The issue of eating disorders has become a growing concern in athletics. In the past few years the question has been raised as to whether adolescent athletes who participate in sports in which weight plays a crucial role may in fact exhibit clinical symptoms of eating disorders such as anorexia nervosa and bulimia nervosa (12,17,29). The research does not provide a clear answer.
One sport in which concern has been raised regarding eating disorders (specifically the eating disorder bulimia nervosa) is wrestling. Many techniques wrestlers use to lose weight—such as self-induced vomiting, laxative and diuretic misuse, excessive exercise, and fasting—are symptomatic of bulimia nervosa. Bulimia nervosa is a clinically diagnosable syndrome, with specific diagnostic criteria having been established by the American Psychiatric Association (3). The estimated prevalence of bulimia nervosa in the general population is 1–3% (5), and the estimated prevalence for adolescent boys and young adult men is approximately 0.2% (8).
It is well-established that wrestlers often employ questionable weight-loss techniques to make weight (1,2,19,26,29,31). These behaviors have been a cause for concern for many years. Position statements have been issued by both the American Medical Association Committee on the Medical Aspects of Sports (2) and the American College of Sports Medicine (1) discouraging the food and fluid practices used by wrestlers. These weight-loss methods include vomiting after eating, the use of laxatives, diuretics, and diet pills, fluid restriction, dehydration, sweating, and fasting.
Several recent studies have pointed out that these weight loss behaviors are similar to the behaviors of bulimics (13,22,25,30,34). Oppliger et al. (25) surveyed 713 high school wrestlers and identified 1.7% as answering questions consistent with all five DSM-III-R criteria for bulimia nervosa. Lakin et al. (22) surveyed 716 wrestlers and classified 1.4% as bulimic. Woods et al. (34) questioned 49 high school wrestlers and reported that 8% used vomiting as a method of weight control and 27% binged during the season. Thiel et al. (30) investigated 25 wrestlers using the Eating Disorder Inventory (EDI) and found that 16% of the wrestlers had pathologic EDI profiles. In general, these studies have focused on the existence of eating disordered behaviors and have not fully examined the psychological aspects of bulimia nervosa. In fact, only two studies (22,25) have based their results on DSM-III-R diagnostic criteria for bulimia nervosa. It seems clear that many wrestlers engage in a variety of bulimic-like behaviors. However, it is not clear whether these wrestlers suffer from the psychological correlates of bulimia. There is no question that standards need to be in place to ensure that wrestlers are losing weight safely and not creating threats to their physical health. It now needs to be determined whether wrestlers are creating threats to their mental health as well.
It is important to remember that bulimia nervosa is a psychiatric disorder; it is not merely a pattern of disordered eating. Johnson and Tobin (20) discuss the importance of considering the context and pattern of behavior before determining whether the behavior is pathological. For example, they point out that eating a large amount of food may not be classified as binge eating unless there is a sense of loss of control during the eating episode or the eating behavior is being used to manage difficult emotions. Some athletes, such as football players, may simply eat large amounts of food because they are trying to gain weight. They also point out that the more definite a pattern of purging is (e.g., the athlete who purges only before meets), the less likely it is to be evidence of an eating disorder. Beals and Manore (4) state that “individuals diagnosed with clinical eating disorders of anorexia nervosa and bulimia nervosa exhibit severe emotional distress and/or specific psychopathologies that go beyond concerns about weight and the use of weight reduction methods.” King (21) studied eating disorders in a general practice population. He remarks that it is extremely important to “distinguish between behaviors (no matter how extreme) aimed principally at maintaining a low weight for reasons such as a vocation, and the central psychopathology of an eating disorder.” He stresses the importance of determining whether the athlete’s abnormal eating is a transient behavior associated with the specific demands of the sport or is actually symptomatic of an eating disorder. These views suggest that it is not enough to merely document behaviors: the emotional and psychological state of the athlete must also be examined.
The problem of bulimia nervosa and wrestling is clearly not resolved. The research points to three major issues. First, studies of wrestling and bulimia nervosa have not gone beyond the administration of eating disorder questionnaires. Several researchers have pointed out the limitations of eating disorder questionnaires and the need to use clinical interviews to determine more accurate prevalence data (15,16,22,25). Lakin et al. (22) states that additional studies are needed “… to substantiate diagnoses of bulimia by clinical interviews.” Second, many of the questionnaires have assessed only physical behaviors and ignored the accompanying psychological problems. The mere fact that bulimia nervosa is included in the DSM-IV makes it obvious that it is a mental disorder and must be examined as such. It is not enough to merely document the presence of inappropriate weight loss techniques. Finally, it has been suggested that wrestlers’ eating patterns may be transient behaviors that are present only during the season and are therefore not truly indicative of bulimia nervosa. Research is needed that examines wrestlers’ behaviors both during the season and in the off-season.
This study will attempt to address each of these research concerns. Is the problem one of eating disorders or disordered eating? The key to this study is the use of a follow-up interview. This interview provides prevalence data that is more accurate than data based solely on eating disorder questionnaires (33). Another important aspect of this study is the use of a standardized eating disorder questionnaire (the EDI) (18) to assess the number of subjects “at risk” for bulimia nervosa. Using a standardized questionnaire allows for comparisons with normative data and also makes it possible for replication in future research. Only two previous studies (13,30) used standardized questionnaires for which normative data was available. The third key aspect is the inclusion of nonwrestlers as a comparison group. The two studies using DSM-III-R criteria to determine risk for bulimia did not use a comparison group (22,25). This makes it difficult to make statements about the differences between wrestlers and nonwrestlers. The final crucial condition of the study is the retesting of the wrestlers. Research examining eating patterns of wrestlers in the off-season is sparse (23,24,29). None of these studies make an attempt to classify wrestlers as “at risk” according to the DSM-III-R or DSM-IV criteria. It is extremely important to determine whether wrestlers’ behaviors are transient; bulimia nervosa is not characterized by transient behaviors.
Two groups of subjects were used for this study. One group consisted of 85 male junior high and high school wrestlers in the Phoenix area. The wrestlers were recruited from local high school and junior high wrestling teams. Twenty-nine of the subjects were high school wrestlers; the remaining 56 subjects were junior high wrestlers. The average weight of the wrestlers was 136.7 pounds, and the average age was 15 yr. The other group consisted of 75 male nonwrestlers; this group was recruited from the same local high schools and junior high schools. The nonwrestlers were used as a comparison group. Twenty-nine of the nonwrestlers were high school students, the remaining 46 nonwrestlers were junior high students. Approximately 57% of the nonwrestlers were athletes who participated in a sport other than wrestling. These sports included track and field, soccer, basketball, football, and baseball. The average weight of the nonwrestlers was 147 pounds, and the average age was 15 yr. All subjects were male and between 12 and 18 yr of age. Subjects signed informed consent forms before participating. Due to the age of the subjects, parental consent was also obtained. The study was also examined and subsequently approved by the Human Subjects Review Board.
The EDI (18) was used as a screening instrument to identify subjects at risk for bulimia nervosa. The questionnaire consists of 64 items and contains eight subscales. Each item is presented in a 6-point format requiring subjects to answer whether each item applies “always,” “usually,” “often,” “sometimes,” “rarely,” or “never.” The first three subscales (Drive for Thinness, Bulimia, and Body Dissatisfaction) examine attitudes and behaviors relating to body image, eating, and dieting. It is these three subscales that contain questions that address the five DSM-IV diagnostic criteria for bulimia nervosa. The remaining five subscales (Interpersonal Distrust, Perfectionism, Interoceptive Awareness, Maturity Fears, and Ineffectiveness) measure psychological traits that have been identified as key aspects of the psychopathology of anorexia nervosa. The questions included in these subscales do not address the DSM-IV diagnostic criteria for bulimia nervosa, and therefore were not included in the analysis. The EDI has been shown to have good reliability and validity (14,18). Garner (14) recommends administering the test in its entirety to preserve this reliability and validity. Therefore, all subjects completed the questionnaire in its entirety, but only the first three subscales were used for analysis.
The Eating Disorder Examination (EDE) (9) is a 62-item semistructured clinical interview and was used to interview those clients identified as “at risk.” The interview is designed to assess the current state of the patient; therefore, all questions refer to the previous 4 wk. The examination has been shown to have adequate reliability and validity (9,10). The goals of the authors (9) were to maximize reliability and create an instrument that was more sensitive than a self-report questionnaire. At least one study (33) found the EDE to be a superior measure of the core pathology of bulimia nervosa. It was used in the present study as the second stage of a two-stage screening process, an approach that is recommended by Garner (14).
Parental consent letters were distributed by teachers to all students in school physical education classes or by coaches to all members of the wrestling teams. Students who returned parental permissions were requested to complete an EDI (18). Wrestlers and nonwrestlers who scored above specific cutoffs were asked to participate in an interview. To preserve confidentiality, coaches and teachers were told that a random sample of subjects would be asked to participate in an interview; they were not aware that only subjects scoring above the cutoffs were interviewed or even that these cutoffs existed. The interview followed the format of the EDE (9). The lead author conducted the interviews, and a psychologist specializing in eating disorders reviewed notes from all “positive response” interviews and corroborated the results. All subjects participating in the interview received an informational pamphlet on eating disorders.
The wrestlers completed the questionnaire twice. The first time the questionnaire was administered was during the wrestlers’ competitive season. The questionnaires were administered within 3 d before competition. The second time the questionnaire was administered was 4–6 wk after the conclusion of the season. This was considered the off-season. The in-season wrestlers and the off-season wrestlers were the same group. Only two wrestlers did not participate in the off-season testing; both had left the school and could not be reached. The same procedure was followed both during the season and in the off-season. The control group completed the questionnaire only once, as they had no off-season or in-season.
Cutoff scores specific to the three subscales (Drive for Thinness, Bulimia, and Body Dissatisfaction) and based on normative data (14,27) were used to identify the number of subjects considered to be “at risk.” Rosen et al. (27) established norms for adolescent girls and boys. The norms for boys were replicated by Shore and Porter (28). Subjects scoring above a 3 on the Drive for Thinness subscale, above a 4 for the Bulimia subscale, or above a 10 on the Body Dissatisfaction subscale were classified as “at risk” and asked to participate in an interview.
Three primary comparisons were made between groups: in-season wrestlers versus nonwrestlers, off-season wrestlers versus nonwrestlers, and in-season wrestlers versus off-season wrestlers. A stem-and-leaf diagram of the data revealed that the data were not normally distributed for any of the groups. This was due to a high number of zero scores on the subscales. Therefore, a chi-square nonparametric test was used to determine differences for the comparisons between in-season wrestlers and nonwrestlers, and between off-season wrestlers and nonwrestlers. A McNemar nonparametric test was used to compare in-season wrestlers with off-season wrestlers. This test is appropriate for comparing related samples, whereas the chi-square is not.
Each group was broken down into four categories for comparison. The first category was the total number of subjects classified as “at risk.” In other words, the subjects within each group who were above the cutoff on any one of the three targeted subscales (Drive for Thinness, Bulimia, or Body Dissatisfaction). Each subscale was then examined separately. The next three categories were the number of subjects above the preset cutoff on the Drive for Thinness subscale, the number of subjects above the preset cutoff on the Bulimia subscale, and the number of subjects above the preset cutoff on the Body Dissatisfaction subscale.
The data were also compared with norms established by Rosen et al. (27) and Shore and Porter (28). Each of the three groups (in-season wrestlers, nonwrestlers, and off-season wrestlers) was compared to established norms for adolescent boys on each of the three subscales—Drive for Thinness, Bulimia, and Body Dissatisfaction.
To determine how much overlap there was between the subscales in terms of “at risk” classifications, the data were examined for high scores on multiple subscales. For each of the three groups (in-season wrestlers, nonwrestlers, and off-season wrestlers), it was determined how many subjects within each group scored above the cutoff on two of the three subscales and how many scored above the cutoff on all three of the subscales.
The majority of the wrestlers (85%) began wrestling while they were in junior high (seventh, eighth, or ninth grade). Approximately half (48%) reported that they wrestle outside of the school season or spend more than 4 months of the year wrestling. On average, the wrestlers stated that they normally gained 9 pounds in the off-season. At the time of the retest, however, the average amount of weight gained was only 2 pounds. The most likely reason for the discrepancy is that at the time of the retest the wrestlers had only been out of season for approximately 4–6 wk. Furthermore, 37% of the wrestlers were participating in club, or freestyle, wrestling, and 33% were participating in a sport other than wrestling. These wrestlers may not have gained all the weight that they were expecting to or may gain more weight later in the off-season. The wrestlers also reported that the largest amount of weight they had lost to wrestle in a specific weight class was 6.3 pounds. Wrestlers who were classified as “at risk” reported an average loss of 7.4 pounds, whereas wrestlers who were not classified as “at risk” reported an average loss of 5.7 pounds. The difference was only 1.7 pounds, which is less than a full weight class. Considering this small difference, it seems unlikely that the largest amount of weight lost at one time to make a specific weight class contributes to the “at risk” classification.
Twelve of the 22 nonwrestlers (56%) who were classified as “at risk” were athletes who participated in another sport. Four of these athletes (2 football players and 2 basketball players) stated that they felt as though weight was of some importance in their sport. Thirty-one of the 53 nonwrestlers (58%) who were not classified as “at risk” were athletes who participated in another sport. Nine of these athletes (7 football players and 2 basketball players) stated that they felt as though weight was of some importance in their sport. As the number of athletes versus nonathletes who were classified as “at risk” is almost 50/50 (12/10), it seems unlikely that the subjects’ athletic status contributed to their “at risk” classification. Also, the athletes who felt that weight was important in their sports were football players and basketball players, two sports in which weight loss is not normally an issue. In fact, with football, the issue is normally weight gain.
Subjects classified as “at risk”.
It was expected that the number of in-season wrestlers classified as “at risk” for bulimia nervosa would be significantly different than the number of nonwrestlers classified as “at risk.” The results of this comparison are shown in Table 1. Thirty-six percent of in-season wrestlers scored above the preset cutoffs on either the Drive for Thinness subscale, the Bulimia subscale, or the Body Dissatisfaction subscale. In comparison, only 29% of nonwrestlers scored above the preset cutoffs. Although this comparison was in the expected direction, it did not reach significance (χ2 (1, N = 160) = 0.92, P = 0.34).
It was hypothesized that off-season wrestlers and nonwrestlers would not be significantly different in their “at risk” classification as measured by the three subscales of the EDI. The results of this comparison are also shown in Table 1. Twenty-nine percent of nonwrestlers scored above the preset cutoffs on either Drive for Thinness, Bulimia, or Body Dissatisfaction subscales. In comparison, only 19% of off-season wrestlers scored above the preset cutoffs. This was not a significant difference.
It was predicted that any bulimic behaviors shown by the wrestlers during the season would subside in the off-season. Therefore, the same wrestlers were tested both during the season and in the off-season, and their scores were compared. Table 1 reveals a significant difference between the number of in-season wrestlers classified as “at risk” for bulimia due to EDI results and the number of off-season wrestlers classified as “at risk.” Thirty-one percent of the in-season wrestlers scored above the preset cutoffs on either Drive for Thinness, Bulimia, or Body Dissatisfaction as compared with only 19% of off-season wrestlers (P < 0.01). This result supports the predictions.
Subjects classified as “at risk” by EDI subscale.
The three subscales of the EDI that were used to determine “at risk” status each address a different aspect of eating disorders. To gain further insight into the differences between wrestlers and nonwrestlers, the subscales were examined separately for significant differences between groups. The results are shown in Table 2. A higher percentage of in-season wrestlers scored above the preset cutoff on each of the three subscales, but only the Drive for Thinness subscale was found to be statistically significant. Twenty-seven percent of in-season wrestlers scored above the cutoff on the Drive for Thinness subscale, as compared with 13% of the nonwrestlers (χ2 (1, N = 160) = 4.71, P < 0.05).
The subscales were also examined separately for differences between off-season wrestlers and nonwrestlers. No significant differences were found between off-season wrestlers and nonwrestlers on any of the three subscales.
An examination of the three subscales reveals only one significant difference between in-season wrestlers and off-season wrestlers. A higher percentage of in-season wrestlers scored above the preset cutoff on each of the three subscales, but a significant difference was observed only on the Drive for Thinness subscale. Twenty-seven percent of in-season wrestlers scored above the cutoff, as compared with 15% of the off-season wrestlers (P < 0.05).
“At risk” classification.
It was possible for a subject to be classified as “at risk” by scoring above the cutoff on only one of the three subscales. In fact, 71% of in-season wrestlers, 91% of nonwrestlers, and 50% of off-season wrestlers who were classified as “at risk” fall into this category. However, some subjects scored above the cutoff on two subscales, and others scored above the cutoff on all three. In this study, 23% of wrestlers who were classified as “at risk” scored above the cutoff on two subscales, and 6% scored above the cutoff on all three. In the off-season, 38% of wrestlers who were classified as “at risk” scored above the cutoff on two subscales, and 12% scored above the cutoff on all three. In comparison, 9% of nonwrestlers who were classified as “at risk” scored above the cutoff on two subscales, and no nonwrestler scored above the cutoff on all three. None of these differences are significant, but the trend is that more wrestlers than nonwrestlers score above the cutoff on more than one subscale.
All subjects classified as “at risk” were interviewed. This included 31 in-season wrestlers, 16 off-season wrestlers, and 22 nonwrestlers. For an “at risk” subject to be classified as bulimic, the EDE interview had to show that he met all five DSM-IV criteria for bulimia nervosa. Interviews revealed that none of the nonwrestlers met any of the criteria. For both the in-season and off-season, one wrestler met two of the five criteria (i.e., the presence of recurrent inappropriate compensatory behavior to prevent weight gain—in his case, self-induced vomiting and self-evaluation unduly influenced by body shape and weight). Another wrestler in both the in-season and off-season met only one criterion (i.e., self-evaluation unduly influenced by body shape and weight). No other wrestlers met any of the criteria at any point.
Subjects versus normative data.
An important purpose of the study was to compare in-season wrestlers, nonwrestlers, and off-season wrestlers to national norms established by Rosen et al. (27) and Shore and Porter (28). It was not possible to make comparisons based on the “at risk” classification. Subjects in this study were classified as “at risk” by scoring above the cutoff on one or more subscales. Therefore, the number of scores above the cutoffs on all three subscales was not necessarily indicative of the number of subjects classified as “at risk.” For example, although there were 42 wrestlers’ scores above the three subscale cutoffs, this translated into only 31 “at risk” subjects, due to the fact that some wrestlers scored above the cutoff on two subscales, and others scored above the cutoff on three subscales. In the normative data, the percentile ranks presented by Rosen et al. (27) and Shore and Porter (28) are given for each individual subscale. It is not possible to determine how many subjects scored above cutoffs on multiple subscales, and therefore a direct comparison of the “at risk” classification was not done. However, the scores of in-season wrestlers, nonwrestlers, and off-season wrestlers were compared with normative data for each individual subscale.
Normative data reveal that approximately 11% of adolescent boys scored higher than a three on the Drive for Thinness subscale. In comparison, 13% of nonwrestlers and 15% of off-season wrestlers scored higher than a three. This was not a significant result. However, findings were significant when comparisons were made between in-season wrestlers and the established norm. Twenty-seven percent of in-season wrestlers scored above a three on Drive for Thinness in comparison with the normative statistic of 11% (χ2 (1, N = 783)= 17.15, P < 0.05). An examination of the reported means provides further support for this trend. In-season wrestlers, nonwrestlers, and off-season wrestlers recorded means of 2.5, 1.7, and 1.7, respectively, on the Drive for Thinness subscale. In comparison, Rosen et al. (27) reported a normative mean of 1.7.
The established norms reveal that 10% of adolescent boys scored above a four on the Bulimia subscale. In this study, 15% of in-season wrestlers, 11% of nonwrestlers, and 12% of off-season wrestlers scored above a four on this subscale. None of these differences were significant, although as the means illustrate, they were in the expected direction. The in-season wrestlers, nonwrestlers, and off-season wrestlers recorded means of 2.1, 1.7, and 1.5, respectively. In comparison, Shore and Porter (28) reported a normative mean of 1.7.
The established norms for the Body Dissatisfaction subscale reveal that 11% of adolescent boys scored above a 10 on the subscale. In this study, only 7% percent of in-season wrestlers, 8% of nonwrestlers, and 6% of off-season wrestlers scored above a 10. None of these differences were significant. All groups in this study scored below national norms on the Body Dissatisfaction subscale. In-season wrestlers, nonwrestlers, and off-season wrestlers recorded means of 3.9, 3.6, and 2.9, respectively. In comparison, Rosen et al. (27) reported a normative mean of 4.3.
This study examined the differences in bulimic-like behaviors between in-season wrestlers, nonwrestlers, and off-season wrestlers. It was expected that the EDI would classify more in-season wrestlers than nonwrestlers as being “at risk” for bulimia nervosa. This was not the case. There were no significant differences between the number of in-season wrestlers who were classified “at risk” and therefore were interviewed and the number of nonwrestlers who were interviewed. This result indicates that participating in the sport of wrestling does not place these athletes at a higher risk for developing bulimia nervosa. However, specific subscales of the EDI examine different components of disordered eating behaviors. Thus, it was important to determine if differences existed between groups on each of the three subscales.
Drive for thinness.
The Drive for Thinness subscale of the EDI addresses the existence of an intense drive to be thinner or an overwhelming fear of fatness. This has been described as the cardinal feature of eating disorders (6,7). There are seven items within this subscale; the items examine excessive concern with dieting, preoccupation with weight, and fear of weight gain. As expected, in-season wrestlers and nonwrestlers differed significantly on this subscale. Wrestlers are very conscious of their weight while they are in-season. Making weight is directly tied to success—if a wrestler does not make weight, he does not wrestle. If a wrestler is even 1 pound over, that pound must be lost before weigh-in. For many wrestlers, dieting, concern with weight, and worries about weight gain are a part of the sport. It was expected that in-season wrestlers would have more concerns and issues with weight than the nonwrestlers.
In general, interviews with in-season wrestlers revealed that their concerns with weight were due entirely to the demands of wrestling. Almost all in-season wrestlers attempting to lose weight in the last 3 months were doing so specifically to “make weight.” They were generally not concerned with body shape, other than expressing a normal desire to look good, and many stated that they only worried about their weight on the day before a match. For these wrestlers losing weight had the following negative consequences: 1) if they did not lose the weight, they would not compete; and 2) the process of losing weight was not an enjoyable one (e.g., the most common weight loss method the wrestlers mentioned was to restrict food intake while at the same time adding extra bouts of exercise in heavy clothing or plastics).
Other researchers report similar preoccupations with weight (22,25,30,34). Thiel et al. (30) examined wrestlers and lightweight rowers and classified 26% as “weight-preoccupied” according to their elevated scores on the Drive for Thinness subscale of the EDI. Although Thiel et al. (30) used a different cutoff score than was used in this study, the trend is comparable. Enns et al. (13) administered two questionnaires (the Eating Attitudes Test and a restricted eating questionnaire) and found a significant difference between wrestlers and other athletes only on those questions that related to weight fluctuations and dieting. This supports the results of the interviews. The concerns of the wrestlers in this study were focused on weight loss and dieting specifically to “make weight.” This is indicative of the demands of the sport rather than the development of bulimia nervosa.
Previous studies have shown the existence of bulimic-like behaviors among wrestlers (22,25,30,34). These behaviors include self-induced vomiting, laxative and diuretic misuse, and binge eating. The Bulimia subscale of the EDI primarily addresses thoughts and patterns of binge eating. This subscale also contains seven items; six of these items are concerned with eating patterns, and one item addresses the issue of self-induced vomiting. More in-season wrestlers than nonwrestlers scored above the preset cutoff; however, this difference was not significant. This result was not expected. It was thought that a significantly higher number of in-season wrestlers would score above the cutoff.
One explanation for the lack of significant differences between groups in the present study may be that the subscale questions focused on eating patterns, rather than on the existence of inappropriate compensatory behaviors. Many male adolescents consider themselves to be “big eaters”; in fact, being able to eat a large amount of food is often considered an admirable trait. The lead author frequently overheard comments such as “Oh, I stuff myself with food all the time” and observed agreement among peers to this comment. Subsequent interviews did not identify any in-season wrestlers or nonwrestlers as having objective bulimic episodes. Although many subjects revealed that they ate a lot of food and liked to “pig out” with their friends, no one felt as though they were out of control while eating or as though they couldn’t stop eating once they had started. Perhaps if the subscale had contained more questions regarding inappropriate compensatory behaviors, the differences between wrestlers and nonwrestlers would have been statistically significant.
The one question that did address these compensatory behaviors was “I have thought of trying to vomit in order to lose weight.” Of the eight nonwrestlers who scored above the cutoff on the bulimia subscale, none of them had a positive answer to this question (a positive answer would have been “always,” “often,” or “very often”). In comparison, of the 13 wrestlers scoring above the cutoff, eight (62%) of them gave a positive answer. This seems to suggest that among male adolescents who think they overeat, only wrestlers feel the need to use a drastic weight control method such as self-induced vomiting. The eight wrestlers, who reported that they had thought about vomiting to lose weight, comprise 9% of the total sample of in-season wrestlers (N = 85). This is similar to the 8% of wrestlers who reported engaging in self-induced vomiting in both the Woods et al. (34) and Oppliger et al. (25) studies, and higher than the 4% reported in the Lakin et al. (22) study. The follow-up interview found that of the eight wrestlers who had thought about vomiting only two had actually done so, and for one of the two it was only a one-time occurrence. Follow-up interviews were not conducted in other wrestling studies (22,25,34).
One of the key criteria for bulimia nervosa is the presence of binge eating. Other studies, which have reported the percentage of wrestlers meeting all five DSM-III-R criteria, have asked wrestlers whether they binge and how often it occurs (22,25). However, they have allowed the wrestlers to elicit their own interpretation of what constitutes a binge. The DSM-IV criteria provide very specific characteristics of a binge. In the present study, although 13% of in-season wrestlers scored above the cutoff on the Bulimia subscale, the follow-up interview revealed that not one was engaging in objective bulimic episodes, as defined by the DSM-IV criteria. It is crucial to determine whether the wrestlers meet the clinical definition of a binge and not their own personal definition. In addition, no comparison group was used in studies (22,25). This makes it difficult to determine whether these wrestlers are in fact different from a group of nonwrestlers.
The Body Dissatisfaction subscale of the EDI consists of nine questions that address dissatisfaction with the overall shape and size of one’s body. Garner (14) states that it is a primary factor in the initiation and continuation of an eating disorder. There were no significant differences between in-season wrestlers and nonwrestlers on this measure. In fact, a slightly higher percentage of nonwrestlers scored above the cutoff for this subscale. This result is not surprising. The wrestlers practice 5 d·wk−1 for about 2 h. One of the goals of their coaches is to get the athletes physically fit and ready for competition. It seems more likely that participation in the sport would improve body image, rather than cause it to decline. Many of the wrestlers stated in the interviews that it was in the off-season when they worried about their body shape, because they were no longer working out regularly they were concerned about getting out of shape. These results are comparable to those reported by Thiel et al. (30).
An important goal of this study was to determine whether the wrestlers’ concerns with weight were transient, or whether they were carried over into the off-season. The results of this study indicate that the concerns are in fact transient. Although in-season wrestlers and nonwrestlers differed significantly on the Drive for Thinness subscale, these same wrestlers tested in the off-season did not differ significantly from nonwrestlers. This suggests that the wrestlers “at risk” behavior was confined to the duration of the wrestling season. Transient behaviors are not characteristic of an eating disorder. In fact, for a person to be classified as bulimic, behaviors and psychological distress must have been present for at least a 3-month period. A significant difference was found between the number of wrestlers qualifying for an interview during the season and the number of wrestlers qualifying for an interview in the off-season. As expected, there were no significant differences between off-season wrestlers and nonwrestlers on either the Bulimia subscale or the Body Dissatisfaction subscale.
It must be pointed out that this study fails to address maturation effects or the effects of taking the EDI twice. The control group was tested only once; therefore, it cannot be ruled out that some of the changes in wrestlers’ EDI scores in-season versus off-season are not due to the retesting effects. However, the fact that the wrestlers’ responses on the Bulimia subscale and the Body Dissatisfaction subscale were extremely consistent, suggests that this was not the case. Two studies (32,33) illustrate the strong test-retest reliability of the EDI. Crowther et al. (11) report retest reliability coefficients ranging from 0.41 to 0.75. The time between tests was 1 yr. Wear and Pratz (32) report retest reliability coefficients for all subscales other than Maturity Fears that are all above 0.80. The time between tests was 3 wk.
“High risk” subjects.
It is difficult to make direct comparisons to other studies examining the issue of wrestlers’ bulimic-like tendencies due to the variety of questionnaires that have been used. It is possible, however, to observe a similar trend in results between this study and two others (22,25) when examining “high risk” subjects. In this study, scoring above the cutoff on all three subscales meant that the subject had answered questions that addressed all five DSM-IV criteria. Two of the 85 in-season wrestlers (2%), and no nonwrestlers, fell into this category. This is similar to the 1.7% of wrestlers in the Oppliger et al. (25) study and the 1.4% of wrestlers in the Lakin et al. (22) study who answered questions consistent with all five DSM-III-R criteria. This suggests that the results found in this study are comparable to results found by other researchers. The key point to make is that the interviews with the two wrestlers failed to classify them as being bulimic. This emphasizes the need for follow-up interviews in future studies. Obviously, a classification of “at risk” does not always translate into a diagnosis of bulimia nervosa. Interviews make it possible to clarify subjects’ answers on the questionnaire, and to explore their behavior in more detail.
Comparison with normative data.
Another goal of this study was to determine whether differences existed between in-season wrestlers, nonwrestlers, and off-season wrestlers in comparison with normative data established for male adolescents (27). The results indicate that there are no significant differences on any of the three subscales between nonwrestlers and established norms (14,27) and between off-season wrestlers and established norms. The mean scores on the three subscales for the off-season wrestlers and the nonwrestlers further illustrate the similarities between these groups and the established norms. A significant difference does exist, however, between in-season wrestlers and the established norms on the Drive for Thinness subscale. This reinforces the idea that any bulimic-like tendencies that wrestlers exhibit during the season are transient and that once wrestlers are in the off-season, they are no different than the average male adolescent. Again, such transience is not characteristic of an eating disorder.
It is not clear from this study what makes some wrestlers more likely than others to score high enough on the EDI to be placed in the “at risk” category. Descriptive data were collected with the intention of gaining some insight into this issue. Unfortunately, this insight was not provided—the data showed “at risk” wrestlers to be very similar to those not “at risk.” The greatest difference shown between wrestlers “at risk” and those not “at risk” was in the “largest amount of weight lost to wrestle in a specific weight class.” Those wrestlers considered “at risk” lost an average of 7.4 pounds, whereas the other wrestlers lost an average of 5.7 pounds. As was previously pointed out, however, the difference between the groups (1.7 pounds) is not even the difference between two weight classes (5 pounds). Therefore, this does not appear to be a significant factor. It might be expected that as the length of time a wrestler has participated in the sport increases, the likelihood that he will develop bulimic-like behaviors would also increase. However, no significant differences were observed between high school wrestlers and junior high wrestlers, either in-season or in the off-season. Although only one wrestler exhibited bulimic-like behaviors, the scores on the Drive for Thinness subscale show that many wrestlers have concerns about body weight. It would be interesting to find out why some wrestlers more than others demonstrate a greater preoccupation with body weight.
Wrestlers are more weight conscious than nonwrestlers. This does not mean, however, that they can be classified as having an eating disorder. Their sport, by its very nature, may place them in an “at risk” category. The follow-up interviews and the off-season testing have shown, however, that for most wrestlers the “at risk” behavior and weight concerns do not go beyond the wrestling season and do not reach the severity level required by the DSM-IV criteria. Bulimia nervosa is a psychological disorder; this study suggests that wrestling does not place its athletes in danger of being diagnosed with such a disorder. This does not mean that weight loss techniques and strategies should not be a concern for those involved with wrestling. Other studies have shown convincingly that wrestlers are engaging in a number of weight-control behaviors that could be injurious to their physical health. It is crucial that these behaviors are eliminated, and wrestlers know how to lose weight safely and effectively. The focus of future research should therefore be on the physical health of wrestlers rather than on eating disorders such as bulimia.
1. American College of Sports. Medicine position stand on weight loss in wrestlers. Med. Sci. Sports Exerc. 28:ix–xii, 1996.
2. American Medical Association. Committee on the medical aspects of sports: wrestling and weight control. JAMA 201:541–543, 1967.
3. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders
, 4th Ed., 1993, pp. 539–550.
4. Beals, K. A., and M. M. Manore. The prevalence and consequences of subclinical eating disorders in female athletes. Int. J. Sport Nutr. 4:175–195, 1994.
5. Brownell, K. D., and J. Rodin. Prevalence of eating disorders in athletes. Eating, Body Weight, and Performance in Athletes: Disorders of Modern Society, K. D. Brownell, J. Rodin, and J. H. Wilmore (Eds.). Malvern, PA: Lea & Febiger, 1994, pp. 128–145.
6. Bruch, H. Eating Disorders: Obesity, Anorexia Nervosa and the Person Within. New York: Basic Books, 1973, pp. 3–87.
7. Bruch, H. Anorexia nervosa: therapy and theory. Am. J. Psychiatry 139:1531–1538, 1982.
8. Carlat, D. J., and C. A. Camargo. Review of bulimia nervosa
in males. Am. J. Psychiatry 148:831–843, 1991.
9. Cooper, Z., and C. G. Fairburn. The Eating Disorder Examination
: a semi-structured interview for the assessment of the specific psychopathology of eating disorders. Int. J. Eat. Disord. 6:1–8, 1987.
10. Cooper, Z., P. J. Cooper, and C. G. Fairburn. The validity of the Eating Disorder Examination
and its subscales. Br. J. Psychiatry 154:807–812, 1989.
11. Crowther, J. H., R. S. Lilly, and P. A. Crawford. The stability of the Eating Disorder Inventory
. Int. J. Eat. Disord. 12:97–101, 1992.
12. Dummer, G. M., L. W. Rosen, W. W. Heusner, P. J. Roberts, and J. E. Counsilman. Pathogenic weight control behavior of young, competitive swimmers. Physician Sportsmed. 15:75–86, 1987.
13. Enns, M. P., A. Drewnowski, and J. A. Grinker. Body composition, body size estimation, and attitudes towards eating in male college athletes. Psychosom. Med. 49:56–64, 1987.
14. Garner, D. M. Eating Disorder Inventory-2: Professional Manual.
Odessa, FL: Psychological Assessment Resources 1991, pp. 1–70.
15. Garner, D. M., and P. E. Garfinkel. Sociocultural factors in anorexia nervosa. Lancet 2:674–679, 1978.
16. Garner, D. M., and P. E. Garfinkel. Sociocultural factors in the development of anorexia nervosa. Psychol. Med. 10:647–656, 1980.
17. Garner, D. M., P. M. Garfinkel, W. Rockert, and M. P. Olmstead. A prospective study of eating disturbances in the ballet. Psychother. Psychosom. 48:170–176, 1987.
18. Garner, D. M., M. A. Olmstead, and J. Polivy. Development and validation of a multidimensional eating disorder inventory
for anorexia nervosa and bulimia. Int. J. Eat. Disord. 2:15–34, 1983.
19. Hursch, L. M. Food and water restriction in wrestlers. JAMA 241:915–916, 1979.
20. Johnson, C., and D. L. Tobin. The diagnosis and treatment of anorexia nervosa and bulimia among athletes. Athl. Train. 26:119–128, 1991.
21. King, M. B. Eating disorders in a general practice population: prevalence, characteristics, and follow-up at 12 to 18 months. Psychol. Med. 14:(Suppl.)12–22, 1989.
22. Lakin, J. A., S. N. Steen, and R. A. Oppliger. Eating behaviors, weight loss methods, and nutrition practices among high school wrestlers. J. Community Health Nurs. 7:223–234, 1990.
23. Melby, C. L., W. D. Schmidt, and D. Corrigan. Resting metabolic rate in weight-cycling collegiate wrestlers compared with physically active, noncycling control subjects. Am. J. Clin. Nutr. 52:409–414, 1990.
24. Nitzke, S. A., S. J. Voichick, and D. Olson. Weight cycling practices and long-term health conditions in a sample of former wrestlers and other collegiate athletes. J. Athl. Train. 27:257–261, 1992.
25. Oppliger, R. A., G. L. Landry, S. W. Foster, and A. C. Lambrecht. Bulimic behavior among interscholastic wrestlers: a statewide survey. Pediatrics 91:826–831, 1993.
26. Ribisl, P. M. Rapid weight reduction in wrestling. J. Sport Med. 3:55–57, 1975.
27. Rosen, J. C., N. T. Silberg, and J. Gross. Eating attitudes test and Eating Disorders Inventory: norms for adolescent girls and boys. J. Consult. Clin. Psychol. 56:305–308, 1988.
28. Shore, R. A., and J. E. Porter. Normative and reliability data for 11 to 18 year olds on the Eating Disorder Inventory
. Int. J. Eat. Disord. 9:201–207, 1990.
29. Steen, S. N., and K. D. Brownell. Patterns of weight loss and regain in wrestlers: has the tradition changed? Med. Sci. Sports Exerc. 22:762–768, 1990.
30. Thiel, A., H. Gottfried, and F. W. Hesse. Subclinical eating disorders in male athletes. Acta Psychiatr. Scand. 88:259–265, 1993.
31. Tipton, C. M., and T. K. Tcheng. Iowa wrestling study: weight loss in high school students. JAMA 214:1269–1274, 1970.
32. Wear, R. W., and O. Pratz. Test-retest reliability for the Eating Disorder Inventory
. Int. J. Eat. Disord. 6:767–769, 1987.
33. Wilson, G. T., and D. Smith. Assessment of bulimia nervosa
: an evaluation of the Eating Disorders Examination. Int. J. Eat. Disord. 8:173–179, 1989.
34. Woods, E. R., C. D. Wilson, and R. P. Masland. Weight control methods in high school wrestlers. J. Adolesc. Health Care 9:394–397, 1988.