Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Understanding risk factors for disordered eating symptomatology in athletes: A prospective study

Abstract

Disordered eating and eating disorders have huge impact on athletic health and performance. Understanding risk factors for disordered eating development is paramount to protecting the health and performance of these athletes. This project tested a model longitudinally to test whether body dissatisfaction (mediated by negative affect) and societal pressures (mediated by internalisation) predicted bulimic symptomatology at 1 year. The study recruited 1017 male and female athletes in a range of sports at three time points over a year. Cross-lag meditation modelling in MPLUS was utilised to test the hypothesised model. Results indicated that societal pressures mediated by general internalisation led to bulimic symptomatology and that gender and sport type do moderate the relationships. However, measurement issues indicate that scales not originally created for athletes may not reliably measure athletes’ experience. This research highlights how understanding how to better assess risk factors and disordered eating related concepts in athletes is a key next step. The study is unique in its longitudinal design and in its sampling of a wide range of sports in both male and female athletes.

Citation: Stoyel H, Stride C, Shanmuganathan-Felton V, Serpell L (2021) Understanding risk factors for disordered eating symptomatology in athletes: A prospective study. PLoS ONE 16(9): e0257577. https://doi.org/10.1371/journal.pone.0257577

Editor: Chiara Milanese, University of Verona, ITALY

Received: February 5, 2021; Accepted: September 6, 2021; Published: September 24, 2021

Copyright: © 2021 Stoyel et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the manuscript and its Supporting Information files.

Funding: No funding was provided for the study from any grants or organizations.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Eating disorders, including Anorexia Nervosa, Bulimia Nervosa and Binge Eating Disorder, are psychiatric illnesses that have significant negative impact on physical and mental well-being [1–3]. Eating disorders sit at the end of a continuum that also includes healthy or intuitive eating at the other end, and disordered eating in the middle [4, 5]. Disordered eating, defined as a subclinical level of issues with food restriction, bingeing and purging behaviours [6] has a much higher prevalence rate than clinically diagnosable eating disorders in both the general and sporting population [7–9].

Exercise is often a symptom or maintenance factor for an eating disorder, however, it is also an integral part of sport [10, 11], which means it is difficult to distinguish pathological from ‘normal’ exercise in athletes. Furthermore, the link between body shape and physical exercise is undeniable, making sport a complex feeding ground for disordered eating development [12–14]. Biomechanical differences between sports demand specific body shapes to increase the chance for successful performance, such as broad shoulders for rowing and minimal body mass for distance running. As such, it is often expected that athletes will have a higher prevalence rate of disordered eating than nonathletes, though rates appear to vary dependent on gender and sport type [15–18]. Athletes who participants in lean sports, those in which success depends on a lean body shape, are at a greater risk [6, 19–25]. Gender plays a moderating role, with prevalence higher for males in anti-gravitational sports such as cycling, and for females in lean or aesthetic sports such as dance [12, 26, 27].

Eating disorders and disordered eating can have long term impact on both health and on quality of life, and as such, understanding how to predict the development of eating disorders and disordered eating to facilitate prevention is paramount [9, 28, 29]. In 2007, Petrie and Greenleaf—based on Stice’s [1994] dual pathway model of predicting bulimia—developed a detailed model that aims to predict disordered eating in athletes [30]. In Petrie and Greenleaf’s model there are eight risk factors discussed: Sport pressures and societal pressures are the predictors, internalisation, body dissatisfaction, negative affect, modelled behaviours, which are thought to mediate the pathway to disordered eating symptomatology, which is deconstructed into restrained eating and binge eating and bulimia.

Stoyel and colleagues conducted a systematic review of the literature testing tenets of Petrie and Greenleaf’s model, which indicated broad support for many of the relationships posited in the model [18]. However, contradictory findings between studies were due to the range of different designs, samples and measurement tools. Only three were of longitudinal design meaning that it was difficult to investigate causation. In general, there has been a lack of longitudinal research in this topic area.

Based on this review, Stoyel and colleagues tested Petrie and Greenleaf’s model using structural equation modelling on a large dataset of male and female athletes from multiple sports [31]. Petrie and Greenleaf’s model offered an inadequate fit to the data: using theoretically-driven adjustments, a new, more parsimonious model was developed [Fig 1], which offered a satisfactory fit. This revised model, hereafter known as the T1 model, is similar to Stice’s 1994 dual pathway model; body dissatisfaction and social pressures predict bulimic symptomatology in athletes, with these effects operating indirectly via the twin mediators of negative affect and internalisation respectively [31].

thumbnail
Download:
Fig 1. T1 model.

This model has been adapted slightly from its original published version. Social pressures is now termed Societal Pressures.

https://doi.org/10.1371/journal.pone.0257577.g001

While the T1 model provided some insight into how disordered eating may develop in athletes, it employed a cross-sectional design. Therefore, the aim of the current study was to test the utility and applicability of the T1 model employing a longitudinal design–by collecting data on the same cohort at two further time points, six months apart, enabling the testing of a cross-lagged type mediation model [32]. The primary hypothesis is that the model devised at T1 will remain of good fit across time, when extended to this cross-lagged form as illustrated in Fig 2 below. The second hypothesis is that each path will constitute a significant effect, with Body Dissatisfaction at time 1 having a positive and significant effect on Bulimia at Time 3, mediated by Negative Affect at Time 2; and Social Pressures at Time 1 having a positive and significant relationship with Time 3 Bulimia mediated by Internalisation at Time 2. Finally, as gender and type of sport play a moderating role in the susceptibility of an athlete to developing disordered eating, it is pertinent to explore these factors [33, 34]. As such it is also hypothesised that the predictor to mediator relationships in the model will be significantly moderated by both gender and lean/nonlean sport type.

thumbnail
Download:
Fig 2. Cross-lag mediation model.

https://doi.org/10.1371/journal.pone.0257577.g002

Methods

Procedure

Data collection took place at three time points over the course of a year. Time point one (T1) as described by Stoyel and colleagues was a three-week period in February 2019, time point two (T2) was a three-week period in October 2019, and time point three (T3) was a three-week period in February 2020 [31]. The study received ethical approval from the Clinical Educational and Health Psychology Department at University College London, reference for this approval: CEHP/2018/573.

All data was collected online, using Opinio (opinion.ucl.ac.uk) for T1 and Qualtrics (www.uclpsych.eu.qualtrics.com) for T2 and T3. This research engaged a volunteer sample via social media and the author’s connections as a sport psychologist at local sporting clubs. The inclusion criteria, set at T1, stated that participants had to be over the age 18; had to consider themselves to be an athlete (determined with a simple yes/no answer to the question “Do you identify as an athlete?”); had to be training for a minimum of ten hours a week; and had to be actively competing. These criteria were set such that those included in the study were athletes, rather than just regular exercisers.

At T1 1017 participants responded (and their data were used in the cross-sectional study described in citation [31]). Participants only completed the questionnaires if they had provided written consent via a consent form which they completed online. T1 participants who had given consent for follow-up contact were invited via email to participate at T2. At T2, 879 responses were collected. Those who responded at T2 were then asked to participate again at T3, at which 744 responses were collected. From T1 to T3 there was a 26.8% attrition rate. At each time point, participants gave informed consent, and received a £5 voucher for completing the questionnaire. All those participants who completed the questionnaire too quickly based on analysis of those in the outer quartiles of a normal distribution curve (under eight minutes at T1 and under six minutes at T2 and T3 as the second and third timepoint questionnaires were shorter) were removed, giving a final analysis sample of 802 observations, of whom 802 responded at T1, 551 at T2 and 469 at T3.

Participants

This research engaged a volunteer sample via social media and the author’s connections as a sport psychologist at local sporting clubs. The inclusion criteria at T1 were that participants had to be over the age of 18; had to consider themselves to be an athlete (determined with a simple yes/no answer to the question “Do you identify as an athlete?”); had to be training for a minimum of ten hours a week; and had to be actively competing. These criteria were set such that those included in the study were athletes, rather than just regular exercisers.

The sample was made up of 54.9% males and 45.1% females. The majority were aged between 18 and 26 (84.9%) with 15.1% of the sample aged 27 years or older at T1. Other sample characteristics are given in Table 1. At T1, main sports were basketball 19.6%, swimming 17.6% distance running 11.2%, football 11.8%, dancing 8.3%, tennis 7.5%, volleyball 8.5%, 7.6% other track and field event, with other sports represented with smaller percentages including hockey, badminton, lacrosse, cricket, cycling, golf, triathlon, rugby, boxing, and rowing. Body Mass Index (BMI) and EDE-Q global scores (used for diagnostic purposes of eating disorders) are also presented in Table 1. The mean EDE-Q global scores and is considered healthy and not indicative of an eating disorder. The range of BMI scores is also presented, with the percentages for those that fall outside the healthy range (below 18.5 and above 25) also included. While, the extreme BMI scores do indicate an unhealthy weight to height ratio for some participants and many fall outside the set indicators for healthy BMI, it is important to note that BMI is an insensitive and inaccurate indicator for those with large muscle mass as is likely the case with a sample of athletes [35].

thumbnail
Download:
Table 1. Descriptive statistics and clinical information.

https://doi.org/10.1371/journal.pone.0257577.t001

Measures

To capture both social pressures and internalisation the third edition of the sociocultural attitudes towards appearance was utilised (SATAQ; [36]). The SATAQ internalisation dimension has both general and athlete subscales and has exhibited good measurement properties in previous studies [37]. Likewise, the SATAQ subscales of Pressures and Information [along with the social media questions] were used to measure the risk factor Social Pressures in Petrie and Greenleaf’s model, alongside additional questions that the research team developed in an attempt to modernise the scale by capturing social media pressures. These new items matched the typical wording of the SATAQ [36]. For example, “Social media is an important source of information about fashion and ‘being attractive.’” Specifically, the SATAQ subscales of Pressures and Information (along with the social media questions) were used to measure the risk factor Societal Pressures in Petrie and Greenleaf’s model with both the General and Athlete Internalisation subscales of the SATAQ used to measure Internalisation. A full listing of items is given in Table 3. Response coding for these items were Definitely Disagree, Mostly Disagree, Neither Agree nor Disagree, Mostly Agree and Definitely Agree.

The ten negative items in the Positive and Negative Affect Schedule (PANAS) were used to measure Negative Affect [38, 39]. A full listing of items is given in Table 3. Response coding asked about emotions and feelings and the extent to which they were felt was from Very slightly or not at all (1) to Extremely (5) over the past week.

To capture the experiences of body dissatisfaction and bulimia related symptomatology, the 9-item Eating Disorder Inventory Body Dissatisfaction (EDI-BD) and the 7-item Bulimia (EDI-B) subscales were utilised respectively [40, 41]. A full listing of items is given in Table 3. Response coding for both the subscales was Never, Rarely, Sometimes, Often, Usually, Always. Scoring was such that numerical scores of 000123 were applied respectively (and the reverse for reversed scored).

The across-time reliabilities of all scales can be found in Table 2 and the items for each scale can be found in Table 3.

thumbnail
Download:
Table 2. Across time reliabilities.

https://doi.org/10.1371/journal.pone.0257577.t002

thumbnail
Download:
Table 3.

https://doi.org/10.1371/journal.pone.0257577.t003

Data preparation

The data was cleaned and prepared in IBM SPSS (version 25) before formal analysis. As the first step in data preparation, across-time reliabilities were calculated (see Table 2). The reliabilities for all the scales except for Bulimia (EDI-B) and Negative Affect (PANAS Negative Items) were below an acceptable standard for use in analysis. Therefore, two steps were taken. Firstly, all those participants who completed the questionnaire too quickly based on analysis of those in the outer quartiles of a normal distribution curve were removed giving a final analysis sample of 802 observations, of whom 802 responded at T1, 551 at T2 and 469 at T3. ‘Too quickly’ was defined as under eight minutes at T1 and under six minutes at T2 and T3 (as the second and third timepoint questionnaires were shorter).

Secondly, reverse-scored items were eliminated. As seen in Table 2, this removal did improve the reliabilities of the scales to be utilised in analysis. Details of each scale, those items completed by participants and then subsequently used in analysis can be found in the item map (see Table 3).

Data analysis

Firstly, confirmatory factor analyses (CFA) to examine the structural validity, the temporal (across time) invariance, and the multigroup invariance (between genders and lean/non-lean sports) of the proposed measurement model and the scales within it. Given the multiple time points and large number of items within each scale, a global CFA of all scales at once would have resulted in an unsatisfactorily low case-free parameter ratio, hence the Social Pressure, Negative Affect, Bulimia, Body Dissatisfaction and Internalisation measures were each considered separately. Having made any necessary adjustments, reliability analyses were then conducted (Cronbach’s alpha coefficient) to check the internal consistency of the proposed scales.

Secondly, having calculated mean scale (i.e. composite) scores for each construct measured, a series of path analysis models was then fitted to test the hypothesised cross-lagged mediation model, and its moderation by gender and lean/nonlean sport (Fig 2). When fitting the cross-lagged mediation model, analysis began with a model in which paths were free to differ across time. The stability of paths across time (stationarity) was then tested by fixing equal across time, in sequence: T1 to T2 and T2 to T3 autoregressive paths between the same variables across time; T1 to T2 and T2 to T3 paths from the predictors (social pressures and body dissatisfaction) to the mediators (Internalisation and Negative Affect); and T1 to T2 and T2 to T3 paths from the mediators to the outcome (Bulimia). The aim was to show that these fixings did not significantly depreciate model fit, i.e. there was no evidence of variation in these relationships across time. The indirect effects from societal pressures and body dissatisfaction to the outcome of bulimia were then calculated to test whether mediation via internalisation and negative affect respectively was occurring as hypothesised [42].

Finally, the moderation hypotheses were tested by adding the interaction effects of both gender and lean/non-lean sport with societal pressures on mediator internalisation, and with body dissatisfaction on mediator negative affect.

Mplus was used to fit and test between models, with Full Information Maximum likelihood estimation employed on all cases [43]. A p level of < 0.005 used throughout.

Item Map (Table 3)

Results

Table 4 gives the results of the CFA for each of the five constructs of the model, showing both the fit of proposed number of factors with no restrictions in loadings and intercept parameters across time (configural temporal invariance), and the comparisons with models in which equivalent loadings and then intercepts were fixed equal across time (metric temporal and scalar temporal invariance). To achieve a satisfactory fit for the configural invariance model, a small number of items were dropped (detailed in Table 3 above), almost all negatively worded items. Table 5 gives the equivalent tests when considering the multigroup invariance of these respective measurement models between genders and between participants of lean and non-lean sports.

thumbnail
Download:
Table 4. Confirmatory factor analyses.

https://doi.org/10.1371/journal.pone.0257577.t004

thumbnail
Download:
Table 5. Group Invariance for gender and lean/nonlean.

https://doi.org/10.1371/journal.pone.0257577.t005

For each of Negative Affect, Body Dissatisfaction, and Bulimia the proposed 1 factor models provided an adequate fit to the data, which was not significantly compromised by fixing factor loadings equal across time. This suggest that the understanding of the items as measures of the respective concepts was consistent across time. Comparing the configural and metric invariance for Negative Affect shows a chi-square difference from 571.00 to 594.18 and no change in the degrees of freedom which was not a significant change with p = 0.18. Scalar temporal invariance, how item responses were understood across time, was not achieved for Bulimia and Body Dissatisfaction–however the fit of the Body Dissatisfaction with scalar invariance, though significantly worse than the metric invariance model, was still satisfactory with the CFI greater than 0.90 for metric and configural models (see Table 4).

For internalisation, a two-factor model of general and athlete dimensions offered a satisfactory fit with a CFI of 0.91 for both configural models and outperformed a potentially competing factor model. When applying metric invariance, the fit of the two-factor model was not significantly reduced with the CFI remaining at 0.91, but scalar invariance was not achieved. Similarly, the three-factor model for Social Pressures achieved a good level of fit 0.91, outperformed a one factor model, and achieved metric invariance (see Table 4).

For each of the above measures, the satisfactory metric invariance models were taken forward to run a series of multigroup CFAs in which loadings and then intercepts were allowed to vary between male/female athletes, or between lean/non-lean sport athletes. Results of invariance tests are given in Table 5. The Social Pressures, Internalisation, Body Dissatisfaction and Negative Affect measurement models all exhibited both metric and scalar invariance between groups for both gender and lean/non-lean sport. Bulimia failed to achieve metric invariance between genders, though the fit of the metric invariance model, whilst significantly weaker than the configural invariance model, was itself still satisfactory.

Structural equation modelling

With evidence for metric invariance and good fit for most scales, the CFAs were subsequently extended to a structural equation model (SEM) also within MPlus. The next step was to test the hypothesised path model, illustrated in Fig 2.

This baseline model with all paths free, model 1, and all of the subsequent models failed to achieve a good fit (see Table 6).

thumbnail
Download:
Table 6. Path analysis model comparisons, testing stationarity of hypothesised model.

https://doi.org/10.1371/journal.pone.0257577.t006

When adding constraints across time to investigate the stationarity of effects, neither fixing the autoregressive pathways (blue paths, Fig 2) nor the predictor to mediator paths (black and light green paths, Fig 2) resulted in a significantly weaker fit. However, fixing the mediators to dependent variable paths (purple and dark green) equal across time significantly reduced the model fit.

Within the final, most constrained model (model 4), there were several significant pathways despite the model lacking an acceptable fit. The pathway from all types of societal pressures to general internalisation was significant. The pathways from internalisation to bulimia were also significant (see Table 7 and Fig 3).

thumbnail
Download:
Fig 3. Visual of indirect, mediated, and direct effects.

https://doi.org/10.1371/journal.pone.0257577.g003

thumbnail
Download:
Table 7. Significant pathways in model 4.

https://doi.org/10.1371/journal.pone.0257577.t007

While the most constrained model, model 4, failed to establish a good fit, it did show that, with all paths set equal over time, there is a significant indirect effect of Societal Pressures (operationalised as SATAQ pressures, SATAQ information, and social media questions) on Bulimia, operating via General Internalisation (see Table 8).

thumbnail
Download:
Table 8. Indirect and direct effects (model 4).

https://doi.org/10.1371/journal.pone.0257577.t008

Several direct effects were also significant across time also shown in Table 8 and Fig 3. Furthermore, Table 9 shows what parts of the model account for percent change in Bulimia at T3.

thumbnail
Download:
Table 9.

https://doi.org/10.1371/journal.pone.0257577.t009

Moderation analyses

Group invariance for lean vs nonlean and gender were tested by taking forward the metric temporal invariance CFA for each scale to assess whether the questions (via metric invariance) and response code (via scalar invariance) are understood the same way across time for all sports/genders (see Table 8). Using chi square of difference testing it was found that all scales had metric and scalar invariance for lean vs nonlean and gender apart from gender invariance for the Bulimia scale.

Finally, to test the final hypothesis and to further extend the model, moderation by gender and lean/nonlean sport was analysed. Moderation of the path from predictors to mediators (black and light green paths in Figs 2 and 3) by gender and lean/nonlean sport showed both gender and lean or nonlean sport consistently moderates the path from Social Pressures (all types) to General Internalisation and from Body Dissatisfaction to Negative Affect. The slope effect of Social Pressures and Body Dissatisfaction is less positive for those in lean sports and for men.

Discussion

The main aim of this study was to test that T1 model of disordered eating in athletes [31] longitudinally using structural equation modelling in the form of a cross-lagged mediation model.

It was hypothesised that the T1 model would show parsimonious goodness of fit across time in the form of a cross-lag mediation model, but this primary hypothesis failed to be supported, with the model failing to achieve a good overall fit. Specifically, it was hypothesised that Body Dissatisfaction at T1 would have a positive and significant relationship with Bulimia at T3 as mediated by Negative Affect at T2, however, this was not supported. It was also predicted that Societal Pressures at T1 would have a positive, significant relationship with Bulimia at T3, mediated by Internalisation at T2. Part of this hypothesis was supported in that Societal Pressures—in the forms of pressures, information, and social media—significantly predicted bulimic symptomatology a year later, mediated by general internalisation. This suggests that athletes are not only exposed to societal pressures in the form of overt pressure, and information by the society at large, in both mass media and social media, but that it is the internalisation or incorporation of these messages into one’s self-worth that predicts bulimic symptomatology. This finding is consistent with results that show that internalisation of the thin ideal, promoted by society and media, in a clinical or nonathlete population is linked to the development of eating disorders [44, 45]. The non-significance of the body dissatisfaction pathway was unexpected. It is possible that the use of general, rather than athlete-specific, body dissatisfaction scales explains this result. Scales that examine general body dissatisfaction may not capture how athletes specifically feel about their body weight and shape. An athlete may feel that their body is satisfactory for societal beauty standards, but not for sport performance and it is possible that the measurement tools in this study did not capture that distinction. Further study of the impact of body dissatisfaction across time is warranted as previous work has been almost exclusively cross sectional (e.g.[23]).

It was hypothesised that the relationships in the model would be significantly moderated by gender and lean/nonlean sport type. This hypothesis was supported with gender and lean/non sport participation both significant moderators, with both gender and lean/nonlean sport participation moderating the pathway from the predictors of societal pressure and body dissatisfaction to the respective T2 mediators of internalisation and negative affect. Therefore, when considering the susceptibility of athletes to developing disordered eating, the type of sport and the gender of that athlete must be taken into account, as the influence of the predictors of social pressure and body dissatisfaction was stronger for female athletes and, surprisingly. for those in nonlean sport. These findings are consistent with previous work that female athletes are more susceptible to the risk factors and correlates of disordered eating as well as disordered eating itself [46, 47]. However, previous work has overwhelmingly, but not always, found that lean sport participants to be more at risk than nonlean (see for example Milligan and Pritchard 2006 versus Kong and Harris 2015 and Torstveit et al 2008). Despite comparisons between lean and nonlean sport, nonlean sport participants are still susceptible to developing disordered eating [21]. The findings in this study perhaps suggest an issue with the classifications of lean and nonlean sports, which was done using the first author’s own expertise as an applied Sport Psychologist, but there is no gold-standard way of determining which sports are lean and which are nonlean [22].

Limitations

Despite its strengths, the current project had several limitations, the first being the poor reliabilities of measurement tools across time. Following the systematic review of Stoyel and colleagues [18], scales with high reliability and frequent usage in the clinical population were chosen in an effort to create consistency. Thus, the first step in exploring why this model did not fit across time was to examine the reliabilities of the scales in this population. Previously, in published clinical, nonathlete samples all scales had reliabilities with α > 0.8 [38, 40, 48–50]. However, in the current study, the across-time reliabilities in an athlete sample were poor and worsened across time. Preparation of the data set was needed to increase the reliabilities so that reliable analysis was possible. Interestingly, due to drop out, the make-up of the sample became increasingly elite-heavy across time, perhaps indicating that the scales are less reliable with higher-level competitive athlete samples. Furthermore, confirmatory factor analyses revealed issues with metric and scalar invariance, that while not so ill-fitting that structural equation modelling was invalid, this does show that gold standard measures from non-athlete studies do not perform as well within this athlete sample.

Further limitations include the fact that recruitment was done via social media and participants were rewarded with a voucher. These decisions may have encouraged careless or dishonest responding and meant some items and participants had to be removed from analysis. Additionally, increasing the number of time points either within the year period, or extending the research for a longer period would have allowed for even more sophisticated latent growth curve analysis, as ideally four or more time points are needed for nonlinear analysis [51].

Implications

This research can inform applied practice, by suggesting that societal pressures and their internalisation could be the focus of prevention interventions for athletes. These findings suggest that mitigating the harmful effects of societal pressures in the form of both information and more overt pressure from an athlete’s social sphere and social media is key. While it is not possible to impede all incoming messages by society, a focus for sport psychologists, clinical psychologists, coaches and athletes should be to prevent such messages being internalised into one’s own self-worth. With societal pressure playing a central role in the development of disordered eating, the current study suggests that all types of athletes and both genders are at risk [52–54]. Thus, screening for disordered eating should include all athletes, not just females or those in lean/aesthetic sports. Furthermore, focussing efforts on creating a social environment for all types of athletes that is supportive for wellbeing may be an important aspect of prevention efforts [55, 56]. This finding reiterates that athletes do not exist in a sporting bubble that goes untouched by society’s expectations, and furthermore it suggests that interventions developed and tested in nonathlete samples may be able to be adapted for athletes [57]. When working with athletes to mitigate pressures related to disordered eating development we must be aware of those pressures outside of and tangential to the sporting realm.

Future research should focus efforts on developing valid and reliable questionnaires to operationalise disordered eating, eating disorders, and related concepts specifically for athletes. In sum, how we operationalise, measure, and therefore diagnose disordered eating and eating disorders in athletes needs to be tailored to them specifically [58]. It is not necessarily that athletes develop disordered eating symptomatology due to their participation in sport, but the ways we measure and convey disordered eating and related concepts and subsequently intervene may need to be adjusted for athletes. The current scales have items that are not specific for athletes and may therefore under or overestimate disordered eating. For instance, asking an athlete an item such as “Have you had a definite fear that you might gain weight?” as in the EDE-Q, for an athlete who is headed into a lightweight rowing race or wrestling match the following week where there are weight requirements in order to enter the competition, that fear may not be based in disordered eating cognitions but would present as a symptomatic response. In short, there is power in words, and the wording must be appropriate to the context.

References

  1. 1. Association AP. American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders: Diagnostic and Statistical Manual of Mental Disorders. Fifth Addi. Arlington, VA, US; 2013. pmid:22809622
  2. 2. Hudson JI, Hiripi E, Pope HG Jr, Kessler RC. The prevalence and correlates of eating disorders in the National Comorbidity Survey Replication. Biol Psychiatry. 2007;61(3):348–58. pmid:16815322
  3. 3. Klump KL, Bulik CM, Kaye WH, Treasure J, Tyson E. Academy for eating disorders position paper: eating disorders are serious mental illnesses. Int J Eat Disord. 2009;42(2):97–103. pmid:18951455
  4. 4. Tylka TL, Subich LM. Exploring the construct validity of the eating disorder continuum. J Couns Psychol. 1999;46(2):268.
  5. 5. Tylka TL, Wilcox JA. Are intuitive eating and eating disorder symptomatology opposite poles of the same construct? J Couns Psychol. 2006;53(4):474.
  6. 6. Krentz EM, Warschburger P. Sports-related correlates of disordered eating in aesthetic sports. Baron Cohen, De Bruin , De Bruin , Doninger , Dosil , Faul , Garner , Garner , Harris , Hausenblas , Hausenblas , Helfert , Helfert , Hinton , Keel , Kerr , Lane , Loland , McCreary , Meermann , Muscat , Okano , Paul , Petrie , Petrie , Petrie , Ricciardelli , Russell , Salbach-An C, editor. Psychol Sport Exerc [Internet]. 2011;12(4):375–82. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=psyc8&NEWS=N&AN=2011-11343-006
  7. 7. Greenleaf C, Petrie TA, Carter J, Reel JJ. Female collegiate athletes: Prevalence of eating disorders and disordered eating behaviors. Barker Blaydon, Brelsford , Cafri , Carter , Cohen , Drewnowski , Etzel , Greenleaf , Harris , Hausenblas , Hausenblas , Hausenblas , Hausenblas , Hausenblas , Johnson , Johnson , Krane , Lester , Lester , Mintz , Montgomery , Nelson , Pasman , Pernick , Petrie , Petrie , Petrie B, editor. J Am Coll Heal [Internet]. 2009;57(5):489–95. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=psyc6&NEWS=N&AN=2009-04011-002 pmid:19254889
  8. 8. Croll J, Neumark-Sztainer D, Story M, Ireland M. Prevalence and risk and protective factors related to disordered eating behaviors among adolescents: relationship to gender and ethnicity. J Adolesc Heal. 2002;31(2):166–75.
  9. 9. Smink FRE, Van Hoeken D, Hoek HW. Epidemiology of eating disorders: incidence, prevalence and mortality rates. Curr Psychiatry Rep. 2012;14(4):406–14. pmid:22644309
  10. 10. Goodwin H, Haycraft E, Meyer C. Disordered eating, compulsive exercise, and sport participation in a UK adolescent sample. Abbott Anderson, Brewerton , Cohen , Cook , Cumming , Davis , Eisenmann , Epling , Foundation , Garner , Gillison , Godin , Goodwin , Hausenblas , Hausenblas , Haycraft , Hubbard , Kerr , Kirk , Martinsen , McCabe , McCarthy , Meyer , Patel , Penas-Lledo , Petrie , Seigel , Stice A, editor. Eur Eat Disord Rev [Internet]. 2016;24(4):304–9. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=psyc13a&NEWS=N&AN=2016-09539-001 pmid:26892196
  11. 11. Mond JM, Hay PJ, Rodgers B, Owen C, Beumont PJ V. Relationships between exercise behaviour, eating‐disordered behaviour and quality of life in a community sample of women: when is exercise ‘excessive’? Eur Eat Disord Rev Prof J Eat Disord Assoc. 2004;12(4):265–72.
  12. 12. Bratland-Sanda S, Sundgot-Borgen Solfrid; ORCID: http://orcid.org/0000-0002-4202-5439 JAI-O http://orcid.org/Bratlan.-S. Eating disorders in athletes: Overview of prevalence, risk factors and recommendations for prevention and treatment. Ackland Baum, Beals, Bennell, Biesecker, Bonci, Bratland-Sanda, Byrne, Cafri, Currie, De Souza, Domine, Drinkwater, Fairburn, Forsberg, Garthe, Greenleaf, Gruber, Hulley, Johnson, Kanayama, Lock, Martinsen, Mazzeo, Murray, Nattiv, Nichols, Nichols, Perni B, editor. Eur J Sport Sci [Internet]. 2013;13(5):499–508. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=psyc10&NEWS=N&AN=2013-33810-010 pmid:24050467
  13. 13. Byrne S, McLean N. Eating disorders in athletes: a review of the literature. J Sci Med Sport. 2001;4(2):145–59. pmid:11548914
  14. 14. Cooper H, Winter S. Exploring the conceptualization and persistence of disordered eating in retired swimmers. Anderson Arthur-Cameselle, Arthur-Cameselle , Arthur-Cameselle , Baer , Beals , Brocki , Busanich , Carradice , Carter , Creswell , Fade , Fairburn , Fairburn , Gapin , Greenleaf , Grogan , Haase , Janghorban , Jennings , Jones , Krentz , Larkin , Lincoln , Luce , Luce , Mond A, editor. J Clin Sport Psychol [Internet]. 2017;11(3):222–39. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=psyc14&NEWS=N&AN=2017-56608-004
  15. 15. Anderson CM, Petrie TA, Neumann CS. Psychosocial correlates of bulimic symptoms among NCAA division-I female collegiate gymnasts and swimmers/divers. J Sport Exerc Psychol. 2011;33(4):483–505. pmid:21808076
  16. 16. Bachner-Melman R, Zohar AH, Ebstein RP, Elizur Y, Constantini N. How anorexic-like are the symptom and personality profiles of aesthetic athletes?. Med Sci Sports Exerc [Internet]. 2006;38(4):628–36. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=med5&NEWS=N&AN=16679976 pmid:16679976
  17. 17. Gapin JI, Kearns B. Assessing prevalence of eating disorders and eating disorder symptoms among lightweight and open weight collegiate rowers. Brownell Black, Carter , Cattanach , Faul , Foreyt , Fulkerson , Garner , Garner , Garner , Greenleaf , Greenleaf , Hausenblas , Hausenblas , Hausenblas , Karlson , Mintz , Montenegro , Petrie , Petrie , Petrie , Powers , Reinking , Saarni , Sanford-Martens , Spillane , Stice B, editor. J Clin Sport Psychol [Internet]. 2013;7(3):198–214. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=psyc10&NEWS=N&AN=2013-31664-003
  18. 18. Stoyel H, Slee A, Meyer C, Serpell L. Systematic review of risk factors for eating psychopathology in athletes: A critique of an etiological model. Eur Eat Disord Rev. 2020; pmid:31793151
  19. 19. Byrne S, McLean N. Elite athletes: effects of the pressure to be thin. J Sci Med Sport [Internet]. 2002;5[2]:80–94. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=med4&NEWS=N&AN=12188089 pmid:12188089
  20. 20. Kong P, Harris LM. The sporting body: Body image and eating disorder symptomatology among female athletes from leanness focused and nonleanness focused sports. Bartholomew Ackard, Bratland-Sanda , Byrne , Combs , Currie , de Bruin , de Bruin , Eisenberg , Feldman , Field , Findlay , Francisco , Fulkerson , Garner , Garner , Garner , Goldschmidt , Greydanus , Hargreaves , Harper , Holm-Denoma , Homan , Hudson , Levine , Lim , Lukacs M B, editor. J Psychol Interdiscip Appl [Internet]. 2015;149(2):141–60. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=psyc12&NEWS=N&AN=2014-56235-002 pmid:25511202
  21. 21. Milligan B, Pritchard M. The Relationship between Gender, Type of Sport, Body Dissatisfaction, Self Esteem and Disordered Eating Behaviors in Division I Athletes. Abell Bruch, Cash, Cooper, Crowther, Davis, Dick, Feingold, Fernanadez-Aranda, Friestad, Garner, Garner, Gleason, Green, Johnson, Kirk, Krane, Lindeman, Lucas, Mintz, Muth, Neumark-Sztainer, Patel, Perry, Petrie, Picard, Pope, Raudenbush, Rosen, Rosenber B, editor. Athl Insight Online J Sport Psychol [Internet]. 2006;8(1):1–15. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=psyc5&NEWS=N&AN=2006-12900-002
  22. 22. Reinking MF, Alexander LE. Prevalence of Disordered-Eating Behaviors in Undergraduate Female Collegiate Athletes and Nonathletes. Ashley Blair, Davis, Drinkwater, Evers, Garner, Garner, Garner, Johnson, Karlson, Loucks, Marten DiBartolo, Otis, Pate, Putukian, Putukian, Rosen, Sanborn, Smolak, Sundgot-Borgen, Warren, West, Yeager, Ziegler B, editor. J Athl Train [Internet]. 2005;40(1):47–51. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=psyc4&NEWS=N&AN=2005-03889-001 pmid:15902324
  23. 23. Rousselet M, Guerineau B, Paruit MC, Guinot M, Lise S, Destrube B, et al. Disordered eating in French high-level athletes: Association with type of sport, doping behavior, and psychological features. Arroyo Beals, Bratland-Sanda, Brisseau-Gimenez, Byrne, Coelho, Espelage, Ferrand, Forsberg, Garner, Gomes, Maimoun, Martinsen, Martinsen, Petrie, Plateau, Pope, Reel, Reinking, Rosendahl, Rouveix, Shanmugam, Shanmugam, Smolak, Sundgot-Borgen, Sundgot-Bor B, editor. Eat Weight Disord [Internet]. 2017;22(1):61–8. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=psyc13a&NEWS=N&AN=2016-55191-001
  24. 24. Torstveit MK, Rosenvinge JH, Sundgot-Borgen J. Prevalence of eating disorders and the predictive power of risk models in female elite athletes: A controlled study. Bauer Beals, Beals, Beglin, Black, Burckes-Miller, Byrne, Byrne, Cooper, De Souza, Fairburn, Fairburn, Fairburn, Fairburn, Favaro, Fogelholm, Garner, Garner, Hoek, Johnson, Johnson, McNulty, Otis, Rosenvinge, Rosenvinge, Shisslak, Siervo, Smolak, Sundgot B, editor. Scand J Med Sci Sports [Internet]. 2008;18(1):108–18. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=psyc6&NEWS=N&AN=2008-02206-003 pmid:17490455
  25. 25. Krentz EM, Warschburger P. Sport-related correlates of disordered eating: A comparison between aesthetic and ballgame sports. Ackard Aiken, Beals, Berry, Doninger, Field, Frazier, Garner, Hausenblas, Helfert, Helfert, Hinton, Krentz, Meermann, Muscat, Pasman, Petrie, Petrie, Petrie, Raudenbush, Rosendahl, Schneider, Shisslak, Smolak, Steffen, Stice, Stice, Stoutjesdyk, Sundgot- A, editor. Int J Sport Psychol [Internet]. 2011;42(6):548–64. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=psyc8&NEWS=N&AN=2012-03633-003
  26. 26. Sundgot-Borgen J. Prevalence of eating disorders in elite female athletes. Int J Sport Nutr Exerc Metab. 1993;3(1):29–40. pmid:8499936
  27. 27. Sundgot-Borgen J, Torstveit MK. Prevalence of eating disorders in elite athletes is higher than in the general population. Clin J Sport Med. 2004;14(1):25–32. pmid:14712163
  28. 28. Steinhausen H-C. The outcome of anorexia nervosa in the 20th century. Am J Psychiatry. 2002;159(8):1284–93. pmid:12153817
  29. 29. Stice E. Risk factors for eating pathology: Recent advances and future directions. 2001;
  30. 30. Petrie TA, Greenleaf CA. Eating disorders in sport: From theory to research to intervention. Abood Agras, Andersen, Ball, Baron, Beals, Beals, Beals, Beals, Beals, Black, Black, Blaydon, Blouin, Brownell, Brownell, Brownell, Carter, Cashel, Cattarin, Cohen, Crandall, Dale, Davis, Davis, Davis, Davis, DePalma, DiBartolo, Dick, Doninger, Drewnowsk A, editor. Handb Sport Psychol 3rd ed [Internet]. 2007;352–78. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=psyc5&NEWS=N&AN=2007-01666-024
  31. 31. Stoyel H, Shanmuganathan-Felton V, Meyer C, Serpell L. Psychological risk indicators of disordered eating in athletes. PLoS One. 2020;15(5):e0232979. pmid:32407345
  32. 32. Cole DA, Maxwell SE. Testing mediational models with longitudinal data: questions and tips in the use of structural equation modeling. J Abnorm Psychol. 2003;112(4):558. pmid:14674869
  33. 33. Anderson C, Petrie TA. Prevalence of Disordered Eating and Pathogenic Weight Control Behaviors among NCAA Division I Female Collegiate Gymnasts and Swimmers. Res Q Exerc Sport [Internet]. 2012 Mar 1;83(1):120–4. Available from: http://search.ebscohost.com/login.aspx?direct=true&AuthType=ip,shib&db=eric&AN=EJ990075&site=ehost-live&scope=site
  34. 34. Bratland-Sanda S, Sundgot-Borgen J, Ro O, Rosenvinge JH, Hoffart A, Martinsen EW. ‘I’m not physically active—I only go for walks’: Physical activity in patients with longstanding eating disorders. Ainsworth Borg, Bouten, Boyd, Cash, Davis, Davis, Fairburn, Freedson, Furnham, Harris, Mond, Plasqui, Sallis, Schmidt, Shrout, Silberstein, Sundgot-Borgen, Szabo, Timperio, Torstveit, Vansteelandt, Westerterp B, editor. Int J Eat Disord [Internet]. 2010;43(1):88–92. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=psyc7&NEWS=N&AN=2009-25187-013
  35. 35. Rothman KJ. BMI-related errors in the measurement of obesity. Int J Obes. 2008;32(3):S56–9. pmid:18695655
  36. 36. Heinberg LJ, Thompson JK, Stormer S. Development and validation of the sociocultural attitudes towards appearance questionnaire. Int J Eat Disord. 1995;17(1):81–9. pmid:7894457
  37. 37. Thompson JK, Van Den Berg P, Roehrig M, Guarda AS, Heinberg LJ. The sociocultural attitudes towards appearance scale‐3 (SATAQ‐3): Development and validation. Int J Eat Disord. 2004;35(3):293–304. pmid:15048945
  38. 38. Crawford JR, Henry JD. The Positive and Negative Affect Schedule (PANAS): Construct validity, measurement properties and normative data in a large non‐clinical sample. Br J Clin Psychol. 2004;43(3):245–65.
  39. 39. Watson D, Clark LA, Tellegen A. Development and validation of brief measures of positive and negative affect: the PANAS scales. J Pers Soc Psychol. 1988;54(6):1063. pmid:3397865
  40. 40. Garner DM, Olmstead MP, Polivy J. Development and validation of a multidimensional eating disorder inventory for anorexia nervosa and bulimia. Int J Eat Disord. 1983;2(2):15–34.
  41. 41. Thiel A, Gottfried H, Hesse FW. [Body image of the male athlete. A study of the psychological health of wrestlers and rowers of the lower weight class]. Das Korpererleb mannlicher Sport Eine Untersuchung zur Psych Gesundh von Ringern und Ruderern der unteren Gewichtsklassen [Internet]. 1993;43(12):432–8. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=med3&NEWS=N&AN=8146262 pmid:8146262
  42. 42. Hayes AF. Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. Guilford publications; 2017.
  43. 43. Muthén LK, Muthén B. Mplus. Compr Model Progr Appl Res user’s Guid. 2018;5.
  44. 44. Stice E, Schupak-Neuberg E, Shaw HE, Stein RI. Relation of media exposure to eating disorder symptomatology: an examination of mediating mechanisms. J Abnorm Psychol. 1994;103(4):836. pmid:7822589
  45. 45. Van Diest AMK, Perez M. Exploring the integration of thin-ideal internalization and self-objectification in the prevention of eating disorders. Body Image. 2013;10(1):16–25. pmid:23182310
  46. 46. Petrie TA, Greenleaf C. Eating disorders. Anderson Beals, Becker, Bonci, Cohen, Crowther, de Bruin, Galli, Greenleaf, Greenleaf, Hoch, Hudson, Isomaa, Johnson, Murray, Nichols, Olivardia, Petrie, Petrie, Reel, Smith, Stice, Stice, Sundgot-Borgen, Thompson, Torstveit, Torstveit, Tylka A, editor. Routledge companion to Sport Exerc Psychol Glob Perspect Fundam concepts [Internet]. 2014;837–51. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=psyc11&NEWS=N&AN=2013-21899-052
  47. 47. Torres-McGehee TM, Monsma E V, Gay JL, Minton DM, Mady-Foster AN. Prevalence of eating disorder risk and body image distortion among National Collegiate Athletic Association Division I varsity equestrian athletes. Ackard Black, Black, Bonci, Bulik, Carter, Caulfield, Cohen, Cox, Davis, Fairburn, Garner, Goitlieb, Greenleaf, Greenleaf, Hallinan, Hausenblas, Henriques, Hulley, Johnson, Lerner, Leydon, Mossavar-Rahmani, Nattiv, Peterson, Peterson, Rao, Rome, Salbach, B, editor. J Athl Train [Internet]. 2011;46(4):431–7. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=psyc8&NEWS=N&AN=2012-23312-002
  48. 48. Clausen L, Rosenvinge JH, Friborg O, Rokkedal K. Validating the Eating Disorder Inventory-3 (EDI-3): A comparison between 561 female eating disorders patients and 878 females from the general population. J Psychopathol Behav Assess. 2011;33(1):101–10. pmid:21472023
  49. 49. Thiel A, Paul T. Test–retest reliability of the Eating Disorder Inventory 2. J Psychosom Res. 2006;61(4):567–9. pmid:17011367
  50. 50. Calogero RM, Davis WN, Thompson JK. The Sociocultural Attitudes Toward Appearance Questionnaire (SATAQ-3): Reliability and normative comparisons of eating disordered patients. Cash Cattarin, Fairburn, Garner, Levine, Slice, Stice, Thompson, Thompson, Thompson, Thompson, Thompson, Tiggemann C, editor. Body Image [Internet]. 2004;1(2):193–8. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=psyc4&NEWS=N&AN=2005-01513-006 pmid:18089151
  51. 51. Duncan TE, Duncan SC. The ABC’s of LGM: An introductory guide to latent variable growth curve modeling. Soc Personal Psychol Compass. 2009;3(6):979–91. pmid:20577582
  52. 52. Shanmugam V, Jowett S, Meyer C. Application of the transdiagnostic cognitive-behavioral model of eating disorders to the athletic population. Ainsworth Bardone-Cone, Bardone-Cone, Bentler, Berry, Bodell, Bollen, Bowlby, Bowlby, Bowlby, Brannan, Brennan, Broberg, Browne, Bruch, Bulik, Button, Button, Byrne, Campbell, Carr, Cassidy, Cassin, Chassler, Cheung, Collins, Courtney, Currie, Davis, Dav A, editor. J Clin Sport Psychol [Internet]. 2011;5(2):166–91. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=psyc8&NEWS=N&AN=2011-13706-005
  53. 53. Steinfeldt JA, Gilchrist GA, Halterman AW, Gomory A, Steinfeldt MC. Drive for muscularity and conformity to masculine norms among college football players. Abell Addis, Andersen, Andersen, Brewer, Brewer, Brooks, Buckley, Cafri, Cafri, Cialdini, Cloud, Cornelius, Davis, Dixson, Farquhar, Festinger, Foley, Franzoi, Franzoi, Galli, Gilchrist, Gilchrist, Grove, Harrison, Helgeson, Hruby, Kahn, Labre, Leit, Leo A, editor. Psychol Men Masc [Internet]. 2011;12(4):324–38. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=psyc8&NEWS=N&AN=2011-17133-001
  54. 54. Lentillon-Kaestner V. Eating disorders among male athletes: A psychosocial perspective. Ackard Baum, Bern, Blayton, Botta, Brehms, Buddeberg-Fischer, Byrne, Chamay-Weber, Chambry, Choquet, Conraux, Dale, Davis, Davis, DiGioacchino DeBate, Doise, Dosil, Ferrand, Ferrand, Filaire, Filaire, Filaire, Garcia Hejl, Goldfield, Goldfield, Goldfield AP, editor. Contemp Top trends Psychol Sport [Internet]. 2014;177–200. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=psyc11&NEWS=N&AN=2014-02199-008
  55. 55. Mountjoy M, Brackenridge C, Arrington M, Blauwet C, Carska-Sheppard A, Fasting K, et al. International Olympic Committee consensus statement: harassment and abuse (non-accidental violence) in sport. Br J Sports Med. 2016;50(17):1019–29. pmid:27118273
  56. 56. Crow RB, Macintosh EW. Conceptualizing a meaningful definition of hazing in sport. Eur Sport Manag Q. 2009;9(4):433–51.
  57. 57. Shanmugam V, Jowett S, Meyer C. Eating psychopathology in athletes and nonathletes: The effect of situational and dispositional interpersonal difficulties. Allen Bruch, Calam, Derogatis, Dunkley, Dunkley, Fairburn, Fairburn, Fairburn, Frederick, Gerner, Gilbert, Goodwin, Greendorfer, Hayes, Hinrichsen, Imber, Jacobi, Johnson, Jones, Jones, Jowett, Kaye, Latzer, Leary, Leff, Lepine, Mayer, McIntosh, Meyer, M A, editor. J Clin Sport Psychol [Internet]. 2014;8(4):319–38. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=psyc11&NEWS=N&AN=2015-15708-001
  58. 58. Giles S, Fletcher D, Arnold R, Ashfield A, Harrison J. Measuring well-being in sport performers: where are we now and how do we progress? Sport Med. 2020;1–16. pmid:32103451