Early specialisation in young gymnasts: a mixed methods study protocol

Abstract

Over the past few decades, we have seen an increase in the popularity of organised sports among youth, especially gymnastics. Along with this increase in the popularity of organised sports, sports specialisation is also on the rise. Some argue that specialising as early as possible is essential for better performance and skill development. However, it has been associated with negative mental health effects in children and adolescents, as well as a higher risk of overuse injuries in young athletes. Although sports specialisation has been a popular research topic in recent years, many questions regarding its underlying factors and its impacts (positive or negative) on injuries, psychosocial health and performance remain unanswered. The purpose of this study is to present the protocol for a study that aims to describe sports specialisation in young gymnasts and to understand the association between sports specialisation, injury, psychological health and physical performance. This study will use a convergent mixed methods design. There will be a qualitative phase where we will conduct interviews and focus group discussions with athletes and different actors in the field. This will be conducted alongside a prospective cohort study over an entire sports season. Athletes will be assessed at the start, middle and end of the season for skill acquisition (core strength and endurance, power and dynamic postural control), psychosocial variables (stress, personality traits, anxiety and quality of life) and sports participation. They will be followed throughout the season using online weekly surveys to monitor training load and injuries.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Early sport specialisation appears to be associated with increased overuse injuries and negative psychological outcomes in various sports. However, the mechanisms underlying these associations are not sufficiently understood, as is the role of early specialisation in gymnastics training.

WHAT THIS STUDY ADDS

• This study will provide valuable and new insights into different stakeholders on early sport specialisation in gymnastics. The robust design will allow us to prospectively investigate the impact of early sport specialisation on skill acquisition, psychological variables and overuse injuries. Moreover, controlling for confounding variables will allow a better understanding of the mechanisms through which early sports specialisation affects dependent variables.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• The results will allow us to better understand how current training practices in gymnastics impact skill acquisition, psychosocial development and overuse injuries. Findings will inform training choices and policies, ultimately promoting inclusive and sustainable athletic development for gymnasts.

Introduction

Participation in organised sports during adolescence has shown positive effects on stress, coping abilities, well-being and motivation to be active.^{1 2} In organised sports, gymnastics is a very popular discipline.³ In addition to the benefits of physical activity, gymnastics participation can positively affect bone health,⁴ muscle health⁵ and motor skill training.⁶ Due to the complex motor skill training required, gymnasts tend to start participation in their discipline earlier than other sports and have a higher rate of early sport specialisation.^{7 8}

The definition of early sports specialisation varies across studies. However, it is usually characterised by six components: (1) year-round training, (2) high intensity of training, (3) focus on a single sport, (4) excluding other sports, (5) deliberate practice and (6) young age threshold.⁹ Most studies converge towards a definition in which early specialisation involves dedicating a significant amount of time and effort to one sport from a young age, typically 13 years old, to become an elite athlete.¹⁰ A framework to explore early specialisation in sports has been developed recently, as seen in figure 1.¹¹

Figure 1. Adaptation of the framework for exploring early specialisation in sport. Adapted from Mosher et al.¹¹.

This model suggests early specialisation occurs through antecedent conditions, which can have a near or far influence on athletes. While sports participation has predominantly been associated with positive physical and psychological outcomes, early sport specialisation also bears the potential for negative outcomes, including injury risk, psychosocial development and specific performance-related skill acquisition.

The necessity to specialise early to become an elite athlete is still unclear.11,13 A recent meta-analysis suggests that multisport practice in childhood and adolescence, starting their main sport later, is more strongly related to becoming a world-class athlete.¹² Moreover, long-term higher performance is associated with athletes getting involved with their main sport at an older age, irrespective of sex or gender.¹³ In gymnastics, because of the technical nature of the sport and the fact that athletes typically peak before full maturity, specialisation is common from a very young age.⁸ Some studies have found that sport specialisation in young gymnasts was not associated with higher fitness or functional movement competency.⁷ In contrast, others have found that sport-specific training in youth improved physical fitness and did not negatively impact cognitive and academic performances.¹⁴

While research interest in early sport specialisation is growing, its positive and negative impacts remain largely unknown. This could be partly because studies have used retrospective or cross-sectional designs, which may not reflect the process through which early sport specialisation occurs and its impact on physical and psychosocial development. These designs also seldom integrate the perspectives of coaches and other stakeholders essential to athletes’ development, leading to a gap in our understanding of how gymnastics’ early sport specialisation is viewed.¹⁵ Furthermore, previous studies have rarely investigated factors that could confound the effect of sports specialisation on skill acquisition, psychological health and injuries.¹¹ Although it has been depicted as harmful to youth, these claims remain to be investigated in robust research designs such as long-term prospective studies, allowing us to track the changes proposed to occur with sports specialisation.^{11 15 16}

Aims

This study describes the research protocol of convergent mixed methods design aimed at exploring early sport specialisation in youth and adolescents participating in gymnastics and its impacts through four objectives.

The first objective is to explore the experiences and perceptions of coaches, trainers, federation representatives, parents, adolescent athletes and retired adolescent and young adult athletes with early sport specialisation and its positive or negative impacts on gymnasts’ development.

The second objective is to evaluate whether early sports specialisation impacts skill acquisition throughout a gymnastics season.

The third objective is to investigate associations between early sports specialisation and (1) psychosocial factors and (2) injuries during a sports season.

The fourth objective is to investigate the moderator effect of psychosocial factors on the association between injuries during the season and early sports specialisation.

Hypotheses

We hypothesise that there will be differences in skill acquisition (eg, level of execution of gymnastics-specific movements) throughout the season based on the athletes’ sport specialisation background. Furthermore, early sport specialisation is expected to be associated with psychosocial variables (eg, stress, athletic identity, coping mechanisms, anxiety and quality of life) and injury incidence. We expect psychosocial factors to influence the associations between sports specialisation and injuries throughout the season. Due to conflicting findings in the literature, no specific direction has been hypothesised concerning the relationships between variables.

Methods

Research design

A convergent mixed methods design will achieve the study’s aims and generate comprehensive data, as illustrated in figure 2.¹⁷ The first objective will be reached through qualitative study using semistructured interviews and focus group discussions. This will allow for a better understanding of athletes’ and stakeholders’ opinions on training practices and early specialisation in gymnastics. This phase will follow the guidelines outlined in the Consolidated Criteria for Reporting Qualitative Research checklist.¹⁸ Concurrent with this phase, quantitative methods involving a prospective cohort study will be used to achieve the second, third and fourth objectives. This second phase will follow the guidelines of the Strengthening the Reporting of Observational Studies in Epidemiology checklist.^{19 20} The qualitative data will provide insights into the results obtained, help contextualise the quantitative findings and allow for the triangulation of results related to skill acquisition, psychosocial health and injuries.²¹

Figure 2. Convergent mixed methods design. AIMS, Athletic Identity Measurement Scale; CISS, Coping Inventory for Stressful Situations; ESS, Exposure to Sport Scale; GFMT, Gymnastics Functional Measurement Tool; PedsQL, Pediatric Quality of Life Inventory; RESTQ-36, Recovery and Stress Questionnaire for Sports; STAI, State-Trait Anxiety Inventory—Children & State-Trait Anxiety Inventory.

Field involvement

This project has been developed in collaboration with three athlete-partners who are competitive gymnasts at the national and elite levels in Canada and the Netherlands. They are also involved in medicine and health sciences research or training outside of gymnastics. They have been invited to participate in the project’s elaboration, comment on the interview guides and questionnaires, and contribute to recruitment efforts. Their involvement adds practical knowledge and experience to the team’s combined expertise. Moreover, the athlete-partners will be invited to advise on the relevance of the results for athletes and coaches and advise on the best ways to communicate results with the field. They will be of invaluable help in ensuring that the conceptualisation and results of the study apply to the context of gymnastics. The support from relevant gymnastics organisations (GymCan, Gymnastique Québec, DutchGymnastics and Fédération Internationale de Gymnastique) will allow the project to benefit from insights from experts in gymnastics and physical activity.

Qualitative phase

Study settings

Qualitative data will be collected over the 2024–2025 and 2025–2026 sports seasons. The semistructured interviews and focus group discussions will explore the experiences of sports specialisation and how it impacts athletes. As early specialisation can be controversial,²² semistructured interviews will allow for personal insights and experiences in an open environment that will encourage participants to express their opinions freely. Focus group discussions will also be used since the stakeholder interactions could lead to new perspectives and emerging themes.

JBP and CSB developed the interview guide. It is based on the early specialisation framework (figure 1). It explores participants’ experience in gymnastics, the factors that led to them specialising or not (for athletes) and the impacts (positive or negative) on athletes. The interview guides vary between athletes, retired athletes, coaches/trainers/federation representatives and parents of athletes. All guides were pilot tested with the athlete-partners and adjusted as necessary. Focus group discussions will occur following the semistructured interviews, as the individual interviews could inform their content. The focus group discussions will include participants from different backgrounds and clubs. The interviews and discussions are conducted in person at gymnastic centres or online (Zoom (Zoom Video Communications, San Jose, California, USA)), depending on the participants’ preferences. JBP conducts interviews in French and English. Interviews in Dutch are conducted by an athlete-partner with experience in qualitative research and whose native language is Dutch. To ensure consistency in data collection, the athlete-partner interviewer was trained based on previous interviews conducted in English. The themes that will emerge from this phase will help target the needs of the participants regarding the dependent variables evaluated in the prospective cohort study and will inform methodological choices for the quantitative phase.²¹

Participants

Athletes aged 12–19, competing at tiers 2, 3 or 4 of the participation classification framework irrespective of their specialisation²³; retired athletes aged between 16 and 30; full-time coaches, trainers, federation representatives and parents will be recruited from gymnastics organisations in Canada (Montréal area) and the Netherlands (Amsterdam area). Although athletes aged 12–15 cannot legally consent to research and will require parental consent, their inclusion will ensure comprehensive results, given that they are within the age range where sport specialisation occurs and where dropout in sports is at its highest.24,26

Recruitment

Participants will be recruited through gymnastics associations and federations. The research team will contact representatives, coaches and trainers to participate via the federation (n=2) or the gymnastics associations (n=8). Following maximum variance sampling, gymnastics organisations agreeing to relay the study information will invite athletes (current or retired), athletes’ parents, coaches and federation members. They will be invited to contact the lead researcher if they are interested in participating. Respondent-driven methods (eg, snowball sampling) will also enhance recruitment strategies. Follow-ups could be sent to specific participant profiles to ensure all stakeholders are represented. Inclusive and gender-neutral language will be used in posters, booklets and presentations to appeal to all athletes.

Analysis

The interviews and discussions in focus groups will be audio recorded, transcribed ad verbatim, translated to English if needed, anonymised and entered into ATLAS.ti. CSB and JBP will carry out the reflexive thematic analysis.^{27 28} The first step will be for the coders (CSB and JBP) to familiarise themselves with the data. This will initially be done by JBP, who will conduct most of the interviews and review the transcriptions. CSB and JBP will then actively read the transcriptions to detect initial meanings or patterns. The second step will be to generate initial codes in a data-driven manner. Once all data have been coded, the third step will be to sort the codes into themes. The fourth step will be to review these themes. Themes will be reviewed on the coded data level and within the entire data set, ensuring that the themes form a coherent pattern and adequately represent the data set. The fifth step will be to define and name themes. Subthemes will be identified and classified within themes. A short description of each theme will be presented to the entire research team for discussion. The final step will be to produce the report presenting the themes obtained from the data.

Particular attention will be given to reflexivity. CSB and JBP will discuss how their backgrounds and preconceived ideas might influence their analysis of the data and the construction of themes.²⁹ To limit this influence, themes will be presented to the research team for their reflection and input. An audit trail will also be developed to document data collection and analysis decisions. This auditing will be independent of funding agencies but not from investigators. A description of the coders’ background will be included in the manuscript presenting qualitative results.

Sample size

Semistructured interviews will be conducted until data saturation is reached, meaning no new topics or concepts emerge during interviews.¹⁷ When no new topics or concepts emerge, additional interviews will be conducted until confirmation that the data collection can stop, that is, when there is a repetition of topics and concepts.³⁰ We aim to recruit at least five participants from each group: athletes, retired athletes, coaches, parents and federation representatives. At least one participant from each country should be recruited per group. Based on previous studies, we expect to conduct approximately 20 semistructured interviews.31,33 Subsequently, all participants will be invited to join a focus group discussion. These discussions will include between 6 and 10 participants, including a facilitator.

Quantitative phase

Study settings

The quantitative data collection will occur over a full sports season, from September to June, for 40 weeks. A first evaluation for all dependent variables will be conducted at the start of the season (T0—September), a second in the middle (T1—February) and a third at the end (T2—June). The evaluations will take place at the gymnastics centres for the convenience of athletes. Before the physical evaluation, questionnaires will be answered online through a secured platform: RedCap (Canada) or LimeSurvey (the Netherlands). The physical evaluation will take place in blocks of 4 hours and will be supervised by a physiotherapist. Multiple evaluation stations, overlooked by coaches, will ensure that more athletes can be evaluated in the same time frame. We expect two blocks of evaluation per gymnastic club (four in Canada and four in the Netherlands) to evaluate 73 participants, approximately 18 athletes in each association.

Participants

We will recruit competitive athletes aged between 12 and 19 years¹¹ from different gymnastics organisations during the summer and at the start of the school year in Canada and the Netherlands. Athletes who practise multiple sports and gymnastics will be included. Participants with injuries that prevent them from participating in gymnastics will be excluded, considering that this may affect their performance in the physical evaluation and their psychological health at baseline.

Recruitment

The researchers will pursue recruitment within the same organisations involved in the qualitative phase. Investigators of the study will give a short presentation of the study to the collaborating gymnastics clubs. The gymnastics clubs will then relay the information to athletes and/or parents of athletes to recruit participants for the quantitative portion. Coaches will be invited to discuss the research project with the athletes and athletes’ parents. Posters will also be left in the different gymnastic organisations. In addition to primary recruitment for this phase, any athletes who participated in the qualitative phase will also be allowed to participate in the quantitative phase. Athletes interested in the study will be invited to contact the postdoctoral researcher in charge of recruitment.

Measurement

The independent variable, sport specialisation, will be described using the Exposure to Sport Scale (ESS).³⁴ It comprehensively evaluates current and historical training load and gathers information on sports exposure duration, frequency and intensity (practice, competition and play). It has received preliminary validation and draws from a previously validated assessment tool, the Developmental History of Athletes Questionnaire.³⁵ Sport specialisation will be defined in three categories: (1) athletes who do not specialise (ie, who practise more than one sport), (2) athletes who specialised in gymnastics early (ie, practice only gymnastics for at least 8 months per year since before the age of 13) and (3) athletes who specialised late (ie, practice only gymnastics for at least 8 months per year, starting at or after the age of 13).¹¹

All dependent variables have been selected based on the framework for exploring early specialisation in sports (figure 1).¹¹ All tools and measurement scales used have excellent psychometric properties.736,44

Dependent variables will be measured during the physical evaluations at the start (T0—September), middle (T1—February) and end of the sports season (T2—June). To account for athletes’ growth, each evaluation will include the collection of data on height, weight, sitting height and leg length. The skill acquisition tests will assess core strength, endurance, power and dynamic postural control. The time spent in the plank and bridge positions will measure core strength and endurance.⁷ Power and dynamic postural control will be measured using the Gymnastics Functional Measurement Tool (GFMT), which has a female and male version.^{7 36} The GFMT consists of 10 items, each worth 10 points, for a possible total score of 100. The female version of this tool has excellent inter-rater reliability (intraclass correlation coefficient (ICC)=0.98), excellent test–retest reliability (ICC=0.97) and a good to excellent association with the competition level of the gymnasts (r²=0.60 to r²=0.97).^{36 45} The male version of this tool has excellent test–retest reliability (ICC=0.97) and has a good association with the competition level of the gymnasts (r²=0.63).

Psychological health will be evaluated through stress, health-related aspects of personality, anxiety and quality of life. Stress will be measured using the Recovery and Stress Questionnaire for Sports (RESTQ-Sports-36).³⁷ The RESTQ-Sports-36 consists of 36 items, rated on a 7-point Likert scale. The average interitem correlation ranges from 0.21 to 0.60, proving its reliability.⁴⁶ Personality will be evaluated using the Athletic Identity Measurement Scale (AIMS)³⁸ and the Coping Inventory for Stressful Situations (CISS).³⁸ The AIMS consists of 7 items measured on a 7-point Likert scale and reported good internal consistency in adolescents.^{38 47} The CISS-21 consists of 21 items measured with a 5-point Likert scale.⁴⁸ Anxiety will be evaluated with the State-Trait Anxiety Inventory—Children (STAIC).³⁹ It consists of 40 items. The first 20 items, the state scale, identify different feelings rated on three levels, that is, I feel… (a) very cheerful, (b) cheerful, (c) not cheerful. The last 20 items, the trait scale, are statements rated as hardly ever, sometimes, or often. It has been shown that the STAIC can discriminate well between children and adolescents who have an anxiety disorder and those who do not.³⁹ Quality of life will be evaluated with the Paediatric Quality of Life Inventory for Children and Young Adults (PedsQL).^{41 42} The PedsQL assesses four aspects of functioning: physical, emotional, social and school. There is a total of 23 items rated on a 5-point Likert scale. It has been shown to have adequate internal consistency in youth (α=0.80–0.88).⁴¹

Weekly injury surveillance will be done using an updated version of the self-report Oslo Sports Trauma Research Centre Overuse Injury Questionnaire by parents in collaboration with the athletes (athletes 12–15 years old) or by athletes (16 years or older).⁴³ The weekly injury surveillance questionnaire will also monitor training hours and the occurrence of competitions. One physiotherapist in Canada and one in the Netherlands will evaluate time-loss injuries to clarify the injury’s context, mechanism and region. These evaluations will be conducted over the telephone and should last less than 20 minutes.

Analysis

The second objective of this study is to evaluate whether early specialisation impacts skill acquisition throughout a gymnastics season. Repeated measures analyses of variance (ANOVA) will be used to compare the T0, T1 and T2 skill acquisition variables between athletes of different specialisation backgrounds, as defined by the ESS: (1) non-specialisers, (2) early specialisers and (3) late specialisers. Analyses of covariance (ANCOVAs) will be conducted to account for sex, age and gymnastic association.

The third objective of this study is to investigate associations between early sports specialisation and (1) psychosocial factors and (2) injuries during a sports season. Two separate multiple linear regressions will be conducted to complete this objective. The first linear regression will explain individual psychosocial factors with early sport specialisation as an independent variable while controlling for age, gender, country of participation, number of gymnastics training hours and whether they were injured in the previous year. The second linear regression will aim to explain the overuse injury incidence throughout the season using early sport specialisation as an independent variable when controlling for age, gender, number of gymnastics training hours and whether they were injured in the previous year.

The fourth objective is to investigate the moderator effect of psychosocial factors on the association between injuries during the season and early sports specialisation. To complete this objective, an additional linear regression will be conducted where the following variables will be added as moderators within the model described previously: recovery and stress (RESTQ-36), athletic identity (AIMS), coping strategies (CISS) and anxiety (STAI).

Sample size

To detect an effect size f of 0.1 in a repeated measures ANOVA with within–between interaction, for an alpha of 0.05, power of 80% and accounting for a 15% attrition rate, 146 participants will be recruited. We aim to recruit 73 participants in Canada and the Netherlands. We aim to recruit a representative sample of the diversity of sex and gender to allow for stratified analyses considering sex and gender identity.⁷

Data management

Each participant will be assigned a random number ID. A password-protected master sheet linking the number ID to the participants’ information will be kept by one of the investigators in a protected encrypted file (VeraCrypt). This sheet will not be accessible to anyone outside the research team. All physical performance and injury data will be kept in a pseudonymised database, accessible only by research team investigators. Individual data will be shared with the athlete participants and their parents for those under 16 years old. Coaches may also have access to the individual data of their athletes’ physical performance (core strength and endurance, GFMT scores) if the participant signifies agreement on the consent form. This access to physical performance test results will be granted to optimise the athletes’ and coaches’ training choices. Coaches will not have access to injury data given by the athlete to the research team or information disclosed during interviews, focus group discussions or questionnaires.

Dissemination

To ensure maximal impact from the study results, our knowledge transfer strategy rests on three main sources: the scientific community, research institutes and athlete/sport partners. In the scientific community, we will submit abstracts to national (Canada and the Netherlands) and international conferences. Scientific articles will be published in high-impact sports medicine journals. The results will be disseminated using presentations and publications to reach researchers, practitioners and clinicians. We will also offer presentations to gymnastics organisations. To reach coaches, trainers, federation representatives, parents and athletes on a larger scale, results will be presented in a scientific outreach context in Canada and Europe (ie, conferences for coaches and staff members of gymnastics organisations and other sports associations that could be interested in the findings of this study). This will be done in collaboration with the researchers of our team, the research institutes, the sports organisation and the athlete-partners supporting this project. All abstracts and manuscripts will be written by authors who have collaborated with the creation of this project and who hold specific expertise relevant to the study.

Discussion

This study aims to present a study describing early sport specialisation in youth and adolescents participating in gymnastics and exploring the potential impacts of early sport specialisation on physical performance, psychological health and injuries. It is the first study to use mixed methods to investigate this problem.

This study has several strengths that foster innovation in the field of sport specialisation. The team comprises researchers with diverse backgrounds, approaches and expertise, which provides a strong foundation for this project. The involvement of athlete-partners and gymnastics organisations also ensures that this project and its results will be relevant to the gymnastics community. The project’s design is informed by several studies’ conclusions: (1) qualitative research is needed to give a voice to stakeholders seldom heard on early specialisation, (2) prospective cohort studies are needed since the vast majority of studies in early sport specialisation have been cross-sectional or retrospective and (3) in addition to skill acquisition and injuries, psychological variables should be investigated.^{10 11 15}

Strengths and limitations

The applied methodological strategies will strengthen the trustworthiness of our findings. To increase the credibility of our findings, we will use triangulation of sources by including different groups (athletes, retired athletes, coaches, trainers, federation representatives and parents). We will have independent coders and ‘critical friends’ to improve this study’s confirmability and dependability.⁴⁹

Following the athletes for one sport season might appear short to study the mid-term and long-term impacts of early sport specialisation on skill acquisition, psychological health and injury incidence. Since very few studies have prospectively investigated this phenomenon in gymnasts, identifying short-term impacts will inform future longitudinal studies. Moreover, sample size calculations were adjusted to account for the modest changes in skill acquisition and psychological health that are expected over a single season.

The exclusion of injured athletes from the quantitative phase of the project could introduce a collider bias. However, this exclusion is necessary as injuries could affect the results of the physical performance variables (core strength, endurance, power and dynamic postural control).

A recall bias and social desirability bias could be introduced in the qualitative and quantitative phases of the study. To limit the impact of social desirability bias in the qualitative phase, participants will be informed at the beginning of interviews that their answers will remain confidential. To limit the impact of recall bias during the quantitative phase, the short questionnaire on injury and training hours will be administered weekly, limiting the recall period.

Coaches participating in the physical evaluation will be trained with the GFMT to minimise evaluator bias. Moreover, physiotherapists evaluating time-loss injuries will follow a standardised questionnaire and be in frequent contact.

Potential benefits

As a first step towards a better understanding of early sports specialisation and its effects on gymnasts, this study will shed light on the experiences of athletes and their entourage with gymnastics training and its associated challenges. This study will also explore the consequences of early sports specialisation on physical and psychological variables and injuries in a prospective setting using reliable and valid assessment tools. These findings will advance scientific knowledge and inform policies promoting inclusive and sustainable athletic development for gymnasts.

Data availability statement

No data are available.