Quantitative MRI Analysis of the Talocrural and

Talonavicular Joints in Ballet Dancers

Toshimi Tando, MD, PhD, Saya Horiuchi, MD, Hon J. Yu, PhD, Alex Luk, PhD, Jeffrey A.
Russell, PhD, AT, FIADMS, Kelli Sharp, DPT, and Hiroshi Yoshioka, MD, PhD

Abstract with a 3.0 T MR scanner. Medial and compared to female non-dancers,

The ankles of ballet dancers are lateral subchondral bone distance though the differences were not sig-
routinely under heavy loading that between the tibia and talus (MSBD nificant (TN angle: 16.6° vs. 22.3°,
may lead to osteoarthritic changes. It and LSBD), axial navicular-talus p = 0.09, and NC: 0.186 vs. 0.165,
would be clinically useful to identify axis angle (ANT angle), sagittal ta- p = 0.28). There were no significant
such pathology as early as possible lar neck angle against the posterior differences in T1rho or T2 values of
in a dancer’s career. Therefore, the talocalcaneal joint (TN angle), and talonavicular joint cartilage. These
purpose of this study was to com- curvature of navicular surface at the results show that the bony anatomy
pare quantitative measurements in talonavicular joint were measured on of dancers’ ankles may adapt to the
magnetic resonance (MR) images sagittal images. The medial subchon- stresses placed on them by ballet.
of the talocrural and talonavicular dral bone distance was found to be

A
joints in ballet dancers and healthy significantly larger in female dancers nkle and foot injuries (includ-
non-dancers for use in formulating than female non-dancers (4.05 mm ing impingement syndromes,
prediction of chronic injury and vs. 2.75 mm, p < 0.05), whereas sprains, tendinopathy, and
degenerative joint disease in these lo- there were no significant differences stress fractures) occur frequently in
cations. Quantitative measurements in LSBD (2.63 mm vs. 2.63 mm, p ballet dancers.1 Female ballet danc-
in MR images of the talocrural and = 0.87). Axial navicular talus angles ers require large ranges of motion of
talonavicular joints were compared in female dancers were significantly the ankle in dorsiflexion and plantar
in 10 female ballet dancers, 10 larger than those in female non-danc- flexion, especially en pointe, a posi-
healthy female non-dancers, and nine ers (38.9° vs. 24.3°, p < 0.05). There tion that requires maximum plantar
male ballet dancers. Fat-suppressed was a tendency for the TN angle to flexion2 and results in an incongruent
density-weighted proton, T1rho, and be smaller and navicular curvature talocrural joint.3 In professional ballet
T2 mapping images were acquired (NC) to be larger in female dancers dancers, ankle range of motion is larg-
er than in non-dancers.2,4 Therefore,
it is always possible that the capsule
Toshimi Tando, MD, PhD, Saya Horiuchi, MD, Hon J. Yu, PhD, Alex Luk, around the talocrural joint in ballet
PhD, and Hiroshi Yoshioka, MD, PhD, Department of Radiological Sciences, dancers is distended or loose due to
University of California, Irvine, California, USA. Jeffrey A. Russell, PhD, AT,
FIADMS, School of Applied Health Sciences and Wellness, Laboratory for
years of weightbearing of this joint
Science and Health in Artistic Performance, Ohio University, Athens, Ohio, USA. in forced maximum dorsiflexion and
Kelli Sharp, DPT, Department of Dance, The Claire Trevor School of the Arts, plantar flexion.5 Because the capsule
and Department of Physical Medicine and Rehabilitation, School of Medicine, around the talocrural joint is one of
University of California, Irvine, California, USA. the factors related to stability of the
joint, a distended or loose capsule, as
Correspondence: Jeffrey A. Russell, PhD, AT, FIADMS, Division of Athletic
Training, School of Applied Health Sciences and Wellness, Ohio University,
indicated by enlarged joint distance,
Grover Center E182, Athens, Ohio 45701, USA; jeff.russell@ohio.edu. suggests instability of the joint.
Talocrural joint instability increases
This work was supported by the University of California, Irvine, Institute for the risk of degenerative joint disease.
Clinical and Translational Science (ICTS) School of Medicine/Claire Trevor Furthermore, chronically altered joint
School of Arts Grant, and the ICTS Voucher Program. mechanics, including malalignment
Copyright © 2021 J. Michael Ryan Publishing, Inc.
38 https://doi.org/10.12678/1089-313X.031521f
 Journal of Dance Medicine & Science • Volume 25, Number 1, 2021 39

and incongruity, may lead to develop- al.15 reported that quantitative T2 numbers of left and right ankles were
ment of osteoarthritis. measurement may potentially be scanned in each female group, while
When in the en pointe position used as a noninvasive imaging tool four right ankles and five left ankles
of ballet, axial mechanical loading as for early detection of cartilage lesions were scanned in the male group. One
a result of the hyper-plantar flexion in the tibiotalar joint. Therefore, left ankle scan of a male dancer was
is applied to the talonavicular joint. the purpose of this study was to excluded due to incorrect angular
Hence, it is no surprise that a high compare quantitative measurements positioning.
prevalence of osteoarthritis of the in MR images of the talocrural and
talocrural, subtalar, and first meta- talonavicular joints of ballet danc- MR Imaging Protocol
tarsophalangeal joints is reported in ers and healthy non-dancers for use All MR acquisitions were performed
retired female dancers,6 but to date in evaluating prediction of chronic on a 3.0 T scanner (Achieva, Philips
little is known about early osteoar- injury and degenerative joint disease Healthcare, The Netherlands) using
thritic changes in articular cartilage of at these locations. an 8-channel, dedicated ankle-foot
the talonavicular joint nor about de- coil. Three sagittal images were ac-
velopmental morphological changes Methods and Materials quired: fat-suppressed (FS) proton
in pre-professional or professional Subjects density-weighted imaging (PDWI),
dancers. Ten female volunteer non-dancers T1rho mapping, and T2 mapping.
Compositional magnetic reso- (mean age: 20.4 ± 1.65 years; range: Participants were in the supine posi-
nance imaging (MRI) sequences, 19 to 24 years), 10 female ballet tion with feet first and the ankle to be
such as T1rho mapping, T2 map- dancers (mean age: 21.7 ± 3.77 years; imaged in neutral position. All MR
ping, delayed gadolinium-enhanced range: 19 to 30 years), and nine images were scanned with true sagit-
MRI of cartilage, sodium imaging, male ballet dancers (mean age: 23.1 tal angulation parallel to the static
and diffusion-weighted imaging ± 6.71 years; range: 18 to 40 years) magnetic field (B0). Additionally,
are techniques that allow for the were enrolled from a conservatory axial and coronal FS PDWIs were
quantification of the biochemical style university dance program. Mean obtained.
composition of cartilage.7 Several years and mean hours per week of The acquisition parameters were as
publications have demonstrated the dance training of the female and male follows:
capabilities of compositional MRI dancers were 15.9 years (range: 9 to • Sagittal FS PDWI: 2D turbo spin-
assessment techniques for detecting 28 years), 16.8 hours (range: 5 to 24 echo; repetition time (TR)/echo
the initial stages of cartilage degen- hours), and 13.2 years (range: 3.5 to time (TE) = 3958/30 ms, number
erative change in the knee prior to 22 years), 25.3 hours (range: 8 to 25 of excitations (NEX) = 1, and total
imaging findings seen on morpho- hours), respectively. No subject had acquisition time = 2 min, 30 s.
logical MRI.8-10 Of these techniques, obvious physical limitations. The • Axial FS PDWI: 2D turbo spin-
T1rho and T2 mappings have been study protocol was approved by the echo; TR/TE = 4582/30 ms, NEX
shown to be feasible for quantifying authors’ university institutional re- = 1, and acquisition time = 4 min,
ankle cartilage with high reproduc- view board, and all subjects provided 7 s.
ibility.11-14 In ballet dancers, Cha et written informed consent. Equal • Coronal FS PDWI: 2D turbo spin-

Figure 1 Coronal FS PDWI MR images of the talocrural joint of female non-dancer (A) and female dancer (B). MSBD (arrow
heads) and LSBD (arrows) were measured in one-third medial and lateral portion of the talar dome.
40 Volume 25, Number 1, 2021 • Journal of Dance Medicine & Science

Figure 2 Axial FS PDWI MR images of

the talocrural (A and C) and talonavicular
(B and D) joints. In A and C the solid line
is the axis of the talar body parallel to the
medial and lateral joint line. In B and D
the solid line is the axis of the navicular
perpendicular to the dashed line marking
the medial and lateral posterior ends of
the talonavicular joint. ANT angles were
measured between the axes of the talar
body and navicular bone. (A and B female
non-dancer; C and D female dancer.)

Figure 3 Sagittal FS PDWI of the hind

foot. TN angle was defined as the angle
between the longitudinal axis of the talar
head and the line parallel to the posterior
talocalcaneal joint surface of the talus (A
and B). NC was measured as the ratio
(b/a) of (a) length of the superior and
inferior ends of the talonavicular joint
and (b) depth from the bottom of the
navicular surface to the line above (C and
D). (A and C female non-dancer; B and
D female dancer.)

echo; TR/TE = 3000/30 ms, NEX ms, and acquisition time = 3 min, matrix = 512 x 512, and number of
= 1, and acquisition time = 3 min, 51 s for each TSL. slices = 20.
48 s. • T2 mapping: 2D turbo spin-echo;
• T1rho mapping: 3D FS PROSET TR/TE = 3000/13, 26, 39, 52, Measurement of Distances and
(Principle of Selective Excitation 65, 78, 91 ms, NEX = 1 and total Angles
Technique); TR/TE = 7.5/3.8 ms, acquisition time = 11 min. The measurement of medial and
flip angle = 10°, NEX = 1, spin-lock All images were obtained with a field lateral subchondral bone distance be-
frequency = 575 Hz, time of spin- of view (FOV) of 140 x 140 mm, tween the tibia and talus (MSBD and
lock (TSL) = 0, 10, 20, 40 and 80 slice thickness/gap = 3/0 mm, image LSBD), axial navicular-talus axis angle
 Journal of Dance Medicine & Science • Volume 25, Number 1, 2021 41

(ANT angle), talar neck angle against 6 years of experience (TT) and one ment.17 The 95% confidence interval
the posterior talocalcaneal joint (TN board-certified radiologist (SH) sub- (95% CI) was determined for each
angle), and curvature of navicular specialized in musculoskeletal radiol- ICC. MSBD and LSBD, differences
surface (NC) at the talonavicular joint ogy with 7 years of experience. They between MSBD and LSBD, ANT
on sagittal images were performed as reviewed the images independently angle, NT angle, and NC and T1rho
per their definitions described below using Medical Image Processing, / T2 values of the talonavicular joint
by one board-certified orthopedic Analysis, and Visualization software segmentation between female danc-
surgeon with 6 years of experience (MIPAV; Center for Information ers and female non-dancers, and
(TT) and one board-certified radi- Technology, National Institutes of between female and male dancers,
ologist (SH) subspecialized in mus- Health, Bethesda, Maryland, USA). were analyzed using the Wilcoxon
culoskeletal radiology with 7 years of Cartilage segmentation was per- signed-rank test. Statistical analyses
experience. formed on T1rho images of TSL = were performed using R statistical
The images were displayed in 20 ms and T2 images at TE of 39 ms, software version 3.3.2 for Mac (The
OsiriX Lite software (Pixmeo SARL, respectively, because of their relatively R Project for Statistical Comput-
Geneva, Switzerland). The MSBD high signal to noise ratio as well as ing, Vienna, Austria). Values of p
and LSBD were measured at the to obtain a superior contrast of the < 0.05 were considered statistically
medial and lateral one-third of the cartilage with respect to the surround- significant.
talar dome on coronal images (Fig. ing tissues. Sagittal FS-PDWI images
1). The ANT angle was measured were also used as an anatomical refer- Results
as the angle between the axis of the ence of the talonavicular joint carti- Intra-Rater and Inter-Rater
talar body parallel to the medial and lage. Mean cartilage values of T1rho Reliabilities
lateral joint line and the short axis and T2 were estimated using the The ICCs of the MSBD, LSBD, dif-
of the navicular perpendicular to three middle slices. The talonavicular ferences between MSBD and LSBD,
the line defined by the medial and joint cartilages were segmented by ANT angle, TN angle, NC, T1rho
lateral proximal ends of the talona- manually placing vertices along the and T2 values are shown in Table
vicular joint surface on axial images boundaries. Both readers repeated 1. ICCs were good to excellent in
(Fig. 2). The TN angle and NC were the measurements after 1 month in a all measurements except for intra-
measured on sagittal images. The random order to determine intra-rater rater reproducibility of LSBD (0.56;
TN angle was measured as the angle reproducibility. Both readers were 95CI 0.26, 0.77) and inter-rater
between the longitudinal axis of the blinded to subject information during reproducibility of LSBD (0.54; 95CI
talar head and the line parallel to the cartilage segmentation. 0.23, 0.76), TN angle (0.41; 95CI
posterior talocalcaneal joint surface 0.06,0.67), and NC (0.57; 95CI 0.26,
of the talus. The NC was measured Statistical Analysis 0.77), which were fair.
as the ratio b/a of (a) length of the Inter- and intra-rater reliability and
line between the superior and infe- the variability of the segmented areas Medial and Lateral Subchondral
rior ends of the talonavicular joint, (pixels) were assessed using intraclass Bone Distance Between Tibia and
and (b) depth from the bottom of correlation coefficients (ICCs). 16 Talus (MSBD and LSBD)
the navicular surface to the line (a) These were interpreted as follows: 0 Medial subchondral bone distance
above (Fig. 3). Both readers repeated to 0.4, poor agreement; 0.4 to 0.6, values were significantly larger in
the measurements in a random order fair agreement; 0.6 to 0.75, good female dancers than in non-dancers
after 1 month to determine intra- agreement; > 0.75, excellent agree- (4.05 mm vs. 2.75 mm, p < 0.01),
rater reproducibility. Both readers
were blinded to subject information
during all of the measurements. Table 1 Intraclass Correlation Coefficients with 95% Confidence Intervals
for Intra-Rater and Inter-Rater Reliability
Image Processing and Cartilage
Segmentation of the Talonavicular Intra-Rater Reliability Inter-Rater Reliability
Joint MSBD 0.69 (0.45, 0.84) 0.97 (0.95, 0.99)
Images were transferred in Digital LSBD 0.56 (0.26, 0.77) 0.54 (0.23, 0.76)
Imaging and Communications in Difference (MSBD-LSBD) 0.74 (0.52, 0.87) 0.84 (0.68, 0.92)
Medicine (DICOM) format to a
ANT angle 0.87 (0.74, 0.93) 0.91 (0.83, 0.96)
computer that was used to perform all
post-processing and analyses. Manual TN angle 0.99 (0.97, 0.99) 0.41 (0.06, 0.67)
cartilage segmentations of talonavicu- NC 0.94 (0.87, 0.97) 0.57 (0.26, 0.77)
lar joint cartilage on T1rho and T2 T1rho 0.94 (0.88, 0.97) 0.97 (0.95, 0.99)
images were performed by one board-
T2 0.87 (0.74, 0.94) 0.91 (0.81, 0.95)
certified orthopedic surgeon with
42 Volume 25, Number 1, 2021 • Journal of Dance Medicine & Science

whereas no significant differences were vs. 37.8°, p = 0.720), (Table 3). There p = 0.604; and NC: 0.186 vs. 0.176,
present for LSBD (2.63 mm vs. 2.63 was no joint effusion in any dancer’s p = 0.604).
mm, p = 0.87), (Table 2). The dif- talonavicular joint.
ferences between MSBD and LSBD T1rho and T2 Values
were significantly larger in female Talar Neck Angle Against There were no significant differences
dancers than non-dancers (1.42 mm Posterior Talocalcaneal Joint (TN in T1rho or T2 values of the talona-
vs. 0.117 mm, p < 0.01). There were Angle) and Navicular Surface vicular joint cartilage between female
no significant differences in MSBD Curvature at Talonavicular Joint dancers and non-dancers (T1rho:
or LSBD between female and male (NC) 53.8 ms vs. 52.4 ms, p = 0.74; T2:
dancers (MSBD: 4.05 mm vs. 3.71 There was a trend toward smaller TN 51.5 ms vs. 51.6 ms, p = 0.48),
mm, p = 0.40; LSBD: 2.63 mm vs. angle and larger NC in female dancers (Table 4). Additionally, there were
3.07 mm, p = 0.0564; differences: compared to their non-dancer coun- no significant differences in T1rho or
1.42 mm vs. 0.645 mm, p = 0.0789), terparts, but these differences were T2 values between female and male
and there was no joint effusion in the not significant (TN angle: 16.58° vs. dancers (T1rho: 53.8 ms vs. 49.07
talocrural joint of any dancer. 22.34°, p = 0.09; NC: 0.186 vs. 0.165, ms, p = 0.0947; T2: 51.5 ms vs. 50.7
p = 0.28), (Table 3). There was also ms, p = 0.604).
Axial Navicular-Talus Axis Angles a tendency toward smaller TN angle
(ANT Angles) and larger NC in male dancers than Discussion
Axial navicular talus angles in female female non-dancers (TN angle: 18.85° The purpose of this study was to
and male dancers were significantly vs. 22.34°, p = 0.278; and NC: 0.176 compare quantitative measurements
larger than those in female non- vs. 0.165, p = 0.968), although the in MR images of the talocrural and
dancers (female dancers vs. female difference between male dancers and talonavicular joints in ballet danc-
non-dancers: 38.9° vs. 24.3°, p < 0.01; female non-dancers was smaller than ers and healthy non-dancers for use
male dancers vs. female non-dancers: between female dancers and non- in evaluating prediction of chronic
37.8° vs. 24.3°, p = 0.0220). There dancers. There were no significant injury and degenerative joint disease.
were no significant differences be- differences between female and male There were two significant differences
tween female and male dancers (38.9° dancers (TN angle: 16.58° vs. 18.85°, between female dancers and female
non-dancers. First, MSBD was sig-
nificantly larger in female dancers
Table 2 Comparison of MSBD, LSBD, and Difference (MSBD-LSBD) compared to female non-dancers,
Among Female Non-Dancers, Female Dancers, and Male Dancers whereas there was no significant dif-
MSBD (mm) LSBD (mm) Difference (mm) ference in LSBD between these two
Female non-dancers 2.75 ± 0.325* 2.63 ± 0.563 0.117 ± 0.744* groups. The results show widening
of the medial talocrural joint space
Female dancers 4.05 ± 0.903 2.63 ± 0.447 1.42 ± 0.954
in the dancers studied, which may be
Male dancers 3.71 ± 0.669 3.07 ± 0.425 0.645 ± 0.652 an indicator of potential instability.
*Statistically significant difference at p < 0.05 compared to female dancers. Second, ANT angles in female dancers
were significantly larger than those in
female non-dancers, suggesting that
Table 3 Comparison of ANT Angle, TN Angle, and NC Among Female the navicular was more adducted
Non-Dancers, Female Dancers, and Male Dancers compared to the talus in the dancers.
ANT angle (°) TN angle (°) NC This may lead to gross foot deformity
Female non-dancers 24.3 ± 10.4 22.34 ± 7.46 0.165 ± 0.0324 and potentially cause foot pain.
Medial subchondral bone distance
Female dancers 38.9 ± 8.92* 16.58 ± 8.29 0.186 ± 0.0336
widening may provide a benefit to
Male dancers 37.8 ± 11.9* 18.85 ± 8.33 0.176 ± 0.0413 ballet dancers. In the talocrural joint,
*Statistically significant difference at p < 0.05 compared to female non-dancers. motion occurs around an axis of
rotation of the talocrural joint that
passes through the body of the talus
Table 4 Comparison of T1rho and T2 Values of Talonavicular Joint and the tips of the medial and lateral
Cartilage Among Female Non-Dancers, Female Dancers, and malleoli.18 In fact, because the lateral
Male Dancers malleolus is inferior and posterior to
T1rho (ms) T2 (ms) the medial malleolus, the axis of rota-
Female non-dancers 52.38 ± 7.59 51.56 ± 8.58 tion of the talocrural joint is inclined
Female dancers 53.76 ± 4.20 51.48 ± 4.33 slightly superiorly and anteriorly as it
passes laterally to medially through
Male dancers 49.07 ± 6.19 50.70 ± 3.90 the talus and both malleoli.19 The axis
 Journal of Dance Medicine & Science • Volume 25, Number 1, 2021 43

deviates about 10° from a pure medial- position, excessive mid-foot plantar pointe position. Nonetheless, because
lateral axis in the frontal (coronal) flexion may occur because of body there are some ligaments (e.g., dorsal
plane and 6° in the horizontal (axial) weight loading in extreme plantar talonavicular ligament, or plantar cal-
plane. Because of the pitch of the axis flexion of the talocrural joint.24 As a caneonavicular, the so-called “spring
of rotation, dorsiflexion is associated result, bone alignment in dancers may ligament”) that firmly support the talo-
with slight abduction and eversion tend toward a high longitudinal arch. navicular joint, that joint itself is natu-
and plantar flexion with slight ad- However, we did not measure the rally very stable.27 Thus, the navicular
duction and inversion.19 To achieve sagittal arch of the foot in this study. may not need a deeper curvature of its
the en pointe position, female ballet The talar neck is naturally deviated surface, nor any accommodating bone
dancers need extreme plantar flexion, downward relative to the talar body morphological changes, to stabilize the
resulting in adduction and inversion and makes a 24° angle of inclina- talonavicular joint. Further investiga-
of the talus. In this position more tion.25 When en pointe, mid-foot tion is needed to test this hypothesis.
medial talocrural joint space (i.e., plantar flexion occurs,4 likely due to T1rho and T2 relaxation times
larger MSBD) is favorable to obtain- the force of a dancer’s body weight. have been widely investigated as
ing extreme plantar flexion. We measured the TN angle because imaging techniques that can identify
Another explanation for the larger the repetitive flexion force to the cartilage degeneration.28-32 T1rho re-
MSBD in female dancers may be that mid-foot in dance may cause mor- laxation time is sensitive to changes
repetitive extreme plantar flexion can phological changes of the talus in in the proteoglycan content of extra-
cause loosening or distension of the dancers, such as plantar inclination of cellular matrix,31 while T2 relaxation
posterior medial capsule, since female the talar head and neck axis compared time is sensitive to increases in water
ballet dancers undertake high levels of with non-dancers. The results were content and alteration in collagen
physically demanding en pointe train- not significantly different between content and architecture.28,29 Because
ing beginning at a young age. Dancers female dancers and non-dancers; on the talonavicular joint receives a larger
must have developmental distraction the contrary, female dancers tended to load of body weight in female danc-
of the posterior medial capsule if exhibit a smaller TN angle, where the ers than non-dancers, degenerative
they are to execute repeated extreme direction of the talar axis on sagittal changes may occur in the talonavicu-
plantar flexion. This type of posterior images was deviated more dorsally. lar joint cartilage earlier in dancers.
capsule adaptation has been reported Body weight load may be directed We measured T1rho and T2 values
previously.5 more dorsally, or perpendicular to the of the talonavicular joint cartilage and
In this study, the navicular of talonavicular joint, in the en pointe compared values of female dancers
female dancers was more adducted position.25 This may, in turn, alter the with those of female non-dancers to
against the talar body than it was in talus bone morphology for preventing identify the extent to which dancers
female non-dancers. In the en pointe subtalar subluxation.26 As just one might develop early osteoarthritic
position, load is applied primar- perspective, this tendency of morpho- changes. However, the results were not
ily along the first, second, and third logical change of the talar head and significant. Although it is clear that
metatarsals and then transmitted to neck may be a factor for acquiring there is more longitudinal load due
the cuneiforms, navicular, and talus. more firm stabilization of the talona- to body weight in the talonavicular
This axis of loading may account for vicular joint in the en pointe position. joint of female dancers, the degree of
navicular adduction, particularly in However, this matter requires further degeneration in dancers was similar
dancers, where the force vector tends study with a biomechanical approach. to that of non-dancers in our cohort.
to run medially along the first meta- In any case, TN angle may not be This may be because the participants
tarsal rather than along the third. correlated to chronic injury such as in this study were too young to have
The anatomy of the longitudinal degenerative joint disease. developed degenerative changes.
arch of the dancer’s foot also may ex- Next, we measured navicular cur- Joint stability considerations also
plain this result. In pes planus, there vature because we hypothesized that may help explain our findings. A
are some reports that the navicular the curvature of the navicular surface significant association between joint
is abducted.20,21 On the other hand, in female dancers would be deeper injury and resultant joint instability
in pes cavus Louie et al.22 reported than that in their non-dancer coun- with concomitant development of
that the navicular was more adducted terparts in order to firmly support the osteoarthritis has been clearly estab-
against the talar body. For ballet danc- talus at the talonavicular joint when lished.33-36 The knee displays a higher
ers, it may be more important to have body weight is loaded en pointe. Our prevalence of osteoarthritis than other
a high arched foot for aesthetic reasons results suggest that this is indeed the joints37 secondary to its wide range of
and because this enables them to get case; however, the difference was not motion and relatively unstable bony
to a high demi-pointe position (with significant. A possible reason for this is configuration compared to other
the toes hyperextended) and to plantar that the navicular is a key component major lower extremity joints, such
flex over the box of the shoe when they of weightbearing and is repeatedly as the hip and ankle. Therefore, the
dance en pointe.23 In the en pointe loaded with body weight in the en similar levels of talonavicular joint
44 Volume 25, Number 1, 2021 • Journal of Dance Medicine & Science

cartilage degeneration in our two high field strength magnetic reso- anterior talofibular ligament. Clin
university-age female samples are not nance imaging. Acta Radiol. 2016 Radiol. 2013 Jul;68(7):e355-61.
unexpected, as the talonavicular joint Aug;57(8):978-84. 14. Haraguchi N, Ota K, Nishida N, et
has a limited range of motion and high 4. Russell JA, Shave RM, Kruse DW, al. T1ρ mapping of articular cartilage
level of stability with firm ligamentous et al. Ankle and foot contributions grafts after autologous osteochondral
to extreme plantar- and dorsiflexion transplantation for osteochondral
support. in female ballet dancers. Foot Ankle lesions of the talus: a longitudinal
Our study has several limitations. Int. 2011 Feb;32(2):183-8. evaluation. J Magn Reson Imaging.
First, we included only females as 5. Rogers J, Dijkstra P, Mccourt P, 2018 Aug;48(2):398-403.
non-dancer controls. Future research et al. Posterior ankle impinge- 15. Cha JG, Yi JS, Han JK, Lee YK.
with control groups of both sexes ment syndrome: a clinical review Comparison of quantitative cartilage
will be needed to explore differences with reference to horizontal jump T2 measurement and qualitative MR
between male dancers and male non- athletes. Acta Orthop Belg. 2010 imaging between professional ballet
dancers. Second, the small sample size Oct;76(5):572-9. dancers and healthy volunteers. Ra-
in our case series limits the generaliz- 6. van Dijk CN, Lim LS, Poortman A, diology. 2015 Jul;276(1):199-206.
ability of our results. Third, because et al. Degenerative joint disease in 16. Shrout PE, Fleiss JL. Intraclass
of the thin cartilage layers present in female ballet dancers. Am J Sports correlations: uses in assessing rat-
Med. 1995 May-Jun;23(3):295-300. er reliability. Psychol Bull. 1979
the talus and navicular, we were not 7. Crema MD, Roemer FW, Marra Mar;86(2):420-8.
able to divide cartilage into talar and MD, et al. Articular cartilage in the 17. Rosner B. Fundamentals of Biosta-
navicular components. Although the knee: current MR imaging tech- tistics (8th ed). Boston: Cengage
depth-associated spatial variation in niques and applications in clinical Learning, 2016.
T2 value is known,12 we were not able practice and research. RadioGraph- 18. Brockett CL, Chapman GJ. Biome-
to segment cartilage in the deep and ics. 2011 Jan-Feb;31(1):37-61. chanics of the ankle. Orthop Trauma.
superficial layers. Fourth, our analysis 8. Pan J, Pialat JB, Joseph T, et al. 2016 Jun;30(3):232-8.
was performed only with MR images; Knee cartilage T2 characteristics and 19. Neumann DA. Kinesiology of the
clinical information such as symp- evolution in relation to morphologic Musculoskeletal System: Foundations
toms, physical examination findings, abnormalities detected at 3-T MR for Rehabilitation (2nd ed). Maryland
and historical data were not included imaging: a longitudinal study of Heights, Maryland: Mosby, 2010.
the normal control cohort from the 20. Chi TD, Toolan BC, Sangeorzan BJ,
in this study. Finally, because we con- Osteoarthritis Initiative. Radiology. Hansen ST Jr. The lateral column
ducted a cross-sectional survey study, 2011 Nov;261(2):507-15. lengthening and medial column sta-
longitudinal results from continued 9. Stahl R, Blumenkrantz G, Carbal- bilization procedures. Clin Orthop
physical activities are unknown. lido-Gamio J, et al. MRI-derived Relat Res. 1999 Aug;365:81-90.
T2 relaxation times and cartilage 21. Sangeorzan BJ, Mosca V, Hansen ST
Conclusion morphometry of the tibio-femoral Jr. Effect of calcaneal lengthening on
In conclusion, this study showed joint in subjects with and without relationships among the hindfoot,
larger MSBD and ANT in female osteoarthritis during a 1-year follow- midfoot, and forefoot. Foot Ankle.
dancers, which may be related to up. Osteoarthritis Cartilage. 2007 1993 Mar-Apr;14(3):136-41.
specific female dancers’ performance, Nov;15(11):1225-34. 22. Louie PK, Sangeorzan BJ, Fassbind
especially in the en pointe position. 10. Dunn TC, Lu Y, Jin H, et al. T2 MJ, Ledoux WR. Talonavicular joint
relaxation time of cartilage at MR coverage and bone morphology be-
This quantitative assessment offers imaging: comparison with severity of tween different foot types. J Orthop
additional information with which to knee osteoarthritis. Radiology. 2004 Res. 2014 Jul;32(7):958-66.
understand cartilage degeneration of Aug;232(2):592-8. 23. Lai JC, Kruse DW. Assessing readi-
the ankle and pre-markers of osteo- 11. Kim HS, Yoon YC, Sung KS, et al. ness for en pointe in young bal-
arthritis in university ballet dancers. Comparison of T2 relaxation values let dancers. Pediatr Ann. 2016
in subtalar cartilage between patients Jan;45(1):e21-5.
References with lateral instability of the ankle 24. Bickle C, Deighan M, Theis N. The
1. Peace KA, Hillier JC, Hulme A, joint and healthy volunteers. Eur effect of pointe shoe deterioration on
Healy JC. MRI features of posterior Radiol. 2018 Oct;28(10):4151-62. foot and ankle kinematics and kinet-
ankle impingement syndrome in bal- 12. Park SY, Yoon YC, Cha JG, Sung ics in professional ballet dancers.
let dancers: a review of 25 cases. Clin KS. T2 relaxation values of the Hum Mov Sci. 2018 Aug;60:72-7.
Radiol. 2004 Nov;59(11):1025-33. talar trochlear articular cartilage: 25. Kelikian AS. Sarrafian’s Anatomy of
2. Russell JA, Kruse DW, Nevill AM, comparison between patients with the Foot and Ankle. Philadelphia: Lip-
et al. Measurement of the extreme lateral instability of the ankle joint pincott Williams & Wilkins, 2011.
ankle range of motion required by and healthy volunteers. AJR Am J 26. Menetrey J, Fritschy D. Subta-
female ballet dancers. Foot Ankle Roentgenol. 2016 Jan;206(1):136- lar subluxation in ballet dancers.
Spec. 2010 Dec;3(6):324-30. 43. Am J Sports Med. 1999 Mar-
3. Russell JA, Yoshioka H. Assessment 13. Lee S, Yoon YC, Kim JH. T2 map- Apr;27(2):143-9.
of female ballet dancers’ ankles ping of the articular cartilage in the 27. De Dea M, Loizou CL, Allen GM,
in the en pointe position using ankle: correlation to the status of et al. Talonavicular ligament: preva-
 Journal of Dance Medicine & Science • Volume 25, Number 1, 2021 45

lence of injury in ankle sprains, histo- Med. 2001 Sep;46(3):419-23. Dahl LL, Roos EM. The long-term
logical analysis and hypothesis of its 31. Duvvuri U, Reddy R, Patel SD, consequence of anterior cruciate
biomechanical function. Br J Radiol. et al. T1rho-relaxation in articular ligament and meniscus injuries: os-
2017 Mar;90(1071):20160816. cartilage: effects of enzymatic deg- teoarthritis. Am J Sports Med. 2007
28. Mosher TJ, Dardzinski BJ. Carti- radation. Magn Reson Med. 1997 Oct;35(10):1756-69.
lage MRI T2 relaxation time map- Dec;38(6):863-7. 35. Onur TS, Wu R, Chu S, et al. Joint
ping: overview and applications. 32. Eckstein F, Burstein D, Link TM. instability and cartilage compression
Semin Musculoskelet Radiol. 2004 Quantitative MRI of cartilage and in a mouse model of posttraumatic
Dec;8(4):355-68. bone: degenerative changes in os- osteoarthritis. J Orthop Res. 2014
29. Goodwin DW, Zhu H, Dunn JF. teoarthritis. NMR Biomed. 2006 Feb;32(2):318-23.
In vitro MR imaging of hyaline Nov;19(7):822-54. 36. Blalock D, Miller A, Tilley M, Wang
cartilage: correlation with scanning 33. Brown TD, Johnston RC, Saltzman J. Joint instability and osteoarthritis.
electron microscopy. AJR Am J CL, et al. Posttraumatic osteoar- Clin Med Insights Arthritis Muscu-
Roentgenol. 2000 Feb;174(2):405-9. thritis: a first estimate of incidence, loskelet Disord. 2015 Feb 19;8:15-
30. Akella SV, Regatte RR, Gougoutas prevalence, and burden of dis- 23.
AJ, et al. Proteoglycan-induced ease. J Orthop Trauma. 2006 Nov- 37. Zhang Y, Jordan JM. Epidemiology
changes in T1rho-relaxation of ar- Dec;20(10):739-44. of osteoarthritis. Clin Geriatr Med.
ticular cartilage at 4T. Magn Reson 34. Lohmander LS, Englund PM, 2010 Aug;26(3):355-69.