Risk Factors for Disordered Eating in Female University Athletes.

Berry, Tanya R. ; Howe, Bruce L.

There has been concern expressed that participation in sport may encourage the onset of disordered eating which, in turn, may lead to a clinical eating disorder. This study examined social pressure, self-esteem, body image, and competition anxiety as risk factors for disordered eating in female university level athletes from selected sports. In addition, the relationships of body fat percentage and body mass index to indicators of disordered eating were studied. Canonical correlations showed significant relationships between restrained eating and all the risk factors as well as a significant relationship between high body mass index and restrained eating. Individual regression analysis showed that restrained eating was significantly predicted by body image and social pressure. Emotional eating was significantly predicted by body image. There were no significant differences apparent in the symptoms for eating disorders among the various sport teams. It was concluded that psychological predisposing factors to e ating disorders do occur in athletes but more research is needed to identify the extent of the interrelationship between the variables. It was shown that all athletes, regardless of sport, could be considered at risk if the predisposing factors identified in this research are present. Finally, it is recommended that more research be undertaken to examine the role of coaches and peers in the development of disordered eating.

Eating disorders such as anorexia nervosa (AN) and bulimia nervosa (BN) are often characterised as diseases of young, middle class, white women (Thompson & Sherman, 1993). The American Psychological Association in the Diagnostic and Statistical Manual, fourth edition (DSM-IV), has outlined strict, clinical, diagnostic criteria for these eating disorders. However it has been recognized, that within identifiable population groups, some individuals may exhibit disordered eating and distorted body image that do not necessarily meet the DSMIV criteria for a clinical diagnosis (Beals & Manore, 1994). It has been further hypothesized that these sub clinical syndromes may be more common than the full clinical syndromes themselves (Shaw & Garfinkel, 1990). Athletes have been identified as one of the groups that display high levels of disordered eating, although this finding has not been consistent, as in one study at least, it has been found that they may be at a lower risk than members of the general population (Wil kins, Boland, & Albinson, 1991). In contrast, it has also been claimed that eating disorders may be more pervasive than expected, particularly in specific sports (Beals & Manore, 1994).

It has been reported that sports with an emphasis on aesthetics, sports that emphasize a lean body build, and sports with weight classes have higher incidences of participants with eating disorder symptoms. Researchers investigating high-performance athletes across Canada found that female athletes participating in weight-matched sports (lightweight rowing and judo) and athletes in sports that emphasize leanness (gymnastics and diving) had significantly higher scores on an eating disorder inventory than athletes in nonweight-restricted sports (volleyball and heavyweight rowing) (Stoutjesdyk & Jevne, 1993). One researcher, using elite Norwegian female athletes as subjects, found that the sports with the greatest number of athletes using pathogenic weight control methods were those that were considered aesthetic, were weight dependent, or required endurance (Sundgot-Borgen, 1993). Similarly, Sundgot-Borgen (1994) found that the prevalence of eating disorders among elite female Norwegian athletes was significan tly higher in aesthetic and weight dependent sports than in other sport populations. Taub and Blinde (1992) compared adolescent female athletes with nonathletes on behavioral and psychological traits associated with eating disorders as well as the use of pathogenic weight loss techniques such as vomiting and diet aids. They found that athletes showed higher perfectionism and bulimia than nonathletes, but found no differences among various sport teams. Research in this area should continue to look at all sporting groups and athletes for evidence of eating disorder symptomology.

There have been several variables identified as possibly contributing to patterns of disordered eating in athletes. Among these are self-esteem, body image, and social pressure. Although some researchers have identified social pressure as a possible variable contributing to the problem (Sundgot-Borgen, 1994b; Rosen & Hough, 1988; Williamson, Netemeyer, Jackman, Anderson, Funsch, & Rabalais, 1995), the role that social factors, such as coach or peer pressure, may play in the development of disordered eating has not been clearly identified. In addition, the pressure on an athlete to meet weight restrictions or to conform to a certain body type may come from a number of different sources: general societal pressure; peer, trainer or coach pressure, as well as the judging criteria used in certain sports (Beals & Manore, 1994). This variable needs to be examined further as a possible contributing factor to disordered eating in athletes.

Among the important psychological factors that have been associated with eating disorders are self-esteem and body image. Low self-esteem has been shown to be common in individuals with eating disorders, and because of its association with heightened self-awareness, it may be a precipitating factor in the development of an eating disorder (Lindeman, 1994). Further, low self-esteem had a strong negative effect on dieting and bingeing behavior in adolescent girls (Neumark-Sztainer et al., 1996).

The role that self-esteem plays in the development of eating disorders in athletes is less clear. Davis and Cowles (1989) found that strenuous exercise may increase the possibility of eating disorders developing among female athletes in sports with an emphasis on a lean body, but their results were inconclusive as to whether it is the more emotionally vulnerable athletes who gravitate to these sports or whether it is the activity that results in lower emotional wellbeing. Thompson and Sherman (1993) wrote that sport could have three possible roles in affecting eating disorders: that sport can attract already at risk individuals, that participation can result in the disorder, or that sport can precipitate an eating disorder in those who are predisposed to its development. It may be then, that individuals with low self-esteem who are just starting or who are already involved in competitive sports are at risk for the development of an eating disorder particularly when you factor in the added pressures of the sp orts arena.

It has been stated that body-image disturbances are present in all individuals with eating disorders (Molinari, 1995). Typically, it has been reported that athletes show higher self-esteem and body satisfaction than nonathletes (Wilkins et al., 1991). However, athletes may represent a special population in their perceptions of body image. As a group, athletes typically most closely embody the societal ideal of a thin, trim build (Beals & Manore, 1994). It has been found that athletes have higher body image scores than nonathletes but that there is an inverse relationship between percent body fat and body image scores for athletes and nonathietes alike (Wilkins et al., 1991). Brownell, Rodin and Wilmore (1992), however, concluded that "the greater extent to which an athlete's body deviates from the 'ideal' for a particular sport, then, the greater the risk that the athlete will develop an eating disorder" (p. 122).

Body image has been found to be a significant mediating variable for competition anxiety, social factors and low self-appraisal, leading to eating disorder symptoms (Williamson et al., 1995). Brownell et al., (1992) suggested that the competitive nature of sports may be associated with the development of eating disorders. A study by Furst and Tenebaum (1984) reported that athletes with higher anxiety also had higher body dissatisfaction. This result was explained by the nature between anxiety and subjective levels of perceived satisfaction and success. Lower anxiety was reported by those athletes who were satisfied with their activity level, regardless of the level at which they were involved. The authors suggest that future research should consider that higher level athletes may not be more satisfied with their involvement and because these athletes may be comparing themselves to different standards of achievement. Greater body dissatisfaction may then result. These findings also suggest that several variab les may need to be present for the development of an eating disorder in an athlete.

Although research is needed in this area, there are issues that cause potential difficulties because of the nature of the problem itself. Most significantly of these is the subjective nature of the testing procedure through the use of questionnaires. Sundgot-Borgen (1993) found that athletes tend to under-report eating disorders because of the potential negative consequences such as not being allowed to compete. Shaw and Garfinkel (1990) point out that perceived loss of control is difficult to assess reliably and that the precise reporting of binge eating may be difficult. For example, what one person might consider a binge may be a normal meal to another.

From these findings, it is apparent that many questions still exist about an athlete's susceptibility to developing dieting problems and disordered eating, particularly about the role that social influences can have in unhealthy eating practices. The purpose of this research therefore, was to examine social pressure, self-esteem, competition anxiety, and body image as risk factors in the development of disordered eating patterns in athletes. Because of problems associated with research in the eating disorders, a secondary purpose was to determine if there was a relationship between low body mass index or low body fat and eating disorder symptoms.

Method

Subjects

Subjects were 46 female University varsity athletes competing in the 1996- '97 athletic season who volunteered to participate in this study. The subjects' ages ranged from 17 to 24 years and represented the sports of field hockey (n = 8), swimming (n = 8), soccer (n = 10), rowing (n = 8) and basketball (n = 12).

Procedures

All subjects were weighed and measured at a University fitness-testing center. Height and weight were taken to determine subject's body mass index (BMI). During the same testing session, skin fold measurements were taken at the bicep, tricep, suprailiac and subscapular sites. Percent body fat was calculated using the Durnin and Womersly method (Baumgartner & Johnson, 1982). The same physical data for the basketball players was collected by other researchers as part of a concurrent study.

Immediately following the physical measurements, subjects completed a series of questionnaires on self-esteem, body image, competition anxiety, social influence and eating disorder symptoms. The basketball players completed the questionnaires during a testing session prior to one of their regular practices.

Self-esteem was measured using Rosenberg's Self-Esteem Scale. The test consists of ten items, scored on a four point Likert scale, which measures the self-acceptance aspect of self-esteem. A higher score indicates lower self-esteem. It has an alpha level of .81 showing high internal validity (Neumark-Sztainer et al., 1996).

Competition anxiety was tested using Marten's Sport Competition Anxiety Test (Martens, 1977). A higher score indicates higher competition anxiety. It is a widely used questionnaire showing high internal and construct validity (Martens, & Simon, 1976; Cox, Qiu & Liu,1993). Corcoran (1989) reported strong validity of the SCAT, noting that there was a correlation of .56 between the SCAT and how a player sees him or herself just before competition.

Body image was assessed using the Body Shape Questionnaire developed by Cooper, Taylor, Cooper and Fairburn (1987) to assess concerns about body shape in young western women. It is a 34-item questionnaire scored on a six point Likert scale. A higher score is indicative of disturbances with body image. It shows high internal and external validity (Cooper & Taylor, 1988).

The demographic questionnaire included questions on the subject's experience and goals in her sport. The questions regarding social and peer influence on dieting behavior, were adapted by the researcher from Neumark-Sztainer et al., (1995), a non-athletic study of eating disorders.

Eating disorder symptoms were tested using the Dutch Eating Behavior Questionnaire (DEBQ). The test was developed by van Strien, Frijters, Bergers and Defares (1986) as a measure of eating behaviors. It is a 33 item questionnaire scored on a five point Likert scale with three subscales of eating behavior: restrained eating, emotional eating and external eating. High scores on the restrained eating scale have shown high correlations with subjects who were watching their weight as well as with anorexic and bulimic patients (Wardle, 1987). Bulimic patients scored significantly higher on the external scale than controls or anorexics, with anorexics scoring significantly lower than controls; similarly, bulimic patients scored higher on emotional eating than controls with anorexics scoring lower than controls (Wardle, 1987). The measure has been shown to have high internal consistency and validity (Gorman & Allison, 1995).

Analyses

Data were collapsed across all subjects for a canonical correlation analysis comparing the dependent variables (scores on the restrained, emotional and external eating scales of the DEBQ) to the variables of self-esteem, competition anxiety, social influences, and body image. A second canonical correlation was run between the dependent variables and body fat and BMI. To examine the sport identity questions, subjects were grouped according to the sport in which they participate and a multivariate analysis of variance (MANOVA) was run to see if there were any differences on the scores of the dependent variables. A second MANOVA was run to determine if there were any differences on the dependent variables when the subjects were grouped as team or individual athletes.

Results

Demographics

The average age of the athletes was 19.85 years with an average of 7.33 years of involvement in their sport. The average body fat percentage was 25.68, while the average BMI was 22.35.

Canonical Correlations

The canonical correlation analysis between the dependent variables and the set of risk factors showed one significant correlation (.844, p[less than].05), accounting for 71.3% of the variance. With a cut-off correlation of.4, the eating disorder symptoms that were most correlated with the canonical variate were restrained and emotional eating. Among the set of risk factors body image, self-esteem, competition anxiety and social pressure were all significantly correlated with the canonical variate. The first canonical variate indicated that those athletes who showed high restrained eating (.972) and high emotional eating (.499) also exhibited low body image (.963), high social pressure (.723), low self-esteem (.428), and high competition anxiety (.520) (see Table 1).

The canonical correlation run between the dependent variables and BMI and body fat showed one significant canonical correlation relating the two sets of variables (.433, p[less than].05), accounting for 18.8% of the variance. With a cut-off correlation of .4, restrained eating was the eating disorder symptom most correlated with the canonical variate relating the dependent variables to the physical measurements. Both body fat and BMI were significantly correlated with the canonical variate. The significant canonical variate indicated that those athletes who showed high restrained eating (.855) also had high body fat (.516) and high BMI (.988) (see Table 2).

Individual Regression Analyses

Individual regression analysis showed that restrained eating was significantly predicted by body image, t (45 ) = 4.89,p [less thasn].001 and social pressure, t (45) = 2.746, p [less than] .01. Emotional eating could be significantly predicted by body image, t (45 ) = 2.076, p [less than] .05.

Individual regression analysis showed that BMI was a significant predictor of restrained eating, t (45) = 2.O2,p [less than] .05.

MANOVAS

The two MANOVAs showed no significant effect of sport team membership on the measures of disordered eating.

Discussion

The results of this study showed that body image and social pressure from coaches and peers were significant predictors of restrained eating, supporting previous research. The strong correlation of low body image with eating disorder symptoms, as well as the significant predictive nature of this variable for eating disorder symptoms, supports muchof the other research in this area. Among athletes, Williamson et al. (1995) found that concern with body size was a strong and primary predictor of eating disorder symptoms. Other researchers have consisstrong and primary of eating disorder symptoms. Other researchers have consistently found that body image is a significant predictor of disordered eating in nonathletes (Molinari, 1995; Cooper et al., 1987).

Less research has looked at the role that coaches and peers can play in the development of disordered eating, but the present study supports the available research showing that coaches may play a significant role. For example, Sundgot-Borgen (1994b) reported that a significant number of athletes who were dieting to improve their performance had been told by their coaches to lose weight. This same author makes the point that coaches are often key figures in an athlete's life, particularly for younger athletes, and that the athlete may feel driven to lose weight in order to meet the expectations of the coach. Further, Rosen and Hough (1988) reported that 75% of female gymnasts who were told they were overweight by their coaches began using unhealthy weight control methods. Following this research, Sundgot-Borgen (1994b) found that eating disorders were more prevalent in those athletes who started on an unsupervised dieting program after being told to lose weight by their coach. Other researchers found that eat ing disorder symptomology was significantly predicted by coach and peer influence (Williamson et al., 1995; Rosen & Hough, 1988; Neumark-Sztanier et al., 1996).

It has been speculated that the greater an athlete's body shape deviates from what is considered "ideal" for her sport, the greater the risk of the development of an eating disorder (Wilson & Eldredge, 1992). Supporting this position, the results of the present study indicated that athletes with higher BMIs tended to score higher on the restrained eating scale. It should be noted however, that Davis and Cowles (1989) have reported that athletes with BMI scores in the low unhealthy range reported a desire to lose weight. The present study found a small but significant correlation between high body mass index and low body image that provides support for Wilson and Eldredge's (1992) hypothesis. It may be that athletes with a higher body mass index than their peers, although still within a healthy range, felt greater body dissatisfaction and therefore the tendency to engage in less healthy weight loss behaviors. Again, there is some evidence that coaches may play a role in this area. One study found that coaches tended to make subjective evaluations of their athletes' weight; that is that they made their judgments on appearance rather than more objective indicators (Griffin & Harris, 1996). They also found that coaches tended to rate females as needing to lose weight and males needing to gain weight.

There was also a strong relationship with between self-esteem and eating disorder symptoms in athletes in the present study, supporting much of the previous literature (Lindeman, 1994; Neumark-Sztainer et al., 1996; Davis & Cowles, 1989). Further, it was found that self-esteem was highly correlated with competition anxiety which may be because sport can be a large part of a university athlete's life. However, it should be recognized that other factors unrelated to the sport experience could be major influences on an athlete's mental state. For example, Sundgot-Borgen (1994a) has reported that traumatic events can be a precipitating factor in the development of an eating disorder.

When the relationship between sport team membership and eating disorder symptoms was examined, no significant differences in eating disorder symptomologies were found. This does not preclude the fact, however, that there were individual athletes in this study who scored high on the subscales of the DEBQ, indicating unhealthy eating behavior. Previous researchers have found that there are some sport classes that may have a greater number of athletes who are at a higher risk for developing an eating disorder (Stoutjesdyk & Jevne, 1993; Sudgot-Borgen, 1993; Pasman & Thompson, 1988; Sykoro, Grilo, Wilfley, & Brownell, 1993), this study clarifies the position that potential exists for athletes from a wide range of sports to show symptoms of disordered eating.

Many researchers have identified self-selection of subjects as a problem in the area of research of eating disorders (Shaw & Garfinkel, 1990; Sundgot-Borgen, 1993; Wilmore, 1991). This study was potentially affected in this way because it used volunteers only and it was a possibility that athletes who had an eating disorder or a concern for their eating habits did not choose to participate and those individuals who did volunteer may have been less concerned. However, through interviews with subjects it was clear that a number of the subjects participated because they wanted to find out about their body fat percentage, which in itself may have been indicative of a level of disturbance with body image.

In conclusion, the purpose of this study was to attempt to identify risk factors for the development of eating disorders in athletes. All of the risk factors examined: self-esteem, competition anxiety, social pressure and body image were significantly correlated with indicators of unhealthy diet practices. In addition, indicators of unhealthy dieting practices were found across all sporting groups tested and it is concluded that any athlete should be considered potentially at risk for an eating disorder when some of all of the factors outlined above are present. More directly, it is proposed that there should be a closer examination of the role a coach or peer can play in the development of an eating disorder, as there is some evidence that a coach may play a significant role in these behaviors. Finally, it is strongly recommended that coaches should receive educational programs that stress nutrition, weight control and patterns of disordered eating in female athletes.

References

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author.

Baumgartner, T., & Johnson, A. (1982). Measurement for evaluation in physical education and exercise science (3rd ed.). Dubuque, IA: Wm. C. Brown.

Beals, K. A., & Manore, M. M. (1994). The prevalence and consequences of subclinical eating disorders in female athletes. International Journal of Eating Disorders, 4, 175-195.

Brownell, K. D., Rodin, J., & Wilmore, J. H. (1992). Eating, body weight and performance in athletes. Philadelphia: Lea & Febiger.

Cooper, P. J., & Taylor, M. J. (1988). Body image disturbance in bulimia nervosa. British Journal of Psychiatry, 153,32-36.

Cooper, P. J., Taylor, M. J., M. J., Cooper, Z., & Fairburn, C. G. (1987). The development and validation of the body shape questionnaire. International Journal of Eating Disorders, 6, 485-494.

Cox, R. H., Qiu, Y., & Liu, Z. (1993) Overview of sport psychology. In Singer, R. N., Murphey, M., & Tennant, L. K. (Eds), Handbook of research on sport psychology(pp. 3-31). New York: Macmillan.

Davis, C., & Cowles, M. (1989). A comparison of weight and diet concerns and personality factors among female athletes and non-athletes. Journal of Psychosomatic Research, 33(5), 527-536.

Furst, D. M., & Tenebaum, G. (1984). A correlation of body-cathexis and anxiety in athletes and nonathletes. International Journal of Sport Psychology 15, 160-168.

Gorman, B. S., & Allison, D. B. (1995). Measures of restrained eating. In Allison, D. B. (Ed.), Handbook of assessment methods for eating behaviors and weight-related problems (pp. 149-181). London: Sage Publications.

Griffin, J., & Harris, M. B. (1996). Coaches' attitudes, knowledge, experiences, and recommendations regarding weight control. Sport Psychologist, 10, 180-194.

Lindeman, A. K. (1994). Self-esteem: Its application to eating disorders and athletes. International Journal of Sport Nutrition, 4, 237-252.

Martens, R. (1977). Sport competition anxiety test. Champaign, IL: Human Kinetics.

Martens, S. R. & Simon, J. A. (1976). Comparison of three predictors of state anxiety in competitive situations. Research Quarterly, 47, 381-387.

Molinari, E. (1995). Body-size estimation in anorexia nervosa. Perceptual and Motor Skills, 81, 23-31.

Neumark-Sztainer, D., Beutler, R., & Palti, H. (1996). Personal and socioenvironmental predictors of disordered eating among adolescent females. Journal of Nutrition Education, 28, 195-201.

Pasman, L, & Thomspon, J. K. (1988). Body image and eating disturbance in obligatory runners, obligatory weightlifters, and sedentary individuals. International Journal of Eating Disorders, 7, 759-769.

Rosen, L W., & Hough, D. O. (1988). Pathogenic weight-control behaviours in female college gymnasts. Physician and Sportsmedicine, 16, 141-146.

Shaw, B. F., & Garfinkel, P. E. (1990). Research problems in the eating disorders. International Journal of Eating Disorders, 9, 545-555.

Stoutjesdyk, D., & Jevne, R. (1993). Eating disorders among high performance athletes. Journal of Youth and Adolescence, 22, 271-281.

Sundgot-Borgen, J. (1993). Prevalence of eating disorders in elite female athletes. International Journal of Sport Nutrition, 3, 29-40.

Sundgot-Borgen, J. (1994a). Risk and trigger factors for the development of eating disorders in female elite athletes. Medicine and Science in Sports and Exercise, 26, 414-419.

Sundgot-Borgen, J. (1994b). Eating disorders in female athletes. Sports Medicine, 17, 176-188.

Sykoro, C., Grilo, C. M., Wilfley, D. E., & Brownell, K. D. (1993). Eating, weight, and dieting disturbances in male and female lightweight and heavyweight rowers. International Journal of Eating Disorders, 14, 203-211.

Taub, D. E., & Blinde, E. M. (1992). Eating disorders among adolescent female athletes: influence of athletic participation and sport team membership. Adolescence, 27, 833-847.

Thompson, R. A., & Sherman, R. T. (1993). Helping athletes with eating disorders. Bloomington, IN: Human Kinetics.

van Strien, T., Frijters, J. E. R., Bergers, G. P. A., & Defares, P. B. (1986). Dutch Eating Behavior Questionnaire in normal subjects and women with eating disorders. Journal of Psychosomatic Research, 31, 161-169.

Wardle, J. (1987). Eating style: a validation of the Dutch Eating Behaviour Questionnaire in normal subjects and women with eating disorders. Journal of Psychosomatic Research, 31, 161-169.

Wilkins, J. A., Boland, F. J., & Albinson, J. (1991). A comparison of male and female university athletes and nonathletes on eating disorder indices: are athletes protected? Journal of Sport Behavior, 14, 129-143.

Williamson, D. A., Netemeyer, R. G., Jackman, L. P., Anderson, D. A., Funsch, C. L. & Rabalais, J. Y. (1995). Structural equation modelling of risk factors for the development of eating disorders in female athletes. International Journal of Eating Disorders, 17, 387-393.

Wilmore, J. H. (1991). Eating and weight disorders in the female athlete. International Journal of Sport Nutrition, 1, 104-117.

Wilson, G. T., & Eldredge, K. L. (1992). Pathology and development of eating disorders: implications for athletes. In Brownell, K. D., Rodin, J., & Wilmore, J. H. Eds), Eating body weight, and performance in athletes: Disorders of modern society. Philadelphia: Lea & Febiger.