An upsurge of scientific interest in the area of disordered eating among athletes has been observed, with more qualitative methodologies focusing on how athletes experience this condition [1
]. Disordered eating is shown to have a higher prevalence among athletes than general population, even reaching 45% in female athletes [2
] and this condition is especially emphasized in sports where weight controlling behaviour is expected, such as martial arts, running, synchronized swimming and gymnastics [3
]. Moreover, patterns related to eating disorders might be difficult to detect among some athletes [5
], which is a problem because eating disorder consequences are oftentimes far-reaching and could impair metabolism, bone health, cardiovascular health and mental health [6
]. In addition, a decrease in food intake frequently leads to lower energy availability among athletes, a state that causes hormonal alterations, menstrual irregularities and impaired bone health, which consequently has a negative impact on sport performance as well [7
Examination of athletes’ personality characteristics should fulfil the gaps in the knowledge regarding eating disorder-related outcomes [8
]. For example, elevated perfectionism is often observed among those with eating disorder symptoms and is considered to be multidimensional [9
], consisting of adaptive (or healthy; perfectionistic strivings) and maladaptive (or unhealthy; perfectionistic concerns) facets [9
]. In the sport context, adaptive perfectionism was associated with greater training performance [11
] and lower levels of training distress [12
], while maladaptive perfectionism seems to be a risk factor for exercise dependence [13
]. Additionally, Gotwals, Stroeber, Dunn and Stoll [14
] noted that perfectionistic strivings in athletes are primarily adaptive, occasionally neutral and rarely maladaptive, when controlling for perfectionistic concerns. However, although adaptive perfectionistic strivings are usually considered essential for the greater sport achievements [15
], Flett and Hewitt [16
] emphasized that interpreting perfectionistic strivings as adaptive after controlling for perfectionistic concerns might not be a veracious approach in a way that athletes rarely exhibit only perfectionistic strivings without also being highly self-evaluative. A recent meta-analysis by Limburg, Watson, Hagger and Egan [17
] showed that both dimensions of perfectionism are often associated with various psychopathological conditions. Likewise, both adaptive and maladaptive perfectionism appear to be elevated in individuals with eating disorder [10
]. Furthermore, perfectionism, achievement motivation and competitiveness were addressed as vulnerabilities for disordered eating development in competitive female athletes [18
]. Perfectionism was also related to dieting and bulimia symptoms among female college athletes [19
] and was shown to be the greatest risk factor for disordered eating among female athletes by Hopkinson and Lock [20
]. It is possible that perfectionists, who often set very high standards and struggle to cope even with small failures, are more likely to develop eating disorder once enrolled in a competitive athletic environment [8
]. Also, this relationship seems to be highlighted among female athletes rather than males and Somasundaram and Burgess [21
] showed that sport type might moderate the relationship between perfectionism dimensions and disordered eating among female athletes. Therefore, the role of adaptive perfectionism might be ambiguous and further examination of its contribution to disordered eating in both female and male athletes would be useful.
A core symptom of eating disorders is body dissatisfaction [22
] but this relationship seems to be less straightforward in athletic population [23
]. Exposure to high standards and constant evaluation of physical appearance in some sports can lead to negative body image if these standards are not met [24
]. For example, female collegiate athletes symptomatic for an eating disorder reported more perceived pressure to be thin than their asymptomatic peers [25
]. This is most commonly noticed in so-called “aesthetic” sports, such as gymnastics, dancing, figure skating [26
] and competitive bodybuilding where increased preoccupation with weight and shape, along with body dissatisfaction has been reported in the literature [27
]. Collectively, these studies indicate that being dissatisfied with one’s own body represents a vulnerability for the development of unhealthy eating habits in female athletes but little is known about this relationship among males. Furthermore, body dissatisfaction could be the factor through which athletes who are perfectionists develop eating disturbances or athletes could experience symptoms of disordered eating due to perfectionistic tendencies solely. Previous studies have proven the mediating effect of body satisfaction between perfectionism facets and unhealthy eating behaviour [28
]. However, not always were the athletes of both genders included or the role of sport type examined and recent findings provide support for the fact that eating disorders are common but in research often neglected among men [30
Therefore, it would be important to comprehend associations between constructs related to unhealthy eating patterns and what triggers these issues in order to assure better eating disorder prevention among athletes, which would lead to greater health- and performance-related outcomes. Finally, due to dissenting findings in the area of perfectionism research, it would be beneficial to further clarify its relationship with both adaptive and maladaptive outcomes in athletic population. Therefore, the aim of this study was to examine whether adaptive and maladaptive perfectionism predict disordered eating symptoms among female and male athletes of team and individual sport and whether body satisfaction can change the significance of this relationship as a mediator between perfectionism and disordered eating.
Findings in the present study indicate that perfectionism and body satisfaction explain dieting among female athletes to some extent, as evidenced by the results of the mediation analysis which yielded a significant indirect effect. Nevertheless, due to confidence intervals being very close to value zero, these effects might not be large enough to be considered as important for explaining dieting behaviour among females. For the male athletes, none of the predictors contributed to the understanding of their dieting behaviour. In addition, age, sport type and duration of sport participation did not contribute to dieting behaviour in case of male and female participants in the present study.
Male and female athletes of the current study did not differ in the prevalence of dieting behaviour. Generally, the prevalence of eating disorders is higher in females than in males and among athletes compared to non-athletes [2
]. Precisely, Bratland-Sanda and Sundgot-Borgen [2
] reported a prevalence range of 0–19% in male and 6–45% in female athletes. The findings of the present study are thus somewhat unexpected since women generally exhibit a higher risk for developing an eating disorder, mostly due to expectations from the society and thin-idealization [37
]. Nordin-Bates et al. [38
], in their longitudinal study, reported that the percentage of male and female dancers who scored above the cut-off in disordered eating attitudes changed between time points of data collection. Cross-sectional studies, like the present one, however, cannot compare time points due to obvious limitation and one-point data collection may not yield stable findings. With this in mind, prevalence of male and female disordered eating symptoms might have differed if we collected data longitudinally.
Adaptive and maladaptive perfectionism were associated solely with body satisfaction among female athletes in the present study but not with dieting behaviour directly. Although Ferrand and Brunet [29
] also reported no significant contribution of adaptive perfectionism to eating disorder symptoms among male cyclists, Somasundaram and Burgess [21
] showed that evaluative concerns (maladaptive perfectionism) have higher association with disordered eating than strivings (adaptive perfectionism) in both female athletes and non-athletes. Further, Hopkinson and Lock [20
] reported that professional athletes exhibited more eating disorder symptoms than recreational athletes, mostly because of elevated general perfectionism than intensity of training. To our knowledge, in the area of disordered eating, no research has yet demonstrated the potentially beneficial role of adaptive perfectionism dimension among athletes. Hence, the results of the current study are thus in contradiction with previous studies as we found adaptive perfectionism to be positively associated with body satisfaction, which is negatively linked to dieting. Perhaps the adaptive perfectionism dimension captured in present study is highly determined by the self-report tool being used, as different perfectionism questionnaires focus on different aspect of this trait. In fact, adaptive perfectionism measured by PANPS might be more representative of positive feelings associated with achieved outcomes, not strivings per se [39
]. However, inconsistent findings warrant more scientific interest in this area.
Satisfaction with one’s body was shown to be the largest independent predictor of dieting in the present study and their relationship was negative. This relationship between body satisfaction and disordered eating is not entirely straightforward in the case of the athletes as female college athletes actually reported higher body satisfaction than non-athletes but also elevated perfectionism related to unhealthy eating attitudes, which was accentuated among athletes in judged sports [19
]. However, Ferrand, Magnan, Rouveix and Filaire [40
] showed that dissatisfaction with body weight was positively associated with both self-oriented perfectionism and disordered eating in synchronized swimmers and operated as a mediator between these concepts. Therefore, it would be important to further establish the relationship between perfectionism dimensions and body satisfaction in the area of disordered eating but also take into a consideration the context in which athletes perceive their body [23
]. With that being said, athletes recruited for the current study were on-going comprehensive education about physical activity (including psychological aspects related to it) which potentially increased their awareness and consequently might have served as a protective factor against development of disordered eating symptoms. However, possible influence of participants’ education was not examined in the present study.
Sport type did not predict dieting behaviour in the current study, which is somewhat unexpected finding since sport type was shown to play a significant role in motivation for reshaping body and development of body dissatisfaction [41
]. However, when measuring disordered eating as an outcome, Haase [42
] showed that athletes of individual sports experienced a higher level of disordered eating symptoms than team sport athletes, whilst in another study, sport type did not play a role in disordered eating among male and female athletes [20
]. Therefore, the role of sport type seems to be inconsistently supported in the literature and our results agree with those in the study by Hopkinson and Lock [20
]. Perhaps using more sport categories in the present study, such as aesthetic and fighting sports, would have demonstrated somewhat different findings than the ones we observed.
Several limitations of the present study should be noted. The data was based on self-reported measures, which often contain a certain amount of dishonesty and socially desirable answers. Cross-sectional design of this study seriously limits the causal inference between perfectionism, body satisfaction and dieting. Furthermore, the convenient sample in the present study consisted of athletes who were also enrolled in a university program (kinesiology students) which might limit the generalizability to other athletes who might not be familiar with potential issues related to perfectionism and body satisfaction. In addition, potential restrictions came from the measurement instruments used for data collection since some researchers (e.g., [39
]) have suggested that PANPS should not be used as a measure of perfectionism due to a questionable factor validity of the “positive perfectionism” subscale. Furthermore, EAT-26 was shown to have a low positive predictive value due to the rarity of targeted conditions [43
] while two out of three EAT-26 subscales demonstrated low internal consistency and omitting them while investigating disordered eating may not be the best approach. Finally, the number of female participants was notably lower than males, which might have affected some of the analysis were gender-split was included.