Disordered eating among collegiate female athletes: The role of athletic seasonal status and self-objectification
List of Tables vi Table of Content
Abstract vii Chapter I: Introduction 1 Lean Versus Non-Lean Sports 3 Differences within Lean Sport Groups 4 Seasonal Status 6 The Present Study 8 Chapter II: Review of the Literature 9 Prevalence of Eating Disorders 9 Eating Disorders in Athletes 11 Lean Versus Non-Lean Sport Athletes 13 Differences among Lean Sport Athletes 18 Identification of “At-Risk” Athletes: The Role of Seasonal Status 22 Objectification Theory 26 Summary 32 Chapter III: Method 34 Design 34 Participants 34 Measures 37 Eating Disorder Diagnostic Scale 37 ATHLETE Questionnaire 39 Eating Attitudes Test 40
Self-Objectification Questionnaire 41 Demographics Questionnaire 42 Procedure 43 Statistics and Data Analysis 45 Chapter IV: Results 47 Preliminary Analyses 47 ANOVA Analyses 49 In-Season Sport Differences 49 Off-Season Sport Differences 50 Sport by Season Interactions 51 Regression Analyses 53 Self-Objectification as a Mediator of Sport Type and Disordered Eating 53 Chapter V: Discussion 55 Discussion of Results 56 Differences Between Sports: General Eating Disorder Symptoms and Diagnoses 56 Differences between Sports: Athlete-Specific Eating Disorder Symptoms 56 Seasonal Status 59 Clinical Implications 60 Limitations 63 Future Research 65
References 67 Appendix 76
List of Tables Table 1. In-Season Sport Differences ANOVA Summary Table Table 2. Means by Sport In-season Data Summary Table Table 3. Off-season Sport Differences ANOVA Summary Table Table 4. Means by Sport Off-season Data Summary Table Table 5. EAT Season by Sport Two-Way ANOVA Summary Table Table 6. ATHLETE Season by Sport Two-Way ANOVA Summary Table Table 7. EDDS Season by Sport Two-Way ANOVA Summary Table
ABSTRACT Over the past several decades there has been a significant increase in attention to the eating related beliefs and behaviors of female college athletes, particularly in determining whether certain subgroups of athletes are at greater risk than others. At seemingly greatest risk for eating disorders are athletes involved in sports where leanness is emphasized or a thin physique is required for performance or aesthetics. However, it remains unclear if differences exist between aesthetic lean and non-aesthetic lean sport athletes. It is possible that seasonal status is associated with the transience of eating disorder symptoms and the motivation to engage in eating disordered behaviors. The purpose of this study is to examine the differences in eating disorder symptoms and diagnoses across three groups of athletes (i.e., aesthetic lean, non-aesthetic lean, and non-lean). Second, this study examined changes in eating disorder symptoms and diagnoses across athletic seasonal status. Finally, level of self-objectification was expected to mediate these between-sport relationships. Participants were 282 female athletes from 14 American universities. Participants completed measures of eating disorder diagnoses, general and athlete-specific eating disorder symptoms, and self-objectification in and out of the athletic season. In the test of sport differences, results revealed that non-aesthetic lean sport athletes reported fewer athlete-specific eating disorder symptoms than the other two groups during the athletic season. Further, non-lean sport athletes reported more athlete-specific eating disorder symptoms at both time points. When examining the effect of seasonal status, athletes reported fewer eating disorder symptoms during the athletic season than in the off- season.
These results have important clinical implications. Because of the expectation that lean individuals are at greatest risk for the development of eating disorders, non-lean sport athletes may not be receiving the attention and treatment needed to improve their eating attitudes and behaviors. For lean-sport athletes, it is important for clinicians and coaching staff members to understand the drive and methods for obtaining and maintaining a low body weight. Finally, increases in disordered eating during the off- season could have significant ramifications, specifically that disordered eating behaviors and eating disorders may be overlooked due to the timing of the screening and coaching staff supervision.
Chapter I Introduction Eating disorders represent an important risk to the mental and physical health of college-aged women in the United States. Although epidemiological studies have found that anorexia affects .5% and bulimia affects 1-3% of the female population in general (American Psychiatric Association, 2000), rates among college students appear higher. In fact, prevalence rates of clinical eating disorders in college-age women range from 1.0% to 4.2% for anorexia and 1.0% to 6.5% for bulimia (Mintz, O’Halloran, Mulholland, & Schneider, 1997; Pope, Hudson, Yurgelun-Todd, & Hudson, 1984; Pyle, Halvorson, Neuman, Mitchell, 1986; Striegel-Moore et al., 2003). An even greater percentage (60%) of college women report engaging in disordered eating behaviors i , such as chronic dieting and binge eating (Mintz & Bentz, 1988). In fact, in a sample of high school and college females, many participants reported skipping meals (59%), eating less than 1200 calories a day (37%), eliminating fats (30%) and carbohydrates (26.5%) from their diets, and fasting for more than 24 hours (26%) (Tylka & Subich, 2002). Although the majority of dieters will not develop an eating disorder, a study of first-year college women found that 15% of those identified as “at risk” in the fall semester moved into the probable bulimia category by the spring semester (Drewnowski, Yee, Kurth, & Krahn, 1994). Over the past several decades there has been a significant increase in attention to the eating related beliefs and behaviors of female college athletes. Athletes represent a unique population with regards to disordered eating because they experience a greater need to maintain a healthy lifestyle due to the physical demands of their sport. However, many athletes engage in disordered eating behaviors, often in order to maintain a specific 1
weight. A number of studies have demonstrated that athletes are at an increased risk for the development of eating disorders, including anorexia nervosa, bulimia nervosa, and eating disorders not otherwise specified (ED-NOS) (Smolak, Murnen & Ruble, 2000; Sundgot-Borgen & Larsen, 1993). Specifically, research examining the prevalence of eating disorders among female athletes has found the pervasiveness of anorexia and bulimia to be as high as 2.9% and 9.2%, respectively (Johnson, Powers, & Dick, 1999). In one of the most rigorous studies to date, Sundgot-Borgen (1993) surveyed 522 elite female athletes in the Norwegian Confederation of Sports using a combined self-report and interview format. Results indicated that a significantly higher number of athletes (18%) than non-athlete controls (5%) had suffered from an eating disorder. Several studies indicate that athletes who do not meet criteria for a diagnosable eating disorder still tend to score higher on various measures of eating disorder symptomatology than non-athletes. For example, Sundgot-Borgen and Larsen (1993) examined the use of pathogenic weight-control methods among female elite athletes and non-athletic controls and found that significantly more athletes (11%) than controls (7%) used pathogenic weight-control methods (e.g., laxative use, diet pills, and dieting). Athletes also reported different reasons for dieting than the controls. While the controls reported dieting to improve their appearance, the athletes generally reported dieting to enhance performance. Further, dieting athletes were found to be more likely than dieting controls to use pathogenic weight-control methods. Similarly, Rosen, McKeag, Hough, and Curley (1986) examined 182 female collegiate athletes to determine the prevalence of pathogenic weight control methods, and found that 32% of participants practiced at least one pathogenic weight control behavior on a daily basis. More specifically, regular
use of diet pills occurred in 25% of the participants and 14% and 16% engaged in self- induced vomiting and laxative abuse, respectively. Although a good deal of research supports the idea that athletes are at a greater risk than non-athletes for the development of an eating disorder, this research is not unequivocal. For example, Marten, DiBartolo, and Shaffer (2002) found that female athletes had fewer eating disorder symptoms and less body image disturbance than non- athletes, while Sanford-Martens, Davidson, Yakushko, and Martens (2005) found no differences between athletes and non-athletes in terms of likelihood of experiencing a clinical or “subclinical” eating disorder. Perhaps the most authoritative research on the topic comes from a meta-analysis of 92 studies comparing eating disorder symptoms between athletes and non-athletes (Hausenblas & Carron, 1999). Results of this study revealed significant, although small, effect sizes indicating that female athletes self- reported more bulimic (ĝ = .16) and anorexic (ĝ = .12) symptomatology than non- athletes. However, female athletes did not demonstrate a greater drive for thinness than non-athletes, a core component of eating disorders. Therefore, although female athletes in general are at greater risk than non-athletes for various eating disordered behaviors, the magnitude of the risk may be relatively small. Lean Versus Non-Lean Sports In addition to the issue of whether athletes are at higher risk for the development of eating disorders than non-athletes is the question of whether certain subgroups of athletes are at greater risk than others. At seemingly greatest risk for eating disorders are athletes involved in sports where leanness is emphasized or a thin physique is required for performance or aesthetics (Books-Gunn, Burrow, & Warren, 1988; Davidson,
Earnest, & Birch, 2002; Garner, Olmsted, Polivy, & Garfinkel, 1987). For example, Sundgot-Borgen and Larsen (1993) found that, overall, a similar proportion of female athletes and non-athletes were classified as being at risk of developing eating disorders. However, when examining only the athletes involved in sports in which leanness or a specific weight is considered important, a higher proportion of athletes than non-athletes were considered to be at risk of developing an eating disorder. At least two meta-analyses have examined the specific differences in eating disordered behaviors among athletes participating in different types of sports. Hausenblas and Carron (1999) examined the differences between aesthetic (e.g., figure skating), endurance (e.g., running), and ball-game (e.g., soccer) sports and found that athletes in aesthetic sports reported greater drive for thinness (ĝ = 0.09) than ball-game athletes and greater anorexic symptomatology (ĝ = 0.38) than both endurance athletes (ĝ = -0.04) and ball-game athletes (ĝ = -0.17). However, the groups did not differ on measures of bulimic symptomatology. Smolak, Murnen, and Ruble (2000) also conducted a meta-analysis of 34 studies examining the relationship between female athletes and eating problems. When compared to non-athletes, elite athletes who participated in lean sports (i.e., sports where a low body weight is important for athletic success; e.g., gymnastics, running) were found to be at increased risk for the development of eating problems (d = .52). The authors concluded that athletes participating in lean sports, dancers, and athletes competing at an elite level were most at risk for the development of an eating disorder. Differences within Lean Sport Groups Although evidence exists to suggest that lean sport athletes may be at greater risk for the development eating disorders than non-lean sport athletes, it remains unclear if
differences exist between different types of lean sport athletes. Specifically, the eating beliefs and behaviors of athletes involved in lean sports in which low body weight is deemed central to performance for aesthetic purposes (aesthetic lean sports: e.g., figure skating) may differ in some fundamental ways from athletes involved in lean sports in which low body weight is deemed central for performance purposes (non-aesthetic lean sports: e.g., cross-county running). While athletes in both sport types are under pressure to maintain a low body weight, aesthetic lean sports have unique elements in that they tend to involve subjective judgment of others to determine athletic success and tend to require tight fitting, feminine attire. Current research that has examined these groups separately has resulted in inconsistent findings. Hulley and Hill (2001) surveyed elite female distance runners and found that runners were at increased risk for eating disorders when compared to national norms. However, Warren, Stanton, and Blessing (1990) examined competitive female athletes to identify risk of disordered eating patterns and found that competitive female cross-country runners were at slightly less risk for body dissatisfaction than the non-athletes, whereas competitive female gymnasts were at greater risk for weight preoccupation. One factor that may lead to differences between these two groups of athletes is the level of internalization of messages related to low weight demands and body objectification. Fredrickson and Roberts (1997) proposed objectification theory to understand the consequences of being female in a society that sexually objectifies the female body. This theory posits that as a result of the pervasiveness of sexual objectification in our culture, women internalize these objectifying messages and, at some level, treat themselves as objects to be looked at and evaluated. This theory
maintains that as the internalization of objectifying messages increase, so too does one’s risk of developing depression, sexual dysfunction, and eating disorders. Sexual objectification occurs when a woman’s body is separated from her identity as a person and reduced to the status of a mere instrument (Fredrickson & Roberts, 1997). Although objectification theory assumes that all women internalize this objectification of their bodies at some level, the extent to which women self-objectify is largely context dependent and some groups are at greater risk than others (Fredrickson & Roberts, 1997). Of particular interest to the current study is the difference in level of self-objectification in different groups of lean sport athletes. While non-aesthetic lean sport athletes tend to experience pressure to be thin to enhance objective performance (e.g., a higher strength to weight ratio increases speed in endurance runners), aesthetic lean sport athletes tend to experience pressure to be thin to gain the subjective approval of others (e.g., judges). According to self-objectification theory, aesthetic lean sport athletes likely have greater levels of self-objectification and may be at greater risk for engaging in disordered eating behaviors. Seasonal Status Another unanswered question regarding eating disordered behaviors and lean sport athletes is the effect of seasonal status (i.e., in- versus off-season). It is possible that seasonal status is associated with the transience of eating disorder symptoms and the motivation to engage in eating disordered behaviors. Johnson and Tobin (1991) discuss the importance of considering the context and pattern of behavior before determining whether the behavior is pathological. Among athletes the training demands associated with seasonal status may be an important contextual variable. Theoretically, those who
meet clinical diagnostic criteria for an eating disorder should not experience transient symptoms across seasonal status. In contrast, some athletes may engage in various eating disordered behaviors during their competitive season due to the environmental pressures associated with their sport, but such symptoms may dissipate during the off-season. To date, minimal research has been done examining the influence of seasonal status on disordered eating behaviors of athletes. One study by Dale and Landers (1999) began to address this issue. They acknowledged the behavioral similarities between wrestlers and individuals suffering from bulimia and examined whether an increased risk of bulimia existed for wrestlers. The Eating Disorder Inventory (EDI) was administered to a group of junior high and high school wrestlers both in and out of the athletic season and to a group of non-wrestlers. Results indicated that there were no significant differences between the number of in-season wrestlers and non-wrestlers classified as at- risk for bulimia. However, significant differences did exist on the Drive for Thinness subscale between in-season wrestlers and non-wrestlers, and between in-season wrestlers and off-season wrestlers. Although a significant difference was demonstrated between in- season wrestlers and non-wrestlers, the same wrestlers tested off-season did not differ significantly from non-wrestlers. The authors concluded that the concerns of the wrestlers are transient, which is not characteristic of an eating disorder. Other related literature has demonstrated the relationship between seasonal status and health-related behaviors among intercollegiate athletes. For example, Martens, Dams-O’Connor, and Duffy- Paiement (2006) assessed for off- versus in-season differences in alcohol consumption among a sample of intercollegiate athletes and found that college athletes’ drinking and negative alcohol-related consequences decreased during their competitive season.
It is proposed that a significant increase in disordered eating behaviors during the athlete’s competitive season would not correlate with a significant increase in the formal diagnoses of eating disorders among all athletic sport types. This change of eating disorder symptoms with seasonal status would support the notion that these unhealthy eating behaviors are not indicative of an eating disorder, but rather are risks primarily to the athletes’ physical health. Understanding the risks associated with various sport types to physical and mental health has significant implications for prevention and treatment issues. The Present Study The main purpose of the present study is twofold. First, this study will examine the differences in eating disorder symptoms and diagnoses across three groups of athletes (i.e., aesthetic lean sport athletes, non-aesthetic lean sport athletes, and non-lean sport athletes). It is hypothesized that the proportion of aesthetic lean sport athletes who meet criteria for eating disorder diagnoses will be greater than non-aesthetic lean sport athletes or the comparison group. It is also hypothesized that eating disorder symptoms will be greater for both groups of lean sports than for the control group. Second, this study will examine changes in eating disorder symptoms and diagnoses across athletic seasonal status. It is hypothesized that the incidence of eating disorder diagnoses will remain stable regardless of seasonal status, but that eating disorder symptoms will increase in the athletic season. It is also hypothesized that increases in symptoms will be greater for non- aesthetic lean sport athletes and the comparison group than for aesthetic lean sport athletes. Level of self-objectification is expected to mediate these between-sport relationships.
Chapter II Review of the Literature In order to provide a background and context for the proposed study, this chapter will review the literature relevant to eating disorders and athletes. First, the prevalence and significance of eating disorders will be discussed, followed by a review of the literature examining female college athletes as a population at increased risk. Next, a brief review of the differences between various subgroups of athletes will be presented, with subsections focusing on lean vs. non-lean sport differences and differences within lean sports athletes. Differences between athletes groups will be explored in two ways. Existing literature on the effect of athletic seasonal status on changes in athlete health behaviors will be presented. Finally, objectification theory will be offered as a possible explanation for the proposed differences in eating disorder symptoms and diagnoses among athletes. Relevant empirical support will be presented. Prevalence of Eating Disorders Eating disorders, such as anorexia nervosa and bulimia, have been found to affect .5% and 1-3% of the female population, respectively, in the United States (American Psychiatric Association, 2000). The average age of onset for these disorders is approximately 18 years old, with women representing 90% of the diagnoses (American Psychiatric Association, 2000). Eating disorders are most common in college-educated women from middle- and upper-class families, and the prevalence has been found to be directly related to the degree of westernization of the country (Johnson & Tobin, 1991). This suggests that the disorders are highly affected by socio-cultural factors. College women may be particularly susceptible to the development of these disorders, given that
entrance into college is a time of adjustment to a new environment with less adult guidance and higher academic demands (Kirk, Singh, & Getz, 2001). Numerous studies have supported the notion that college-aged women may be at increased risk for the development of eating disorders. Epidemiological investigations reporting prevalence rates of clinical eating disorders in college-age women range from 1.0% to 4.2% for anorexia and 1.0% to 6.5% for bulimia (Mintz, O’Halloran, Mulholland, & Schneider, 1997; Pope, Hudson, Yurgelun-Todd, & Hudson, 1984; Pyle, Halvorson, Neuman, Mitchell, 1986; Striegel-Moore et al., 2003). A much higher percentage of college women do not meet criteria for eating disorders, but report engaging in disordered eating behaviors (e.g., 61%; Mintz & Bentz, 1988). Rates for fasting, binge eating, purging, and use of diet pills, laxatives, and diuretics to lose weight have been found to range from 1.4% for laxative use to 26% for fasting in college women (French & Jeffery, 1994; Pyle, et al., 1986; Tylka & Subich, 2002). The relatively common use of disordered eating behaviors appears to represent a somewhat normative discontent with body weight and shape in the female college population (Thompson, 1990). That is, women’s dissatisfaction with their bodies is so widespread that frequent dieting, drive for thinness, and fear of fat have become a relatively typical response for adolescent and young adult women (Johnson & Tobin, 1991; Markey & Markey, 2005). Although the majority of dieters will not pass the threshold to develop a clinical eating disorder, Drewnowski, Yee, Kurth, and Krahn (1994) found that 15% of those college-aged women identified as “at-risk” in the fall semester moved into the probable bulimia category by the spring semester. Although female college students in general seem to be at risk for eating disorder symptoms,
research has shown that certain subgroups are at even greater risk. Specifically, it has been hypothesized that participation in athletics further increases the risk for the development of eating disorders. Eating Disorders in Athletes Female collegiate athletes represent a unique population with regards to disordered eating. Although they experience a greater need to maintain a healthy lifestyle due to the physical demands of their sport, athletes often feel significant pressure to strive for low body weight, putting them at greater risk for engaging in disordered eating behaviors (Kirk et al., 2001). Athletes may experience explicit and implicit pressure from coaches, parents, fans, or teammates to maintain a low weight. A significant amount of research demonstrates that athletes are at an increased risk for the development of eating disorders (Smolak, Murnen & Ruble, 2000; Sundgot-Borgen & Larsen, 1993). Studies examining prevalence rates of eating disorder diagnoses in athletes have found that athletes are at increased risk for clinical eating disorders. For example, with the use of self-report surveys, Johnson, Powers, and Dick (1999) found 2.9% and 9.2% of female athletes were identified as having clinically significant problems with anorexia and bulimia, respectively. Other research examining the prevalence of eating disorders among female athletes has found the pervasiveness of anorexia and bulimia in female college athletes to be as high as 4.2% and 39.2%, respectively (Burckes-Miller & Black, 1988). Sundgot-Borgen (1993) proposed that the prevalence of eating disorder diagnoses in the athlete population may be inaccurate due to the self-report nature of the measures. Utilizing a clinical interview in a sample of 522 athletes and 448 non-athletes, 18% of
athletes and 5% of non-athletes were found to be suffering from a diagnosable eating disorder. Research indicates that athletes who do not meet criteria for a diagnosable eating disorder still tend to engage in more disordered eating behaviors than non-athletes. For example, several studies have found that athletes diet more than non-athletes, more frequently use diet pills, laxatives, and diuretics, and more frequently engage in binge eating and purging (Davis & Cowles, 1989; Johnson et al., 1999; Sundgot-Borgen & Larsen, 1993). Use of these weight loss methods may indicate an increased risk for the development of an eating disorder (Fitzgibbon, Sánchez-Johnsen, & Martinovich, 2003) Although a good deal of research supports the idea that athletes are at greater risk than non-athletes for the development of eating disorders, this research is not unequivocal. Several studies have found no differences between athletes and non-athletes in clinical eating disorders (e.g., Fulkerson, Keel, Leon, & Door, 1999; Sanford-Martens et al., 2005), while others suggest that athletes have less eating disorder symptomatology and healthier psychological functioning than non-athletes (e.g., DiBartolo & Shaffer, 2002; Kurtzman, Yager, Landsverk, Wiesmeier, & Bodurka, 1989). Perhaps the most authoritative research on the topic comes from two meta- analyses, which provide some clarity of the inconsistent literature. Hausenblas and Carron (1999) conducted a meta-analysis of 92 studies with 560 effect sizes comparing eating disorder symptoms between athletes and non-athletes. Results of this study revealed significant, although small, effect sizes indicating that athletes report more eating disorder symptoms than non-athletes. Specifically, female athletes self-reported more bulimic (ĝ = .16) and anorexic (ĝ = .12) symptomatology than non-athletes.
However, female athletes did not demonstrate a greater drive for thinness, a core component of eating disorders. One explanation for this finding is that due to current societal pressures for women to meet an idealized standard of beauty, both female athletes and non-athletes experience a perpetual feeling of discontent with regards to their body, normalizing a drive for thinness (Cooley & Toray, 2001). Smolak et al. (2000) conducted a meta-analysis of 34 studies examining the overall relationship between female athletes and eating problems and found similar results. A small effect was found, demonstrating that athletes reported more eating problems than non-athletes (d = .07, p < .01). When examining the effects of competition level, however, differences were greater among higher competitive levels, with college student athletes reporting significantly greater eating problems than non-athletes (d = .15, p < .001). Lean versus Non-Lean Sport Athletes As research indicates that, in general, athletes are at greater risk for the development of eating disorders and experiencing disordered eating behaviors than non- athletes, it is logical to assess potential differences among various subgroups of athletes. Although a good deal of research has been conducted examining individual sports, much of the existing literature has directly compared different subgroups of athletes. One major challenge in this type of research, which contributes to the inconsistency of findings, is developing a classification system for grouping athletes. At seemingly greatest risk for eating disorders are those sports where leanness is emphasized or a thin physique is required for performance or aesthetic purposes (Books-Gunn, Burrow, & Warren, 1988; Davidson, Earnest, & Birch, 2002; Garner, Olmsted, Polivy, & Garfinkel, 1987). As
such, a common classification system for studying eating disorders in athletes is to compare athletes involved in lean sports to athletes involved in non-lean sports. However, there has been an inconsistent use of categorical labels across such studies, such as lean versus nonlean, thin body build versus normal build, and fostering the attainment of a thin physique versus low weight not as central (Hausenblas & Carron, 2002). The lack of consistency in terminology across studies has led to a continued lack of clarity in differences between sport groups. In spite of these variations in categorical labels, researchers have examined sport type differences. However, along with the inconsistency of labels come differences in how sports are classified in terms of eating disorder risk. For example, Petrie (1996) examined the differences among three groups: two groups of athletes - one group participated in “lean sports” (i.e., swimming, diving, cross-country, wrestling, and gymnastics) and the other group participated in “non-lean sports” (i.e., volleyball, softball, fencing, tennis, basketball, golf, rifle/pistol, track and field, and field hockey) and a non-athletic control group. Petrie found non-significant differences between the three groups on EDI subscales of Bulimia, Perfectionism, Interpersonal Distrust, and Maturity Fears. However, results suggested that lean sport athletes had higher Drive for Thinness scores than the non-lean sport athletes and non-athletes. Although these findings appear straightforward, comparing results across studies fails to provide clarity and consistency. In a seemingly comparable study, Sundgot-Borgen and Corbin (1987) examined the extent to which female athletes are preoccupied with weight and are at-risk for eating disorders, as measured by the EDI. They too examined three groups: two groups of
female athletes - one group (n = 35) participated in sports that “emphasized leanness” (i.e., ballet dancers, body builders/weight trainers, cheerleaders, and gymnasts) a second group (n = 32) participated in sports that “did not emphasize leanness” (i.e., swimming, track and field, and volleyball), and a non-athletic control group (n = 101). It is important to note that while Petrie (1996) identified swimming as a sport that emphasizes leanness, Sundgot-Borgen and Corbin (1987) identified it as one that does not emphasize leanness. Data from this study were compared to mean EDI scores for individuals diagnosed with anorexia. Results suggested that although no differences existed between athletes and non-athletes in the total number of EDI subscale scores above the means for individuals with anorexia, differences did exist between athlete groups when comparing the total number of subscale scores above that mean. The athletes in sports that emphasized leanness (21%) had more subscale scores above the means of individuals diagnosed with anorexia than did those in sports not emphasizing leanness (11%; χ 2 (1) = 11.65, p < .05). As demonstrated above, a major limitation in the literature examining the differences between sport types is that classification systems are not consistent across studies (Hassenblas & Carron, 2002). This not only creates difficulty in comparing studies, but likely fails to ascertain meaningful differences in causes of disordered eating. For example, in the common classification system of lean versus non-lean sport athletes, athletes involved in wrestling, figure skating, and distance-running are often categorized as one, without regard for the fact that the underlying reasons for disordered eating behavior may differ considerably across sports. Although wrestlers are under pressure to maintain a specific weight class, figure skaters are under pressure to maintain a low