Altered Tendon Characteristics and Mechanical Properties Associated with Insertional Achilles ...

12 Pages·2016·0.77 MB·English

Digital Commons @ George Fox University
Faculty Publications - School of Physical Therapy School of Physical Therapy
2014
Altered Tendon Characteristics and Mechanical
Properties Associated with Insertional Achilles
Tendinopathy
Ruth L. Chimenti
Adolph S. Flemister
Joshua Tome
James M. McMahon
Marie A. Flannery
See next page for additional authors
Follow this and additional works at:https://digitalcommons.georgefox.edu/pt_fac
Part of thePhysical Therapy Commons
Recommended Citation
Chimenti, Ruth L.; Flemister, Adolph S.; Tome, Joshua; McMahon, James M.; Flannery, Marie A.; Xue, Ying; and Houck, Jeff R.,
"Altered Tendon Characteristics and Mechanical Properties Associated with Insertional Achilles Tendinopathy" (2014).Faculty
Publications - School of Physical Therapy. 51.
https://digitalcommons.georgefox.edu/pt_fac/51
This Article is brought to you for free and open access by the School of Physical Therapy at Digital Commons @ George Fox University. It has been
accepted for inclusion in Faculty Publications - School of Physical Therapy by an authorized administrator of Digital Commons @ George Fox
University. For more information, please [email protected].
Authors
Ruth L. Chimenti, Adolph S. Flemister, Joshua Tome, James M. McMahon, Marie A. Flannery, Ying Xue, and
Jeff R. Houck
This article is available at Digital Commons @ George Fox University:https://digitalcommons.georgefox.edu/pt_fac/51
RUTH L. CHIMENTI, PT, PhD, DPT1 • ADOLPH S. FLEMISTER, MD2 • JOSHUA TOME, MS3 • JAMES M. MCMAHON, PhD4
MARIE A. FLANNERY, PhD, RN, AOCN4 • YING XUE, DNSc, RN4 • JEFF R. HOUCK, PT, PhD5
Altered Tendon Characteristics and
Mechanical Properties Associated With
Insertional Achilles Tendinopathy
I
nsertional Achilles tendinopathy (IAT) is a debilitating condition time off work due to weight-bearing re-
that is challenging to treat clinically. The disability associated with strictions, and risk of complications.15,24
Moreover, even a year after surgery, many
IAT interferes with daily weight-bearing activities. After failure
patients continue to report some pain
of conservative care, patients seeking operative care reported that
and functional limitations.15,24 Despite
pain limited their ability to walk and to approximately half of patients with IAT the severity and chronicity of this prob-
work full time.15 While the standard of progress to having surgery.26 Surgery is lem, there is little research to guide the
care includes a trial of physical therapy, costly in terms of financial resources, development of new physical therapy in-
terventions for IAT.
Tendon degeneration results in altered
STUDY DESIGN: Case-control laboratory study. alterations in tendon properties, and participant
TT tendon characteristics (shape, composi-
demographics. Pearson correlation was used to
TTOBJECTIVES: To compare tendon charac- tion) in persons with IAT. A larger diam-
teristics (shape, composition) and mechanical examine the association between severity of ten-
eter and lower echogenicity at the tendon
properties (strain, stiffness) on the involved side of don pathology and severity of symptoms (Victorian
insertion are due to the loss of parallel
participants with insertional Achilles tendinopathy Institute of Sport Assessment-Achilles).
(IAT) to the uninvolved side and to controls, and to collagen structure, loss of fiber integrity,
RESULTS: The side with IAT had a larger tendon
examine if severity of tendon pathology is associ- TT fatty infiltration, and/or capillary pro-
diameter (P<.001), lower echogenicity (P<.001),
ated with severity of symptoms during function. liferation.6,9,19 Ultrasound imaging is a
higher strain (P = .011), and lower stiffness (P =
BACKGROUND: Despite the severity and valid, reliable, and sensitive tool to detect
TT .007) compared to the side without IAT and the
chronicity of IAT, the quality of theoretical evidence altered tendon characteristics due to ten-
controls. On the involved side of participants with
available to guide the development of exercise dinopathy.6,10,31 Aström and colleagues6
IAT, a lower echogenicity correlated with higher
interventions is low. While tendon pathology
demonstrated that tendon thickness,
severity of symptoms (r = 0.603, P = .010).
of midportion Achilles tendinopathy has been
measured both by ultrasound (r = 0.51)
described, there are few studies specific to IAT. CONCLUSION: Ultrasound imaging combined
TT and magnetic resonance imaging (MRI)
METHODS: Twenty individuals with unilateral with dynamometry can discriminate alterations in
TT (r = 0.49), correlated with the histopatho-
IAT and 20 age- and sex-matched controls volun- tendon shape, composition, and mechanics in par-
logical score of Achilles tendon pathology
teered to participate. Ultrasound imaging was used ticipants with IAT. Future clinical trials for IAT may
to quantify changes in tendon shape (diameter) consider strategies to alter tendon characteristics in midportion and insertional Achilles
and composition (echogenicity). A combination and restore tendon mechanics. J Orthop Sports tendinopathy. Additionally, Collinger and
of ultrasound and dynamometry was used to Phys Ther 2014;44(9):680-689. Epub 7 August colleagues9 demonstrated that a darker
measure tendon mechanical properties (strain 2014. doi:10.2519/jospt.2014.5369 tendon on ultrasound imaging (ie, lower
and stiffness) during passive ankle rotation toward
echogenicity) indicated altered tendon
dorsiflexion. Generalized estimating equations TTKEY WORDS: ankle, plantar flexors, tendon,
composition in persons with rotator cuff
were used to examine the association between IAT, triceps surae, ultrasound
tendinopathy compared to healthy con-
1Department of Biomedical Engineering, University of Rochester Medical Center, Rochester, NY. 2Department of Orthopaedics and Rehabilitation, University of Rochester Medical
Center, Rochester, NY. 3Department of Physical Therapy, Center for Foot and Ankle Research, Ithaca College-Rochester Campus, Rochester, NY. 4School of Nursing, University of
Rochester, Rochester, NY. 5Physical Therapy Program, George Fox University, Newberg, OR. This research was supported by a Sproull Fellowship from the University of Rochester.
The protocol for this study was approved by the University of Rochester Research Subjects Review Board and the Ithaca College All-College Review Board for Human Subjects
Research. The authors certify that they have no affiliations with or financial involvement in any organization or entity with a direct financial interest in the subject matter or
materials discussed in the article. Address correspondence to Dr Ruth Chimenti, University of Rochester, Department of Biomedical Engineering, Goergen Hall 215, Rochester,
NY 14642. E-mail: [email protected] Copyright ©2014 Journal of Orthopaedic & Sports Physical Therapy®
T
44-09 Chimenti.indd 680 8/15/2014 7:11:08 PM
trols. While these findings support the
Characteristics of Participants With IAT
validity of ultrasound imaging, they are TABLE 1
and Matched Control Participants*
not specific to characterizing tendon al-
terations in IAT. Moreover, information
on tendon characteristics from imaging IAT (n = 20) Control (n = 20) P Value
may have prognostic value.4,12,26,35 Nichol- Age, y 58.6 ± 7.8 58.2 ± 8.5 .863
son et al26 demonstrated that grading of Sex (female), % 55 55 1.00
tendon diameter and severity of altered Weight, kg 87.5 ± 17.5 80.3 ± 16.0 .187
composition on MRI predicted the suc- Height, m 1.7 ± 0.1 1.7 ± 0.1 .999
cess of nonoperative care for IAT. There Body mass index, kg/m2 30.4 ± 5.4 27.9 ± 5.3 .158
may be similar value to examining IAT VISA-A, % 47.6 ± 26.8 100.0 ± 0.0 .001†
using ultrasound imaging, which has the Duration of symptoms, mo‡ 10 (4-24) … NA
added benefits of decreased cost and in- Abbreviations: IAT, insertional Achilles tendinopathy; NA, not applicable; VISA-A, Victorian Institute
creased availability. However, no studies of Sport Assessment-Achilles.
*Values are mean ± SD unless otherwise indicated.
have examined ultrasound imaging in
†Statistically significant difference (P≤.05).
patients with IAT. ‡Values are median (interquartile range).
Tendinopathy is associated with
changes in mechanical properties of on the impact of IAT on tendon me- chanical properties on the involved side
the tendon in patients with midportion chanical properties. Eccentric exercise of participants with IAT to those of the
Achilles tendinopathy.5,8,10,31 However, is commonly prescribed to patients with uninvolved side and to those of age- and
similar studies of patients with IAT are tendinopathy, based on the high success sex-matched healthy controls. Similar to
lacking. The insertional form of Achilles rate (82%-100%) of the intervention for participants with midportion Achilles ten-
tendinopathy differs from its midportion athletes with midportion Achilles ten- dinopathy, IAT was hypothesized to be as-
form, not only in the location of symp- dinopathy.1,11,13,22,27 However, eccentric sociated with higher measures of pathology
toms at the tendon’s calcaneal insertion exercise has been less effective (60% in terms of tendon characteristics (greater
but also in the bony deformity commonly success rate) for sedentary individuals tendon diameter, abnormal composition)
associated with IAT. Approximately 80% with Achilles tendinopathy,30 and for and mechanical properties (greater strain,
of persons with IAT have Haglund’s de- individuals with IAT (32% or less suc- lower stiffness). Additionally, higher mea-
formity,16 and more than 70% have cal- cess rate).11,29 While these studies focus sures of pathology are hypothesized to be
cifications within the tendon.17 Yet the on return to the previous level of self- associated with greater severity of symp-
impact of bony deformity on tendon me- reported function, few studies focus on toms in participants with IAT.
chanical properties in persons with IAT the mechanical function of the tendon.
METHODS
is unknown. A combination of ultrasound Arampatzis and colleagues2,3 demon-
imaging and isometric dynamometry can strated that dose-specific isometric ex-
capture in vivo Achilles tendon strain, ercises increased tendon stiffness in Participants
that is, tendon elongation, and stiffness healthy adults. Exercise dose and mode Twenty individuals with chronic
(force required per unit of tendon elonga- have not been examined in a similar way (greater than 3 months) unilateral
tion) during clinically relevant tasks, such to determine the effect on tendon me- IAT and 20 age- and sex-matched
as an isometric contraction or a passive chanical properties in participants with controls volunteered to participate. There
stretch.5,8,18 Evaluation of mechanical Achilles tendinopathy. Although im- were no differences in demographic vari-
properties has demonstrated increased provement in tendon mechanical proper- ables between groups (TABLE 1). All par-
tendon strain and decreased stiffness in ties is desirable, no literature exists that ticipants with IAT were diagnosed by an
participants with midportion Achilles characterizes in vivo tendon mechanical orthopaedic surgeon and had tenderness
tendinopathy compared to controls.5,8,31 properties in IAT participants. To inves- to palpation along the distal 2 cm of the
While a similar increase in strain and tigate the influence of specific exercise Achilles tendon at its insertion. Partici-
decrease in stiffness are logically hypoth- dosages on mechanical properties, stud- pants with IAT had chronic symptoms,
esized in IAT, there is no literature that ies that first establish the mechanical with a median duration of 10 months
reports on these factors in patients with properties of the tendon in participants (range, 3 months to 15 years), and report-
IAT. with pathology and decreased function ed greater symptom severity than controls
The quality of evidence guiding ex- are needed. (TABLE 1). Individuals were excluded if they
ercise interventions for IAT is low, and The purpose of this study was to had a history of foot or ankle surgery, were
could be improved with information compare tendon characteristics and me- unable to safely climb stairs indepen-
44-09 Chimenti.indd 681 8/15/2014 7:11:08 PM
FIGURE 1. Tendon characteristics for uninvolved (top row) and involved (bottom row) sides of a participant with insertional Achilles tendinopathy. (A and B) Longitudinal
ultrasound images of tendon diameter; (C and D) cross-sectional ultrasound images of echogenicity measured by the mean grayscale value; (E and F) corresponding histograms
demonstrating the mean grayscale value (scale from 0, black to 255, white) for images C and D. A darker tendon results in a lower mean grayscale value, which indicates tendon
degeneration.
dently, were pregnant, or had comorbidi- Measures longitudinal (FIGURES 1A and 1B) and cross-
ties that may alter tendon structure (eg, All measures were performed in a single sectional (FIGURES 1C and 1D) images, using
diabetes, neurological condition). The session in the afternoon, as in the morn- B-mode imaging on a Philips HD11 XE
sample size was chosen based on a power ing symptoms are often most severe. digital ultrasound machine (Koninklijke
analysis, assuming a minimum power Participants were instructed to avoid Philips Electronics NV, Amsterdam, the
of 0.80 and a 2-tailed alpha level of .05. high-level activities (eg, running, going to Netherlands) with a linear-array probe
Using a t test as a conservative estimate, the gym) before their session, to minimize (L12-3) at a frequency of 10 to 12 MHz.
16 participants with IAT were needed to the risk that study participation might ag- For these measurements, the participant
detect statistically significant differences gravate their symptoms. Data collection was in prone and the plantar surface of
between sides for tendon stiffness, based began with motion analysis (a portion of the foot was stabilized against a board
on pilot data indicating a large effect size these data were used to measure Achil- oriented perpendicular to the floor. All
(Cohen d = 1). For a correlational analysis, les tendon moment arm), followed by ultrasound images were stored digitally
a sample size of 20 participants with IAT measurements of tendon characteristics and processed using ImageJ (National
was needed to detect a moderate (0.6 or and mechanical properties. Completion Institutes of Health, Bethesda, MD). The
greater) correlation between tendon char- of self-report measures and rest periods tendon insertion was defined as the most
acteristics/mechanical properties and were interspersed as needed, based on distal point of the tendon on the calcane-
symptom severity. the participant’s activity tolerance. The us in a sagittal view. If there were spurs or
All participants were informed of the examiners were not blinded to group sta- calcifications in the area of the insertion,
study procedures and signed a consent tus (IAT or control) of participants. the best distal tendon point was estimat-
form approved by the University of Roch- ed. The error in estimating this point in
ester Research Subjects Review Board Tendon Characteristics 10 persons with IAT was low (intraclass
and the Ithaca College All-College Review Tendon characteristics (diameter, echo- correlation coefficient [ICC] = 0.981;
Board for Human Subjects Research. genicity) were measured by obtaining standard error of measurement [SEM],
44-09 Chimenti.indd 682 8/15/2014 7:11:09 PM
the anterior tibialis muscles was moni-
tored by a 2-channel Bagnoli desktop
electromyography system (Delsys Inc,
Natick, MA). Single-differential surface
electrodes (input impedance greater than
1015 Ω // 0.2 pF; common-mode rejec-
tion ratio, –92 dB; noise, 1.2 μV; pre-
amplifier gain, 10 V/V  1%) were used.
The contact material was 99.9% Ag, in
the shape of 2 bars with contact dimen-
sions of 10 × 1 mm and spaced 1 cm apart.
The electrodes were placed over the mo-
tor point of the medial gastrocnemius
and the anterior tibialis muscles. The
FIGURE 2. (A) Tendon mechanical properties were measured during passive (visually confirmed using
raw electromyography signal was viewed
electromyography) rotation of the ankle by the isokinetic dynamometer. (B and C) Ultrasound images
using a 2-channel digitizing oscilloscope
demonstrating gastrocnemius muscle–Achilles tendon junction displacement as the ankle is rotated from (B) 10°
of plantar flexion to (C) 10° of dorsiflexion. The difference in length between the white horizontal lines represents (Tektronix, Inc, Beaverton, OR). The raw
the linear displacement of the muscle-tendon junction during passive ankle rotation. data were viewed on the oscilloscope to
visually inspect if muscle activity was
0.31 mm). On a longitudinal view, ten- was moved in the sagittal plane from 10° greater than resting level. If participants
don diameter was measured at the widest of plantar flexion to 10° of dorsiflexion had less than 10° of dorsiflexion (n = 1) or
anterior/posterior point within 2 cm of at a rate of 5°/s, and the passive plantar demonstrated greater than their resting-
the tendon insertion (FIGURES 1A and 1B). flexion torque was recorded at a rate of level muscle activity (n = 2), then they
A cross-sectional image was obtained 1000 Hz. Similar to previous studies,5,8,18 were excluded from analysis of tendon
2 cm proximal to the tendon insertion. ultrasound imaging was used to track mechanical properties.
Echogenicity was quantified as the mean the linear displacement of the gastroc- To calculate tendon strain, tendon
grayscale value for the central 1 cm of the nemius muscle–Achilles tendon junction elongation was first calculated by sub-
entire depth of the tendon cross-section (MTJ), with the probe at a fixed point on tracting the displacement of the MTJ
(FIGURES 1C and 1D). The grayscale value the calf (FIGURES 2B and 2C). A first set of from the total elongation of the muscle-
for a pixel could range from 0 (black) to 6 rotations was used for preconditioning, tendon unit (tendon elongation = total
255 (white). The mean grayscale value followed by a second set of 6 rotations, muscle-tendon-unit elongation – muscle
for all pixels in the region of interest was from which images were used to calculate elongation). The displacement of the
quantified using ImageJ, with a lower tendon mechanical properties. B-mode MTJ was recorded using ultrasound im-
value indicating greater pathology. All ultrasound images were collected when aging. Because the knee joint was immo-
ultrasound images had the same gain set- the ankle was in (1) 10° of plantar flexion bile, total muscle-tendon-unit elongation
ting of 56 and were collected at a 3-cm at the start and end of the second set and was estimated from the change in ankle
depth, so that differences in ultrasound (2) 10° of dorsiflexion during the third, angle. Previous studies have validated
settings would not influence the mean fourth, and fifth repetitions. The location a similar approach to estimating total
grayscale values between participants. of the MTJ in each ankle position was av- muscle-tendon-unit elongation.14,18 The
eraged, so that distance between the MTJ muscle-tendon-unit elongation also de-
Tendon Mechanical Properties locations in 10° of plantar flexion and pended on the position of the heel cup on
The combination of ultrasound imaging 10° of dorsiflexion could be calculated. A the isokinetic dynamometer, which had
and dynamometry was used to measure video was taken on the second repetition 3 adjustments. Depending on the size of
tendon mechanical properties (strain, to confirm that the images captured the the foot, the heel cup was adjusted to ap-
stiffness). Each participant’s ankle was full range of MTJ displacement. Passive proximate the participant’s ankle axis of
positioned and secured in a Biodex Sys- torque values from the third, fourth, and rotation with that of the dynamometer.
tem 4 (Biodex Medical Systems, Inc, fifth repetitions of the second set were For each adjustment, the linear transla-
Shirley, NY) in sitting, with the seatback used for analysis. tion of the Achilles insertion was estab-
reclined to 45°, the chair height adjusted Participants were instructed to relax lished using 3-D motion capture (n = 3).
so that the tibia was parallel to the floor, as the foot plate moved the ankle. To en- A 3-D motion-capture system (Optotrak
and the hip in neutral adduction/abduc- sure that the condition was passive, the motion analysis system, model 3020;
tion (FIGURE 2A). The participant’s ankle activity of the medial gastrocnemius and Northern Digital Inc, Waterloo, On-
44-09 Chimenti.indd 683 8/15/2014 7:11:09 PM
these 2 points, when the participant was al curve in FIGURE 3). Similar to Kawakami
500
in a prone position with the knee straight et al,18 tendon stiffness was approximated
IAT, uninvolved side
400 y = 29.0x – 144.8 and the foot perpendicular to the floor using the coefficient of the linear portion
Healthy control
e, N 300 y = 44.5x – 46.1 (due to tibial inclination, the ankle was of the curve. Because individuals with
Forc 200 IAT, in approximately 10° of plantar flexion), IAT often experience symptoms dur-
involved side was measured and used as the tendon ing activities that require dorsiflexion,15
100 y = 17.6x – 56.0
length. Between-day test-retest reliability mechanical properties toward the end
0
of tendon length measurement was good, range of ankle motion, which coincides
0 5 10 15 20
with an error of 0.04% of original ten- with the linear portion of the curve, were
Tendon Elongation, mm don length (n = 6; ICC = 0.910; SEM, 7.9 of particular interest in the current study.
mm). An abbreviated 2-point method (point 1,
FIGURE 3. Force elongation curve for pilot work in To examine the resistance to stretch 10° of ankle plantar flexion; point 2, 10°
which tendon mechanical properties were examined
during passive ankle rotation in dor- of ankle dorsiflexion) was developed to
from 30° of PF to maximum DF. The points on each
siflexion, tendon force was calculated. capture mechanical properties in the lin-
curve correspond to force and elongation at 10°
of DF and 10° of PF (ie, methods used in current To derive Achilles tendon force, mo- ear region of the force-elongation curve
study). Stiffness is the slope of the corresponding line tion analysis combined with ultrasound (lines connecting 2 points along curves in
highlighted in bold on the linear equations. The curves imaging was used to estimate Achilles FIGURE 3). This abbreviated methodology
represent the data for 1 healthy male and 1 male with
tendon moment arm. In a neutral stand- in preliminary work with young, healthy
unilateral IAT. Abbreviations: DF, dorsiflexion; IAT,
ing position, for each participant, the adults was initially compared to the ap-
insertional Achilles tendinopathy; PF, plantar flexion.
medial malleolus and lateral malleolus proach described by Kawakami et al.18
tario, Canada) was used to track ankle were digitized, and the center between Converting our data into the same units
dorsiflexion at a rate of 60 Hz. Infrared these 2 points was defined as the ankle for stiffness reported by Kawakami et al18
diodes were placed at the tendon inser- joint center. A point was also digitized (Nm/%), the average tendon stiffness of
tion and actively emitted light for the on the posterior aspect of the heel. The young, healthy adults (n = 6; mean  SD
cameras to track 3-D motion as the an- moment arm was calculated by the an- stiffness, 3.6  1.6 Nm/%) was within 1
kle was moved in the sagittal plane. The terior/posterior distance between the SD of the mean reported by Kawakami et
accuracy of tracking an infrared diode ankle joint center and the digitized heel al18 (3.4  2.5 Nm/%).
is up to 0.1 mm with the research-grade point (measured with motion capture)
motion-capture system. After collection, minus the anterior/posterior thickness of Reliability
the kinematic data were smoothed using the skin and subcutaneous tissue (mea- Test-retest reliability was examined on
a fourth-order Butterworth filter with a sured with ultrasound imaging). Tendon different days within the same week. As
zero phase lag and a cutoff frequency of force was calculated by dividing passive only 1 physical therapist collected and
6 Hz. Total muscle-tendon-unit elonga- plantar flexion torque (measured by the processed tendon variables, the following
tion was (mean  SD) 16.2  1.9 mm for dynamometer during ankle rotation) by statistics represent intrarater reliability.
setting 1, 18.7  1.1 mm for setting 2, and the moment arm of the Achilles tendon To minimize risk of aggravating symp-
19.5  1.5 mm for setting 3. (tendon force = passive torque of ankle/ toms in participants with IAT, repeat
Strain was calculated to take into ac- moment arm of Achilles tendon). Ten- testing was performed in 6 young, healthy
count differences in anthropometrics, don stiffness was calculated by dividing adults (control participants). Test-retest
with tendon elongation normalized to the change in tendon force by the change reliability and SEM for echogenicity (ICC
each individual’s tendon length (ten- in strain as the ankle was rotated from = 0.98; SEM, 3.0 units), diameter (ICC
don strain = tendon elongation/tendon plantar flexion to dorsiflexion (tendon = 0.996; SEM, 0.1 mm), strain (ICC =
length). Two points were used to define stiffness = change in tendon force/change 0.946; SEM, 0.3%), and stiffness (ICC
the length of the tendon. First, the most in tendon strain). = 0.998; SEM, 0.7 N/mm) in the con-
distal portion of the tendon insertion was The current study focused on measure- trol participants were high. A difference
found using ultrasound, and a thin wire ment of tendon mechanical properties in between the involved and uninvolved
was placed under the transducer at the the linear region of the force-elongation sides greater than 1.96 SEM was used to
level of the insertion. A mark was made curve (FIGURE 3). In a previous study by describe how frequently a comparison
on the skin at the position of the wire. Kawakami et al,18 the authors measured between sides indicated altered tendon
Second, using the same procedure, the mechanical properties throughout a characteristics on the involved side. This
medial gastrocnemius MTJ was found greater arc of ankle motion to describe cutoff value was chosen because it repre-
using ultrasound and a second mark both the toe and linear regions of the sents the 95% confidence interval for a
placed on the skin. The distance between force-elongation curve (entire polynomi- score on the uninvolved side.
44-09 Chimenti.indd 684 8/15/2014 7:11:10 PM
In addition, the reliability of process-
Outcomes of the Generalized
ing the ultrasound images was examined.
Estimating Equation Analyses Comparing
A set of 10 images from the involved and
Sides With Insertional Achilles Tendinopathy
10 images from the uninvolved sides TABLE 2
(Involved Side) to Sides Without Insertional
were blinded. The echogenicity, ten-
Achilles Tendinopathy (Uninvolved
don diameter, and tendon elongation
Side and Control Participants)
were measured for 20 images twice, and
then unblinded and compared for accu-
racy. The reliability of processed ultra- Dependent Variable β* P Value†
sound images was high (ICC>0.9) for all Diameter, mm 2.5 (1.8, 3.3) <.001
measures. Echogenicity –11.2 (–17.1, –5.3) <.001
Tendon strain, % 1.0 (0.2, 1.7) .011
Symptom Severity Tendon stiffness, N/mm –12.3 (–21.1, –3.4) .007
Symptom severity was measured with *Values in parentheses are 95% confidence interval.
†Statistically significant difference (P≤.05).
the Victorian Institute of Sport Assess-
ment-Achilles questionnaire (VISA-A).
This self-report measure sums the score analysis was necessary prior to perform- ticipants with IAT. Statistical significance
of 8 items to assess stiffness, pain dur- ing comparisons using the more typical for all tests was defined as a 2-tailed P
ing activity, and ability to participate in analysis of variance (ANOVA) because of value less than or equal to .05.
sports. The VISA-A has been demon- the lack of independence between sides.
RESULTS
strated to be a valid, reliable, and sen- A significant Wald chi-square test was
sitive measure in persons with Achilles then followed with an ANOVA to make
tendinopathy.21,28 comparisons across sides and groups. The side with IAT had altered
Demographic variables (body mass in- tendon characteristics (larger ten-
Statistical Analysis dex, age, sex) were included as covariates don diameter and lower echogenici-
Comparisons between the side with IAT in the GEE analyses. Significance was ty) compared to the side without IAT. On
and the side without IAT were tested defined as a 2-tailed P value less than or average, the side with IAT had a 2.5-mm
using generalized estimating equations equal to .05. larger tendon diameter and an 11-unit
(GEEs). A GEE analysis uses the gener- One-way ANOVAs were used to test lower echogenicity compared to the side
alized linear model, which accounts for whether differences found in the GEE without IAT (TABLE 2). These alterations
the correlation between sides within an analyses were driven by side (within- were consistent in pairwise comparisons
individual.33 This analysis allowed for subject effect), group (between-subject between sides and between groups. Spe-
inclusion of all data (both sides of all effect), or both. To test the effect of side, cifically, the involved side of participants
participants), and uniquely identified the involved side was compared to the with IAT had a larger tendon diameter
differences between the side with IAT uninvolved side in the participants with than the uninvolved side and the cor-
(involved side of participants with IAT, IAT. To test the effect of group, the in- responding side in control participants
n = 20) and the side without IAT (unin- volved side (right or left) in those with (P<.001 between sides and between
volved side of participants with IAT, n = IAT was matched for analysis with the group comparisons) (TABLE 3). Similarly,
20; both sides of controls, n = 40 sides). corresponding side of an age- and sex- the involved side had lower echogenicity
Interpretation of the regression coeffi- matched control participant. It was hy- than the uninvolved side (P = .002) and
cients (β) derived from the GEE analysis pothesized that the differences between controls (P = .003) (TABLE 3). A higher
is similar to standard linear regression. the side with IAT and the side without body mass index was also associated with
For example, β for the dichotomous inde- IAT would be driven by both within- and a lower echogenicity (GEE analysis: β =
pendent variable represents, on average, between-subject effects. If a demographic –1.9; 95% confidence interval: –2.5, –1.3;
how much greater/lower the side with covariate was statistically significant in P<.001; between-side analysis of covari-
IAT is compared to the side without IAT. the GEE analysis, it was included as a co- ance, P = .017; between-group analysis of
Within the GEE analysis, the significance variate in the ANOVA. covariance, P<.001). For 85% (17/20) of
of the independent variable (side with Pearson correlations were used to participants with IAT, the involved side
IAT versus side without IAT) in predict- examine the associations between ten- had a larger tendon diameter than the
ing a dependent variable (echogenicity, don properties (diameter, echogenicity, uninvolved side. Compared to the unin-
diameter, strain, stiffness) was tested strain, stiffness) of the involved side and volved side, 65% (13/20) of participants
using a Wald chi-square test. The GEE symptom severity (VISA-A) among par- had lower echogenicity on the involved
44-09 Chimenti.indd 685 8/15/2014 7:11:10 PM
side was associated with greater symp-
Pairwise Comparisons of the Involved tom severity. Consistent with hypotheses
Side of Participants With IAT to the that related IAT to tendon pathology, the
TABLE 3
Uninvolved Side and to the Corresponding involved side of participants with IAT
Side in Matched Controls* had a larger tendon diameter and lower
echogenicity than the uninvolved side
Participants With Unilateral IAT and controls. Increased tendon diameter
was the most consistent marker of IAT
Involved Side Uninvolved Side Matched-Control Side
(n = 20 tendons) (n = 20 tendons) (n = 20 tendons) pathology of the ultrasound variables
Diameter, mm 6.4 ± 1.7 4.4 ± 0.7† 3.6 ± 0.7† assessed in the current study. Although
Echogenicity 67.3 ± 21.8 77.5 ± 19.0† 84.8 ± 9.6† tendon mechanical properties showed
Tendon elongation, mm 6.5 ± 2.9 5.2 ± 1.9† 4.5 ± 1.6† higher strain and lower stiffness, similar
Tendon length, mm 189.8 ± 28.2 189.0 ± 27.3 194.2 ± 23.3 to previous studies of midportion Achilles
Tendon strain, % 3.4 ± 1.5 2.8 ± 1.0† 2.2 ± 0.6† tendinopathy,5,8 the clinical significance
Passive force, N 194.9 ± 56.8 206.6 ± 59.6 187.6 ± 48.3 of the measured changes in mechanical
Tendon stiffness, N/mm 33.5 ± 12.6 43.7 ± 19.1† 46.4 ± 20.0† properties requires further study.
Abbreviation: IAT, insertional Achilles tendinopathy. Tendon echogenicity is a feasible,
*Values are mean ± SD. quantifiable measure that was associated
†Significant difference (P≤.05) from involved side of participants with IAT with an analysis of
with the severity of IAT symptoms, as
variance.
measured by the VISA-A scale. A lower
echogenicity (lower mean grayscale
side (difference between sides, greater IAT had higher strain on the involved value) indicates less organization and
than 1.96 SEM). side, and 65% (11/17) of participants had homogeneity of the tendon microstruc-
The side with IAT had altered tendon lower stiffness on the involved side com- ture (greater severity of pathology).9 This
mechanical properties (higher strain and pared to the uninvolved side. disorganized microstructure is believed
lower stiffness) compared to the side On the involved side of participants to represent tendon degeneration as-
without IAT. The GEE analysis demon- with IAT, there was a significant associa- sociated with tendinopathy. However,
strated that, on average, the side with IAT tion between symptom severity (VISA- validity studies connecting histopathol-
had a 1% greater strain and a 12-N/mm A) and tendon characteristics. A lower ogy with measures of gray and white
lower stiffness compared to the side with- echogenicity was associated with greater pixel patterns from ultrasound imaging
out IAT (TABLE 2). These alterations were symptom severity (r = 0.603, P = .010). (mean grayscale value) would improve
consistent in pairwise comparisons be- There was a marginally nonsignificant the understanding of echogenicity. To
tween sides and between groups. The in- negative correlation between tendon date, studies examining the relationship
volved side in participants with IAT had diameter and symptom severity (r = between histopathology and ultrasound
greater strain than the uninvolved side –0.456, P = .066). There was no clear evi- measures of tendon composition have
(P = .047) and the corresponding side dence of a significant correlation between been limited by categorical (normal ver-
of participants in the control group (P symptom severity and tendon strain (r = sus abnormal) descriptions of the severity
= .006) (TABLE 3). Similarly, the involved –0.329, P = .197) or stiffness (r = 0.165, of tendon pathology.6,10 However, a cat-
side in participants with IAT had lower P = .527). egorical measure limits the potential use
stiffness than the uninvolved side (P = of echogenicity for diagnostic grading of
DISCUSSION
.030) and the corresponding side of those IAT severity or predicting success with
in the control group (P = .021) (TABLE 3). conservative care among a spectrum of
There was not sufficient evidence of dif- This is the first study, to our patients with IAT. Because echogenicity
ferences in tendon length (between sides, knowledge, to demonstrate that the was associated with body mass index
P = .349; between groups, P = .603) or involved side of individuals with and IAT symptom severity, further stud-
tendon force (between sides, P = .376; be- IAT has altered tendon characteristics ies investigating the relationship between
tween groups, P = .677) in pairwise com- and mechanical properties, as measured these variables and echogenicity and ten-
parisons (TABLE 3). There were significant by ultrasound imaging, compared to the don histopathology are encouraged. Fur-
differences between sides (P = .033) and contralateral uninvolved side and con- ther research is needed to examine why a
groups (P = .020) in tendon elongation trols without IAT. Additionally, greater higher body mass index is associated with
(TABLE 3). Compared to the uninvolved impairment in tendon characteristics a lower echogenicity, and if this may be a
side, 59% (10/17) of participants with (lower echogenicity) on the involved risk factor for developing Achilles tendi-
44-09 Chimenti.indd 686 8/15/2014 7:11:11 PM
nopathy. In this analysis, body mass index of ankle excursion,18 and determining the influence of dorsiflexion range of motion,
was a covariate, minimizing its influence tendon length,7 can affect values when Achilles tendon strain, and its influence
on the findings of this study. comparing across studies. For example, it on continuing symptoms associated with
Tendon diameter was another simple would be anticipated that during an iso- IAT are necessary.
and reliable ultrasound measure that was metric contraction, as performed in the Despite some differences in methods
strongly associated with IAT. In 85% of studies of Arya and Kulig5 and Child et and type of tendon problems across stud-
participants with IAT, ultrasound imag- al,8 the strain would be greater than dur- ies, the effect of tendinopathy on stiffness
ing was sensitive to differences in tendon ing a passive task,32 as performed in the was similar to that reported in other stud-
diameter between the involved and unin- current study. Nonetheless, the similar ies. The current study examined tendon
volved sides. Given the consistency of this group effects across different methodolo- stiffness during a passive task (20° ankle
measure to detect differences in diameter, gies and types of Achilles tendinopathy rotation toward dorsiflexion) and found a
this variable may be particularly useful in support the idea that tendon strain is 27.8% decrease in tendon stiffness in par-
the development of prognostic criteria. increased in this population. While these ticipants with IAT compared to controls.
Nicholson and colleagues26 developed a between-group differences in strain ap- Similarly, during an active task (maxi-
grading system of tendon pathology us- pear relatively small, the magnitude of mum voluntary isometric contraction),
ing MRI at the Achilles insertion. Par- the difference may increase with dynamic Arya and Kulig5 found a 24.9% decrease
ticipants with a tendon diameter greater tasks, such as running, stair climbing, or in Achilles tendon stiffness in partici-
than 8 mm and abnormal signal inten- end-range calf stretches. For example, the pants with midportion Achilles tendinop-
sity at the insertion were more likely to strain associated with single-leg hopping athy. A passive task was chosen for the
undergo surgery. For example, 88 of 92 (Lichtwark and Wilson,20 8.3%) is nearly current study to obtain similar loading
participants with a diameter greater than 4 times the amount of strain demon- between the IAT and control groups. Yet,
8 mm and greater signal abnormality strated by controls for the current study individuals with IAT may show greater
chose surgery. In contrast, participants during a 20° passive ankle dorsiflexion differences in tendon characteristics un-
with a tendon diameter of less than 8 mm movement (2.2%) (TABLE 3). Alterations der higher loads. Stiffness in participants
and less signal abnormality continued in strain during more challenging tasks with IAT may also be assessed with an
with rehabilitation (2 of 16 participants may affect dynamic muscle-tendon func- active test in which stiffness is measured
chose surgery). Ultrasound imaging as tion. Although a link between tendon me- during a maximal isometric effort. This
performed in the current study may be chanical properties and muscle-tendon was not done in the current study, be-
equally effective in determining progno- function is possible, no evidence of this cause differences in strength would have
sis and is much less expensive to employ relationship between strain and symptom confounded the comparison between
clinically than MRI. Prognostic informa- severity (r = –0.329, P = .197) was found groups (tendon strain is not independent
tion could allow patients to make a more in the current study. from force). Future studies, however, may
informed choice about their care, when Based on the results of this study, the consider using various submaximal loads
weighing the costs of conservative reha- use of passive stretching of the Achilles to evaluate tendon stiffness under more
bilitation versus the risks associated with tendon as part of a conservative treat- functional loading conditions.
surgical intervention. While prospective ment program for IAT is inconsistent Therapeutic strategies designed to re-
data are necessary to establish prognos- with the finding of higher tendon strain. verse alterations in mechanical properties
tic factors, this cross-sectional study has The finding of excessive elongation of the may improve clinical outcomes for IAT.
established that there is a spectrum of Achilles tendon calls into question the Although eccentric exercise is common-
tendon diameter thickness across par- regular administration of calf stretching ly prescribed for IAT,11,16,29 in controls,
ticipants with IAT. for tendons that already exhibit increased isometric exercise stimulated improve-
Insertional and midportion Achilles strain. Although commonly recommend- ment in tendon mechanical properties.
tendinopathies may both have similar ef- ed, based on anecdotal experience, there Two studies of healthy participants have
fects on tendon strain. The current study is little empirical evidence to support demonstrated increased Achilles tendon
found a difference in strain (1.2%) be- stretching in patients with IAT.34 Typical- stiffness after a 14-week isometric exer-
tween participants with IAT and controls ly, the goal of calf stretching is to increase cise training protocol.2,3 Conversely, stud-
that was similar in magnitude to that of dorsiflexion range of motion. Stretching ies demonstrating the effect of eccentric
previous studies comparing participants the calf muscle also may increase strain exercise at 6 weeks showed a decrease in
with midportion Achilles tendinopathy to of the Achilles tendon. Data from the Achilles tendon stiffness.23,25 To date, re-
controls (Arya and Kulig,5 0.8%; Child et current study demonstrate that strain is search on the effects of exercise on Achil-
al,8 1.8%). Methodological differences, typically increased, not decreased, in par- les tendon mechanical properties has
such as a passive or active task,32 range ticipants with IAT. Further studies of the been limited to a healthy population. The
44-09 Chimenti.indd 687 8/15/2014 7:11:12 PM