Comparison of Eating Disorders Symptoms and Body Image Between Individidual and Team Sport Adolescent Athletes During the COVID-19 Pandemic

COVID-19 Lowered Physical Activity and Increased Psychological Distress

The global outbreak of COVID-19 has resulted in closure of multiple recreational facilities (e.g., gyms) around the world. Many individuals are not able to participate in their regular individual or group physical activities. According to the World Health Organization (WHO), under such conditions, many people tend to be less physically active, have longer screen time, irregular sleep patterns, as well as worse diets, potentially resulting in weight gain, and loss of physical fitness, which can have a distinct impact on athletes (1). COVID-19 has also been associated with significantly higher levels of psychological distress, which can have a negative effect on people’s functioning in different domains, overall well-being, and quality of life (2).

COVID-19 Disruption of Routines

In particular, COVID-19 has significantly disrupted the routines of school and sports for adolescents (e.g., in-person training activities, organized sporting events), which can affect their usual eating behaviors, body image, and exercise patterns. Adolescence is a developmental growth period with a normally expected increased focus on body weight, which can lead to disordered eating (DE), eating disorders (ED), and body image concerns (3). Thus, adolescents (and especially athletes, who have more regimented eating and exercise behaviors) may be particularly vulnerable to these concerns during this unusual lockdown period.

The DE to ED Spectrum

Results from international studies that have addressed DE (altered dieting practices) during the first year of the pandemic suggest that there is large inter-individual variability in dietary changes (4, 5). For individuals experiencing persistent increases in unhealthy diet choices, there are important public health implications associated with Body Mass Index (BMI), gender, quality of life, living conditions, physical activity, and other dietary behaviors (4, 5).

DE can range on a spectrum from minor deviations away from normal eating patterns to major alterations in eating behaviors. These major alterations, often contributed to by a combination of genetic (family history) and environmental causes (e.g., COVD-19 stress), can result in individuals meeting the full diagnostic criteria for a primary ED, as defined by the Diagnostic and Statistical Manual of Mental Disorders-5 (DSM-5) (6), outlined below. EDs are associated with certain risk factors which may be especially developmentally relevant for adolescent athletes (e.g., age of ED onset is usually in adolescence, ED effects on the developing body can significantly interfere with an adolescent’s athletic performance).

Anorexia Nervosa (AN)

Per the DSM-5 (6), AN is defined as: a restriction of energy intake, leading to a significantly low body weight; intense fear of gaining weight, or persistent behavior that interferes with weight gain; and disturbance in body image, or lack of recognition of the seriousness of the current low body weight. Research indicates that individuals with a family history of AN, have childhood anxiety traits, and pursue avocations that encourage

thinness (e.g., elite athletes), are at an increased risk for developing AN (6).

Bulimia Nervosa (BN)

Also per the DSM-5 (6), BN is defined as: recurrent episodes of binge eating (feeling lack of control when eating a large amount of food in 2 hours); recurrent compensatory behaviors to prevent weight gain (e.g., vomiting); these behaviors occur at least once a week for 3 months; self-evaluation is unduly influenced by body shape/weight; and the disturbance does not occur exclusively during episodes of AN. Research indicates that individuals with a family history of BN, have childhood obesity, subsequently attempted to restrict food and/or overexercise (e.g., through sports), but ended up binging later, are at increased risk for developing BN (6).

Body Image

A cross-cutting symptom of the above primary types of EDs is the concern about body image, which is the internal representation of an individual's external appearance (e.g., self-evaluations of one’s physical appearance) (7). Given that appearance is an integral part of one’s identity, and it plays a role in social situations, body image is an important aspect of life for most people, and particularly for those individuals who are at risk for, or are experiencing, an ED (8). When these concerns are combined with environmental pressures (e.g., to be an elite athlete, competing with peers on measures of appearance and performance), body image can become even more ingrained, as it is linked to their identity (e.g., as an athlete).

Sport-Specific Pressures

Another aspect of an athlete’s identity includes how they respond to the unique pressures of their specific sport. Petrie and Greenleaf formulated a theoretical etiological model of eight mediating factors in the development of DE in athletes (9, 10). One factor is the type of sport (individual vs. team sports; lean vs. non-lean sports). Individual sports tend to be ones that emphasize leanness (11), with the belief that lower body weight improves performance (e.g., track and field, martial arts, wrestling) (11). Alternatively, team sports tend to be ones that are non-lean (12), in that they do not require a low body weight in order for an athlete to be competitive (e.g., basketball, soccer, volleyball) (12). Lean sports may increase risk for DE because athletes may engage in pathogenic weight-control behaviors to achieve a lower body weight (13). A recent systematic review (14) of the prevalence of DE in athletes categorized by emphasis on leanness and activity reported that six out of the seven included studies found a significant increase in DE rates among lean sport types, compared to non-lean sport types (14).

Sports can be further divided into six subgroups: aesthetic, weight-dependent, endurance, ball game, power, and technical sports (15). Of these categories, aesthetic, weight-dependent, and endurance sports are typically considered lean sports, whereas ball game, power, and technical sports are considered non-lean sports. In particular, aesthetic sports, which are judged by a complex set of rules favoring appearance (e.g., ballet, gymnastics), have been shown in sport-specific studies to have a higher prevalence of DE and ED compared to other sport subgroups (16–18). Weight-dependent sports, which divide competitors into different categories based on their weight (e.g., wrestling, martial arts), have similarly been shown to have a higher ED pathology than those in other sport subgroups (19, 20). Endurance sports typically associate lower body weight with a higher level of competition (e.g., track and field-running, cross-country skiing) (21). As a result, athletes may utilize DE to achieve a body weight that is too low, resulting in an energy imbalance with the high metabolic requirements of their aerobic training (22, 23).

Another sport-specific factor to consider is peer competition, which can exacerbate DE, EDs, and distorted body image (24). While both team and individual sports require competition, cooperation is much more prevalent in team than in individual sports. Team athletes have to compete (for starting roles) while also cooperating (for team success) with the same teammates. Thus, it is possible that team athletes’ additional goal of cooperation more evenly balances out the goal of competition, compared to individual athletes’ sole requirement for competition (25). Additionally, peer competition can lead athletes to deny the seriousness of their concerns, and if they do seek help from a doctor, it is more likely due to complaints of decreased performance, rather than the symptoms of an ED, per se (25).

Gender-Specific Pressures

Although EDs are often thought of as feminine illnesses, epidemiological studies indicate that males are also at-risk for developing EDs (26–28). Most notably, adolescent males may be less likely to seek treatment than females due to an overall higher degree of shame and stigma related to their EDs (29–31). Further, physicians and other health care providers may be less likely to recognize disordered eating symptoms in adolescent males due differences in symptom presentation (e.g., male focus on muscle, vs. female focus on weight) (30, 31).

These gender-specific pressures can overlap with the above-noted sport-specific pressures. A systematic review (32) of risk factors for eating psychopathology included a study which examined the relationship between gender, type of sport (lean vs. non-lean), body dissatisfaction, and self-esteem, with DE behaviors in Division 1 college athletes (32). They found that participating in lean sports was associated with increased DE and body dissatisfaction for male athletes, but not female athletes (33). This unique vulnerability for male athletes, in addition to the factors related to lower likelihood of treatment for males (e.g., stigma, shame, provider misconceptions), suggests that male athletes should be targeted by coaches for prevention and early intervention for DE, ED, and distorted body image (34).

Adolescent Athletes-Additional COVID-19 Challenges

In addition to the above-noted challenges (i.e., pubertal development, focus on exercise, food, and body image, sport-specific pressures, gender-specific pressures) outside of COVID-19 (3), the pandemic has created additional concerns for adolescent athletes with EDs. Touyz and colleagues reported on the exacerbation of EDs during the beginning of lockdown (in 2020) (35). For many individuals with AN, lack of access to the usual forms of exercise (e.g., going to the gym) heightened concerns about gaining weight, leading to further restriction and emaciation, which increased their potential for experiencing more serious COVID-19 complications (35). If the body is too physically compromised from AN and COVID-19 (e.g., respiratory weakness, heart conditions, decreasing endurance, inability to perform skills), it is possible the person may not recover enough to participate in athletics again. For many individuals with BN during this same time period, increased access to food while in quarantine at home, without the normal athletic exercise routine to balance out the calorie intake, resulted in reported weight gain that they fear may hinder their athletic performance (35).

Furthermore, recent research on the psychological impact of the COVID-19 pandemic specifically on elite athletes suggests that levels of perceived stress and negative emotions during the lockdown (in 2021) are higher in athletes who experience changes in their: 1) motivation to compete (decrease from pre-pandemic); 2) stress-management (lowered coping abilities from pre-pandemic); and 3) post-pandemic performance expectations (maintain their usual high pre-pandemic standards, without allowance for COVID-19 interruptions) (36). Conversely, positive emotions during the lockdown are higher in athletes who experience no changes in their competition motivation, stress-management, or post-pandemic performance expectations (36).

The exacerbation of both physical and psychological symptoms of adolescent athletes with ED during the pandemic have unfortunately been further impacted by delays in timely treatment. Healthcare systems have reported prolonged waitlists, increased referrals, and deprioritized status for sub-acute conditions, contributing to less services for everyone (37, 38). While access to evidenced based treatment (EBT) for EDs was challenging for many individuals before COVID-19, the lockdown has imposed further barriers (e.g., travel, insurance limits).

Aim of the Present Study

COVID-19 thus presents specialized issues for adolescent athletes, particularly those at risk for, or experiencing, DE, ED, and associated negative body image. Sport-specific pressures, especially those related to individual sports (e.g., strict weight requirements, revealing sports attire, aesthetic judgments), may exacerbate these concerns. An additional factor to consider is the gender of the athletes, with males typically experiencing more barriers to treatment than females. The aim of the present study was therefore to compare ED symptoms and body image between individual and team sport adolescent male athletes who were geographically located in one of the most affected areas of Iran during COVID-19.

Participants

Participants of the study included 124 Iranian male adolescent athletes residing in the Mazandaran province (one of the most affected areas of Iran during COVID-19), who were recruited through the three University study sites nearby (Farhangian University, Tarbiat Modares University, and Islamic Azad University). The athletes played in 1 of 6 sports (individual or group): 1) Taekwondo (individual); 2) Track and Field (individual); 3) Wrestling (individual); 4) Volleyball (group); 5) Basketball (group); or 6) Soccer (group).

Study inclusion criteria were: 1) aged 12 to 19 years old; 2) no history of physical or mental illness, assessed pre-study through participant self-report via the WHO Global Physical Activity Questionnaire (WHO-GPAQ) (39), and the WHO-Composite International Diagnostic Interview (WHO-CIDI) (40), respectively; 3) sports history for at least 3 years, assessed via the WHO-GPAQ; and 4) written consent by the adolescents and their parents for participation in the research. The study exclusion criterion was: 1) lack of regular physical activity, assessed via the WHO-GPAQ. This criterion was included to differentiate between a lack of regular physical activity due to normal patterns (i.e., sedentary lifestyle), vs. lack of regular physical activity due to changes during the COVID-19 pandemic assessment period (i.e., reduced exercise).

Using the Krejcie and Morgan Sampling Method to simplify the process of determining the sample size for a finite population (41), 124 athletes were selected to participate in this study, then divided equally into two groups: the individual sport group (n = 62), and the team sport group (n = 62). Data were collected from June through August 2020 during a COVID-19 quarantine period. Throughout this time, all athletes practiced and played at home, or in socially isolated environments, depending on the training that the coaches designed for them. Thus, the athletes had continued exercise during the lockdown, albeit with modified routines and formats.

Procedure

The governing Medical Ethics and History of Medicine Research Center for each Iranian university study site reviewed and approved the study before enrollment. Participants were recruited via a study research coordinator who had contacted the respective officials of the 6 sports groups. The study procedures were explained, and informed consent was obtained from all participants and their parents prior to study initiation.

Measures

De-identified study data (sociodemographic variables, questionnaires) were collected online at the participants’ homes during the quarantine period of COVID-19. The Persian (Farsi) translations of the questionnaires were used for the participants’ comfort and familiarity with the language. These translated versions were evaluated, and found to be reliable and valid for the Iranian population (please see below for the psychometrics of each questionnaire).

Eating Attitudes Test-26 (EAT-26)

The EAT-26 (50) is a 26-item self-report measure with 3 subscales: 1) thoughts and behaviors related to dieting (Dieting); 2) preoccupation with food and impulses to binge and purge (Bulimia and Food Preoccupation); and 3) attempts to control food intake (Oral Control). Response agreement is rated on a 4-point Likert scale (0 = “Never,” to 3 = “Always”), with total scores ranging from 0–78 (clinical cut-off score of ≥ 20 indicates disordered eating). Garner and colleagues reported that the EAT-26 items have a reliability of 0.88, the overall scale has a content validity of 0.91, and the 3 subscales have a content validity of 0.78 (50). In the present study, the Persian (Farsi) translation version of the EAT-26 (51) had a reliability Cronbach's α of 0.90. The test-retest reliability for the EAT-26 3 subscales was good (r = 0.84–0.89, P < 0.01).

Body-Esteem Scale for Adolescents and Adults (BESAA)

The BESAA (52) is a 23-item self-report measure with 3 subscales: 1) feelings regarding one’s appearance (Appearance); 2) perceptions of others’ evaluations about one’s appearance (Attribution); and 3) weight satisfaction (Weight). Response agreement is rated on a 5-point Likert scale (1 = “Never,” to 5 = “Always”), with total scores ranging from 23–115 (higher scores indicate more positive body-esteem). Mendelson and colleagues reported that the BESAA items have an internal consistency of 0.89, a reliability of 0.88, and a content validity of 0.72 (52). In the present study, the Persian (Farsi) translation version of the BESSA (53) demonstrated good model fit statistics (chi-square/df = 3.41, P < 0.001) and good internal consistency for the 3 subscales: BE-weight (α = 0.85), BE-appearance (α = 0.76), and BE-attribution (α = 0.69). The test-retest reliability for the BESSA 3 subscales was moderate (r = 0.57–0.68, P < 0.01).

Data Analyses

All data were entered into the Statistical Package for Social Science (SPSS) 24 software for analyses (54). The Kolmogorov-Smirnov Test (55) was used to check the normality of the EAT-26 and BESSA data (normality is the first statistical assumption required for the subsequent Multiple Analysis of Variance (MANOVA) (56). MANOVA was chosen instead of a single t-test with multiple dependent variables in order to limit the joint error rate (when you perform a series of tests, the joint probability of rejecting a true null hypothesis increases with each additional test; instead, of you perform one MANOVA, the error rate equals the significance level). The M-Box Test of Equality of Covariance Matrices (57) was used to check for the equality of covariance matrices for the overall differences in EAT-26 and BESSA scores between the individual and team athlete groups (equality of covariance matrices is the second statistical assumption required for the subsequent MANOVA). Levene’s Test of Homogeneity of Variance (58) was used to check the homogeneity of variance for the EAT-26 and BESSA specific subscale scores of the individual and team athlete groups (homogeneity of variance is the third statistical assumption required for the subsequent MANOVA). With the above statistical assumptions being met, the MANOVA was first utilized to compare overall differences between the individual and team athlete groups on the dependent variables in general (all EAT-26 and BESSA subscale scores). Then, the MANOVA was further refined to detail the specific dependent variables (individual EAT-26 and BESSA subscale scores) that were significantly different between the individual and team athlete groups.

Individual and Team Athlete Groups Sociodemographic Variables

The sociodemographic variables of the individual (n = 62) and team (n = 62) athlete groups are presented in Table 1. In summary, compared to team athletes, individual athletes had a significantly higher mean age of 14.1 years (Standard Deviation, SD = 0.6 years), and mean Body Mass Index (BMI) of 21.1 (SD = 2.1), as calculated by the Centers for Disease Control and Prevention (CDC) BMI Percentile Calculator for Child and Teen (59) (p = 0.001).

Mean EAT-26 and BESSA Subscale and Total Scores of the Individual and Team Athlete Groups

Table 2 notes the differences (mean and SD) between the individual and team athlete groups on EAT-26 and BESSA subscale and total scores. In summary, individual athletes had significantly higher mean scores compared to team athletes on the EAT-26 Dieting, Oral Control, and Total Eating Disorders subscales (p = 0.001). Team athletes had significantly higher mean scores compared to individual athletes on the EAT-26 Bulimia and Food Preoccupation subscale, and the BESSA Appearance, Weight, Attribution, and Total Body Image subscales (p = 0.001).

Statistical Tests for the Assumptions of the Overall MANOVA

The M-Box Test of Equality of Covariance Matrices (57) for the overall differences in EAT-26 and BESSA scores between the individual and team athlete groups is shown in Supplemental Table 1. Levene’s Test of Homogeneity of Variance (58) for the homogeneity of variance in the EAT-26 and BESSA subscale scores between the individual and team athlete groups is presented in Supplemental Table 2. The MANOVA for the overall differences in EAT-26 and BESSA scores of the individual and team athlete groups is noted in Supplemental Table 3. In summary, 49.7% of the variance related to the difference between the individual and team athlete groups is due to the interaction of all of the dependent variables.

MANOVA Specific Dependent Variables

Table 3 details the specific dependent variables (individual EAT-26 and BESSA subscale scores) in the MANOVA that had significant differences between the individual and team athlete groups. In summary, the individual athlete group had significantly higher scores on the EAT-26 Bulimia and Food Preoccupation subscales (p = 0.019), and significantly higher scores on the BESSA Appearance (p = 0.001), Weight (p = 0.001), and Attribution (p = 0.001) subscales, compared to the team athlete group. However, there were no significant differences between the two athlete groups on the EAT-26 Dieting and Oral Control subscales.

Clinical Implications

In the current study, there were several important findings and clinical implications related to the athletic settings that the participants were in (individual vs. team sports). It should be noted that there were no significant differences between the individual and team athlete groups on the EAT-26 subscale scores for Dieting and Oral Control, which indicates that both groups are at an equal risk for these concerns, and should be assessed and monitored similarly. However, the individual athlete group had significantly higher EAT-26 subscale scores for Bulimia and Food Preoccupation, as well as significantly higher BESAA subscale scores for Appearance, Weight, and Attribution, compared to the team athlete group. This suggests that the individual athlete group is at a relatively higher risk for these concerns during COVID-19, and that coaches should pay special attention to these particular athletes, in order to provide early intervention, and hopefully mitigate the risk of long-term consequences.

Limitations

The findings of the current study must be viewed within the context of certain limitations.

First, the study utilized a sample of Iranian male adolescent athletes during the beginning of COVID-19 quarantine (June-August 2020); therefore, results cannot be generalized to other populations. Future research should endeavor to conduct similar studies in other populations, in order to compare, and potentially replicate, the results. Second, the level of reported ED pathology in both individual and team athlete groups during this short assessment period was relatively low, compared to pre-COVID-19 research indicating that athletes have three times the rate of EDs as the general population (65–70). Given this known risk factor, we can hypothesize that these groups will have experienced worsened ED and body image as the pandemic has continued. Third, the cross-sectional assessment of ED and body image in the current study preclude causal inferences for these relationships. Prospective research could employ a more experimental, longitudinal design, in order to elucidate the effects of each of these variables. Fourth, due to the early assessment period for this study, structured measurements of pandemic barriers (e.g., impact on physical and mental health for individuals, family, and friends) had not yet been developed. As these scales are now available, they should be utilized in conjunction with other instruments of interest. Fifth, the participant self-reports were not supported by additional investigator conducted diagnostic interviews. Subsequent research should endeavor to also include investigator interviews (e.g., by tele-interview during COVID-19) to support any self-report measures. Finally, we did not conduct sport-specific analyses due to concerns of low statistical power. It is possible that the characteristics of the athletes within one or two sports may be responsible for the differences that were seen between groups, and the conclusions might thus be more specific than we presented. Future research could increase the overall and sport-specific subgroup samples in order to provide additional cell sizes for more robust analyses.