2.1. Study design

This study used a qualitative design based on in-depth interviews to understand dancers’ perspectives. Some dancers with current limitation to dance because of pain were included. Most or all dancers will experience pain and injury at some point, and we acknowledge that this will shape their perspective of the relationship between posture and movement. Further, we included ballet and contemporary dancers. We considered these two issues to be critical to ensure diverse perspectives were captured. We used semi-structured group interviews to collect data about professional dancers’ beliefs and conceptualisation of posture and movement. Themes from narratives provided by participants were derived via thematic analysis using an inductive approach.

2.2. Participant selection

Thirty-six individuals who self-identified as elite ballet or contemporary dancers were recruited via social media, word-of-mouth, dance community notice boards and through the networks of professional dancers in New York City. Recruited participants were encouraged to invite colleagues to consider participation. Selective sampling was used to ensure a range of ages, dancing styles (both ballet or contemporary, and from different companies) and history of musculoskeletal pain (e.g., current and previous pain, different regions of the body) were represented. To be included, participants had to be professional ballet or contemporary dancers earning a primary wage as a dancer or full-time students in a university setting with dance as their current major, and 18 years of age or older. The study was approved by the institutional Medical Research Ethics Committee (The University of Queensland: 2019001451) and conducted in July 2019. Participants provided written informed consent.

2.3. Procedure of data collection

Focus group discussion has been shown to be useful for gathering information from people with similar backgrounds to discuss a specific topic of interest and was used here [20]. Five face-to-face semi-structured focus groups were directed by three researchers (PH, ML and MH) and each was conducted with a different group of seven or eight dancers. The topics for discussion were developed by PH, ML and MH and involved informal discussion with a group of dancers and dancer teachers prior to the formal focus groups. The discussion during the focus groups followed an interview outline (see Supplement S1 File for semi-structured focus group guide). PH led the focus groups. He has experience in leading focus groups and qualitative research but has limited experience in dance which enabled him to remain impartial regarding the main purpose of this study. During the focus group, with input from ML and MH, PH proposed issues for discussion according to the interview outline, and if necessary, clarified meaning and probed more deeply. Due to the exploratory nature of this study free-flowing conversations were allowed to enrich the depth and breadth of the data. The script was piloted with a group of 5 dancers 12 months prior to the focus groups. During these discussions, dancers were asked to reflect on issues such as ‘when’, ‘how’ and ‘why’ they think about their posture and movement.

Each focus group discussion lasted for approximately one hour and was audio-recorded. Saturation, determined through iterative data analysis during the period of data collection, was reached when repetitive information with sufficient richness and variability related to study aims could be found. Demographic data regarding age, gender, dance experience and age were collected via a paper form provided to the participants on arrival at the focus group.

2.4. Theoretical underpinning

Thematic analysis as described by Braun and Clarke [21] was conducted to generate understanding of dancers’ views and experiences on posture control from broad perspectives. An inductive approach was applied to minimise the influence of pre-existing scientific exploration and our own theoretical interpretation of results. This procedure first identified codes by selecting elements of the raw data that conveyed meaningful information. Themes and sub-themes emerged from these codes as broader categories that represent patterned response.

2.5. Data analysis

The research team consisted of two final year physiotherapy students (AG and XZ) and three senior academics (PH, MH and ML). MH and ML are experienced physiotherapists and researchers who have worked with professional dancers and dance-related injuries for over 25 years. MH is a former professional dancer. PH is an expert in movement control, pain and rehabilitation. The recorded interviews were transcribed verbatim using Otter software (AISense, Mountain View, USA). Two investigators (AG and XZ) examined the transcribed interviews to correct any errors from the automatic transcriptions. Participants were de-identified during this process. As AG and XZ familiarised themselves with the data, they identified and sorted significant parts of dancers’ opinions into codes and then categorised them into themes using data management software (Nvivo version 12, Lumivero, Denver USA). The relevance of the themes was considered for the purpose of this study and only themes related to dancers’ conceptualisations and beliefs of posture and movement were extracted. PH independently read the transcripts and also identified relevant themes. AG, XZ and PH met to identify themes, to discuss accuracy of theme identification, and to refine the terminology used to describe them. Four major themes were identified. AG and PH developed graphical representation of the themes to consider subthemes, and to consider the connections within and between the themes. The graphical representation of the initial themes were presented to MH and ML as the starting point for a series of four on-line discussions to gain their insight within the contexts of their expertise in dance-related research and clinical experience. These discussions resulted in modification and refinement of the themes. Consolidated Criteria for Reporting Qualitative Studies (COREQ) questionnaire was used to ensure rigour in reporting results [22].

3.1. Theme 1: Posture being conscious or subconscious

Participants acknowledged that posture during dance and non-dance was controlled both consciously and subconsciously (Fig 1).

3.2. Theme 2: Posture in dancer mode and non-dancer mode

As dancers reflected upon their postures, they conceptualised two different contexts for posture – “dancer mode” and “non-dancer mode”, which alters in its form based on contextual demand (Fig 2).

3.3. Theme 3: Posture as an ideal

This theme considers the conceptualisation of posture in general and posture as an ideal state (Fig 3).

3.4. Theme 4: Adjustment and improvement of posture

This theme focused on how dancers adjust or “improve” their postures and three non-mutually exclusive subthemes were observed and discussed (Fig 4).

4.1. Postural control is a continuous motor learning process

Dancers reinforced that when learning a novel or unfamiliar movement and posture, the level of attention transitions from conscious to subconscious after repetitive task practice. This concurs with the widely accepted models of motor learning [23], which consider three progressive stages: cognitive, associative, and autonomous, with reduced attention as motor learning progresses [24]. Although it might be expected that professional dancers would perform at the autonomous level with low attentional demand, they acknowledged that despite repetitive practice, absence of ongoing conscious attention over time can “degrade” the quality of posture from its desirable form. Dancers suggested that the permanence of motor skill acquisition cannot be assumed, and high-level refinement required continuous conscious effort for maintenance.

That dancers describe attention to specific postures and movements (internal focus of attention) that might appear to contradict the concept that movement performance is better when attention is directed towards the external effects of an action (external focus of attention) [25]. There are two issues to consider. First, dancers implied that this was not considered in every task but described a concept of “checking in” periodically. Second, continued engagement of cognitive awareness to fine-tune movement concurs with observations of Ericsson [19] that suggest learning differs between skills needed for everyday tasks and those needed for expert performance. For everyday activities the goal is to learn a movement with satisfactory quality in minimum time and become autonomous. In contrast, expert performance involves increasingly complex mental representations to attain higher levels of control of performance that remain within the cognitive and associative phases [26]. Dancers also identified greater conscious awareness in “dancer” than “non-dancer” mode to improve the quality and efficiency of dancing. These findings suggest dancers acknowledge the necessity to maintain a constant high degree of conscious awareness to continuously optimise and refine the precision of posture, despite their advanced expertise in movement.

4.2. Posture and movement are highly individualised

Dance is generally considered to involve achievement of specific idealised dance postures, yet dancers argue that individual variation is required to achieve this goal. This is not unique to dancers. Recent data show individual differences in muscle activation patterns in gait, despite similar performance of the task [27]. Of note, dancers argued that a feature of maturity as a dancer was the realisation that they can, and should, find their own optimal solution to match their own biomechanics, kinematics and physiology and not merely duplicate others to achieve the dance aesthetics. Some examples have been researched. Individual differences in shape of the acetabulum, and neck of femur can limit the range of hip external rotation and requires compensation by other segments, such as greater lumbar lordosis, external rotation of the knee, or valgus heel with forefoot pronation [28]. Although these compensations might enable the position required by dance, this may place additional stress on segments that compensate. Absence of compensation is also likely to be problematic; dancers who lack sufficient ankle plantarflexion for pointe or demi-pointe positions, will often attempt to force plantarflexion and place greater stress on the posterior and lateral ankle structures, and predisposes themselves to varus malalignment at the foot-ankle complex [29]. Consideration of individual variation and whether to compensate or train requires consideration.

Dancers also indicated that achievement of a desired posture depends on elements of fitness beyond the specific muscles and joints involved in the task, such as physical capacity and “core stability”, which refers to training of trunk muscles to maintain control of the proximal regions of the body, particularly the spine, during limb movements. Consistent with this proposal are data that show improvement in ballet dancers’ alignment of demi-plié and passé dance positions after participation in a summer intensive dance program that trained lower abdominal muscle strength [30]. This concept reinforces that optimum state of posture requires attention to features beyond those of each individual segment.

4.3. The dynamicity and adaptability of posture

Dancers emphasised the static and dynamic nature of posture. A common concept was that posture is inherently dynamic and can never be static. In quasi-static situations, such as standing, dancers reinforced that maintenance of alignment relies on constant contribution of muscle control. This principle is congruent with evidence that even in quiet stance, small motions of spine and pelvis are required to counteract the disturbance to postural equilibrium, including those caused by breathing [31].

Dancers used the term “kinaesthetic potential” to describe the concept that maintenance of alignment in a dynamic manner enables availability of adjustments during motion such that if perturbations during movements can easily be corrected. This concurs with the view that describes stability of a dynamic system as “the ability to maintain the desired trajectory despite kinetic, kinematic, or control disturbances.” [32]. This dynamic control is particularly important for dancers who move through large or perhaps “unnatural” trajectories.

Postural control depends on the environment, intended goals, biomechanical constraints, and sensory conditions [33] and must be adaptable to changes in these elements. Some dancers indicated that components of dance-related posture and movement are preserved in everyday activities, either consciously or subconsciously, with positive and negative consequences for health. Although some data suggest gait patterns used by dancers generate greater strain on lateral ankle structures [34], other work shows that when dancers transfer their landing technique to non-ballet-specific tasks this is associated with a low rate of ACL injuries [35]. Whether this translation between contexts is beneficial or harmful has yet to be clarified.

4.4. Relationship between posture/movement, injury and pain

Dancers expressed a strong belief that “poor” posture places them at risk of pain and injury, and that “good” postures can prevent injury and improve efficiency of dancing. A common perspective was that sometimes injury and pain were induced by incongruency between the dance aesthetic and musculoskeletal health. That is, that postures and movements that they considered to be good/ideal in non-dance contexts might protect them from injury, but with the caveat that some aspects of posture and movement that were required for dance placed them at risk because they were too “extreme” and “unnatural”. Findings from several studies support this view: systematic review of evidence suggests specific spinal alignment increases risk for injury in professional ballet dancers [28,36]; hyperflexibility of hip joint and extreme hip motions increase susceptibility for hip impingement and instability [37]; and gait patterns commonly used by dancers induce greater medial shear force in pre-swing gait phases than that of non-dancers [34].

It is difficult to extrapolate studies that question the protective nature of posture and movement [14,15]. to dancers. The relationship needs to be considered in the context of demands and cummulative load of professional dance. For example, pain and injury may be related to demand, but outcomes might be worse if posture and movement were less ideal. Additionally, it is possible that the relationship is more complex than a focus on a single negative aspect of posture (as is often assumed in previous research). A relationship to injury might be more likely in professional dancers because of the substantial stress on the body as a result of the demands of dance, for example, the load placed on the lumbar spine to achieve greater turnout during routine movements and positions [28] and demanding training loads [36]. Regardless of the evidence, prevention of injury is a primary motivator for dancers to attend carefully to postures and movements. Further investigation is required to examine whether such attention prevents injury.

A major motivator for dancers to maintain unique postures and movements is the aesthetic demands of dance. Some dancers defined ideal posture as based on aesthetic standards, striving for an “unnatural perfection”. This implies dancers are willing to compromise what they consider to be optimal biomechanics for artistic purposes, exemplified by the increasing elevation angle of the lifting leg, which is argued to increase stress on adjacent regions and induce pain [37,38]. Thus, dancers express a quandary; to either attain specific postures to prevent injury, or exaggerate movement for aesthetics with possible impact on susceptibility to injury [37]. This duality was criticised by some dancers. To address this, some dancers expressed a need to seek and establish postures and movements that consider the uniqueness of their own body structure to achieve their perceived ideal posture as discussed below.

4.5. Strengths and limitations

The strength of this study is the inclusion of a large group of dancers from highly competitive elite dance companies and University programs. The findings of this study need to be considered in light of several limitations. First, only 22% of the participants identified as ballet dancers and views might be biased to those of contemporary dancers. Of relevance here, contemporary dance is not restricted by classical ballet conventions and can include more fluid movements [39]. As it was impossible identify the identity of all contributors in the transcriptions it is not possible to relate each comment to the dance training or experience of the individual responsible. Second, the study involved ballet and contemporary dancers from New York City, and findings may not be transferable to individuals in other environments where there might be different concepts of dance training, different cultures, and so on. Third, a weakness of focus group as a data collection method is the potential for dominance by outspoken individuals, as their opinions can potentially influence the other participants’ answers.

4.6. Practical and clinical applications and implications

Insights into dancers’ perspectives of posture and movement have potential relevance for conceptualisation and rehabilitation of posture and movement in non-dancers. Although our qualitative study cannot determine whether the perspectives of dancers are accurate, it does constitute the opinion of “experts” in posture and movement and justifies consideration of several issues.

First, the conceptualisation of dancers that ongoing conscious effort is required to fine tune the “quality” of their posture and movement and prevent deterioration of their desired form suggests that when a patient/client is asked to change their posture or movement, it might be unrealistic for clinicians to expect permanence and that ongoing maintenance might be required.

Second, considering the individual variation in anatomy, physiology, and biomechanics, rather than aiming for a single ideal, individualisation of training and individualisation of the goal is likely to be necessary.

Third, conceptualisation of posture as dynamic contrasts the common clinical focus on training individuals to maintain static postures [40]. This reinforces that posture involves a balance of movement and stiffness that depends on the demands of the task and are never completely static [5,41]. For function, training of posture is likely to benefit from be dynamic within diverse environments, with varying sensory feedback, and different levels of difficulty, in order to enhance posture’s adaptability and transferability.

Fourth, although this study does not resolve the debate whether posture and movement are relevant for prevention and management of pain and injury, it highlights potential explanations for the apparent contradiction between dancer’s high rates of pain and injury, despite enviable quality. Namely, the contrasting issues of optimisation of posture and movement to prevent injury, and risk derived from demand to achieve unnatural tasks.