The acute adaptations of normal and pathological human achilles tendons to eccentric and concentric exercise

KEY WORDS Achilles Tendon, Tendinopathy, Eccentric Exercise, Rehabilitation, Treatment, Diametral Strain, Morphology, Tendon Thickness, Ultrasound, Sonography, Echogenicity, Time-Depende

THE ACUTE ADAPTATIONS OF NORMAL AND PATHOLOGICAL HUMAN ACHILLES

TENDONS TO ECCENTRIC AND CONCENTRIC EXERCISE

Thesis submitted in fulfilment of the requirements for the award:

Doctor of Philosophy

2011

Nicole Lorraine Grigg BAppSc (Human Movement) BHlthSc (Hons) (Human Movement)

Institute of Health and Biomedical Innovation School of Human Movement Studies Queensland University of Technology

KEY WORDS

Achilles Tendon, Tendinopathy, Eccentric Exercise, Rehabilitation, Treatment, Diametral Strain, Morphology, Tendon Thickness, Ultrasound, Sonography, Echogenicity, Time-Dependent Conditioning, Walking, Electromyography, Ground Reaction Force, Power Spectral Density, Frequency, Motor Output Variability, Triceps Surae, Steadiness

ABSTRACT

Eccentric exercise is the conservative treatment of choice for mid-portion Achilles tendinopathy While there is a growing body of evidence supporting the medium to long term efficacy of eccentric exercise in Achilles tendinopathy treatment, very few studies have investigated the short term response of the tendon to eccentric exercise Moreover, the mechanisms through which tendinopathy symptom resolution occurs remain to be established The primary purpose of this thesis was to investigate the acute adaptations

of the Achilles tendon to, and the biomechanical characteristics of, the eccentric exercise protocol used for Achilles tendinopathy rehabilitation and a concentric equivalent The research was conducted with an orientation towards exploring potential mechanisms through which eccentric exercise may bring about a resolution of tendinopathy symptoms Specifically, the morphology of tendinopathic and normal Achilles tendons was monitored using high resolution sonography prior to and following eccentric and concentric exercise, to facilitate comparison between the treatment of choice and a similar alternative To date, the only proposed mechanism through which eccentric exercise is thought to result in symptom resolution is the increased variability in motor output force observed during eccentric exercise This thesis expanded upon prior work

by investigating the variability in motor output force recorded during eccentric and concentric exercises, when performed at two different knee joint angles, by limbs with and without symptomatic tendinopathy

The methodological phase of the research focused on establishing the reliability of measures of tendon thickness, tendon echogenicity, electromyography (EMG) of the Triceps Surae and the standard deviation (SD) and power spectral density (PSD) of the vertical ground reaction force (VGRF) These analyses facilitated comparison between the error in the measurements and experimental differences identified as statistically significant, so that the importance and meaning of the experimental differences could be established One potential limitation of monitoring the morphological response of the Achilles tendon to exercise loading is that the Achilles tendon is continually exposed to additional loading as participants complete the walking required to carry out their necessary daily tasks The specific purpose of the last experiment in the methodological

phase was to evaluate the effect of incidental walking activity on Achilles tendon morphology The results of this study indicated that walking activity could decrease Achilles tendon thickness (negative diametral strain) and that the decrease in thickness was dependent on both the amount of walking completed and the proximity of walking activity to the sonographic examination Thus, incidental walking activity was identified

as a potentially confounding factor for future experiments which endeavoured to monitor changes in tendon thickness with exercise loading

In the experimental phase of this thesis the thickness of Achilles tendons was monitored prior to and following isolated eccentric and concentric exercise The initial pilot study demonstrated that eccentric exercise resulted in a greater acute decrease in Achilles tendon thickness (greater diametral strain) compared to an equivalent concentric exercise, in participants with no history of Achilles tendon pain This experiment was then expanded to incorporate participants with unilateral Achilles tendinopathy The major finding of this experiment was that the acute decrease in Achilles tendon thickness observed following eccentric exercise was modified by the presence of tendinopathy, with a smaller decrease (less diametral strain) noted for tendinopathic compared to

healthy control tendon Based on in vitro evidence a decrease in tendon thickness is

believed to reflect extrusion of fluid from the tendon with loading This process would appear to be limited by the presence of pathology and is hypothesised to be a result of the changes in tendon structure associated with tendinopathy Load induced fluid movement may be important to the maintenance of tendon homeostasis and structure as

it has the potential to enhance molecular movement and stimulate tendon remodelling

On this basis eccentric exercise may be more beneficial to the tendon than concentric exercise Finally, EMG and motor output force variability (SD and PSD of VGRF) were investigated while participants with and without tendinopathy performed the eccentric and concentric exercises Although between condition differences were identified as statistically significant for a number of force variability parameters, the differences were not greater than the limits of agreement for repeated measures Consequently the meaning and importance of these findings were questioned Interestingly, the EMG amplitude of all three Triceps Surae muscles did not vary with knee joint angle during

the performance of eccentric exercise This raises questions pertaining to the functional importance of performing the eccentric exercise protocol at each of the two knee joint angles as it is currently prescribed EMG amplitude was significantly greater during concentric compared to eccentric muscle actions Differences in the muscle activation patterns may result in different stress distributions within the tendon and be related to the different diametral strain responses observed for eccentric and concentric muscle actions

TABLE OF CONTENTS

KEY WORDS II ABSTRACT III

TABLE OF CONTENTS 6

LIST OF FIGURES 12

LIST OF TABLES 15

LIST OF ABBREVIATIONS 16

LIST OF PUBLICATIONS 18

STATEMENT OF AUTHORSHIP 19

ACKNOWLEDGEMENTS 20

1.0 INTRODUCTION 22

2.0 LITERATURE REVIEW 26

2.1 Introduction 26

2.2 Achilles Tendon Anatomy 26

2.2.1 Musculature 26

2.2.2 Fibre Architecture 28

2.2.3 Vascularity 29

2.2.4 Summary 30

2.3 Tendon 31

2.3.1 Composition and Structure 31

2.3.2 Mechanical Properties 36

2.3.3 The Influence of Structure on Mechanical Properties 40

2.4 Load Induced Changes in Tendon Morphology and Interstitial Fluid Flow 41

2.4.1 In Vitro Evidence and Proposed Mechanisms 41

2.4.2 In Vivo Observations 46

2.5 Tendinopathy 48

2.5.1 Pathophysiology 49

2.5.2 Aetiology 52

2.6 Treatment for Tendinopathy 55

2.7 Evidence for the Efficacy of Eccentric Exercise: A Systematic Review 56

2.7.1 Introduction 56

2.7.2 Method 56

2.7.3 Results 58

2.7.4 Discussion 64

2.8 Characteristics of Eccentric and Concentric Exercise 68

2.8.1 Force Fluctuations 68

2.8.2 EMG: Within Muscle Characteristics 72

2.8.3 EMG: Between Muscle Characteristics 75

2.9 Summary 76

3.0 RESEARCH AIMS AND OBJECTIVES 78

3.1 Methodological Phase 79

3.2 Experimental Phase 80

4.0 THE RELIABILITY OF SONOGRAPHIC MEASUREMENT FOR MONITORING TENDON THICKNESS AND ECHOGENICITY 81

4.1 Abstract 81

4.2 Introduction 82

4.3 Method 83

4.3.1 Participants 83

4.3.2 Procedure 84

4.3.3 Sonographic Imaging 85

4.3.4 Sonographic Image Analysis 86

4.3.5 Data Analysis 86

4.4 Results 87

4.5 Discussion 92

4.6 Conclusions 94

5.0 ERRORS IN THE MEASUREMENT OF ELECTROMYOGRAPHIC AND GROUND REACTION FORCE PARAMETERS DURING TRICEPS SURAE EXERCISE 95

5.1 Abstract 95

5.2 Introduction 95

5.3 Method 96

5.3.1 Procedure 96

5.3.2 Instrumentation 97

5.3.3 Data Analysis 99

5.3.4 Statistical Analysis 101

5.4 Results 101

5.5 Discussion 102

5.6 Conclusion 103

6.0 INCIDENTAL WALKING ACTIVITY IS SUFFICIENT TO INDUCE TIME-DEPENDENT CONDITIONING OF THE ACHILLES TENDON 104

6.1 Abstract 104

6.2 Introduction 105

6.3 Methods 106

6.3.1 Participants 106

6.3.2 Procedure 107

6.3.3 Sonographic Imaging 107

6.3.4 Image Analysis 108

6.3.5 Activity Data Reduction 109

6.3.6 Statistical Analysis 110

6.4 Results 110

6.5 Discussion 112

6.6 Conclusion 114

7.0 ECCENTRIC CALF MUSCLE EXERCISE PRODUCES A GREATER ACUTE REDUCTION IN ACHILLES TENDON THICKNESS THAN CONCENTRIC EXERCISE 115

7.1 Abstract 115

7.2 Introduction 116

7.3 Methods 118

7.3.1 Participants 118

7.3.2 Procedure 118

7.3.3 Exercise Protocol 118

7.3.4 Sonographic Imaging 119

7.3.5 Image Analysis 119

7.3.6 Statistical Analysis 120

7.4 Results 120

7.5 Discussion 122

7.6 Conclusion 125

8.0 THE DIAMETRAL STRAIN RESPONSE ELICITED BY ECCENTRIC EXERCISE IS REDUCED IN TENDINOPATHY 126

8.1 Abstract 126

8.2 Introduction 127

8.3 Method 130

8.3.1 Participants 130

8.3.2 Procedure 131

8.3.3 Exercise Protocol 131

8.3.4 Biomechanics 132

8.3.5 Sonographic Imaging 133

8.3.6 Biomechanical Analysis 134

8.3.7 Sonographic Image Analysis 134

8.3.8 Statistical Analysis 136

8.4 Results 137

8.4.1 Baseline Tendon Characteristics 137

8.4.2 Biomechanics 138

8.4.3 Achilles Tendon Diametral Strain 141

8.4.4 Diametral Strain and Dorsiflexion Angle 145

8.4.5 Achilles Tendon Echogenicity 145

8.5 Discussion 145

8.5.1 Baseline Tendon Characteristics 145

8.5.2 Achilles Tendon Diametral Strain 146

8.5.3 Biomechanics of Exercise Performance 148

8.5.4 The Time Course of Achilles Tendon Diametral Strain 149

8.5.5 Achilles Tendon Echogenicity 150

8.5.6 Limitations 150

8.6 Conclusions 151

9.0 THE FREQUENCY POWER SPECTRA OF GROUND REACTION FORCE RECORDED

DURING ECCENTRIC EXERCISE IS ALTERED IN ACHILLES TENDINOPATHY 152

9.1 Abstract 152

9.2 Introduction 153

9.3 Method 155

9.3.1 Participants 155

9.3.2 Procedure 156

9.3.3 Exercise Protocol 156

9.3.4 Instrumentation 157

9.3.5 Data Analysis 159

9.3.6 Statistical Analysis 161

9.4 Results 162

9.4.1 Knee Joint Angle 162

9.4.2 SD of the VGRF 162

9.4.3 PSD of the VGRF 163

9.4.4 EMG 168

9.5 Discussion 170

9.5.1 Knee Joint Angle 170

9.5.2 Eccentric versus Concentric Exercise 172

9.5.3 Limb 174

9.5.4 Limitations 176

9.6 Conclusions 176

10.0 GENERAL DISCUSSION 178

10.1 Methodological Issues 179

10.2 Experimental Findings 181

10.2.1 Acute Morphological Adaptations to Eccentric and Concentric Exercise

181

10.2.2 Biomechanical Characteristics of Eccentric and Concentric Exercise 184

10.3 Limitations 186

10.4 General Conclusions 186

10.5 Future Research 187

11.0 APPENDICES 190 Appendix 1: EMG and VGRF Reliability Data 190 12.0 REFERENCES 192

LIST OF FIGURES

Figure 2.1: Tendon hierarchal structure, adapted and modified from (Ker, 2007; Wang,

2006) 34

Figure 2.2: Matrix to fibril stress transfer and distribution, adapted from Ker (1999). 37

Figure 2.3: Typical tendon stress strain curve, adapted from Wang, (2006). 39

Figure 4.1: Sonographic images of the Achilles tendons of control (left) and

symptomatic (right) limbs, with the measurement site (40 mm) identified by the magenta line, the grey scale profile plotted in yellow and the anterior and posterior tendon borders identified by the green lines 86

Figure 4.2: Bland and Altman plots illustrating the bias and 95% limits of agreement for

the sonographic measurements of sagittal Achilles tendon thickness (mm) at the 20 (left) and 40 (right) mm sites for control (), asymptomatic ( ) and symptomatic ( ) limbs 89

Figure 4.3: Bland and Altman plots illustrating the bias and 95% limits of agreement for

the mean tendon grey-scale (U) at the 20 (left) and 40 (right) mm sites for control ( ), asymptomatic ( ) and symptomatic ( ) limbs 91

Figure 5.1: Force plate block configuration. 98

Figure 5.2: Modified Helen Hayes marker system. 99

Figure 5.3: An example of the VGRF frequency power spectra for two exercise

repetitions in one participant 100

Figure 6.1: Sagittal diameter (d) of the Achilles tendon measured at a standard reference

point 2 cm from the superior aspect of the calcaneus (c) 108

Figure 6.2: Relationship between Achilles tendon diametral strain and total activity, for

each individual, normalised to a percentage of the maximum possible activity within each 0–12 (∆) and 12–24 ( ) hour period 111

Figure 6.3: Relationship between Achilles tendon diametral strain and total weighted

activity, for each individual, normalised to a percentage of the maximum possible activity within each 0–12 ( ) and 12–24 ( ) hour period 111

Figure 7.1: Mean Achilles tendon sagittal thickness normalised for pre-exercise values

following eccentric ( ) and concentric ( ) calf muscle exercise Error bars

represent 95% confidence intervals Exponential fit representative of the eccentric recovery time course (-) For clarity, data points have been shifted either side of the specified time points, at which they were observed 122

Figure 8.1: Sonographic images of the Achilles tendons of control (left) and

symptomatic (right) limbs, with the measurement site (40 mm) identified by the magenta line, the grey scale profile plotted in yellow and the anterior and posterior tendon borders identified by the green lines 135

Figure 8.2: Mean load duration (s) for control, asymptomatic and symptomatic limbs

during the performance of eccentric () and concentric ( ) exercise Error bars represent 95% confidence intervals *significantly different from control

limb (p < 05) 139

Figure 8.3: Mean maximum sagittal ankle joint angles for control, asymptomatic and

symptomatic limbs during the performance of concentric ( ) and eccentric ( ) loading exercises and mean minimum sagittal ankle joint angles for

control, asymptomatic and symptomatic limbs during the performance of concentric (∆) and eccentric () loading exercises Error bars represent 95%

confidence intervals *significantly different from control limb (p < 05) 141

Figure 8.4: Mean Achilles tendon diametral strain (%) at the 20 mm site, following

concentric (triangle) and eccentric (square) exercise for control (no shading), asymptomatic (solid grey) and symptomatic (solid black) limbs Error bars represent 95% confidence intervals 144

Figure 8.5: Mean Achilles tendon diametral strain (%) at the 40 mm site, prior to and

following concentric (triangle) and eccentric (square) exercise for control (no shading), asymptomatic (solid grey) and symptomatic (solid black) limbs Error bars represent 95% confidence intervals 144

Figure 9.1: Force plate block configuration. 158

Figure 9.2: Mean SD of the VGRF for control (), asymptomatic ( ) and symptomatic

( ) limbs during the performance of eccentric and concentric exercise

*significantly different from the eccentric condition (p < 05) 163

Figure 9.3: Mean power within the 9 (top) and 10 (bottom) Hz window for eccentric ( )

and concentric ( ) muscle actions across the six exercise sets Error bars represent 95% confidence intervals *significantly different from eccentric

exercise set one, two and three (p < 05) 165

Figure 9.4: Mean power within the 9 (top), 10 (middle) and 11 (bottom) Hz window for

control (), asymptomatic ( ) and symptomatic () limbs during the

performance of eccentric and concentric muscle actions Error bars represent

95% confidence intervals *significantly different from eccentric exercise (p <

.05) 167

Figure 9.5: Mean EMG amplitude (V) of LG (top), MG (middle) and SOL (bottom)

during concentric ( ) and eccentric () exercise across exercise sets 1 -6

Error bars represent 95% confidence intervals *significantly different from

concentric exercise set one, two and three (p < 05) 169

LIST OF TABLES

Table 2.1: Intervention study characteristics including mean ± SD (range) for participant

age and symptom duration 61

Table 2.2: Mean ± SD pain scores from the eight contrasts. 63

Table 4.1: Bias and 95% limits of agreement for sagittal thickness measurement of the

Achilles tendon expressed as diametral strain (%) 90

Table 5.1: 95% limits of agreement for the EMG amplitude of the MG, LG and SOL

muscles, the power (N2/Hz) within each frequency window and the standard deviation (N) of the VGRF 102

Table 7.1: Mean (standard deviation) sagittal thickness (mm) of Achilles tendons

allocated to eccentric or concentric exercise protocols 121

Table 8.1: Mean (SE) baseline (PRE) Achilles tendon thickness (mm) and echogenicity

(0-200 U) of control, asymptomatic and symptomatic limbs at the 20 and 40

mm measurement sites 138

Table 8.2: Mean (SE) maximum and minimum sagittal knee joint angles (degrees)

across exercise sets 140

Table 8.3: Mean (SE) tendon echogenicity (0-200 U) across all time points and muscle

actions for control, asymptomatic and symptomatic limbs at the 20mm and 40mm measurement sites 145

Table 9.1: Mean (SE) maximum and minimum sagittal knee joint angles (degrees) across

exercise sets 162

Table 9.2: Mean (SE) power spectral density (N2/Hz) of eccentric and concentric muscle

actions for the frequency windows 2-6 Hz 164

Table 11.1: Bias and 95% limits of agreement for EMG amplitude (V). 190

Table 11.2: Bias and 95% limits of agreement for the power (N2/Hz) within each

frequency window and for the standard deviation (SD) of the VGRF 191

LIST OF ABBREVIATIONS

1-RM One repetition Maximum

3-4SET Sonographic imaging conducted between the third and fourth exercise

sets 8hPOST Sonographic imaging conducted 8 hours after exercise completion

24hPOST Sonographic imaging conducted 24 hours after exercise completion

A_max Maximum sagittal ankle joint angle (maximum dorsiflexion)

A_min Minimum sagittal ankle joint angle (maximum plantarflexion)

ADC Apparent Diffusion Coefficient

AIC Akaike’s Information Criterion

ASIS Anterior Superior Iliac Spine

HP Hydroxylysylpyridinolone (enzymatic collagen cross link)

IMMPOST Sonographic imaging conducted immediately upon the completion of

exercise K_max Maximum sagittal knee joint angle

K_min Minimum sagittal knee joint angle

LG Lateral Gastrocnemius

LP Lysylpyridinoline (enzymatic collagen cross link)

MG Medial Gastrocnemius

MMP Matrix Metalloproteinase

MRI Magnetic resonance Imaging

MVC Maximum Voluntary Contraction

NRS Numerical Rating Scale

NSAIDs Non-steroidal Anti-inflammatory Drugs

PEG2 Prostaglandin E

PRE Sonographic imaging conducted prior to exercise

2

PSD Power Spectral Density

S\V Surface Area to Volume Ratio

SD Standard Deviation

SE Standard Error

SOL Soleus

TIME Duration of exercise repetition loading

VAS Visual Analogue Scale

VEGF Vascular Endothelial Growth Factor

VGRF Vertical Ground Reaction Force

VISA-A Victorian Institute of Sport Assessment Achilles

LIST OF PUBLICATIONS

Published Papers

Grigg, N L., Wearing, S C., & Smeathers, J E (2009) Eccentric calf muscle exercise

produces a greater acute reduction in Achilles tendon thickness than concentric

exercise British Journal of Sports Medicine, 43(4), 280-283

Grigg, N L., Stevenson, N J., Wearing, S C., & Smeathers, J E (2010) Incidental

walking activity is sufficient to induce time-dependent conditioning of the

Achilles tendon Gait & Posture, 31(1), 64-67

Grigg, N L., Smeathers, J E., & Wearing, S C (2010) Concurrent Validity of the Polar

s3 Stride Sensor for Measurement of Walking Stride Velocity Research Quarterly for Exercise and Sport, In Press

Conference Proceedings and Presentations

Grigg, N., Smeathers, J., & Grimshaw, P (2007) Unanticipated cutting in female netball

players: Valgus knee angle and the potential for non-contact ligament injury

Australian Conference of Science and Medicine in Sport Adelaide, Australia

Grigg, N L., Smeathers, J E., Wearing, S C., & Urry, S R (2008) Tendon

rehabilitation: isolated eccentric loading invokes a greater reduction in Achilles

tendon thickness than concentric exercise Australian Conference of Science

and Medicine in Sport Hamilton Island, Australia

Stevenson, N., Smeathers, J., Grigg, N., & Wearing, S (2008) Modeling activity

dependent diametral strain in Achilles tendon Australian Conference of Science

and Medicine in Sport Hamilton Island, Australia

Grigg, N L., Stevenson, N J., Smeathers, J E., & Wearing, S C (2009) Incidental

activity induces time-dependent conditioning of the Achilles tendon XXII

Congress of the International Society of Biomechanics Cape Town, South

Africa

STATEMENT OF AUTHORSHIP

The work contained in this thesis has not been previously submitted to meet requirements for an award at this or any other higher education institution To the best of

my knowledge and belief, the thesis contains no material previously published or written

by another person except where due reference is made

Signature:

Date:

ACKNOWLEDGEMENTS

First and foremost to my supervisors Dr James Smeathers and Associate Professor Scott Wearing, thank you, just doesn’t seem quite enough Thank you for sharing your knowledge in such an elegant and patient manner I couldn’t have asked for better supervisors The time and effort you have put into my work and the way in which you have fostered my learning, to facilitate my development as a scientist, is appreciated ever so much I have been privileged to work with two people who possess such intelligence and good nature To James, Cara calls you my intellectual friend, the impromptu and informal conversations we had on regular occasions, particularly in

2010, were informative and helpful I’m grateful for the time you have made for me To Scott, you have not only been an amazing academic mentor but a good friend As a trusted friend you really did help me through the toughest time of my short life to date and for that I will always be grateful

To all the people who provided technical support to the project: Diana Battistutta and Dimitrios Vagenas for their assistance with all matters statistical; Tim Gurnett for his assistance with the evolution of the MATLAB code used for image analysis and John O’Toole for helping with the MATLAB code used to quantify the power spectral density

of the vertical ground reaction force At this point I must mention the one and only Dr Nathan Stevenson Nathan developed the original MATLAB code for image analysis More recently, Nathan developed the code which facilitated segmentation of EMG, force and kinematic data into individual exercise repetitions and provided the code for the EMG analysis It was an absolute honour and privilege to work with Nathan, not only because of his wonderful MATLAB skills but his infectious personally I have such fond memories of the times during which we worked together

To Michael Cole, Cara Graepel, Jamie Sheard, Feng Qui, Amanda Rojek, Melissa Tucker, Manas Sivaramakrishnan and Tara Fernandez, you guys have been like a safety harness, you have kept me safe and allowed me to enjoy the experience, as I rode the rollercoaster which was PhD Thank you It has been a lot of fun

Finally I would like to thank all the kind people who participated in the research Their generous donation of time allowed this research to be completed, my personal goals to

be achieved and information added to the body of knowledge

1.0 INTRODUCTION

The Achilles tendon is the largest and strongest tendon in the human body, acting to transmit force produced by the Triceps Surae muscle group to the calcaneus, a transmission necessary for locomotory movement (O'Brien, 2005) Consequently the Achilles tendon is exposed to high tensile loads with extreme frequency The peak tensile force within the Achilles tendon during the stance phase of the walking gait cycle

is approximately two times bodyweight (1430 N) (Finni, Komi & Lukkariniemi, 1998) and Achilles tendon forces during running and jumping can be up to eight times bodyweight (3100-5330 N) (Magnusson & Kjaer, 2003) Although forces encountered

by the Achilles tendon are high, with respect to the strength of the tendon, the loads encountered usually equate to less than one third of the ultimate tensile stress (100 MPa) and induce strains of less than 4% (Wang, Iosifidis & Fu, 2006) While the original tendon length will be restored after exposure to strains of up to 4%, it is likely that even

at low strain levels, the fibrils and collagen cross-links will sustain routine damage The tendon must, therefore, continually undergo a process of remodelling to maintain structural homeostasis and mechanical integrity (Ker, 2007; Wang et al., 2006)

With continuous remodelling comes the potential for the process to fail and the tendon

to become degenerative Achilles tendon degeneration resulting in symptoms such as pain and a sensation of stiffness particularly at the onset of loading and clinical signs including local tenderness, swelling, reduced ankle dorsiflexion and increased tendon thickness and hypoechogenicity evident on ultrasound imaging, is commonly termed tendinopathy (Abate et al., 2009; Alfredson & Cook, 2007; Fredberg & Stengaard-Pedersen, 2008; Kingma, de Knikker, Wittink & Takken, 2007; Kjaer, 2004; Wang et al., 2006) Whilst the pathogenesis of tendinopathy is poorly understood, the condition is believed to arise from exposure to repetitive high magnitude loading without sufficient time for remodelling between loading sessions, and the interaction between loading and certain physiological and or environmental factors (Abate et al., 2009; Fredberg & Stengaard-Pedersen, 2008; Wang et al., 2006)

The lack of understanding as to the pathogenesis of tendinopathy has led to the development of an array of conservative treatment options (Alfredson & Cook, 2007; Rees, Wilson & Wolman, 2006; Wang et al., 2006) Eccentric exercise is one such treatment and is reported to be the treatment of choice for Achilles tendinopathy (Rees, Maffulli & Cook, 2009) In contrast to many other treatment options eccentric exercise has been the subject of many scientific investigations Such investigations have almost exclusively focused on the improvement of symptoms or changes in some physiological characteristics following the completion of a 12 week eccentric exercise program Overall, the available evidence suggests that eccentric exercise is beneficial in the treatment of tendinopathy, due to methodological limitations however, the level of evidence for the suggested efficacy remains relatively low (Kingma et al., 2007; Wasielewski & Kotsko, 2007; Woodley, Newsham-West & Baxter, 2007) Importantly, the mechanisms through which eccentric exercise may produce a reduction in tendinopathy symptoms have received little attention within the literature and still remain to be determined Understanding how eccentric exercise reduces the symptoms

of Achilles tendinopathy is of critical importance, as this understanding may enhance the effectiveness and efficiency of existing eccentric exercise protocols and lead to the development of new treatments

While the medium to long term effects of eccentric training on Achilles tendinopathy symptoms has been investigated extensively, the short term response of the tendon to eccentric exercise has received little attention within the literature Direct measurement

of how the tendon responds to eccentric exercise and contrasting this response with that

of a control condition may provide a more objective measure of the tendons response to eccentric exercise than a patients’ perception of symptoms Moreover, some insight into the underlying mechanisms for eccentric exercise efficacy may be gained by directly measuring the response of the tendon to eccentric exercise and a control condition An equivalent concentric exercise protocol provides the ideal contrast, as there is a small amount of evidence to suggest that concentric exercise is less effective in reducing tendinopathy symptoms than eccentric exercise (Mafi, Lorentzon & Alfredson, 2001; Niesen-Vertommen, Taunton, Clement & Mosher, 1992), and the two exercises would

be expected to expose the tendon to similar tensile loads In addition, characteristics of the exercises including muscle activation and motor output force variability maybe monitored, in order to identify characteristics unique to eccentric exercise

The aim of this thesis was, therefore, to examine the acute morphological response of normal and tendinopathic human Achilles tendons to both isolated eccentric and concentric Triceps Surae exercise, using sonographic imaging Furthermore, the biomechanical characteristics of these exercises were examined via surface electromyography (EMG) and monitoring fluctuations in the motor output force

The body of the thesis is divided into two sections; (1) the methodological section in which the errors associate with sonographic imaging, EMG, motor output force and incidental walking activity were quantified; (2) the experimental section in which the morphological response to, and characteristics of, eccentric and concentric exercise were quantified in participants with and without Achilles tendinopathy

The methodological section spans three chapters Chapter 4 investigates the observer reliability for the measurement of anterior-posterior Achilles tendon thickness from sagittal sonographic images, incorporating errors associated with image capture and digitisation Chapter 4 also investigates the reliability of image echogenicity quantification which is mostly dependent on the capture of the image Chapter 5 explores the error associated with quantification of EMG and force parameters over multiple exercise repetitions The two reliability chapters provide estimates of the level

intra-of error associated with each intra-of the measurement variables Thus, any differences observed across experimental conditions can be compared to the level of error, allowing the meaning of the difference to be identified independently from the statistical findings The third chapter in the methodological section (Chapter 6) examines the effect of a potential confounding factor, that of incidental walking activity, on the Achilles tendon diametral strain It was necessary to understand whether the loading associated with incidental walking was sufficient to modify the Achilles tendon diameter If this was the

case, incidental walking would be a source of error in future studies, and would need to

be controlled or accounted for

The experimental section of the thesis also consists of three chapters Chapter 7 reports the findings of a pilot experiment conducted to determine whether the morphological response of the Achilles tendon was dependent on exercise mode i.e isolated eccentric

or isolated concentric exercise The pilot experiments revealed that eccentric and concentric exercises did elicit different morphological responses and thus informed the design of subsequent experiments within the experimental section Chapter 8 examines the morphological response to eccentric and concentric exercise in participants with and without Achilles tendinopathy to determine whether pathological tendons responded in a similar manner to healthy tendons Chapter 9 considers the electromyographic and motor output force characteristics of the eccentric and concentric exercises in order to evaluate the potential mechanistic differences between the two exercises and whether these characteristics of exercise performance were dependent on the presence of tendinopathy

Findings from the individual studies are collectively discussed, in light of the research limitations and with reference to directions for future research, within Chapter 10

2.0 LITERATURE REVIEW

2.1 Introduction

This review commences with an overview of Achilles tendon anatomy followed by a synopsis of the general structure and mechanical properties of tendon finishing with a discussion on how the structure gives rise to the mechanical properties Tendon response

to tensile load is examined with specific reference to morphology and interstitial fluid movement Given the lack of published information pertaining to interstitial fluid

movement and the acute morphological response of tendon in vivo, information is obtained from in vitro investigations Current theories surrounding the pathology and

aetiology of tendinopathy are discussed Conservative treatment methods are listed and the evidence for the efficacy of eccentric exercise is systematically reviewed Finally, the force profile and electromyographic characteristics of eccentric and concentric exercise are detailed Where possible, the acute morphological response and tendinopathy have been discussed with specific reference to the Achilles tendon It must

be noted that this review focuses on the mid-portion of the Achilles tendon The effect of tensile load on or pathology of, the enthesis and myotendinous junction is beyond the scope of this review

2.2 Achilles Tendon Anatomy

The Achilles tendon is the largest and strongest tendon in the body It is a conjoined tendon formed from the integration of the three tendons arising from the Gastrocnemius, Soleus and Plantaris muscles (Gray, 2000; O'Brien, 2005) Collectively this group of muscles is known as the Triceps Surae The following section describes the anatomy of the muscles from which the Achilles tendon arises, the fibre architecture and the vascularity of the Achilles tendon

2.2.1 Musculature

The Gastrocnemius is a bi-articular, fusiform muscle, predominantly composed of fast twitch type II muscle fibres (O'Brien, 2005) As the Gastrocnemius is bi-articular it acts both to plantarflex the ankle and flex the knee (Gray, 2000; O'Brien, 2005) The

Gastrocnemius muscle consists of two heads, a medial and a lateral The medial head arises from the popliteal surface superior to the medial femoral condyle and posterior to the medial supracondylar line and adductor tubercle The lateral head arises from the posterior lateral surface of the lateral femoral condyle, superior and posterior to the lateral epicondyle and from a portion of the lateral lip of the linea aspera In addition both heads arise from inferior portions of the oblique popliteal ligament and knee joint capsule The Gastrocnemius heads are attached to the femoral condyles by strong flat tendons, each of which spread out to form aponeurosises, on the anterior surface of the muscle The medial head is the larger and extends more distally than the lateral head The muscle fibres of the medial and lateral Gastrocnemius unite in the mid calf region forming a tendinous raphe The tendinous raphe expands and broadens forming the aponeurosis located on the anterior aspect of the muscle Moving in a distal direction this aponeurosis gradually narrows before uniting with the aponeurosis of the soleus to form the Achilles tendon (Gray, 2000; O'Brien, 2005)

The Soleus is a broad, flat pennate muscle which is deep to, wider than and extends distally to a greater extent than the Gastrocnemius (O'Brien, 2005) The primary muscle action of the Soleus is to plantarflex the ankle joint The Soleus is a postural muscle mainly composed of slow twitch type I fibres The Soleus maintains posture and balance while standing, by contracting as the centre of gravity passes in front of the transverse axis of the knee joint, counteracting the tendency of the body to tilt forward at the ankle joint (O'Brien, 2005) The Soleus arises from the head and superior third of the posterior surface of the fibula, the fibrous arch between the fibula and tibia and the middle third of the medial tibial boarder The Soleus consists of two aponeurotic lamellae with the majority of multi-pennate muscle fibres located between the two aponeurotic lamellae Muscle fibres attach to the aponeurosis which covers the posterior surface of the muscle and gradually becomes thicker and narrower before joining the tendon of the Gastrocnemius to form the Achilles tendon (Gray, 2000; O'Brien, 2005)

The Plantaris muscle arises from the popliteal surface superior to the lateral femoral condyle and from the oblique popliteal ligament of the knee joint (Gray, 2000; O'Brien,

2005) The muscle belly is between five and ten centimetres long and the thin tendon extends distally crossing obliquely to the medial side between the Gastrocnemius and Soleus muscles The tendon inserts into the medial anterior boarder of the Achilles tendon (Gray, 2000; O'Brien, 2005)

2.2.2 Fibre Architecture

The proximal end of the Achilles tendon originates from the fusion of the thickened and narrowed portions of Gastrocnemius and Soleus aponeuroses (Gray, 2000; Jozsa & Kannus, 1997; O'Brien, 2005) The Achilles tendon is approximately 15 cm long, with the Soleus portion of the tendon ranging between 3 and 11 cm long while the Gastrocnemius portion of the tendon can range between 11 and 15 cm in length (O'Brien, 2005) The Achilles tendon is broad and flat at the origin becoming increasingly round and narrow forming an elliptical shape towards the calcaneal insertion Although the aponeurosis of the Gastrocnemius and Soleus muscles have fused proximally forming the Achilles tendon, the anterior aspect of tendon can continue

to receive soleus muscle fibres up to three centimetres from the insertion Approximately four centimetres above the insertion, the Achilles tendon flattens and expands (Gray, 2000; Jozsa & Kannus, 1997; O'Brien, 2005) Between 1.2 and 2.5 cm from the calcaneus, the tendon forms a wide delta shaped attachment In this region an area of fibrocartilage exists on the anterior aspect between the tendon and the calcaneus (O'Brien, 2005) The delta shaped fibrocartilagous attachment inserts on to the mid posterior region of the calcaneus (Gray, 2000; Jozsa & Kannus, 1997; O'Brien, 2005) At this fibrocartilaginous insertion the calcaneus has no true periosteum, however superficial fibres of the paratenon become continuous with the fibrous tissue of the calcaneus and pass from the lower boarder of the calcaneus to join the plantar fascia (O'Brien, 2005)

The Achilles tendon does not have a true synovial sheath, rather a peritendinous sheath commonly referred to as the paratenon (Jozsa & Kannus, 1997) Proximally the paratenon is continuous with the fascia of the muscle and distally the paratenon blends with the fibrous outer tissue of the calcaneus (Jozsa & Kannus, 1997; O'Brien, 2005)

The contribution of tendon fibres from each muscle and the orientation of the tendon fibres can vary between individuals (Jozsa & Kannus, 1997; O'Brien, 2005) In general, three classifications of fibre contribution have been identified The most common, apparent in approximately 50% of individuals, involves two thirds of the tendon fibres originating from the Soleus and one third from the Gastrocnemius In approximately 35% of individuals, half of the tendon fibres originate from the Gastrocnemius and half from the Soleus In approximately 15% of individuals, two thirds of the tendon fibres originate from the Gastrocnemius, with the remaining third from the Soleus The tendon fibrils spiral rotating through 90° as they distend distally Rotation occurs such that fibres that were positioned on the medial aspect of the tendon pass across the superficial surface of the tendon towards the lateral side and fibres originally on the lateral aspect of the tendon move across the deep side of the tendon towards the medial side The degree

to which tendon fibres rotate is determined by the level of fusion between the Gastrocnemius and Soleus tendons The more distal the tendons fuse, the less tendon fibre rotation that occurs This spiralled fibre arrangement is thought to create an area of stress concentration and most commonly occurs between two and five centimetres from the calcaneal insertion (Jozsa & Kannus, 1997; O'Brien, 2005)

The twisted fascicles originating from the medial Gastrocnemius form the posterior layer

of the Achilles tendon and are further divided into medial and lateral portions (Szaro, Witkowski, Smigielski, Krajewski & Ciszek, 2009) The anterior layer of the tendon is composed of fascicles originating from the lateral Gastrocnemius and Soleus, with fascicles from the lateral Gastrocnemius constituting the lateral aspect and fascicles originating from the Soleus making up the central and medial portion According to Szaro et al (2009) the fascicles originating from the three muscles can be easily separated and therefore, are able to work somewhat independently

2.2.3 Vascularity

The Achilles tendon receives blood supply through the myotendinous junction, the calcaneal insertion and along the length through the paratenon (Jozsa & Kannus, 1997; O'Brien, 2005) Due to the length of the tendon, the majority of blood supply is sourced

through the paratenon (Jozsa & Kannus, 1997; O'Brien, 2005) The paratenon on the anterior aspect of the tendon supplies the majority of blood vessels to the tendon Larger blood vessels in the paratenon run transversely towards the tendon At the surface of the tendon, the larger blood vessels branch several times into smaller vessels and then run longitudinally, parallel to the long axis of the tendon (Chen et al., 2009; Jozsa & Kannus, 1997; O'Brien, 2005) The small vessels of the paratenon enter the tendon and run along the endotenon Arterioles, flanked by two venules, run longitudinally and form

a uniform mesh like vascular system along the length of the tendon as capillaries loop from arterioles to venules on the surface of the fascicles (Jozsa & Kannus, 1997; O'Brien, 2005) The blood supply for the distal tendon (up to 4 cm from the insertion) and the proximal tendon (7 cm from the insertion up to the aponeurosis) is derived from

a recurrent branch of the posterior tibial artery, while the blood supply for the portion of the tendon (4 to 7 cm proximal to the insertion) is derived from the peroneal artery (Chen et al., 2009) The Achilles tendon demonstrates three different zones of vascularity When cross sections of the Achilles tendon were assessed, the number of vessels per square centimetre in the distal portion of the tendon was reported to be 56.6 vessels/cm2, in the mid portion 28.2 vessels/cm2 and in the proximal portion 73.4 vessels/cm2 (O'Brien, 2005; Zantop, Tillmann & Petersen, 2003) The numbers of vessels per square centimetre and angiographic analysis have demonstrated that the mid-portion of the tendon which is supplied by the peroneal artery is characterised by a zone

mid-of decreased vascularity in comparison to the proximal and distal tendon regions (Chen

et al., 2009; Jozsa & Kannus, 1997; O'Brien, 2005; Zantop et al., 2003)

Located on the posterior aspect of the lower leg the Triceps Surae muscle group is composed of the Gastrocnemius, Soleus and Plantaris muscles The Gastrocnemius is characterised by two heads, medial and lateral, both of which are a bi-articular acting to plantarflex the foot and flex the knee The Soleus is located deep to the Gastrocnemius and is mono-articular acting to plantarflex the foot The Plantaris is a small muscle located between the Gastrocnemius and Soleus The Achilles tendon is formed from the aggregation of fascicles originating from the medial and lateral heads of the

Gastrocnemius, the Soleus and the Plantaris muscles The lateral rotation of fibres results in an area of concentrated stress within the mid-portion of the tendon The mid-portion of the tendon is further characterised by decreased vascularity with respect to tendon regions more proximal to the muscle aponeurosis or the calcaneal insertion

2.3 Tendon

2.3.1 Composition and Structure

Tendon is attached at one end to a muscle and at the other to bone and transmits the force generated by muscle to the bony skeleton facilitating joint movement The mechanical behaviour of tendon tissue will determine the efficiency by which muscle force it transmitted to bone (Culav, Clark & Merrilees, 1999; Ker, 2007; Magnusson, Hansen & Kjaer, 2003; Wang, 2006) This efficiency will of course, be a function of the tendon structure and the ability of this structure to adapt to loading and resist failure The following is a description of tendon structure, mechanical properties and their relationship, with the aim of identifying structural properties associated with adaptation

to loading

Tendon is composed of five primary components water, collagen, elastin, proteoglycans and cells The proportion of each component can vary and this variability is reflected by the reported tendon constituents rarely adding up to 100% (Ker, 2002) In general tendon

is composed of approximately 70% water Of the dry mass collagen constitutes approximately 60-85%, elastin two percent, cells two percent and proteoglycans less than one percent Of the collagen, 95% is collagen type I and of the remainder, collagen type III and V are most prevalent (Culav et al., 1999; Kjaer, 2004; Magnusson et al., 2003; Wang, 2006)

Collagen is produced at the cellular level via the formation of three chains of amino

acids Collagen type I is composed of two 1 and one 2 chains which are products of

separate genes (Kjaer, 2004) Procollagen is principally produced by the ribosomes of the endoplasmic reticulum within the tenocytes Translation of the procollagen mRNA occurs in the ribosome and procollagen is assembled in the endoplasmic reticulum

Following translation, the pro chains undergo post translational reactions First

hydroxylation converts residual peptides to 4-hydroxyproline or 3-hydroxyproline, the

collagen peptides are glycoslated and the formation of intra and inter chain disulfide bonds occurs Each procollagen chain contains C-propeptide regions which fold and through the process of trimerization where chains combine to form a single triple

helical procollagen molecule (Kjaer, 2004) The chains are composed of a repeating

sequence of three amino acids Glycine-X-Y Glycine is the smallest amino acid, occupies the central core of the triple helix and is critical for the correct folding of the three chains into the helical configuration (Culav et al., 1999) The helical structure of the collagen molecule inherently resists tension Additionally, the 4-hydroxyproline residues which appear frequently at positions X and Y, promote the formation of intermolecular cross links upon which, the stability and quality of the collagen molecule

is largely based (Kjaer, 2004)

Procollagen is secreted from the cell into the extra cellular matrix (ECM) through the

golgi apparatus and the rate of secretion is dependent upon the rate of chain folding

into the triple helix The soluble procollagen molecule contains non-helical NH2 and COOH terminal peptides at respective ends of the molecule Following secretion into the ECM procollagen is converted into tropocollagen via the cleavage of these terminal peptides by aminoprotease and carboxyprotease enzymes respectively (Culav et al., 1999; Kjaer, 2004)

The tropocollagen molecules which are approximately 300 nm in length and 1.5 nm wide then self assemble into fibrils via end to end fusion This fusion is the result of the polarity of the ends of the tropocollagen molecules (Kjaer, 2004; Magnusson et al., 2003) The end to end fusion of collagen molecules forms fibril segments which are prerequisites for fibril formation and gradually increase in length and diameter Increased decorin formation occurs subsequent to end to end fusion, forming cross-links between collagen molecules and increasing the diameter of fibrils (Kjaer, 2004) In addition, collagen molecules are cross-linked with the aid of the lysyl oxidase enzyme, which forms lysylpyridinoline (LP) and hydroxylysylpyridinolone (HP) links between

fibrils at the sites of the 4-hydroxyproline protein residues (Wang, 2006) Fibrils are the smallest structural unit, are arranged in a quarter stagger array and appear to have a consistent aspect ratio despite fibril diameter varying between 10 and 500 nm The collagen fibrils are organised in a hierarchal manner where fibrils combine to form a fibre, which is surrounded by endotenon, fibres group together to form fascicles, also enveloped by endotenon and faicles form the tendon which is encapsulated by the epitenon (Figure 2.1) Tropocollagen molecules, fibrils, fibres and fascicles are arranged

in parallel and run longitudinally with the long axis of the tendon (Magnusson et al., 2003; Wang, 2006)

The endotenon is a thin network of connective tissue inside the tendon which binds tendon fibres together and into fascicles The endotenon facilitates movement of fibre bundles and distribution of blood vessels, nerves and lymphatics deep into the tendon (Jozsa & Kannus, 1997; Magnusson et al., 2003; O'Brien, 2005; Wang, 2006) The entire tendon is surrounded by a fine connective tissue sheath; the epitenon On the deep surface the epitenon is continuous with the endotenon and on the superficial surface the epitenon is continuous with the paratenon (Jozsa & Kannus, 1997; O'Brien, 2005) The paratenon is composed of loose areolar connective tissue This connective tissue is rich

in mucopolysaccharides, demonstrates a well defined collagenous fibre system, and contains elastic fibres and synovial cells which line the inner surface of the paratenon The paratenon acts as an elastic sleeve and functions to reduce friction permitting free movement of the tendon against surrounding tissue The paratenon and epitenon is referred to collectively as the peritendon (Jozsa & Kannus, 1997; O'Brien, 2005) (Figure 2.1)

Figure 2.1: Tendon hierarchal structure, adapted and modified from (Ker, 2007; Wang,

2006)

Although proteoglycans account for a very small proportion of tendon dry mass their concentration is related to tendon water content and collagen stability There are two groups of proteoglycans the small and the large (Yoon & Halper, 2005) Proteoglycans are composed of a small protein core to which one or more glycosaminoglycan (GAG) chains are covalently bonded The protein core of the small leucine rich proteoglycans is characterised by repeated sequences of 20-30 amino acids with a high concentration of leucine Small proteoglycans are so named, as the molecule is relatively small, due to the attachment of only one to two chondroitin or dermatam sulphate or several keratine sulphate GAG chains Large proteoglycans are rich in chondroitin sulphate and keratine sulphate GAG chains (Yoon & Halper, 2005) There are six major GAG chains all of which are composed of repeating disaccharide units GAG chains are negatively charged, have the tendency to attract ions, creating an osmotic imbalance and causing GAGs to attracting water from surrounding areas (Culav et al., 1999) Thus, large proteoglycans are negatively charged, hydrophilic and usually located between collagen fibrils and fibres Due to the high fixed charge density and charge to charge repulsive forces large proteoglycans are stiff and provide collagen fibrils with the capacity to resist transverse, between fibre, compressive force resulting from tensile load In addition, they facilitate diffusion of water soluble molecules, play a role in cell migration and the

binding of GAG chains to the positive regions on collagen molecules adds to mechanical stability (Culav et al., 1999; Rees, Dent & Caterson, 2009; Yoon & Halper, 2005) Aggrecan is the most common large proteoglycan in tendon, possessing up to 160 GAG side chains and it binds to hyaluronan via a glycoprotein Hyaluronan is a long GAG chain which is not attached to a protein core (Culav et al., 1999) This complex produces osmotic pressure, tendon hydration and a swelling pressure on the collagen matrix ideal for resisting compression For this reason aggrecan and large proteoglycans are found in higher concentrations within the fibrocartilaginous enthesis where aggrecan is produced

in response to compression (Kjaer, 2004; Yoon & Halper, 2005) Elsewhere within regions of tension aggrecan contributes to viscoelasticity, by reducing the viscosity of the tendon and acts as a lubricant allowing fibrils to slide over each other and stretch to dissipate sudden loading (Rees, Dent et al., 2009; Vogel, 2004; Yoon & Halper, 2005)

In regions of tension such as the mid-portion of the Achilles tendon, 90% of proteoglycans are small (Yoon & Halper, 2005) Decorin is the most abundant proteoglycan and is a small leucine rich proteoglycan with one dermatan sulphate GAG chain attached (Vogel, 2004) Decorin is produced by fibroblasts, chondrocytes and endothelial cells in response to tensile load and binds to all collagen types through the protein core (Kjaer, 2004; Yoon & Halper, 2005) Decorin enables tendon remodelling

in response to tensile force by facilitating the formation of fibrils from tropocollagen molecules (Yoon & Halper, 2005) Decorin functions to link neighbouring collagen molecules to increase fibril diameter The decorin protein core attaches to the collagen molecule and the GAG chain extends orthogonally Collagen molecules are linked via the binding of two decorin GAG chains originating from adjacent collagen molecules (Culav et al., 1999; Rees, Dent et al., 2009; Screen, Shelton, Chhaya et al., 2005) Each collagen molecule has multiple decorin binding sites suggesting that decorin may enhance structural integrity by assisting in load transfer between collagen molecules (Screen, Shelton, Chhaya et al., 2005) This hypothesis is, however, controversial A recent study designed to determine the mechanical role of GAG chains, found the removal of 93% of GAG chains had no influence on the elastic modulus or energy loss

of rat tail tendon fascicles (Fessel & Snedeker, 2009) Fibromodulin, another small

leucine rich proteoglycan is also known to function as a modulator of fibrillogenesis Fibromodulin is expressed in high levels in tendon, binds to type I collagen at a different site to decorin and promotes the formation of large mature fibrils Small leucine rich proteoglycans may also regulate cell proliferation and activity as they are known to bind

to transforming growth factor (TGF ) (Rees, Dent et al., 2009; Yoon & Halper, 2005)

Elastin fibres are composed of an elastin core and microfibrils located around the periphery Elastin also contains two amino acids which form cross-links between adjacent elastin chains and along with the hydrophobic properties of the protein are considered to provide the ability to undergo extension and recoil Although, the exact mechanism of extensibility and recoil is not understood (Culav et al., 1999; Magnusson

et al., 2003), elastin is important in the recovery of tendon length and crimp configuration following tensile load (Wang, 2006)

The dominant tendon cell type is the fibroblast, although fibroblasts found within tendon are commonly referred to as tenocytes Tenocytes function to maintain and remodel the ECM via synthesis of components such as collagen and proteoglycans in response to mechanical stimuli (Kjaer, 2004; Magnusson et al., 2003; Wang, 2006) Tenocytes align

in rows in the direction of crimp and attach predominantly to a single collagen fibre Tenocytes possess numerous long cell processes which communicate not only with cells

on the same fibre but with rows of cells on adjacent fibres Tenocytes are linked by gap junctions located at the ends of the cell processes and on the cell body, which provide a direct pathway for cell to cell communication (Kjaer, 2004; Screen, Shelton, Chhaya et al., 2005) This arrangement of tenocytes and associated cell processes ensures that the cells do not contribute to the mechanical properties of the tendon although the proportion of tendon strain experienced by the tenocytes is unknown (Ker, 2007; Screen, Shelton, Chhaya et al., 2005)

2.3.2 Mechanical Properties

Mechanically, tendon can be considered to be a composite material with longitudinally arranged parallel fibrils embedded in a matrix (Ker, 1999, 2007) Tendon fibrils self

assemble in the matrix and although, fibril length is unknown, it is likely that fibrils do not directly attach to the points of load application at either end of the tendon Therefore,

at some point load must be transferred between fibrils via the matrix When the tendon is loaded, shear stress is created between the tendon matrix and the surface of neighbouring collagen fibrils This shear stress is transferred to the tendon fibrils resulting in fibril tensile stress Despite unknown fibril length, fibrils are known to possess a high aspect ratio Shear stress is transferred to the central parts of the fibril via the matrix This transfer is possible even though the matrix has a modulus several orders

of magnitude lower than that of the fibrils, as the fibril surface area is much greater than the cross sectional area Also the ends of the collagen fibril are tapered so the tip of the fibril will experience zero tensile stress, while the matrix at this point possesses maximum stress Stress within the fibril builds up over the transfer length, while the stress in the matrix declines (Figure 2.2) In addition, matrix shear stress is likely transferred to the fibrils via GAGs and other cross-links between fibrils (Ker, 1999, 2007)

Figure 2.2: Matrix to fibril stress transfer and distribution, adapted from Ker (1999)

Tendon displays a non-linear viscoelastic response to dynamic loading, which is thought

to largely reflect tendon structure At stresses below 20 MPa, the tendon stress-strain curve is characterised by an initial non-linear or toe region (Figure 2.3) (Ker, 2007; Wang, 2006) This initial non-linear region, which accounts for strains in the order of two percent, is thought to represent straightening of the normally wavy configuration of

tendon fibres, termed crimp This is supported by research which visualised crimp straightening as rat Gastrocnemius tendon was stretched via progressively increasing the joint angle The number of visible crimps reduce significantly by 47% along with a reduction in crimp angle (Franchi et al., 2007) The wave length of tendon crimp is typically in the order of 0.01 mm but can vary The angle and length of the crimp will affect the tendon mechanical properties, within the toe region of the stress strain curve,

as tendons with smaller crimp angles will experience greater strain at lower loads compared to those with a larger crimp angle (Ker, 2007; Wang, 2006)

From approximately 15 to 25 MPa the fibrils are effectively straight and the subsequent region of the stress strain curve is essentially linear (Ker, 2007; Wang, 2006) Extending

up to four percent strain, this linear region of the curve represents the physiological or elastic range of tendon, in which fibre extension and between fibre sliding occurs (Screen, Shelton, Chhaya et al., 2005) Within this region, the tendon behaves elastically and will return to the original length upon removal of load, with approximately 90 percent of strain energy returned (Magnusson et al., 2003; Wang, 2006) If the tendon is loaded above four percent strain, microscopic disruption of the fibre structure occurs resulting in plastic deformation, which is permanent and energy is unrecoverable (Magnusson et al., 2003) In general, adult tendons demonstrate an elastic modulus of approximately 1500 MPa, a hysteresis of 10% and an ultimate tensile stress of 100 MPa (Ker, 2007)

Figure 2.3: Typical tendon stress strain curve, adapted from Wang, (2006)

The variability of tendon mechanical properties in vivo can be great even for tendons of

the same anatomical site and species (Ker, 1999, 2002, 2007) Despite this variability a broad view of mammalian tendons cannot identify any consistent variation in the elastic modulus, hysteresis or strength between tendons of animals with different masses or for tendons which function as springs (Ker, 2002) This finding is unusual as tendons are

exposed to different levels of stress in vivo depending on position in the body and

function Tendons which do not act as springs in general experience a maximum stress

in life in the region of 13 MPa while the maximum stress of the human Achilles tendon

is approximately 67 MPa Taking into account that the lower limit of ultimate tensile strength is 100 MPa, it appears that the majority of tendons are free from the risk of rupture, as a result of a single action, unless the tendon is pre-damaged (Ker, 2002)

It is generally thought that it is the fatigue quality (time to rupture under given loading conditions) of the tendon which correlates with the stress the tendon experiences in life Tendons exposed to higher stresses in life generally demonstrate a longer time to failure

at a given load, however, if tendons are stressed at their specific in life stress levels then the time to failure of all tendons is approximately equal (Ker, 2002) As stress in life is accumulated over a much longer period of time, than the time to failure, the matrix must undergo adaptation to maintain homeostasis and prevent rupture Thus, tendons exposed

to high and ongoing stress resulting in damage, undergo adaptation to maintain homeostasis and develop improved fatigue quality While tendons exposed to low levels

of stress are less fatigue resistant as the body only responds, in terms of adaptation, to routine damage (Ker, 2002, 2007) As the human Achilles tendon is exposed to high and repetitive stress, the tendon likely undergoes a continuous process of adaptation to maintain homeostasis and function

2.3.3 The Influence of Structure on Mechanical Properties

Tendons of different location and function in the body can vary greatly in fatigue quality (Ker, 2002) This raises the question, how does tendon structure differ to create variability in fatigue quality Factors which are most likely to affect the fatigue quality

of tendon include the water content and collagen cross-linking (Ker, 2002) An increase

in tendon water content has been associated with increased fibril spacing, a lower maximum force to failure, increased creep and an increase in fibril sliding (Ker, 2002; Screen, Shelton, Chhaya et al., 2005) Increased fibril sliding indicates that the integrity

of collagen cross-link binding at the fibril level is reduced resulting in a reduction in the ability to resist elongation and failure at a lower stress (Screen, Shelton, Chhaya et al., 2005) In contrast, a reduction in tendon water content has been shown reduce creep (Hoffman, Robichaud, Duquette & Grigg, 2005; Thornton, Shrive & Frank, 2001), increase stiffness and resistance to fatigue (Ker, 2002) and a higher modulus is associated with an increase in the number of collagen cross-links (Ker, 2002; Magnusson et al., 2003)

Further evidence of the importance of collagen cross-linking comes from evaluations of highly stressed horse tendons where the proportion of type I collagen is actually lower than other types of tendons and the proportion of the non fibrillar type IV collagen, which is thought to link type I collagen fibrils, is higher (Ker, 2002) Additionally, in the