Embed Size (px)
Citation preview
i
A Case-control Study of Insertional Achilles Tendinopathy Impairments:
Tendon Characteristics, Dorsiflexion Range of Motion, and Plantar Flexion Strength
by
Ruth L Chimenti
Submitted in Partial Fulfillment of the
Requirements for the Degree
Doctor of Philosophy
Supervised by
Professor James McMahon &
Professor Jeff Houck
Health Practice Research
School of Nursing
University of Rochester Rochester, New York
2014
ii
Biographical Sketch
The author was born in Topeka, KS. She attended Emory University from 2002 to
2006, where she majored in Spanish and had a minor in Dance. She attended the
physical therapy program at Washington University from 2006 to 2009. While there she
participated in the T32 Predoctoral Interdisciplinary Clinical Research Training Program
under the mentorship of Professor Linda Van Dillen. After graduating with her clinical
doctorate, she worked at The Rehabilitation Institute of St. Louis, an outpatient
orthopaedic clinic, for one year before returning to school for a Doctorate in Philosophy
in Health Practice Research. She received a Sproull Fellowship from the University of
Rochester, and attended an interdisciplinary Health Practice Research program in the
School of Nursing under the mentorship of Professor James McMahon. During this time
she did research under the mentorship of Professor Jeff Houck in the Movement
Analysis Laboratory at Ithaca College- Rochester campus. She received a Florence P.
Kendall Doctoral Scholarship in 2010 from the Foundation of Physical Therapy.
The following publications were a result of work conducted during doctoral study:
Houck J, Neville C, Chimenti R . The Foot and Ankle: Physical Therapy Patient
Management Utilizing Current Evidence. (2011). Orthopaedic Section Independent
Study Course series 21.2, Current Concepts for Orthopaedic Physical Therapy.
Barske H, Chimenti R , Martin L, Tome J, Flemister AS, Houck J. Clinical Outcomes,
Static and Dynamic Assessment of Foot Posture after Lateral Column Lengthening
(LCL) Procedure. (2013). Foot & Ankle International Journal. 34(5): 673-83.
Chimenti RL , Flemister AS, Tome J, McMahon JM, Flannery MA, Xue Y, Houck JR.
Insertional achilles tendinopathy alters tendon characteristics and mechanical properties.
(In revision, 2014). Journal of Orthopaedic and Sports Physical Therapy.
iii
Acknowledgements
This work was completed as part of a multi-disciplinary and multi-institutional
effort. I would like to thank my advisor Dr. Jeff Houck for challenging me to strive toward
a high academic standard and also for keeping the fun in research. I would also like to
thank my advisor Dr. James McMahon, who helped to foster my interest in statistics and
to complete this work on time. I appreciate Dr. Deborah Nawoczenski, Dr. Marie
Flannery and Dr. Ying Xue for both their intellectual guidance and moral support
throughout this program. Also, I would like to thank all of the participants who generously
donated their time to this study.
The Health Practice Research program, under the former guidance of Dr.
Margaret Kearney and current guidance of Dr. Bethel Powers, allowed for the
development of a program of study that combined the resources of several departments
at the University of Rochester and Ithaca College. I am grateful to my classmates and
professors at the School of Nursing have given me an appreciation for how multiple
disciplines can learn from each other and work together. I would like to thank the
Department of Orthopaedic Surgery, Foot and Ankle team for their support of my
dissertation project. In particular, I thank Dr. A. Samuel Flemister for his consistent
commitment to this project and contribution of his expertise. Additionally, participation in
courses, journal clubs and lab meetings in the Biomedical Engineering Department has
been an experience that has greatly complimented my progress in the PhD program.
The faculty and students’ interest in translational research has facilitated opportunities,
such as working in Dr. Mark Buckley’s laboratory and the honor of having Dr. Diane
Dalecki as the chair of the dissertation defense. Also the data could not have been
collected without the resources at the Ithaca College, Department of Physical Therapy. I
iv
would like to thank the faculty, staff and students who have helped me each step of the
way. In particular, I would like to thank Josh Tome who helped with every data collection
and with troubleshooting technical problems, which made the process much smoother.
Finally, I would like to thank my friends and family for their encouragement
throughout the PhD program. And I am grateful for my husband Peter. He has given me
the confidence and love needed to look past the bumps in (foot of snow on) the road and
focus on the many joys along the way.
v
Abstract
Purpose: The purpose of this study was to test clinical theories on how impairments in
tendon characteristics, dorsiflexion (DF) range of motion (ROM) and plantar flexion (PF)
strength were related to insertional achilles tendinopathy (IAT). The association between
these impairments and function was also examined.
Methods: Twenty individuals with IAT (age= 58.6± 7.8 years, 55% female) and 20 age-
and gender-matched controls (age= 58.2± 8.5 years, 55% female) volunteered for this
study. Tendon characteristics (diameter, echogenicity, strain, stiffness) were measured
using a combination of ultrasound imaging and an isokinetic dynamometer. Three
dimensional motion analysis was used to quantify ROM and functional strength during
clinical tests and stair ascent. Plantar flexion isometric strength was also documented.
Sides with IAT (n=20) were compared to sides without IAT (n=60) using Generalized
Estimating Equations (GEE). The correlations between impairments and self-reported
function (assessed using the Victorian Institute Sports Assessment- Achilles
questionnaire) were also examined.
Results: Sides with IAT had a larger tendon diameter (P<0.001), lower echogenicity
(P<0.001), higher strain (P=0.007) and lower stiffness (P=0.001) than sides without IAT.
There was not evidence of differences between groups in clinical tests of DF (P=0.414),
but there were significant differences in the percent of available DF used during stair
ascent (P=0.042). Sides with IAT had lower isometric PF strength than sides without IAT
(P=0.010) and used lower ankle power (P<0.001) during stair ascent than sides without
IAT. Impairments in echogenicity, and functional use of DF and PF strength were
associated with lower function (P<0.05).
vi
Conclusions: Clinical theories associating tendon degeneration with IAT were
supported by the alterations in tendon ultrasound imaging. Further, the impingement
theory of IAT tendinopathy was supported. Tendon pathology, defined using ultrasound
imaging, and impairments influence functional status in persons with IAT.
Clinical implications : Although prospective research is needed, ultrasound imaging is
promising as a clinical marker of IAT severity. Evaluation of DF ROM and PF strength
are important because of their link to function. However, because decreased DF ROM is
not typically associated with IAT and IAT tendons show increased stiffness, routine
stretching may be overprescribed and strengthening overlooked.
vii
Contributors and Funding
The conception, implementation and writing of this work was supervised by a
dissertation committee consisting of Professors James McMahon (University of
Rochester, School of Nursing), Ying Xue (University of Rochester, School of Nursing),
Marie Flannery (University of Rochester, School of Nursing), Deborah Nawoczenski
(Ithaca College, Department of Physical Therapy), and Jeff Houck (George Fox
University, Department of Physical Therapy). In addition, Dr. A. Samuel Flemister
contributed to the conception, design and interpretation of the data. Josh Tome
contributed to the design, data collection and analysis of the data. All other work
conducted for the dissertation was completed by the student independently. Graduate
study was supported by a Sproull Fellowship from the University of Rochester and a
Florence P. Kendall scholarship from the Foundation of Physical Therapy.
viii
Table of Contents
Chapter 1: Introduction Clinical Significance Tendon Characteristics Specific aim 1 Dorsiflexion Range of Motion
Specific aim 2 Plantar Flexion Strength Specific aim 3 Model of insertional achilles tendinopathy impairments Exploratory aim 4
1
4
5
6
7
7
8
9
10
Chapter 2: Background on impairments associated with achilles tendinopathy Etiology
Tendon Characteristics
Pathology
Mechanical properties
Dorsiflexion Range of Motion
Capacity
Performance
Plantar Flexion Strength
Capacity
Performance
Model of insertional achilles tendinopathy impairments
11
18
18
19
20
20
20
21
21
22
23
ix
Chapter 3: Methods Design
Sample Self-report measures
Demographics
Victorian Institute of Sport Assessment- Achilles
Numerical Rating Scale
Instrumented measures Ultrasound
Dynamometer
Kinematics and Kinetics
Kinematic and Kinetic Model
Procedures
Recruitment Protection of human subjects Setting Tendon Pathology Tendon Mechanical Properties
Dorsiflexion range of motion capacity
Plantar flexion strength capacity
Range of motion and strength performance
Statistical Analyses
Specific aim 1
Specific aim 2
26
27
30
30
30
31
31
32
32
32
34
35
36
36
36
39
51
57
59
63
63
65
x
Specific aim 3 Exploratory Aim 4 Chapter 4. Results
Sample Characteristics
Missing data
Specific Aim 1- Tendon Characteristics
Reliability
Specific aim 1a (Tendon pathology)
Specific aim 1b (Tendon pathology)
Specific aim 1c (Tendon mechanical properties)
Specific aim 1d (Tendon mechanical properties)
Specific Aim 2- Dorsiflexion Range of Motion
Reliability
Specific aim 2a-b (Range of motion capacity)
Specific aim 2c-d (Range of motion performance)
Post-hoc aim 2e-f (Range of motion performance)
Plantar flexion strength
Reliability
Specific aim 3a (Strength capacity)
Specific aim 3b (Strength performance)
Post-hoc aim 3c (Strength performance)
Model of insertional achilles tendinopathy impairments
Exploratory aim 4a (Tendon characteristics and Function)
Exploratory aim 4b (Range of motion and Function)
67
69
70
71
71
72
73
74
75
75
75
76
77
80
81
81
82
83
84
86
86
87
68
xi
Exploratory aim 4c (Strength and Function)
Exploratory aim 4d (Tendon characteristics and Range of motion)
Exploratory aim 4e (Tendon characteristics and Strength)
Chapter 5. Discussion
Tendon Characteristics
Dorsiflexion Range of Motion
Plantar flexion strength
Model of insertional achilles tendinopathy impairments
Limitations
Statistical models
Conclusions
References
Appendices
88
89
89
92
96
101
103
105
107
108
111
119
xii
List of Tables
Table Title Page
Table 1 Risk factor and corresponding percentage of individuals with insertional achilles tendinopathy
15
Table 2 Inclusion and exclusion criteria for the study sample 28
Table 3 Sample size calculations based on pilot data 29
Table 4 Three types of impairments with associated focus, variables and instrumentation used for measurement
32
Table 5 Tendon pathology on involved and uninvolved sides of persons with achilles tendinopathy in pilot work and a published study
39
Table 6 Achilles tendon elongation and strain in healthy adults 43
Table 7 Tendon strain in participants with insertional achilles tendinopathy and controls
51
Table 8 Two common types of clinical dorsiflexion capacity measures 52
Table 9 Measures of dorsiflexion capacity in participants with insertional achilles tendinopathy and controls
56
Table 10 Maximum isometric plantar flexion torque and associated pain in participants with insertional achilles tendinopathy
59
Table 11 Weight-bearing dorsiflexion capacity in participants with insertional achilles tendinopathy, controls and healthy adults
62
Table 12 Table 13 Table 14 Table 15
Plantar flexion strength capacity in participants with insertional achilles tendinopathy and controls Characteristics of participants with insertional achilles tendinopathy and healthy matched controls The echogenicity of the involved side in participants with insertional achilles tendinopathy compared to two groups: the uninvolved side and controls The diameter of the involved side in participants with insertional achilles tendinopathy compared to two groups: the uninvolved side and controls.
63
70
73
74
xiii
Table 16 Table 17 Table 18 Table 19 Table 20 Table 21 Table 22 Table 23 Table 24 Table 25
The strain of the involved side in participants with insertional achilles tendinopathy compared to two groups: the uninvolved side and controls Tendon stiffness of the involved side in participants with insertional achilles tendinopathy compared to two groups: the uninvolved side and controls Non-weight-bearing and weight-bearing dorsiflexion capacity of the involved side in participants with insertional achilles tendinopathy, the uninvolved side of participants with IAT and controls Dorsiflexion used during stair ascent of the involved side in participants with insertional achilles tendinopathy, the uninvolved side of participants with IAT, and controls Plantar flexion motion used during stair ascent of the involved side in participants with insertional achilles tendinopathy compared to two groups: the uninvolved side and controls Maximum isometric plantar flexion torque of the involved side in participants with insertional achilles tendinopathy compared to two groups: the uninvolved side and controls Peak ankle moment of stair ascent of the involved side in participants with insertional achilles tendinopathy, the uninvolved side, and controls Ankle power during stair ascent on the involved side of participants with IAT compared to the uninvolved side and to the same side (right or left) of controls Correlations between impairments and function on the involved side in participants with insertional achilles tendinopathy Clinical treatment recommendations
74
75
77
79
81
83
84
86
90
105
xiv
List of Figures
Figure
Title Page
Figure 1 Model of insertional achilles tendinopathy impairments 10
Figure 2 External and internal factors that contribute to tendon characteristics
12
Figure 3 a) Radiograph of lateral view of ankle demonstrating Haglund’s deformity and retrocalcaneal osteophytes in a person with IAT b) Magnetic resonance image of retrocalcaneal bursitis
14
Figure 4
Figure 5
Anatomy of achilles tendon insertion a) Neutral ankle position with Haglund’s deformity as the shaded portion on the posterosuperior aspect of calcaneus b) Representation of Haglund’s deformity impinging on achilles tendon in a dorsiflexed position
16
17
Figure 6 Eccentric heel lowering exercise 24
Figure 7 Two segment kinematic model 34
Figure 8 Participant performing stair ascent as viewed by a) Motion Monitor software, and b) video camera
35
Figure 9 Tendon characteristics for uninvolved (top row) and involved (bottom row) sides of a participant with IAT: a-b) Longitudinal ultrasound images of tendon diameter; c-d) Cross-sectional ultrasound images of echogenicity measured by the mean grayscale value; e-f) Corresponding histograms demonstrating the mean grayscale value (scale from 0=black to 255=white) for images c and d.
38
Figure 10 The achilles tendon complex consists of the aponeurosis, covering the soleus muscle, and the free tendon
41
Figure 11 Tendon elongation from rest at to maximum isometric plantar flexion torque
43
Figure 12 Adaptation of figures from Maganaris and Rees articles 45
xv
Figure 13 a) Configuration of ultrasound imaging, isokinetic dynamometer and electromyography during measurement of tendon mechanical properties. Ultrasound images of demonstrating gastrocnemius muscle- achilles tendon junction displacement as the ankle is rotated from a) 10⁰ plantar flexion to b) 10⁰ dorsiflexion
48
Figure 14 Force-elongation curve from 30⁰ planar flexion (PF) to maximum dorsiflexion (DF) for a healthy control (max DF=25⁰), the involved and uninvolved sides of an IAT subject (max DF=15⁰). The points on each curve correspond to force and strain at 10º DF and 10º PF, and stiffness is the slope of the corresponding line highlighted in bold on the linear equation.
50
Figure 15 The three phases of stair ascent include, a) weight acceptance, b) pull-up, c) forward continuance.
61
Figure 16
Stair ascent ankle motion (dorsiflexion is positive) of: a) involved side of participants with insertional achilles tendinopathy, and b) healthy controls .
62
Figure 17
Figure 18
Figure 19
Figure 20
Figure 21
Figure 22
Ankle power during stair ascent performed by a) the involved side of participants with insertional achilles tendinopathy, and b) healthy controls Degrees of ankle motion (dorisflexion is positive) used during stair ascent by percentage of stance by participants with insertional achilles tendinopathy on the involved and uninvolved sides as well as by controls Ankle moment by percentage of stance for participants with insertional achilles tendinopathy (involved and uninvolved sides) and controls Ankle power by percentage of stance for participants with insertional achilles tendinopathy (involved and uninvolved sides) and healthy controls a) Tendon composition (Echogenicity) and b) shape (Diameter) vs. self-reported function (VISA-A) on the involved side of participants with insertional achilles tendinopathy a) Weight-bearing dorsiflexion (DF) capacity and b) Percentage of DF capacity used during stair ascent vs. self-reported function (VISA-A) on the involved side in participants with insertional achilles tendinopathy
63
79
84
86
87
88
xvi
Figure 23
Figure 24
a) Isometric plantar flexion strength and b) ankle power vs. self-reported function (VISA-A) on the involved side in participants with insertional achilles tendinopathy Revised model of insertional achilles tendinopathy impairments
88
91
1
Chapter 1: Introduction
One of the most frequent sites of tendinopathy is the achilles tendon, which
connects the gastrocnemius and soleus muscles to the calcaneus (heel bone). The
estimated lifetime incidence of achilles tendinopathy in the general population is 5%
(Kujala, Sarna, & Kaprio, 2005). Approximately 1/3 of achilles tendinopathy cases have
pain at the site where the tendon attaches to the bone, known as insertional achilles
tendinopathy (IAT) (Karjalainen et al., 2000; Khan et al., 2003; Kujala et al., 2005;
Nicholson, Berlet, & Lee, 2007). Tendinopathy is degeneration occurring within the
tendon, which is defined by a loss of parallel collagen structure, loss of fiber integrity,
fatty infiltration and capillary proliferation (Klauser et al., 2013; Movin, Gad, Reinholt, &
Rolf, 1997). While these histopathological changes define tendinopathy, the diagnosis
for IAT is based on clinical exam. Symptoms used to diagnosis IAT are tenderness of
the tendon within 2cm of the insertion and pain that is aggravated by activity. Insertional
achilles tendinopathy was chosen as a focus for the current study, because it has a
worse prognosis than other types of achilles tendinopathy and there is currently little
available research to improve care.
Clinical Significance
Insertional achilles tendinopathy is a disabling and relatively common foot and
ankle problem. In an epidemiologic study at a foot and ankle orthopaedic surgery clinic,
12% (85/697) of patients had IAT (Waldecker, Hofmann, & Drewitz, 2012). The disability
reported by these patients is generally high. For example, pre-operative subjects in a
surgical intervention study for IAT reported moderate to severe pain and a maximum
walking distance of 1-3 blocks (Johnson, Zalavras, & Thordarson, 2006). Moreover, 56%
(14/25) were unable to work full-time (Johnson et al., 2006). The standard of care for IAT
2
is non-operative, e.g. physical therapy, for 3 to 6 months; unfortunately the prognosis is
poor with 24% to 53% of patients progressing to surgical intervention (Karjalainen et al.,
2000; Kvist, 1991; Nicholson et al., 2007). Surgery can significantly decrease pain and
improve function (Johnson et al., 2006; McGarvey, Palumbo, Baxter, & Leibman, 2002),
however it is also an invasive procedure with risks of poor wound healing and major
complications, e.g. thromboembolic disease. In addition surgery is not an option for
some individuals due to co-morbidities or financial constraints. Given that up to half of
individuals with IAT fail non-operative care and that there is a lack of data on this
disease (Nicholson et al., 2007), there are currently no good evidence-based treatment
options for individuals who wish to avoid surgery.
Insertional achilles tendinopathy is challenging to treat clinically. Approximately
80% of individuals with IAT have concurrent bony deformity and/or bursitis at the tendon
insertion (Jonsson, Alfredson, Sunding, Fahlstrom, & Cook, 2008), which may
complicate recovery. Clinical theories propose that bony deformities compress the
tendon when the ankle is in dorsiflexion (DF), resulting in repetitive trauma to the tendon
(Fowler & Philip, 1945; Jonsson et al., 2008). Because daily tasks, such as climbing
stairs (Andriacchi, Andersson, Fermier, Stern, & Galante, 1980; Reeves, Spanjaard,
Mohagheghi, Baltzopoulos, & Maganaris, 2008), frequently require near end range DF
range of motion (ROM), rehabilitation for IAT symptoms may be difficult without surgery
to remove bony deformities.
While non-operative treatment for IAT is often unsuccessful, one published
plantar flexor (i.e. gastrocnemius and soleus muscles in the calf) strengthening protocol
that minimized DF has been effective. Plantar flexor strengthening was originally
developed for individuals with another form of achilles tendinopathy. The rationale for
3
this treatment is that eccentric loading of the achilles tendon from a plantar flexed to a
dorsiflexed position during strengthening may alter tendon characteristics by promoting
tendon healing (Cook & Purdam, 2009). However, two studies conducted using this
traditional strengthening protocol found that ≤32% of participants with IAT reported
decreased pain and improved function (Fahlstrom, Jonsson, Lorentzon, & Alfredson,
2003; Rompe, Furia, & Maffulli, 2009). Yet a modified strengthening protocol by Jonsson
et al. (2008) had a much higher success rate with 67% (18/27) of participants with IAT
reporting decreased pain and return to pre-injury activity level. The authors attributed the
success of this modified eccentric training program to the avoidance of lowering the heel
into DF, a position which may contribute to bony impingement on the tendon (Jonsson et
al., 2008). Yet to date the success of this non-operative intervention for IAT has not
been replicated nor has the effectiveness of other novel IAT-specific exercise programs
been examined. Therefore the level of evidence supporting the use of current physical
therapy interventions and facilitating the development of novel exercise interventions is
low.
Although the theory behind the mechanism of eccentric strengthening combines
ideas about altering tendon characteristics, avoiding DF ROM, and improving plantar
flexion (PF) strength, surprisingly there is minimal evidence to support these ideas. To
date most non-operative treatments are based on impairments identified in other types
of tendinopathies, and then applied to IAT (Wiegerinck, Kerkhoffs, van Sterkenburg,
Sierevelt, & van Dijk, 2013). However, IAT is unique from other achilles tendon problems
in terms of the population it affects, which tends to be the middle-aged and overweight
with a high incidence of concurrent bony pathology at the insertion. There is a lack of
data on tendon characteristics, DF ROM and PF strength in persons with IAT compared
4
to healthy controls. Therefore, the purpose of this case-control study was to examine the
association between IAT and impairments in tendon characteristics, DF ROM, and PF
strength. The first three specific aims tested clinical hypotheses linking impairments in
tendon characteristics, DF ROM and PF strength to IAT. The fourth exploratory aim
tested one theoretical model, examining the relationships among identified impairments
and self-reported function.
Tendon Characteristics
Tendinopathy is a degenerative process within the tendon, which results in an
altered composition and shape (Rees, Wilson, & Wolman, 2006). Alterations in the
tendon structure due to tendinopathy can be detected with ultrasound imaging (De Zordo
et al., 2010; Khan, Cook, Bonar, Harcourt, & Astrom, 1999; Sconfienza, Silvestri, &
Cimmino, 2010). In order to grade the severity of tendinopathy on imaging, two tendon
characteristics are assessed: 1) composition (lower echogenicity is considered
pathological) and 2) shape (increased diameter is considered pathological) (Archambault
et al., 1998). While the findings of several studies suggest that individuals with IAT have
abnormal tendon composition and increased tendon diameter (Astrom et al., 1996;
Karjalainen et al., 2000; Nicholson et al., 2007), there has only been one case-control
study to examine these signs of tendinopathy with ultrasound imaging and only 4/20
subjects had IAT (Astrom et al., 1996). Thus our knowledge about tendon pathology in
persons with IAT is primarily based on assumptions from findings in other types of
tendinopathy and case series studies (Astrom et al., 1996; Karjalainen et al., 2000;
Nicholson et al., 2007).
Another sign of tendon degeneration that has not been examined in persons with
IAT is mechanical properties. The mechanical properties of tendon strain and stiffness
5
indicate the magnitude of tendon elongation and resistance to a specified load. This data
provides information on how the tendon responds to functional loading conditions, such
as stretching into DF and transmitting PF force. Studies have demonstrated that
individuals with midportion achilles tendinopathy have increased tendon strain, defined
as elongation normalized to tendon length, and decreased stiffness, defined as
resistance to stretch per unit of elongation, compared to controls (Arya & Kulig, 2010;
Child, Bryant, Clark, & Crossley, 2010; Sconfienza et al., 2010). Because midportion and
IAT share similar signs of tendon pathology (hypoechogenicity and increased diameter),
the hypothesis of the current study was that individuals with IAT will have similar
impairments in tendon mechanical properties (strain, stiffness).
Specific Aim 1. This aim examined if tendon pathology (echogenicity and
diameter) and mechanical properties (strain and stiffness) differs between sides with IAT
(involved side of participants with IAT, n=20) and sides without IAT (uninvolved side of
participants with IAT, n=20 and both sides of controls, n=40 sides). Additionally, these
dependent variables are compared between sides (involved vs. uninvolved in the IAT
group) and between groups (IAT vs. controls).
Hypotheses.
H1a. Sides with IAT have a lower echogenicity than sides without IAT in both case and
control groups.
H1b. Sides with IAT have a greater diameter than sides without IAT in both case and
control groups.
H1c. Sides with IAT have a greater strain than sides without IAT in both case and control
groups.
6
H1d. Sides with IAT have a lower stiffness than sides without IAT in both case and
control groups.
Dorsiflexion Range of Motion
Individuals with IAT have pain and difficulty with weight-bearing activities, which
generally require ankle DF. There are two clinical hypotheses on why DF is painful. One
hypothesis proposes that a short and stiff gastrocnemius muscle contributes to IAT
symptoms. An isolated gastrocnemius contracture results in limited DF capacity, which is
common and occurs in approximately one quarter of healthy adults (DiGiovanni et al.,
2002). Limited DF capacity could contribute to an increased pull on the achilles tendon
during functional activities, and thus, over time, contribute to the development of IAT.
According to this idea, DF ROM is limited in patients with IAT, as defined by measures of
DF capacity, i.e. during static non-weight-bearing and weight-bearing measures, and DF
performance, i.e. during a dynamic functional task.
Another hypothesis is that impingement of bony deformity onto the tendon
insertion contributes to pain and progression of IAT. According to this idea, the further
one moves into end-range DF the more that bony deformity impinges on the achilles
tendon insertion (Fowler & Philip, 1945; Jonsson et al., 2008). This hypothesis can be
tested by examining what percentage of end-range DF is used during a functional task.
The percentage of end-range DF is operationally defined by normalizing dynamic DF
performance during stair ascent to the static weight-bearing DF capacity. Surprisingly,
while there are two clinical hypotheses on how DF ROM is linked to IAT, there are no
studies documenting DF ROM in persons with IAT. Examining DF ROM with static and
dynamic measures will test these two ideas.
7
Specific Aim 2. This aim tested two clinical ideas on how DF ROM is linked to
IAT. According to the first, individuals with IAT have a gastrocnemius contracture on the
involved side (involved side of participants with IAT, n=20), which results in limited DF
ROM, as evidenced by lower DF capacity and performance measures, than sides
without IAT (uninvolved side of participants with IAT, n=20 and both sides of controls,
n=40 sides). According to the second idea, individuals with IAT have bony impingement
on the achilles tendon insertion during daily activities. Assuming that this compression
causes IAT, sides with IAT may use a greater percentage of end-range DF than sides
without IAT. Additionally, these dependent variables are compared between sides
(involved vs. uninvolved in the IAT group) and between groups (IAT vs. controls).
Hypotheses based on potential isolated gastrocnemius contracture.
H2a. Sides with IAT have a lower non-weight-bearing DF capacity than sides without IAT
in both case and control groups.
H2b. Sides with IAT have a lower weight-bearing DF capacity than sides without IAT in
both case and control groups.
H2c. Sides with IAT exhibit lower DF during performance of stair ascent than sides
without IAT in both case and control groups.
Hypothesis based on potential bony impingement of tendon.
H2d. Sides with IAT use a greater percentage of end-range DF during performance of
stair ascent than sides without IAT in both case and control groups.
Plantar flexion strength
Individuals with IAT report difficulty with activities that challenge plantar flexion
(PF) strength, such as prolonged walking and climbing stairs. Yet there is no research
documenting that PF strength is impaired in persons with IAT. Decreased PF strength
8
has been linked to midportion achilles tendinopathy (Mahieu, 2006; Silbernagel,
Gustavsson, Thomee, & Karlsson, 2006). A prospective study of male military recruits by
Mahieu et al. (2006) reported that decreased PF strength was a risk factor for midportion
achilles tendinopathy. Mahieu et al. (2006) found that the 10 recruits who developed
midportion achilles tendinopathy over the 6-weeks of basic training had lower baseline
PF strength than the other 59 recruits who did not develop achilles tendinopathy. The
authors interpreted this finding by suggesting that a history of developing greater PF
strength also resulted in a stronger achilles tendon, which was able to sustain the high
loads during basic training (Mahieu, 2006). It is also hypothesized that decreased PF
strength may develop secondary to achilles tendinopathy. A cross-sectional study by
Silbernagel et al. (2006) found that among individuals with midportion achilles
tendinopathy, the involved side had lower PF strength compared to the uninvolved side.
The findings of this study further support the association between PF strength and
midportion achilles tendinopathy, but it remains unclear if the decreased strength was
pre-existing the development of achilles tendinopathy or a secondary impairment due to
achilles tendinopathy. Regardless of the cause, decreased PF strength may be a
primary factor contributing to the disability associated with IAT. However, to date no
studies have examined if PF strength is impaired in persons with IAT.
Specific Aim 3. The purpose of this aim was to examine if PF strength (capacity
and performance) is lower on sides with IAT (involved side of participants with IAT,
n=20) than sides without IAT (uninvolved side of participants with IAT, n=20 and both
sides of controls, n=40 sides). Additionally, these dependent variables were compared
between sides (involved vs. uninvolved in the IAT group) and between groups (IAT vs.
controls).
9
Hypotheses.
H3a. Sides with IAT have a lower isometric PF torque than sides without IAT in both
case and control groups.
H3b. Sides with IAT exhibit a lower ankle moment during stair ascent than sides without
IAT in both case and control groups.
H3c. Sides with IAT exhibit a lower ankle power during stair ascent than sides without
IAT in both case and control groups.
Model of insertional achilles tendinopathy impairme nts
Currently measures of tendon pathology (hypoechogenicity and diameter) and
decreased self-reported function are the only evidence-based impairments associated
with IAT (Astrom et al., 1996; Johnson et al., 2006; Karjalainen et al., 2000; Nicholson et
al., 2007). However, it is unknown if greater impairment of tendon characteristics are
associated with a greater decrease in function. It is also unknown if other hypothesized
impairments associated with IAT (DF ROM and PF strength) impact function.
Understanding how impairments in tendon characteristics, DF ROM and PF strength are
related to function would provide a clinical rationale for developing interventions specific
to impairments identified in persons with IAT.
Further, altered tendon characteristics may impact DF ROM and PF strength.
Although the tendon is only one component of the muscle-tendon unit, decreased
stiffness may be associated with increased DF ROM. An increase in DF ROM could
allow for greater compression of bony deformity against the tendon, contributing to pain
and decreased function. Alternatively, decreased tendon stiffness may decrease PF
strength, due to alterations in the muscle force-length curve, which impairs function.
Examining the relationship between these impairments elucidates if these are dependent
10
or independent predictors of function. Below is a model for testing the relationship
between impairments and function (Figure 1).
Exploratory Aim 4. The purpose of this aim was to examine the association
between self-reported function and impairments observed on the involved side in
participants with IAT.
Hypotheses.
H4a. Impairment in tendon characteristics is associated with lower self-reported function.
H4b. Impairment in DF ROM is associated with lower self-reported function.
H4c. Impairment in PF strength is associated with lower self-reported function.
H4d. Impairment in tendon characteristics, which are associated with lower function, is
associated with impairment in DF ROM.
H4e. Impairment in tendon characteristics, which are associated with lower function, is
associated with impairment in PF strength.
Figure 1 . Model of insertional achilles tendinopathy impairments
11
Chapter 2. Background on Impairments Associated wit h Achilles Tendinopathy
This chapter begins with an overview defining achilles tendinopathy and its
etiology. Insertional deformity is discussed in this section due to its theoretical
importance on specific aims 1 and 2. Literature informing what is known and what data is
lacking about tendon characteristics (aim 1), DF ROM (aim 2) and PF strength (aim 3) in
persons with IAT is described in turn. This chapter concludes with a theoretical model
linking these impairments to each other and to self-reported function in persons with IAT.
Etiology
The understanding of tendinopathy has dramatically changed over the past two
decades due to findings from histological research. Previously the diagnosis for a
chronically painful tendon was “tendonitis,” a term that also implied the presence of
inflammation. However, histological studies of individuals with chronic tendinopathy,
including IAT, have demonstrated an absence of inflammatory markers within the tendon
(Astrom et al., 1996; Movin et al., 1997). The majority of tendon pathologies are now
considered to be due to chronic degeneration and thus are called “tendinosis.” The more
general term “tendinopathy” does not distinguish between “tendinitis” and “tendinosis,”
but is used throughout this paper because it is the term most commonly used in the
literature when referring to IAT.
While the etiology of achilles tendinopathy is unknown, a variety of internal and
external contributing factors have been proposed (Figure 2). External factors are the
focus of current study because they can be modified with non-operative care. For
example, certain prescription drugs, such as fluroquinolones and steroids, are no longer
recommended for patients with tendon problems due to their association with increased
risk of tendon injury (Irwin, 2010). Physical therapy interventions often manipulate
12
external factors, such as environment and training, in order to reduce symptoms and
promote tissue healing. For example, individuals with IAT are often given advice on
shoewear modifications, such as wearing an open-backed shoe that will not rub against
the heel and provoke symptoms. Also in rehabilitation physical activity level is increased
gradually, rather than suddenly, in order to allow for the tendon to adapt to the new
activity. Based on clinical experience, physical therapists prescribe exercises to improve
DF ROM and PF strength in patients with achilles tendinopathy. However, the link
between DF ROM and PF strength and tendon characteristics in persons with IAT is
based on anecdotal and theoretical, rather than empirical, evidence. Understanding how
these external factors relate to tendon characteristics would be useful for development of
novel physical therapy interventions for IAT.
Figure 2. External and internal factors that contribute to tendon characteristics
Demographic and disease related internal factors may not be modifiable, but may
affect the relationship between external factors and IAT. Foot shape, including both pes
13
cavus (rigid supinated foot) and pes planus (flexible pronated foot), have been
hypothesized to asymmetrically load the achilles tendon due to an altered alignment of
tendon insertion with calcaneal inversion (associated with pes cavus) or eversion
(associated with pes planus) (Irwin, 2010; Saltzman & Tearse, 1998). Aging is
associated with increased cross-sectional area of the achilles tendon and decreased
stiffness (Stenroth, Peltonen, Cronin, Sipila, & Finni, 2012), thus some markers of
tendinopathy may also describe changes associated with normal aging. Also, women
exhibit lower tendon stiffness compared to men, thus both age and gender contribute to
tendon mechanical properties (Stenroth et al., 2012). Comorbidities, such as
hypertension, diabetes and obesity, are associated with increased risk of tendon injury
(Irwin, 2010; Thomas et al., 2010). Additionally, there are other factors, such as
neovascularization and neural pain, that are theorized to contribute to achilles
tendinopathy (Rees et al., 2006). While the relationship between internal factors and
achilles tendinopthy has not been definitively estabilished in the literature, a low level of
evidence suggests that foot shape, age, gender and co-morbidities may all be
considered potential risk factors for altered tendon characteristics and, thus, also for IAT.
Another key internal factor that drives both operative and non-operative care for
IAT is insertional deformity. Tendinopathy at the achilles insertion often occurs alongside
bony and/or soft-tissue deformity, such as Haglund’s deformity, retrocalcaneal bursitis
and bone spurs (Irwin, 2010) (Figure 3). These bony and soft-tissue deformities are
distinct from IAT, which refers specifically to achilles tendon degeneration. However, the
concurrence of IAT, Haglund’s deformity and retrocalcaneal bursitis is prevalent enough
that the term “Haglund’s triad” has been used to describe these three conditions
occurring together (Brunner et al., 2005; DeVries, Summerhays, & Guehlstorf, 2009;
14
Schneider, Niehus, & Knahr, 2000; Sofka, Adler, Positano, Pavlov, & Luchs, 2006).
Insertional deformities are considered to contribute to trauma and pain at the insertion,
and so surgical intervention for IAT often removes bony abnormalities and retrocalcaneal
bursitis (Anderson, Suero, O'Loughlin, & Kennedy, 2008; Brunner et al., 2005; DeVries
et al., 2009; Hu & Flemister, 2008; K. A. Johnson & Strom, 1989; McGarvey et al., 2002;
Schneider et al., 2000; Yodlowski, Scheller, & Minos, 2002). Current surgical practice of
removing insertional deformity is based on the idea that insertional deformity contributes
to achilles tendinopathy, yet this assumption is actually supported by a low level of
evidence.
Figure 3. a) Radiograph of lateral view of ankle demonstrating Haglund’s deformity and retrocalcaneal osteophytes in a person with IAT (Hu & Flemister, 2008) b) Magnetic resonance image of retrocalcaneal bursitis (Crawford & Desruisseau, 2010)
b a
15
Table 1. Risk factors and corresponding percentage of individuals with insertional achilles tendinopathy (IAT)
Risk Factor Disease status
IAT Controls
Retrocalcaneal bursitis
(15+27)/ (18 + 34) = 81% (Jonsson et al., 2008;
Karjalainen et al., 2000)
(15+0)/ (100+50) = 10% (Falsetti et al., 2004; Soila,
Karjalainen, Aronen, Pihlajamaki, & Tirman, 1999)
Bone Spurs 25/34 = 74% (Jonsson et al., 2008)
12/50= 24% (Falsetti et al., 2004)
Haglund’s deformity
27/34 = 79% (Jonsson et al., 2008)
10/60 = 17% (Chauveaux, Liet, Le Huec, &
Midy, 1991)
While insertional deformity and tendinopathy frequently occur together, there are
no prospective studies examining the causative role of insertional deformity on IAT. The
best available evidence to examine the relationship between insertional deformity and
achilles tendinopathy is a combination of findings from several case series studies
(Table 1). However, these studies used different imaging modalities to determine the
presence of deformity, so it is unclear if differences in percentages of the risk factor are
due partially to differences in imaging. Also, some studies only reported findings by the
number of feet rather than individuals (Chauveaux et al., 1991; Jonsson et al., 2008;
Soila et al., 1999), which confounds the estimation of prevalence since the rate of
bilateral involvement is unknown. Despite the potential bias in these samples, there
appears to be a higher percentage of individuals with IAT who have insertional deformity
compared to individuals without IAT (Table 1). Lacking better data, the risk ratio of IAT
associated with retrocalcaneal bursitis is 4.0= , with bone spurs is 3.5=
, and with Haglund’s deformity is 5.9= (calculated from Table 1).
Thus, current evidence supports the surgical practice of removing insertional deformity
16
because individuals with insertional deformity are 3.5 to 5.9 times more likely to have
IAT than individuals without these risk factors. However, the risk ratio calculated from
prior work is not ideal, and a prospective study design may have a different outcome.
The anatomy of the achilles tendon insertion helps explain how insertional
deformity can protect against or contribute to IAT (Figure 4). As the achilles tendon
descends from the plantar flexor muscles to insert on the rough posterior surface of the
calcaneus, the posterosuperior peak of the calcaneus is separated from the achilles
tendon by the retrocalcaneal bursa (Chauveaux et al., 1991). Thus, the anterior wall of
the distal 2 cm of the achilles tendon is composed of the retrocalcaneal bursa,
proximally, and the calcaneus, distally (de Palma, Marinelli, Meme, & Pavan, 2004).
During DF the posterosuperior surface of the calcaneus functions as a pulley, this
converts some of the tension into compression over the anterior surface of the insertion.
Thus, the anatomy of the achilles tendon insertion is designed to dissipate forces at the
insertion.
Figure 4. Anatomy of achilles tendon insertion
17
Haglund’s deformity and retrocalcaneal bursitis are enlargements of structures
designed to decrease stress at the tendon insertion. However instead of dissipating
stress, Haglund’s deformity creates greater impingement on the tendon as the ankle
moves into DF (Figure 5). Simultaneously, while the retrocalcaneal bursa is designed to
flatten during DF to reduce compression in the retrocalcaneal space, bursitis decreases
the ability of the structure to adapt. Therefore, retrocalcaneal bursitis may also contribute
to the amount of compression occurring against the anterior side of the tendon. The idea
of insertional deformity causing trauma to the tendon is also supported by the
observation on MRI that the anterior side of the achilles tendon is most commonly
affected in IAT (de Palma et al., 2004). In summary, the key components of this
impingement theory are that 1) bony impingement occurs in DF, and 2) the location of
tendon pathology is adjacent to bony deformity.
Figure 5. a) Neutral ankle position with Haglund’s deformity as the shaded portion on the posteriosuperior aspect of calcaneus b) Representation of Haglund’s deformity impinging on achilles tendon in a dorsiflexed position
a) b)
18
Tendon Characteristics
Pathology. Several studies have reported that pathology associated with chronic
tendinopathy can be detected with ultrasound imaging (De Zordo et al., 2010; Khan et
al., 1999; Sconfienza et al., 2010). Moreover, imaging of tendon pathology may have
prognostic value (Archambault et al., 1998; Fredberg & Bolvig, 2002; Nicholson et al.,
2007; Zanetti et al., 2003). A retrospective chart review by Nicholson, Berlet and Lee
(2007) reviewed MRI images of individuals with IAT treated non-operatively (8
individuals, 16 tendons) and operatively (74 individuals, 83 tendons). The severity of
tendon pathology was graded as: Grade I indicating a tendon diameter of 6-8 mm and
non-uniform degeneration; Grade II indicating a diameter of >8mm with uniform
degeneration of <50% of tendon width; and Grade III indicating a tendon diameter >8mm
and uniform degeneration of >50% tendon width (Nicholson et al., 2007). The authors
found that the percentage of tendons that underwent surgery varied by severity of
pathology, as follows: 12.5% (2/16 tendons) with Grade I, 90.8% (59/65 tendons) with
Grade II, and 70.4% (19/27) with Grade III (Nicholson et al., 2007). Yet there are several
limitations to this retrospective study. Magnetic resonance imaging (MRI) was
unavailable for many patients who chose non-operative care, and it is unstated if the
person who graded the severity on MRI was blind to subjects’ choice of care (Nicholson
et al., 2007). Despite limitations, the findings of this study suggest that severity of tendon
pathology (changes in composition and shape) may be associated with prognosis in
terms of the progression to surgical intervention.
While MRI is considered the gold standard for assessing tendon pathology, an
ultrasound exam can measure similar parameters in a more time- and cost-effective
manner (Astrom et al., 1996). While tendon thickness >6mm is commonly used as a
19
diagnostic criterion for IAT (Karjalainen et al., 2000; Khan et al., 2003; Nicholson et al.,
2007), no studies have examined the validity of this cut-off value nor attempted to
measure the severity of IAT using quantitative ultrasound measures. If ultrasound
measures of tendon pathology are associated with IAT, then ultrasound can be used
clinically to supplement diagnostic and prognostic decisions.
Mechanical Properties. Tendinopathy is also associated with changes in the
tendon mechanical properties, which has been demonstrated using a variety of
ultrasound imaging methods (Arya & Kulig, 2010; Child et al., 2010; De Zordo et al.,
2010; Sconfienza et al., 2010). A combination of ultrasound imaging and isometric
dynamometry can capture in vivo achilles tendon strain and stiffness during clinically
relevant tasks, such as an isometric contraction or a passive stretch (Arya & Kulig, 2010;
Child et al., 2010; Kawakami, Kanehisa, & Fukunaga, 2008). In vivo evaluation of tendon
mechanical properties demonstrated increased tendon strain and decreased stiffness in
participants with midportion achilles tendinopathy compared to controls (Arya & Kulig,
2010; Child et al., 2010; Sconfienza et al., 2010). While a similar increase in strain and
decrease in stiffness are anticipated in IAT, no current studies exist in patients with IAT.
Further, current therapeutic strategies do not focus on restoring tendon mechanical
properties. It is possible that restoring tendon mechanical function in addition to reducing
symptoms during rehabilitation is an important element of preventing re-occurrence of
tendinopathy.
In summary, case-control studies on individuals with midportion achilles
tendinopathy demonstrate that individuals with midportion achilles tendinopathy have
increased tendon strain and decreased stiffness compared to healthy controls (Arya &
Kulig, 2010; Child et al., 2010). However the possibility of increased tendon strain and
20
decreased stiffness in persons with IAT has not been tested. Examining the relationship
between mechanical properties and IAT tests the theory that this type of tendinopathy
also alters mechanical functioning of the tendon. Furthermore, this information can be
used to develop novel exercise interventions to decrease symptoms and improve tendon
mechanical properties.
Dorsiflexion range of motion
Capacity. One clinical hypothesis proposes that a gastrocnemius contracture
often contributes to the development of IAT. This idea is based on a case-control study
by Digiovanni et al. (2002) which compared static non-weight-bearing DF capacity in
persons with forefoot and midfoot pain to healthy adults. Based on differences between
groups, DiGiovanni et al. (2002) defined <5⁰ DF as a threshold for increased risk of foot
and ankle problems. This definition of hypomobility had 77% specificity. In other words,
among controls only 8/34 had <5⁰ DF, while 22/34 participants with foot pain had <5⁰ DF
(DiGiovanni et al., 2002). While the findings of this study are used to support the idea
that limited DF capacity is linked to foot and ankle problems, this study did not examine
individuals with rearfoot problems, such as achilles tendinopathy. It is unknown if limited
DF capacity is associated with IAT, and if non-operative treatments, such as stretching,
are appropriate to be generally prescribed for patients with IAT.
Performance . It has been demonstrated that limited DF capacity results in
decreased DF during stair descent (Moseley, Crosbie, & Adams, 2003). A study by
Moseley, Crosbie and Adams (2003) compared ankle kinematics during stair descent
between 9 men with ankle hypomobility (non-weight-bearing DF capacity <1 standard
deviation (SD) below the mean) and 9 men with hypermobility (>1 SD above the mean).
The authors found that the group with hypomobility used about 10⁰ less DF during stair
21
descent (mean ± SD= 27.9º± 4.8) than the group with hypermobility (39.4º± 5.7). One
weakness of this study is the inconsistency in the definition of DF ROM. Subjects were
grouped into “hypomobility” and “hypermobility” based on a non-weight-bearing DF
capacity measure taken with the knee straight, which would indicate gastrocnemius
muscle tightness. However, the authors examined a functional task requiring weight-
bearing DF with the knee bent, which would primarily test soleus muscle tightness. While
the study findings support a potential relationship between DF capacity and
performance, it is unknown how this relationship may change with different measures of
capacity (non-weight-bearing vs. weight-bearing) and measures of performance (knee
position during measurement of capacity mirrors position during performance).
Another clinical hypothesis is that bony deformity impinges against the tendon in
DF. According to this idea, greater use end-range DF during functional tasks contributes
to greater pain and difficulty due to bony impingement on the tendon. One way to test
this theory is to examine what percentage of end-range DF is used during performance
of a functional task. According to the impingement hypothesis, individuals with IAT use a
greater percentage of end-range DF than controls. Because neither of these clinical
hypotheses have been tested in persons with IAT, it is unknown if one of or both
hypotheses are true in persons with IAT.
Plantar flexion strength
Capacity . Individuals with IAT report pain and difficulty with tasks that challenge
PF strength, such as climbing stairs and prolonged walking. While DF ROM can be one
contributing factor, decreased PF strength can also contribute to the difficulty performing
these tasks. Midportion achilles tendinopathy has been linked to decreased PF strength
(Mahieu, 2006; Silbernagel et al., 2006). Mahieu et al. (2006) measured isokinetic PF
22
strength in military recruits at the start of training. The authors found that the 10 military
recruits who developed midportion achilles tendinopathy over the 6-weeks of basic
training had lower baseline PF strength during the isokinetic testing than the other 59
recruits who did not develop midportion achilles tendinopathy (Mahieu, 2006).
Additionally, a cross-sectional study by Silbernagel el al. (2006) of 42 patients with
midportion achilles tendinopathy also found that individuals with midportion achilles
tendinopathy have decreased PF strength on the involved side compared to the
uninvolved side. In contrast, other studies of individuals with midportion achilles
tendinopathy have demonstrated no differences in isometric plantar flexion strength
compared to controls (Arya & Kulig, 2010; Child et al., 2010). However, in the study by
Arya and Kulig (2010) there was a trend (P=0.073) for the midportion achilles
tendinopathy group to have lower isometric PF strength than controls. Additionally, in the
study by Child et al. (2010) all participants were athletes running at least 12 miles per
week, and so this sample did not have very severe symptoms. Given the chronicity and
severity of IAT, this form of achilles tendinopathy is likely associated with PF weakness.
Performance . Functional tasks require a dynamic interaction between DF ROM
and PF strength, and impairment in either could impair functional ability. Diminished
ability of the plantar flexor muscles to eccentrically control DF ROM can result in
increased tendon impingement and pain. Thus, plantar flexion (PF) strength deficits may
contribute to disease progression. Also, due to potential pain with contraction of the
plantar flexors during functional activities, such as stair climbing, individuals with IAT
may avoid use of PF strength and develop weakness secondary to IAT. Decreased PF
strength can contribute to further progression of IAT and/or be a primary factor
contributing to the disability associated with IAT.
23
Model of insertional achilles tendinopathy impairme nts
The only evidence-based impairments associated with IAT are tendon pathology
and decreased function (Astrom et al., 1996; Johnson et al., 2006). However, the link
between these impairments is unclear. Tendon characteristics can directly impact
functional ability to perform a variety of functional tasks. Alternatively, altered tendon
characteristics can affect DF ROM and/or PF strength, which can also decrease
functional ability. While it is important to understand how individuals with IAT differ from
controls, it is also important to understand if these differences impact function. The
model of insertional achilles tendinopathy impairments (Figure 1) outlines hypothesized
relationships between impairments and function on the involved side in persons with
IAT. Assessing the relationship between impairments identified in persons with IAT can
validate the clinical relevance of identified impairments.
Current clinical interventions are based on the assumption that mechanically
loading the tendon through the use of DF ROM and/or PF strengthening will improve
function. However, it has not been tested if impairments in DF ROM and PF strength are
related to function in persons with IAT. For example, a modified PF strengthening
exercise protocol resulted in decreased pain and increased function in persons with IAT
(Jonsson et al., 2008). Jonsson et al. (2008) proposed that eliminating DF from the
exercise minimized compression between the tendon and insertional deformity (Figure
6). However, the study by Jonsson et al. (2008) only had self-report measures as an
outcome. Thus while the intervention, designed to mechanically load the tendon through
PF strengthening while avoiding DF, was able to improve function, it remains unclear
how this improvement in function occurred.
24
Figure 6. Eccentric heel lowering exercise (Jonsson et al., 2008)
In addition, there may be relationships between tendon characteristics, DF ROM
and PF strength. If a pathological tendon has 1.3% greater strain than healthy tendon
(Arya & Kulig, 2010; Child et al., 2010), then this can result in increased DF ROM on the
involved side. For example, based on pilot data in which the tendon in controls strained
about 3% over a 20⁰ ankle excursion (See preliminary data in Methods, Table 7) (R.
Chimenti, J. Tome, A. S. Flemister, & J. Houck, 2014), an additional 1.3% tendon strain
in persons with IAT could theoretically contribute an additional 8⁰ (20⁰/3.1%=6.5⁰/%,
1.3%*6.5⁰/%=8.4⁰) in DF capacity. Also, tendon characteristics may relate to PF
strength. This idea has been supported by research demonstrating that immobilization of
the plantar flexors decreased achilles stiffness (Matsumoto, Trudel, Uhthoff, & Backman,
2003). In contrast, increased PF strength, after a standardized isometric exercise
protocol, was associated with increased stiffness (Arampatzis, Peper, Bierbaum, &
Albracht, 2010). Thus, decreased PF strength in persons with IAT may contribute to
impairments in mechanical properties. While there are many potential relationships
between impairments associated with IAT and function, these theories have not been
25
tested. In order to develop evidence-based theories on IAT impairments, baseline data is
useful to examine proposed associations between identified impairments and self-
reported-function.
Summary
There is little empirical evidence on impairments (tendon characteristics, DF
ROM, PF strength) associated with IAT. Therefore both non-operative and operative
care have been driven primarily by theory and anecdotal evidence. The purpose of the
current study is to test the theories described in this chapter on impairments associated
with IAT. Moreover, in order to inform clinical decision making on which impairments are
functionally relevant, the association between these impairments and self-reported
function is examined.
26
Chapter 3. Methods
Design
The level of evidence supporting current non-operative care for individuals with
IAT is low. Most studies on persons with IAT consist of case series studies, which lack a
control group, documenting the effect of an intervention through self-report measures of
pain and function. To provide baseline data for higher level intervention studies, a case-
control study design was used to compare tendon characteristics, DF ROM and PF
strength on sides with IAT (n=20 involved sides of participants with IAT) to sides without
IAT (n=60: 20 uninvolved sides of participants with IAT and 40 sides of 20 controls).
Healthy controls were matched to IAT subjects based on age- and gender, because
these factors have been shown to influence the dependent variables of tendon
characteristics, DF ROM, and PF strength (Stenroth et al., 2012; Vandervoort et al.,
1992; Vandervoort & McComas, 1986). Both sides of participants with IAT and controls
were included in the primary analyses (sides with IAT vs. sides without IAT) to account
for the normal correlation between sides within an individual.
Further analyses were done to examine if findings from the primary analyses
were consistent with comparisons between sides (involved vs. uninvolved) and between
groups (IAT vs. controls). Comparison between the involved and uninvolved sides within
participants with IAT allowed for comparison of the influence of IAT without the
confounding influence of demographic factors on which participants were not matched.
However, a weakness of a between sides comparison is that the uninvolved side may
not represent normal. For example, the uninvolved side may compensate for the
involved side and, thus, have increased strength compared to normal, or a
compensatory motor pattern may affect both sides. Therefore, the involved side was
27
also compared to controls. Assuming that there may be some limb asymmetry in
controls (Sadeghi, Allard, Prince, & Labelle, 2000), the limb side of the control chosen
for analysis was matched to the side that was involved for the similarly aged IAT
participant. Thus for the between group comparison, IAT participants and controls were
matched on age, gender and limb side (right vs. left).
Sample
A total of 40 individuals were recruited to participate. All subjects were adults
aged 21 or older and spoke English. The sample included two groups: 1) 20 individuals
with chronic unilateral IAT, and 2) 20 healthy adults matched based on age (+/- 4 years)
and gender. Participants with IAT were diagnosed by an orthopaedic surgeon and/or
physical therapist. The diagnosis for IAT was based on clinical exam, and was defined
by symptoms for at least 3 months, tenderness to palpation within 2cm of the achilles
insertion, and symptoms that were aggravated by activity. For both groups, individuals
were excluded if they had foot/ankle surgery or were unable to safely climb stairs
independently without a railing. Individuals with systemic disorders that could affect the
results were excluded, in specific these conditions were rheumatologic disorders,
neurological conditions, or diabetes. In addition, pregnant women were excluded based
on the effect of estrogen of tendon and ligament structure (Irwin, 2010). Individuals were
excluded from the control group if they had a history of achilles tendinopathy or if they
had lower extremity pain that limited physical activity in the past 6 months.
28
Table 2 . Inclusion and exclusion criteria for the study sample Inclusion criteria Exclusion criteria
- Aged 21 years or older - History of foot/ankle surgery - Speak English - Unable to safely climb stairs
independently without handrails - Rheumatologic disorder
- Neurological condition - Diabetes
- Pregnancy Insertional achilles
tendinopathy Controls Insertional achilles
tendinopathy Controls
- Symptoms of IAT for >3 months
- Unilateral achilles tendinopathy
- Match Insertional achilles tendinopathy group based on age and gender
- Other injury that primarily limits activity
- History of achilles tendinopathy
- Lower extremity pain limiting activity in the past 6 months
A sample size of 40 participants (at least 20 sides per comparison) was chosen
based on a power analysis indicating that a sample size of 20 participants with IAT was
needed to detect statistically significant Pearson correlations ≥0.6 with a power of 0.80
(analysis for aim 4). Power analyses were also performed, assuming a minimum power
of 0.80 and a two-tailed alpha level of 0.05, for an outcome variable from each specific
aim. Power analyses were performed using paired/independent samples t-test as an
estimate of the sample size needed for repeated-samples/one-way ANCOVA’s. Data
from pilot work for each specific aim were used to estimate variability and effect sizes
(Table 3). For tendon characteristics (aim 1) and PF strength (aim 3), the number of
sides needed per comparison was consistent with a sample size of 40. However, for the
within sides comparison for DF capacity (aim 2) a sample size of 102 was needed for a
statistically significant result due to the high variability in weight-bearing DF for
participants with IAT. Based on this power analysis and methodological reasons (See
Methods, DF capacity, preliminary studies), the methods to measure DF capacity were
29
changed. Therefore, excluding the between sides comparison for aim 2, a sample size
of 40 participants was sufficient to detect statistically significant differences for the aims
of the current study.
Table 3. Sample size calculations based on pilot data Impairment Focus Variable Side, group Mean±
SD Number of participants needed
Tendon characteristics
Mechanical properties
Passive stiffness (Nm/%)
Involved side, IAT
(n=9 sides)
1.8± 0.6
Uninvolved side, IAT
(n=5 sides)
2.6± 0.5
N=14
Controls
(n=6 sides)
3.8± 1.7
N=26
DF ROM Capacity Weight-bearing
(⁰)
Involved side, IAT
(n=6 sides)
24.6± 7.0
Uninvolved side, IAT
(n=4 sides)
26.6± 3.0
N= 102
Controls
(n=4 sides)
38.7± 5.0
N= 10
PF Strength Capacity Isometric PF torque
(Nm)
Involved side, IAT
(n=6 sides)
73.2± 18.9
Uninvolved side, IAT
(n=6 sides)
94.0± 23.2
N=26
Controls
(n=3 sides)
108.3±16.1
N=12
30
Self-Report Measures
Demographics. A brief medical history was taken including: BMI, co-morbidities,
treatment history, history of steroid injections/use, insertional bony deformities (when
available from x-ray) [See appendix for guide of verbal interview, “Study Subjects Intake
Sheet”]. These demographic variables were used to describe the sample and considered
as potential risk factors for developing IAT. Further, this information was used in
analyses as a covariate and/or for identifying outliers.
Victorian Institute of Sport Assessment- Achilles Questionnaire (VISA-A).
The VISA-A is an 8 item scale that measures self-reported achilles tendon stiffness,
pain, and return to sport (Robinson et al., 2001) [See appendix]. Questions from the
scale were read to the participant and scored by the examiner. The scale was developed
through review of published items, interviewing clinicians to find unpublished items used
in the clinic, and patient interviews. Most items are scored on an 11 point scale with 0
indicating extreme pain, stiffness or functional limitation and 10 indicating no limitation.
The score on the VISA-A is the sum of the 8 items (1 item on sport participation is worth
30 points). The reliability and validity of the scale was assessed in 2 groups: individuals
with a diagnosis of achilles tendinopathy (n=59), and individuals without achilles
tendinopathy (n=87). The scale has a high test-retest reliability of 0.81, with the tests
given one week apart. In terms of construct validity, the scale had only fair correlation (r
=0.57-0.58) with two other scales: Percy and Conochie, for assessment of achilles
tendon rupture; Curwin and Stanish, for hindfoot problems. However, these two scales
may not be an appropriate gold standard, given that there is no published data on
validity or reliability. Construct validity may be better demonstrated by the ability of the
measure to discriminate between individuals with achilles tendinopathy (mean± SD; non-
31
surgical patients=64±17%, surgical patients= 44±28%) and individuals without achilles
tendinopathy (university students= 96±7%, runners=98± 3%) (Robinson et al., 2001).
The VISA-A was used in the current study due to its high reliability and validity tested
within the population of interest, as well as the ability to use this measure to compare
severity of symptoms to other studies.
Numerical Rating Scale. The Numerical Rating Scale (NRS) scale was used to
verbally assess pain at the achilles tendon insertion from 0, indicating no pain, to 10
indicating the worst pain imaginable (Williamson & Hoggart, 2005). As in clinical
practice, pain rating was monitored during each task, and was recorded prior to and
immediately after the testing session for participants with IAT. If a subject reported >4/10
pain, then the task was discontinued. For controls, the procedures were not anticipated
to provoke pain and the NRS was not used. This measure facilitated quick verbal
communication between the participant and the researcher about pain without the need
for pen, paper, and a clipboard at each assessment point. The NRS was used in the
current study, because the primary use of this scale was to assess subject discomfort
throughout the testing session.
Instrumented Measures
Instrumented measures were used to test hypotheses for specific aims (Table 4).
A combination of instrumented and self-report measures (VISA-A) were used to test
hypotheses for the fourth exploratory aim. Pilot work was performed to assess the
feasibility of performing the proposed data collection measures. Unless otherwise
referenced, data and images presented in the following section were generated from
pilot work as described by the methods below. The author conducted all testing and
completed all calculations.
32
Table 4. Three types of impairments with associated focus, variables and instrumentation used for measurement Impairment Focus Variable Instrumentation
Specific Aim 1: Tendon
Characteristics
Pathology Echogenicity Ultrasound
Diameter Ultrasound
Mechanical properties
Strain Ultrasound
Stiffness Ultrasound and Dynamometer
Specific Aim 2: Dorsiflexion (DF) Range of Motion
(ROM)
Capacity Non-weight-bearing Motion Capture
Weight-bearing Motion Capture
Performance Stair ascent Motion Capture
Percentage of end-range used
Motion Capture
Specific Aim 3: Plantar flexion (PF) Strength
Capacity Isometric PF torque Dynamometer
Performance Ankle moment Motion Capture and
Force Plates Ankle power
Ultrasound . B-mode imaging on a Phillips HD11 Digital Ultrasound machine with
a linear array probe (L12-3) was used at a frequency of 10-12 MHz. The gain was set at
56 and the depth at 3 cm for all images. Ultrasound images were stored digitally and
processed using Image J (http://rsbweb.nih.gov/).
Dynamometer . The Biodex System 4 (Biodex Medical Systems, Shirley, NY)
was used to measure passive PF torque for the calculation of tendon stiffness as well as
isometric PF strength. Participants were in a sitting position with the seatback reclined to
45⁰, the chair height was adjusted so that the tibia was parallel to the floor, and the hip
was in neutral adduction/abduction.
Kinematics and Kinetics . The parameters used for kinematics and kinetics
were based on previous studies in the foot and ankle literature (Houck, Neville, Tome, &
Flemister, 2009; Houck, Tome, & Nawoczenski, 2008). A 9 camera Optotrak Motion
33
Analysis System (Northern Digital, Inc., Waterloo, Ontario, model 3020) was used to
track lower extremity motion during stair ascent at a rate of 60 Hz and measurement of
non-weight-bearing DF ROM. Infrared diodes (IREDs) were attached to the lower
extremity based on the kinematic model (described below), and actively emitted light for
the cameras to track 3D motion. The accuracy of tracking an IRED is up to 0.1 mm with
this research-grade motion capture system (Maletsky, Sun, & Morton, 2007). After
collection, the kinematic data was smoothed using a 4th order, zero phase lag, and
Butterworth filter with a cut-off frequency of 6 Hz.
A floor mounted force plate (Kistler, Instrument Corp., Amherst, NY, model 9286)
recorded the 3D ground reaction force during stair ascent. Force plate data was
recorded at a sampling rate of 1,000 Hz. The accuracy of the force plate recordings are
± 2 % full scale on all channels (Kistler, 1984). The start of stance phase was defined by
vertical ground reaction forces above 20 N (4.5 pounds) and the end of stance phase
was defined by the vertical ground reaction forces lower than 20 N. The stance phase of
all subjects was normalized by defining 101 points from 0, start of stance, to 100, end of
stance. The Innsport Training Software (Innsport Training, Chicago, IL) was used to
integrate the IRED motion capture and force data to determine relative angular
displacement and ground reaction force data.
A force gauge (Model SML-25 Interface, Scottsdale, AZ) was used to standardize
the measurement of non-weight-bearing DF ROM. The force gauge was connected in
series with a metatarsal pad designed to distribute pressure for a force applied at the
head of the second metatarsal. The voltage associated with this force was synchronized
with kinematic data so that the applied torque to the metatarsals was standardized
across participants.
34
Kinematic and Kinetic Model. The kinematic model for stair ascent consisted of
4 segments, or bones, which were assumed to be rigid bodies: hallux, first metatarsal,
calcaneus and the tibia (Figure 7). Each segment was tracked by a set of 3 IREDs on a
thermoplastic molded platform designed to contour the bone, which was placed on the
skin. In addition, an IRED was placed on the fifth metatarsal. For the current study, only
kinematic data of the calcaneus relative to the tibia in the sagittal plane (plantar
flexion/dorsiflexion) was used for analysis. Digitized points were used to establish ankle
and knee joint centers through the Innsport Training Software (Innsport Training,
Chicago, IL), as well as the midpoint between the first and fifth metatarsal heads. The
coordinate systems were established for each segment with the y-axis oriented
superior/inferior (positive is superior), x-axis oriented anterior/posterior (positive is
anterior), and the z-axis oriented medial/lateral (positive is toward the subject’s right)
(Figure 8). These coordinate systems were used to calculate relative joint angles, such
as ankle PF/DF defined as calcaneal motion around the z-axis relative to the tibia.
Figure 7 . Two segment kinematic model
Tibia segment
Calcaneus segment
35
Figure 8 . Participant performing stair ascent as viewed by a) Motion Monitor software, and b) video camera
The kinetic model combined kinematic data with force plate data (ground reaction
force). Ankle power was defined as the ankle joint moment (moment arm * ground
reaction force) times the calcaneal velocity in the sagittal plane (angular rotation per
second). For the current study it was relevant that angular rotation into PF only
contributes to power when at least 20 N of force was registered on the force plate.
Therefore, ankle power was quantifying the rate of PF in a weight-bearing position.
Procedures
Recruitment . Patients with IAT were identified by orthopedists and physical
therapists at the University of Rochester. These patients were given a flyer that
described the study and had the email/phone number of the study coordinator. Patients
who were interested in the study, and gave verbal consent to be contacted, were
contacted by the study coordinator. Controls were recruited through word of mouth,
36
distribution of flyers at local recreation centers and researchmatch.org. Individuals who
qualified for the study were scheduled for a testing session at their convenience.
Protection of human subjects . All subjects were informed of the study
procedures and signed a consent form approved by the University of Rochester,
Research Subjects Review Board, and the Ithaca College, All College Review Board for
Human Subjects Research.
Setting . All study procedures occurred at the Movement Analysis Laboratory of
Ithaca College- Rochester Center, Department of Physical Therapy.
Tendon pathology.
Data collection. Ultrasound measures of altered tendon composition and shape
due to tendon pathology has been validated with histological studies (Astrom et al.,
1996; Klauser et al., 2013). In a prospective study by Astrom et al. (1996) 27
consecutive patients with achilles tendinopathy, 4 with IAT, were examined with
ultrasound, magnetic resonance imaging (MRI), and histopathological examination from
tendinous tissue excised during surgery. Ultrasound and MRI were used to examine the
presence of abnormalities (i.e. hypoechogenicity on ultrasound and high signal intensity
on MRI) and the maximum diameter of the tendon. The ultrasound and MRI techniques
had a similar accuracy in identifying abnormalities on the involved versus uninvolved
sides (Sensitivity: ultrasound= 71%, MRI= 75%; Specificity: ultrasound=92%, MRI=
86%). In addition, the diameter measured both by ultrasound (r= 0.51) and MRI (r=0.49)
correlated with the histopathological score graded by an experience pathologist (Astrom
et al., 1996). The findings of this study support that ultrasound imaging can be used to
measure severity of tendon pathology (abnormal composition and shape) in persons
with IAT.
37
One limitation to the measurement of tendon composition in the study by Astrom
et al. (1996) was that it was graded on a dichotomous scale (normal or abnormal).
Having a dichotomous measure of tendon composition limits the ability to examine for a
potential continuous relationship with prognosis or mechanical tendon properties. One
method of attaining a continuous measure of abnormality on ultrasound is the mean
grayscale value, which quantifies echogenicity (Collinger, Fullerton, Impink, Koontz, &
Boninger, 2010; Collinger, Gagnon, Jacobson, Impink, & Boninger, 2009). Imaging
software can be used to compute a mean grayscale value of all the pixels in a defined
region of interest. The grayscale values for each pixel ranges from 0 (black or
hypoechoic) to 255 (white or hyperechoic). On ultrasound a healthy tendon will typically
appear as alternating light and dark bands due to the reflection pattern from hyperechoic
collagen fibers (Figure 9a,c). With tendinopathy the tendon has altered collagen
alignment and areas of hypoechogenicity (Figure 9b,d). Comparing the mean grayscale
value (echogenicity) of a healthy tendon (Figure 9e) to a tendon with IAT (Figure 9f), the
tendon with pathology has a lower mean grayscale value. This lower mean grayscale
value in the pathological tendon may be due to the loss of the hyperechoic collagen
fibers, an altered organization of collagen fibers, and/or increased ground substance and
lipids. Measuring echogenicity has not been validated in persons with achilles
tendinopathy, yet a previous study demonstrated that individuals with rotator cuff
tendinopathy had a lower mean grayscale value than healthy controls (Collinger et al.,
2010). This method also appears to have moderate to good intra-rater reliability and a
standard error of measurement (SEM) of 7.3 to 10.6 for two examiners assessing the
echogenicity for the biceps and supraspinatus tendons (Collinger et al., 2009).
38
Figure 9 . Tendon characteristics for uninvolved (top row) and involved (bottom row) sides of a participant with IAT: a-b) Longitudinal ultrasound images of tendon diameter; c-d ) Cross-sectional ultrasound images of echogenicity measured by the mean grayscale value; e-f) Corresponding histograms demonstrating the mean grayscale value (scale from 0=black to 255=white) for images c and d.
In the current study, tendon characteristics (diameter, mean grayscale value)
were measured by obtaining longitudinal (Figure 1a-b) and cross-sectional images
(Figure 1c-d) with the subject in prone and plantar surface of the foot perpendicular to
the floor. A Phillips HD11 Digital Ultrasound machine with a linear array probe (L12-3)
was used at a frequency of 10-12 MHz to capture tendon characteristics. All ultrasound
images were stored digitally and processed using Image J (http://rsbweb.nih.gov/).
Ultrasound measures were completed by the primary author, whose technique was
based on information from discussion with researchers, clinicians and self-study
(literature, online lectures (http://sonoworld.com/)).
On a longitudinal view, tendon diameter was measured at the widest point within
2 cm of the tendon insertion (Figure 9a-b). Abnormal composition was quantified as the
mean grayscale value for the central 1 cm of the tendon on the cross-sectional image
(Figure 9c-d). All ultrasound images had the same gain setting of 56 with a 3 cm depth,
39
so that differences in ultrasound settings would not influence mean grayscale values
between subjects. Abnormal composition was indicated by a lower mean grayscale
value compared to controls (Figure 9 e-f). The mean grayscale value of the histogram
was calculated by the Image J program for the region of interest selected.
Preliminary studies. Based on pilot work, ultrasound imaging can be used to
detect differences in tendon pathology between the involved and uninvolved sides in
participants with IAT. The difference in echogenicity between sides (mean difference=
15.2) was more than double the variance of the involved side (SD= 6.8) (Table 5). Also,
the diameter measured in pilot work was similar to published data on tendon
characteristics of individuals with achilles tendinopathy (midportion and insertional) and
healthy controls (Astrom et al., 1996) (Table 5).
Table 5. Tendon pathology on the involved and uninvolved side of persons with achilles tendinopathy in pilot work and a published study Involved Uninvolved Mean grayscale value (Echogenicity) a 64.5± 6.8 79.7± 3.3 Diameter with ultrasound, pilot work b 8.0± 1.8 mm 5.3± 1.3 mm Diameter with ultrasound, Astrom study c 8.0± 3.0 mm 6.0± 1.0 mm Diameter with MRI, Astrom study c 9.0± 2.0 mm 6.0± 1.0 mm a Findings from pilot data with 3 individuals with unilateral IAT and 1 person with bilateral involvement. A lower value indicates a greater amount of pathology. b Findings are from pilot data with 5 individuals with unilateral IAT and 2 individuals with bilateral involvement c Astrom study included 23 individuals with midportion and insertional achilles tendinopathy and 14 healthy controls (Astrom et al., 1996)
Tendon mechanical properties.
Data collection. Ultrasound imaging is used to measure in vivo tendon
mechanical properties, i.e. strain and stiffness, by tracking the displacement of the
gastrocnemius muscle-achilles tendon junction (Arya & Kulig, 2010; Child et al., 2010).
This methodology has several advantages, including that it: captures mechanical
function of the tendon within the naturally surrounding tissue system; is non-invasive;
40
and has no known side effects (Ter Haar, 2011). However, a disadvantage is the
difficulty of isolating the mechanical properties of one tissue structure, such as the distal
2 cm of the achilles tendon where most pathology occurs in persons with IAT. One
method of circumventing this limitation is inserting a metal needle into the free tendon to
track the displacement of a particular point within the achilles tendon (Magnusson et al.,
2003). Using this method Magnusson et al. (2003) found that a greater amount of tendon
elongation occurred along the “free tendon” (5.9± 0.9 mm) than “aponeurosis” portion
(2.1± 0.6 mm) [Figure 10]. However, the risk of this procedure may outweigh the
scientific value of this invasive measure in a population with achilles tendinopathy
symptoms. Other methods using ultrasound imaging, such as sonoelastography, are
currently being developed and validated (De Zordo et al., 2010; Klauser, Faschingbauer,
& Jaschke, 2010; Sconfienza et al., 2010), and could be used in the future to capture
localized strain patterns in patients with IAT. Nevertheless, while differential strain
patterns likely exist along the achilles tendon complex, the tendon functions as a unit
during functional activities, which require DF ROM and PF strength. Therefore,
examining the achilles tendon complex, hereafter referred to simply as the achilles
tendon, best fits the purpose of the current study to assess the relationship between
tendon characteristics, DF ROM and PF strength in persons with IAT.
41
Figure 10. The achilles tendon complex consists of the aponeurosis, covering the soleus muscle, and the free tendon
Several different methods have been used to assess achilles tendon elongation
and strain (elongation normalized to resting tendon length) with ultrasound imaging. In
general these methods can be divided into passive and active measures. Interestingly,
despite differences in measurement condition, both passive and active measures appear
to require a similar amount of tendon elongation (active= 6 to 11 mm elongation, passive
= 8 to 10 mm elongation) (Table 6). Thus, passive and active methods may reflect
similar mechanical properties (Theis, Mohagheghi, & Korff, 2012). For passive
conditions the ankle is rotated from PF to DF, while an ultrasound probe placed on the
gastrocnemius muscle-tendon junction is used to measure tendon elongation over a
specified range of motion. An advantage of the passive method is the ability to capture
multiple measurements throughout the range of motion, which allows for a description of
the curvilinear relationship between tendon elongation and passive torque. However, the
disadvantage is that this measure alone requires about 40 minutes to complete (~20
minutes per side during pilot work). In contrast, active measures examine tendon
42
elongation at the gastrocnemius muscle-tendon junction and force generation during an
isometric contraction. Similar to the passive method, force is divided by the tendon
elongation/strain in order to calculate stiffness. The advantage of an active measure is
that it is much quicker because only 2 measurements of the tendon are needed: rest and
contraction. The disadvantages are that it is unable to capture the curvilinear properties
of the tendon and that viscoelastic properties of the tissues are less controlled because
the rate of contraction varies by subject. In addition, pilot work suggests that individuals
with IAT have decreased isometric PF strength compared to controls. Since tendon
elongation linearly increases with PF torque, a comparison of tendon elongation
between participants with IAT and controls would be confounded by differences in
strength (Figure 11). For this reason, the current study examined tendon mechanics
under a passive condition.
43
Table 6. Achilles tendon elongation and strain in healthy adults
Study Sample Testing
Condition Achilles Tendon
N % M Age (y)a Elongation (mm)a
Strain (%)a
(Kawakami et al., 2008)
12 50 19±38 Passive from 30ºPF to 30º DF in 10º increments
10ºDF= 8± 3
10ºDF= 4.5± 1.6
(Muraoka, Muramatsu, Takeshita, Kawakami, & Fukunaga, 2002)
8 100 26±3 Passive from 36ºPF to 7º DF
7ºDF= 10±~2
7ºDF= 5.9±~1
(Arya & Kulig, 2010)
12 100 45±7 100% Isometric MVC in 0ºDF
11.0± 0.9 4.36± 0.31
(Child et al., 2010)
16 100 35± 9 100% Isometric MVC in 0ºDF
NA 3.4±1.8
(Mahieu. et al., 2004)
21 43 31 100% Isometric MVC in 0ºDF
6.1± 2 NA
(Magnusson et al., 2003)
5 100 32±5 100% Isometric MVC in 0ºDF
8.0± 1.1 NA
(Muramatsu et al., 2001)
7 100 26±3 90% Isometric MVC in 0ºDF
~9± 1 (Fig 4)
~5±1 (Fig 5)
a Data presented as mean± standard deviation when available
Figure 11 . Tendon Elongation from rest to maximum Isometric PF torque at 0⁰DF (Chimenti, Tome, Flemister, & Houck, 2012)
44
Another way to describe mechanical properties is to examine the rate at which a
tendon elongates in relation to the force through the tendon. This relationship can be
depicted as a force-deformation curve (Figure 12). From this curve the rate of change, or
slope, is used to describe the tendon stiffness. A greater stiffness, i.e. increased slope,
indicates that a greater amount of force is needed per unit of tendon deformation, i.e.
elongation. It is difficult to compare values of achilles tendon stiffness between studies
because the calculation often varies depending on if/how the author normalized the
measure. For example, Kawakami et al. (2008) chose to normalize tendon elongation to
the resting length of the tendon, while Theis, Mohagheghi & Korff (2012) did not
normalize to elongation. In order to compare preliminary studies to a similar technique
used by Kawakami et al, stiffness for preliminary data is reported with the units Nm/%.
However, data in the results are reported with the units of N/mm to better align with
common scientific practice.
45
Figure 12. Adaptation of figures from Maganaris and Rees articles (Maganaris, Narici, & Maffulli, 2008; Rees et al., 2006)
Different portions of the force-deformation curve reflect different passive
mechanical properties of the tendon (Figure 12) (Maganaris et al., 2008; Rees et al.,
2006). As the tendon is initially stretched from its resting length, there is a “toe region”
during which there is a curvilinear increase in force per unit deformation. This initial
curvilinear relationship is due to a natural crimp in the tendon, so there is a relatively
large amount of tendon elongation in comparison to the amount of force. As elongation
increases beyond the “toe region” the relationship between force and elongation
becomes more linear. Finally, when the loading approaches tendon failure then the
relationship again returns to curvilinear as the tendon structure begins to rupture
(Maganaris et al., 2008; Rees et al., 2006). The Kawakami method captures the toe and
linear regions of the force-deformation curve, and then uses the coefficient of the linear
46
portion of the curve to approximate tendon stiffness (Kawakami et al., 2008). However,
the “toe region,” where crimp in the tendon is designed to minimize force on the tendon
with small motions, is not of interest to the current study. Functional activities require the
tendon to operate at higher forces and deformations, and thus mechanical properties at
the “linear” portion of the curve were examined.
The combination of ultrasound imaging and isometric dynamometry were used to
measure tendon mechanical properties (strain, stiffness). Each participant’s ankle was
positioned and secured in Biodex System 4 (Biodex Medical Systems, Shirley, NY). The
participant was positioned in sitting with the seatback reclined to 45⁰, the chair height
adjusted so that the tibia was parallel to the floor and the hip in neutral adduction/
abduction (Figure 13a). The participant’s ankle was then rotated from 10⁰ PF to 10⁰ DF
at a rate of 5º/s and the passive PF torque recorded by the Biodex. Ultrasound imaging
was used to track the linear displacement of the gastrocnemius muscle-achilles tendon
junction with the probe at a fixed point on the calf (Figure 13 b-c) (Arya & Kulig, 2010;
Child et al., 2010; Kawakami et al., 2008). A first set of 6 rotations was used for pre-
conditioning, followed by a second set of 6 rotations from which images were used to
calculate tendon mechanical properties. To ensure that the condition was passive,
electromyography of the medial gastrocnemius and the anterior tibialis muscles was
used to monitor muscle activity.
In order to calculate tendon strain, tendon elongation was first calculated by
subtracting the displacement of the musculotendinous junction, recorded using
ultrasound imaging, from the total elongation of the muscle-tendon unit (Formula 1).
Because the knee joint was immobile, total elongation of the muscle tendon unit was
estimated from the change in ankle angle. Previous studies validated a similar approach
47
(Grieve, Cavanagh, & Pheasant, 1978; Kawakami et al., 2008). In addition, the muscle
tendon unit elongation also depended on the position of the heel cup on the isokinetic
dynamometer, which had three adjustments. Depending on the size of the foot, the heel
cup was adjusted to approximate the participant’s ankle axis of rotation with that of the
isokinetic dynamometer. For each adjustment, the linear translation of the achilles
insertion was established using 3D motion capture (N=3). Total muscle tendon unit
elongation for setting 1= 16.2 mm± 1.9, for setting 2= 18.7 mm± 1.1, and setting 3= 19.5
mm± 1.5.
(1) Tendon elongation= Total muscle-tendon unit elongation – Muscle elongation
Strain was calculated to take into account differences in anthropometrics with
tendon elongation normalized by each individual’s resting tendon resting length (Formula
2). Two points were used to define the resting length of the tendon. First the tendon
insertion was found using ultrasound, then a mark placed on the skin. Second the medial
gastrocnemius muscle-tendon junction was found using ultrasound and second mark
placed on the skin. The distance between these two points, when the participant was in
a position with the knee straight and the ankle in 0⁰ DF, was measured and used as the
resting tendon length.
(2) Tendon strain= _Elongation_ Resting length
To examine the resistance to stretch during the passive ankle rotation, tendon
force was calculated. To derive achilles tendon force, motion analysis combined with
ultrasound imaging was used to estimate achilles tendon moment arm. In a neutral
standing position, for each participant the medial and lateral malleolus were digitized and
the center between these two points was defined as the ankle joint center. A point was
also digitized on posterior aspect of the heel. The moment arm was calculated by the
48
anterior-posterior distance between the ankle joint center and the digitized heel point
(measured with motion capture) minus the thickness of the skin, subcutaneous tissue
and ½ the tendon diameter (measured with ultrasound imaging). Passive tendon force
was calculated by dividing passive torque (measured by the dynamometer during ankle
rotation) by the moment arm of the achilles tendon (4). To examine both the magnitude
of elongation and the resistance to stretch, tendon stiffness was calculated by dividing
the change in passive tendon force during ankle rotation by the change in elongation
(Formula 4).
(3) Tendon force= __Passive torque of ankle___ Moment arm of achilles tendon
(4) Tendon stiffness= __Tendon force__ Tendon elongation
Figure 13 . a) Configuration of ultrasound imaging, isokinetic dynamometer and electromyography during measurement of tendon mechanical properties. Ultrasound images of the gastrocnemius muscle- achilles tendon junction (MTJ) displacement as the ankle is rotated from b) 10⁰ plantar flexion to c) 10⁰ dorsiflexion.
Preliminary studies. The focus of pilot work was to examine the feasibility of
novel ultrasound imaging measures. In preliminary work, tendon mechanical properties
49
were examined using methods similar to those used by other researchers (Arya & Kulig,
2010; Child et al., 2010; Kawakami et al., 2008). Kawakami et al. (2008) measured
mechanical properties through the majority of ankle range of motion (30⁰ PF to
maximum DF) to describe both the toe and linear regions of the force-strain curve
(similar to the entire polynomial curve in Figure 14). Tendon stiffness was approximated
using the coefficient of the linear portion of the curve in healthy adults (Kawakami et al.,
2008).
An abbreviated method was developed to only assess tendon mechanical
properties in the linear region. This was based on the rationale that individuals often
report symptoms toward the end-range of ankle DF. Therefore, mechanical properties
towards the end range of ankle DF, which coincides with the “linear” portion of the curve,
was of interest to the current study. An abbreviated two-point method was developed to
capture mechanical properties in the linear region of the torque-strain curve (Lines
connecting 2 points along curves in Figure 14). This abbreviated methodology was
initially compared against the approach described Kawakami et al (Kawakami et al.,
2008). Converting the data into the same units for stiffness reported by Kawakami et al
(Nm/%), the average tendon stiffness of young healthy controls (N= 6, stiffness= 3.8 ±
1.7 Nm/%) was within 1 SD of the mean reported by Kawakami et al. (2008) (3.4±2.5
Nm/%). Given the similar estimation of tendon stiffness between these two methods, the
current study measured tendon mechanical properties only in the linear region of the
force-elongation curve (Figure 14).
50
Figure 14. Force-elongation curve from 30⁰ planar flexion (PF) to maximum dorsiflexion (DF) for a healthy control (max DF=25⁰), the involved and uninvolved sides of an IAT subject (max DF=15⁰). The points on each curve correspond to force and strain at 10º DF and 10º PF. Stiffness is the slope of the corresponding line highlighted in bold on the linear equation.
Preliminary findings support hypotheses that the involved side of individuals with
IAT compared to the uninvolved and controls demonstrate increased tendon strain and
decreased stiffness (Table 7). The stiffness of the involved side of IAT subjects was
nearly half the value reported by Kawakami et al. (2008) for healthy adults (3.4±2.5
Nm/%) , and is >1 SD below the control group mean of the current study. Test-retest
reliability for strain (N= 6, ICC= 0.946, SEM= 0.3 %) and stiffness (N= 6, ICC= 0.998,
SEM= 0.7 N/mm) in controls was acceptable.
51
Table 7 . Tendon strain in participants with insertional achilles tendinopathy (IAT) and controls Involved side, IAT
n= 9 sides a Uninvolved side, IAT n= 5 sides a
Controls n= 6 sides b
Strain (%) 5.2 ± 1.1 4.1± 1.6 3.1± 1.2 Stiffness (Nm/%) 1.8± 0.6 2.6± 0.5 3.8±1.7
a5 individuals with unilateral and 2 individuals with bilateral IAT (Mean age= 56.6± 16.7 years, 3M and 4F) b3 healthy controls (Mean age= 22.7± 0.6 years, 1M and 2F)
Dorsiflexion range of motion capacity.
Data collection. While there are several different methods of measuring DF
ROM, the current study used methods that are, theoretically, most likely to detect
impairment in persons with IAT. Measurement of DF ROM can be performed with the
knee flexed, where the gastrocnemius muscle-tendon unit is on slack when the ankle is
dorsiflexed, or with the knee straight, where the gastrocnemius muscle-tendon unit is
stretched when the ankle is dorsiflexed. While measures of DF ROM are often
interpreted as examining gastrocnemius (knee straight) or soleus (knee bent) muscle
tightness, the measures are not specific to the muscle. Resistance to DF ROM includes
structures around the ankle joint, including the joint capsule (cartilage around the
talocrural joint); tendinous tissues (tendon and aponeurosis); epi-, peri-, and
endomysium (connective tissue around entire muscle, fascicles, and fibers); sarcolemma
(cell membrane of muscle fibers) and endosarcomeric structures (titin) (Muir, Chesworth,
& Vandervoort, 1999; Muramatsu et al., 2001). While manipulation of the knee joint
angle may bias the measure toward including or excluding the gastrocnemius muscle,
clinical measures of DF ROM are not specific to one structure. Because “isolated
gastrocnemius tightness” is believed to contribute to achilles tendinopathy (DiGiovanni et
al., 2002), DF ROM was measured with the knee straight for the current study.
52
For clinical relevance, procedures for measuring DF ROM in the current study
were based on the two common types of DF ROM measures used in the clinic: 1) non-
weight-bearing, and 2) weight-bearing (Table 8). A non-weight-bearing maximum angle
is traditionally one of the most common orthopaedic measures of DF ROM (Hoppenfeld
& Hutton, 1976). Clinically, the ankle is dorsiflexed to the maximum tolerated stretch,
and a goniometer is used to measure DF (angle between the tibia and foot). In research,
the amount of force used to dorsiflex the ankle is standardized in order to minimize the
influence of individual differences in maximal stretch tolerance. A study by Moseley,
Crosbie, and Adams (2001) examined passive ankle dorsiflexion in 300 healthy men and
women between 15 and 34 years old in order to establish normative values for ankle
hypo- and hypermobility. Using a standardized 12 Nm of torque, examiners passively
dorsiflexed the ankle. With the knee extended, DF ROM was normally distributed
(mean= 18.1⁰, SD= 6.9⁰); hypomobility was defined as <4.3⁰ DF and hypermobility as
>31.9⁰ DF (Moseley, Crosbie, & Adams, 2001).
Table 8 . Two common types of clinical dorsiflexion capacity measures Type Method Advantages Disadvantages Non-weight-bearing
DF angle at a specified PF torque in a non-weight-bearing position
Established normative values, good specificity, consistent with clinical exam findings of foot pathology a
Fair sensitivity, variable (fair to excellent) reliability b
Weight-bearing
Maximum DF angle in a weight-bearing position
Good to excellent reliability, face validity c
Lacks normative values and validity studies; depends on stretch tolerance
a (DiGiovanni et al., 2002) b (Kim et al., 2011) c (Munteanu, Strawhorn, Landorf, Bird, & Murley, 2009)
53
This definition of hypomobility is consistent with the definition by Digiovanni et al.
(2002) in a case-control study examining DF in persons with forefoot and midfoot pain
compared to healthy adults. The sample included 68 individuals, 34 patients and 34
controls, between 21 and 76 years old (DiGiovanni et al., 2002). The torque of 10 Nm,
which is similar to the amount of torque chosen by Moseley, Crosbie and Adams (2001),
was justified as a typical amount of force used to examine DF ROM by two orthopaedists
and a rehabilitation medicine specialist. DiGiovanni et al. (2002) defined hypomobility as
<5⁰ DF for two reasons: 1) good specificity, and 2) consistent results with clinical exam
in controls. This definition of hypomobility had 76% specificity; in other words, among
controls only 8/34 tested positive for an isolated gastrocnemius contracture indicating a
low false positive rate. Classification of hypomobility based on the standardized 10 Nm
torque was compared with results of a common clinical test (i.e. Silfverskiold) defining
hypomobility as decreased DF ROM when the knee is straight compared to when the
knee is bent. Comparing this clinical test to the 10 Nm standardized test, the results
were the same for 94% (32/34) of the controls (DiGiovanni et al., 2002). In summary, the
validity of defining hypomobility as <5º DF with the knee extended is supported by the
distribution of normative data, good specificity, and consistency with clinical exam
(DiGiovanni et al., 2002; Moseley et al., 2001).
However, there are limitations to the non-weight-bearing DF ROM measure in a
population with pathology. In the DiGiovanni et al. (2002) study, the <5⁰ DF criteria only
had fair sensitivity, with 65% (22/34) of participants having a positive test, and a lower
consistency of 76% (26/34) with clinical test in the patient group. Therefore, the test has
a high false negative rate in participants with foot pain.
54
A method that may have greater reliability than a non-weight-bearing measure is a
weight-bearing measure using an inclinometer. The non-weight-bearing measure
requires clinical proficiency in terms of consistently aligning the goniometer with bony
landmarks and joint center. In contrast the weight-bearing measure requires less skill,
given that an examiner only has to measure the difference in tibial inclination from
standing to maximum dorsiflexion stretch. The simplicity of the inclinometer measure
may contribute to greater reliability using an inclinometer (intrarater ICC=0.85 to 0.91,
interrater ICC= 0.92 to 0.95) compared to a goniometer (intrarater ICC=0.74 to 0.84,
interrater ICC= 0.34 to 0.40) (Kim et al., 2011; Munteanu et al., 2009). Because
functional tasks, such as walking and stairs, are performed in weight-bearing, the
weight-bearing static measure also has face validity. However, because this is a
relatively new method of assessing DF ROM, studies have focused on establishing
reliability of this technique rather than validity and normative values. Because the non-
weight-bearing measure has relatively high validity while the weight-bearing measure
has relatively high reliability, it is unclear which technique would be more sensitive to
detecting DF ROM impairment in persons with IAT. Therefore, DF ROM was measured
both in non-weight-bearing and weight-bearing in the current study.
For research-level accuracy, DF ROM was defined as calcaneal rotation in the
sagittal plane relative to the tibia with a three dimensional motion capture system. Non-
weight-bearing DF capacity: Subject was in supine position with the knee straight. The
examiner applied 12 Nm of torque to the plantar surface of the metatarsal heads using a
customized instrumentation system (Moseley et al., 2001). A platform, with an imbedded
50 lb load cell, output the voltage associated with the force applied to the metatarsal
heads [Force= (-9.91*Voltage)+37.19, equation developed from calibration trial with
55
known weight]. The applied force was synchronized with the kinematic data through
Innsport Training Software (Innsport Training, Chicago, IL). In order to keep the torque
consistent between subjects despite differences in foot size, the distance from the ankle
joint center to the midpoint of first and fifth metatarsal heads defined the moment arm of
the applied force for each individual. Weight-bearing DF capacity: Participants were
asked to align their second toe and heel along a line. The participant was then asked to
bring one foot forward, similar to a lunge, and lean forward until maximum tolerated DF
ROM. A chair was used for balance as needed as a one second interval of kinematic
data were collected in this position.
Preliminary studies. Based on pilot work, different measures may provide
contradictory information about DF ROM in persons with IAT. For this reason,
procedures in the current study differed from pilot work, in that the current study
measured DF ROM (calcaneus relative to the tibia) within the motion capture system.
For example, in pilot work non-weight-bearing DF capacity was similar across groups,
while the weight-bearing measure indicated a potential limitation in DF capacity in
persons with IAT (Table 9)(R. Chimenti, J. Tome, A. Flemister, & J. Houck, 2014). One
potential reason for these differences may be the greater collapse of the midfoot in a
weight-bearing position. When DF is defined as the angle of the foot relative to the tibia,
then motion from both a collapse of the midfoot and calcaneal rotation contribute to the
angle. Weight-bearing measures may have a greater potential for collapse of the midfoot
than a non-weight-bearing measure due to a greater force being put through the foot due
to body weight than a manual pressure. The idea that greater DF ROM during weight-
bearing measures is due to increased midfoot motion rather than rearfoot motion was
tested by Chizewski and Chiu (2012). In this study, 48 subjects performed a partial squat
56
until they reached their maximum ankle DF. On average, the maximum DF capacity,
defined as the calcaneal rotation in the sagittal plane relative to the tibia, was 19º (range
12º to 29º) (Chizewski & Chiu, 2012). This value is much lower than the value reported
by studies examining weight-bearing DF capacity with the knee bent which range from
38.8±5.2 (Konor, Morton, Eckerson, & Grindstaff, 2012) to 50.4±8.1 (Bennell et al.,
1998). Because the primary interest in the current study is DF capacity related to
elongation of the gastrocnemius muscle-tendon unit, all measures of DF ROM were
taken using the motion capture system and are defined as calcaneal rotation in relation
to the tibia.
Based on pilot work, it appears that the weight-bearing measure is more
sensitive to limitation in DF capacity in participants with IAT than the non-weight-bearing
measure. It should be noted that pilot data on DF capacity included motion from the
forefoot, and thus may have less accuracy than the proposed methods for the current
study (Table 9). The current study measured non-weight-bearing DF capacity due to its
prevalent use clinically and the potential inaccuracy in pilot work due to instability of the
midfoot. However, findings from the weight-bearing measure were more influential in
guiding current hypotheses.
Table 9. Measures of dorsiflexion (DF) capacity in participants with insertional achilles tendinopathy (IAT) and controls DF capacity Involved side, IAT
(n= 9 sides) Uninvolved side, IAT (n= 4 sides)
Controls (n= 4 sides)
Non-weight-bearing (⁰) 15.0± 6.2 15.2± 4.0 17.0± 7.1 b Weight-bearing (º)a 24.6± 8.5 27.0± 3.6 38.7± 5.0 c
a Weight-bearing DF ROM was not collected on all pilot subjects and data in this row is based on 4 individuals (3 M, 1F) with unilateral IAT, one person with bilateral involvement, aged 62.5± 9.5 years b 4 healthy adults (1 M, 3F) aged 26.8± 8.2 years c 4 healthy adults (2M, 2F) aged 41± 14 years
57
Plantar flexion strength capacity.
Data collection. In order to test if PF weakness is associated with IAT, a testing
protocol to measure maximum PF strength was modified for this patient population. A
key factor influencing muscle performance is the muscle-tendon unit length. In a study of
28 young healthy males and females, maximum isometric PF strength occurred when
the knee was extended and the ankle was in 15º DF (Winter & Challis, 2008). For the
gastrocnemius and soleus, muscle force increases as the length of the muscle-tendon
unit increases (Winter & Challis, 2008). Physiologically, this relates to an ideal muscle
cell length, in which actin-myosin overlap is optimized for force production (Winter &
Challis, 2008). Therefore, measurement of maximum PF force would ideally be
performed with the muscle-tendon unit in a lengthened position. However, study
participants can have limited DF capacity or this range of motion of may be painful. For
this reason, maximum PF strength was tested at a neutral ankle position of 0º DF to
maximize participant comfort. Therefore, the current study did not capture maximum PF
strength due to an inability to have all subjects achieve 15⁰ DF with the knee extended.
However, a position of 0⁰ DF is close to the ideal position and was achievable for all
participants.
For the measurement of PF strength capacity, participants were instructed to
press on the foot plate of the isokinetic dynamometer as if they were pressing on the gas
pedal. Participants had one practice repetition. The average of the maximum isometric
PF torque achieved on the following 3 repetitions (5 second contraction, 30 second rest
break) was used to define PF strength capacity. Testing always started with the
uninvolved side for participants with IAT, so that the practice repetition could be reduced
58
or skipped on the involved side if the subject reported pain >4/10 on the NRS scale. For
controls, the first side tested was randomly chosen.
Preliminary studies. While the maximum number of single limb heel raises is
one of the most common clinical measures of PF strength (Kendall, Wadsworth, Kendall,
& McCreary, 1983), this measure was not sensitive to impairment in persons with IAT.
Three out of the first three pilot subjects with IAT were unable to perform a single
unilateral heel raise on the involved side (unpublished pilot data). Therefore, using the
single limb heel raise test in the current study would likely result in a floor effect and
diminish the ability to find associations with other impairments.
The current study measured PF strength with an isometric test in 0⁰ DF.
However, the measured strength capacity may be limited by either weakness or pain.
Out of the 8 pilot subjects with pain ratings during isometric PF strength, half (4/8) had
pain when testing the involved side (Table 10). Pain may confound the ability to measure
a true maximum PF strength, because PF strength capacity may be limited by pain
rather than weakness. Considering participants with unilateral involvement (6/8), the
involved side always had a lower maximum isometric PF strength than the uninvolved
side. If there was a direct relationship between pain and PF strength performance, then
one would assume that the difference between the involved and uninvolved sides would
directly increase with pain rating. Examining Table 10 this pattern was not evident from
the small pilot sample. For example, both subjects 053012 and 072012 had a 30 Nm
difference in maximum PF strength, but subject 053012 had no pain and the latter
reported the highest pain rating from this sample. While pain cannot be eliminated as a
confounding factor from this measure, the difference between the involved and
uninvolved sides suggests that pathology is also a key factor influencing strength.
59
Table 10 . Maximum isometric plantar flexion torque (Nm) and associated pain in participants with insertional achilles tendinopathy. Subject Pain Involved Uninvolved Difference 090512* 0/10 99.6 103.1 3.5 090612* 0/10 121.6 128.5 6.9 091112 0/10 74.6 85.9 11.3 053012 0/10 90 120 30 072612 4/10 60 100 40 092812 4/10 98.1 105.8 7.7 082312 5/10 46.7 52.2 5.5 072012 6/10 70 100 30 * Subjects had bilateral involvement and the weaker side is the “Involved” side.
Range of motion and strength performance.
Data collection. Participants practiced a few trials climbing the stairs until they
felt comfortable walking while instrumented with the IREDs. A metronome was used to
cue subjects to walk at the same speed. The participants stepped up onto a stabilized
block, which had a height comparable to a standard 17 cm step height. This task was
repeated at least three times for each side. Movement of the foot and ankle was
recorded with a video camera and motion capture system.
Several variables were calculated to quantify DF ROM and PF strength performance.
The percentage of end-range DF used during stair ascent was calculated by normalizing
the amount of DF used during stair ascent to each individual’s capacity for weight-
bearing DF (Formula 5). Ankle moment was calculated by the motion monitor software
by multiplying the ground reaction force (body mass * acceleration) by the moment arm
(distance from the center of pressure to the ankle joint center) (Formula 6). Ankle power
was calculated as the ankle joint moment multiplied by the calcaneal velocity in the
sagittal plane (angular rotation per second), and normalized to body weight (Formula 7).
(5) Percentage of end-range DF = Stair ascent DF * 100 used during stair ascent DF capacity
(6) Ankle moment= Ground reaction force * Moment arm
60
(7) Ankle power= Ankle moment * Calcaneal velocity Body weight
Stair negotiation speed was based on a study by Oh-Park (2011) examining
normative values associated with the stairs in healthy adults (n=513, mean
age=80.8±5.1 years). On average, participants ascended 3 steps in 2.36± 0.2 seconds,
which corresponds to a cadence range of 96 to 116 (Oh-Park, Wang, & Verghese,
2011). In order to accommodate individuals with IAT that may have difficulty with stairs,
a cadence at the lower end of this normative range was used for the current study. The
cadence of 100 bpm was also used by another researcher examining the effect of ankle
flexibility on stair descent kinematics and kinetics in a group of persons aged 15 to 34
years old (Moseley et al., 2003). Therefore, this cadence for stair ascent should be
within the normative range for participants in the current study.
For the purpose of analysis, stair ascent was divided into 3 phases as defined by
McFadyen and Winter (1998). Each phase is illustrated below by a control participant
performing the transition to stair ascent on the left leg (Figure 15) (McFadyen & Winter,
1988). The first phase is weight acceptance during which body weight is being
transferred from the right leg, which commences swing phase, to the left leg, which
supports body weight (Figure 15a). The second phase is pull-up of the right leg to the
step (Figure 15b). During pull-up the left ankle moves into DF as the right leg lifts toward
the step. The third phase is forward continuance during which the center of mass is
being moved forward and upward as the left foot pushes off of the ground and the right
leg accepts body weight (Figure 15c). Successful completion of forward continuance
depends on adequate ankle power on the left side to push-off against the ground. The
61
swing phase was not analyzed, since during this time there are minimal forces on the
achilles tendon.
Figure 15 . The three phases of stair ascent include, a) weight acceptance, b) pull-up, c) forward continuance. An arrow indicates the limb side for analysis.
Preliminary studies. The task of stair ascent was of particular interest to the
current study because it simultaneously challenges DF ROM and PF strength when the
knee is relatively straight. This knee position is relevant because it was hypothesized
that individuals with IAT had isolated gastrocnemius tightness. Based on pilot work,
limitation in weight-bearing DF capacity is more likely to be present in persons with IAT
when the knee is straight rather than bent (Table 11). Therefore the task of stair ascent,
rather than descent, was chosen to examine potential impairments in DF performance in
participants with IAT. It was feasible to collect DF during stair ascent (Figure 16), and
these findings were used to guide hypotheses and estimates of effect size for statistical
analysis.
Table 11 . Weight-bearing dorsiflexion (DF) capacity in participants with insertional achilles tendinopathy (IAT), controls and healthy adults DF in weight-bearing (º)
Involved side, IAT (n= 5 sides) a
Uninvolved side, IAT (n= 3 sides)
Controls (n= 3 sides) b
Healthy adults
Knee bent 36.5± 15.8 42.5± 9.2 39.8± 5.5 38.8± 5.2 b Knee straight 24.6± 8.5 27.0± 3.6 34.8± 4.7 38.7± 5 .0c
a4 individuals (3 M, 1F) with unilateral IAT, one person with bilateral involvement, aged 62.5± 9.5 years b4 healthy adults (2M, 2F) aged 41± 14 years b20 individuals (7 M, 13 F) aged 24±3 years (Konor et al., 2012) c30 individuals (10 M, 20 F) aged 22.1±5.6 years (Munteanu et al., 2009)
62
Figure 16 . Stair ascent ankle motion (dorsiflexion is positive) of: a) involved side of participants with insertional achilles tendinopathy (IAT), and b) healthy controls.
The task of stair ascent was also used to examine functional PF strength. Stair
ascent is a challenging functional task for the plantar flexor muscle-tendon unit. As
previously described, pilot subjects had difficulty performing a single limb heel raise on
the involved side, and stair ascent is functionally similar to a single limb heel raise. If
someone has difficulty using the plantar flexors to push off the ground, then an
alternative movement strategy is needed to complete the task. For example, a “pull-off”
strategy is when the hip and knee extensors of the uninvolved side are used to pull the
other side up, rather than depending on the plantar flexor muscles of the involved side to
push off the ground. Preliminary work suggests that individuals with IAT may use a “pull-
off” strategy during stair ascent. This can be seen by examining the lower ankle power
on the involved side of individuals with IAT compared to healthy controls (Figure 17,
Table 12). This pilot work represents the current best estimate of functional PF strength
in persons with IAT.
a b
IAT Controls
63
Figure 17 . Ankle power during stair ascent performed by a) the involved side of individuals with insertional achilles tendinopathy (IAT), and b) healthy controls Table 12 . Plantar flexion strength in participants with insertional achilles tendinopathy (IAT) and controls Involved side, IAT Controls
(n= 4 sides) Ankle Power (Watts/Kg body weight) 2.89 ± 1.00 a 4.21 ± 0.92 Isometric PF (Nm) 79.95± 28.67 b 110.00± 12.25
a 4 participants with unilateral IAT and 1 participant with bilateral involvement b 6 participants with unilateral IAT and 2 participant with bilateral involvement Statistical Analyses
Specific Aim 1. This aim examined if tendon pathology (echogenicity and
diameter) and mechanical properties (strain and stiffness) differs between sides with IAT
(involved side of participants with IAT, n=20) and sides without IAT (uninvolved side of
participants with IAT, n=20 and both sides of controls, n=40 sides). Additionally, these
dependent variables are compared between sides (involved vs. uninvolved in the IAT
group) and between groups (IAT vs. controls).
Hypotheses.
H1a. Sides with IAT have a lower echogenicity than sides without IAT in both case and
control groups.
a b IAT Controls
64
H1b. Sides with IAT have a greater diameter than sides without IAT in both case and
control groups.
H1c. Sides with IAT have a greater strain than sides without IAT in both case and control
groups.
H1d. Sides with IAT have a lower stiffness than sides without IAT in both case and
control groups.
Statistical analysis. Comparisons between sides with IAT and sides without IAT
were tested using generalized estimating equations (GEE). A GEE analysis uses the
generalized linear model, which can account for the correlation between sides within an
individual. This analysis allowed for inclusion of all data (both sides of all participants),
and uniquely identified differences between sides with IAT (involved side of participants
with IAT, n=20) and sides without IAT (uninvolved side of participants with IAT, n=20 and
both sides of controls, n=40 sides). Interpretation of the regression coefficients (B)
derived from the GEE analysis are similar to standard linear regression. Within the GEE
analysis, the significance of the independent variable (sides with IAT vs. sides without
IAT) in predicting a dependent variable (H1a, echogenicity; H1b, diameter; H1c, strain;
H1d, stiffness) was tested using a Wald Chi-Square test. Demographic variables (BMI,
Age, Sex) were included as covariates in the GEE analyses. Significance was defined as
a two-tailed P-value ≤0.05.
For significant differences between sides with IAT and sides without IAT,
additional statistical tests were performed independent from the GEE analyses. One-way
ANCOVA’s were used to test if differences found in the GEE analyses were driven by
side (within subject effect), by group (between subject effect), or both. To ensure
homogeneity of regression slopes, the correlation for significant covariates from the GEE
65
analysis was examined for each side used in the ANCOVA comparisons (i.e. involved
side of IAT, uninvolved side of IAT, matched control side). In this bivariate analysis, if the
statistical significance was <0.20 for the correlation between the demographic variable
and the dependent variable for both sides used in the comparison, then it was included
in the ANCOVA analysis. To test the effect of side, the involved side was compared to
the uninvolved side in participants with IAT. To test the effect of group, the same side
(right or left) that was involved for the IAT participant was chosen as the matched side
for analysis for the corresponding age- and gender-matched control. It was hypothesized
that the differences between sides with IAT and sides without IAT would be driven by
both within and between subject effects.
Specific Aim 2. This aim tested two clinical ideas on how DF ROM is linked to
IAT. According to the first, individuals with IAT have a gastrocnemius contracture on the
involved side (involved side of participants with IAT, n=20), which results in limited DF
ROM, as evidenced by lower DF capacity and performance measures, than sides
without IAT (uninvolved side of participants with IAT, n=20 and both sides of controls,
n=40 sides). According to the second idea, individuals with IAT have bony impingement
on the achilles tendon insertion during daily activities, and thus sides with IAT use a
greater percentage of end-range DF than sides without IAT. Additionally, these
dependent variables are compared between sides (involved vs. uninvolved in the IAT
group) and between groups (IAT vs. controls).
Post-hoc analyses for aim 2. Although the initial focus on ankle motion during
stair ascent was DF, it became clear that there were also differences between groups in
plantar flexion (PF) motion during stair ascent. In fact the greatest differences in ankle
motion between groups occurred when the ankle was moving into PF during the weight
66
acceptance and forward continuance phases of stair ascent. For this reason, two
additional post-hoc analyses were added to examine maximum ankle PF during weight
acceptance and forward continuance phases of stair ascent.
Hypotheses based on potential isolated gastrocnemius contracture.
H2a. Sides with IAT have a lower non-weight-bearing DF capacity than sides without IAT
in both case and control groups.
H2b. Sides with IAT have a lower weight-bearing DF capacity than sides without IAT in
both case and control groups.
H2c. Sides with IAT exhibit lower DF during performance of stair ascent than sides
without IAT in both case and control groups.
Hypothesis based on potential bony impingement of tendon.
H2d. Sides with IAT use a greater percentage of end-range DF during performance of
stair ascent than sides without IAT in both case and control groups.
Post-hoc aim H2e . Sides with IAT will use less PF motion during the weight-acceptance
phase of stair ascent than sides without IAT.
Post-hoc aim H2f . Sides with IAT will use less PF motion during the forward
continuance phase of stair ascent than sides without IAT.
Statistical analysis. As described above, comparisons between sides with IAT
and sides without IAT were tested using generalized estimating equations (GEE). Within
the GEE analyses, the significance of the independent variable (sides with IAT vs. sides
without IAT) in predicting a dependent variable (H2a, non-weight-bearing DF capacity;
H2b, weight-bearing DF capacity; H2c, stair ascent DF; H1d, percentage of DF capacity
used during stair ascent) was tested using a Wald Chi-Square test. One-way ANCOVA
was used to test if differences observed in the GEE analysis was driven by side (within
67
subject effect), by group (between subject effect), or both. Statistically significant
covariates from the GEE analysis that did not violate the assumption of homogeneity of
regression slopes for ANCOVA were included in a multivariate model. It was
hypothesized that the differences between sides with IAT and sides without IAT would
be driven by both within and between subject effects.
Specific Aim 3. The purpose of this aim was to examine if PF strength (clinically
and functionally) is lower on sides with IAT (involved side of participants with IAT, n=20)
than sides without IAT (uninvolved side of participants with IAT, n=20 and both sides of
controls, n=40 sides). Additionally, these dependent variables were compared between
sides (involved vs. uninvolved in the IAT group) and between groups (IAT vs. controls).
Hypotheses.
H3a. Sides with IAT have a lower isometric PF torque than sides without IAT in both
case and control groups.
H3b. Sides with IAT exhibit a lower ankle moment during stair ascent than sides without
IAT in both case and control groups.
H3c. Sides with IAT exhibit a lower ankle power during stair ascent than sides without
IAT in both case and control groups.
Statistical analysis. Similar to Aims 1-2, comparisons between sides with IAT
and sides without IAT were tested using generalized estimating equations (GEE). Within
the GEE analysis, the significance of the independent variable (sides with IAT vs. sides
without IAT) in predicting a dependent variable (H3a, PF strength capacity; H3b, ankle
moment during stair ascent; H3c, ankle power during stair ascent) was tested using a
Wald Chi-Square test. One-way ANCOVA was used to test if differences observed in the
GEE analysis was driven by side (within subject effect), by group (between subject
68
effect), or both. Statistically significant covariates from the GEE analysis that did not
violate the assumption of homogeneity of regression slopes for ANCOVA were included
in a multivariate model. It was hypothesized that the differences between sides with IAT
and sides without IAT were driven by both within and between subject effects.
Exploratory Aim 4. The purpose of this aim was to examine the association
between self-reported function and impairments observed on the involved side in
participants with IAT.
Hypotheses.
H4a. Impairment in tendon characteristics is associated with lower self-reported function.
H4b. Impairment in DF ROM is associated with lower self-reported function.
H4c. Impairment in PF strength is associated with lower self-reported function.
H4d. Impairment in tendon characteristics, which are associated with lower function, is
associated with impairment in DF ROM.
H4e. Impairment in tendon characteristics, which are associated with lower function, is
associated with impairment in PF strength.
Statistical analysis. Each hypothesis for aim 4 was tested independently.
Scatter plots for each hypothesis were used to inspect the data for linear relationships.
Pearson correlations were used to describe the magnitude and sign of linear
relationships.
69
Chapter 4. Results
The dependent variables of tendon characteristics (Aim 1), DF ROM (Aim 2) and
PF strength (Aim 3) on sides with IAT were compared to sides without IAT (the
uninvolved side and age- and gender-matched controls) using GEE analyses. In
addition, pairwise comparisons (ANCOVA’s) were used to examine the effect of side
(involved vs. uninvolved in participants with IAT) and group (IAT vs. controls). Because
of limited empirical data on IAT, hypotheses were primarily derived from theories on the
etiology of IAT, studies on midportion achilles tendinopathy, anecdotal clinical
experience and preliminary studies (See Chapters 2 and 3). Aim 4 explores the
potential associations (Pearson correlation) between impairments associated with IAT
(tendon characteristics, DF ROM, PF strength) and self-reported function within a
theoretical model (See Chapter 2).
The results begin with sections on sample characteristics and missing data, and
are followed by results for each aim including an analysis of reliability, GEE analyses
and ANCOVA’s. The effects of order (first vs. second side tested) and side (right vs. left)
were in turn each initially included as covariates in the GEE analyses. Since neither
effect was a significant predictor (P<0.05) of dependent variables, these within subject
effects were not included in the final statistical models for subsequent analyses.
Sample Characteristics
Twenty individuals with chronic, unilateral IAT and 20 age- and gender-matched
controls were recruited from the University of Rochester, Department of Orthopaedics
and Physical Therapy, Foot and Ankle teams. The median duration of symptoms in
participants with IAT was 10 months (range: 3 months to 15 years). The sample
consisted of individuals who were mostly in their 50’s and 60’s with a roughly equal
70
number of male and female participants (Table 13). With exception for the VISA-A, there
were no statistically significant differences in demographics between the groups (Table
13). However, due to a small sample and limited power, true differences between groups
might not have been detected. For example, although not reaching statistical
significance, the IAT group on average weighed ~7 kg (15 lbs) more than the control
group. Thus, categorically the IAT sample was obese, while the control sample was
overweight based on the average body mass index (BMI).
Table 13. Characteristics of participants with insertional achilles tendinopathy (IAT) and healthy matched controls IAT (n=20) Control ( n=20) P-value Age (yr) 58.6± 7.8 58.2± 8.5 0.863 Female (%) 55% 55% 1.000 Weight (kg) 87.5± 17.5 80.3± 16.0 0.187 Height (m) 1.7± 0.1 1.7± 0.1 0.999 BMI (kg/m2) 30.4± 5.4 27.9± 5.3 0.158 VISA-A (%) 47.6± 26.8 100±0.0 <0.001
Values expressed as mean ± standard deviation or as otherwise indicated Statistically significant differences are in bold (P ≤ 0.05) Missing data
The measurement of tendon mechanical properties and non-weight-bearing DF
capacity occurred under a passive condition. Participants were excluded from analysis if
this condition was not achieved. During measurement of tendon mechanical properties,
electromyography of the medial gastrocnemius and the anterior tibialis muscles was
used to monitor muscle activity. If participants had less than 10⁰ dorsiflexion (n=1) or
demonstrated greater than their resting level muscle activity (n=2), then they were
excluded from analysis of tendon mechanical properties. This resulted in a total of 17
sides with IAT for analysis of tendon mechanical properties.
During the measure of non-weight-bearing DF capacity, activity of the anterior
tibialis muscle was visually monitored. If a participant was unable to keep the anterior
71
tibialis muscle relaxed then the test was not completed. A total of 13 participants (5
participants with IAT and 8 controls), who were unable to relax the anterior tibialis
muscle during testing, were excluded from the non-weight-bearing DF capacity analysis
since the methods depended on achieving an equivalent passive force into DF for all
participants (See Methods, DF ROM capacity, Procedures). Since this was such a large
proportion of the total sample, the weight-bearing DF measure was compared between
participants who could relax (n=27 IAT and controls) and participants who could not
relax (n= 13 IAT and controls). There were no differences in weight-bearing DF between
participants who could relax (mean± SD= 26.4⁰± 5.8) and participants who could not
relax (25.1⁰± 4.8, P= 0.486). Moreover, there was a strong correlation between non-
weight-bearing and weight-bearing DF (sides with IAT: r= 0.71, P= 0.005; sides without
IAT: r= 0.48, P= 0.015). Even though there was a large group of participants that were
excluded from the passive measure, participants who had difficulty relaxing did not
appear to differ in DF ROM from participants who were included in the analysis.
Specific Aim 1- Tendon characteristics
Reliability . Test-retest reliability was examined on different days within the same
week. To minimize risk of aggravating symptoms in participants with IAT, repeat testing
was done with 6 young, healthy adults. Test-retest reliability and standard error of
measurement (SEM) for echogenicity (ICC=0.98, SEM= 3.0 units), diameter (ICC=
0.996, SEM= 0.1 mm), strain (ICC= 0.946, SEM= 0.3 %) and stiffness (ICC= 0.998,
SEM= 0.7 N/mm) in controls was high. A difference between the involved and
uninvolved sides >1.96 SEM was used to describe how frequently a comparison
between sides indicated altered tendon characteristics on the involved side. This cut-off
72
value was chosen because it represents the 95% confidence interval for a subject’s
score on the uninvolved side.
In addition, the reliability of processing the ultrasound images was examined. A
set of 10 images from the involved and 10 images from the uninvolved sides were
blinded. The echogenicity, diameter and tendon elongation were measured for 20
images twice, and then un-blinded and compared for accuracy. The reliability of
processed ultrasound images was high (ICC>0.9 for all measures).
Specific aim 1a (Tendon pathology). A lower echogenicity, which indicates
greater tendon pathology, was associated with IAT. The GEE analysis demonstrated
that the presence of IAT (B= -11.13, 95% Confidence Interval (CI)= -17.11 to -5.16, P<
0.001) was associated with a lower echogenicity. The ANCOVA’s indicated that this
relationship between IAT and echogenicity was consistent for both comparisons
(involved vs. uninvolved side, P= 0.002; IAT vs. controls, P= 0.003; Table 14).
Compared to the uninvolved side, 65% (13/20) of participants had lower echogenicity on
the involved side (difference between sides > 1.96 SEM).
Covariates. A lower echogenicity was associated with a higher BMI and older
age. The GEE analysis demonstrated that a higher BMI (B= – 1.93, 95% CI= -2.54 to -
1.33, P< 0.001) and older age (B= -0.56, 95% CI= -1.09 to -0.03, P= 0.039) were
associated with a lower echogenicity. The correlation between BMI and echogenicity
was similar for all groups (involved side for IAT: r= -0.48, P= 0.034; uninvolved side for
IAT: r= -0.638, P= 0.002; matched side for controls: r= -0.71, P= 0.001). Similar to the
GEE results, a higher BMI was associated with a lower echogenicity in the ANCOVA
analyses (Table 14). The correlation between age and echogenicity was similar for the
involved (r= -0.51, P= 0.023) and uninvolved sides (r= -0.43, P= 0.060) in participants
73
with IAT. However, there was insufficient power to demonstrate that this was a
statistically significant covariate in the ANCOVA (Table 14). In contrast, the correlation
between age and echogenicity (r= 0.15, P= 0.517) had a significance >0.2, and so it was
not included as a covariate in the ANCOVA. There was not sufficient evidence that
gender was associated with echogenicity (B= 6.14, 95% CI= -1.42 to 13.704, P= 0.112),
and so it was not included as a covariate in the ANCOVA analyses.
Table 14. The echogenicity of the involved side in participants with insertional achilles tendinopathy (IAT) compared to two groups: the uninvolved side and controls.
Involved side Comparison side P-value
Group BMI Age
IAT (n=20) 66.7± 20.5
Uninvolved side, IAT (n=20) 77.4± 17.9 0.002 0.017 0.065
Matched side, Control (n=20) 84.8± 9.6 0.003 <0.001 NA
Values expressed as mean ± standard deviation Statistically significant differences are in bold (P ≤ 0.05)
Specific aim 1b (Tendon pathology). A larger tendon diameter, which
indicates more severe tendon pathology, was associated with the presence of IAT. Sides
with IAT had a larger tendon diameter than sides without IAT. The presence of IAT (B=
2.32, 95% CI= 1.66 to 2.99, P< 0.001) predicted a 2.3 mm larger tendon diameter
compared to sides without IAT. Consistent with the GEE analysis, sides with IAT had a
larger tendon diameter than the uninvolved side and controls (P< 0.001 for both
comparisons; Table 15). For 85% (17/20) of participants with IAT the involved side had a
larger tendon diameter than the uninvolved side (difference between sides > 1.96 SEM).
There was not sufficient evidence that BMI (B= 0.06, 95% CI= -0.01 to 0.12, P= 0.071),
Age (B= 0.01, 95% CI= -0.03 to 0.06, P= 0.536) or sex (B= 0.10, 95% CI= -0.61 to 0.80,
P= 0.790) predicted tendon diameter in the GEE analysis, and these variables were not
included as covariates in the ANOVA analyses.
74
Table 15. The diameter (mm) of the involved side in participants with insertional achilles tendinopathy (IAT) compared to two groups: the uninvolved side and controls.
Involved side Comparison side P-value
IAT (n=20) 6.4± 1.6
Uninvolved side, IAT (n=20) 4.3± 0.8 <0.001
Matched side, Control (n=20) 3.6± 0.7 <0.001
Values expressed as mean ± standard deviation Statistically significant differences are in bold (P ≤ 0.05)
Specific aim 1c (Tendon mechanical properties) . A higher tendon strain,
which indicates greater impairment in tendon mechanical properties, was associated
with IAT. Sides with IAT had a higher tendon strain than sides without IAT. The presence
of IAT was associated with a 0.76% higher tendon strain (B=0.76, 95% CI= 0.20 to 1.31,
P= 0.007) compared to sides without IAT. Both compared to the uninvolved side
(P=0.047) and to controls (P= 0.006), the involved side of participants with IAT had
higher tendon strain (Table 16). Compared to the uninvolved side, 59% (10/17) of
participants had higher strain on the involved side (difference between sides > 1.96
SEM). There was not sufficient evidence that BMI (B=0.04, 95% CI= -0.01 to 0.09, P=
0.126), age (B= 0.03, 95% CI= -0.01 to 0.07, P= 0.089) or sex (B= -0.41, 95% CI= -1.07
to 0.24, P= 0.217) predicted tendon diameter in the GEE analysis, and these variables
were not included as covariates in the ANOVA analyses.
Table 16. The strain (%) of the involved side in participants with insertional achilles tendinopathy (IAT) compared to two groups: the uninvolved side and controls
Involved side Comparison side P-value
IAT (n=17) 3.4± 1.5
Uninvolved side, IAT (n=17) 2.8± 1.0 0.047
Matched side, Control (n=20) 2.3± 0.8 0.006
Values expressed as mean ± standard deviation Statistically significant differences are in bold (P ≤ 0.05)
75
Specific aim 1d (Tendon mechanical properties). A lower tendon stiffness,
which indicates greater impairment in tendon mechanical properties, was associated
with IAT. Sides with IAT had lower stiffness than sides without IAT. The presence of IAT
predicted a 12 N/mm lower stiffness than sides without IAT (B= -12.44, 95% CI=-19.62
to -5.27, P= 0.001). Sides with IAT had lower stiffness than the uninvolved side and
controls (involved vs. uninvolved side, P= 0.030; IAT vs. controls, P= 0.021). The
absolute values for tendon stiffness and ANOCA comparisons are listed in Table 17. For
participants with IAT, 65% (11/17) of participants had lower stiffness on the involved side
than the uninvolved side (difference between sides > 1.96 SEM). There was not
sufficient evidence that BMI (B= 0.83, 95% CI= -0.31 to 1.96, P= 0.156), age (B= 0.05,
95% CI= -0.83 to 0.94, P= 0.910) or sex (B= 1.78, 95% CI= -7.22 to 10.78, P= 0.698)
predicted tendon diameter in the GEE analysis, and these variables were not included
as covariates in the ANOVA analyses.
Table 17. Tendon stiffness (N/mm) of the involved side in participants with insertional achilles tendinopathy (IAT) compared to two groups: the uninvolved side and controls
Involved side Comparison side P-value
IAT (n=17) 33.5± 12.6
Uninvolved side, IAT (n=17) 43.7± 19.2 0.030
Matched side, Control (n=20) 44.4± 14.4 0.021
Values expressed as mean ± standard deviation Statistically significant differences are in bold (P ≤ 0.05) Specific Aim 2 – Dorsiflexion Range of Motion
Reliability . To examine how consistently the examiner was able to measure non-
weight-bearing DF capacity, this measure was repeated twice within the same testing
session in 5 controls and 5 participants with IAT on the uninvolved side. The reliability of
this measure was high (ICC=0.900). The reliability of weight-bearing DF capacity
76
examined how consistently the participants assumed the same position when instructed
to move to the maximum tolerated DF. The reliability of this measure was also high
(ICC=0.788). To compare classification of limited DF capacity in the current study to the
literature, the definition of cut-off values was based on SD’s of the control group (rather
than SEM’s) to be consistent with previous studies.
Specific aim 2a-b (Range of motion capacity) . There was no evidence
supporting an association between IAT and DF capacity. The 95% confidence interval of
the difference in DF capacity between sides with IAT and sides without IAT included 0
(Non-weight-bearing: B= 1.20, 95% CI= -1.69 to 4.09, P=0.414; Weight-bearing: B= -
1.07, 95% CI= -3.73 to 1.59, P=0.431). Descriptive data for absolute DF capacity is
listed in Table 18.
Based on the definition of <5⁰ as limited non-weight-bearing DF capacity by
Digiovanni et al. (2002) and Moseley et al. (2001), only 1/15 participants with IAT and
0/12 controls were classified as limited non-weight-bearing DF capacity. Since there is
no definition of limited DF in a weight-bearing position, DF capacity less than 2 SD’s
from the mean of the control group in the current study was used. This is the same
manner that Moseley et al. (2001) used to define limited non-weight-bearing DF capacity
. Using the controls as a standard for normal, limited weight-bearing DF capacity was
defined as <17.3⁰ (µ – 2SD, Table 18). Based on this definition, 2/20 sides with IAT and
1/20 sides of controls were classified as having limited weight-bearing DF capacity.
Covariates. Non-weight-bearing and weight-bearing measures provided mixed
evidence supporting an association between DF capacity with gender and BMI. Male
gender was associated with a lower non-weight-bearing DF capacity (B= -3.05, 95% CI=
-5.47 to -0.63, P= 0.014), but there was not sufficient evidence to support an association
77
with weight-bearing DF (B= -0.16, 95% CI= -2.46 to 2.13, P= 0.890). A higher body mass
index was associated with a lower weight-bearing DF capacity (B= -0.27, 95% CI= -0.48
to -0.06, P= 0.013), but there was not sufficient evidence to support an association with
non-weight-bearing DF capacity (B= -0.15, 95% CI= -1.69 to 4.09, P= 0.414). There was
not sufficient evidence that age (Non-weight-bearing: B= 0.07, 95% CI= -0.08 to 0.22,
P= 0.352; Weight-bearing: B= -0.09, 95% CI= -0.21 to 0.03, P= 0.129) was associated
with DF capacity.
Table 18. Non-weight-bearing and weight-bearing dorsiflexion (DF) capacity of the involved side in participants with insertional achilles tendinopathy (IAT), the uninvolved side of participants with IAT and controls
DF capacity ( ⁰⁰⁰⁰) Involved side,
IAT (n=15) Uninvolved side,
IAT (n=15) Matched side, Control (n=12)
Non-weight-bearing 15.4± 5.3 14.5± 6.4 14.3± 2.1
Involved side, IAT (n=20)
Uninvolved side, IAT (n=20)
Matched side, Control (n=20)
Weight-bearing 24.8± 5.5 25.4± 5.0 27.5± 5.1
Values expressed as mean ± standard deviation
Specific aim 2c-d (Range of motion performance). There were mixed results
supporting the association between IAT and DF ROM performance variables. The 95%
confidence interval of the difference in stair ascent DF between sides with IAT and sides
without IAT included 0 (B= 1.10, 95% CI= -0.70 to 2.90, P= 0.232; Figure 18). Yet IAT
was associated with 8% higher use of end-range DF (B= 7.78, 95% CI= 0.28 to 15.28,
P= 0.042) than sides without IAT. For the comparison between sides in participants with
IAT, there was not a statistically significant difference in the percentage of end-range DF
used between sides (P= 0.141, Table 19). However, the involved side of participants
with IAT used 13% more end-range DF to ascend stairs than controls (P= 0.041, Table
19).
78
Covariates. Older age was associated with higher DF during stair ascent. Older
age was associated with a 0.18⁰ and a 1% increase in DF per year of age (Stair ascent
DF: B= 0.18, 95% CI= 0.08 to 0.27, P< 0.001; Percentage of DF capacity used during
stair ascent: B= 0.96, 95% CI= 0.56 to 1.35, P< 0.001). Given that age was consistently
associated with percentage of DF capacity used for all sides (involved side for IAT: r=
0.47, P= 0.036; uninvolved side for IAT: r= 0.45, P= 0.044; matched side for control: r=
0.48, P= 0.031), it was included as a covariate for ANCOVA analyses. Similar to the
GEE analysis, age was a significant covariate for between sides and between groups
comparisons (Table 19).
A higher BMI was also associated with a higher percentage of end-range DF
used during stair ascent (B= 1.01, 95% CI= 0.32 to 1.71, P= 0.004), but there was not
sufficient evidence to support an association with the degrees of stair ascent DF (B=
0.09, 95% CI= -0.09⁰ to 0.27⁰, P= 0.320). There was an association between BMI and
percentage of end-range DF used for participants with IAT (involved side: r= 0.44, P=
0.050; uninvolved side: r= 0.381, P= 0.097). However, there was insufficient power with
the ANCOVA analysis to support this association (Table 19). In controls, the statistical
significance of the association between BMI and percentage of end-range DF used was
<0.2 (r= 0.19, P=0.413), and so it was not included as a covariate in the comparison
between groups (Table 19). There was no evidence to support that sex (stair ascent DF:
B=0.65⁰, 95% CI= -1.28⁰ to 2.58⁰, P= 0.511; Percentage of end-range DF used: B= -
0.16, 95% CI= -7.85 to 7.54, P= 0.968) predicted DF performance.
79
Table 19. Dorsiflexion (DF) used during stair ascent in participants with insertional achilles tendinopathy (IAT) on the involved and uninvolved sides and in controls
DF during pull-up phase of stair ascent ( ⁰⁰⁰⁰) P-value
Group Age BMI
Involved side, IAT (n=20) 11.6± 4.7
Uninvolved side, IAT (n=20) 10.7± 3.9 NA NA NA
Matched side, Control (n=20) 10.0± 3.7 NA NA NA
Percentage of DF capacity used during pull-up phase of stair ascent (%)
Involved side, IAT (n=20) 50.2± 22.6
Uninvolved side, IAT (n=20) 43.6± 17.4 0.141 0.047 0.950
Matched side, Control (n=20) 37.7± 16.6 0.041 0.003 NA
Abbreviations: NA, not applicable since ANCOVA’s not performed Values expressed as mean ± standard deviation Statistically significant differences are in bold (P ≤ 0.05) Figure 18. Degrees of ankle motion (dorsiflexion is positive) used during stair ascent by percentage of stance for participants with insertional achilles tendinopathy (IAT) on the involved and uninvolved sides as well as by controls. Asterisks indicate when during stance there were differences between groups and/or sides.
Weight Acceptance
3 Phases of Stair Ascent:
Pull-up of contralateral limb
Forward Continuance
*
*
80
Post-hoc aim 2e-f (Range of motion performance). Less PF motion during
stair ascent, which indicates impairment in terms of remaining in a dorsiflexed position,
was associated with IAT. The presence of IAT was associated with lower ankle PF
during the weight acceptance (B= 2.47, 95% CI= 0.77 to 4.17, P= 0.004) and forward
continuance (B= 5.56, 95% CI= 3.43 to 7.69, P< 0.001) phases of stair ascent. This
pattern of decreased ankle PF for the involved side of individuals with IAT was
consistent when compared to the uninvolved side (weight acceptance phase: P= 0.044,
forward continuance phase, P< 0.001; Table 20) and to controls (weight acceptance
phase: P= 0.031, forward continuance phase, P= 0.017; Table 20).
Covariates. Older age was associated with less PF motion during stair ascent.
Each additional year of age was associated with 0.2⁰ less PF during weight acceptance
(B= 0.15, 95% CI= 0.04 to 0.25, P=0.007) and forward continuance (B=0.18, 95% CI=
0.00 to 0.36, P= 0.047) phases of stair ascent, assuming all other factors remain
constant. The correlation between age and PF motion was consistent for all sides used
for comparison during the weight acceptance phase (involved side for IAT: r= 0.56, P=
0.010; uninvolved side for IAT: r= 0.38, P= 0.102; matched side for controls: r= 0.36, P=
0.123). Similar to the GEE analysis, age was a significant predictor of PF motion during
the weight acceptance phase of gait for both ANCOVA analyses (Table 20). However,
the correlation between age and PF motion was not statistically significant during
forward continuance phase (involved side for IAT: r= 0.25, P= 0.299; uninvolved side for
IAT: r= 0.05, P=0.838; matched side for controls: r= 0.10, P= 0.681). Therefore, it was
not included in the pairwise comparisons.
Gender was inconsistently associated with PF motion, and there was insufficient
evidence to support an association between BMI and PF motion. There was not
81
sufficient evidence to support an association between sex and PF motion during weight
acceptance phase (B=0.13, 95% CI= -1.77 to 2.02, P=0.897). Yet, male gender was
associated with 6⁰ less PF motion during forward continuance phase (B= -5.78, 95% CI=
-9.46 to -2.11, P=0.002). However, since the statistical significance of this correlation
was greater than 0.2 for the involved side of IAT participants (r= -0.21, P=0.373), sex
was not included as a covariate in the ANOVA analyses. There was not sufficient
evidence to support an association between BMI (weight acceptance: B=0.04, 95% CI= -
0.15 to 0.23, P= 0.699; forward continuance: B= -0.04, 95% CI= -0.26 to 0.02, P= 0.732)
and PF motion.
Table 20. Plantar flexion (PF) motion used during stair ascent of the involved side in participants with insertional achilles tendinopathy (IAT) compared to two groups: the uninvolved side and controls
PF during weight acceptance phase ( ⁰⁰⁰⁰)
Comparison side P-value
Group Age
IAT (n=20) -1.1± 4.2
Uninvolved side, IAT (n=20) -3.1± 3.0 0.044 0.007
Matched side, Control (n=20) -3.8± 3.4 0.031 0.003
PF during forward continuance phase ( ⁰⁰⁰⁰)
IAT (n=20) -7.6± 3.8
Uninvolved side, IAT (n=20) -13.4± 5.5 <0.001 NA
Matched side, Control (n=20) -12.1± 7.0 0.017 NA
Values expressed as mean ± standard deviation Specific Aim 3 – Plantar flexion strength
Reliability . The test-retest reliability of measuring maximum isometric PF torque
was assessed in 6 young, healthy adults. The reliability of this measure was good with
an ICC of 0.928 and a SEM of 10 Nm. Similar to the cut-off values for tendon
characteristics, a difference between the involved and uninvolved sides >1.96 SEM was
82
used to describe how frequently a comparison between sides indicated PF weakness on
the involved side.
Specific aim 3a (Strength capacity). A lower isometric plantar flexion torque,
which indicates impairment in PF strength capacity, was associated with IAT. The
presence of IAT predicted a 9 Nm lower maximum isometric PF torque compared to
sides without IAT (B= -8.55, 95% CI= -15.03 to -2.07, P= 0.010), when covariates were
held constant. For the comparison between the involved side and controls, the presence
of IAT (P= 0.051) was associated with a lower PF torque (Table 21). For the comparison
between the involved side and controls, the presence of IAT (P= 0.051) was associated
with a lower PF torque. For 5/20 participants with IAT, the involved side was weak
compared to the uninvolved side (difference between sides >1.96 SEM).
Covariates. A lower isometric PF torque was also associated with female gender
in the GEE and ANCOVA analyses. Female gender (B= -23.16, 95% CI= -46.00 to -0.32,
P= 0.047) was associated with lower isometric PF torque. Female gender was
associated with a lower maximum isometric PF torque on sides with IAT (r= -0.63, P=
0.003) and sides without IAT (uninvolved: r= -0.57, P= 0.009; controls: r= -0.42, P=
0.066). For the comparison between the involved and uninvolved sides, being female
(P=0.022) was associated with a lower isometric PF torque (Table 21). For the
comparison between the involved side and controls, female gender (P=0.011) was
associated with a lower PF torque (Table 21).
Age was inconsistently associated with isometric PF strength, and there was
insufficient evidence to support an association between BMI and PF strength. Older age
(B= -2.11, 95% CI= -3.49 to -0.73, P= 0.003) was associated with lower isometric PF
torque. Older age was associated with lower maximum isometric PF torque on sides with
83
IAT (r= -0.50, P= 0.024) and sides without IAT (uninvolved, r= -0.55, P= 0.012; controls:
r= -0.56, P= 0.008). For the comparison between the involved side and controls, older
age (P=0.008) was associated with a lower PF torque (Table 21). However, there was
insufficient power to detect age as a statistically significant covariate in the comparison
to controls (Table 21). The association between BMI and isometric plantar flexion
strength (B= -1.17, 95% CI= -2.76 to 0.41, P= 0.147) was not significant within the GEE
analysis, and, therefore, was not included in the ANCOVA analyses.
Table 21. Maximum isometric plantar flexion torque (Nm) of the involved side in participants with insertional achilles tendinopathy (IAT) compared to two groups: the uninvolved side and controls
Involved side Comparison side P-value
Group Age Gender
Involved side, IAT
65.3± 40.0
Uninvolved side, IAT 73.5± 37.8 0.028 0.081 0.022
Matched side, Control 85.9± 38.0 0.051 0.008 0.011
Values expressed as mean ± standard deviation Statistically significant differences are in bold (P ≤ 0.051)
Specific aim 3b (Strength performance). There was not conclusive evidence
that a more negative ankle moment, indicating impairment in the ability of the plantar
flexors to control DF, was associated with IAT. The peak ankle moment occurred during
the forward continuance phase of stair ascent (Figure 19) and descriptive data on this
variable are provided in Table 22. The presence of IAT was associated with a 0.04
Nm/kg increase in ankle moment compared to sides without IAT (B= 0.04, 95% CI= -
0.09 to 0.17, P= 0.538); this difference was not statistically significant.
Covariates. BMI, age and sex were all associated with ankle moment. A higher
BMI was associated with a lower ankle moment (B= -0.02, 95% CI= -0.01 to 0.03, P=
0.001). Older age was associated with a higher moment (B= 0.01, 95% CI= 0.00 to 0.02,
84
P= 0.013). Female gender was not associated with a higher moment (B= -0.02, 95% CI=
-0.15 to 0.12, P= 0.823).
Figure 19 . Ankle moment by percentage of stance for participants with insertional achilles tendinopathy (involved and uninvolved sides) and controls
Table 22. Peak ankle moment of stair ascent of the involved side in participants with insertional achilles tendinopathy (IAT), the uninvolved side, and controls Moment (Nm/kg)
Involved side, IAT
Uninvolved side, IAT
Matched side, Control
-1.5± 0.4 -1.5± 0.3 -1.7± 0.3 Values expressed as mean ± standard deviation
Specific aim 3c (PF strength performance) . A lower ankle power, which
indicates impairment in PF strength, was associated with IAT (Figure 20). The presence
of IAT was associated with a 0.8 W/kg lower ankle power compared to sides without IAT
(B= -0.76, 95% CI: -1.13 to -0.30, P<0.001). Compared to the uninvolved side (P=
Weight Acceptance
3 Phases of Stair Ascent:
Pull-up of contralateral limb
Forward Continuance
85
0.006) and to controls (P= 0.006), the involved side had lower ankle power during stair
ascent (Table 23).
Covariates. A lower ankle power was associated with older age. Per year of
increased age, ankle power decreased by 0.04 Nm/kg*s (B= -0.04, 95% CI= -0.07 to
0.00, P= 0.036). In the ANCOVA analyses, age was associated with ankle power in
participants with IAT (involved side: r= -0.36, P= 0.115; uninvolved side: r= -0.32, P=
0.176), but this association was not statistically significant in controls (r= -0.09,
P=0.702).
A lower ankle power was associated inconsistently with female gender, and there
was not sufficient evidence of an association between ankle power and BMI. Being
female predicted a 0.67 W/kg decrease in ankle power (B=0.67, 95% CI= 0.17 to 1.17,
P= 0.009). Yet there was not sufficient evidence to support that sex was associated with
ankle power in participants with IAT on the involved side (r= -0.10, P= 0.670), and so it
was not included as a covariate in the ANCOVA comparisons. The association between
BMI and ankle power was not significant within the GEE analysis (B= -0.04, 95% CI: -
0.09 to 0.00, P=0.061).
86
Figure 20 . Ankle peak power by percentage of stance for participants with insertional achilles tendinopathy (involved and uninvolved sides) and healthy controls. Asterisk indicates when during stance there were differences between groups and sides.
Table 23. Ankle power (W/kg) during stair ascent on the involved side of participants with IAT compared to the uninvolved side and to the same side (right or left) of controls
Involved side Comparison side P-value
Group Age
IAT (n=20) 2.7± 0.9
Uninvolved side, IAT (n=20) 3.5± 1.0 0.006 0.079
Matched side, Control (n=20) 3.6± 1.0 0.006 NA
Values expressed as mean ± standard deviation Statistically significant differences are in bold (P ≤ 0.05)
Specific Aim 2– Model of insertional achilles tendi nopathy impairments
Exploratory aim 4a (Tendon characteristics and func tion) . On the involved
side of participants with IAT, a lower echogenicity was associated with lower function (r=
3 Phases of Stair Ascent:
Weight Acceptance
Pull-up of contralateral limb
Forward Continuance
*
87
0.62, P= 0.004; Figure 21a). The negative correlation between tendon diameter and
function was significant at the P= 0.051 level (r= -0.44, P= 0.051; Figure 21b). There was
not conclusive evidence of a correlation between either tendon strain (r=-0.33, P= 0.197)
or stiffness (r=0.17, P=0.530) with function.
Figure 21 . a) Tendon composition (Echogenicity) and b) shape (Diameter) vs. self-reported function (VISA-A) on the involved side of participants with insertional achilles tendinopathy
Exploratory aim 4b (Range of motion and function). Both DF ROM capacity
and performance variables were associated with self-reported function. Weight-bearing
DF capacity was positively correlated (r= 0.47, P= 0.035, Figure 22a) and the
percentage of end-range DF used during stair ascent was negatively correlated (r= -
0.53, P= 0.017, Figure 22b) with function. The relationships indicated that function
decreased as DF capacity decreased or as the percentage of DF capacity used
increased. Non-weight-bearing DF capacity (r= 0.22, P= 0.459) and degrees of stair
ascent DF (r= -0.24, P= 0.302) were not significantly correlated with function.
a) b)
88
Figure 22 . a) Weight-bearing dorsiflexion (DF) capacity and b) Percentage of DF capacity used during stair ascent vs. self-reported function (VISA-A) on the involved side in participants with IAT
Exploratory Aim 4c (Strength and function). Both PF strength capacity and
performance variables were associated with self-reported function. The maximum
isometric PF strength (r= 0.47, P= 0.037, Figure 23a) and ankle power (r= 0.55, P=
0.012, Figure 23b) were positively associated with function. The direction of the
relationship indicated that self-reported function increased as PF strength capacity
(isometric PF torque) and performance (ankle power) increased.
Figure 23 . a) Isometric plantar flexion strength and b) ankle power vs. self-reported function (VISA-A) on the involved side in participants with IAT
a) b)
a) b)
89
Exploratory aim 4d (Tendon characteristics and Rang e of motion). Since
percentage of end-range DF used during stair ascent was most strongly correlated to
function of the DF ROM variables, the correlation between percentage of end-range DF
and tendon characteristics was examined. The percentage of end-range DF correlated
with some tendon characteristics. Use of a greater percentage of end-range DF was
associated with lower echogenicity (r= -0.46, P= 0.006), but this association was not
statistically significant with tendon diameter (r= 0.369, P=0.109). Use of a greater
percentage of end-range DF was associated with greater impairment in terms of
mechanical property of strain (r=0.639, P=0.006), but this association was not
statistically significant with stiffness (r= -0.433, P=0.082).
Exploratory Aim 4e (Tendon characteristics and Stre ngth). Since ankle
power during stair ascent was most strongly correlated to function out of the PF strength
variables, the correlation between ankle power and tendon characteristics was
examined. There were no significant correlations between ankle power and tendon
pathology (echogenicity: r= 0.19, P= 0.422; diameter: r= -0.29, P= 0.213). Neither were
there significant correlations between ankle power and mechanical properties (strain: r=
-0.32, P= 0.210; stiffness: r=0.10, P= 0.690). The lack of correlation between PF
strength and tendon characteristics led to a revised model of insertional achilles
tendinopathy impairments (Table 24, Figure 24).
90
Table 24 . Correlations between impairments and function on the involved side in participants with insertional achilles tendinopathy DF ROM Tendon PF strength
Function Variable Weight-bearing
% end-range
DF Echo Strain Max PF
torque
Peak ankle power
Weight-bearing DF
-0.559, 0.010
0.244, 0.300
-0.539, 0.026
0.236, 0.316
0.365, 0.113
0.474, 0.035
% of DF capacity used
-0.461,
0.041 0.639, 0.006
-0.349, 0.132
-0.306, 0.189
-0.527, 0.017
Echo
-0.062, 0.812
0.374, 0.105
0.190, 0.422
0.620, 0.004
Strain
0.062, 0.813
-0.320, 0.210
0.165, 0.527
Max PF torque
0.325,
0.162 0.469, 0.037
Peak ankle power
0.549,
0.012
Abbreviations: DF, dorsiflexion; ROM, range of motion; PF, plantar flexion; %, percentage; Echo, echogenicity; Max, maximum Values presented as r, P-value Statistically significant values are in bold
Overall, differences in tendon characteristics, DF ROM and PF strength were
found between sides with IAT and sides without IAT. In addition, greater impairment in
these variables was associated with lower self-reported function in participants with IAT.
However, statistically significant differences were not found between sides with IAT and
sides without IAT for certain variables examined in aim 2 (DF capacity, degrees of DF
used to ascent stairs) and aim 3 (ankle moment). For statistically significant differences
between sides with IAT and sides without IAT in the GEE analyses, the same pattern of
differences were found with the pairwise comparisons. One exception to this pattern was
that there was not a statistically significant difference between the involved and
91
uninvolved side in the percentage of end-range DF used during stair ascent. An
unanticipated finding was that sides with IAT differed from sides without IAT in the
amount of PF used during stair ascent. The clinical significance of these findings and
how they may be used to guide future research will be discussed in the following
chapter.
Figure 24 . Revised model of insertional achilles tendinopathy impairments with asterisks indicating statistically significant correlations
92
Chapter 5. Discussion
The findings of the current study are the first to quantify abnormalities in tendon
characteristics with ultrasound imaging, DF ROM and PF strength in persons with IAT.
Further, the association between the severity of these impairments and self-reported
function was elucidated in a theoretical model. This chapter is divided into six sections to
discuss the findings of each study aim as well as statistical analyses and study
limitations. In addition, a table summarizing clinical recommendations based on findings
from each aim is located at the end of the chapter.
Tendon characteristics
The new findings from this study are that IAT is associated with ultrasound
imaging measures of pathology. Additionally, the severity of pathology on the involved
side was associated with lower self-reported function. Consistent with hypotheses on
tendon characteristics, the involved side of participants with IAT had a larger tendon
diameter and lower echogenicity than the uninvolved side and than controls. These
findings support the potential diagnostic and prognostic value of ultrasound measures in
IAT. While the magnitude of alterations in tendon mechanical properties were similar to
other types of tendinopathy, the clinical significance of these alterations requires further
study.
In the current study, ultrasound measures detected altered tendon structure due
to IAT. Further, measures of echogenicity, as defined here, were associated with
function. A lower echogenicity (i.e. a lower mean grayscale value) indicated less
organization and altered composition of the tendon microstructure (Collinger et al.,
2010). This altered composition is believed to represent tendon degeneration
associated with tendinopathy. However, validity studies connecting tendon structure and
93
histopathology with measures of gray and white pixel patterns from ultrasound imaging
(e.g. mean grayscale value) would improve the understanding echogenicity. Further,
use of the mean grayscale value may underestimate the magnitude of pathology and
variability of tendon composition, which new measures may be able to better capture. To
date, studies examining the relationship between ultrasound measures of tendon
composition and histopathology have been limited by categorical (normal vs. abnormal)
descriptions (Astrom et al., 1996; Klauser et al., 2013). However, a categorical measure
limits the potential to use echogenicity as a means of diagnostic grading of IAT severity
or as a means of predicting success with conservative care among a spectrum of IAT
patients (Table 25). Because echogenicity was associated with BMI and function in
persons with IAT, further studies associating this continuous measures of the gray and
white pixel patterns from ultrasound and tendon histopathology are encouraged. Further
research is needed to examine why a higher BMI was associated with a lower
echogenicity, and if this is a risk factor for developing tendinopathy.
Tendon diameter was another reliable ultrasound measure that was strongly
associated with IAT. For 85% of participants with IAT the involved side was thicker than
the uninvolved side. Because diameter was the most consistent marker of IAT among
the tendon characteristics measured in the current study, this variable may be
particularly useful in the development of diagnostic and prognostic criteria. Nicholson et
al. (2007) developed a grading system for achilles tendon pathology based on MRI, in
which the severity was associated with prognosis for conservative care. The grading
system was based on tendon diameter and the presence of degeneration in the tendon,
as follows; Grade I: AP diameter of 6-8 mm and non-uniform degeneration; Grade II:
diameter of >8 mm with uniform degeneration of <50% of tendon width; and Grade III:
94
tendon diameter >8 mm and uniform degeneration of >50% tendon width (Nicholson et
al., 2007). Nicholson et al. (2007) found that individuals with grade I IAT pathology were
less likely (13%, 2/16 tendon) to get surgery than individuals with Grade II (91%, 59/65)
or Grade III (70%, 19/27) pathology (Nicholson et al., 2007). Ultrasound imaging may be
equally as effective in determining prognosis, as well as more available and less
expensive to employ clinically than MRI. For example, based on the <8 mm criteria used
by Nicholson et al. (2007), 16/20 participants in the current study would have a positive
prognosis with conservative care. While prospective data are necessary to establish
prognostic factors, this cross-sectional study establishes that there is a spectrum of
tendon diameter thickness across participants with IAT (Table 25).
Insertional and midportion forms of achilles tendinopathy may have similar
effects on tendon strain. The difference in strain (1.2%) between participants with IAT
and controls in the current study was similar in magnitude to previous studies comparing
participants with midportion achilles tendinopathy to controls (Arya et al.(2010) = 0.8%
and Child et al. (2010) = 1.8%) . While these differences between groups in strain
appear relatively small, the magnitude of the difference may increase with dynamic
tasks, such as running or end-range calf stretches. For example, the strain associated
with single leg hopping (Lichtwark et al. (2005) = 8.3%) is nearly 4 times the amount of
strain demonstrated by controls for the current study during a 20⁰ passive ankle
excursion (2.2%, Table 16). Thus, alterations in strain during more challenging tasks
may affect dynamic muscle-tendon function. While a link between tendon mechanical
properties and muscle-tendon function is possible, there was no conclusive evidence of
a link between strain and self-reported function (r=-0.329, P=0.197) in the current study.
95
Passive stretching of the achilles tendon as a part of physical therapy care for
IAT is inconsistent with the finding of increased tendon strain (Table 25). The finding of
excessive elongation of the achilles tendon calls into question the regular administration
of calf stretching for tendons that already exhibit increased strain. Although commonly
recommended based on anecdotal experience, there is little empirical evidence to
support stretching in patients with IAT (Verrall, Schofield, & Brustad, 2011). Typically,
the goal of calf stretching is to increase DF ROM. Stretching may further reinforce
increased strain of the achilles tendon, which data from the current study demonstrate
was already increased in participants with IAT. Studies on the influence of DF ROM and
its influence on continuing symptoms associated with IAT are necessary (See discussion
of DF ROM below).
Despite some differences in methods and type of tendon problems, the effect of
tendinopathy on stiffness was similar to other studies on tendinopathy. The current study
examined tendon stiffness during a passive task (20⁰ ankle rotation) and found a 27.8%
decrease in tendon stiffness in participants with IAT compared to controls. Similarly,
during an active task (maximum isometric voluntary contraction) Arya et al. (2010) found
a 24.9% decrease in achilles tendon stiffness in participants with midportion achilles
tendinopathy. A passive task was chosen for the current study to obtain similar loading
between the IAT and control groups. Yet, individuals with IAT may show greater
differences in tendon characteristics under higher loads. Stiffness in participants with IAT
could also be tested with an active test where stiffness is measured during a maximal
isometric effort. This was not done in the current study because differences in strength
(See Chapter 2, PF strength Capacity, Preliminary Studies, Table 10) would confound
the comparison between groups (i.e. tendon strain is not independent from force). Future
96
researchers, however, may consider using various submaximal loads to evaluate tendon
stiffness under more functional loading conditions. For example, data from this study
suggest loads as high as 25 Nm could be achieved by 90% of participants with IAT (See
Discussion, PF strength).
Based on cross-sectional data from the current study demonstrating decreased
tendon stiffness in IAT, therapeutic strategies that address abnormal mechanical
properties may theoretically benefit participants with IAT (Table 25). While eccentric
exercise is commonly prescribed for IAT (Fahlstrom et al., 2003; Jonsson et al., 2008;
Rompe et al., 2009), in controls isometric exercise, not eccentric exercise, stimulated
improvement in tendon mechanical properties. Two studies of healthy participants have
demonstrated increased achilles tendon stiffness after a 14 week isometric exercise
training protocol (Arampatzis, Karamanidis, & Albracht, 2007; Arampatzis et al., 2010).
Conversely studies demonstrating the effect of eccentric exercise at 6 weeks showed a
decrease in achilles tendon stiffness (N. N. Mahieu et al., 2008; Morrissey et al., 2011).
To date research on the effects of exercise on Achilles tendon mechanical properties
has been limited to a healthy population. The ability of exercise to alter the tendon
mechanical properties of participants with tendinopathy remains unexplored, yet may
yield a more specific exercise approach that both reduces pain and restores tendon
mechanical behavior.
Dorsiflexion Range of Motion
Individuals with IAT often have pain and difficulty with stairs, but it was unknown
if this was related to a limited capacity or altered performance of DF ROM. Contrary to
the theory of gastrocnemius tightness contributing to IAT, limited DF ROM was not a
common impairment in participants with IAT in the current study. In contrast, the theory
97
of bony impingement was supported in that participants with IAT used a higher
percentage of end-range DF during stair ascent, and greater use of DF was correlated
with lower function. Although participants with IAT and controls had a similar DF
capacity, participants with IAT used a greater percentage of their total available DF ROM
during stair ascent. Additionally older age was associated with use of higher percentage
end-range DF during stair ascent, indicating that both pathology and age affect
functional performance.
There were no statistically or clinically significant differences in non-weight-
bearing measures of DF capacity between sides with IAT and sides without IAT. Despite
a smaller sample size due to difficulties in achieving a relaxed DF ROM test, the mean
non-weight-bearing DF capacity found in the current study (n=15, 14.3⁰±2.1) was similar
to that found by Digiovanni et al. (2002) (n=34, 13.1⁰±8.2) and slightly lower than
Moseley et al. (2001) (N=298, 18.1⁰±6.9). A younger sample of healthy adults (15 to 34
years old) may have contributed to the difference in DF between Moseley et al. (2001)
and the current study. Nevertheless, using the definition of limited non-weight-bearing
DF proposed by Digiovanni et al. (2002) and Moseley et al. (2001), only 7% (1/15) of
participants with IAT and 0% (0/12) of controls had limited (<5⁰) non-weight-bearing DF
capacity. The percentage of participants with limited DF capacity were similar between
non-weight-bearing and weight-bearing measures, in which 10% (2/20) of participants
with IAT and 5% (1/20) controls had limited weight-bearing DF capacity. In contrast,
Digiovanni et al. (2002) reported 65% (22/34) of participants with foot pain and 24%
(8/34) of controls had limited non-weight-bearing DF capacity. Thus, while limited non-
weight-bearing DF capacity may be commonly associated with forefoot and midfoot pain,
it was not commonly associated with IAT in the current study. This finding indicates that
98
stretching exercises were not needed to improve DF ROM for the majority of persons
with IAT in this sample (Table 25).
Similarly, limited weight-bearing DF capacity was not common in participants with
IAT. However, there was a correlation between weight-bearing DF capacity and self-
reported function. There are currently no studies with a large sample size on weight-
bearing DF capacity with the knee straight, and thus mean estimates for DF capacity in
healthy adults range widely from 22.8⁰ to 33.6⁰ (Denegar, Hertel, & Fonseca, 2002;
Krause, Cloud, Forster, Schrank, & Hollman, 2011; Munteanu et al., 2009; Sidaway et
al., 2012; Williams, Caserta, & Haines, 2013). A comparable measure was achieved in
the current study by using an approach which isolated calcaneal motion relative to the
tibia, avoiding the inclusion of arch deformation. Despite this methodological difference,
the mean weight-bearing DF capacity found in the current study (27.5±5.1) was within
the range reported in the literature (22.8⁰ to 33.6⁰) (Denegar et al., 2002; Krause et al.,
2011; Munteanu et al., 2009; Sidaway et al., 2012; Williams et al., 2013). Of note for
clinical implications, the maximal weight-bearing calf stretch was the most painful task
during study participation for persons with IAT. The maximum DF was limited by pain for
half of participants with IAT, who reported ≥4/10 pain during the weight-bearing calf
stretch (Table 25). It is likely that the magnitude of force on the tendon is related to the
amount of pain during a stretch. The external moment applied during non-weight-bearing
was 12 Nm, compared to the weight-bearing test where much higher external moments
are possible. The combination of higher loads and pain likely contributed to the
significant correlation observed between weight-bearing DF capacity and self-reported
function on the VISA-A, which were not observed with the non-weight-bearing DF test.
99
Thus, limited DF capacity in a weight-bearing position may indicate functional limitation
more than gastrocnemius tightness in persons with IAT.
Performance measures of DF also did not support the clinical idea of
gastrocnemius tightness in participants with IAT. It was anticipated that participants with
IAT would demonstrate a movement pattern during stair ascent that was consistent with
limited DF. However, there was not sufficient evidence to indicate a difference between
groups in stair ascent DF. In addition, the control values obtained in this study
(10.0⁰±3.7) were similar to a previous study by Andriacchi et al. (1980) (10.0⁰±7.6),
which used a similar kinematic model of the ankle . Based on the discomfort reported by
participants with IAT during functional activities that require DF, such as walking, stair
ascent and descent, further analysis across the stance phase of the step task was
pursued.
Another perspective on use of DF during stair ascent is the ability of participants
to move out of DF and into PF. Based on observance of the kinematics during stair
ascent, differences between groups in PF motion were examined during the weight-
acceptance and forward continuance phases of stair ascent. During weight acceptance
7/20 participants with IAT did not achieve a position of PF, and thus started performance
of the task in a position of greater DF. Similarly, during the forward continuance phase
when the ankle is plantar flexing to move the body’s center of mass up and forward,
participants with IAT used almost 50% less PF motion on the involved side (-7.6⁰± 3.8)
compared to the uninvolved side (-13.4⁰±5.5). A lack of PF indicates that participants
with IAT are using a different motor strategy to complete the task of stair ascent. This
altered strategy could be due to plantar flexor weakness (Table 25). Decreased PF
strength could decrease the ability to move into PF and to control the amount of DF,
100
which can increase bony impingement. Alternatively, participants with IAT may choose
not to use the plantar flexors in order to avoid a potentially painful muscle contraction
that would pull on the tendon.
The percentage of end-range DF used during stair ascent also suggests that
persons with IAT had an altered motor strategy. The lower weight-bearing DF capacity
combined with the higher DF during stair ascent (both non-significant trends), resulted in
a significantly higher percentage of end-range DF used during stair ascent on the
involved side of IAT participants compared to the uninvolved side and to controls. The
average difference between groups in the percentage used was relatively small (~10%),
suggesting that most persons with IAT still had ~50% of their available DF capacity that
was not utilized. However, for some individuals with IAT, stair ascent required near end-
range DF ROM. More specifically, 4 individuals with IAT used more than 70% of their
available DF capacity during stair ascent, whereas only 1 control exceeded 70% of their
available DF capacity. The higher percentage of end-range DF utilized on the involved
side was not significantly different compared to the uninvolved side in persons with IAT.
This finding supports the idea that unilateral tendinopathy may result in bilateral motor
deficits (Heales, Lim, Hodges, & Vicenzino, 2013), and the uninvolved side may not
adequately represent a goal for achieving “normal” gait. The negative correlation
between percentage of DF capacity used and function underscores the clinical relevance
of minimizing functional use of end-range DF for patients with IAT (Table 25).
There are multiple factors that could contribute to why end-range DF is painful.
The pull along the length of the tendon and the compression of the tendon and soft-
tissue structures against the posterior-superior surface of the calcaneus are
hypothesized to result in pain (Irwin, 2010). Yet because the average difference between
101
participants with IAT and controls was only about 10% and the peak percentage used
(<80%) was less than end-range, this mechanism was only partially supported in this
study. When extending the analysis to all phases of the step cycle, it became apparent
that IAT resulted in reduced PF, suggesting strategies to restore PF and limit DF may
benefit individuals with IAT. In fact, anecdotally, one participant with IAT reported that
he had started to change the way he walked to reduce end-range DF ROM in order to
minimize symptoms. Interestingly, this participant demonstrated the lowest percentage
DF capacity during stair ascent (9.5%) out of the entire sample. Patients with IAT may
benefit from interventions, such as heel lifts and motor control training, to decrease use
of end-range DF (Table 25). While few participants demonstrated limited DF capacity on
the involved side, stretching may have therapeutic benefit by increasing tolerance to
stretch (Law et al., 2009). As noted above though, stretching into greater DF may have
detrimental effects on tendon characteristics (via impingement) and mechanical
properties (via increased strain). New studies may be used to determine the prognostic
value for various target DF ROM cut offs and the influence of stretching on IAT clinical
variables such as self-report and performance based function.
Plantar flexion strength
This is the first study with findings of PF weakness being associated with IAT and
function. In agreement with previous studies, older age and female gender were
associated with lower PF strength. Therefore, difficulty with functional tasks related to
deficits in PF strength associated with IAT may be further magnified by the effect of
aging and gender.
This is the first study to demonstrate that reduced PF strength is associated with
IAT. The involved side had about 14% (73.5-65.3/73.5, Table 21) lower isometric PF
102
strength than the uninvolved side and nearly 25% (85.9-65.3/85.9, Table 21) lower
isometric PF strength than controls. The results of the current study are consistent with
previous studies on the effect of gender and age on isometric PF strength (Vandervoort
et al., 1992; Vandervoort & McComas, 1986). Each additional year of age was
associated with a 2 Nm decline in PF strength. In addition, female gender was also
associated with a 23 Nm decline in PF strength, and, thus, women with IAT often had
very low PF strength. For example, among the 5 participants with weakness on the
involved side compared to the uninvolved side (difference between sides ≥ 20 Nm), 4
were women. Given that PF strength capacity was associated with function, together
these results indicate that IAT pathology, older age and gender could all contribute to
lower PF strength capacity and functional ability.
Despite lower PF strength capacity, there were no differences between groups in
the peak ankle moment of stair ascent. The magnitude of the ankle moment represents
the amount of force required to prevent the ankle from collapsing into DF. The findings of
the current study are similar to a previous study; the moment for each group in the
current study (Involved side= -127.6 Nm ± 36.2, Uninvolved side= -126.3 Nm ± 23.0,
Controls= -133.8 Nm ± 35.5, units of data converted to match cited study) was within
one SD of values reported by Andriacchi et al. (1980) (-137.2 Nm ± 34.0). This finding is
consistent with the similar degrees of DF used during stair ascent by all groups. Thus
despite weakness on the involved side, participants with IAT placed the same demand
on the plantar flexors during forward continuance as controls.
Reduced ankle power was demonstrated by participants with IAT during stair
ascent, and this reduced PF strength was unique to the involved side. Despite
differences in the ankle kinematic model, the ankle power demonstrated by controls in
103
the current study (3.6 W/kg ± 1.0, Table 23) was similar to the power reported by Alcock
et al. (2014) (3.8 W/kg ± 1.1). This finding is consistent with decreased use of PF motion
during forward continuance phase and decreased isometric plantar flexion strength on
sides with IAT compared to sides without IAT. Because participants with IAT and
controls performed the task at the same cadence, the difference in power between
groups indicated a difference in rate of PF on the involved side. Moreover, due to
reduced ankle power on the involved side, participants with IAT had to compensate by
using increased power by proximal muscles or power by muscles on the uninvolved side
to pull the body mass up the step. Thus, participants with IAT appear to use a pull-up
motor strategy during stair ascent, and this strategy is unique to the involved side.
The variation of impairment in PF strength within the sample facilitated
examination of the relationship between PF strength and self-reported function. Lower
PF strength capacity (r=0.469, P=0.037) and performance (r=0.549, P=0.012) were
moderately associated with lower self-reported function on the VISA-A. Therefore, lower
PF strength in persons with IAT may be severe enough to contribute to difficulty in
performing daily tasks. Alternatively, individuals with IAT may use lower ankle power
during stair ascent in order to reduce loading on the achilles tendon and minimize
symptoms. In either case, these findings support the prescription of PF strengthening in
a pain free manner to improve strength and/or tolerance of muscle use (Table 25).
Model of insertional achilles tendinopathy impairme nts
In addition to tendon characteristics, DF ROM and PF strength impairments were
associated with function in participants with IAT. A higher percentage of end-range DF
used during stair ascent was associated both with greater tendon pathology (lower
echogenicity) and altered mechanical properties (increased strain). These findings
104
indicate that altered tendon characteristics may contribute to altered use of DF ROM
and/or vice versa. Further research is needed to examine if an intervention that
decreases the percentage of end-range DF used during daily tasks could improve
function by not only decreasing position-related symptoms but also minimizing further
tendon degeneration (Table 25).
There was no conclusive evidence of an association between PF strength and
other impairments (DF ROM, tendon characteristics). Theoretically, PF weakness could
result in poor eccentric control of ankle dorsiflexion during daily tasks, which could result
in increased DF and increased pain. Yet although IAT was associated with higher use of
DF capacity and lower ankle power during stair ascent, these two impairments were not
significantly correlated in the current study. Additionally, there was insufficient evidence
to support an association between PF strength and tendon characteristics. This finding
does not support the theory that stronger plantar flexors have better tendon
characteristics (Mahieu, 2006). This is not to state that strengthening cannot improve
function in persons with tendinopathy nor that strengthening cannot alter tendon
characteristics. However, it is likely that PF strength and tendon characteristics
contribute independently to an individual’s self-reported function.
105
Table 25 . Clinical Treatment Recommendations Study Aim (s) Clinical recommendation 1a-b, 4a Tendon characteristics have diagnostic and prognostic value, but more
research is needed to determine objective criteria and cut-off values for ultrasound imaging.
1c, 2a-c Stretching should be prescribed with caution to patients with IAT, because: (1c) Sides with IAT had greater strain than sides without IAT, and thus passive stretches may reinforce abnormal tendon characteristics; (2a-c) The difference in DF capacity and performance between sides with IAT and sides without IAT was not statistically or clinically significant, indicating that gastrocnemius stiffness was not a common impairment in persons with IAT; and (2b) Given that half of the sample with IAT reported pain with a weight-bearing calf stretch, if this stretch is prescribed unnecessarily then it may aggravate symptoms.
2d-f, 4b, 4d Persons with IAT demonstrate an altered movement pattern of increased ankle DF relative to controls, which results in increased risk for bony impingement at the tendon insertion. Greater use of end-range DF is associated with lower function and may contribute to disease progression. This finding supports the prescription of heel lifts and shoewear modification to facilitate ankle plantar flexion.
1d, 2f, 3a, 3c,4c
Strengthening exercises may improve both tendon mechanical properties and function in patients with IAT, because (1d) Sides with IAT had lower tendon stiffness than sides without IAT, and tendon loading with a strengthening program (such as isometric) may be able to restore tendon mechanical properties; and (2f, 3a, 3c, 4c) Persons with IAT commonly have reduced PF strength capacity and performance, which can contribute to decreased function.
Limitations
Both a strength and a limitation of the current study is that this was the first study
to test clinical theories on alterations in achilles tendon properties from ultrasound
imaging, impairments in DF ROM and PF weakness in participants with IAT. More
research is needed to examine the reproducibility (ultrasound measures are new and
have only been demonstrated to have high intra-rater reliability, as opposed to inter-rater
reliability) and generalizability of the findings. Key limitations to this study are the sample
size and cross-sectional design. Many variables and relationships were examined in the
106
current study through a variety of pre-planned and post-hoc analyses in a sample of 20
individuals with chronic unilateral IAT. While the current findings are the highest level of
evidence to date on impairments in persons with IAT, this is study only demonstrates
association and not causation. Given the severity of disability that IAT can cause and the
current low success rate of non-operative interventions, there is high potential for
clinically based diagnostic, prognostic and intervention studies in the future to improve
care for this population. This study provides baseline data to develop clinical studies that
can test the clinical recommendations suggested above. While these clinical
recommendations are based on more evidence than previous clinical theories, more
studies are needed to test the ideas used to inform these recommendations.
Another limitation was that a few participants were unable to complete the
protocol for testing tendon mechanical properties. The current method involved rotating
the ankle to 10º DF, because it was near end-range yet also comfortable for pilot
subjects with IAT. However, one participant in the current study was unable to achieve
10⁰ DF and two participants had difficulty relaxing throughout this motion. Different
methods could examine mechanical properties throughout the individual’s available DF
motion, rather than a set range of motion for all participants. Additionally, estimates of
total muscle-tendon unit elongation are difficult. Error could result from poor stabilization
of the foot in the heel cup of the isokinetic device and/or skin artifact. These potential
errors were minimized in the current study by examining tendon mechanical properties
over a relatively small range of motion. Another limitation of the current study methods
is that they assess tendon mechanical properties of the entire tendon, rather than locally
at the tendon insertion. Because the tendon functions as a unit during exercise and daily
activities, the current findings are clinically applicable. However, development of
107
methodology specific to the tendon insertion may have greater sensitivity to detecting
abnormal mechanical properties among individuals with IAT. Finally, the ultrasound
imaging methods used to assess tendon characteristics and mechanical properties were
reliable and able to detect differences between sides with IAT and sides without IAT.
The cut-off values provided here establish baseline data to build toward development of
diagnostic and prognostic criteria. More research is needed to examine the intra-rater
reliability and clinical feasibility of these measures.
A limitation in the measure of DF ROM and PF strength capacities is that pain
tolerance also influences these measures. For the measures of weight-bearing DF and
PF strength capacity, half and one fifth (respectively) of the sample reported pain ≥ 4/10.
Therefore, it is unknown if limited capacity was due to impairment or pain tolerance.
Capacity for DF was also measured in non-weight-bearing, which was able to minimize
the effect of pain tolerance; however, there was only one measure of PF strength
capacity. While the effect of pain on capacity is challenging to assess statistically on a
group level, a physical therapist can incorporate the potential influence of pain into
clinical evaluation of impairment for an individual patient.
A limitation of the current study is that it was focused on impairments and altered
motor strategy at the ankle joint. Further research is needed to understand how
impairments associated with IAT impact the function of more proximal joints that have to
compensate for reduced ankle function. The findings of the current study may help
guide the development of exercises for the foot and ankle specific to IAT patients, but
still neglects the assessment of impairments that may occur more proximally in patients
with IAT. Also, in terms of functional performance this study was focused on the
transitional step from the ground to stair ascent. Further research is need to examine if
108
the altered movement patterns identified in the current study similarly affect other
common daily tasks, such as gait, continuous stair ascent and stair descent.
Statistical Models
In the physical therapy literature, case-control study hypotheses are commonly
tested using t-tests or ANCOVA, since interpretation of these statistics are commonly
taught in an evidence-based curriculum. However, other statistical techniques, such as
GEE, may be more appropriate and yield higher power to detect significant differences.
Both a GEE and ANCOVA analyses were performed in the current study, which allowed
for a direct comparison of the two statistical models. The GEE had several advantages
over an ANCOVA. First of all, the GEE analysis was able to use all data when testing if
sides with IAT (involved side of IAT) differed from sides without IAT (uninvolved side IAT
and both sides of controls). In contrast, a one-way ANCOVA (1x3: involved side IAT,
uninvolved side IAT, matched side control) would violate the assumption that the
variables are independent and only includes one side of controls in the analysis. A two-
way ANCOVA (2x2: independent variables= side (involved, uninvolved) and group (IAT,
controls) could include both sides of the control, but the definition of side is violated by
the control group which does not have an “involved” side. Therefore, two separate
analyses were needed to compare: 1) sides (involved side vs. uninvolved side in
participants with IAT, and 2) groups (IAT vs. controls). Yet, this separation of analyses in
the ANCOVA was advantageous when examining if significant differences in the GEE
analysis is due to differences between sides, groups or both. Another advantage to the
GEE analysis was the statistical power to examine the effect of demographic covariates
on the dependent variable. In contrast, demographic covariates that were statistically
significant in the GEE analysis, were commonly non-significant in the ANCOVA analysis
109
(see Tables 14, 19, 21, 23). This indicates that use of ANCOVA may fail to detect when
demographic covariates are affecting the dependent variable. Given the relatively small
sample size (N<50) that is common in physical therapy literature, the GEE analysis may
be a useful means of maximizing use of data to detect statistically significant findings.
Conclusions
This study confirms that ultrasound imaging is able to characterize tendon
diameter changes and altered composition. The strong association between altered
composition and function merits the further evaluation of echogenicity. This ultrasound
imaging measure is a relatively quick and cheap measurement that could augment
current clinical evaluation. This is the first study to demonstrate that alterations in tendon
mechanical properties are associated with IAT. This is consistent with previous studies
documenting abnormal tendon mechanical properties in participants with midportion
achilles tendinopathy. This is also the first study to demonstrate a correlation between
tendon characteristics and DF ROM, indicating that these impairments are inter-related.
Future intervention studies designed to influence DF ROM in persons with achilles
tendinopathy may be enriched by also examining the effect on tendon characteristics
and vice versa.
Previous researchers have not examined the link between measures of DF ROM,
PF strength and self-reported function. These data improve the foundation for future
clinical trials to examine prognostic factors and plan intervention studies for persons with
IAT. For example, contrary to common clinical practice of prescribing calf stretches to
patients with IAT, limited DF ROM capacity was not a common impairment in
participants with IAT. However, DF ROM was correlated to function indicating some level
110
of DF ROM was necessary. The role of calf stretching should be evaluated for persons
with IAT in future studies including its effect on pain and tendon characteristics.
Finally, this is the first study to demonstrate motor deficits in persons with IAT.
Participants with IAT were more likely to use a greater percentage of DF capacity,
remain in a relatively dorsiflexed ankle position, and exhibit lower ankle power during
stair ascent. Therefore daily activities may aggravate symptoms due to greater and
prolonged used of DF (contributing to impingement) and due to decreased use of the
plantar flexors (contributing to weakness). These findings support the importance of
incorporating strategies to improve impairments in DF ROM and PF strength capacity
with function.
111
References Alcock, L., O'Brien, T. D., & Vanicek, N. (2014). Biomechanical demands differentiate
transitioning vs. continuous stair ascent gait in older women. Clinical Biomechanics (Bristol, Avon), 29(1), 111-118. doi: 10.1016/j.clinbiomech.2013.10.007
Anderson, J. A., Suero, E., O'Loughlin, P. F., & Kennedy, J. G. (2008). Surgery for retrocalcaneal bursitis: a tendon-splitting versus a lateral approach. Clinical Orthopaedics and Related Research, 466(7), 1678-1682. doi: 10.1007/s11999-008-0281-9
Andriacchi, T. P., Andersson, G. B., Fermier, R. W., Stern, D., & Galante, J. O. (1980). A study of lower-limb mechanics during stair-climbing. The Journal of bone and joint surgery.American volume, 62(5), 749-757.
Arampatzis, A., Karamanidis, K., & Albracht, K. (2007). Adaptational responses of the human Achilles tendon by modulation of the applied cyclic strain magnitude. Journal of Experimental Biology, 210(Pt 15), 2743-2753. doi: 10.1242/jeb.003814
Arampatzis, A., Peper, A., Bierbaum, S., & Albracht, K. (2010). Plasticity of human Achilles tendon mechanical and morphological properties in response to cyclic strain. Journal of Biomechanics, 43(16), 3073-3079. doi: 10.1016/j.jbiomech.2010.08.014
Archambault, J. M., Wiley, J. P., Bray, R. C., Verhoef, M., Wiseman, D. A., & Elliott, P. D. (1998). Can sonography predict the outcome in patients with achillodynia? Journal of Clinical Ultrasound, 26(7), 335-339.
Arya, S., & Kulig, K. (2010). Tendinopathy alters mechanical and material properties of the Achilles tendon. Journal of Applied Physiology, 108(3), 670-675. doi: 10.1152/japplphysiol.00259.2009
Astrom, M., Gentz, C. F., Nilsson, P., Rausing, A., Sjoberg, S., & Westlin, N. (1996). Imaging in chronic achilles tendinopathy: a comparison of ultrasonography, magnetic resonance imaging and surgical findings in 27 histologically verified cases. Skeletal Radiology, 25(7), 615-620.
Bennell, K. L., Talbot, R. C., Wajswelner, H., Techovanich, W., Kelly, D. H., & Hall, A. J. (1998). Intra-rater and inter-rater reliability of a weight-bearing lunge measure of ankle dorsiflexion. The Australian journal of physiotherapy, 44(3), 175-180.
Brunner, J., Anderson, J., O'Malley, M., Bohne, W., Deland, J., & Kennedy, J. (2005). Physician and patient based outcomes following surgical resection of Haglund's deformity. Acta Orthopaedica Belgica, 71(6), 718-723.
Chauveaux, D., Liet, P., Le Huec, J. C., & Midy, D. (1991). A new radiologic measurement for the diagnosis of Haglund's deformity. Surgical and radiologic anatomy : SRA, 13(1), 39-44.
Child, S., Bryant, A. L., Clark, R. A., & Crossley, K. M. (2010). Mechanical properties of the achilles tendon aponeurosis are altered in athletes with achilles tendinopathy. American Journal of Sports Medicine, 38(9), 1885-1893. doi: 10.1177/0363546510366234
112
Chimenti, R., Tome, J., Flemister, A., & Houck, J. (2014). Use of end-range dorsiflexion range of motion in people with insertional achilles tendinopathy. Paper presented at the American Physical Therapy Association Combined Sections Meeting, Las Vegas, NV.
Chimenti, R., Tome, J., Flemister, A. S., & Houck, J. (2012). Achilles Tendon Elongation during Passive and Active Conditions in People with Insertional Achilles Tendinopathy. Section on Research Retreat, Java City, NY.
Chimenti, R., Tome, J., Flemister, A. S., & Houck, J. (2014). Tendon characteristics and mechanical properties in people with Insertional Achilles Tendinopathy. Paper presented at the Platform presentation at: American Physical Therapy Association Combined Sections Meeting, Las Vegas, NV.
Chizewski, M. G., & Chiu, L. Z. (2012). Contribution of calcaneal and leg segment rotations to ankle joint dorsiflexion in a weight-bearing task. Gait and Posture, 36(1), 85-89. doi: 10.1016/j.gaitpost.2012.01.007; 10.1016/j.gaitpost.2012.01.007
Collinger, J. L., Fullerton, B., Impink, B. G., Koontz, A. M., & Boninger, M. L. (2010). Validation of grayscale-based quantitative ultrasound in manual wheelchair users: relationship to established clinical measures of shoulder pathology. American Journal of Physical Medicine & Rehabilitation / Association of Academic Physiatrists, 89(5), 390-400. doi: 10.1097/PHM.0b013e3181d8a238
Collinger, J. L., Gagnon, D., Jacobson, J., Impink, B. G., & Boninger, M. L. (2009). Reliability of quantitative ultrasound measures of the biceps and supraspinatus tendons. Academic Radiology, 16(11), 1424-1432. doi: 10.1016/j.acra.2009.05.001
Cook, J. L., & Purdam, C. R. (2009). Is tendon pathology a continuum? A pathology model to explain the clinical presentation of load-induced tendinopathy. British Journal of Sports Medicine, 43(6), 409-416. doi: 10.1136/bjsm.2008.051193
Crawford, T., & Desruisseau, R. (2010). MRI Reveals Cause of Heel Pain. Retrieved May 25, 2012, from http://www.diagnosticimaging.com/case-studies/content/article/113619/1695359
de Palma, L., Marinelli, M., Meme, L., & Pavan, M. (2004). Immunohistochemistry of the enthesis organ of the human Achilles tendon. Foot & ankle international / American Orthopaedic Foot and Ankle Society [and] Swiss Foot and Ankle Society, 25(6), 414-418.
De Zordo, T., Chhem, R., Smekal, V., Feuchtner, G., Reindl, M., Fink, C., . . . Klauser, A. S. (2010). Real-time sonoelastography: findings in patients with symptomatic achilles tendons and comparison to healthy volunteers. Ultraschall in der Medizin, 31(4), 394-400. doi: 10.1055/s-0028-1109809
Denegar, C. R., Hertel, J., & Fonseca, J. (2002). The effect of lateral ankle sprain on dorsiflexion range of motion, posterior talar glide, and joint laxity. Journal of Orthopaedic and Sports Physical Therapy, 32(4), 166-173. doi: 10.2519/jospt.2002.32.4.166
DeVries, J. G., Summerhays, B., & Guehlstorf, D. W. (2009). Surgical correction of Haglund's triad using complete detachment and reattachment of the Achilles tendon. The Journal of foot and ankle surgery : official publication of the
113
American College of Foot and Ankle Surgeons, 48(4), 447-451. doi: 10.1053/j.jfas.2009.03.004
DiGiovanni, C. W., Kuo, R., Tejwani, N., Price, R., Hansen, S. T., Jr., Cziernecki, J., & Sangeorzan, B. J. (2002). Isolated gastrocnemius tightness. The Journal of bone and joint surgery.American volume, 84-A(6), 962-970.
Fahlstrom, M., Jonsson, P., Lorentzon, R., & Alfredson, H. (2003). Chronic Achilles tendon pain treated with eccentric calf-muscle training. Knee Surgery, Sports Traumatology, Arthroscopy, 11(5), 327-333. doi: 10.1007/s00167-003-0418-z
Falsetti, P., Frediani, B., Acciai, C., Baldi, F., Filippou, G., Prada, E. P., . . . Marcolongo, R. (2004). Ultrasonographic study of Achilles tendon and plantar fascia in chondrocalcinosis. The Journal of rheumatology, 31(11), 2242-2250.
Fowler, A., & Philip, J. F. (1945). Abnormality of the calcaneus as a cause of painful heel its diagnosis and operative treatment. British Journal of Surgery, 32(128), 494-498. doi: 10.1002/bjs.18003212812
Fredberg, U., & Bolvig, L. (2002). Significance of ultrasonographically detected asymptomatic tendinosis in the patellar and achilles tendons of elite soccer players: a longitudinal study. American Journal of Sports Medicine, 30(4), 488-491.
Grieve, D. W., Cavanagh, P. R., & Pheasant, S. (1978). Prediction of gastrocnemius length from knee and ankle position. Biomechanics, VI-A, 405-412.
Heales, L. J., Lim, E. C., Hodges, P. W., & Vicenzino, B. (2013). Sensory and motor deficits exist on the non-injured side of patients with unilateral tendon pain and disability--implications for central nervous system involvement: a systematic review with meta-analysis. British Journal of Sports Medicine. doi: 10.1136/bjsports-2013-092535
Hoppenfeld, S., & Hutton, R. (1976). Physical examination of the spine and extremities. New York: Appleton-Century-Crofts.
Houck, J. R., Neville, C., Tome, J., & Flemister, A. S. (2009). Foot kinematics during a bilateral heel rise test in participants with stage II posterior tibial tendon dysfunction. The Journal of orthopaedic and sports physical therapy, 39(8), 593-603. doi: 10.2519/jospt.2009.3040
Houck, J. R., Tome, J. M., & Nawoczenski, D. A. (2008). Subtalar neutral position as an offset for a kinematic model of the foot during walking. Gait & posture, 28(1), 29-37. doi: 10.1016/j.gaitpost.2007.09.008
Hu, C. T., & Flemister, A. S. (2008). Insertional Achilles Tendinopathy: Surgical Options. Foot and Ankle Surgeries: Operative Techniques and Evidence Based Outcomes, 18(4), 247-253. doi: 10.1053/j.oto.2009.02.002
Irwin, T. A. (2010). Current concepts review: insertional achilles tendinopathy. Foot & ankle international / American Orthopaedic Foot and Ankle Society [and] Swiss Foot and Ankle Society, 31(10), 933-939. doi: 10.3113/FAI.2010.0933
Johnson, Zalavras, C., & Thordarson, D. B. (2006). Surgical management of insertional calcific achilles tendinosis with a central tendon splitting approach. Foot and Ankle International, 27(4), 245-250.
114
Johnson, K. A., & Strom, D. E. (1989). Tibialis posterior tendon dysfunction. Clinical Orthopaedics and Related Research, (239)(239), 196-206.
Jonsson, P., Alfredson, H., Sunding, K., Fahlstrom, M., & Cook, J. (2008). New regimen for eccentric calf-muscle training in patients with chronic insertional Achilles tendinopathy: results of a pilot study. British Journal of Sports Medicine, 42(9), 746-749. doi: 10.1136/bjsm.2007.039545
Karjalainen, P. T., Soila, K., Aronen, H. J., Pihlajamaki, H. K., Tynninen, O., Paavonen, T., & Tirman, P. F. (2000). MR imaging of overuse injuries of the Achilles tendon. American journal of roentgenology, 175(1), 251-260.
Kawakami, Y., Kanehisa, H., & Fukunaga, T. (2008). The relationship between passive ankle plantar flexion joint torque and gastrocnemius muscle and achilles tendon stiffness: implications for flexibility. Journal of Orthopaedic and Sports Physical Therapy, 38(5), 269-276. doi: 10.2519/jospt.2008.2632
Kendall, F. P., Wadsworth, G. E., Kendall, H. O., & McCreary, E. K. (1983). Muscles, testing and function (Vol. 3). Baltimore: Williams & Wilkins.
Khan, K. M., Cook, J. L., Bonar, F., Harcourt, P., & Astrom, M. (1999). Histopathology of common tendinopathies. Update and implications for clinical management. Sports Medicine, 27(6), 393-408.
Khan, K. M., Forster, B. B., Robinson, J., Cheong, Y., Louis, L., Maclean, L., & Taunton, J. E. (2003). Are ultrasound and magnetic resonance imaging of value in assessment of Achilles tendon disorders? A two year prospective study. British Journal of Sports Medicine, 37(2), 149-153.
Kim, P. J., Peace, R., Mieras, J., Thoms, T., Freeman, D., & Page, J. (2011). Interrater and intrarater reliability in the measurement of ankle joint dorsiflexion is independent of examiner experience and technique used. Journal of the American Podiatric Medical Association, 101(5), 407-414.
Kistler. (1984). Multicomponent Measuring Force Plate for Biomechanics and Industry Type 9287.
Klauser, A. S., Faschingbauer, R., & Jaschke, W. R. (2010). Is sonoelastography of value in assessing tendons? Seminars in musculoskeletal radiology, 14(3), 323-333. doi: 10.1055/s-0030-1254521
Klauser, A. S., Miyamoto, H., Tamegger, M., Faschingbauer, R., Moriggl, B., Klima, G., . . . Jaschke, W. R. (2013). Achilles tendon assessed with sonoelastography: histologic agreement. Radiology, 267(3), 837-842. doi: 10.1148/radiol.13121936
Konor, M. M., Morton, S., Eckerson, J. M., & Grindstaff, T. L. (2012). Reliability of three measures of ankle dorsiflexion range of motion. International Journal of Sports Physical Therapy, 7(3), 279-287.
Krause, D. A., Cloud, B. A., Forster, L. A., Schrank, J. A., & Hollman, J. H. (2011). Measurement of ankle dorsiflexion: a comparison of active and passive techniques in multiple positions. Journal of sport rehabilitation, 20(3), 333-344.
Kujala, U. M., Sarna, S., & Kaprio, J. (2005). Cumulative incidence of achilles tendon rupture and tendinopathy in male former elite athletes. Clinical journal of sport medicine : official journal of the Canadian Academy of Sport Medicine, 15(3), 133-135.
115
Kvist, M. (1991). Achilles tendon injuries in athletes. Annales Chirurgiae et Gynaecologiae, 80(2), 188-201.
Law, R. Y., Harvey, L. A., Nicholas, M. K., Tonkin, L., De Sousa, M., & Finniss, D. G. (2009). Stretch exercises increase tolerance to stretch in patients with chronic musculoskeletal pain: a randomized controlled trial. Physical Therapy, 89(10), 1016-1026. doi: 10.2522/ptj.20090056
Maganaris, C. N., Narici, M. V., & Maffulli, N. (2008). Biomechanics of the Achilles tendon. Disability and Rehabilitation, 30(20-22), 1542-1547. doi: 10.1080/09638280701785494
Magnusson, S. P., Hansen, P., Aagaard, P., Brond, J., Dyhre-Poulsen, P., Bojsen-Moller, J., & Kjaer, M. (2003). Differential strain patterns of the human gastrocnemius aponeurosis and free tendon, in vivo. Acta Physiologica Scandinavica, 177(2), 185-195.
Mahieu. (2006). Intrinsic risk factors for the development of achilles tendon overuse injury: a prospective study. The American Journal of Sports Medicine, 34(2), 226-235. doi: 10.1177/0363546505279918
Mahieu, N. N., McNair, P., Cools, A., D'Haen, C., Vandermeulen, K., & Witvrouw, E. (2008). Effect of eccentric training on the plantar flexor muscle-tendon tissue properties. Medicine and Science in Sports and Exercise, 40(1), 117-123. doi: 10.1249/mss.0b013e3181599254
Mahieu., Witvrouw, E., Stevens, V., Willems, T., Vanderstraeten, G., & Cambier, D. (2004). Test-retest reliability of measuring the passive stiffness of the achilles tendon using ultrasonography. Isokinetics and Exercise Science, 12, 185-191.
Maletsky, L. P., Sun, J., & Morton, N. A. (2007). Accuracy of an optical active-marker system to track the relative motion of rigid bodies. Journal of Biomechanics, 40(3), 682-685. doi: 10.1016/j.jbiomech.2006.01.017
Matsumoto, F., Trudel, G., Uhthoff, H. K., & Backman, D. S. (2003). Mechanical effects of immobilization on the Achilles' tendon. Archives of Physical Medicine and Rehabilitation, 84(5), 662-667.
McFadyen, B. J., & Winter, D. A. (1988). An integrated biomechanical analysis of normal stair ascent and descent. Journal of Biomechanics, 21(9), 733-744.
McGarvey, W. C., Palumbo, R. C., Baxter, D. E., & Leibman, B. D. (2002). Insertional Achilles tendinosis: surgical treatment through a central tendon splitting approach. Foot and Ankle International, 23(1), 19-25.
Morrissey, D., Roskilly, A., Twycross-Lewis, R., Isinkaye, T., Screen, H., Woledge, R., & Bader, D. (2011). The effect of eccentric and concentric calf muscle training on Achilles tendon stiffness. Clinical Rehabilitation, 25(3), 238-247. doi: 10.1177/0269215510382600
Moseley, A. M., Crosbie, J., & Adams, R. (2001). Normative data for passive ankle plantarflexion--dorsiflexion flexibility. Clinical Biomechanics (Bristol, Avon), 16(6), 514-521.
Moseley, A. M., Crosbie, J., & Adams, R. (2003). High- and low-ankle flexibility and motor task performance. Gait and Posture, 18(2), 73-80.
116
Movin, T., Gad, A., Reinholt, F. P., & Rolf, C. (1997). Tendon pathology in long-standing achillodynia. Biopsy findings in 40 patients. Acta Orthopaedica Scandinavica, 68(2), 170-175.
Muir, I. W., Chesworth, B. M., & Vandervoort, A. A. (1999). Effect of a static calf-stretching exercise on the resistive torque during passive ankle dorsiflexion in healthy subjects. The Journal of orthopaedic and sports physical therapy, 29(2), 106-113; discussion 114-105.
Munteanu, S. E., Strawhorn, A. B., Landorf, K. B., Bird, A. R., & Murley, G. S. (2009). A weightbearing technique for the measurement of ankle joint dorsiflexion with the knee extended is reliable. Journal of science and medicine in sport / Sports Medicine Australia, 12(1), 54-59. doi: 10.1016/j.jsams.2007.06.009
Muramatsu, T., Muraoka, T., Takeshita, D., Kawakami, Y., Hirano, Y., & Fukunaga, T. (2001). Mechanical properties of tendon and aponeurosis of human gastrocnemius muscle in vivo. Journal of applied physiology (Bethesda, Md.: 1985), 90(5), 1671-1678.
Muraoka, T., Muramatsu, T., Takeshita, D., Kawakami, Y., & Fukunaga, T. (2002). Length change of human gastrocnemius aponeurosis and tendon during passive joint motion. Cells, tissues, organs, 171(4), 260-268.
Nicholson, C. W., Berlet, G. C., & Lee, T. H. (2007). Prediction of the success of nonoperative treatment of insertional Achilles tendinosis based on MRI. Foot and Ankle International, 28(4), 472-477. doi: 10.3113/FAI.2007.0472
Oh-Park, M., Wang, C., & Verghese, J. (2011). Stair negotiation time in community-dwelling older adults: normative values and association with functional decline. Archives of Physical Medicine and Rehabilitation, 92(12), 2006-2011. doi: 10.1016/j.apmr.2011.07.193
Rees, J. D., Wilson, A. M., & Wolman, R. L. (2006). Current concepts in the management of tendon disorders. Rheumatology (Oxford, England), 45(5), 508-521. doi: 10.1093/rheumatology/kel046
Reeves, N. D., Spanjaard, M., Mohagheghi, A. A., Baltzopoulos, V., & Maganaris, C. N. (2008). The demands of stair descent relative to maximum capacities in elderly and young adults. Journal of electromyography and kinesiology : official journal of the International Society of Electrophysiological Kinesiology, 18(2), 218-227. doi: 10.1016/j.jelekin.2007.06.003
Robinson, J. M., Cook, J. L., Purdam, C., Visentini, P. J., Ross, J., Maffulli, N., . . . Victorian Institute Of Sport Tendon Study, G. (2001). The VISA-A questionnaire: a valid and reliable index of the clinical severity of Achilles tendinopathy. British Journal of Sports Medicine, 35(5), 335-341.
Rompe, J. D., Furia, J., & Maffulli, N. (2009). Eccentric loading versus eccentric loading plus shock-wave treatment for midportion achilles tendinopathy: a randomized controlled trial. American Journal of Sports Medicine, 37(3), 463-470. doi: 10.1177/0363546508326983
Sadeghi, H., Allard, P., Prince, F., & Labelle, H. (2000). Symmetry and limb dominance in able-bodied gait: a review. Gait and Posture, 12(1), 34-45.
117
Saltzman, C. L., & Tearse, D. S. (1998). Achilles tendon injuries. The Journal of the American Academy of Orthopaedic Surgeons, 6(5), 316-325.
Schneider, W., Niehus, W., & Knahr, K. (2000). Haglund's syndrome: disappointing results following surgery -- a clinical and radiographic analysis. Foot & ankle international / American Orthopaedic Foot and Ankle Society [and] Swiss Foot and Ankle Society, 21(1), 26-30.
Sconfienza, L. M., Silvestri, E., & Cimmino, M. A. (2010). Sonoelastography in the evaluation of painful Achilles tendon in amateur athletes. Clinical and Experimental Rheumatology, 28(3), 373-378.
Sidaway, B., Euloth, T., Caron, H., Piskura, M., Clancy, J., & Aide, A. (2012). Comparing the reliability of a trigonometric technique to goniometry and inclinometry in measuring ankle dorsiflexion. Gait and Posture, 36(3), 335-339. doi: 10.1016/j.gaitpost.2012.01.019
Silbernagel, K. G., Gustavsson, A., Thomee, R., & Karlsson, J. (2006). Evaluation of lower leg function in patients with Achilles tendinopathy. Knee surgery, sports traumatology, arthroscopy : official journal of the ESSKA, 14(11), 1207-1217. doi: 10.1007/s00167-006-0150-6
Sofka, C. M., Adler, R. S., Positano, R., Pavlov, H., & Luchs, J. S. (2006). Haglund's syndrome: diagnosis and treatment using sonography. HSS journal : the musculoskeletal journal of Hospital for Special Surgery, 2(1), 27-29. doi: 10.1007/s11420-005-0129-8
Soila, K., Karjalainen, P. T., Aronen, H. J., Pihlajamaki, H. K., & Tirman, P. J. (1999). High-resolution MR imaging of the asymptomatic Achilles tendon: new observations. American journal of roentgenology, 173(2), 323-328.
Stenroth, L., Peltonen, J., Cronin, N. J., Sipila, S., & Finni, T. (2012). Age-related differences in Achilles tendon properties and triceps surae muscle architecture in vivo. Journal of Applied Physiology, 113(10), 1537-1544. doi: 10.1152/japplphysiol.00782.2012
Ter Haar, G. (2011). Ultrasonic imaging: safety considerations. Interface focus, 1(4), 686-697. doi: 10.1098/rsfs.2011.0029
Theis, N., Mohagheghi, A. A., & Korff, T. (2012). Method and strain rate dependence of Achilles tendon stiffness. Journal of electromyography and kinesiology : official journal of the International Society of Electrophysiological Kinesiology. doi: 10.1016/j.jelekin.2012.06.004
Thomas, J. L., Christensen, J. C., Kravitz, S. R., Mendicino, R. W., Schuberth, J. M., Vanore, J. V., . . . Baker, J. (2010). The diagnosis and treatment of heel pain: a clinical practice guideline-revision 2010. The Journal of foot and ankle surgery : official publication of the American College of Foot and Ankle Surgeons, 49(3 Suppl), S1-19. doi: 10.1053/j.jfas.2010.01.001
Vandervoort, A. A., Chesworth, B. M., Cunningham, D. A., Paterson, D. H., Rechnitzer, P. A., & Koval, J. J. (1992). Age and sex effects on mobility of the human ankle. Journal of Gerontology, 47(1), M17-21.
118
Vandervoort, A. A., & McComas, A. J. (1986). Contractile changes in opposing muscles of the human ankle joint with aging. Journal of applied physiology (Bethesda, Md.: 1985), 61(1), 361-367.
Verrall, G., Schofield, S., & Brustad, T. (2011). Chronic Achilles tendinopathy treated with eccentric stretching program. Foot and Ankle International, 32(9), 843-849.
Waldecker, U., Hofmann, G., & Drewitz, S. (2012). Epidemiologic investigation of 1394 feet: coincidence of hindfoot malalignment and Achilles tendon disorders. Foot and Ankle Surgery, 18(2), 119-123. doi: 10.1016/j.fas.2011.04.007
Wiegerinck, J. I., Kerkhoffs, G. M., van Sterkenburg, M. N., Sierevelt, I. N., & van Dijk, C. N. (2013). Treatment for insertional Achilles tendinopathy: a systematic review. Knee Surgery, Sports Traumatology, Arthroscopy, 21(6), 1345-1355. doi: 10.1007/s00167-012-2219-8
Williams, C. M., Caserta, A. J., & Haines, T. P. (2013). The TiltMeter app is a novel and accurate measurement tool for the weight bearing lunge test. Journal of Science and Medicine in Sport, 16(5), 392-395. doi: 10.1016/j.jsams.2013.02.001
Williamson, A., & Hoggart, B. (2005). Pain: a review of three commonly used pain rating scales. Journal of Clinical Nursing, 14(7), 798-804. doi: 10.1111/j.1365-2702.2005.01121.x
Winter, S. L., & Challis, J. H. (2008). Reconstruction of the human gastrocnemius force-length curve in vivo: part 2-experimental results. Journal of Applied Biomechanics, 24(3), 207-214.
Yodlowski, M. L., Scheller, A. D., Jr., & Minos, L. (2002). Surgical treatment of Achilles tendinitis by decompression of the retrocalcaneal bursa and the superior calcaneal tuberosity. The American Journal of Sports Medicine, 30(3), 318-321.
Zanetti, M., Metzdorf, A., Kundert, H. P., Zollinger, H., Vienne, P., Seifert, B., & Hodler, J. (2003). Achilles tendons: clinical relevance of neovascularization diagnosed with power Doppler US. Radiology, 227(2), 556-560. doi: 10.1148/radiol.2272012069
119
Appendix I: IRB approval and study measures
120
121
122
123
Study Subject Intake Sheet:
Date:__________________
Race/Ethnicity_______________________
Height____________________
Involved side_____Right _______Left
NRS, worst:_____________
NRS, best:______________
Date of start of achilles tendinopathy sx________________________
Type/ Duration of rehabilitation________________________________________________________
_________________________________________________________________________________
What treatments have you had?
Injections?_____ How many?___________ When was your last injection?__________
Night splints?_____
Inserts?_______
Heel lifts?__________
Stretching?_______ What type?__________
Formal physical therapy?____________
Type/ Duration Exercise?_____________________________________________________________________
__________________________________________________________________________________________
General Health Information: (e.g. hypertension, diabetes, neurological problems)_______________________________
_________________________________________________________________________________________
Medications:______________________________________________________________
124
125
126
127
Appendix II: Foot & ankle publications
128
The Foot and Ankle: Physical Therapy Patient Management Utilizing Current Evidence. Houck J, Neville C, Chimenti R . This monograph is part of the Orthopaedic Section Independent Study Course series 21.2, Current Concepts for Orthopaedic Physical Therapy, 3rd edition, 2011.
Static and Dynamic Assessment of Foot Posture after Lateral Column Lengthening Procedure. Barske H, Chimenti R , Tome J, Martin E, Flemister AS, Houck J. Foot Ankle Int. 2013 May;34(5):673-83. doi: 10.1177/1071100712471662. Epub 2013 Jan 25. Introduction : After LCL, the medial longitudinal arch is restored radiographically. First metatarsal dorsiflexion (1st 10 MT DF) has not yet been used to examine the impact of LCL on foot function during gait. Methods : Thirteen patients with a stage II flatfoot who had undergone unilateral LCL surgery, and 13 matched control subjects completed self-reported pain and functional scales and well as a clinical exam. Foot kinematic data were collected during gait using three dimensional motion analysis techniques. A custom force transducer was used to establish the maximum passive range of motion of 1st 16 MT DF at 40 n of force. Results : Most patients continued to report pain and dysfunction postoperatively. On average, the talus-first metatarsal angle decreased by 19.6º on the lateral view after surgery. During gait, sagittal plane motion of the first metatarsal was 2.8º less on the operated than the non-operated side. For the operated side, the relative amount of 1st MT DF during gait reached 187.8 % of maximum passive motion, while the non-operated side reached 250.2 % of the maximum. During midstance, 67% of subjects on the involved side and 8% of subjects on the uninvolved side were below maximum passive 1st MT DF. Conclusion : Patients undergoing LCL for correction of Stage II AAFD experience mixed outcomes and similar foot kinematics as the uninvolved limb despite radiographic correction of deformity. The surgery appears to have a larger effect in terms of the relative range at which the 1st MT DF motion is occurring. The operated side had 62.4% lower peak value than the non-operated side. Postoperatively, patients displayed a decreased duration of time in a low arch posture during stance. Longitudinal data are necessary to make a more valid comparison of the effects of surgical correction measured using radiographs and dynamic foot posture during gait.
129
Foot and Ankle Kinematics of Adult-Acquired Flatfoot Deformity and Age on Single Limb Heel Rise Test Chimenti, RL. Tome, J. Flemister, A.S. Houck, J.R. J Orthop Phys Ther. (In press 2014) Study design: Cross-sectional, laboratory study Objective: To compare single limb heel rise performance and foot/ankle kinematics between persons with stage II Adult-Acquired Flatfoot Deformity (AAFD) and healthy controls. Background: The inability to perform a single limb heel rise is considered a functional diagnostic test for AAFD. However, it is unknown what foot motions contribute to poor performance of this task. Methods: Fifty individuals participated in this study (20 with stage II AAFD, age=57.6±11.3 years; 15 older controls, age=56.8± 5.3 years; 15 younger controls, age=22.2± 2.4 years). Forefoot (sagittal plane) and rearfoot (sagittal and frontal plane) kinematics were collected using a three dimensional motion analysis system. Heel rise performance (heel height) and kinematics (joint angles, excursions) were evaluated. One-way and two-way ANOVA’s were used to examine differences in heel rise performance and kinematics between groups. Results: Lower heel height was associated with AAFD and older age (P<0.001). Persons with AAFD demonstrated higher degrees of first metatarsal dorsiflexion (P<0.001), lower ankle plantar flexion (P<0.001) and higher subtalar eversion (P=0.027) than older controls. Persons with AAFD demonstrated lower ankle (P<0.001) and first metatarsal (P<0.001) excursions than older controls, but no difference in subtalar excursion (P=0.771). Conclusions: Persons with stage II AAFD did not achieve sufficient heel height during a single leg heel rise. Both forefoot and rearfoot kinematics in the sagittal plane, not frontal plane, contributed to the lower heel height in participants with stage II AAFD.
