Musculoskeletal injuries are considered to be very common among dancers. Incidence of injury varies from 17% to 95 % (Bronner et al., 2003). It has also been stated that injury incidence is 4.4 per 1000 hours, and 6.8 injuries per dancer within a year (Allen et al., 2012). Overuse injuries are predominant, such as tendinosis, stress fractures and apophyseal injuries that manifest at the lower limb, hip and spine (Allen et al., 2012; Hincapié et al., 2008; Jacobs et al., 2012; Motta-Valencia, 2006; Steinberg et al., 2013). Thus, mastering dance technique requires repetitive physical loading that may exceed the limits of anatomical and physiological capabilities, and lead to injuries. Further understanding of the biomechanics of dance is essential to identify the specific musculoskeletal demands placed on the body and to uncover pathomechanics that may lead to injury (Motta-Valencia, 2006). To reduce both the injury incidence and prevalence it is crucial to enhance effective injury prevention strategies, rehabilitation, and biomechanically safe and efficient technique training practices as well dance pedagogy.

Research to date in dance biomechanics has focused on impact (such as jumps) and high pressure (e.g. pirouettes) movements as well as describing typical ballet movements. Other areas of interest include point shoes, motor control, dancing regimens, dance floor, motor skills of novice and professional dancers, and differences between barre work and center floor (Kadel and Couillandre, 2007; Koutedakis et al., 2008; Krasnow et al. 2011; Krasnow, 2012; Laws, 1985; Ward 2012). To our knowledge, there is little research on the effect of the commonly found apophyseal injury on dancer’s movement patterns and the challenges these injuries may pose to technical demands of dance and dance pedagogy. In the present case report, the aim is to compare the kinematics of the symptomatic (i.e. ischial apophyseal symptoms) and contralateral lower limbs, as well as associated reported pain of a ballet dancer with a left ischial apophysitis during ballet exercises. The purpose of this study is to inform dancers, their teachers and health professionals about movement changes that may present with this common but complex and underreported injury in a dance context.

A 27-year-old professional male classical dancer with left ischial apophysitis volunteered to participate in the study. The diagnosis was made by a specialist physiotherapist with relevant PhD, supported by clinical history, physical assessment, MRI findings and medical specialist’s (sports physician) opinion. The participant also presented with concomitant, but less symptomatic, left heel and achilles pain, and longstanding right groin pain primarily at the origin of adductor muscles (i.e. right pubis and the symphysis pubis).

The pain in the ischial tuberosity had been ongoing for about 6–8 months, and was described as vague. Self-reported pain via numeric rating scale (NRS; self-reported pain score in integers: 0–10) (Ferreira-Valente et al., 2011; Jensen et al., 1986) was usually 3, and 6 on its highest. The pain was aggravated mainly by stretching and during rapid movements. The participant reported that both the injuries at the pelvis had been treated with relative rest to relieve the pain (e.g. modification in training load and limitation of pain aggravating activities), therapeutic exercises, ice and ultrasound. The dancer stated that the injuries restricted range of motion (ROM) in general.

Data collection took place on an indoor synthetic track surface in the Motion Laboratory at the University of the Sunshine Coast (Australia). Prior to the testing, the parti-cipant was provided with a Research Project Information Sheet and given the opportunity to ask questions about the test protocols before signing a consent form. The participant also completed a medical screening questionnaire prior to the test. The study was approved by the institutional Human Research Ethics Committee.

Three-dimensional (3-D) kinematic data were collected at a sampling rate of 500 Hz, with a nine-camera motion analysis system (Qualysis Motion Capture System; Qualysis AB, Gothenburg, Sweden). Kinetic data (ground reaction force, GRF, x, y, z) were recorded via two force plates (Bertec; Bertec Corporation, Ohio, USA), sampling at 2000 Hz. Anthropometric parameters (inclu-ding height and weight) were also measured.

A total of 40 individual retro-reflective markers (16 mm) and four clusters (4 markers each) were attached at specific anatomical locations on the head, upper and lower limbs, and trunk according to the protocol of the University of the Sunshine Coast. Specific marker sites were right and left temples, right and left zygomatic bones, right and left lateral sides of acromions, manubrium of sternum and inferior part of sternum (sternal body), 7^th cervical spinous process (C7), 6^th and 12^th thoracic spinous process (T6, T12), 2^nd and 4^th lumbar spinous process (L2, L4), sacrum (S2), right and left posterior and anterior superior iliac spines, both the greater trochanters, clusters of markers for right and left thigh segments, both the lateral and medial femoral epicondyles in the right and left, right and left lateral and medial sides of tibial condyles, right and left tibial tuberosities, cluster of markers for shanks, right and left malleoli, both the heels (LHEEL, RHEEL) and 1^st and 5^th metatarsal bones in the right and left. (Bishop and Kerr, 2010) The acromion landmark was modified from lateral acromion to the spine of scapula (~ posterior to the acromioclavicular joint) to improve marker tracking, particularly as the arms were held in second position throughout the exercises.

The test battery commenced with a static capture (standing in anatomical position) that was followed by dynamic captures (i.e. ballet tasks). The subject performed a total of seven different tasks including demi-plié (i.e. small flexion of the knees), grand plié (i.e. full flexion of the knees), battement développé (i.e. gradual unfolding of the leg to front, side and back), grand battement (i.e. rapid lift of the straight held gesture leg into hip level or higher to front, side and back), sauté (small jump), grand sissonne (à la côté) (i.e. jump with elevation of the gesture leg to side), and grand jeté (en avant) (i.e. large jump where legs are thrown to 90 degrees, gesture leg in front) (see Figure 1). The selected elementary ballet movements represent the most common vocabulary used in ballet dance technique, depict weight bearing (both on two feet and on one foot) and require variation in range of motion, complexity, force and velocity.

All single lower limb exercises were repeated on both legs. Five representative trials of each task were on in the first position of the feet. Hands were held in the second position throughout the exercises. Seven attempts were allowed to obtain five representative trials. Acceptable trials were defined as those in which the participant completed the task on the force plates as required.

To minimize the risk of further injury, the participant was given the opportunity to warm-up prior to and after the attachment of the markers, and to cool down after the test. Rest intervals were allowed for 30 seconds between the tasks as required. The participant was also given a practice trial before each task, to be more familiar with the surface and the force plates. After completing each task, the participant reported pain via NRS.

Raw data for markers (trajectories) were labelled according to respective anatomical landmarks and trimmed of redundant pre and post task data. Gaps under 100 ms (i.e. 50 frames at 500 Hz) were gap-filled. The 3D coordinate data were then modelled using standard biomechanical software (Visual3D; C-motion 4.96.3, Inc. Maryland, USA) to construct a 7 segment rigid body model of the pelvis and lower limbs (left and right: thigh, shank and foot). Data were smoothed using a 4 Hz second order low pass digital filter prior to the construction of the model. A global reference system (GRS) was established with the positive y-axis in the intended direction of travel (anterior-posterior), the x-axis perpendicular to the intended direction of travel (positive direction to the right, mediolateral) and the positive z-axis pointing vertically upwards (vertical). The movement phases were defined by start and end points (creating movement events such as start and end of plié, deepest point of plié). Pain (NRS) and the following kinematic and kinetic parameters were defined for analysis: joint angle (6 DOF ankle, knee and hip; pelvis), and ground reaction force (3D; x, y and z axis).

Data reflecting relative weight-bearing were produced for the three ballet movements that were performed on both feet: demi-plié, grand plié, and sauté. In all three of these movements, weight-bearing was asymmetrical. Generally, increased or preferential weight-bearing was observed on the left lower limb, for example in demi-plié (see Figure 2).

With the comparison of the kinematics, dissimilarities between lower limbs were observed consistently in both open and closed kinetic chain tasks but particularly in the closed chain tasks. Differences in movement planes were noticeable particularly in frontal (particularly knee adduction / abduction) and horizontal (particularly hip rotations) planes. Differences in sagittal plane movements were observed, particularly in dorsiflexion. A reduction in movement amplitude (i.e. range of movement, height of the gesture leg) of the left hip flexion was noted during grand battement en avant and grand jeté en avant. Intertwined recurring dissimilarities were apparent in rotations: as increased external rotation was observed in the left hip, the right knee showed increased knee adduction and internal rotation (see Figure 3).

Increased anterior tilt in the pelvis was also frequently apparent throughout most of the test battery. Consistent pelvic tilt strategies were observed during higher amplitude movements (e.g. battement développés and grand battements), where posterior tilt was apparent in devant (i.e. front) and á la second (i.e. side), and anterior tilt in á la derriere (i.e. back). The grand jeté en avant highlighted a different movement strategy (see Figure 4) where left hip flexion (i.e. leading lower limb) was performed in posterior pelvic tilt, whereas right hip flexion was associated with anterior pelvic tilt.

Pain increased with the progress of the test battery, within the limits of the participant’s usual practice and performance (see Table 1). Prior to commencement and during the demi-plié, the self-reported pain was 1 (NRS 0–10). It increased in grand plié to 3 during the extension phase of the movement. Pain was maintained and plateaued at 3 during the battement développé; with the exception of the left leg à la derriere (i.e. as leading leg), which resulted in pain of 5 in the right pubic area. Similarly, in the grand battement series, pain was 2 in movements other than leading leg. Directions of devant, and à la second reproduced pain of 4 in the left as leading leg. During à la derriere pain decreased to 3. In sauté and sissonne the pain was reported to be 3. While performing grand jeté en avant pain increased: pain was 4 while right leg in front (around the pubic bone), and 5 while having left leg in front (ischial tuberosity). It is noteworthy that the symptomatic lower limb (i.e. the left; en avant) in grand jeté reached a reduced ROM of only 45 degrees of hip flexion, while the right was executed into 90 degrees.

This article focused on the three-dimensional biomechanics of ballet exercises. Asymmetrical ground reaction forces were consistent with preferential weightbearing on the left (i.e. symptomatic) lower limb. In addition, differences between the symptomatic (i.e. left) and contralateral (i.e. right) lower limb kinematics were observed in all movement planes. Both increased knee adduction and medial rotation, and to a lesser extent anterior tilt in the pelvis, were consistent with commonly described compensatory movements for inadequate and uneven hip turnout (lateral rotation). These compensatory strategies in the kinetic chain refer to anterior tilt in the pelvis, external rotation of the knee and hyper pronation (Bennell et al., 1999; Grossman et al., 2008; Hamilton et al., 2006; Negus et al., 2005) that could be observed in this study as well.

Increased anterior tilt in the pelvis was also frequently apparent throughout most of the test battery. Consistent pelvic tilt strategies were observed during (e.g. battement développés and grand battements), where posterior tilt was apparent in devant and á la second, and anterior tilt in á la derriere. The position of the pelvis during the higher amplitude movements were consistent with the previous literature that suggests that in similar positions such as grand rond de jambe en l`air and développé arasbesque the pelvis should attend actively with the leading leg when 90 degrees or more is accomplished; either in tilting posteriorly (while leading leg is in front), anteriorly (leading leg in back), or laterally (leading leg to the side) (Wilson et al., 2007). On the contrary grand jeté en avant indicated a different movement strategy in the gesture leg. This might be in relation to the traction forces, rapid movement and demand of large range of motion. Overall, pain was aggravated with progress of the test battery, and it was at its highest during fast movements with the requirement of large amplitude (i.e. grand jeté avant NRS 5, grand battement devant and à la second NRS 4) in the gesture leg that exerted traction on the apophyseal area (Adirim and Cheng, 2003; Eich et al., 1992; Kujala et al., 1997; Ogden, 2003; Paluska, 2005).

It is essential to acknowledge that specific motion analysis and movement description concerning the ischial apophysitis must be interpreted whilst remaining mindful of the other injuries in the dancer. However, this is not an unusual array of concomitant issues and a case study offers an ideal mode for exploration of this complex presentation. It is also important to note that the tasks performed in this study were relatively elementary ballet movements. Perhaps more vigorous and complex tasks, or movements with larger ROM (range of motion), may produce different movement changes than those presented here. It is acknowledged that the limitations of this case study (e.g. one participant with multiple injuries) demands that the reader should only generalize findings with caution. As with all case studies, the methodology does not allow for demonstration of causation or association.

The dance medicine literature is heterogeneous and a young discipline (Hincapié et al., 2008). Large gaps remain in the literature with respect to specific dance injuries and their relationships, including widely varying prevalence and incidence rates and a lack of published evidence on injuries. This may be due to a dance culture that underreports injury, difficulties in recruiting elite dancers for research or the difficulties in controlling experimental variables and cohort homogeneity in a target population with high rates of concomitant pathology. In the face of these challenges to clinical research in dance, case studies offer opportunities to break new ground for research evidence, and describe these complex situations to inform injury prevention and rehabilitation.

During the test the pain related to the left ischial apophysitis was associated with reduced amplitudes especially in fast ballet movements that required large range of motion in flexion and adduction in the left hip joint. Several recurring dissimilarities between the lower limbs could be found in the joint angles. Thus the case study indicates that there might be differences in movement patterns between the limbs in a dancer with ischial apophysitis. This may limit dance technique and performance.