer, R.D. (1987). Continuous, real-time, noninvasive monitor of blood pressure:
Peňaz methodology applied to the finger. Journal of clinical monitoring 3(4), 282-
287.
Borgers, A.J., van den Born, B.-J.H., Alkemade, A., Schattenkerk, D.W.E., van
Lieshout, J.J., Wesseling, K.H., et al. (2014). Determinants of vascular and
cardiac baroreflex sensitivity values in a random population sample. Medical &
biological engineering & computing 52(1), 65-73.
Callister, R., Suwarno, N.O., and Seals, D.R. (1992). Sympathetic activity is influenced
by task difficulty and stress perception during mental challenge in humans. The
Journal of physiology 454(1), 373-387.
Carter, J.R., and Goldstein, D.S. (2011). Sympathoneural and adrenomedullary
responses to mental stress. Comprehensive Physiology 5(1), 119-146.
Carter, J.R., and Ray, C.A. (2008). Sympathetic responses to vestibular activation in
humans. American Journal of Physiology-Regulatory, Integrative and
Comparative Physiology 294(3), R681-R688.

93
Chalacheva, P., Kato, R.M., Sangkatumvong, S., Detterich, J., Bush, A., Wood, J.C., et
al. (2015). Autonomic responses to cold face stimulation in sickle cell disease: a
time‐varying model analysis. Physiological reports 3(7), e12463.
Chalacheva, P., Kato, R.M., Shah, P., Veluswamy, S., Denton, C.C., Sunwoo, J., et al.
(2019). Sickle cell disease subjects have a distinct abnormal autonomic
phenotype characterized by peripheral vasoconstriction with blunted cardiac
response to head-up tilt. Frontiers in physiology 10, 381.
Chalacheva, P., Khaleel, M., Sunwoo, J., Shah, P., Detterich, J.A., Kato, R.M., et al.
(2017). Biophysical markers of the peripheral vasoconstriction response to pain
in sickle cell disease. PloS one 12(5), e0178353.
Coates, T.D., Chalacheva, P., Zeltzer, L., and Khoo, M.C. (2018). Autonomic nervous
system involvement in sickle cell disease. Clinical hemorheology and
microcirculation 68(2-3), 251-262.
Dampney, R.A. (2015). Central mechanisms regulating coordinated cardiovascular and
respiratory function during stress and arousal. American Journal of Physiology-
Regulatory, Integrative and Comparative Physiology 309(5), R429-R443.
Dampney, R.A. (2017). Resetting of the baroreflex control of sympathetic vasomotor
activity during natural behaviors: description and conceptual model of central
mechanisms. Frontiers in neuroscience 11, 461.
Darbari, D.S., Onyekwere, O., Nouraie, M., Minniti, C.P., Luchtman-Jones, L., Rana, S.,
et al. (2012). Markers of severe vaso-occlusive painful episode frequency in
children and adolescents with sickle cell anemia. The Journal of pediatrics
160(2), 286-290.
Darbari, D.S., Sheehan, V.A., and Ballas, S.K. (2020). The vaso‐occlusive pain crisis in
sickle cell disease: Definition, pathophysiology, and management. European
journal of haematology 105(3), 237-246.
Darbari, D.S., Wang, Z., Kwak, M., Hildesheim, M., Nichols, J., Allen, D., et al. (2013).
Severe painful vaso-occlusive crises and mortality in a contemporary adult sickle
cell anemia cohort study. PloS one 8(11).
Durocher, J.J., Klein, J.C., and Carter, J.R. (2011). Attenuation of sympathetic
baroreflex sensitivity during the onset of acute mental stress in humans.
American Journal of Physiology-Heart and Circulatory Physiology 300(5), H1788-
H1793.
Eaton, W.A. (2020). Hemoglobin S polymerization and sickle cell disease: A
retrospective on the occasion of the 70th anniversary of Pauling's Science paper.
American journal of hematology 95(2), 205-211.
Eaton, W.A., Hofrichter, J., and Ross, P.D. (1976). Delay time of gelation: a possible
determinant of clinical severity in sickle cell disease.

Eeftinck Schattenkerk, D.W., Van Lieshout, J.J., Van Den Meiracker, A.H., Wesseling,
K.R., Blanc, S., Wieling, W., et al. (2009). Nexfin noninvasive continuous blood

94
pressure validated against Riva-Rocci/Korotkoff. American journal of
hypertension 22(4), 378-383.
Elgendi, M. (2012). On the analysis of fingertip photoplethysmogram signals. Current
cardiology reviews 8(1), 14-25.
Ergul, S., Brunson, C.Y., Hutchinson, J., Tawfik, A., Kutlar, A., Webb, R.C., et al.
(2004). Vasoactive factors in sickle cell disease: in vitro evidence for endothelin‐
1‐mediated vasoconstriction. American journal of hematology 76(3), 245-251.
Frenette, P.S. (2002). Sickle cell vaso-occlusion: multistep and multicellular paradigm.
Current opinion in hematology 9(2), 101-106.
Frenette, P.S., and Atweh, G.F. (2007). Sickle cell disease: old discoveries, new
concepts, and future promise. The Journal of clinical investigation 117(4), 850-
858.
Gil, K.M., Carson, J.W., Porter, L.S., Ready, J., Valrie, C., Redding-Lallinger, R., et al.
(2003). Daily stress and mood and their association with pain, health-care use,
and school activity in adolescents with sickle cell disease. Journal of Pediatric
Psychology 28(5), 363-373.
Gil, K.M., Carson, J.W., Porter, L.S., Scipio, C., Bediako, S.M., and Orringer, E. (2004).
Daily mood and stress predict pain, health care use, and work activity in African
American adults with sickle-cell disease. Health Psychology 23(3), 267.
Gizdulich, P., Imholz, B., van den Meiracker, A., Parati, G., and Wesseling, K. (1996).
Finapres tracking of systolic pressure and baroreflex sensitivity improved by
waveform filtering. Journal of hypertension 14(2), 243-250.
Gizdulich, P., Prentza, A., and Wesseling, K.H. (1997). Models of brachial to finger
pulse wave distortion and pressure decrement. Cardiovascular research 33(3),
698-705.
Gladwin, M.T., and Vichinsky, E. (2008). Pulmonary complications of sickle cell disease.
New England journal of medicine 359(21), 2254-2265.
Hall, J.E., and Guyton, A.C. (2011). Guyton and Hall textbook of medical physiology.
Philadelphia: Saunders/Elsevier,.
Hassell, K.L. (2010). Population estimates of sickle cell disease in the US. American
journal of preventive medicine 38(4), S512-S521.
Hassell, K.L., Eckman, J.R., and Lane, P.A. (1994). Acute multiorgan failure syndrome:
a potentially catastrophic complication of severe sickle cell pain episodes. The
American journal of medicine 96(2), 155-162.
Hissen, S.L., Macefield, V.G., Brown, R., and Taylor, C.E. (2019). Sympathetic
baroreflex sensitivity is inversely related to vascular transduction in men but not
women. American Journal of Physiology-Heart and Circulatory Physiology
317(6), H1203-H1209.
Hofrichter, J., Ross, P.D., and Eaton, W.A. (1974). Kinetics and mechanism of
deoxyhemoglobin S gelation: a new approach to understanding sickle cell
disease. Proceedings of the National Academy of Sciences 71(12), 4864-4868.

95
Hoshikawa, Y., and Yamamoto, Y. (1997). Effects of Stroop color-word conflict test on
the autonomic nervous system responses. American journal of physiology-Heart
and circulatory physiology 272(3), H1113-H1121.
Katona, P.G., McLean, M., Dighton, D.H., and Guz, A. (1982). Sympathetic and
parasympathetic cardiac control in athletes and nonathletes at rest. Journal of
Applied Physiology 52(6), 1652-1657.
Khaleel, M., Puliyel, M., Shah, P., Sunwoo, J., Kato, R.M., Chalacheva, P., et al. (2017).
Individuals with sickle cell disease have a significantly greater vasoconstriction
response to thermal pain than controls and have significant vasoconstriction in
response to anticipation of pain. American journal of hematology 92(11), 1137-
1145.
Khoo, M.C., and Chalacheva, P. (2019). Respiratory modulation of peripheral
vasoconstriction: a modeling perspective. Journal of Applied Physiology 127(5),
1177-1186.
Klabunde, R.E. (2011). Cardiovascular physiology concepts. Lippincott Williams &
Wilkins.
Korhonen, I., and Yli‐Hankala, A. (2009). Photoplethysmography and nociception. Acta
Anaesthesiologica Scandinavica 53(8), 975-985.
Lanzkron, S., Carroll, C.P., and Haywood Jr, C. (2013). Mortality rates and age at death
from sickle cell disease: US, 1979–2005. Public health reports 128(2), 110-116.
Lombard, J.H., and Cowley Jr, A.W. (2012). "Neural control of blood vessels," in Primer
on the autonomic nervous system. Elsevier), 187-191.
Mahdi, N., Al-Ola, K., Khalek, N.A., and Almawi, W.Y. (2010). Depression, anxiety, and
stress comorbidities in sickle cell anemia patients with vaso-occlusive crisis.
Journal of pediatric hematology/oncology 32(5), 345-349.
Makani, J., Cox, S.E., Soka, D., Komba, A.N., Oruo, J., Mwamtemi, H., et al. (2011).
Mortality in sickle cell anemia in Africa: a prospective cohort study in Tanzania.
PloS one 6(2), e14699.
Manwani, D., and Frenette, P.S. (2013). Vaso-occlusion in sickle cell disease:
pathophysiology and novel targeted therapies. Blood, The Journal of the
American Society of Hematology 122(24), 3892-3898.
Marchi, A., Bari, V., De Maria, B., Esler, M., Lambert, E., Baumert, M., et al. (2016).
Simultaneous characterization of sympathetic and cardiac arms of the baroreflex
through sequence techniques during incremental head-up tilt. Frontiers in
physiology 7, 438.

Martina, J.R., Westerhof, B.E., van Goudoever, J., de Beaumont, E.M.H., Truijen, J.,
Kim, Y.-S., et al. (2012). Noninvasive continuous arterial blood pressure
monitoring with Nexfin®. The Journal of the American Society of
Anesthesiologists 116(5), 1092-1103.

96
Martins, W., Lopes, H.F., Consolim-Colombo, F.M., Gualandro, S.d.F.M., Arteaga-
Fernández, E., and Mady, C. (2012). Cardiovascular autonomic dysfunction in
sickle cell anemia. Autonomic Neuroscience 166(1-2), 54-59.
Mestre, J.R., Hernandez, A., Agramonte, O., and Hernandez, P. (1997). Cardiovascular
autonomic dysfunction in sickle cell anemia: a possible risk factor for sudden
death? Clinical Autonomic Research 7(3), 121-125.
Morrison, S.F. (2001). Differential control of sympathetic outflow. American Journal of
Physiology-Regulatory, Integrative and Comparative Physiology 281(3), R683-
R698.
Mozzarelli, A., Hofrichter, J., and Eaton, W.A. (1987). Delay time of hemoglobin S
polymerization prevents most cells from sickling in vivo. Science 237(4814), 500-
506.
Nebor, D., Bowers, A., Hardy-Dessources, M.-D., Knight-Madden, J., Romana, M.,
Reid, H., et al. (2011). Frequency of pain crises in sickle cell anemia and its
relationship with the sympatho-vagal balance, blood viscosity and inflammation.
Haematologica 96(11), 1589-1594.
Osterziel, K.J., Hänlein, D., Willenbrock, R., Eichhorn, C., Luft, F., and Dietz, R. (1995).
Baroreflex sensitivity and cardiovascular mortality in patients with mild to
moderate heart failure. Heart 73(6), 517-522.
Panepinto, J.A., Brousseau, D.C., Hillery, C.A., and Scott, J.P. (2005). Variation in
hospitalizations and hospital length of stay in children with vaso‐occlusive crises
in sickle cell disease. Pediatric blood & cancer 44(2), 182-186.
Pappano, A.J., and Wier, W.G. (2012). Cardiovascular physiology. Elsevier Health
Sciences.
Parati, G., Di Rienzo, M., and Mancia, G. (2000). How to measure baroreflex sensitivity:
from the cardiovascular laboratory to daily life. Journal of hypertension 18(1), 7-
19.
Pearson, S.R., Alkon, A., Treadwell, M., Wolff, B., Quirolo, K., and Boyce, W.T. (2005).
Autonomic reactivity and clinical severity in children with sickle cell disease.
Clinical Autonomic Research 15(6), 400-407.
Penaz, J. (Year). "Photoelectric measurement of blood pressure, volume and flow in the
finger. Digest 10th Int", in: Conf. Med. Biol. Eng).
Persson, P.B., DiRienzo, M., Castiglioni, P., Cerutti, C., Pagani, M., Honzikova, N., et al.
(2001). Time versus frequency domain techniques for assessing baroreflex
sensitivity. Journal of hypertension 19(10), 1699-1705.
Platt, O.S., Thorington, B.D., Brambilla, D.J., Milner, P.F., Rosse, W.F., Vichinsky, E., et
al. (1991). Pain in sickle cell disease: rates and risk factors. New England
Journal of Medicine 325(1), 11-16.
Porta, A., Bari, V., Maria, B.D., Cairo, B., Vaini, E., Malacarne, M., et al. (2018).
Peripheral resistance baroreflex during incremental bicycle ergometer exercise:
characterization and correlation with cardiac baroreflex. Frontiers in physiology 9,
688.

97
Porter, L.S., Gil, K.M., Carson, J.W., Anthony, K.K., and Ready, J. (2000). The role of
stress and mood in sickle cell disease pain: An analysis of daily diary data.
Journal of Health Psychology 5(1), 53-63.
Rauh, R., Posfay, A., and Mück‐Weymann, M. (2003). Quantification of inspiratory‐
induced vasoconstrictive episodes: a comparison of laser Doppler fluxmetry and
photoplethysmography. Clinical physiology and functional imaging 23(6), 344-
348.
Rees, D.C., Williams, T.N., and Gladwin, M.T. (2010). Sickle-cell disease. The Lancet
376(9757), 2018-2031.
Reisner, A., Shaltis, P.A., McCombie, D., Asada, H.H., Warner, D.S., and Warner, M.A.
(2008). Utility of the photoplethysmogram in circulatory monitoring. The Journal
of the American Society of Anesthesiologists 108(5), 950-958.
Robertson, D., Biaggioni, I., Low, P.A., and Polinsky, R.J. (2011). Primer on the
autonomic nervous system. Academic Press.
Robinson, A.T., Babcock, M.C., Watso, J.C., Brian, M.S., Migdal, K.U., Wenner, M.M.,
et al. (2019). Relation between resting sympathetic outflow and vasoconstrictor
responses to sympathetic nerve bursts: sex differences in healthy young adults.
American Journal of Physiology-Regulatory, Integrative and Comparative
Physiology 316(5), R463-R471.
Sangkatumvong, S., Coates, T., and Khoo, M. (2008). Abnormal autonomic cardiac
response to transient hypoxia in sickle cell anemia. Physiological measurement
29(5), 655.
Sangkatumvong, S., Khoo, M.C., Kato, R., Detterich, J.A., Bush, A., Keens, T.G., et al.
(2011). Peripheral vasoconstriction and abnormal parasympathetic response to
sighs and transient hypoxia in sickle cell disease. American journal of respiratory
and critical care medicine 184(4), 474-481.
Serjeant, G.R. (2013). The natural history of sickle cell disease. Cold Spring Harbor
perspectives in medicine 3(10), a011783.
Settels, J.J. (2014). "Noninvasive arterial pressure monitoring," in Monitoring
Technologies in Acute Care Environments. Springer), 87-107.
Shah, P., Khaleel, M., Thuptimdang, W., Sunwoo, J., Veluswamy, S., Chalacheva, P., et
al. (2020). Mental stress causes vasoconstriction in subjects with sickle cell
disease and in normal controls. haematologica 105(1), 83-90.
Stoica, P., and Moses, R.L. (2005). Spectral analysis of signals.
Taylor, L.E.V., Stotts, N.A., Humphreys, J., Treadwell, M.J., and Miaskowski, C. (2010).
A review of the literature on the multiple dimensions of chronic pain in adults with
sickle cell disease. Journal of pain and symptom management 40(3), 416-435.
Treadwell, M.J., Alkon, A., Styles, L., and Boyce, W.T. (2011). Autonomic nervous
system reactivity: children with and without sickle cell disease. Nursing research
60(3), 197.

98
Veluswamy, S., Shah, P., Denton, C.C., Chalacheva, P., Khoo, M.C., and Coates, T.D.
(2019). Vaso-Occlusion in Sickle Cell Disease: Is Autonomic Dysregulation of the
Microvasculature the Trigger? Journal of Clinical Medicine 8(10), 1690.
Veluswamy, S., Shah, P., Khaleel, M., Thuptimdang, W., Chalacheva, P., Sunwoo, J., et
al. (2020). Progressive Vasoconstriction with Sequential Thermal Stimulation
Indicates Vascular Dysautonomia in Sickle Cell Disease. Blood.
Vichinsky, E.P., Neumayr, L.D., Earles, A.N., Williams, R., Lennette, E.T., Dean, D., et
al. (2000). Causes and outcomes of the acute chest syndrome in sickle cell
disease. New England Journal of Medicine 342(25), 1855-1865.
Wesseling, K. (1990). "Finapres, continuous noninvasive finger arterial pressure based
on the method of Penaz," in Blood pressure measurements. Springer), 161-172.
Wesseling, K. (1995). Physiocal, calibrating finger vascular physiology for Finapres.
Homeostasis 36, 67-82.
Yang, Y.-M., Shah, A.K., Watson, M., and Mankad, V.N. (1995). Comparison of costs to
the health sector of comprehensive and episodic health care for sickle cell
disease patients. Public health reports 110(1), 80.
Yusuf, H.R., Atrash, H.K., Grosse, S.D., Parker, C.S., and Grant, A.M. (2010).
Emergency department visits made by patients with sickle cell disease: a
descriptive study, 1999–2007. American journal of preventive medicine 38(4),
S536-S541.

Asset Metadata

Creator Thuptimdang, Wanwara (author)

Core Title Role of the autonomic nervous system in vasoconstriction responses to mental stress in sickle cell disease: a bioengineering perspective

School Andrew and Erna Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Biomedical Engineering

Degree Conferral Date 2021-08

Publication Date 07/19/2021

Defense Date 05/11/2021

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag autonomic nervous system,baroreceptor reflex,baroreflex,baroreflex sensitivity,Mental stress,OAI-PMH Harvest,sickle cell disease

Format application/pdf (imt)

Language English

Advisor Khoo, Michael (committee chair)

Creator Email thuptimd@usc.edu,wanwara.th@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC15611964

Unique identifier UC15611964

Legacy Identifier etd-Thuptimdan-9800

Document Type Dissertation

Format application/pdf (imt)

Rights Thuptimdang, Wanwara

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email uscdl@usc.edu

Abstract (if available)

Abstract In sickle cell disease (SCD), flexible red blood cells, after releasing oxygen to the tissues, become rigid structures, producing regional obstructions of capillary flow that can potentially cascade into full-blown painful vaso-occlusive crises (VOC). Anecdotal evidence suggests that VOC is often precipitated by stress, cold exposure or pain. ? Recent studies have shown that individuals with sickle cell disease (SCD) exhibit greater vasoconstriction responses to physical autonomic stressors, such as heat pain and cold pain than normal individuals, but this is not the case for mental stress. ? Mental stress caused neural-mediated vasoconstriction in SCD and control subjects; however, there was substantial variability in the magnitude and the patterns of vasoconstriction responses across groups and individuals. In this thesis, we sought to determine whether the variability of magnitude and the patterns of vasoconstriction responses are related to inter-individual differences in autonomic function. ? In our experimental protocol, fifteen subjects with SCD and 15 healthy volunteers participated in 3 mental stress tasks: N-back, Stroop, and pain anticipation. R-R interval, arterial blood pressure and finger photoplethysmogram (PPG) were continuously monitored before and during these mental stress tasks. The magnitude of vasoconstriction was quantified using changes in PPG amplitude (PPGa) from the baseline period. The patterns of vasoconstriction were characterized from the patterns of PPGa signals during N-back and Stroop tasks. “Block Entropy” was applied to PPGa signals to quantify how close the pattern of vasoconstriction is to sustained vasoconstriction. ? To assess autonomic function, we estimated cardiac and vascular baroreflex sensitivities. Cardiac baroreflex sensitivity (BRSc) was estimated by applying both the “sequence” and “spectral” techniques to beat-to-beat measurements of systolic blood pressure and R-R intervals. The vascular baroreflex sensitivity (BRSv) was quantified using the same approaches, modified for application to beat-to-beat diastolic blood pressure and PPG amplitude (PPGa) measurements. For each subject, BRSc and BRSv were assessed during baseline and stress periods. ? In the first part of this work (Chapter 4), we determined whether the magnitude of vasoconstriction is related to inter-group difference in baseline BRSc and BRSv. We found that baseline BRSc was not different between SCD and non-SCD subjects, was not correlated with BRSv, and was not associated with the magnitude of vasoconstriction responses to mental stress tasks. BRSv in both groups was correlated with mean PPGa, and since both baseline PPGa and BRSv were lower in SCD, these findings suggested that the SCD subjects were in a basal state of higher sympathetically-mediated vascular tone. In both groups, baseline BRSv was positively correlated with the magnitude of vasoconstriction responses to N-back, Stroop and pain anticipation. After adjusting for differences in BRSv within and between groups, we found no difference in the vasoconstriction responses to all 3 mental tasks between SCD and non-SCD subjects. ? In the second part (Chapter 5), we examined whether sustained vasoconstriction was associated with autonomic function assessed during N-back and Stroop tasks. We found that sustained vasoconstriction was not particularly associated with SCD. In addition, the subjects who had sustained vasoconstriction also had low mean PPGa and BRSv during stress tasks. There was a strong correlation between BRSv and the degree of sustainability as indicated by block entropy. These findings suggested that sustained vasoconstriction during mental stress was primarily driven by increase in sympathetic tone. ? Overall, this work demonstrates that inter-individual difference in sympathetic function affects the variability of vasoconstriction responses to mental stress in sickle cell disease.