110
Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic Spinal Cord Injury by Rasha El-Kotob A thesis submitted in conformity with the requirements for the degree of Master of Science in Rehabilitation Science Graduate Department of Rehabilitation Science University of Toronto © Copyright by Rasha El-Kotob 2015

Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

  • Upload
    lemien

  • View
    218

  • Download
    3

Embed Size (px)

Citation preview

Page 1: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

Spinal Cord Injury

by

Rasha El-Kotob

A thesis submitted in conformity with the requirements for the degree of Master of Science in Rehabilitation Science

Graduate Department of Rehabilitation Science University of Toronto

© Copyright by Rasha El-Kotob 2015

Page 2: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

ii

Assessing Heart Rate Variability as a Surrogate Measure of

Cardiac Autonomic Function in Chronic Traumatic Spinal Cord

Injury

Rasha El-Kotob

Master of Science in Rehabilitation Science

Graduate Department of Rehabilitation Science

2015

Abstract

Individuals with a spinal cord injury (SCI) are at greater risk of cardiovascular disease (CVD)

than able-bodied individuals. A major CVD contributing factor is the presence of autonomic

disturbances, but the SCI-related changes in cardiac autonomic function are poorly understood.

Heart rate variability (HRV) has been reported to non-invasively assess the cardiac autonomic

nervous system (ANS). The following thesis involves investigating resting HRV in 56 subjects

with a traumatic chronic SCI with the aim to 1) describe the overall distribution of HRV in SCI;

2) determine whether there are HRV differences based on level and/or severity of injury; and, 3)

determine whether there is a relationship between parasympathetic and sympathetic frequency

measures. The results revealed that HRV is variable between-subjects, there were no significant

HRV differences based on level and/or severity of impairment, and the low frequency-to-high

frequency ratio (LF:HF), may not be an applicable measure in traumatic chronic SCI.

Page 3: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

iii

Acknowledgments

I would like to thank my supervisor, Professor Molly Verrier, for accepting me as a Master’s

student, introducing me to the world of research, providing me with true mentorship, and

stimulating my interests. I would also like to thank my co-supervisor, Dr. Sunita Mathur, for her

constant genuine encouragement and for offering me valuable on-going guidance even beyond

the scope of my research. I took pleasure in conducting my Master’s study and I owe it to both of

my supervisors.

I would like to acknowledge the members of my advisory committee, Dr. Catherine Craven, Dr.

Dave Ditor, and Dr. Paul Oh for their insightful input and even assistance regarding the planning

and execution of my research work.

Thank you to the research staff, graduate students and co-op students at Toronto Rehabilitation

Institute-UHN, Lyndhurst Centre. I was fortunate to work in such a fruitful research environment

with such knowledgeable colleagues. I would also like to especially thank Dr. Masae Miyatani

for allowing me to run a secondary data analysis on her collected data.

I am forever grateful for my parents who have always believed in me and never stopped cheering

me on. Thank you to my siblings, for being the reasons why I smile even during the stressful

times. Finally, I would like to express my gratitude to my university sweetheart and devoted

husband- not only did you support me relentlessly, but also you sincerely showed an immense

interest in my research which undeniably contributed to my eagerness.

I would also like to acknowledge the funder of this research: The Canadian Institute of Health

Research (Grant #: TCA-118348).

Page 4: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

iv

Table of Contents

Acknowledgments .......................................................................................................................... iii

Table of Contents ........................................................................................................................... iv

List of Tables ................................................................................................................................ vii

List of Figures .............................................................................................................................. viii

List of Appendices ......................................................................................................................... ix

Glossary .......................................................................................................................................... x

Chapter 1: ........................................................................................................................................ 1

1 Introduction ................................................................................................................................ 1

1.1 Spinal cord injury ................................................................................................................ 1

1.1.1 Cardiovascular disease in spinal cord injury .......................................................... 2

1.1.2 The function of the autonomic nervous system in spinal cord injury ..................... 3

1.2 Heart rate variability ........................................................................................................... 5

1.3 Literature review ................................................................................................................. 8

1.3.1 The findings on heart rate variability in spinal cord injury .................................... 8

1.3.2 Factors affecting heart rate variability .................................................................. 10

1.4 Study rationale .................................................................................................................. 12

Chapter 2: ...................................................................................................................................... 15

2 Objectives and Hypothesis ....................................................................................................... 15

2.1 Objectives ......................................................................................................................... 15

2.1.1 Primary Objective ................................................................................................. 15

2.1.2 Secondary Objective ............................................................................................. 15

2.1.3 Tertiary Objective ................................................................................................. 16

2.2 Hypothesis ......................................................................................................................... 16

2.2.1 Primary Hypothesis ............................................................................................... 16

2.2.2 Secondary Hypothesis ........................................................................................... 16

Page 5: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

v

2.2.3 Tertiary Hypothesis ............................................................................................... 17

Chapter 3: ...................................................................................................................................... 18

3 Methodology ............................................................................................................................ 18

3.1 Overview ........................................................................................................................... 18

3.2 Study variables .................................................................................................................. 19

3.2.1 Heart rate variability indices and related factors .................................................. 19

3.3 Subject selection. .............................................................................................................. 20

3.3.1 Electrocardiogram recordings ............................................................................... 20

3.3.2 Medications ........................................................................................................... 20

3.4 Heart rate variability analysis ........................................................................................... 21

3.5 Statistical Analysis ............................................................................................................ 24

3.5.1 Objective 1: Heart rate variability frequency distributions ................................... 24

3.5.2 Objective 2: Comparison of heart rate variability based on level and/or

severity of injury ................................................................................................... 24

3.5.3 Objective 3: Assessing the LF and HF indices ..................................................... 25

Chapter 4: ...................................................................................................................................... 26

4 Results ...................................................................................................................................... 26

4.1 Subject selection ............................................................................................................... 26

4.2 Frequency distributions of the heart rate variability indices ............................................. 28

4.3 Heart rate variability comparisons across level and/or severity of injury ........................ 31

4.3.1 Comparison of heart rate variability related factors across cohorts ...................... 35

4.4 Assessing the heart rate variability frequency domain indices: LF, HF and LF:HF ........ 38

4.4.1 Relationship between LF and HF ......................................................................... 38

4.4.2 Relationship between LF, HF, LF:HF and influencing factors ............................ 39

4.4.3 Predicting LF and HF from heart rate variability-related factors ......................... 41

Chapter 5: ...................................................................................................................................... 43

Page 6: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

vi

5 Discussion ................................................................................................................................ 43

5.1 Implications and future directions .................................................................................... 51

5.2 Study limitations ............................................................................................................... 52

Chapter 6: ...................................................................................................................................... 54

6 Conclusions .............................................................................................................................. 54

References ..................................................................................................................................... 55

Appendices .................................................................................................................................... 61

Page 7: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

vii

List of Tables

Table 1. HRV time domain and frequency domain measures* ...................................................... 8

Table 2. Factors reported to influence HRV ................................................................................. 11

Table 3. Potential HRV-related variables selected from the primary data ................................... 19

Table 4. Demographics and vital signs of the participants in total sample and per cohort........... 28

Table 5. Descriptive statistics for each HRV index in the entire sample (N=56) ......................... 29

Table 6. Comparison of HRV indices based on level of injury .................................................... 31

Table 7. Comparison of HRV indices based on severity of injury ............................................... 32

Table 8. Comparison of HRV indices based on level and severity of injury ................................ 33

Table 9. Relationship of the LF and HFindices based on level or severity of injury ................... 39

Table 10. Relationship of the LF and HF indices based on level and severity of injury .............. 39

Table 11. The relationship between LF, HF indices and the scalar HRV-related factors ............ 39

Table 12. Multiple linear regression analysis to predict LF for the entire sample (R2=0.039) .... 41

Table 13. Multiple linear regression analysis to predict HF for the entire sample (R2=0.009) .... 42

Table 14. Comparison of inter-individual variations in HRV between healthy subjects and

chronic traumatic SCI ................................................................................................................... 44

Page 8: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

viii

List of Figures

Figure 1. Possible contributors to greater CVD risk in individuals with SCI............................... 13

Figure 2. Parasympathetic and sympathetic innervations of the heart and peripheral muscles. ... 14

Figure 3. Representative example of HRV analysis using LabChart

(v.7.0) .............................. 22

Figure 4. Representative example of Poincaré Plot before and after the application of 45Hz low

pass filter ....................................................................................................................................... 23

Figure 5. CONSORT flowchart reflecting the inclusion and exclusion of the final data sample. 27

Figure 6. Frequency distribution of LF:HF. .................................................................................. 30

Figure 7. Boxplot representing LF:HF distribution. ..................................................................... 30

Figure 8. Boxplot representation, with and without the outliers, of LF:HF based on level and

severity of SCI .............................................................................................................................. 34

Figure 9. Boxplot representation of LTPAQ-SCI based on level and severity of injury. ............. 36

Figure 10. Boxplot representation of LEMS based on level and severity of injury. .................... 37

Figure 11. Boxplot representation of SCIM-III based on level and severity of injury. ................ 37

Figure 12. The relationship between LF and HF for total sample. ............................................... 38

Figure 13. Possible contributors to greater CVD risk in individuals with chronic traumatic SCI.48

Page 9: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

ix

List of Appendices

Appendix A ................................................................................................................................... 62

Appendix B ................................................................................................................................... 65

Appendix C ................................................................................................................................... 67

Appendix D ................................................................................................................................... 69

Appendix E ................................................................................................................................... 75

Appendix F .................................................................................................................................... 78

Appendix G ................................................................................................................................... 80

Appendix H ................................................................................................................................... 83

Appendix I .................................................................................................................................... 91

Appendix J .................................................................................................................................... 94

Appendix K ................................................................................................................................... 97

Page 10: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

x

Glossary

Absolute VO2 peak: Highest value of oxygen uptake attained during an incremental exercise

test. Expressed in litres of oxygen per minutes (L/min)

Activities of daily living (ADL): refers to daily self-care activities with an individual’s place of

residence, in outdoor environments, or both.

American Spinal Cord Injury Association (ASIA) impairment scale (AIS): a five point scale

(A-E), where A corresponds to a complete injury, B-D is an incomplete injury, E is

normal motor and sensory function

Autonomic nervous system (ANS): the system of nerves and ganglia that innervates the blood

vessels, heart, smooth muscles, viscera, and glands and controls their involuntary

functions, consisting of sympathetic and parasympathetic branches

Body mass index (BMI): An index for assessing overweight and underweight, obtained by

dividing body weight in kilograms (kg) by height in meters squared (m2). A measure of

25 kg/m2 or more is considered overweight

Bootstrapping: The sampling distribution of a statistic is estimated by taking repeated samples

from the data set in order to ensure that analytical models are reliable and will produce

accurate results

Bradycardia: a slow heart rate, usually less than 60 beats per minute (bpm)

Cardiac disease: congenital or acquired disease of only the heart

Cardiorespiratory fitness: the ability of the circulatory and respiratory systems to supply

oxygen to skeletal muscles during sustained physical activity

Cardiovascular disease (CVD): congenital or acquired disease of the heart and blood vessels

Chronic: having for a long duration

Complete Injury: No motor or sensory function in the lowest sacral segments (S4-S5)

Page 11: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

xi

Co-morbidities: Two or more diseases present simultaneously in a patient

Distribution: A graph plotting values of observations on the horizontal axis

Differences in HRV-related factors: Includes differences in age, sex, body mass index (BMI),

waist circumference (WC), time post injury, current smoking status, smoking history,

cardiorespiratory fitness level (absolute VO2 peak, relative VO2 peak, peak heart rate),

leisure time physical activity questionnaire-spinal cord injury (LTPAQ-SCI), lower

extremity motor score (LEMS), spinal cord independence measure (SCIM-III), number of

co-morbidities, family history of heart disease and sleep apnea.

Fast Fourier transform (FFT): Mathematical transformation of a function of time into a

function of frequency.

Heart rate variability: commonly used term to describe the oscillation of the heart rate and is

determined by measuring the R peak to R peak intervals, also referred to as NN intervals,

in an electrocardiograph (EGG)

High frequency (HF): A heart rate variability frequency domain measure representing vagal

modulation of the heart

Hypotension: Low resting blood pressure; in men systolic blood pressure less than 110mmHg

and in women systolic blood pressure less than 100mmHg

Incomplete injury: motor and/or sensory function preservation below neurological level of

injury and includes sacral segments (S4-S5)

Leisure time physical activity questionnaire (LTPAQ): Total number of minutes of physical

activity, not including activities of daily living, performed over the past week

Low frequency (LF): A heart rate variability frequency domain measure representing

parasympathetic and sympathetic, although more indicative of the latter, modulation of

the heart

Page 12: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

xii

Lower extremity motor score (LEMS): Measures voluntary motor strength in five myotomes

each scored out of five for a bilateral total of 50. A score of 30 or more suggests that the

individual is likely to walk

Mean Ranks: The data is ranked from lowest to highest to eliminate the effect of outliers

Neurological level of injury (NLI): This is the lowest segment where motor and sensory

function is normal in both the left and right side of the body

Paraplegia: complete or partial loss of sensation and movement in legs and in part or all of the

trunk due to an injury below the cervical vertebrae

Peak heart rate: Highest value of heart rate attained during an exercise test. Expressed in beats

per minute (bpm).

Physical activity: Any bodily movement produced by the skeletal muscles that increases heart

rate and breathing and requires energy expenditure

Physical capacity: a measure of ability to perform

Physical fitness: A state of physiological well being

Poincaré Plot: A diagram in which each R-R interval is plotted as a function of the previous R-

R interval. The values of each pair of successive R-R interval define a point in the plot.

Proportion of the number of interval differences of the consecutive NN intervals greater

than 50ms (pNN50): The proportion of the number of interval differences of the

consecutive R peak to R peak intervals greater than 50ms derived from an

electrocardiogram. A heart rate variability time domain measure that represents cardiac

parasympathetic modulation

Relative VO2 peak: Highest value of oxygen uptake attained during an incremental exercise

test. Expressed in milliliters of oxygen per kilogram of subject’s bodyweight per minute

(mL/kg/min)

Page 13: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

xiii

Sleep apnea: Sleeping disorder in which breathing repeatedly stops and starts. Obstructive sleep

apnea occurs when the throat muscles relax and Central sleep apnea occurs when the

brain doesn’t send proper signals to the muscles that control breathing.

Spinal cord independence measure (SCIM-III): Measures independence, out of a total of 100,

in performing activities of daily living.

Spinal cord injury: An injury that damages the spinal cord, due to trauma or disease, and results

in complete or partial paralysis.

Statistical Power: The power of a test is the probability that a given test will find an effect

assuming that one exists in the population.

Square root of the mean squared differences (RMSSD): square root of the mean squared

differences of the consecutive R peak to R peak intervals derived from an

electrocardiogram. A heart rate variability time domain measure that represents cardiac

parasympathetic modulation

Surrogate measure: a measurement taken with the intent to gain insight into a variable that is

either impractical to measure directly, or in principle impossible to measure.

Sympatho-vagal balance: The interaction between the sympathetic and parasympathetic

modulation of the heart.

Tachycardia: Rapid heart rate, usually greater than 100 beats per minute (bpm)

Tetraplegia: complete or partial paralysis of all four limbs due to an injury at the cervical

vertebrae

Traumatic: injury occurred due to physical damage to the spinal cord

Waist circumference (WC): a measure of the distance around the abdomen with the aim to

assess abdominal fat for chronic disease risk such as type 2 diabetes, high cholesterol,

high blood pressure, and heart disease

Page 14: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

1

Chapter 1:

1 Introduction

The likelihood of developing cardiac disease is 4.01 times greater in individuals with a spinal

cord injury (SCI) than individuals without a SCI.1 It has been speculated that a possible reason

could be the disruption of the cardiovascular autonomic nervous system (ANS) as evidenced by

the prevalence of autonomic dysreflexia and orthostatic hypotension within the SCI population.1-

5 The underlying physiological mechanisms responsible for the ANS disruptions in SCI have not

yet been fully determined.6 Heart rate variability (HRV) has been reported to non-invasively

measures the modulation of the cardiac ANS and has the potential to assess risk of cardiac

disease.7

Consequently in this study, HRV was assessed to examine the cardiac autonomic

changes in chronic traumatic SCI.

1.1 Spinal cord injury

In 2010, it was reported that around 86, 000 people were living with a SCI in Canada with the

prevalence to increase to 121,000 by 2030.8 Also it was reported that there are around 4,300 new

cases of SCI per year, 42% due to traumatic injuries and 58% due to non-traumatic causes.8

Traumatic SCIs arise due to a physical cause, for instance motor vehicle accidents, falls, or acts

of violence.8 Non-traumatic injuries, on the other hand, occur as a result of diseases, infections or

tumors that disrupt the normal functioning of the spinal cord.8

The cardiac risk factors for non-

traumatic SCI are more challenging to identify than for traumatic SCI since non-traumatic

injuries include tumor-related, congenital/developmental, infectious inflammatory and ischemic

causes.1 In addition, given that most of the SCI studies mainly focus on traumatic injuries,

9 there

is more information available regarding cardiac health within the traumatic group. There are also

demographic differences between the two groups: mean age is higher in the non-traumatic than

in the traumatic group, and although the proportion of females and males is the same in the non-

traumatic group there are three times more males than females in the traumatic group.9

Additionally, there are more incomplete injuries than complete injuries reported within the non-

traumatic group than in the traumatic group.9

There are less secondary complications, such as

spasticity and pressure ulcers, in the non-traumatic group than in the traumatic group.10

Finally,

Page 15: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

2

the traumatic group improves with rehabilitation to a greater extent than the non-traumatic

group.9

The spinal cord, which is located within the spinal canal, provides motor and sensory

information between the brain and the body.11-12

In humans, the spinal cord is comprised of 31

segments: 8 cervicali, 12 thoracic, 5 lumbar, 5 sacral and 1 coccygeal.

11 Each segment receives

sensory information from skin areas referred to as dermatomes, and each segment innervates a

group of muscles referred to as a myotome.12

After a SCI, the International Standards for

Neurological and Functional Classification of Spinal Cord Injury (ISNCSCI) are used to

evaluate the neurological level of impairment (NLI) and severity of the injury [American Spinal

Cord Injury Association (ASIA) impairment scale (AIS)] in terms of motor and sensory function

(Appendix A).13-15

AIS is measured by a five point scale (A-E), where A corresponds to a

complete injury (no motor or sensory function in the lowest sacral segments), B-D is an

incomplete injury (motor and/or sensory function preserved below neurological level of injury

and includes sacral segments) and E is normal motor and sensory function.13-14,12

1.1.1 Cardiovascular disease in spinal cord injury

Cardiovascular disease (CVD) has been identified as the leading cause of morbidity and

mortality accounting for approximately 30-50% of deaths within the SCI population; in contrast

to 5-10% in an age and sex matched able-bodied population.2-5,15-16

There is supporting evidence

in the literature indicating that individuals with a SCI are at an increased risk of cardio-metabolic

syndrome (CMS).17-19

CMS is characterized by having three or more of the following five

clinical features: central obesity (waist circumference men>120cm, women>88cm),

hypertriglyceridemia (≥1.7mmol/l), low plasma high density lipoprotein cholesterol

(men<1.03mmol/l, women<1.29mmol/l), hypertension (≥130/85mmHg or on relevant

medications), and fasting hyperglycemia (≥100mg/dl or on relevant medications).18-20

All of the

aforementioned risk factors lead to atherosclerotic plaque formation and earlier onset of CVD.18-

19 In addition, after a SCI, the often sedentary lifestyle, physical deconditioning and inflated

post-injury inflammatory cytokines contribute to the pro-atherogenic outcome and CVD

i The 8

th cervical nerve emerges between the 7

th cervical (C7) and the first thoracic (T1) vertebrae

9

Page 16: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

3

development.17-19

As the number of CMS risk factors increases, CVD vulnerability greatly

increases.19

Some of the CMS risk factors overlap with the traditional CVD risk factors such as

sex, age, diabetes, blood lipid profile, elevated systolic blood pressure, smoking status,

sedentary lifestyle, unhealthy diet for example diet high in saturated fats, and obesity.5,21

Also,

investigators have reported that the risk of developing CVD increases with level and severity of

injury i.e. higher and complete injury.16,22

Nevertheless, a recent study by Miyatani et al.23

found

that there was greater arterial stiffness, an emerging indicator of coronary artery disease, in

subjects with paraplegia than tetraplegia. Surprisingly, only 48% of the subjects with arterial

stiffness met the diagnostic criteria for CMS.23

Therefore, the CMS and traditional CVD risk

factors do not completely explain why individuals with a SCI are at such great risk.5

Consequently, there must be additional unexplained factors that contribute to the high prevalence

of CVD within the SCI population. Cardiovascular autonomic disruption is common after a SCI

and the impairment has been reported to increase the risk of developing CVD.2-5

In able-bodied

subjects, a poorly balanced cardiovascular ANS measured via the assessment of vital signs, has

been associated with myocardial infarction, congestive heart failure, life threatening arrhythmias,

and atherosclerotic plaque progression.24

1.1.2 The function of the autonomic nervous system in spinal cord injury

In comparison to studies examining motor and sensory dysfunction post SCI, there are fewer

studies examining disturbances in the ANS.16,25

After a SCI, there is a disruption in the ANS

resulting in abnormal regulation of heart rate, blood pressure, bladder, bowel and temperature

regulation, as well as respiratory and/or sexual dysfunction.6,26

The parasympathetic

preganglionic neurons are situated in the brain stem, specifically in the nuclei of four cranial

nerves: oculomotorius (III), facialis (VII), glossopharyngeus (IX), and vagus (X).5-6,17,27

The

vagus nerve supplies most of the internal organs with the exceptions of the genital organs,

bladder, distal intestine and anus, which are innervated by the parasympathetic sacral (S2-S4)

nerves.5-6,17,28

There is no parasympathetic innervation of the peripheral blood vessels except for

the vessels that supply the pelvic organs.5-6,28

As for the sympathetic preganglionic neurons they

are situated in the grey matter of the spinal column at T1-L2.4-6,26,28

Page 17: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

4

With respect to cardiovascular autonomic function, the sympathetic preganglionic neurons at T1-

T5 innervate the heart and the blood vessels of the upper body, while T6-L2 innervate the blood

vessels of the lower body.4-5,11,25-26,29

The parasympathetic innervation of the heart, arises from

the vagal nuclei of the brainstem.4-5,11,25-26,29

Therefore, depending on the level of injury,

sympathetic function may be disrupted resulting in impaired control of heart rate and/or blood

pressure and hypotension and bradycardia are both prevalent after a cervical injury.16-17,25,27

Parasympathetic activity, via beat-to-beat control, decreases heart rate and conversely

sympathetic activity gradually increases heart rate.17,29-30

The level and degree of SCI has been

reported to be directly linked to the extent of cardiovascular autonomic dysfunction.4,16-17,26

For

instance, individuals with a complete cervical injury suffer from an absolute disconnection

between the upper autonomic centres in the brain and the intermediolateral cell column at T1-

L2.17

Early after a SCI, sympathetic activity is quickly disrupted resulting in bradycardia and the vagus

nerve is hypersensitive for at least 2-3 weeks.17

Some treatment approaches that may be required

to maintain an adequate heart rate involve either implanting a temporary pacemaker or

administering atropine which is a competitive muscarinic acetylcholine receptor antagonist.17

Unfortunately, the acute period of cardiac autonomic disruption does not necessarily normalize

and may become a chronic issue, especially among individuals with complete cervical or high

thoracic injuries.5,6,17,25

The disrupted cardiovascular ANS is characterized by a low resting

sympathetic tone and an unaffected resting parasympathetic tone leading to a reduced resting

blood pressure and heart rate and an abnormal cardiovascular response to exercise.26

Furthermore, 91% of individuals with high and complete injuries are more prone to autonomic

dysreflexia (AD) than those with low (below T6) and incomplete injuries (27%).16,26-27

AD arises

from a sensory stimulus below the level of injury and results in episodes of hypertension (20-40

mmHg above baseline) accompanied by a baroreflex mediated bradycardia.5,17,25-26

Similarly,

orthostatic hypotension (OH) is also related to the level and severity of the SCI. The incidence of

OH is as high as 74% in individuals with high (T5 and above) and complete SCI.4,16,25,27

OH is

characterized by a decrease in systolic blood pressure ≥20 mmHg and/or a decrease in diastolic

blood pressure ≥ 10 mmHg from baseline, immediately after transferring from a supine to a

seated position.4-5

The exact mechanisms resulting in both AD and OH are not clearly

Page 18: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

5

understood and are probably multifactorial, however the loss of sympathetic control has been

reported to be a predominant factor.4-6

To evaluate autonomic function, a guideline was recently published (2012) outlining the

International Standards to Document Remaining Autonomic Function after Spinal Cord Injury

(ISAFSCI).6 It is recommended to use the guideline in addition to the ISNSCI, and it can be

administered at any time following the injury.6 In the autonomic standards assessment form

(Appendix B), a general description of the remaining autonomic function is recorded for each

system/organ.6 For the urinary tract, bowel and sexual function there is a grading system similar

to the ISNSCI scoring system.6

Also, the assessment form incorporates self-reported history, if

any, regarding the function of the ANS.6 In terms of assessing the cardiovascular ANS, general

autonomic control of the heart is reported as normal, abnormal (bradycardia, tachycardia and/or

any other dysrhythmias), unknown, or unable to assess. Similarly, autonomic control of the

blood pressure, is described as normal, abnormal (resting systolic blood pressure is below 90

mmHg, OH and/or AD), unknown or unable to assess.6 Nonetheless, it is important to consider

that the use of ISAFSCI has not yet been validated.6 In addition, the autonomic assessment,

particularly for the heart, lacks sensitivity and specificity and does not definitively determine the

degree of cardiac autonomic function/dysfunction.

1.2 Heart rate variability

HRV is the most commonly used term to describe the oscillation of the heart rate and is

determined by measuring the peak R to R intervals, also referred to as NN intervals, on an

electrocardiogram (ECG).31

The sinoatrial (SA) node, located in the right atrium of the heart, is

responsible for generating each heartbeat and its firing rate is modulated by the ANS.29-30,32-34

It

has been reported that HRV analysis can non-invasively reflect cardiac regulation via the ANS

which controls heart rate through parasympathetic and sympathetic innervation of the

heart.24,29,33,35-36

Studies have shown that low HRV is an independent predictor of cardiovascular

dysfunction and cardiovascular risk.31,33-34,36-40

Diminished levels of HRV in able-bodied

subjects, have been associated with heart failure,38

diabetes,38,40

hypertension,38,40

abnormal

cholesterol,40

asymptomatic left ventricular dysfunction,38

fatal arrhythmias,41

and death due to

cardiac causes.31

Page 19: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

6

HRV standardize guidelines were developed in 1996 by a Task Force composed of members

from the European Society of Cardiology and the North American Society of Pacing and

Electrophysiology.7 Based on the recommended guidelines, there are a number of methods that

can be utilized to measure HRV: 1. Time domain measures- subdivided into statistical measures

and geometric measures; 2. Frequency domain measures; and, 3. Non-linear measures.7 In the

published literature, the two most commonly applied measurement methods for HRV are the

time domain statistical measures and the frequency domain measures (Table 1). Investigators

may have a preference towards these methods since some of the parameters are thought to be

physiological markers and therefore can be used to directly assess sympathetic and

parasympathetic modulation of the heart.

The statistical measures, the square root of the mean squared differences of the consecutive NN

intervals (RMSSD) and the proportion of the number of interval differences of the consecutive

NN intervals greater than 50ms (pNN50), are a reflection of cardiac parasympathetic

modulation.31,35,37

As for the frequency domain measures, the high frequency (HF) component

has been reported to reflect cardiac parasympathetic modulation24,31-32,40-41

while the low

frequency (LF) component is controversial. Some claim that LF is both a marker of

parasympathetic and sympathetic modulation24,31-32,40-41

while others claim that it is more

indicative of sympathetic modulation.2,26,30,33,42-45

Further understanding of the simultaneous

actions of the neurotransmitters on the heart rate may assist in elucidating the physiological

interpretations. The LF:HF has been described as a measure of the sympatho-vagal balance of

the cardiac autonomic nervous system; higher ratio indicating greater sympathetic activity and a

lower ratio indicating lower sympatheticactivity.2,29-30,33,42-44

It is important to note that HRV

measures the modulation of the cardiac ANS rather than the mean level of autonomic activity.2,30

Therefore, comparing HRV, for instance between able-bodied individuals and individuals with a

SCI or pre and post exercise intervention, may be more informative than solely reporting the

value on its own.

The HRV time domain measures are calculated directly from the NN intervals on an ECG,7

while the frequency domain measures are derived using either parametric (e.g. autoregressive

model) or non-parametric [e.g. fast Fourier transform (FFT)] mathematical algorithms.7 The FFT

is the most commonly used and recommended measure as it is simple and quick to apply.7,31,33

Page 20: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

7

Using FFT, the NN intervals in the ECG are transformed to provide the amount of variation as a

function of frequency.7,31,33

In the calculated power spectrum, the total power represents the total

variance and each frequency component corresponds to a specific bandwidth7,41

(Table 1). The

frequency components are measured in absolute values of power (ms2).

7

There are currently no widely accepted HRV normative values,37

and this is probably due to the

dynamic nature of the sympatho-vagal system and the intrinsic and/or extrinsic factors that may

influence it.44

The Task Force provided normal HRV values of the frequency measures in

healthy adults extracted from a short term recording. However, a systematic review paper by

Nunan and colleagues37

questions the Task Force’s normal values since they were approximated

from small sample size studies. In comparison with the literature, the Task Force LF and HF

power values were higher; Task Force figures being 1,170ms2 for LF power and 975 ms

2 for HF

power while the literature reported 519 ms2 for LF power and 657 ms

2 for HF power.

37 Also, the

LF:HF ratio reported by the Task Force (1.5-2.0) was lower than the ratio that was extracted

from the literature (2.8).37

Page 21: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

8

Table 1. HRV time domain and frequency domain measures*

Time Domain- Statistical Measures Frequency Domain Measures

Statistical variables include:

Standard deviation of the NN interval

(SDNN)

Square root of the mean squared

differences of the consecutive NN

intervals (RMSSD)

Number of the interval differences of

the consecutive NN intervals greater

than 50ms (NN50)

Proportion of the NN50 (pNN50) which

is calculated by dividing the NN50 by

the total number of NN intervals

Frequency variables include:

Ultra low frequency (ULF)

Bandwidth: below 0.0033 Hz

Very low frequency (VLF)

Bandwidth: 0.0033-0.04 Hz

Low frequency (LF)

Bandwidth: 0.04-0.15 Hz

High frequency (HF)

Bandwidth: 0.15-0.40 Hz

* Table was developed using the Task Force guidelines.7

1.3 Literature review

1.3.1 The findings on heart rate variability in spinal cord injury

Assessing HRV in SCI is valuable as it can quantify the extent of cardiac autonomic dysfunction

that is distinctively experienced by each individual, and can be regularly used to evaluate and

monitor changes in a clinical setting over time.2,29,42

Unlike HRV analysis, most ANS

measurement tools are invasive and/or require specialized expertise and equipment making it

difficult to assess routinely in a clinical setting; for instance administering a sympathetic skin

response testii or measuring resting plasma catecholamine concentrations.

2,4,16 HRV analysis is

currently the only assessment tool that solely examines cardiac autonomic modulation in SCI.

ii Sympathetic skin response involves the momentary change of the electrical potential of the skin with the aim to

assess sympathetic function. The response may be either spontaneously or reflexively induced by applying an

internal or external arousal stimulus46

Page 22: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

9

The psychometric properties of measuring HRV have been minimally studied, but HRV analysis

in SCI has been shown previously to be a reproducible measure (LF:HF and LF, intraclass

correlation coefficient (R)=0.82-0.88; HF, R=0.53).47

Additionally, Claydon and colleagues2

revealed that the HRV frequency indices (HF, LF, LF:HF), measured in the supine position,

correlated with clinical measures of the cardiovascular autonomic function including sympathetic

skin response, orthostatic cardiovascular response, and plasma catecholamine levels.

A number of studies have used time domain measures to assess HRV in chronic SCI. Bunten and

colleagues42

did not observe any differences in the time domain measures between the complete

and incomplete cervical SCI, thoracic (T1-L5) SCI and able-bodied subjects. As well, Wang and

colleagues29

reported no differences in RMSSD and pNN50 when comparing complete cervical

injuries against complete low thoracic injuries (T10-L2). Provided that the cardiac

parasympathetic innervations remain intact after a SCI, the investigators expected that the

parasympathetic time domain markers would not be disrupted.42

Bunten and colleagues42

explain

that there is parasympathetic predominance, but without an increase in parasympathetic activity.

However, Rosado-Rivera and colleagues15

reported that the low paraplegia group (T7-T12)

displayed lower RMSSD values compared to able-bodied, high paraplegia (T2-T5) and

tetraplegia (C4-C8) groups. Therefore, as expected, the low paraplegia group displayed the

highest mean heart rate (83±12bpm) and mean heart rate was significantly higher than the able-

bodied (70±9bpm) group and the tetraplegiac group (69±10bpm).15

Frequency domain measures have also been used to evaluate HRV in chronic SCI. A study by

Claydon and colleagues2 showed that individuals with complete and incomplete cervical SCIs

displayed lower LF values, in comparison to thoracic (T2-T11) SCI and able-bodied subjects.

The reduced LF in individuals with SCI is most likely due to the loss of sympathetic control.2,42

Similarly, Wang and colleagues29

reported low LF values in individuals with complete cervical

injuries versus those with complete thoracic (T10-L1) injuries. Individuals with incomplete

cervical injuries had greater LF values than those with complete cervical injuries given that there

was less damage to the descending sympathetic pathways.30,42

They also showed that the LF

power in the thoracic group was similar to the controls indicating undisrupted sympathetic

cardiac autonomic control. However, when examined per case, the two subjects with high

thoracic SCI (above T5) displayed lower LF values than the controls.2 On the contrary; the study

Page 23: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

10

by Bunten and colleagues42

found that both the thoracic group (T1-L5) and cervical group had

lower LF values than the able-bodied subjects. Similarly, Castiglioni et al.48

reported reduced LF

values in the thoracic group (T5-L4) with respect to the able-bodied group.

As for the HF component, the results by Claydon et al.2 revealed that it was higher in the

complete and incomplete cervical group than in the thoracic SCI and control group. The increase

in vagal tone explained the presence of bradycardia within the cervical group.2,49-50

In addition,

the lower LF:HF outcome in the cervical group compared to the thoracic group and the control

group further suggests that there is parasympathetic predominance after a cervical SCI.2,42

However, Grimm and colleagues30

found that individuals with a complete cervical injury had

lower HF values than those with an incomplete cervical injury, thoracic (below T7) injury and

able-bodied individuals. Wang et al.29

also reported lower HF in a cervical group when

compared to a thoracic group (T10-L1). Both Wang et al.29

and Grimm et al.30

found no

differences in the LF:HF between the cervical and thoracic injuries. They suggested that the lack

of difference in LF:HF indicated that the cardiac ANS was trying to maintain sympatho-vagal

homeostasis.29-30

The study by Bunten and colleagues,42

on the other hand, found no differences

in the HF component between the three groups indicating normal resting vagal tone. In the

thoracic group (T2-T11), Claydon et al.2 found that the HF was lower, and LF:HF was higher

than that observed in the able-bodied group. In addition, Rosado-Rivera et al.15

suggested that a

reduced HF and higher LF:HF is a possible explanation for the prevalence of elevated heart rates

among individuals with high and low paraplegia. Unfortunately, it is unclear why vagal tone is

reduced but some have hypothesized that it could be a compensatory reduction with the aim to

maintain sympatho-vagal balance2,15,29-30

or be due to cardiovascular deconditioning.15

Castiglioni and colleagues,48

on the contrary, reported no differences in the HF values and

LF:HF between the able-bodied and thoracic group (T5-L4). The reasons for the HRV

discrepancies reported in the literature are still uncertain, however, there are a number of factors,

including the experimental paradigm used for ECG collection that could have a major influence.

1.3.2 Factors affecting heart rate variability

The relationship between HRV and potential influencing factors has been examined in previous

literature. Factors including age, sex, obesity, fitness level, sleep apnea, emotional state, and

Page 24: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

11

smoking status have all been confirmed (Table 2). There is a well-established relationship

between age and HRV, with younger individuals showing higher HRV.32,37,40

The decrease of

HRV with increasing age could be due to the reduction of both parasympathetic and sympathetic

activity.33

The relationship between sex and HRV remains unclear. A review by Nunan and

colleagues37

found that the chosen unit of measurement influenced the relationship between sex

and HRV. There is a confirmed link between ANS dysfunction and obesity40,51

as Alrefaie et

al.52

and Mehta et al.53

found a relationship between body mass index (BMI) and HRV while

Farah et al.54

reported a negative correlation between waist circumference and HRV, but no

relationship between BMI and HRV. Farah et al.54

argue that central obesity measured via waist

circumference is a better indicator of cardiac autonomic dysfunction than general obesity

measured by BMI.54

Melanson and colleagues55

examined the effect of endurance training on

HRV in previously sedentary subjects and found that engaging in regular physical activity

increased HRV. As for sleep apnea, Flevari et al.56

hypothesized that sleep disordered breathing

increased autonomic tone. Also, Chalmers and colleagues57

conducted a meta-analysis and found

reduced HRV in individuals with anxiety disorders. Finally, Lee et al.58

reported that smoking

decreases cardiac parasympathetic activity and increases sympathetic function.

Table 2. Factors reported to influence HRV

Factors Influence on HRV

Age Negative relationship between age and

HRV

Sex Females displayed lower time domain

measures than males (8-11% lower)

LF and HF lower in males when expressed

in absolute units (ms2) (14% and 8% lower

than females, respectively)

When the units were normalized, LF was

higher in males (17% higher than females)

and HF was comparable

When expressed in log units, females had

20% lower LF, and 18% lower HF than

males

LF:HF lower in females than males

regardless of the measurement unit

Page 25: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

12

Obesity Individuals with body mass index greater

than or equal to 30 kg/m2 had lower

RMSSD, HF and LF but similar LF:HF in

comparison to non-obese control group

Negative correlation between waist

circumference and RMSSD (r2=0.15) and

PNN50 (r2=0.16)

Fitness level RMSSD and HF increased above baseline

after 12 weeks of moderate-vigorous

intensity exercise

Sleep apnea Constant HF but higher RMSSD, pNN50,

LF, and LF:HF in patients with positional

obstructive sleep apnea

Emotional state People with panic disorder, post-traumatic

stress disorder, generalized anxiety

disorder, social anxiety disorder all showed

lower HF values relative to the control

group

Smoking status Smokers had lower HF and higher LF and

LF:HF than non-smokers

RMSSD the same in smokers and non-

smokers

1.4 Study rationale

The combination of an increased risk of cardio-metabolic syndrome, physical deconditioning,

increase in inflammatory cytokines, cardiac autonomic dysfunction, and barriers to a physically

active lifestyle all lead to an increased risk of developing CVD after a SCI (Figure 1). The

currently available autonomic evaluation guidelines, unlike HRV, provide general information

regarding cardiac ANS activity but do not measure the extent of cardiac autonomic dysfunction.

Since the degree of cardiac autonomic dysfunction in SCI depends on the neurological level of

impairment and severity of injury, resting HRV measures may vary accordingly.

Parasympathetic innervation of the heart is still intact after a SCI, as it arises from the brainstem,

and therefore cardiac autonomic function is thought to be disrupted due to sympathetic damage

(Figure 2). Given the location of sympathetic innervation [T1-T4(T5)], HRV is expected to be

disrupted in individuals with a SCI above the level of T5 and the degree of disruption is expected

to be greater in complete injuries. Unfortunately, due to the few studies in SCI and the

Page 26: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

13

inconsistent HRV findings, the relationship between HRV and SCI remains unclear. The

limitations of current literature are: small sample size, combining different etiology of SCI,

and/or cohort selection (discrepancies in neurological level of impairment). Consequently, the

cardiac autonomic changes contributing to ANS dysfunction in SCI, as measured via HRV, are

yet to be fully determined. In this thesis, resting supine HRV was examined in a large and

representative sample of chronic traumatic SCI while still considering autonomic innervations

based on the anatomy of the cardiac ANS. Chronic SCI, as opposed to acute or sub-acute, is

considered to be a stable state and as a result is the ideal phase to study the adaptive state of ANS

in individuals with a SCI.3 Also, given that there are important etiological, comorbidities and

demographic differences between traumatic and non-traumatic SCI it was decided to examine

HRV and influencing factors in traumatic SCI only.

Figure 1. Possible contributors to greater CVD risk in individuals with SCI. The theoretical

framework summarizes the relationship between SCI and CVD. After a SCI, increased risk of

cardio-metabolic syndrome, elevated levels of inflammatory cytokines, lifestyle changes,

disrupted ANS, and some non-modifiable factors all contribute to overall CVD development

(modified from Figure 1.0 on page 128 in the Rehabilitation Environmental Scan Atlas:

Capturing Capacity in Canadian SCI Rehabilitation.59

)

Page 27: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

14

Figure 2. Parasympathetic and sympathetic innervations of the heart and peripheral

muscles. Sympathetic innervations arise from T1-T4/T5 cord segments. Consequently, level of

injury may affect cardiac autonomic function as measured by HRV.

Page 28: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

15

Chapter 2:

2 Objectives and Hypothesis

2.1 Objectives

2.1.1 Primary Objective

To describe the distribution of HRV indices in a population of individuals with a chronic

traumatic SCI.

Primary HRV index: Low frequency to high frequency ratio (LF:HF).

Secondary HRV indices: Low frequency (LF), high frequency (HF), square root of the

mean squared differences of the consecutive NN intervals (RMSSD), and proportion of

the NN50 (pNN50) which is the percentage of pairs of adjacent NN intervals differing by

more than 50ms.

2.1.2 Secondary Objective

a) To determine whether there is a difference in HRV indices (primary and secondary)

based on level of injury (above T5 and below T5).

b) To determine whether there is a difference in HRV indices (primary and secondary)

based on severity of injury (complete injury and incomplete injury).

c) To determine whether there is a difference in HRV indices (primary and secondary)

based on level and severity of injury (complete and equal to/above T5, complete and

below T5, incomplete and equal to/above T5, and incomplete and below T5).

d) To determine whether there are any differences in the selected HRV-related factors based

on level and/or severity of injury. The selected HRV-related factors are: age, sex, body

mass index (BMI), waist circumference (WC), time post injury, current smoking status,

smoking history, cardiorespiratory fitness level (absolute VO2 peak, relative VO2 peak,

peak heart rate), leisure time physical activity questionnaire-spinal cord injury (LTPAQ-

Page 29: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

16

SCI), lower extremity motor score (LEMS), spinal cord independence measure (SCIM-

III), number of co-morbidities, family history of heart disease and sleep apnea.

2.1.3 Tertiary Objective

a) To determine whether there is a relationship between low frequency (LF) and high

frequency (HF) indices in chronic traumatic SCI.

b) To determine whether there is a relationship between LF, HF, LF:HF and HRV related

factors in chronic traumatic SCI: age, BMI, WC, time post injury, cardiorespiratory

fitness level (absolute VO2 peak, relative VO2 peak, peak heart rate), LTPAQ-SCI,

LEMS, SCIM-III, and number of co-morbidities.

c) To determine whether there is a relationship between age, waist circumference and peak

heart rate and the LF or HF indices in the entire study sample and in individuals with a

complete injury that is equal to/above T5.

d) To determine whether there is a relationship between LF, HF, age at injury and resting

systolic blood pressure in the entire study sample and in individuals with a complete

injury that is equal to/above T5.

2.2 Hypothesis

2.2.1 Primary Hypothesis

There will be a multimodal distribution of the HRV indices based on level and severity of injury.

2.2.2 Secondary Hypothesis

a) Individuals with an injury equal to or above T5 will display lower HRV values than those

with an injury below T5.

b) Individuals with a complete injury will display lower HRV values than those with an

incomplete injury.

Page 30: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

17

c) Individuals with a complete injury equal to or above T5 will display the lowest HRV

values. Alternately, individuals with an incomplete injury below T5 will display the

highest HRV values indicating an undisrupted cardiac ANS.

d) Age, sex, time post injury, current smoking status, smoking history, and family history of

heart disease do not depend on the level or severity of injury and thus will not show any

differences across the cohorts. However, BMI, WC, number of co-morbidities and

presence of sleep apnea will be greater in individuals with a higher level of injury and/or

a complete injury. On the contrary, cardiorespiratory fitness level, LTPAQ-SCI, LEMS,

and SCIM-III will be lower in individuals with a higher level of injury and/or a complete

injury.

2.2.3 Tertiary Hypothesis

a) The LF and HF indices will display a high positive linear relationship since the role of the

ANS is to maintain homeostasis

b) Both LF and the LF:HF will display a positive linear relationship with age, BMI, WC,

time post injury and number of co-morbidities and a negative linear relationship with

cardiorespiratory fitness level (absolute VO2 peak, relative VO2 peak and peak heart

rate), LTPAQ-SCI, LEMS, and SCIM-III. HF will display a negative linear relationship

with age, BMI, WC, time post injury and number of co-morbidities and a positive linear

relationship with cardiorespiratory fitness level, LTPAQ-SCI, LEMS, and SCIM-III.

c) Age, WC and peak heart will predict LF and HF indices in the entire sample and in

individuals with an injury equal to/above T5.

d) In the entire sample and in individuals with an injury equal to/above T5: There will be a

positive linear relationship between LF and age at injury and resting systolic blood

pressure. Whereas there will be negative linear relationship between HF and age at injury

and resting systolic blood pressure.

Page 31: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

18

Chapter 3:

3 Methodology

3.1 Overview

This study was a secondary data analysis of a primary data set from a recently published study

that explored the associations between arterial stiffness and spinal cord impairment.23

The

inclusion criteria of the primary study were English speaking subjects between 18-80 years of

age living in the Greater Toronto Area with a chronic SCI (C1-T12, AIS A-D, ≥2 years post

impairment) of traumatic and non-traumatic etiology.23

The exclusion criteria, of the primary

study, consisted of any subjects with a previous or current history of: angina, myocardial

infarction, atypical chest pain, coronary artery bypass or revascularization, aortic stenosis,

uncontrolled arrhythmia or left bundle branch block, hypertrophic cardiomyopathy, severe

chronic obstructive pulmonary disease requiring oral steroids or home oxygen, diaphragmatic

pacer, and stroke.23

The subjects underwent medical screening, electrocardiogram, and chart

review to ensure that they met the inclusion and exclusion criteria.23

Overall, out of the 125

subjects who were screened, 100 consented to participate, 10 withdrew their consent, and three

did not meet the inclusion criteria and thus a total final sample of 87 subjects met the inclusion

criteria23

; 75 subjects had ECG data collected. Both primary and secondary studies were

approved by the University Health Network Research Ethics Board (REB#:09-019-DE) and the

secondary study was also approved by the University of Toronto Office of Research Ethics

(REB#:30133).

HRV, as measured via ECG, was collected in accordance with the Task Forceiii

guidelines.

Subjects were asked to abstain from caffeine and nicotine, and fast for at least 8 hours prior to

the ECG collection session. The subjects were also instructed to refrain from exercise 24 hours

prior to the session. The ECG data were collected between 9:00am-1:00pm. The subject was

transferred to a supine position onto a bed, in a quiet and temperature controlled (24◦C) room and

iii European Society of Cardiology and the North American Society of Pacing and Electrophysiology Task Force

HRV guidelines developed in 1996

Page 32: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

19

allowed to rest for 20 minutes before collecting continuous 3-lead ECG (lead II system) for ten

minutes, at a sampling rate of 1000Hz (PowerLab/16SP; AD instruments, Inc., Bella Vista,

Australia).23

3.2 Study variables

3.2.1 Heart rate variability indices and related factors

Five HRV indices were selected based on the literature findings: LF:HF (primary index), LF,

HF, RMSSD and pNN50.

In addition to collecting ECG, the demographics and health status of each subject were also

recorded in the primary study. Variables that were hypothesized to have an influence on HRV

were included in this study (see Table 3).

Table 3. Potential HRV-related variables selected from the primary data

Construct of interest Measurement Method

SCI impairments Time post injury (years), neurological level of

injury, severity of injury (complete or

incomplete), and etiology of injury (traumatic

or non-traumatic)

Age Age (years)

Sex Sex (male/female)

Medications Beta blockers, calcium channel blockers and

any other cardiac rhythm drugs

Obesity BMI (kg/m2) and WC (cm)

Smoking status Current smoking status and smoking history

(yes/no)

Page 33: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

20

Family history of heart disease Family history of heart disease (yes/no)

Sleep deprivation Sleep apnea (yes/no)

Cardiorespiratory fitness Absolute VO2 peak (L/min), Relative VO2

peak (ml/kg/min) and peak heart rate (bpm)

Physical status Self-reported physical activity: LTPAQ-SCI

(min/week)

Self-reported independence in ADL’s: SCIM-

III (/100)

Measured motor impairment: LEMS (/50)

Chronic disease Number of co-morbidities (/7)

Abbreviations: BMI, body mass index; WC, waist circumference; LTPAQ-SCI, leisure time

physical activity questionnaire; ADLs, activities of daily living; SCIM-III, spinal cord

independence measure; LEMS, lower extremity motor score

3.3 Subject selection.

3.3.1 Electrocardiogram recordings

All of the ECG recordings were reviewed visually with the assistance of an internist with

expertise in cardiovascular stress testing and ECG monitoring (Dr. P. Oh). For each subject, the

rate and rhythm (normal sinus rhythm, bradycardia, or tachycardia), presence of premature atrial

and/or ventricular contractions, electrical artifact and visual variability observed in the RR

intervals were reviewed. If the subject displayed frequent premature contractions (greater than

ten per minute), arrhythmias, or excessive artifact that prevented the proper analysis of the RR

intervals, they were excluded from the dataset for detailed analysis.

3.3.2 Medications

Medications were reviewed in consultation with a physiatrist (Dr. C. Craven) and internist (Dr.

Oh). Subjects taking medications which could have an influence on HRV (beta blockers, calcium

Page 34: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

21

channel blockers that influence cardiac conductioniv

such as diltiazem and verapamil, and any

other cardiac anti-arrhythmic drugs such as amiodarone, procainanmide, encainide and

flecainide) were excluded from the study.

3.4 Heart rate variability analysis

HRV analysis was conducted using LabChart

(version 7.0). According to the Task Force, the

gold standard for HRV short term recording analysis is a five minute interval. Therefore, the ten

minutes of ECG were divided into three segments of five continuous minutes; first five minutes

(t=0 - t=300 seconds), middle five minutes (t=150 - t=450 seconds) and last five minutes (t=300

- t=600 seconds) with the aim to select the segment with the least noise interference. Each five

minute ECG recording was then reviewed to confirm that all and only the R peaks were marked

(Figure 3a). The Poincaré Plotv was checked to examine the normal and ectopic

vi RR interval

ranges (Figure 3b) and to detect any ectopic islandsvii

. Physiologically, “ectopic” indicates any

cardiac activity not originating from the SA node.60

Ectopic islands were detected in 26.79% of

the subjects and occurred mainly due to technical error or unknown causes. The details of the

ectopic islands per subject are summarized in Appendix C. To omit ectopic islands, according to

the noise-omitting method of Young and colleagues,61

the data was filtered with a 45Hz low pass

filter (Figure 4). The following post-filtered variables were recorded for the three segments:

average heart rate, SDNN, SDANN, RMSSD, NN50, pNN50, total power, VLF, LF, HF, LF:HF,

and noise/ectopic/artifact percentages. After analyzing all three ECG segments, the segments for

each subject with the highest percentage of normal i.e. lowest percentage of ectopic beats, was

included in the analysis. If the percentage of normal and ectopic beats were equal in all three

segments for a particular subject, then a segment was randomly chosen using a computer-based

randomizer (http://www.random.org/). If only the ectopic beats were all equal (0%) then the

highest percentage of normal was chosen (the one closest to 100%).

ivAny calcium channel blocker ending with “ine” only influences blood pressure for example amlodipine and

nifedipine; they decrease blood pressure, but do not affect heart rate v The Poincaré Plot is a LabChart

software feature used to assess the lengths of the RR intervals by plotting the

length of each RR interval against the length of the following RR interval vi

Defined as “ectopic” by the LabChart

program vii

The term “ectopic islands” refers to the clustering of certain data points

Page 35: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

22

Figure 3. Representative example of HRV analysis using LabChart

(v.7.0). a. The threshold

was set and the R peaks were determined. The RR intervals are also referred to as NN intervals.

b. The Poincaré Plot was used to examine the lengths of the RR intervals. The interval ranges

indicates whether each RR interval is within the normal or ectopic range.

a)

b)

Page 36: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

23

Figure 4. Representative example of Poincaré Plot before and after the application of 45Hz

low pass filter. The arrows in the first diagram indicate two clusters of data, referred to as

ectopic islands. After the filter was applied, the ectopic islands were removed and the noise in

the data was reduced from 9.2 % to 0%.

Page 37: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

24

3.5 Statistical Analysis

Statistical analysis was conducted using IBM SPSS Statistics v.22 and is described per study

objective. If the data were not normally distributed a median was reported and if the data were

normally distributed a mean was reported.

3.5.1 Objective 1: Heart rate variability frequency distributions

To describe the distribution of HRV in chronic and traumatic SCI, descriptive statistics were

reported [mean and standard deviation or median and interquartile range (IQR)]. The frequency

distributions, for each HRV index, were also plotted and the distribution was described using

skewness and kurtosis. To assess whether the data were normally distributed, a Kolmogorov-

Smirnov (K-S) test was administered and boxplots were checked for major outliers (a minor

outlier was defined as 1.5xInterquartile range (IQR) outside the central box and a major outlier

as 3.0xIQR outside the central box). If K-S p>0.05 and there were no major outliers in the

boxplots, then the data were considered normally distributed. Furthermore, if the primary HRV

index (LF:HF) displayed any outliers, the characteristics of the subjects who were outliers were

examined with the aim to postulate possible reasons.

3.5.2 Objective 2: Comparison of heart rate variability based on level and/or severity of injury

HRV parameters and the HRV-related factors were compared between: a. Level of injury (below

versus above T5), b. Severity of injury (complete versus incomplete) and c. Level and severity of

injury. For normally distributed data, an independent t-test or ANOVA was administered. For

non-normally distributed data, a Mann-Whitney or Kruskal-Wallis test was used to compare

across cohorts. Following ANOVA or Kruskal-Wallis, if there was a significant difference a

post-hoc test was administered and was adjusted for multiple comparisons. Furthermore, if the

primary HRV index (LF:HF) displayed any outliers, the data was examined to ensure that the

outliers were not responsible for the results observed. For the categorical HRV-related factors, if

in the chi square output the expected frequencies in each cell was greater than five then a

Pearson Chi-Square test was used, if less than five then a Fisher’s exact test was chosen. An

alpha of 0.05 was set as the level of significance.

Page 38: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

25

3.5.3 Objective 3: Assessing the LF and HF indices

The relationship between the LF and HF indices was examined using Spearman’s rho correlation

co-efficient for the entire sample and per cohorts (level and/or severity of injury). For the entire

sample, Spearman’s was also used in order to determine whether there is a relationship between

LF, HF, LF:HF and the scalar HRV-related factors. The relationship between LF, HF, age at

injury and resting systolic blood pressure was assessed using Spearman’s for the entire sample

and in the cohort considered to be the most vulnerable to CVD (complete and equal to/above

T5). The strength of each relationship was assessed using the following descriptors: r=0.0-0.25

little or no relationship, r=0.26-0.50 fair relationship, r=0.50-0.75 moderate to good relationship,

and r>0.75 good to excellent relationship.62

A multiple linear regression analysis was used to

examine the relationship of CVD risk factors (age, waist circumference and peak heart) and the

LF and HF indices. The relationship was assessed for the entire sample and in the cohort most

vulnerable to CVD (complete and equal to/above T5). If the assumption of linearity and

normality were not met, bootstrapping was conducted. An alpha of 0.05 was set as the level of

significance.

Page 39: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

26

Chapter 4:

4 Results

4.1 Subject selection

The primary data set consisted of 75 subjects with non-traumatic and traumatic injuries. The

non-traumatic subjects (n=13) were excluded from the data set based on etiology of injury. After

the resting ECG was reviewed for each subject, three subjects were excluded: Two subjects

displayed frequent premature ventricular contractions (PVCs ≥10/min) and one due to technical

difficulties with ECG data collection. In addition, three subjects were excluded based on the

medications reported: Two subjects were on beta blockers, and one subject was taking both a

beta-blocker and a calcium channel blocker diltiazem (Tiazac XL) which decreases heart rate.

The final sample size included a total of 56 subjects which were then further subdivided based on

level and severity of injury (Figure 5). The characteristics of the participants are summarized in

Table 4.

Page 40: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

27

Figure 5. CONSORT flowchart reflecting the inclusion and exclusion of the final data

sample. A total of 56 subjects were included for analysis.

Assessed for eligibility (N=75)

Analyzed (N=56)

Complete and equal to or above

T5

(N=27)

Complete and below T5

(N=11)

Incomplete and equal to or above

T5

(N=10)

Incomplete and below T5

(N=8)

Excluded (N=19)

- Non-traumatic SCI (N=13)

- ECG: PVCs≥10/min (N=2); Technical error (N=1)

- Medications: β-blockers (N=2); β-blocker and Ca2+ channel blocker (N=1)

Page 41: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

28

Table 4. Demographics and vital signs of the participants in total sample and per cohort

Total

Sample

Complete and

equal to/above

T5

Complete and

below T5

Incomplete

and equal

to/above T5

Incomplete

and below T5

N 56 27 11 10 8

Age (years) 46.75±12.44 46.30 ±10.59 44.45 ±15.01 53.40 ±12.17 43.13 ±14.12

Time post injury

(years)

14.23±9.86 17.26±10.16 16.55±9.54 9.10±7.40 7.25±6.48

Sex

(males/females)

44/12 22/5 9/2 8/2 5/3

BMI (kg/m2) 26.13±4.84 25.34±4.54 25.73±4.30 28.44±6.91 26.48±3.11

WC (cm) 95.65±14.56 94.60±14.89 96.56±14.63 97.37±18.19 95.81±10.05

HR (bpm) 61.67±8.98 59.41±7.54 65.71±9.98 60.52±9.72 65.52±7.07

SBP (mmHg) 109.77±16.8

6

101.19±12.17 114.73±17.05 113.00±16.42 127.88±14.51

DBP (mmHg) 71.38±13.08 64.63±10.46 78.00±12.62 74.00±10.87 81.75±13.47

All values are mean ± standard deviation or as otherwise indicated

Abbreviations: BMI, body mass index; WC, waist circumference; HR, heart rate; SBP, systolic

blood pressure; DBP, diastolic blood pressure

4.2 Frequency distributions of the heart rate variability indices

Table 5 summarizes the LF:HF descriptive statistics for the entire sample. The frequency

distribution was positively skewed and leptokurtic indicating that the values cluster at the lower

end and it is a pointy and heavy-tailed distribution (Figure 6). The results showed that the data

were significantly different than a normal distribution as K-S p<0.001 and there were major

outliers present in the boxplot (two major and two minor) (Figure 7). The characteristics of the

Page 42: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

29

subjects who were outliers are summarized in Appendix D, Table 1. To further examine the

influence of outliers, they were removed and the data were reassessed for normality [N=52,

Mdn=1.08(0.59, 2.45), IQR=1.86]. The frequency distribution of the LF:HF without the outliers

is in Appendix D, Figure 1. The data were still not normally distributed since K-S test p=0.002,

however, the distribution became less positively skewed (+1.13) and leptokurtic (+0.57). The

descriptive statistics of the secondary HRV indices (LF, HF, RMSSD, pNN50) are summarized

in Table 5, and none of the indices were normally distributed. The frequency distributions and

the boxplots for each HRV index are in Appendix E. The LF, HF, and RMSSD distributions

were positively skewed and leptokurtic, whereas the distribution of pNN50 was positively

skewed and platykurtic. The boxplots displayed: Five minor outliers in LF, three major and two

minor outliers in HF, three minor outliers in RMSSD, and none in pNN50. The descriptive

statistics for total power can be found in Appendix F, Table 1.

Table 5. Descriptive statistics for each HRV index in the entire sample (N=56)

LF:HF LF HF RMSSD pNN50

Median

(Lower,

Upper

quartile)

1.21

(0.63,2.85)

460.20 ms2

(207.73,1266.33)

362.32 ms2

(143.69,1086.72)

35.75 ms

(20.03,59.03)

6.62%

(1.23,23.35)

IQR 2.22 1058.60 ms2 943.03 ms

2 39.00 ms 22.12%

Skewness +2.70 +1.68 +2.29 +1.73 +1.02

Kurtosis +8.45 +2.13 +5.16 +3.39 -0.08

p-value† p<0.001* p<0.001* p<0.001* p=0.007* p<0.001*

†Kolmogorov-Smirnov (K-S) test; p≤0.05*

Abbreviations: LF:HF, low frequency to high frequency ratio; LF, low frequency; HF, high

frequency, RMSSD, square root of the mean squared differences of the consecutive NN

intervals, pNN50, proportion of the number of interval differences of the consecutive NN

intervals greater than 50ms; IQR, interquartile range

Page 43: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

30

Figure 6. Frequency distribution of LF:HF. The data was positively skewed (+2.70) and

leptokurtic (+8.45).

Figure 7. Boxplot representing LF:HF distribution. The asterisks represent major outliers

(3.0xIQR outside the central box) and the circles represent minor outliers (1.5x IQR outside the

central box).

Page 44: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

31

4.3 Heart rate variability comparisons across level and/or severity of injury

The HRV indices were compared between two cohorts: Level of injury (above versus below T5)

and severity of injury (complete versus incomplete injury). None of the HRV indices differed

based on level of injury (Table 6) or severity of injury (Table 7). The comparisons of total

power can be found in Appendix F, Tables 2 and 3.

Table 6. Comparison of HRV indices based on level of injury

Median (Lower, Upper quartile)

HRV Index Below T5

(N=19)

Above/Equal to

T5 (N=37)

p –Value†

LF:HF 1.54 (0.76,3.32) 1.10 (0.60,2.71) 0.431

LF (ms2) 667.88

(209.71,1945.83)

330.22

(187.32,330.22)

0.210

HF (ms2) 436.29

(177.26,1173.55)

354.27

(133.73,883.00)

0.869

RMSSD (ms) 34.58

(21.45,61.54)

39.16

(19.48,58.94)

0.869

pNN50 (%) 8.08

(1.49,26.64)

5.24

(1.13,23.07)

0.883

†Mann-Whitney test; p≤0.05*

Abbreviations: LF:HF, low frequency to high frequency ratio; LF, low frequency; HF, high

frequency, RMSSD, square root of the mean squared differences of the consecutive NN

intervals, pNN50, proportion of the number of interval differences of the consecutive NN

intervals greater than 50ms

Page 45: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

32

Table 7. Comparison of HRV indices based on severity of injury

Median (Lower, Upper quartile) or

Mean ± Standard deviation

HRV Index Complete

(N=38)

Incomplete

(N=18)

p –Value

LF:HF 1.60 (0.69,3.06) 0.96 (0.47,1.76) 0.150†

LF (ms2) 460.20

(223.75,1075.85)

351.74

(104.85,2724.76)

0.902†

HF (ms2) 362.32

(139.41,876.10)

390.06

(127.68,2026.22)

0.461†

RMSSD (ms) 38.91 ±25.28 58.77 ±46.95 0.107§

pNN50 (%) 5.00 (1.14,22.10) 8.31 (1.78,26.95) 0.680†

†Mann-Whitney test;

§ Independent t-test; p≤0.05*

Abbreviations: LF:HF, low frequency to high frequency ratio; LF, low frequency; HF, high

frequency, RMSSD, square root of the mean squared differences of the consecutive NN

intervals, pNN50, proportion of the number of interval differences of the consecutive NN

intervals greater than 50ms

Similarly, when examined across the four cohorts, based on level and severity of injury, none of

the HRV indices were significantly different (Table 8). The comparisons in total power can be

found in Appendix F, Table 4. To examine the influence of the outliers in LF:HF (the primary

HRV index) a boxplot with and without the outliers was examined (Figure 8) and the new values

were re-tested for comparisons. There were still no differences across the cohorts (p=0.133)

when the outliers were omitted. The characteristics of the subjects who were the initial outliers

can be found in Appendix D, Table 1. Refer to Appendix G to see the boxplots across level and

severity of injury for each HRV index.

Page 46: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

33

Table 8. Comparison of HRV indices based on level and severity of injury

Median (Lower, Upper quartile) or Mean ± Standard deviation

HRV

Index

Complete and

equal to/above

T5 (N=27)

Complete and

below T5

(N=11)

Incomplete and

equal to/above

T5 (N=10)

Incomplete and

below T5 (N=8)

p –Value

LF:HF 1.10

(0.55,2.91)

2.48

(1.51,4.84)

1.16

(0.72,1.93)

0.62

(0.35,2.00)

0.070†

LF

(ms2)

330.22

(207.07,741.10)

667.88

(455.41,1945.83)

351.74

(102.30,2810.39)

463.81

(122.14,2483.00)

0.379†

HF

(ms2)

370.37

(124.05,879.65)

295.37

(177.26,660.09)

284.53

(125.28,2321.10)

794.90

(144.60,2661.62)

0.816†

RMSSD

(ms)

40.36 ±27.92 35.38 ±17.88 65.60±55.62 50.22 ±35.00 0.373§

(Welch)

0.264§

(Brown-

Forsythe)

pNN50

(%)

4.76

(1.11,22.50)

7.91

(1.49,15.43)

6.94

(1.78,24.70)

11.73

(0.86,28.12)

0.980†

†Kruskal-Wallis test;

§One way ANOVA test; p≤0.05*

Abbreviations: LF:HF, low frequency to high frequency ratio; LF, low frequency; HF, high

frequency, RMSSD, square root of the mean squared differences of the consecutive NN

intervals, pNN50, proportion of the number of interval differences of the consecutive NN

intervals greater than 50ms

Page 47: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

34

Figure 8. Boxplot representation, with and without the outliers, of LF:HF based on level and severity of SCI. For each color-coded

cohort there are two boxplots, the darker shaded boxes represent data with the outliers while the lighter shaded boxes represent the data

after the initial outliers were omitted. The outlier numbers correspond with the subject numbers.

Page 48: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

35

4.3.1 Comparison of heart rate variability related factors across cohorts

SCIM-III (p=0.036), relative VO2 peak (p=0.038) and peak heart rate (p=0.033) were all greater

in individuals with an injury below T5. However, time post injury (p=0.501), BMI (p=0.883),

LTPAQ-SCI (p=0.668), LEMS (p=0.111), number of co-morbidities (p=0.179), age (p=0.221),

waist circumference (p=0.830) and absolute VO2 peak (p=0.088) were all similar. The medians

or means for each factor are summarized in Appendix H, Table 1.

Time post injury (p=0.001) was greater in individuals with a complete injury, whereas LEMS

(p<0.0001) and SCIM-III (p=0.003) were greater in the incomplete cohort. BMI (p=0.352),

LTPAQ-SCI (p=0.106), number of co-morbidities (p=0.622), age (p=0.393), waist circumference

(p=0.720), relative VO2 peak (p=0.496), absolute VO2 peak (p=0.232) and peak heart rate

(p=0.705), were all the similar. The medians or means for each factor are summarized in

Appendix H, Table 2.

There were no differences in the HRV-related categorical factors based on level of injury

(Appendix H, Table 3) or severity of injury (Appendix H, Table 4).

The HRV-related factors were also compared based on level and severity of injury. LTPAQ-SCI

(Figure 9), LEMS (Figure 10) and SCIM-III (Figure 11) all revealed significant differences

between the cohorts. Individuals with a complete injury below T5 reported less leisure time

activity than individuals with an incomplete injury below T5 (p=0.019). Individuals with a

complete injury equal to/above T5 displayed a lower LEMS than individuals with an incomplete

injury equal to/above T5 (p<0.001) and incomplete and below T5 (p<0.001). Also individuals

with a complete injury below T5 revealed a lower LEMS than individuals with an incomplete

injury equal to/above T5 (p=0.001) and incomplete and below T5 (p=0.008). Finally, individuals

with a complete injury equal to or above T5 had a lower SCIM-III score than individuals with an

incomplete injury below T5. On the other hand, time post injury (p=0.016 but was not significant

for each between group comparison), BMI (p=0.798), number of co-morbidities (p=0.583), age

(p=0.271), waist circumference (p=0.958), relative VO2 peak (p=0.302), absolute VO2 peak

(p=0.254 and p=0.154) and peak heart rate (p=0.077 and p=0.096) were similar across the four

cohorts. The medians and means for each factor are summarized in Appendix H, Table 5 and the

Page 49: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

36

follow-up analysis is in Appendix H, Table 6. There were no differences in the HRV-related

categorical factors based on level and severity of injury (Appendix H, Table 7).

Figure 9. Boxplot representation of LTPAQ-SCI based on level and severity of injury. The

complete and below T5 cohort has a lower LTPAQ-SCI score than the incomplete and below T5

cohort (p=0.019).

Page 50: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

37

Figure 10. Boxplot representation of LEMS based on level and severity of injury. The

complete and equal to/above T5 cohort had a lower LEMS than the incomplete and equal

to/above T5 (p<0.001) and incomplete and below T5 (p<0.001) cohorts. Also, the complete and

below T5 cohort had a lower LEMS than the incomplete and equal to/above T5 (p=0.001) and

incomplete and below T5 (p=0.008) cohorts.

Figure 11. Boxplot representation of SCIM-III based on level and severity of injury. The

complete and equal to/above T5 cohort had a lower SCIM-III score than the incomplete and

below T5 cohort (p=0.010).

Page 51: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

38

4.4 Assessing the heart rate variability frequency domain indices: LF, HF and LF:HF

4.4.1 Relationship between LF and HF

A high positive correlation was found between LF and HF for the entire sample (r=0.708,

p<0.0001) (Figure 12) and in each of the cohorts (Table 9 and Table 10). The scatter plots of LF

and HF based on level and/or severity of injury can be found in Appendix I.

Figure 12. The relationship between LF and HF for total sample. There was a positive linear

correlation between LF and HF in individuals with a chronic and traumatic SCI. The coefficient

of determination (R2) indicated that 50.8% of the variation in LF was shared by HF.

Page 52: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

39

Table 9. Relationship of the LF and HFindices based on level or severity of injury

Level of Injury Severity of Injury

Below T5 Above/Equal to T5 Complete

Injury

Incomplete

Injury

Spearman’s Rho† 0.665

p=0.002*

0.730

p<0.0001*

0.590

p<0.0001*

0.849

p<0.0001*

†Spearman’s correlation test; *p≤0.05

Table 10. Relationship of the LF and HF indices based on level and severity of injury

Complete and

equal

to/above T5

Complete and

below T5

Incomplete

and equal

to/above T5

Incomplete

and below T5

Spearman’s

Rho†

0.648

p<0.0001*

0.655

p=0.029*

0.879

p=0.001*

0.833

p=0.010*

†Spearman’s correlation test; *p≤0.05

4.4.2 Relationship between LF, HF, LF:HF and influencing factors

The relationships between LF, HF, LF:HF and the HRV-related factors were examined in the

entire sample (Table 11). LF was negatively correlated with age (r=-0.366, p=0.006) and time

post injury (r=-0.384, p=0.003). Similarly, HF was negatively correlated with age (r=-0.317,

p=0.017) and time post injury (r=-0.344, p=0.009) and was positively correlated with SCIM-III

(r=0.299, p=0.025). LF:HF did not reveal a significant relationship with any of the HRV indices.

The scatter plots of the significant correlations listed in Table 13 can be found in Appendix J.

Table 11. The relationship between LF, HF indices and the scalar HRV-related factors

LF† HF

† LF:HF

Age -0.366 -0.317 -0.012

Page 53: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

40

p=0.006* p=0.017* p=0.930

Time post injury -0.384

p=0.003*

-0.344

p=0.009*

-0.043

p=0.750

BMI -0.093

p=0.497

-0.106

p=0.436

-0.043

p=0.750

WC -0.091

p=0.506

-0.218

p=0.107

0.153

p=0.259

Absolute VO2 peak 0.052

p=0.722

-0.065

p=0.652

0.201

p=0.162

Relative VO2 peak 0.047

p=0.748

-0.006

p=0.969

0.146

p=0.311

Peak heart rate 0.096

p=0.506

0.031

p=0.832

0.084

p=0.564

LTPAQ-SCI -0.010

p=0.940

0.124

p=0.361

-0.182

p=0.180

LEMS 0.078

p=0.569

0.118

p=0.386

-0.105

p=0.442

SCIM-III 0.224

p=0.097

0.299

p=0.025*

-0.133

p=0.330

Number of

comorbidities

-0.204

p=0.131

-0.256

p=0.057

0.084

p=0.539

†Spearman’s correlation test; *p≤0.05

Abbreviations: BMI, body mass index; WC, waist circumference; LTPAQ-SCI, leisure time

physical activity questionnaire-spinal cord injury; LEMS, lower extremity motor score; SCIM-

III, spinal cord independence measure

Page 54: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

41

There was no significant relationship between age at injury, resting systolic blood pressure, LF

or HF when examined in the entire sample (Range of values: Spearman’s rho (r=-0.005 –

r=0.073; p>0.05). Similarly, in individuals with a complete injury at or above T5 (hypothesized

to be the cohort at greatest risk for CVD) also showed no significant relationship between age at

injury, resting systolic blood pressure, LF or HF (Range of values: Spearman’s rho (r=-0.008 –

r=0.170; p>0.05).

4.4.3 Predicting LF and HF from heart rate variability-related factors

The HRV related factors, age, waist circumference, peak heart which are also CVD risk factors,

were assessed using multiple linear regressions to determine whether they can predict LF and/or

HF. Table 12 and Table 13 depict that there were no significant relationships when assessed in

the entire sample. Similarly, when assessed in individuals with a complete injury equal to/above

T5 (expected to be more vulnerable to CVD) there was still no significant relationship; results

are summarized in Appendix K.

Table 12. Multiple linear regression analysis to predict LF for the entire sample (R2=0.039)

Parameter Regression

Coefficient

95% CI p-value

Intercept 830.89 -1533.96,3634.78 0.557

Age -14.67 -39.07,14.16 0.255

WC 4.89 -15.55,22.43 0.604

Peak heart rate 2.48 -6.88,13.38 0.661

*p≤0.05

Abbreviations: WC, waist circumference; CI, confidence interval

Page 55: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

42

Table 13. Multiple linear regression analysis to predict HF for the entire sample (R2=0.009)

Parameter Regression

Coefficient

95% CI p-value

Intercept 1252.59 -2592.42,4551.481 0.483

Age -9.68 -38.97,24.47 0.601

WC 2.32 -14.70,22.20 0.789

Peak heart rate -1.46 -12.51,10.84 0.798

*p≤0.05

Abbreviations: WC, waist circumference; CI, confidence interval

Page 56: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

43

Chapter 5:

5 Discussion

This is the first study to examine the HRV characteristics in a large and representative sample of

individuals with chronic traumatic SCI. The primary and secondary HRV indices were not

normally distributed and revealed high inter-individual variability in HRV which is similar to

what has been reported in the able-bodied population. Nevertheless, it was unclear why some

subjects (n=4) were outliers when LF:HF was examined in the entire sample. Also, the primary

and secondary HRV indices were not significantly different when compared across cohorts. It is

possible that the lack of a relationship between HRV and level and/or severity of injury is due to

other unaccounted factors influencing the ANS, or because between-subject comparisons were

made instead of within-subjects. In addition, even though some of the HRV-related factors were

different across cohorts, the HRV results were still similar when compared based on level and/or

severity of injury. Therefore, the HRV-related factors may have little influence on HRV in

individuals with chronic traumatic SCI in this study, despite being reported to influence HRV in

the able-bodied population. In addition, a strong positive linear relationship was found between

LF and HF and thus the LF:HF ratio may remain unchanged in individuals with chronic

traumatic SCI. Finally, the bivariate and multivariate analysis between LF, HF and the potential

factors that may influence HRV did not display any significant relationships. There are many

biophysiological changes that occur after a SCI and thus it may be challenging to determine

which factors may be influencing the HRV results.

The frequency distributions of the primary and secondary HRV indices were all positively

skewed indicating that the HRV values in this SCI sample were mainly low. The observed high

inter-individual variability in HRV aligns with the able-bodied population. Nunan et al.37

conducted a systematic review to determine the normal values of HRV in healthy adults. The

systematic review was comprised of studies that measured short term HRV, in accordance with

the Task Force guidelines, in healthy adult participants (n≥30). Nunan and colleagues37

found

large inter-individual variations (up to 260, 000% of variation was reported) between the studies,

especially for the frequency domain measures. Nevertheless, according to Malliani et al.44

it

should not be surprising that there are no HRV normative values within the healthy population

Page 57: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

44

since the cardiac ANS is a dynamic system, consists of a large range of values, and is influenced

by a number of internal and external factors. Table 14 depicts the mean and median of the

LF:HF, LF and HF that were reported in healthy subjects based on the systematic review37

and in

chronic traumatic SCI based on the thesis findings. Both the mean and median values depict a

large standard deviation and interquartile range, respectively. Consequently, even though the

HRV values cannot be generalized to the SCI population, the values do indicate that, similar to

able-bodied subjects, there are inter-individual differences in cardiac ANS function within

individuals with chronic traumatic SCI. In addition, the large variations in HRV values

emphasize the importance of reporting HRV as a median as opposed to a mean.

Table 14. Comparison of inter-individual variations in HRV between healthy subjects and

chronic traumatic SCI

Mean ±SD

Median (Lower, Upper quartile)

HRV Indices

Healthy Subjects

(Systematic review)37

Chronic Traumatic SCI

(Current thesis findings)

LF:HF 2.8±2.6

2.1 (1.1, 11.6)

2.2±2.6

1.2 (0.63, 2.85)

LF ms2 519±291

458 (193, 1009)

886±1038

460 (208,1266)

HF ms2 657±777

385 (82, 3630)

834±1138

362 (143,1087)

Abbreviations: SD, standard deviation; LF:HF, low frequency to high frequency ratio; LF, low

frequency; HF, high frequency

Page 58: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

45

The LF:HF is a primary measure of HRV as it is reported to examine the level of sympathetic to

parasympathetic activity and thus assesses the modulation of the cardiac ANS. After the LF:HF

outliers were examined, the characteristics of the subjects who were considered outliers did not

provide any further insights. The subjects did not demonstrate any consistent factors to account

for the high LF:HF as they varied in terms of both level and severity of injury, and displayed

diverse HRV-related factors (e.g. differences in sex, time post injury, smoking status, and

physical fitness). In addition, when the subjects were compared to the median of the rest of the

study sample, each subject displayed different factors that were outside the median range. Based

on the data available it was not possible to determine why they were outliers leading one to

expect that perhaps more sensitive biophysiological differences, for instance arterial stiffness

and/or high levels of inflammatory cytokines, may be responsible.

Given that both ANS dysfunction and CVD have been reported to be linked with level and

severity of SCI,4-6,22

it was hypothesized that HRV will differ across cohorts. However, in this

study there were no differences and thus the results suggest that HRV does not depend on type of

SCI (injury level and completeness). Nevertheless, cardiac autonomic regulation is part of two

larger complex systems: the cardiovascular system and the autonomic nervous system. The

cardiovascular component includes peripheral circulation, often altered in SCI, and influences

cardiac function.63

In addition, heart rate is not only modulated by the autonomic nervous

system, but also by the intrinsic cardiac system, baroreflex function, respiration and humoural

factors.63

The ANS controls many other body functions that may also be disrupted in individuals

with a SCI depending mainly on the level and completeness of pathology44

and thus cardiac

autonomic function may be influenced by other intrinsic ANS dysfunctions. Consequently, the

differences in pathophysiology of cardiac autonomic function in individuals with a chronic

traumatic SCI may not be ideal to examine in isolation without considering other cardiovascular

abnormalities. In addition, despite the HRV differences reported in the literature (described in

Chapter 1, Section 1.3.1) some investigators have emphasized that HRV is not a direct measure

of the parasympathetic and sympathetic nerve activity, but instead quantifies cardiac autonomic

responsiveness.7,45

The interpretation of HRV values raises the question if discrete measurement

of HRV at a single point in time is indicative of cardiac ANS modulation and whether comparing

discrete HRV values between subjects has biological merit.

Page 59: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

46

Based on only level of injury, the results showed that relative VO2 peak, peak heart rate, and

SCIM-III scores, were greater in the subgroup of individuals with an injury below T5. In

agreement with the literature, Simmons and colleagues64

classified level of SCI as a major

determinant of relative VO2 peak. The established general reference value for cardiorespiratory

fitness is significantly higher in people with paraplegia (median: 16.0 ml/kg/min) than those with

tetraplegia (median: 8.8 ml/kg/min).64

Furthermore, in accordance with the literature,

Ravensbergen and colleagues25

reported that a person with a SCI at the level of T5 or above

attains a reduced peak heart rate during exercise due to diminished sympathetic control. Hagen et

al.17

reported that individuals with complete tetraplegia are unable to raise their peak heart rate to

more than 125 bpm. The increased heart rate observed during exercise has been thought to be the

result of vagal withdrawal.4,25

As SCIM-III provides insight regarding physical capacity, it was not surprising that SCIM-III

was related to the level of injury. For instance, an increase in physical capacity signifies that the

individual has greater voluntary functional muscle mass.64

Even though the results indicate

greater cardiorespiratory fitness (relative VO2 peak and peak heart rate) and greater physical

capacity (SCIM-III) in individuals with an injury below T5, cardiorespiratory fitness and

physical capacity appeared to have a minimal impact on the HRV results as they failed to

contribute to the expected higher HRV values. From our analysis we would have to conclude

that, cardiorespiratory fitness level and physical capacity may not have a major influence on

HRV when assessed based on level of SCI alone.

Physical capacity, measured by LEMS and SCIM-III, was higher in incomplete injuries and the

time post injury was longer in complete injuries. Previous findings have shown that level and

severity of SCI may influence HRV results,2 but other SCI characteristics have not been

examined. For instance, it is unknown if and how time post SCI influences HRV. Based on this

study, however, time post injury in addition to physical capacity may not have an influence on

HRV when assessed based on severity of injury since there were no differences observed

between the cohorts.

As expected LTPAQ-SCI, LEMS and SCIM-III, differed based on both level and severity of

injury. Bucholz and colleagues65

found that LTPAQ-SCI has been associated with a decrease of

Page 60: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

47

CVD risk factors in individuals with chronic traumatic SCI. Previous studies have shown that

individuals with a complete and cervical injury report less leisure time physical activity than

those with an incomplete and lower level of injury.65-66

In addition, the differences in physical

activity are probably because individuals with complete tetraplegia are limited in terms of in

which exercises they can participate and require greater assistance with exercise protocols.65-66

Also, individuals with high lesions, especially cervical, often have bradycardia and thus their low

level of cardiac sympathetic function makes it difficult to participate in physical activity.25

The

results did show that individuals with complete injury engaged in less leisure time physical

activity than those with an incomplete injury, but these findings were only observed in

individuals with injuries below the level of T5. A possible reason as to why, contrary to the

literature, severity of injury did not influence the physical activity results in individuals with high

level injuries, could be the method of measuring leisure time physical activity. Most studies use

the Physical Activity Recall Assessment for People with Spinal Cord Injury (PARA-SCI)

questionnaire which involves recording both activities of daily living in addition to leisure time

physical activity and is collected over three days as opposed to over a week. The present study,

however, used the LTPAQ-SCI which recorded the number of minutes of physical activity per

week and did not include activities of daily living.

Expected and consistent with the literature, LEMS were lower in individuals with a complete

injury and/or with an injury above the level of T5 illustrating less voluntary muscle function in

the lower limbs. Similarly, SCIM-III revealed lower scores for individuals with complete and

equal to/above T5 injuries than those with in incomplete and below T5 injuries. Lower physical

capacity and physical activity can be used to determine whether the individual is likely to be

sedentary64

and thus has been reported to contribute to lower HRV values.25

However, in this

study, despite the differences in the physical activity observed amongst the cohorts, the HRV

indices were still similar when compared based on both level and severity of injury. Therefore,

physical activity and capacity may not have a substantial impact on HRV in a chronic traumatic

SCI population.

When revisiting the theoretical framework illustrating the link between SCI and CVD for chronic

traumatic SCI, it is important to include impaired cardiac autonomic modulation as a component

of the disrupted autonomic nervous system (Figure 13). Many studies have reported a diminished

Page 61: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

48

LF in individuals with a cervical and/or thoracic SCI.2,29,42,48

A reduced LF, a purported cardiac

sympathetic marker, indicates a greater challenge to participate in physical activity25

and thus

contributes to sedentary behavior. A decrease in HF, a parasympathetic marker, is problematic

for this population since parasympathetic activity decreases the amount of work on the heart and

thus has been linked to restoring and protecting the cardiovascular system in other populations.67

Therefore, if LF does indeed decrease after a SCI, the positive relationship between LF and HF

may indicate increased risk of CVD development in individuals with a chronic and traumatic

SCI. Since the primary study excluded any subjects with a cardiac disease, including

arrhythmias, we have no indication of how arrhythmias interact with cardiac autonomic

modulation. Furthermore, with the non-modifiable factors of age, sex, genetic history, sedentary

lifestyle, smoking status and obesity being accounted for in this study albeit considering the

collinearity but demonstrating a lack of effect on the HRV indices suggests that other biological

contributors to CVD development will need to be examined in this chronic population.

Figure 13. Possible contributors to greater CVD risk in individuals with chronic traumatic

SCI. The theoretical framework represents the relationship between SCI and CVD while also

considering the findings from the thesis. Impaired cardiac autonomic modulation has been added

as a subgroup of disrupted autonomic nervous system to indicate that impairment also

contributes to CVD development in chronic traumatic SCI. The arrow between the sympathetic

Page 62: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

49

and parasympathetic cardiac activity is bidirectional since the ANS represents a shifting balance

between the two systems (Modified from Figure 1.0).

There is a debate in the literature regarding the relationship between LF and HF in individuals

with a SCI. In this study, LF and HF displayed a positive linear relationship when assessed in the

entire sample and based on level and/or severity of injury. After a cervical SCI, Claydon et al.2

observed lower LF and higher HF, whereas Grimm et al.30

and Wang et al.29

reported both lower

LF and HF. As for the thoracic group, Claydon et al.2 reported no change in LF and a reduced

HF while Bunten et al.42

and Castiglioni et al.48

reported reduced LF but no change in the HF

values. Although the findings from the current study did not show a decrease in any of the HRV

parameters, it did show that in the entire sample, 50.8% of the variation in LF was shared by HF

and thus a decrease or even an increase in both components could result in a similar LF:HF ratio.

The LF and HF relationship was strongest in individuals with an incomplete injury and equal

to/above T5, as 84.3% of the variability in LF was shared by HF. In incomplete injuries the cord

is not completely disconnected from the brain30,42

and with an injury that occurs at or above T5

the sympathetic activity is disrupted,25

thus it is possible that the modulation of ANS system was

altered to a greater extent in this group. There is low resting sympathetic tone in individuals with

a SCI in comparison with able-bodied subjects.26

Consequently, most investigators indicate that a

change in HF is required to align with the low levels of LF since the ANS re-balances to

maintain homeostasis.2,15,29-30

However, the physiological basis of how the ANS re-balances

remains unclear. However, Billman68

challenged the presumption of the ANS re-balancing and

argued that the LF index is not more indicative of sympathetic function but is rather a complex

combination between the two ANS branches along with other unidentified factors. Furthermore,

the correlation between the LF and HF indices does not meet the nine Bradford-Hill criteria69

for

causation and thus it is not certain that a change in LF caused a change in HF. Therefore, the

positive correlation between the LF and HF indices may not necessarily represent a re-balanced

ANS system. Despite the indeterminate physiological reasons for the positive correlation

observed between the LF and HF indices, overall this finding questions whether the LF:HF, the

most common HRV measure, is an appropriate marker of the cardiac sympatho-vagal balance in

individuals with chronic traumatic SCI.

Page 63: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

50

None of the HRV related factors that were tested displayed a significant correlation with the

LF:HF even though some variables did display a significant correlation with the LF and HF

indices. After examining the scatter plots of LF, HF and the significantly correlated HRV-related

factors, it appeared that the variation in LF and HF were only minimally shared by age, SCIM-

III, and time post injury (ranging from almost 1% to 15%). The mean age (46.75±12.44) reported

in the SCI and resting HRV literature is similar to the mean age reported in this study. It was

expected that as age increases sympathetic activity increases, represented by higher LF, and

parasympathetic activity decreases, represented by lower HF.32

Surprisingly, however, an inverse

relationship was observed between LF and age, but only 4.1% of variation in LF was shared by

age. Also, unexpectedly, HF displayed a weak negative relationship with age and only 1% of

variation in HF was shared by age.

A higher degree of physical capacity usually indicates capability to be physically active which in

turn has been linked to the predominance of the parasympathetic function.31,67,70-71

As a result, it

was surprising that the HF was only slightly positively correlated with SCIM-III; where only

0.8% of the variability in HF was shared by SCIM-III. However, SCIM-III may not be a good

indicator of physical capacity. The relationship between HRV and time post injury has not yet

been established for SCI, but the findings revealed that time post injury was negatively

correlated with both LF and HF; 14.7% of the variability in LF was shared by time post injury,

and 12.8% of variability in HF was shared by time post injury. Therefore, the results might be

interpreted that as time post injury increased, both sympathetic and parasympathetic function

decreased illustrating the persistence of CVD risk (for the same reasons mentioned earlier). The

strengths of the relationships between LF, age and time post injury, and between HF, age, time

post injury and SCIM-III were all of fair magnitude at best and thus over interpretation without

further analysis is not warranted.

Similarly, the multiple linear regression analysis did not reveal any significant relationships

between the CVD risk factors (age, waist circumference, peak heart rate) and LF or HF when

examined in the entire sample. The multivariate relationship was also assessed in individuals

with a complete injury at or above T5, the cohort that is most likely to develop CVD, but again

no relationships were observed. Consequently, according to our results, the CVD risk factors,

which are also HRV-related factors, do not have a substantial impact on LF and HF in

Page 64: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

51

individuals with a traumatic and chronic SCI. The presence of a SCI, associated with various

physiological and functional changes, may minimize the impact of the HRV-related factors that

have been observed in the able bodied population.

When age at injury and resting systolic blood pressure were examined to determine if they were

contributors to HRV in both the entire sample and in the cohort with a complete injury equal

to/above T5, age at injury and LF or HF both appeared to have no relationship. Therefore, the

cardiac ANS does not behave differently if a SCI occurred at a younger or older age our range

being 37to 54 years. Resting systolic blood pressure has been reported to be linked with level

and severity of SCI; higher level and complete injuries display lower resting systolic blood

pressure5 and the reduction in blood pressure has been reported to occur due to the reduction in

sympathetic activity after a SCI.3 Again we could not demonstrate these expected findings in the

entire sample or in individuals with a complete injury at or above T5. Consequently, in

individuals with a chronic and traumatic SCI, it appears that the sympathetic and

parasympathetic cardiac activity is not influenced by sympathetic reduced resting blood pressure.

The ANS is a complex biological system and therefore it is difficult to ascertain what and how

other changing ANS activity could be influencing the HRV parameters without conducting

physiological experiments that challenge the ANS system over time.

5.1 Implications and future directions

Cardiac autonomic disturbances are believed to be a major contributor to the development of

CVD within the SCI population. Therefore, quantifying the cardiac parasympathetic and

sympathetic modulation of the heart, via a non-invasive measure, is important for diagnostic,

prognostic and/or rehabilitative purposes. However, HRV values for SCI have not yet been

established and the lack of HRV differences observed in this study indicate that HRV does not

directly reflect the anatomical sympathetic and parasympathetic autonomic innervations and

response of the heart in individuals with a chronic and traumatic SCI. The proposed HRV-related

factors, in addition to age at injury and resting systolic blood pressure did not have an impact on

the HRV results, which further emphasizes the complexity of the ANS. The findings from this

data set suggests limited potential for assessing HRV at a single point in time in individuals with

a chronic traumatic SCI to measure autonomic cardiac function..

Page 65: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

52

There is disagreement in the literature as to whether the subjects should be divided into three

levels i.e. cervical, high thoracic, low thoracic and lumbar, or two levels i.e. above a level of

injury and below a level of injury. Consequently, to check whether there are any HRV

differences between different types of SCI, in addition to a larger sample size, this work could be

repeated with a different cohort selection: cervical (C1-C8), high thoracic (T1-T5), and low

thoracic (T6-T12). The presence of a strictly cervical cohort allows for the assessment of

completely disconnected cardiac sympathetic innervation. In addition, West et al.5

have reported

that in individuals with a chronic SCI, autonomic completeness of the injury, which can be

estimated via catecholamine concentrations as well as blood pressure variability, is more closely

related to the function of the cardiovascular ANS than the neurological completeness of injury.5

Therefore, it may be more important to measure concomitantly the autonomic completeness of

injury and additional autonomic dysfunctions such as orthostatic hypotension and autonomic

dysreflexia to gain further insights. Also, provided that HRV represents the modulation of the

cardiac ANS, it may be a valuable tool to test the responsive of the cardiac ANS to different

testing conditions.45

The testing conditions will eliminate the problem of high HRV inter-subject

variability since the HRV comparisons will be made within-subjects. It may also be useful to

combine HRV assessments with other cardiac measurements to assess the risk of cardiac disease

in individuals with a SCI such as combining left ventricular ejection fractions with HRV

assessments in order to identify cardiac patients as suggested by Kleiger and colleages41

. The

International Standards to document remaining Autonomic Function after Spinal Cord Injury

(ISAFSCI)6 have been recently considered the gold standard for ANS assessment in SCI and

includes measures of the heart rate, blood pressure, sweating, temperature regulation and the

broncho-pulmonary system etc. and could add value to future studies using different testing

paradigms to assist with interpreting the results.

5.2 Study limitations

A major study limitation was that HRV was compared across cohorts and subjects that were

highly variable. Intra–subject comparisons using HRV may be a far better paradigm. Also,

contrary to the literature, no differences were observed in HRV when compared based on level

and severity of injury. The total sample size divided into four cohorts may have been too small to

detect significant differences. Post-hoc power analysis revealed that based on the mean per

Page 66: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

53

cohort of the primary HRV index, LF:HF, a total sample size of 72 would be required to detect a

difference.

Also the completeness of injury, as assessed by the International Standards for Neurological and

Functional Classification of Spinal Cord Injury indicates whether there is sensory or motor

function preserved in the lowest sacral segments (S4-S5).13

Therefore neurological completeness

which was used in this study, provided little information regarding the severity of autonomic

dysfunction after a SCI.

A final study limitation was that the breathing pattern was not monitored or recorded

simultaneously with the ECG data collection in the primary study so the data were not available

for analysis. Respiratory sinus arrhythmia is the natural variation of the heart rate and is driven

by the breathing pattern via vagal influence of the heart.31

The HF bandwidth has been linked to

the respiratory sinus arrhythmia and thus the breathing pattern during data collection may alter

the results of HRV.31-32,35

Nunan and colleages37

showed that the parasympathetic activity was

elevated when testing was done in a resting supine position along with paced breathing.

Billman68

also suggested that all subjects must engage in paced breathing to ensure precise

measurement of HRV. Consequently, depending on whether the subjects in this study engaged in

controlled or spontaneous breathing, the HF component may have disproportionally represented

the parasympathetic modulation of the heart.

Page 67: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

54

Chapter 6:

6 Conclusions

Traditional CVD risk factors, such as age, sex, obesity, and lifestyle, in addition to SCI-related

changes pose an increased risk of CVD development among individuals with a SCI. Autonomic

dysfunction, particularly of the cardiovascular ANS has been recently classified as a major CVD

contributor in SCI and thus requires further investigation. HRV analysis was examined since it

has been hypothesized to have the potential to non-invasively measure of cardiac autonomic

disruption and thus assess cardiac risk in individuals with a SCI. The findings illustrated that

there was an extremely wide range of HRV values in a chronic cross-sectional population thus

making it difficult to develop HRV reference values for this population of SCI. Nonetheless, the

inter-subject variability has also been observed in the able-bodied population which may indicate

that, likewise, individuals with a SCI also experience diverse cardiac ANS function.

Furthermore, despite the fact that cardiac autonomic dysfunction has been shown to be related to

the level and severity of injury, our results revealed no differences across the selected cohorts.

The disparity with the literature might have been due to a number of reasons: 1. Between subject

comparisons were made in spite of the fact that HRV has high inter-subject variability; 2. HRV

is not exclusively linked to level and severity of SCI, and 3. Cardiac autonomic function has

multiple biological complexities that cannot be measured exclusively by heart rate parameters.

Given that the LF:HF, is the most commonly used HRV measure of cardiac sympatho-vagal

balance and the physiological interpretation of the positive relationship between LF and HF

remains undetermined, HRV indices may not be applicable in individuals with a traumatic and

chronic SCI. Further understanding of the biological interpretation of the HRV indices is

required before routinely using HRV in SCI to monitor and/or manage CVD progression.

Page 68: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

55

References

1. Cragg JJ, Noonan VK, Krassioukov A, Borisoff J. Cardiovascular disease and spinal cord

injury results from a national population health survey. Neurology. 2013;81(8):723-8.

2. Claydon VE, Krassioukov AV. Clinical correlates of frequency analyses of

cardiovascular control after spinal cord injury. American Journal of Physiology - Heart &

Circulatory Physiology. 2008 Feb;294(2):H668-78. PubMed PMID: 18024546.

3. Myers J, Lee M, Kiratli J. Cardiovascular disease in spinal cord injury: An overview of

prevalence, risk, evaluation, and management. American Journal of Physical Medicine &

Rehabilitation. 2007 Feb;86(2):142-52. PubMed PMID: 17251696.

4. Sahota IS, Ravensbergen HJ, McGrath MS, Claydon VE. Cerebrovascular responses to

orthostatic stress after spinal cord injury. J Neurotrauma. 2012;29(15):2446-56.

5. West CR, Bellantoni A, Krassiokouv AV. Cardiovascular function in individuals with

incomplete spinal cord injury: A systematic review. Top Spinal Cord Inj Rehabil. 2013

Fall;19(4):267-78. PubMed PMID: 24244092. Pubmed Central PMCID: PMC3816721

Epub 2013/11/19.eng.

6. Krassioukov A, Biering-Sorensen F, Donovan W, Kennelly M, Kirshblum S, Krogh K, et

al. International standards to document remaining autonomic function after spinal cord

injury. J Spinal Cord Med. 2012 Jul;35(4):201-10. PubMed PMID: 22925746. Epub

2012/08/29. Eng

7. Task Force of the European Society of Cardiology and the North American Society of

Pacing and Electrophysiology. Heart rate variability: standards of measurement,

physiological interpretation and clinical use. Task Force of the European Society of

Cardiology and the North American Society of Pacing and Electrophysiology.

Circulation. 1996 Mar 1;93(5):1043-65. PubMed PMID: 8598068. Epub 1996/03/01. eng.

8. Farry A, Baxter D. The incidence and prevalence of spinal cord injury in Canada:

Overview and estimates based on current evidence. Rick Hansen Institute and Urban

Futures Institute; 2010 December.

9. Ones K, Yilmaz E, Beydogan A, Gultekin O, Caglar N. Comparison of functional results

in non-traumatic and traumatic spinal cord injury. Disability & Rehabilitation.

2007;29(15):1185-91. PubMed PMID: 17653992.

10. Chapman J. Comparing medical complications from nontraumatic and traumatic spinal

cord injury. Arch Phys Med Rehabil. 2000;81:1264.

11. Bican O, Minagar A, Pruitt AA. The spinal cord: a review of functional neuroanatomy.

Neurologic Clinics. 2013;31(1):1-18. PubMed PMID: 23186894.

12. Maynard FM, Jr., Bracken MB, Creasey G, Ditunno JF, Jr., Donovan WH, Ducker TB, et

al. International standards for neurological and functional classification of spinal cord

injury. Spinal Cord. 2007;35(5):266-74. PubMed PMID: 9160449.

13. Kirshblum SC, Biering-Sorensen F, Betz R, Burns S, Donovan W, Graves DE, et al.

International Standards for Neurological Classification of Spinal Cord Injury: cases with

Page 69: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

56

classification challenges. Journal of Spinal Cord Medicine.2014;37(2):120-7. PubMed

PMID: 24559416.

14. Kirshblum SC, Burns SP, Biering-Sorensen F, Donovan W, Graves DE, Jha A, et al.

International standards for neurological classification of spinal cord injury (revised

2011). J Spinal Cord Med. 2011 Nov;34(6):535-46. PubMed PMID: 22330108. Pubmed

Central PMCID: PMC3232636. Epub 2012/02/15. eng.

15. Rosado-Rivera D, Handrakis JP, Cirnigliaro CM, Jensen AM, Kirshblum S, Bauman

WA, et al. Comparison of 24-hour cardiovascular and autonomic function in paraplegia,

tetraplegia, and control groups: Implications for cardiovascular risk. Journal of Spinal

Cord Medicine. 2011;34(4):395-403. PubMed PMID: 21903013.

16. West C, Mills P, Krassioukov A. Influence of the neurological level of spinal cord injury

on cardiovascular outcomes in humans: A meta-analysis. Spinal Cord. 2012;50(7):484-92

17. Hagen E, Rekand T, GrÃnning M, Faerestrand S. Cardiovascular complications of spinal

cord injury. Tidsskrift for den Norske laegeforening: tidsskrift for praktisk medicin, ny

raekke. 2012;132(9):1115-20.

18. Nash MS, Cowan RE, Kressler J, Nash MS, Cowan RE, Kressler J. Evidence-based and

heuristic approaches for customization of care in cardiometabolic syndrome after spinal

cord injury. Journal of Spinal Cord Medicine. 2012 Sep;35(5):278-92. PubMed PMID:

23031165.

19. Wahman K, Nash MS, Lewis JE, Seiger A, Levi R. Cardiovascular disease risk and the

need for prevention after paraplegia determined by conventional multifactorial risk

models: The Stockholm spinal cord injury study. J Rehabil Med. 2011;43(3):237-42.

20. Rice BH, Cifelli CJ, Pikosky MA, Miller GD. Dairy components and risk factors for

cardiometabolic syndrome: Recent evidence and opportunities for future research.

Advances in Nutrition: An International Review Journal. 2011 Sep;2(5):396-407

21. Pearson TA, Blair SN, Daniels SR, Eckel RH, Fair JM, Fortmann SP, et al. AHA

guidelines for primary prevention of cardiovascular disease and stroke: 2002 update:

Consensus panel guide to comprehensive risk reduction for adult patients without

coronary or other atherosclerotic vascular diseases. Circulation. 2002 July;106(3):388-91.

22. Furlan JC, Fehlings MG. Cardiovascular complications after acute spinal cord injury:

Pathophysiology, diagnosis, and management. Neurosurgical Focus.2008;25(5):E13.

PubMed PMID: 18980473.

23. Miyatani M, Szeto M, Moore C, Oh PI, McGillivray CF, Craven BC. Exploring the

associations between arterial stiffness and spinal cord impairment: A cross-sectional

study. J Spinal Cord Med. 2014;37(5):556-64.

24. Xhyheri B, Manfrini O, Mazzolini M, Pizzi C, Bugiardini R. Heart rate variability today.

Prog Cardiovasc Dis. 2012 Nov-Dec;55(3):321-31. PubMed PMID: 23217437. Epub

2012/12/12. Eng.

25. Ravensbergen HJ, de Groot S, Post MW, Slootman HJ, van der Woude LH, Claydon VE.

Cardiovascular function after spinal cord injury: Prevalence and progression of

Page 70: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

57

dysfunction during inpatient rehabilitation and 5 years following discharge. Neurorehabil

Neural Repair. 2014 Mar;28(3):219-29. PubMed PMID: 24243916. Epub 2013/11/19.

eng.

26. Krassioukov A. Autonomic dysreflexia: current evidence related to unstable arterial

blood pressure control among athletes with spinal cord injury. Clinical Journal of Sport

Medicine. 2012;22(1):39-45.

27. Wecht JM, Bauman WA. Decentralized cardiovascular autonomic control and cognitive

deficits in persons with spinal cord injury. J Spinal Cord Med. 2013;36(2):74-81.

28. Alexander MS, Biering-Sorensen F, Bodner D, Brackett NL, Cardenas D, Charlifue S, et

al. International standards to document remaining autonomic function after spinal cord

injury. Spinal Cord. 2009;47(1):36-43. PubMed PMID: 18957962.

29. Wang YH, Huang TS, Lin JL, Hwang JJ, Chan HL, Lai JS, et al. Decreased autonomic

nervous system activity as assessed by heart rate variability in patients with chronic

tetraplegia. Arch Phys Med Rehabil. 2000 Sep;81(9):1181-4. PubMed PMID: 10987159.

Epub 2000/09/15. Eng.

30. Grimm DR, De Meersman RE, Almenoff PL, Spungen AM, Bauman WA.

Sympathovagal balance of the heart in subjects with spinal cord injury. Am J Physiol.

1997 Feb;272(2 Pt 2):H835-42. PubMed PMID: 9124446. Epub 1997/02/01. eng.

31. Freeman JV, Dewey FE, Hadley DM, Myers J, Froelicher VF. Autonomic nervous

system interaction with the cardiovascular system during exercise. Progress in

Cardiovascular Diseases. 2006 Mar-Apr;48(5):342-62. PubMed PMID: 16627049.

32. McMillan DE. Interpreting heart rate variability sleep/wake patterns in cardiac patients. J

Cardiovasc Nurs. 2002 Oct;17(1):69-81. PubMed PMID: 12358094. Epub 2002/10/03.

eng.

33. Montano N, Porta A, Cogliati C, Costantino G, Tobaldini E, Casali KR, et al. Heart rate

variability explored in the frequency domain: A tool to investigate the link between heart

and behavior. Neuroscience & Biobehavioral Reviews. 2009 Feb;33(2):71-80. PubMed

PMID: 18706440.

34. Stauss HM. Heart rate variability. American Journal of Physiology - Regulatory

Integrative & Comparative Physiology. 2003 Nov;285(5):R927-31. PubMed PMID:

14557228.

35. Bertsch K, Hagemann D, Naumann E, Schachinger H, Schulz A. Stability of heart rate

variability indices reflecting parasympathetic activity. Psychophysiology. 2012

May;49(5):672-82. PubMed PMID: 22335779.

36. Lauer MS. Autonomic function and prognosis. Cleveland Clinic Journal of Medicine.

2009 Apr;76 Suppl 2:S18-22. PubMed PMID: 19376976.

37. Nunan D, Sandercock GR, Brodie DA. A quantitative systematic review of normal values

for short-term heart rate variability in healthy adults. Pacing Clin Electrophysiol. 2010

Nov;33(11):1407-17. PubMed PMID: 20663071. Epub 2010/07/29. eng.

Page 71: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

58

38. Sandercock GRH, Bromley PD, Brodie DA. Effects of exercise on heart rate variability:

inferences from meta-analysis. Medicine & Science in Sports & Exercise. 2005

Mar;37(3):433-9. PubMed PMID: 15741842.

39. Schroeder EB, Whitsel EA, Evans GW, Prineas RJ, Chambless LE, Heiss G, et al.

Repeatability of heart rate variability measures. Journal of Electrocardiology. 2004

Jul;37(3):163-72. PubMed PMID: 15286929.

40. Thayer JF, Yamamoto SS, Brosschot JF. The relationship of autonomic imbalance, heart

rate variability and cardiovascular disease risk factors. Int J Cardiol. 2010 May

28;141(2):122-31. PubMed PMID: 19910061. Epub 2009/11/17. eng.

41. Kleiger RE, Stein PK, Bigger JT, Jr. Heart rate variability: Measurement and clinical

utility. Ann Noninvasive Electrocardiol. 2005 Jan;10(1):88-101. PubMed PMID:

15649244.

42. Bunten DC, Warner AL, Brunnemann SR, Segal JL. Heart rate variability is altered

following spinal cord injury. Clin Auton Res. 1998 Dec;8(6):329-34. PubMed PMID:

9869550. Epub 1998/12/30. eng.

43. Ditor DS, Macdonald MJ, Kamath MV, Bugaresti J, Adams M, McCartney N, et al. The

effects of body-weight supported treadmill training on cardiovascular regulation in

individuals with motor-complete SCI. Spinal Cord. 2005 Nov;43(11):664-73. PubMed

PMID: 15968298.

44. Malliani A. Heart rate variability: From bench to bedside. European journal of internal

medicine. 2005 Feb;16(1):12-20. PubMed PMID: 15733815. Epub 2005/03/01. Eng.

45. Perini R, Veicsteinas A, Perini R, Veicsteinas A. Heart rate variability and autonomic

activity at rest and during exercise in various physiological conditions. European Journal

of Applied Physiology. 2003 Oct;90(3-4):317-25. PubMed PMID: 13680241.

46. Vetrugno RM, Liguori RM, Cortelli PM, Montagna PM. Sympathetic skin response. Clin

Auton Res. 2003;13(4):256-70.

47. Ditor DS, Kamath MV, Macdonald MJ, Bugaresti J, McCartney N, Hicks AL.

Reproducibility of heart rate variability and blood pressure variability in individuals with

spinal cord injury. Clin Auton Res. 2005;15(6):387-93.

48. Castiglioni P, Di Rienzo M, Veicsteinas A, Parati G, Merati G. Mechanisms of blood

pressure and heart rate variability: An insight from low-level paraplegia. Am J Physiol

Regul Integr Comp Physiol. 2006 Apr;292(4):R1502-9. PubMed PMID: 17122332. Epub

2006/11/24. eng.

49. Claydon VE, Hol AT, Eng JJ, Krassioukov AV. Cardiovascular responses and

postexercise hypotension after arm cycling exercise in subjects with spinal cord injury.

Archives of Physical Medicine and Rehabilitation. 2006;87(8):1106-14.

50. Claydon VE, Krassioukov AV. Orthostatic hypotension and autonomic pathways after

spinal cord injury. J Neurotrauma. 2006;23(12):1713-25

51. Kaikkonen KM, Korpelainen RI, Tulppo MP, Kaikkonen HS, Vanhala ML, Kallio MA,

et al. Physical activity and aerobic fitness are positively associated with heart rate

Page 72: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

59

variability in obese adults. J Phys Act Health. 2014. Ahead of print. PubMed PMID:

24508687. eng.

52. Alrefaie Z. Brief assessment of supine heart rate variability in normal weight,

overweight, and obese females. Ann Noninvasive Electrocardiol. 2014;19(3):241-6.

53. Mehta RK. Impacts of obesity and stress on neuromuscular fatigue development and

associated heart rate variability. Int J Obes. 2014. Ahead of print.

54. Farah BQ, Prado WL, Tenorio TR, Ritti-Dias RM. Heart rate variability and its

relationship with central and general obesity in obese normotensive adolescents. Einstein.

2013;11(3):285-90. PubMed PMID: 24136753.

55. Melanson EL, Freedson PS. The effect of endurance training on resting heart rate

variability in sedentary adult males. European Journal of Applied Physiology. 2001

Sep;85(5):442-9. PubMed PMID: 893734373; 11606013.

56. Flevari A, Vagiakis E, Zakynthinos S. Heart rate variability is augmented in patients with

positional obstructive sleep apnea, but only supine LF/HF index correlates with its

severity. Sleep Breath. 2014. Ahead of print.

57. Chalmers J, Quintana DS, Abbott MJ-A, Kemp AH. Anxiety disorders are associated

with reduced heart rate variability: A meta-analysis. Frontiers in Psychiatry. 2014

July11;5.

58. Lee CL CW. The effects of cigarette smoking on aerobic and anaerobic capacity and

heart rate variability among female university students. International Journal of Women's

Health. 2013 Oct;5:667-79

59. Craven C, Verrier M, Balioussis C, Wolfe D, Hsieh J, Noonan V, et al. Rehabilitation

environmental scan atlas: Capturing capacity in Canadian SCI rehabilitation. Rick

Hansen Institute, Vancouver, BC. 2012.

60. Brennan M, Palaniswami M, Kamen P. Do existing measures of Poincare plot geometry

reflect nonlinear features of heart rate variability? Biomedical Engineering, IEEE

Transactions on. 2001;48(11):1342-7.

61. Young FL, Leicht AS, Young FLS, Leicht AS. Short-term stability of resting heart rate

variability: influence of position and gender. Applied Physiology, Nutrition, &

Metabolism Physiologie Appliquee, Nutrition et Metabolisme. 2011 Apr;36(2):210-8.

PubMed PMID: 21609282.

62. Portney LG, Watkins MP. Foundations of Clinical Research: Applications to Practice.

2nd ed. Upper Saddle River, NJ: Prentice-Hall Inc, 2000.

63. Diego JB, Pedrosa DF, Gava AL. Cardiac Arrhythmias - New Considerations Croatia:

InTech; 2012. Chapter 8, Neurohumoral Control of Heart Rate; p.167-192.

64. Simmons OL, Kressler J, Nash MS. Reference fitness values in the untrained spinal cord

injury population. Arch Phys Med Rehabil. 2014;95(12):2272-8.

Page 73: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

60

65. Buchholz AC, Martin Ginis KA, Bray SR, Craven BC, Hicks AL, Hayes KC, et al.

Greater daily leisure time physical activity is associated with lower chronic disease risk

in adults with spinal cord injury. Appl Physiol Nutr Metab. 2009;34(4):640-7.

66. Martin Ginis KA, Jetha A, Mack DE, Hetz S. Physical activity and subjective well-being

among people with spinal cord injury: A meta-analysis. Spinal Cord. 2010;48(1):65-72.

67. Routledge FS, Campbell TS, McFetridge-Durdle JA, Bacon SL. Improvements in heart

rate variability with exercise therapy. Canadian Journal of Cardiology. 2010 Jun-

Jul;26(6):303-12. PubMed PMID: 20548976.

68. Billman GE. The LF/HF ratio does not accurately measure cardiac sympatho-vagal

balance. Frontiers in Physiology. 2013;4(26):1-5.

69. Bradford-Hill A. The environment and disease: Association or causation? Proc R Soc

Med. 1965;58:295-300.

70. de la Cruz Torres B, Lopez Lopez C, Naranjo Orellana J. Analysis of heart rate

variability at rest and during aerobic exercise: a study in healthy people and cardiac

patients. Br J Sports Med. 2008 Sep;42(9):715-20. PubMed PMID: 18199627. Epub

2008/01/18. eng.

71. Grant CC, Viljoen M, van Rensburg DC, Wood PS. Heart rate variability assessment of

the effect of physical training on autonomic cardiac control. Ann Noninvasive

Electrocardiol. 2012 Jul;17(3):219-29. PubMed PMID: 22816541. Epub 2012/07/24. eng.

Page 74: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

61

Appendices

Page 75: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

62

Appendix A

Page 76: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

63

Page 77: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

64

Page 78: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

65

Appendix B

Page 79: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

66

Page 80: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

67

Appendix C

The table below summarizes the average percentage of noise, ectopic, and artifact for each five minute segment, and the possible reasons

for displaying ectopic islands. The Poincaré plot was examined after the subjects were selected and therefore a total sample of 56 subjects.

Table 1. Summary of the subjects who displayed ectopic islands in their Poincaré plots before the filter was applied

Subject

ID

Complete

(C)/Incomplete

(I) injury

Level of injury Normal (%) Ectopic (%) Artifact (%) Possible Reason

2

C T4

84.03 15.28 0.69

Unclear- Normal ECG &sinus

rhythm

5

I T10

88.66 7.32 4.02

R peak 0.001V (very small

compared to others)

11

I C4

88.59 11.05 0.36

Unstable baseline;

unifocal/premature ventricular

beat # 35, 16, 157, 163, 206, 522

12 I C6 90.94 6.43 2.63

Unclear- Normal ECG &sinus

rhythm

23 C C6 82.70 6.89 10.41 Unstable baseline

26 I T5 38.78 34.51 26.71 Unstable baseline

30 C T4 84.47 14.32 1.21

Unclear- Normal ECG &sinus

rhythm

Page 81: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

68

31 C C5 91.29 7.37 1.34

Unclear-Normal ECG &sinus

rhythm

34 I T12 51.94 48.06 0.33 Read two R peaks for each one

39 C T4 86.33 8.52 5.15

Unclear- Normal ECG &sinus

rhythm

45 I C6

66.72 1.32 31.96

Unclear-Normal ECG &sinus

rhythm

71 I C6

93.24 0.26 6.49

High T waves (at times larger

than R wave due to unstable

baseline)

72 I C4 95.29 3.30 1.41 Unstable baseline

88 I T12 90.66 9.17 0.51 Read two R peaks for each one

91 I T12 85.66 4.78 9.56 Read two R peaks for each one

Page 82: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

69

Appendix D

The following appendix examines the outliers present when the primary HRV index (LF:HF) was plotted. Table 1 describes the

characteristics of the subjects who were outliers when the data was examined for the entire sample and based on level and severity of

injury. Figure 1 is the frequency distribution after the outliers were omitted from the entire sample.

Table 1. The characteristics of the subjects who were outliers when LF:HF was assessed in the entire sample and based on level

and severity of injury, in comparison to the rest of the study sample

Shaded areas indicate that the value is outside the median range

Factors Study sample

(without the

outliers)

Subject 82 Subject 17 Subject 40 Subject

19

Subject 45 Subject 61

Type of SCI Level of Injury NA C6 T6 T7 C7 C6 C2

Severity of Injury NA Complete Complete Complete Complete Incomplete Incomplete

Time post injury

(years)

Mean:

13.78±9.74

Median: 11.00

(5.00, 21.25)

30 22 25 6 22 3

HRV

Frequency

domain

measures

LF (ms2) Mean:

900.76±1070.26

Median: 460.20

(202.78,

26.84 2053.79 499.02 453.74 211.51 1326.25

Page 83: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

70

1230.49)

HF (ms2) Mean:

917.25±1177.12

Median: 439.38

(177.14,

1173.75)

1.94 196.73 60.29 61.84 69.08 425.96

LF:HF Mean:

1.49±1.24

Median: 1.03

(0.57, 2.15)

13.81 10.44 8.28 7.34 3.06 3.11

Demographics Age (years) Mean:

46.04±12.64

Median: 44.00

(37.00, 54.25)

50 43 53 70 54 46

Sex

(male/female)

30/12 Male Male Male Male Male Male

Cardiovascular

status

BMI (kg/m2) Mean:

25.83±4.73

Median: 27.06

(21.97, 29.08)

22.48 31.24 27.77 26.32 38.34 25.63

Waist

Circumference

(cm)

Mean:

94.25±14.22

Median: 94.90

95.50 123.50 99.50 103.20 124.00 98.00

Page 84: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

71

(83.05, 104.70)

Absolute VO2

peak (L/min)

Mean:

1.29±0.61

Median: 1.14

(0.82, 1.59)

1.10 2.37 1.92 1.05 NT 1.09

Relative VO2

peak (ml/kg/min)

Mean:

16.71±7.62

Median: 16.29

(11.38, 22.34)

14.61 23.36 20.68 13.36 NT 13.45

Peak Heart rate

(bpm)

Mean:

130.42±29.08

Median: 128.00

(105.00,

155.00)

94 176 146 121 NT 107

Resting systolic

blood pressure

(mmHg)

Mean:

110.82±16.70

Median: 109.00

(98.00, 122.00)

88 132 100 90 108 88

Resting diastolic

blood pressure

(mmHg)

Mean:

71.02±13.37

Median: 70.00

(59.50, 80.50)

60 90 80 68 80 68

Page 85: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

72

Resting heart rate

(bpm)

Mean:

61.44±8.90

Median: 60.72

(55.51, 67.67)

71.83 55.82 67.71 58.78 77.19 50.48

Family history of

heart disease

(yes/no)

26/24 Yes Yes No Yes Yes Yes

Smoker (yes/no) 13/37 Yes No No No No No

Smoking history

(yes/no)

31/19 Yes No Yes No No No

Number of co-

morbidities

Mean:

1.74±1.24

Median: 2.00

(1.00, 2.25)

2 2 3 2 2 1

Functional

status

LTPAQ-SCI

(min/week)

Mean:

281.00±520.03

Median: 112.50

(0, 420.00)

120 240 30 380 NT 180

LEMS (/50) Mean:

10.84±16.55

Median: 0 (0,

22.75)

0 0 0 0 35 50

Page 86: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

73

SCIM-III (/100) Mean:

61.08±21.28

Median: 67.50

(50.00, 74.25)

33 66 70 25 60 91

General Sleep apnea

(yes/no)

11/39 No No Yes No No No

Height (cm) Mean:

171.73±23.96

Median: 175.63

(167.91,

181.92)

182.88 180.34 183 172.72 183 178

Weight (kg) Mean:

79.24±16.96

Median: 81.20

(65.77, 89.10)

75.2 101.6 93 78.5 128.4 81.2

NA: not applicable; NT: not tested due to subject’s choice

Page 87: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

74

Figure 1. Frequency distribution of the LF:HF plotted without the outliers for the entire sample.

LF:HF

Page 88: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

75

Appendix E

The following appendix includes all the frequency distributions and boxplots for the secondary HRV indices when examined in the entire

sample. The numbers in the boxplots correspond with the SPSS cell numbers and not the subject numbers.

Page 89: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

76

LF (ms2) LF (ms2)

HF (ms2) HF (ms2)

Page 90: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

77

pNN50 (%) pNN50 (%)

RMSSD (ms) RMSSD (ms)

Page 91: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

78

Appendix F

The following appendix includes the descriptive statistics of total power in the entire sample (Table 1)

and comparisons of total power based on level and/or severity of injury (Tables 2, 3 and 4).

Table 1. Descriptive statistics of total power in the entire sample (N=56)

Total Power

Mean ± SD 4233.51±5659.95

Median

(Lower, Upper quartile)

2192.24

(1229.12,5547.04)

IQR 4317.92

Skewness +3.41

Kurtosis 15.36

p-value† p<0.0001*

†Kolmogorov-Smirnov (K-S) test; p≤0.05*

Abbreviations: SD, standard deviation; IQR interquartile range

Table 2. Comparison of median total power based on level of injury

Below T5 (N=19) Above/Equal to

T5 (N=37)

p –Value†

Total Power

(ms2)

2121.44

(1264.10,6013.27)

2263.03

(1112.54,5177.26)

0.789†

†Mann-Whitney test; p≤0.05*

Page 92: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

79

Table 3. Comparison of median total power based on severity of injury

Complete (N=38) Incomplete

(N=18)

p –Value

Total Power

(ms2)

2514.90

(1216.39,5078.26)

2035.08

(1181.18,7889.27)

0.847†

†Mann-Whitney test;

p≤0.05*

Table 4. Comparison of median total power based on level and severity of injury

Complete and

equal to/above

T5 (N=27)

Complete and

below T5 (N=11)

Incomplete and

equal to/above T5

(N=10)

Incomplete and

below T5 (N=8)

p –

Value

Total

Power

(ms2)

2516.98

(922.97,

4228.42)

2512.82

(1429.34,

6013.27)

2105.88

(1181.18,

11434.80)

1741.22

(876.03,

6067.28)

0.805†

†Kruskal-Wallis test; p≤0.05*

Page 93: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

80

Appendix G

The following appendix includes all the boxplots representing the primary and secondary HRV indices based on level and severity of

injury. The numbers in the boxplots correspond with the SPSS cell numbers and not the subject numbers.

LF:H

F

Page 94: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

81

HF

(ms2

)

LF (

ms2 )

Page 95: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

82

pN

N5

0 (

%)

RM

SSD

(m

s)

Page 96: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

83

Appendix H

The following appendix includes a total of seven tables: six tables comparing the scalar and categorical

HRV-related factor based on level and/or severity on injury, and one table summarizing the multiple

comparisons of the significant factors; p≤0.05 indicates that there was significant difference across the

cohorts.

Table 1. Comparison of the HRV-related factors based on level of injury

Median (Lower, Upper

quartile) or Mean ± Standard

deviation

Below T5 Above/equal

to T5

p –Value

Time post

injury

10.00

(6.00,22.00)

16.00

(5.00,22.00)

0.510†

BMI 27.15

(22.71,29.05)

26.58

(21.99,29.46)

0.883†

LTPAQ-SCI 120.00

(30.00,420.00)

120.00

(0.00,405.00)

0.668†

LEMS 7.00

(27.00,0.00)

0.00

(0.00,16.00

0.111†

SCIM-III 71.00

(77.00,64.00)

65.00

(32.00,72.00)

0.036†*

Number of

Comorbidities

2.00

(0.00,2.00)

2.00

(1.00,2.50)

0.179†

Page 97: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

84

Age (years) 43.89 ±14.25 48.22 ±11.33

0.221§

Waist

circumference

(cm)

96.24 ±12.58 95.35 ±15.63 0.830

§

Relative VO2

peak

(ml/kg/min)

19.01 ±7.29 15.36 ±7.13 0.038

§*

Absolute

VO2 peak

(L/min)

1.56 ±0.75 1.15 ±0.44 0.088

§

Peak heart

rate (bpm)

141.37 ±25.14 123.45 ±29.64 0.033

§*

†Mann-Whitney test;

§ Independent t-test; p≤0.05*

Abbreviations: BMI, body mass index; LTPAQ-SCI, leisure time physical activity questionnaire-spinal

cord injury; LEMS, lower extremity motor score; SCIM-III, spinal cord independence measure

Table 2. Comparison of the HRV-related factors based on severity of injury

Median (Lower, Upper

quartile) or Mean ± Standard

deviation

Complete

Injury

Incomplete

Injury

p –Value

Time post

injury

17.50

(6.75,25.00)

6.00

(3.00,11.50)

0.001†***

BMI 26.84

(21.97,28.91)

27.06

(23.98,30.17)

0.352†

Page 98: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

85

LTPAQ-SCI 80.00

(0.00,382.50)

225.00

(40.00,471.25)

0.106†

LEMS 0.00

(0.00,0.00)

33.00

(20.75,41.25)

<0.0001†***

SCIM-III 65.00

(43.50,71.00)

77.00

(55.25,91.00)

0.003†**

Number of

Comorbidities

2.00

(1.00,2.00)

1.50

(0.75,3.00)

0.622†

Age (years) 45.76 ±11.85 48.83 ±13.71

0.393§

Waist

circumference

(cm)

95.16 ±14.64 96.68 ±14.75 0.720

§

Relative VO2

peak

(ml/kg/min)

16.20 ±6.58 17.81 ±8.74 0.469

§

Absolute

VO2 peak

(L/min)

1.23 ±0.50 1.45 ±0.76 0.232

§

Peak heart

rate (bpm)

131.39

±32.30

128.06 ±22.38 0.705

§

†Mann-Whitney test;

§ Independent t-test; p≤0.05*; p≤0.01**; p≤0.001***

Abbreviations: BMI, body mass index; LTPAQ-SCI, leisure time physical activity questionnaire-spinal

cord injury; LEMS, lower extremity motor score; SCIM-III, spinal cord independence measure

Page 99: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

86

Table 3. Comparison of HRV-related categorical characteristics based on level of injury

Below T5 Above/equal

to T5

p-value

Sex

(males/females)

14/5 30/7

p=0.516†

Current smoker

(no/yes)

13/6

29/8

p=0.518†

Smoking

history (no/yes)

7/12

16/21

p=0.645§

Family history

of heart disease

(no/yes)

8/11

17/20

p=0.784§

Sleep apnea

(no/yes)

14/5

30/7

p=0.516†

†Fisher exact test;

§Pearson chi-square; p≤0.05*

Table 4. Comparison of HRV-related categorical characteristics based on severity of injury

Complete

Injury

Incomplete

Injury

p-value

Sex

(males/females)

31/7

13/5

p=0.494†

Current

Smoker

(no/yes)

31/7

11/7

p=0.133†

Page 100: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

87

Smoking

history (no/yes)

15/23

8/10

p=0.724§

Family history

of heart disease

(no/yes)

19/19

6/12

p=0.241§

Sleep apnea

(no/yes)

32/6

12/6

p=0.171†

†Fisher exact test;

§Pearson chi-square; p≤0.05*

Table 5. Comparison of the HRV-related factors based on level and severity of injury

Median (Lower, Upper

quartile) or Mean ± Standard

deviation

Complete and

equal to/above

T5

Complete and

below T5

Incomplete

and equal

to/above T5

Incomplete

and below T5

p –Value

Time post

injury

17.00

(6.00,26.00)

19.00

(7.00,24.00)

6.00

(3.75,17.00)

6.00

(3.00,9.00)

0.016†*

BMI 26.31

(21.98,28.86)

27.15

(20.53,29.05)

26.84

(23.25,34.54)

27.12

(23.43,29.27)

0.798†

LTPAQ-SCI 170.00

(0.00,420.00)

30.00

(0.00,120.00)

75.00

(0.00,317.50)

420.00

(217.50,607.5

0)

0.029†*

LEMS 0.00

(0.00,0.00)

0.00

(0.00,0.00)

37.50

(23.00,42.50)

26.50

(15.25,39.25)

<0.0001†***

Page 101: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

88

SCIM-III 63.00

(25.00,70.00)

66.00

(61.00,71.00)

74.00

(39.00,81.75)

79.00

(66.75,95.25)

0.009†**

Number of

Comorbidities

2.00

(1.00,2.00)

2.00

(0.00,2.00)

1.50

(1.00,3.00)

1.00

(0.00,3.50)

0.583†

Age (years) 46.30 ±10.59 44.45 ±15.01 53.40 ±12.17 43.13 ±14.12 0.271§

Waist

circumference

(cm)

94.60 ±14.89 96.56 ±14.63 97.37 ±18.19 95.81 ±10.05 0.958

§

Relative VO2

peak

(ml/kg/min)

15.44 ±7.49 17.73 ±4.13 15.18 ±6.57 20.76 ±10.31 0.302

§

Absolute

VO2 peak

(L/min)

1.14 ±0.47 1.42 ±0.54 1.18 ±0.37 1.76 ±0.98 0.254

§

(Welch)

0.154§

(Brown-

Forsythe)

Peak heart

rate (bpm)

124.09

±34.04

146.00

±23.52

121.89

±15.78

135.00

±27.49 0.077

§

(Welch)

0.096§

(Brown-

Forsythe)

†Kruskal-Wallis test;

§One way ANOVA test; p≤0.05*; p≤0.01**; p≤0.001***

Abbreviations: BMI, body mass index; LTPAQ-SCI, leisure time physical activity questionnaire-spinal

cord injury; LEMS, lower extremity motor score; SCIM-III, spinal cord independence measure

Page 102: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

89

Table 6. Follow up analysis on the HRV-related factors that were different based on level and

severity of injury

Time Post Injury

Adjusted Sig.

LTPAQ-SCI

Adjusted Sig.

LEMS

Adjusted Sig.

SCIM-III

Adjusted Sig.

Complete and

equal to/above T5

vs. Complete and

below T5

1.000 1.000 1.000 1.000

Complete and

equal to/above T5

vs. Incomplete and

below T5

0.073 0.173 0.000*** 0.010**

Complete and

equal to/above T5

vs. Incomplete and

equal to/above T5

0.152 1.000 0.000*** 0.257

Complete and

below T5 vs.

Incomplete and

below T5

0.156 0.019* 0.008** 0.491

Complete and

below T5 vs.

Incomplete and

equal to/above T5

0.305 1.000 0.001*** 1.000

Incomplete and

below T5 vs.

1.000 0.203 1.00 1.000

Page 103: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

90

Incomplete and

equal to/above T5

†Kruskal-Wallis multiple comparison follow up test; p≤0.05*; p≤0.01**; p≤0.001***

Table 7. Comparison of HRV-related categorical subject characteristics based on level and

severity of injury

Complete and

equal

to/above T5

Complete and

below T5

Incomplete

and equal

to/above T5

Incomplete

and below T5

p-value

Sex

(males/females)

22/5

9/2

8/2

5/3

p=0.701†

Current smoker

(no/yes)

24/3

7/4

5/5

6/2

p=0.055†

Smoking

history (no/yes)

12/15

3/8

4/6

4/4

p=0.761†

Family history

of heart disease

(no/yes)

14/13

5/6

3/7

3/5

p=0.694†

Sleep apnea

(no/yes)

23/4

9/2

7/3

5/3

p=0.441†

†Fisher exact test;

§Pearson chi-square; p≤0.05*

Page 104: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

91

Appendix I

The following appendix includes all the scatter plot of the LF and HF based on level and/or severity of injury.

LF (

ms2

)

LF (

ms2 )

HF (ms2) HF (ms2)

Page 105: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

92

LF (

ms2 )

LF (

ms2 )

LF (

ms2

)

LF (

ms2 )

HF (ms2) HF (ms2)

HF (ms2) HF (ms2)

Page 106: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

93

LF (

ms2 )

LF (

ms2 )

HF (ms2) HF (ms2)

Page 107: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

94

Appendix J

The following appendix includes the scatter plots of the significant correlations between LF or HF and other selected HRV-related factors.

The factors include: age, time post injury and SCIM-III.

LF (

ms2 )

HF

(ms2 )

Page 108: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

95

LF (

ms2 )

HF

(ms2

)

Page 109: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

96

HF

(ms2 )

Page 110: Assessing Heart Rate Variability as a Surrogate Measure of ... · PDF fileii Assessing Heart Rate Variability as a Surrogate Measure of Cardiac Autonomic Function in Chronic Traumatic

97

Appendix K

The following appendix includes the multiple linear regression tables to predict LF (Table 1) or HF

(Table 2) in individuals with a complete injury equal to/above T5.

Table 1. Multiple linear regression analysis to predict LF for individuals with a complete injury

equal to/above T5 cohort (R2=0.261)

Parameter Regression

Coefficient

95% CI p-value

Intercept -663.03 -4198.33,1651.80 0.661

Age -36.25 -76.21,-12.67 0.116

WC 25.33 -8.07,72.51 0.290

Peak heart rate 4.38 -3.94,17.48 0.403

*p≤0.05

Abbreviations: CI, confidence interval; WC, waist circumference

Table 2. Multiple linear regression analysis to predict HF for individuals with a complete injury

equal to/above T5 cohort (R2=0.261)

Parameter Regression

Coefficient

95% CI p-value

Intercept 830.89 -1434.01,3226.77 0.527

Age -14.67 -38.45,13.15 0.279

WC 4.89 -15.55,22.29 0.617

Peak heart rate 2.48 -7.11,12.61 0.637

*p≤0.05

Abbreviations: CI, confidence interval; WC, waist circumference