|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
LONGITUDINAL EXAMINATION OF DISORDERED EATING CORRELATES IN
COLLEGIATE FEMALE GYMNASTS
JESSICA C. HALVORSEN
A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE IN
EXERCISE AND SPORT SCIENCES
UNIVERSITY OF FLORIDA
I would like to thank my committee members, Dr. Heather Hausenblas, Dr. Chris
Janelle, and Dr. Peter Giacobbi, Jr., for their valuable feedback on this project. I am
especially grateful to my advisor, Dr. Hausenblas, for her guidance. I appreciate the
support she has shown me throughout my master's program. I would like to thank Amy,
Nini, Beth, and Sarah for their friendship and support during the development of this
study. I would also like to thank my fiance, Jeremy, for his understanding,
encouragement, and love. Finally, I must thank my parents who have been a source of
inspiration, support (both emotional and financial!), and encouragement throughout my
TABLE OF CONTENTS
A C K N O W L E D G M E N T S .................................................................................................. ii
LIST OF TABLES ....................................................... ............ ....... ....... vi
ABSTRACT ........ .............. ............. .. ...... .......... .......... vii
1 IN TR OD U CTION ............................................... .. ......................... ..
B a ck g ro u n d .................................................................................................... ..... .
Significance of the Study ...................................... ................ .. ........ ..
P u rp o se ...................................................... .......................... 3
H y p o th e se s .............................. ............................................................. ............... 3
H y p o th e sis 1 ...................................................... ................ .. 4
H y p oth esis 2 ....................................................... 4
2 REVIEW OF LITERATURE ......................................................... .............. 5
E ating D isorders D efined ............................................................. ....................... 5
Anorexia Nervosa .................. ........................................... .... ........ 5
B ulim ia N erv osa .................................................................................. .. 7
B inge E ating D disorder ................. .............................................. .............. ...9
Eating Disorders Not Otherwise Specified (EDNOS)......................................10
M easurem ent of E ating D isorders ................................... .................... ...... ............11
E ating A attitudes Test (EA T)........................................................... ..... ........... 11
Eating Disorder Inventory (EDI)..................................................................... 12
T he B ulim ia T est ................................................................... 14
A athletes and Eating D isorders ......................................................... ............. 14
Intra-Individual Characteristics ................................ ................................... 15
P erfectionism ................................................................................................ 15
B ody-im age disturbance........................................ ........................... 16
Social physique anxiety ........................................... .......... ............... 17
S o cial In flu en c e s ........................................................................................... 1 8
Sport Participation .......................................... .... .. ........ ................... 19
Studies finding no group differences for eating disorders .........................20
Studies finding athletes at less risk for eating disorders ...........................21
Studies finding athletes at greater risk for eating disorders .......................22
C om prison of sport type ........................................ ......... ............... 23
Gymnastics ................................. ........................... .... ....... 25
R research Lim stations ........................ .................... ... ............. .. .... .. 26
3 M E T H O D .............................................................................3 0
P a rtic ip a n ts ........................................................................................................... 3 0
M easu res ................................................................................ 30
Demographic Questionnaire for Athletes .....................................................30
Demographic Questionnaire for Nonathletes ...................................................31
D rive for Thinness Subscale................................ ................... 31
Body Dissatisfaction Subscale ................................ ............... 32
Perfectionism Subscale....................................... ......... 32
Social Physique Anxiety Scale .................................................. ........ 32
P ro c e d u re ................................................................ 3 3
Data Collection ................................. ........................... .... ....... 33
A athletes ........................................33.............................
Nonathletes ................ ............ ................. ...............34
Techniques to Increase R response R ate.................................... ............... 35
D ata A n aly sis ............................... .......... .. ........ .......... .......................3 6
Drive For Thinness, Body Dissatisfaction, and Perfectionism Subscales...........36
Social Physique A anxiety Scale .................................................................... 37
4 R E S U L T S ............................................................................. 3 8
R espon se R ate.......................................................38
P articipant D em graphics ..................................................................................... 39
Gymnast Dem ographics............................................ 39
R liability of the Study M measures ............................................ ........................ 40
Group Differences for Participants Who Completed Time 1 Only and Those Who
Com pleted Tim e 1 and Tim e 2 ....................................... .... ................. 41
Mail Administration and Direct Administration Differences of Gymnasts .............42
Comparison of Time Differences ................... ........................... ...............43
Comparison of Group Differences.............................. ......................... 44
5 D ISC U S SIO N ......... .............. ............................................................. 45
O v e rv iew .........................................................................................4 5
R espon se R ate.......................................................4 5
Stu dy F in din g s ................................................................................ 4 8
G group D differences .................................... ..........................................48
Tim e Differences for Gymnasts ........................................ ........................ 49
L im itatio n s ........................................ ........................................................ 5 1
F u tu re D direction s ............................................................................................52
C on clu sion s......................................................................................... .54
A DEMOGRAPHIC QUESTIONNAIRE FOR ATHLETES ........................................56
B INJURY QUESTIONNAIRE FOR GYMNASTS ............... .................... ..........57
C DEMOGRAPHIC QUESTIONNAIRE FOR NONATHLETES ............................58
D DRIVE FOR THINNESS SUBSCALE ............................ ........... ..........................59
E BODY DISSATISFACTION SUBSCALE.....................................................60
F PERFECTION ISM SU B SCALE.................................................................. ........ 61
G SOCIAL PHYSIQUE ANXIETY SCALE......... ........... ....................62
H SCHOOLS CONTACTED FOR PARTICIPATION ...........................................63
I L E T TE R T O C O A C H E S ................................................................ .....................64
J IN F O R M E D C O N SE N T ................................................................. .....................65
K INSTRUCTIONS FOR CONTACT PERSON ................... ......................... 67
L IST O F R E F E R E N C E S ........................................................................ .....................69
BIOGRAPH ICAL SKETCH ...................................................... 76
LIST OF TABLES
1-1. Eating Disorders Not Otherwise Specified Categories.............................................11
4-1. Guidelines to Interpret the Reliability of a Measure ..............................................40
4-2. Internal Consistency (Cronbach's Alpha) Estimates for the Study Measures............41
4-3. Mean (M) and Standard Deviation (SD) Scores for Participants Who Completed
Time 1 Only and Those Who Completed Time 1 an Time 2 ...............................42
4-4. Mean (M) and Standard Deviation (SD) Scores for Gymnasts Who Completed the
Study Via M ail and Direct Administration. .................................. .................43
4-5. Mean (M) and Standard Deviation (SD) Scores ....................................................44
Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science in Exercise and Sport Sciences
LONGITUDINAL EXAMINATION OF DISORDERED EATING CORRELATES IN
COLLEGIATE FEMALE GYMNASTS
Jessica C. Halvorsen
Chair: Heather A. Hausenblas
Major Department: Exercise and Sport Sciences
The purpose of this study was to longitudinally examine the eating disorder
correlates of body dissatisfaction, perfectionism, drive for thinness, and social physique
anxiety of female gymnasts and nonathletes over a six-month period. Participants
completed the Social Physique Anxiety Scale and the Body Dissatisfaction, Drive for
Thinness, and Perfectionism subscales of the Eating Disorder Inventory-2 in September
(preseason) and in March (competitive season). One hundred and eleven nonathletes and
102 gymnasts completed the Time 1 questionnaire. The Time 1 response rate for the
nonathletes was 72% and the response rate for the gymnasts was 98%. For Time 2, 17
nonathletes completed the questionnaire for a response rate of 15%, and 79 gymnasts
completed the Time 2 questionnaire for a response rate of 77%. I found that the gymnasts
and nonathletes did not differ significantly on the eating disorder correlates at Time 1.
Because of the poor response rate of the nonathletes for the Time 2 data collection, time
differences were examined only for the gymnasts. I found no significant time differences
for the gymnasts for the eating disorder correlate measures. The attitudinal, behavioral,
and psychological factors that may have explained the study findings were considered,
and areas for future research were highlighted.
The Diagnostic and Statistical Manual-IV (DSM-IV; American Psychological
Association [APA], 1994) provides diagnostic criteria for the following four eating
disorders: anorexia nervosa, bulimia nervosa, binge eating disorder, and eating disorders
not otherwise specified. These eating disorders are associated with negative biological
(e.g., amenorrhea), psychological (e.g., depression), and social effects (e.g., isolation;
APA, 1994). Certain groups are more at risk for developing eating disorders than others.
These groups include women, Caucasians, chronic dieters, models, and dancers (APA,
1994; Demarest & Allen, 2000; Nagel & Jones, 1992; Striegel-Moore, Silberstein, &
Also, some researchers and clinicians consider athletes to be a high-risk group for
developing an eating disorder (Hausenblas & Carron, 1999; Smolak, Murnen, & Ruble,
2000; Thompson & Sherman, 1999a). Furthermore, it has been suggested that subgroups
of athletes may be at an increased risk for eating disorders (Garner, Rosen, & Barry,
1998; Stoutjesdyk & Jevne, 1993; Sundgot-Borgen, 1994). Athletes participating in
aesthetic (e.g., gymnastics, figure skating), endurance (distance running, swimming), and
weight-dependent sports (wrestling, rowing) may be more at risk than athletes
participating in other types of sports (e.g., basketball, archery; Hausenblas & Carron,
1999). In particular, some clinicians and researchers report that athletes competing in the
aesthetic sport of gymnastics may be at the greatest risk for developing an eating disorder
(Hausenblas & Carron, 1999). However, the research examining eating disorders and
athletes in general and eating disorders and gymnasts in particular, is equivocal; thus
further investigation into gymnastics and eating disorders is warranted.
This study extends the research on athletes and eating disorders by addressing
some of the contradictory findings and methodological limitations. First, many
researchers fail to include an appropriate control group in their studies (Smolak et al.,
2000). A control sample that is matched for gender and age is necessary to attribute group
differences to the grouping variables (Ashley et al., 1996). Thus, I recruited an age-
matched nonathletic female control group.
Second, many studies examine athletes of different levels, however, differences
among athletes of different levels have been found (Picard, 1999). For example, Division
I athletes are different from Division III athletes in risk for disordered eating (Picard),
and it cannot be assumed that elite athletes are comparable to athletes of other levels
(Reel & Gill, 1996). Therefore, I examined athletes competing at one level (e.g., Division
I collegiate gymnastics).
Third, the majority of research has examined eating disorders and athletes using
cross-sectional designs. However, examining eating disorders over time is important. If
athletes that are predispositioned to developing eating disorders gravitate towards certain
types of sports then the disordered eating behavior should not vary throughout the year.
However, if participation in certain sports causes disordered eating, then the time at
which measures are taken is important. Thus, I examined eating disorder correlates
longitudinally over a six-month period, assessing eating attitudes once during the
preseason and once during the competitive season.
Fourth, there is little theoretical basis for the sport category breakdowns typically
used by researchers, and arbitrarily placing sports in categories may not reveal useful or
valid results (Hausenblas & Carron, 2002). Also, most classification systems do not take
into account sport-specific demands. Therefore, I examined a large sample of athletes
from a single sport (i.e., gymnastics) rather than multiple sports, thus eliminating sport
classification problems (Hausenblas & Carron, 2002).
Significance of the Study
This study is significant for two reasons. First, it examines a large sample of
athletes from a single sport. To have a large enough sample size for adequate power, the
majority of the research to date combines athletes from many different sports into one
group because it is often difficult to recruit enough athletes from one sport (Hausenblas &
Carron, 2002). Second, this study is one of the first to examine disordered eating
correlates over time. A longitudinal examination of disordered eating correlates in
athletes is necessary to determine if sport participation results in eating disorder
symptoms (Anderson, 1992; Hausenblas & Carron, 1999).
The purpose of this study was to longitudinally examine the eating disorder
correlates of body dissatisfaction, perfectionism, drive for thinness, and social physique
anxiety of gymnasts and nonathletes over a six-month period. The first assessment was in
October (gymnasts' preseason) and the second assessment was in March (gymnasts'
The following two hypotheses were proposed.
It was hypothesized that gymnasts would have greater disordered eating
symptoms during the competitive season than during the preseason (Dale & Landers,
1999). Gymnasts may experience pressures during the season leading them to experience
more disordered eating symptoms than during the preseason. Such pressures include:
making the lineup, qualifying to post-season competition, to have an aesthetically
pleasing appearance for judges, and coaches' expectations (James & Collins, 1997;
Scanlan, Stein, & Ravizza, 1991).
Second, it was hypothesized that gymnasts would have greater disordered eating
symptoms than controls during the competitive season, but they would not differ from
controls during the preseason (Dale & Landers, 1999; Davis, 1992; Hausenblas &
Carron, 1999; Smolak et al., 2000; Sundgot-Borgen, 1993). According to the hypothesis
that sport participation leads to disordered eating, gymnasts should experience an
increase in disordered eating symptoms when they become involved in their competitive
season (Dale & Landers, 1999). It is also assumed that disordered eating symptoms will
remain stable over time for women not participating in sport because they do not
experience the competitive pressures that may lead to disordered eating symptoms.
REVIEW OF LITERATURE
There are three purposes to this chapter. The first is to present information about
the following four eating disorders described in the DSM-IV: anorexia nervosa, bulimia
nervosa, binge eating disorder, and eating disorders not otherwise specified. The second
purpose is to describe three popular eating disorder measures: The Eating Attitudes Test
(Garner & Garfinkle, 1979), the Eating Disorder Inventory-2 (Garner, 1991; Garner,
Olmstead, & Polivy, 1983), and The Bulimia Test-Revised (Smith & Thelen, 1984;
Thelen, Farmer, Wonderlich, & Smith (1991). The third purpose is to provide a review of
the literature pertaining to three popular explanations linking athletes and eating
disorders. The first explanation states that intra-individual characteristics can lead to
eating disorders, the second proposes that an athlete's social influences can lead to eating
disorders, and the third explanation proposes that participation in certain sports can lead
to disordered eating (Hausenblas & Carron, 1999).
Eating Disorders Defined
There are four criteria that must be present for a clinical diagnosis of anorexia
nervosa (APA, 1994). First, there is a refusal to maintain body weight at or above a
minimally normal weight (e.g., less than 85% of normal weight) for age and height.
Second, there is an intense fear of gaining weight or becoming fat, even though
underweight. Third, there is body-image disturbance. Body image is a person's attitude
towards his or her body, and it has cognitive, behavioral, perceptual, affective, and
subjective components (Bane & McAuley, 1998). The final criterion is amenorrhea in
postmenarcheal women. Amenorrhea is the absence of three or more consecutive
menstrual periods (Highet, 1989).
There are two types of anorexia: binge-eating purging type and restricting-type.
Binge- eating purging type anorexics engage in binge eating or purging behavior during
an episode of anorexia, while restricting-type anorexics do not engage in this behavior
(APA, 1994). Purging behaviors include self-induced vomiting, fasting, and misuse of
laxatives, diuretics, and enemas.
It is estimated that 0.2% to 1.3% of the United States population suffers from
anorexia nervosa, and 90% of the sufferers are women (APA, 1994). Caucasians are more
at risk than ethnic minorities (Cash & Henry, 1995). Also, the more affluent are at greater
risk for developing anorexia nervosa than those that are less affluent (Allaz, Bernstein,
Rouget, Archinard, & Morabia, 1998). Adolescents and young adults are also a high-risk
group (APA). Certain personality characteristics are risk factors for anorexia nervosa.
These include a need for approval, conscientiousness, high personal expectations,
perfectionism, obsessionality, insecurity, self-denial, and being deferential to others
(Garfinkle & Garner, 1982; Strober, 1986).
Most medical complications of anorexia nervosa are a direct result of weight loss,
and these complications often return to normal with weight restoration (Costin, 1999).
Anorexics have brittle nails, thinning hair, and yellow-tinged skin due to dehydration,
nutrient deficiencies, and poor liver function (Walsh, Wheat, & Freund, 2000). They also
experience a fine downy growth of hair on the face, back, and arms, known as lanugo.
This hair aids in insulation (Rock, 1999). More serious complications of anorexia affect
the cardiovascular system, the hematological system, the endocrine system, and the
gastrointestinal system. It has been shown that the time for food to move out of the
stomach and into the intestines is slowed in individuals with anorexia nervosa. This is
why anorexics may complain of early satiety and abdominal pain (Walsh et al., 2000).
They may also experience constipation because of slow food transit time, poor colon
reflex, inadequate food intake, and being sedentary (Costin). The cardiovascular system
is affected by severe weight loss by thinning heart fibers which results in diminished
cardiac capacity. Anorexics also have a slowed heart rate and low blood pressure that can
be dangerous in combination with an arrhythmia (Rock). Anorexia also affects the
hematological system, and causes anemia and a low white blood cell count, called
leukopenia. The effects of anorexia on the endocrine system are amenorrhea and
osteoporosis (e.g., deterioration of bone density; Otis, Drinkwater, Johnson, Loucks, &
Anorexics face a variety of psychosocial consequences as well. Their appearance
determines their self-worth, and they are never satisfied with their body shape (APA,
1994). Anorexics may also display depression, social withdrawal, feelings of
ineffectiveness, irritability, insomnia, and decreased interest in sex (APA). Such
individuals may also meet the DSM-IV criteria for major depressive disorder (APA).
Anorexia is also associated with concerns about eating in public, the need to control
one's environment, inflexible thinking, and restrained initiative and emotional expression
The following criteria must be present for a clinical diagnosis of bulimia nervosa
(APA, 1994). First, there must be recurrent episodes of binge eating. Binge eating is
characterized by eating, in a discreet period of time, an amount of food that is greater
than what most people would eat under similar circumstances. There is also a lack of
control over eating during the episode. Second, there is inappropriate compensatory
behavior to prevent weight gain (e.g., self-induced vomiting, fasting, misuse of laxative,
diuretics, enemas). Third, the binge eating and compensatory behaviors must occur at
least twice a week for three months. Fourth, there is body-image disturbance. Fifth, the
behavior does not occur exclusively during episodes of anorexia nervosa.
There are two types of bulimia nervosa: purging-type and nonpurging-type.
Purging-type bulimics engage in either self-induced vomiting or the misuse of laxatives,
diuretics, or enemas. Nonpurging-type bulimics use other compensatory behaviors such
as fasting or excessive exercise, but they do not regularly engage in self-induced
vomiting or the misuse of laxatives, diuretics, or enemas. It is estimated that between
1.0% and 3.0% of the United States population suffers from bulimia nervosa, and women
are at greatest risk than men (APA, 1994). The risk factors for bulimia nervosa are
similar to those of anorexia nervosa.
The majority of the physical complications resulting from bulimia nervosa are a
result of purging (e.g., self-induced vomiting, laxative abuse, and diuretic abuse). For
example, sialadenosis, an early complication from self-induced vomiting, is a swelling of
the parotid gland near the area between the jawbone and the neck. This can be reduced
with the cessation of vomiting (Costin, 1999). The acid from self-induced vomiting also
erodes the enamel on teeth, causes inflammation of the gums, and increases the incidence
of dental cavities. The esophagus is also damaged by acid in the stomach during
vomiting, and it can rupture during forceful vomiting (Walsh et al., 2000). Self-induced
vomiting also causes the following two electrolyte disorders: hypokalemia (low
potassium) and alkalosis (high blood alkaline level). If severe enough, either of these
disorders can result in cardiac arrhythmia, seizures, and muscle spasms (Rock, 1999).
Laxative abuse can result in abdominal discomfort, constipation, and problems with fecal
retention. If laxative abuse is severe enough, a colonectomy may be necessary (Costin).
Diuretic abuse can result in electrolyte imbalance similar to that of vomiting, and it can
also cause the development of lower leg edema (Walsh et al.).
The psychological features associated with bulimia include increased depression,
low self-esteem, increased anxiety, social withdrawal, and anger. About one third of
bulimics also suffer from substance abuse or dependence (APA, 1994).
Binge Eating Disorder
There are five criteria for binge eating disorder (APA, 1994). First, there are
recurrent episodes of binge eating. Similar to bulimia nervosa, binge eating is the eating
of an amount of food that is greater than what most people would eat under similar
circumstances in a discreet period of time. There is also a lack of control over eating
during the episode, such as a feeling that one cannot stop eating or control how much is
eaten. Second, the binge eating episodes must be associated with three or more of the
following: a) eating more rapidly than normal; b) eating until feeling uncomfortably full;
c) eating large amounts of food when not feeling physically hungry; d) eating alone
because of being embarrassed by how much one is eating; and e) feelings of disgust,
depression, or guilt after overeating. Third, there is marked distress regarding binge
eating. Fourth, the binge eating occurs, on average, at least two days a week for six
months. Finally, the binge eating is not associated with the regular use of inappropriate
compensatory behaviors, and it does not occur during the course of either anorexia
nervosa or bulimia nervosa.
It is estimated that 0.7% to 4.0% of the United States population suffers from
binge eating disorder, with women being 1.5 times more likely to suffer from this
disorder than men (APA, 1994). Individuals with binge eating disorder experience dietary
disinhibition, excessive concern with body shape and thinness, and difficulty interpreting
sensations related to hunger and satiety (Dingemens, Bruna, & van Furth, 2002).
Individuals with binge eating disorder often experience varying degrees of obesity
(APA). Obese individuals with binge eating disorder experience lower self-esteem and
greater depressive symptomatology than obese people without binge eating disorder, and
they are more likely to have become overweight at a younger age and to have made more
failed weight-loss attempts (Dingemans et al., 2002). The shame associated with binge
eating and resulting weight gain can lead to negative self-evaluation and depressive
symptoms (Dingemans et al.).
Eating Disorders Not Otherwise Specified (EDNOS)
The category of eating disorders not otherwise specified is used to diagnose
syndromes that resemble anorexia nervosa or bulimia nervosa, but are either missing an
essential feature or the symptoms are not of the required frequency (APA, 1994).
Although these individuals do not meet the full diagnostic criteria for either anorexia
nervosa or bulimia nervosa, they often experience the same physiological and
psychological disturbances as those suffering from anorexia nervosa or bulimia nervosa
(APA). Descriptions of the EDNOS categories are listed in Table 1-1.
Table 1-1. Eating Disorders Not Otherwise Specified Categories.
Category Description of Disorder
Subthreshold bulimia All other criteria for bulimia nervosa are met except that
episodes are less frequent than twice per week or of a
duration of less than three months
Nonbinging bulimia All other criteria for bulimia nervosa are met except that no
binge eating is evident. Instead, inappropriate
compensatory behavior occurs after eating small or normal
amounts of food
Menstruating anorexia All other criteria for anorexia nervosa are met except that
regular menses occurs
Normal weight anorexia All other criteria for anorexia nervosa are met except that
weight is not 15% less than that expected for age and
Measurement of Eating Disorders
A number of self-report questionnaires have been developed to assess eating
disorder symptoms in clinical and nonclinical populations. The most commonly used
measures are the Eating Attitudes Test (Garner & Garfinkle, 1979), the Eating Disorder
Inventory-2 (Garner, 1991; Garner et al., 1983), and the Bulimia Test-Revised (Smith &
Thelen, 1984; Thelen et al., 1991). These measures are often applied to athlete
populations (Hausenblas & Carron 1999; Petrie, 1993; Smolak et al., 2000). The Eating
Attitudes Test is used to assess anorexic symptomatology, and the Eating Disorder
Inventory assesses cognitions and behaviors associated with eating disorders. Finally, the
Bulimia Test is used to assess bulimic symptomatology. Each of these tests is described
in more detail below.
Eating Attitudes Test (EAT)
Garner and Garfinkle (1979) developed the Eating Attitudes Test-40, which
contains 40 items, to examine attitudes and beliefs associated with anorexia nervosa.
Garner, Olmstead, Bohr, and Garfinkle (1982) revised the scale through factor analysis to
create the Eating Attitudes Test-26. The revised version eliminated 14 redundant items
while maintaining a high correlation (r = .98) with the original version. The shortened
version allows for easier administration than the long version. A cut-off score of 20 for
the Eating Attitudes Test-26 and 30 for the Eating Attitudes Test-40 is used to
differentiate eating disorder prone individuals from nondisordered eaters.
Although the Eating Attitudes Test is frequently used, there are several scale
limitations. First, the distinct cut-off fails to account for a continuum of eating concerns.
For example, a person scoring one point below the cut-off is considered the same as
someone scoring at the bottom of the nondisordered range (Mintz, O'Holloran,
Mulholland, & Schneifer, 1997). Second, although the Eating Attitudes Test
differentiates anorexic individuals from a nonclinical sample, it does not distinguish
between anorexics and bulimics (Williamson, Anderson, Jackman, & Jackson, 1995).
Third, the Eating Attitudes Test often incorrectly classifies eating disordered and normal
individuals (Williamson et al., 1995). Although this measure identifies attitudes and
beliefs associated with anorexia, it does not address some of the eating disorder behaviors
that are diagnostic criteria. Thus, the Eating Attitudes Test cannot be used to make an
eating disorder diagnosis (Schlundt & Johnson, 1990). Finally, the psychometric
properties of the scale have been questioned, especially when used with men
Eating Disorder Inventory (EDI)
The most widely used measure for eating disorder research is the Eating Disorder
Inventory (Garner, 1991; Garner et al., 1983). This multidimensional scale has two
versions, the 64-item Eating Disorder Inventory (Garner et al.) and the 91-item Eating
Disorder Inventory-2 (Garner). The original Eating Disorder Inventory contains eight
subscales that assess attitudinal, behavioral, and psychological correlates of eating
disorders. The Drive for Thinness, Bulimia, and Body Dissatisfaction subscales examine
behaviors and attitudes towards eating, body shape, and weight. The Ineffectiveness,
Perfectionism, Interpersonal Distrust, Interoceptive Awareness, and Maturity Fears
subscales measure general psychological characteristics relevant to individuals with
eating disorders. Researchers often use only a few of the subscales to tailor their
measures to what they are examining (Garner).
The Eating Disorder Inventory has been validated in clinical and nonclinical
samples. The inventory distinguishes between eating disordered and nondisordered
individuals (Williamson et al., 1995). The multidimensional nature of the scale accounts
for a wide range of symptoms, including cognitions and behaviors associated with eating
disorders. Although some of the DSM-IV diagnostic criteria for eating disorders are
addressed (e.g., body image disturbance and fear of weight gain), this measure cannot be
used to diagnose eating disorders. It may, however, serve as an effective screening tool
(Williamson et al.).
In addition to the original eight subscales, the Eating Disorder Inventory-2
contains three provisional subscales: Asceticism, Impulse Regulation, and Social
Insecurity. These provisional constructs are not as well validated; however, they have
adequate internal consistency and add to the attitudinal and behavioral profile for eating
disordered individuals. The Eating Disorder Inventory-2 provides predictive validity
similar to the Eating Disorder Inventory.
The Bulimia Test
Smith and Thelen (1984) developed the Bulimia Test to measure thoughts,
feelings, and behaviors associated with bulimia nervosa. The items were derived from the
DSM-III diagnostic criteria for bulimia nervosa. Thelen, Farmer, Wonderlich, and Smith
(1991) updated this scale to create the Bulimia Test-Revised. This revised version is
based on the DSM-IIIR diagnostic criteria, but also diagnoses bulimic individuals
according to the DSM-IV criteria (Thelen, Mintz, & Bowman, 1996). The Bulimia Test-
Revised effectively discriminates bulimic and nonclinical samples. However, this scale
cannot be used to assess other types of eating disorders. The Bulimia Test-Revised has
adequate internal consistency (r = .97) and high test-retest reliability (r = .95; Thelen et
al., 1991). Williamson et al. (1995), however, reported an unstable factor structure and a
high false-negative rate for this instrument.
Athletes and Eating Disorders
There are three popular explanations linking athletes and eating disorders
(Hausenblas & Carron, 1999). The first explanation is that intra-individual characteristics
that are evident in eating disorder patients (e.g., social physique anxiety, perfectionism,
and body image disturbance) are also characteristics evident in athletes; thus placing
certain individuals at high risk for disordered eating. The second explanation is that social
influences inherent in sport (e.g., coaches, family, team members) may lead athletes to
develop eating disorders. The third explanation is that participation in certain sports (e.g.,
aesthetic sports) may lead to disordered eating because of the sport-specific pressures the
athletes face. Each of these explanations will be described in more detail below.
One possible explanation linking athletes and eating disorders proposes that
athletes who have characteristics evident in eating disorder patients (e.g., social physique
anxiety, perfectionism, and body image disturbance) may be predisposed to develop an
eating disorder (Hausenblas & Carron, 1999). Some of the strongest support for this
explanation has been found for the following general characteristics: perfectionism, body
image, and anxiety about the physique (Garner, Rosen, & Barry, 1998; Thompson &
Perfectionism has consistently been linked to the development of disordered
eating in clinical populations (Garner, 1983). For example, anorexic patients have high
levels of perfectionism (Garner). Similarly, Halmi et al. (2000) found that anorexics had
higher perfectionism scores than a healthy comparison group. Tyrka, Waldron, Graber,
and Brooks-Gunn (2002) found that perfectionism was a strong predictor of anorexic
syndromes in young adults. Similar results have been found with athletic populations
(Fulkerson, Keel, Leon, & Dorr, 1999; Thompson & Sherman, 1999b). For example,
Fulkerson et al. found that female athletes who had high perfectionism scores also had
higher drive for thinness scores and had trouble discerning bodily states, all of which are
characteristics of people with eating disorders.
Recent research has found that perfectionism is multidimensional in nature with
adaptive and maladaptive dimensions, often referred to as normal or positive and neurotic
or negative (Davis, 1997; Haase, Prapavessis, & Owens, 2002; Terry-Short, Owens,
Slade, & Dewey, 1995). Adaptive perfectionism is associated with high strivings and
feelings of achievement while maladaptive perfectionism is associated with impossibly
high personal standards and an intense need to avoid failure (Davis). Terry-Short et al.
(1995) found that eating disordered individuals had high positive and negative
perfectionism scores, while athletes reported low levels of negative perfectionism and
high levels of positive perfectionism similar to that of the eating disordered individuals.
Haase et al. (2002) found a positive correlation between negative perfectionism and
social physique anxiety among elite athletes, while positive perfectionism was unrelated
to social physique anxiety. Female athletes with higher social physique anxiety scores
and higher levels of negative perfectionism report greater disturbed eating attitudes
(Haase et al.). These results indicate that athletes that have high levels of perfectionism,
specifically negative perfectionism, may be more at risk for disordered eating than those
athletes with lower levels of perfectionism.
Body-image disturbance is a diagnostic criterion for eating disorders (APA,
1994). Research with athletes has also found support that body image disturbance is
positively related to eating disorders (Berry & Howe, 2000; Hausenblas & Symons
Downs, 2001). For example, Berry and Howe found that body-image disturbance was
positively correlated with and predicted eating disorder symptoms in female athletes.
Williamson et al. (1995) found similar results in that body image was a strong primary
predictor for eating disorder symptoms among female athletes.
In a meta-analytic review, Hausenblas and Symons Downs (2001) found that
athletes have a more positive body image than nonathletes. Furthermore when examining
across sport types, aesthetic, endurance, and weight-dependent sports did not differ in
body image. Also, college athletes have a more positive body image than club or
recreational athletes (Hausenblas & Symons Downs). Although the meta-analysis
examined body image and sport, it did not investigate the relationship between body
image and eating disorders. It appears that although body image disturbance is positively
related to disordered eating symptoms, athletes tend to have better body image than
nonathletes. Perhaps sport improves body image for those athletes who are not
psychologically predisposed to have body-image disturbance.
Social physique anxiety
The anxiety one feels at the thought of having her physique evaluated socially is
called social physique anxiety (Hart, Leary, & Rejeski, 1989). Although social physique
anxiety and pathological eating are independent constructs, the two could be related due
to their close parallels (Haase et al., 2002). Social physique anxiety is positively
correlated with disordered eating (Cox, Lantz, & Mayhew, 1997) and negative
perfectionism (Haase et al.). Athletes and nonathletes often have similarly high levels of
social physique anxiety (Cox et al., 1997).
Haase and Prapavessis (1998) examined the relationship between social physique
anxiety and disordered eating attitudes in a university population. They found that social
physique anxiety and abnormal eating attitudes were positively correlated. They also
found that as social physique anxiety scores increased and BMI scores decreased,
participants reported less healthy eating attitudes. The same was true for eating attitudes
when social physique anxiety scores decreased and BMI increased (Haase &
Social physique anxiety has also been examined in athletic populations. Krane et
al. (2001) examined the social physique anxiety levels of athletes with different types of
uniforms: baggy, revealing, or mixed and exercisers. There was no difference in social
physique anxiety among the different uniform types and exercisers. However, body
dissatisfaction and drive for thinness were the strongest predictors for social physique
anxiety for both exercisers and athletes (Krane et al.). Similarly, Cox et al. (1997) found
that social physique anxiety was the major predictor of disordered eating behavior, and
the combination of social physique anxiety, % body fat, and gender accounted for about
one third of the variance in eating behavior. However, they found no differences in social
physique anxiety between athletes and nonathletes (Cox et al.).
Finally, a study by Hausenblas and Mack (1999) found that divers reported lower
social physique anxiety scores than the athletic control group (i.e., lacrosse, volleyball, &
soccer) and nonathletes. They also found no differences among the groups in regards to
disordered eating correlates. The results of this study indicate that neither sports in
general nor participation in one particular sport lead to social physique anxiety or
It appears that social physique anxiety is positively related to other disordered
eating correlates. Also, athletes who have high levels of social physique anxiety appear to
be more at risk for disordered eating symptoms than athletes who have lower levels of
social physique anxiety.
A second explanation linking athletes and eating disorders proposes that it is the
influence of an athlete's social setting that can lead them to develop disordered eating.
Coaches, peers, family, and a judged environment have all been shown to have a
significant effect on disordered eating symptoms. For example, Zucker et al. (2001)
found that participants in judged sports had a greater drive for thinness and a higher trend
for eating disorder diagnosis than participants in refereed sports or nonathletes, indicating
that a subjective evaluation environment may be a potential risk factor for disordered
eating. Also, Berry and Howe (2000) found that restrained eating was predicted by
societal pressure from coaches and peers. Similarly, Rosen and Hough (1988) found that
75% of female gymnasts who were told by their coaches that they were overweight began
using unhealthy weight control methods. Williamson et al. (1995) found that social
pressure for thinness from coaches and peers, mediated by over concern with body size,
was a risk factor for developing disordered eating symptoms. Sundgot-Borgen (1994)
found that a significant number of athletes reported that due to the recommendations of a
coach, they were dieting to improve performance. Griffin and Harris (1996) found that
the majority of coaches studied had negative attitudes towards overweight people, lacked
knowledge about healthy weight loss methods, and determined if an athlete needed to
lose weight mainly through visual observations rather than objective measures. This
indicates that a coach's subjective evaluation of an athlete's body could result in
unhealthy dieting. In a study that examined gymnasts, Fender-Scarr (1999) found that
negative comments from the coach concerning the gymnast's appearance were positively
correlated with eating disorder symptoms. She also found that pressure from the family to
do well in the sport was positively correlated with disordered eating (Fender-Scarr).
Similarly, Byely, Archibald, Graber, and Brooks-Gunn (2000) found that in nonathletic
populations, mothers who perceived their daughters as heavy acted on these perceptions
and negatively influenced the body image and dieting habits of their daughters.
In an attempt to determine if sport participation leads to disordered eating,
researchers have often examined if athletes are at greater risk for developing eating
disorders than nonathletes. This research, however, is equivocal. That is, some studies
have found that athletes are at greater risk than nonathletes (Davis, 1992; Hausenblas &
Carron, 1999; Smolak et al., 2000; Sundgot-Borgen, 1993), some have found athletes are
less at risk (DiBartolo & Shaffer, 2002; Hausenblas & Symons Downs, 2001; Kirk,
Singh, & Getz, 2001; Kurtzman, Yager, Landsverk, Wiesmeier & Bodurka, 1989; Petrie,
1996; Snyder & Kivlin, 1975; Wilkins, Boland, & Albinson, 1991; Zucker, Womble,
Williamson, & Perrin, 2001), and some have found no difference between the two groups
(Ashley, Smith, Robinson, & Richardson, 1996; Fulkerson et al., 1999; Hausenblas &
Carron, 1999; Krane et al., 2001; Taub & Blinde, 1994; Warren, Stanton, & Blessing,
1990). Examples of each of these groups are provided below.
Studies finding no group differences for eating disorders
Warren, Stanton, and Blessing (1990) found no differences between Division I
collegiate athletes and nonathletic controls on the Eating Disorder Inventory subscales or
on the Eating Attitudes Test. Ashley et al. found similar results in that athletes involved
in lean sports, other sports and controls did not differ with regard to scores on the Eating
Disorder Inventory-2 subscales. Similarly, Taub and Blinde (1994) found that high
school female athletes, nonathletes, and performance squad members did not have
differing disordered eating patterns or pathogenic weight control usage. Fulkerson et al.
(1999) found no difference in level of risk between high school athletes and nonathletes
for developing eating disorders. Specifically, no significant differences were found
between the two groups for the Bulimia, Drive for Thinness, Body Dissatisfaction, and
Perfectionism subscales of the Eating Disorder Inventory. Krane et al. (2001) found
similar results. Their study compared exercisers and athletes with different types of
uniforms (revealing, baggy, and mixed). No differences between the groups were found
in body dissatisfaction, drive for thinness, bulimia, and social physique anxiety. Harris
and Greco (1990) found that collegiate gymnasts used pathological weight control
measures less often than previously studied samples, and their scores on the Eating
Disorder Inventory did not differ from norms for adolescent girls. Finally, a meta-
analysis found that female athletes and nonathletes did not differ on drive for thinness,
which is the cardinal feature of an eating disorder (Hausenblas & Carron, 1999).
Studies finding athletes at less risk for eating disorders
One study found that of the several subgroups examined, athletes reported the
lowest frequency of bulimic and anorexic symptomotology (Kurtzman et al., 1989).
Athletes also reported the highest desired weights and had the lowest ineffectiveness
scores of all the groups (Kurtzman et al.). Zucker, Womble, Williamson, and Perrin
(2001) found that participants in refereed sports reported lower drive for thinness scores
than participants in judged sports and nonathletes. Participants in refereed sports also had
less concern about body size and shape than participants in judged sports and nonathletes
(Zucker et al., 2001). Similarly, in another study, female nonathletes had higher scores on
the Eating Attitudes Test than female athletes (Kirk et al., 2001). Petrie (1996) found that
nonathletes had greater body dissatisfaction than athletes, and had higher perceived
ineffectiveness scores. Athletes were found to be more satisfied with their bodies, and to
feel more worthwhile, effective, and in control of their lives (Petrie). In a study
comparing male and female athletes and nonathletes, the athletic sample had significantly
lower levels of pathological eating behaviors and attitudes (Wilkins et al., 1991). The
same study found that athletes were less likely to perceive themselves as overweight, less
likely to rely on dieting, to exhibit higher self-esteem and to have a more positive body
image (Wilkins et al.). Athletes have also reported less body-image disturbance and fewer
eating disorder symptoms than nonathletes (DiBartolo & Shaffer, 2002; Fulkerson et al.,
1999; Snyder & Kivlin, 1975).
A meta-analysis examining the body image of athletes and nonathletes found that
although the effect size was small, athletes tended to have a more positive body image
than controls (Hausenblas & Symons Downs, 2001). Aesthetic, ball game, and endurance
sports did not differ from each other. When compared to the control group, college
athletes had more positive body image than club/recreational athletes (Hausenblas &
Studies finding athletes at greater risk for eating disorders
Some studies have found that athletes are at greater risk for developing eating
disorders than nonathletes. For example, Davis (1992) found that high performance
athletes had a greater degree of abnormal dieting behavior and greater concern with body
weight than nonathletes. She also found that even though more athletes than nonathletes
were underweight, the athletes were dissatisfied with their bodies, were dieting, and
wanted to be thinner than the nonathletes did. Finally, she found that excessive weight
preoccupation was most common in gymnasts and synchronized swimmers. Sundgot-
Borgen (1993) found that although there was no significant difference between athletes
and nonathletes classified as at risk for eating disorders, when interviewed, significantly
more athletes than nonathletes were found to have an eating disorder. More specifically,
athletes participating in sports in which leanness was considered important were more
likely to develop eating disorders than both nonathletes and athletes competing in sports
in which leanness was considered less important.
In a meta-analysis examining female athletes and eating problems, Smolak,
Murnen, and Ruble (2000) found that athletes were more at risk for developing eating
disorders than nonathletes were, although the effect was small. They also found that
certain groups of athletes were at greater risk than nonathletes, including: college women,
elite athletes, lean sport athletes, and elite lean sport athletes (Smolak et al., 2000). In
another meta-analysis, Hausenblas and Carron (1999) showed that female athletes self-
reported more bulimic and anorexic symptomatology than nonathletes, and that female
athletes participating in aesthetic sports were at the greatest risk for developing eating
Comparison of sport type
Comparing athletes and nonathletes has not provided clear results, as seen in the
equivocality of the research investigating the relationship between eating disorders and
athletes. A major difficulty researchers face is having enough athletes from a single sport
to achieve sufficient power. Because of this, different sports are often grouped together
based on similar characteristics. One common classification system compares sports in
which a specific weight is demanded or thinness and appearance are emphasized (lean
sports) with sports in which thinness and appearance are not overly important (nonlean
sports; Petrie, 1996). Lean sports include diving, cross-country running, gymnastics,
wrestling, and figure skating. Nonlean sports include volleyball, softball, track and field,
tennis, and fencing. Researchers have found a tendency for women involved in lean
sports to be at greater risk for developing an eating disorder than women in nonlean
sports (Petrie; Picard, 1999; Smolak et al., 2000). For example, Petrie found that female
lean sport athletes were more concerned with dieting and more preoccupied with their
weight than both nonlean sport athletes and nonathletes. Also more specifically, Picard
found that female lean sport athletes reported significantly higher scores on the Eating
Attitudes Test than nonlean sport athletes and nonathletic controls. Lean sport athletes
also indicated all the signs and symptoms of eating disorder patients including self-
discipline, control, and denial (Picard, 1999). Similarly, Smolak and colleagues (2000) in
a meta-analytic review found that cheerleaders and dancers were at greater risk than
nonlean sport athletes, while swimmers and gymnasts were not. In fact, in this study,
gymnasts were slightly less at risk than other lean sport athletes. Because the sport
specific demands vary among lean sports, more specific research categories are necessary
when examining eating disorders and athletes.
Some researchers (Davison, Earnest, & Birch, 2002; Hausenblas & Carron, 1999;
Sundgot-Borgen & Larsen, 1993) have used a six-category breakdown of sports.
Aesthetic sports are those in which appearance plays a role in determining performance
excellence, such as gymnastics, figure skating, and diving. Endurance sports, such as
swimming and cross-country running, are distance sports in which body size affects
performance. Ball game sports use a ball in play, and include basketball, soccer, and
baseball. Weight-dependent sports are those in which specific weight limits or categories
are used; for example, wrestling or lightweight rowing. Physical strength is the main
component of power sports, which include shot put, sprinting, and power lifting.
Technical sports are those in which skill, rather than physical ability is necessary; for
example, shooting or billiards (Hausenblas & Carron).
Much research has investigated the prevalence of disordered eating in aesthetic
sports, of which gymnastics is part. For example, Stoutjesdyk and Jevne (1993) examined
high performance athletes participating in aesthetic, weight-dependent, and nonweight-
dependent sports, and they found that athletes in aesthetic and weight-dependent sports
scored higher on the Eating Attitudes Test than the athletes in nonweight-dependent
sports did. Sundgot-Borgen (1993) found that athletes in aesthetic, weight dependent, and
endurance sports had the highest frequency of pathogenic weight control methods.
Davison and colleagues (2002) found that female athletes involved in aesthetic sports had
higher weight concern at age five and seven than athletes in other types of sports.
Aesthetic sport athletes have also been found to be more weight preoccupied than other
athletes (Petrie, 1996). Research has also found that athletes participating in aesthetic
sports have a higher prevalence of eating disorders than athletes participating in other
sports (Sundgot-Borgen, 1994) and have higher scores on eating disorder inventories than
other types of sports (Beals & Manore, 2002). For example, in a meta-analysis,
Hausenblas and Carron (1999) found that women participating in aesthetic sports self-
reported more anorexic symptomatology and had a higher drive for thinness than women
participating in other sport types. That is, aesthetic sport athletes were at greater risk for
developing eating disorders than athletes in other sports.
Research examining eating pathologies with gymnastics is equivocal. For
example, Rosen and Hough (1988) studied 42 gymnasts, and they found that all the
gymnasts were dieting and over half had used at least one form of pathogenic weight-
control (e.g., self-induced vomiting, laxatives, etc.). Petrie (1993) found that of the 215
gymnasts surveyed, 58% met the criteria for an intermediate disordered eating category
and only 22% of the gymnasts had normal eating behaviors. O'Connor and colleagues
(1995) found that more gymnasts than athletic controls reported an absence of a
menstrual period for three or more months, which is a possible indicator of amenorrhea.
However, Smolak et al. (2000), in a review of the literature, found that gymnasts were no
different than nonathletes with respect to eating problems. Similarly, Harris and Greco
(1990) found that the Eating Disorder Inventory scores for collegiate gymnasts were not
significantly different than norms for adolescent girls. In summary, it appears that
overall, gymnasts may be at high risk for developing disordered eating, however, the
research examining whether or not they are at greater risk than athletes participating in
other types of sports is dated and inconsistent. Thus, further research examining eating
disorders in gymnasts is warranted.
There are several limitations to the literature linking athletes and eating disorders
that reduce the generalizability of the results and the potential for comparison across
studies. First, a limitation is the failure to include an appropriate control group (Smolak et
al., 2000). A control sample that is matched for gender and age is necessary to attribute
group differences to the grouping variables (Ashley et al., 1996). Without a control group
there is no point of reference to compare the results of the testing group with. Second,
there is a lack of standardized eating disorder measures used. Eating disorder scales
measure different aspects of the disorders, so an accurate comparison of symptomatology
between studies using different measures is difficult (Hausenblas & Carron, 1999). A
third limitation is combining different levels of sport. For example, it cannot be assumed
that elite athletes are comparable to athletes of other levels (Picard, 1999; Reel & Gill,
1996). Elite athletes' training regimens, coaching situations, competition experiences, etc.
may be different from those of athletes of different levels (Smolak et al., 2000). For
example, Picard found that Division I and Division III collegiate athletes were
significantly different from each other, and Division I athletes were at greater risk for
eating disorders than Division III athletes. Arbitrarily comparing different levels with no
supporting research provides confounding results.
Fourth, the sport classification systems used to group sport-types together lack
consistency in both terminology and category criteria. This makes comparisons among
categories from separate studies difficult (Hausenblas & Carron, 2002). There is also
little theoretical basis for the category breakdowns, and arbitrarily placing sports in
categories may not reveal useful or valid results (Hausenblas & Carron). Also, most
classification systems do not take into account sport-specific demands. For example,
figure skating and wrestling are often placed in the same category because of the
importance both sports place on thinness. However, wrestlers try to "make weight" while
figure skaters try to look appealing. To overcome the limitations of the sport
classification systems, Hausenblas and Carron recommended examining large samples of
individual sports as opposed to combining small samples of multiple sports. They note
however, that it is a challenge for researchers to have a large enough sample size of a
A fifth limitation of the literature is that few researchers have examined
differences in eating disorder syptomotology longitudinally (Hausenblas & Carron,
1999). If athletes that are predispositioned to developing eating disorders gravitate
towards certain types of sports then the disordered eating behavior should not vary
throughout the year. However, if participation in certain sports causes disordered eating,
then the time at which measures are taken is important. For example, disordered eating
may increase from the off-season to the competitive season because athletes may feel
more pressure to be thin and win, especially before competitions.
Only one study has investigated eating behaviors in athletes longitudinally. Dale
and Landers (1999) examined the eating behaviors of junior high and high school
wrestlers during the season versus the off-season. The disordered eating behavior was
examined once during the season and once during the off-season. No differences were
found in the number of in-season wrestlers and nonwrestlers classified as "at risk" for
bulimia. However, the in-season wrestlers had significantly higher Drive For Thinness
scores than the nonwrestlers. Also, during the in-season the wrestlers reported higher
Drive for Thinness than during the off-season. These results indicate that wrestlers were
at risk for an eating disorder during the in-season, but were no different than controls
during the off-season. It is possible that the wrestling season is a risk factor for the
development of disordered eating symptoms, and if this is true, the same could be
possible for the competitive seasons of other sports.
Finally, a significant problem researchers face when using self-report measures is
the accuracy and honesty of the participant's responses. Several researchers have
identified this as a problem when using measures of disordered eating (O'Connor et al.,
1995; Sundgot-Borgen, 1993; Wilmore, 1991). Although women have been shown to lie
on self-report measures of disordered eating, they tend to underreport symptoms rather
than over report eating disorder symptoms (Berry et al., 2000; Johnson et al., 1999;
Wilmore). This is problematic because it makes it very difficult to get an accurate
representation of the prevalence and severity of disordered eating among female
populations. To maximize honest responses, participants should be informed that: a) their
responses are anonymous, b) no identifying information will be requested of them, c)
their coaches, trainers, etc. will not have access to their responses (Johnson et al.;
In general, sport participation for women is positive. However, there appear to be
certain types of sports or sport environments in which disordered eating symptoms may
be prevalent. Women participating in lean, aesthetic, weight dependent, and endurance
sports appear to be at greater risk for developing disordered eating symptoms than other
types of sports. Whether or not athletes participating in the sport of gymnastics are at
greater risk than other lean sports has yet to be determined due to the equivocal research.
This study seeks to add to the literature by eliminating some of the methodological
concerns of previous research.
Participants were 102 female Division I collegiate gymnasts and 111 female
collegiate nonathletic controls. Only Division I programs were selected to control for the
moderating effect of competitive level on eating disorder symptoms in athletes (Picard,
1999). The gymnasts were uninjured actively competitive members of their team, while
the nonathletic controls were recruited from physical activity courses and were not
members of a competitive athletic team.
Demographic Questionnaire for Athletes
The Demographic Questionnaire for Athletes assessed the following information:
age, height, current weight, ideal weight, race, year in school, socioeconomic status, the
number of hours per week they trained in the gym, the number of hours per week they
spent training outside the gym, and the number of years they had participated in
gymnastics (see Appendix A). The athletes also indicated what type of scholarship they
had, what events they usually competed on, and their level of competition before and
during their competitive collegiate experience. The Time 2 questionnaire also included
questions assessing the gymnasts' injury history throughout the season (See Appendix B).
Content validity was established in two ways. First, this questionnaire was compared to
demographic questionnaires used in previous studies on eating disorders and athletes.
Second, three health sciences experts examined the questionnaire for comprehension and
Demographic Questionnaire for Nonathletes
The Demographic Questionnaire for Nonathletes was the same for Time 1 and
Time 2 and assessed the following information: age, height, current weight, ideal weight,
race, year in school, and socioeconomic status (see Appendix C). Nonathletes indicated
whether they were currently either members of a varsity athletic team or a competitive
athletic team. If they answered yes to either of these questions, they then indicated what
team they were on. Nonathletes also indicated if they had ever been members of a
competitive athletic team. If they answered yes, they indicated what team they had been
on, the dates of participation, and the highest level of competition. Those who indicated
they were current members of a competitive sport team were not included in the analysis.
The identical protocol for the Athlete Questionnaire was used to establish content validity
for the Nonathlete Questionnaire.
Drive for Thinness Subscale
The Drive for Thinness subscale of the Eating Disorder Inventory-2 (Garner,
1991) assesses extreme concern with dieting and fear of weight gain, which are the
cardinal features of an eating disorder (see Appendix D). Each item is answered on a six-
point Likert scale anchored at each end by never (1) and always (6). Scores are
determined by summing the items composing this subscale. A higher score is indicative
of increased disordered eating symptomatology, and a score of 14 or greater indicates that
the individual is at high risk for an eating disorder (Garner). Items include, "I feel
extremely guilty after overeating" and "I am preoccupied with the desire to be thinner".
The Drive For Thinness subscale has adequate reliability and validity (Garner).
Body Dissatisfaction Subscale
The Body Dissatisfaction Subscale of the Eating Disorder Inventory-2 (Gamer,
1991) assesses dissatisfaction with the size and shape of specific body parts (e.g., hips,
buttocks; see Appendix E). This subscale includes statements such as "I think that my
thighs are just the right size" and "I feel satisfied with the shape of my body". Each item
is answered on a six point Likert scale anchored at each end by never (1) and always (6).
Scores are determined by summing the items composing the subscale. A higher score
indicates high body dissatisfaction (Garner). This subscale has adequate reliability and
The Perfectionism Subscale of the Eating Disorder Inventory-2 (Gamer, 1991)
assesses the extent to which one believes his or her achievements must be superior and
the belief that others expect superior achievement (see Appendix F). Items include
statements such as, "I hate being less than the best at things" and "I feel that I must do
things perfectly or not do them at all". Each item is answered on a six-point Likert scale
anchored at each end by never (1) and always (6). Scores are determined by summing the
items composing the subscale. A higher score is indicates high perfectionistic tendencies
(Garner). The reliability and validity of the Perfectionism subscale is adequate (Gamer).
Social Physique Anxiety Scale
The Social Physique Anxiety Scale is a nine-item self-report instrument (Martin,
Rejeski, Leary, McAuley & Bane, 1997) that measures self-presentational anxiety related
to the physique (see Appendix G; Hart, Leary & Rejeski, 1989). Participants indicate the
degree to which statements like "Unattractive features of my physique/ figure make me
nervous in this social setting" and "I wish I wasn't so uptight about my physique/ figure"
are characteristic of them. Each item is answered on a five-point Likert scale anchored at
each end by "not at all characteristic (1)" and "extremely characteristic (6)". A total score
is determined by summing the responses. A high score is indicative of high social
physique anxiety. This validity and reliability of the Social Physique Anxiety Scale is
adequate (Martin et al., 1997).
Athletes were recruited from seven United States universities (See Appendix H).
The researcher contacted 30 Division I gymnastics coaches, seven of whom agreed to
participate, and explained the rationale and protocol for the study. Coaches were
informed that the study examined how athletes' self-perceptions and nutritional attitudes
changed over the season, and that the participants would be asked to spend 10-15 minutes
completing a questionnaire packet at the following two times: October 2002 and March
2003. The packet included the Demographic Questionnaire For Athletes, the Drive for
Thinness Subscale, the Body Dissatisfaction Subscale, the Perfectionism Subscale, and
the Social Physique Anxiety Scale, which were all contained in an envelope that could be
sealed. There were two methods of data collection: 1) direct administration or 2) mail
Athletes who completed the study via direct administration were given the
Informed Consent (see Appendix I) and the questionnaire packet either individually or in
small groups. Before being given the packet, each participant signed the consent form. To
be able to match the participants' first set of data with their second set, participants were
assigned a code number based on their birth date and the last four digits of their social
security number. No other identifying information was requested. Participants placed
their completed questionnaires in the envelope and sealed it before returning it to the
For those athletes who completed the study by mail, the researcher sent packets,
including instructions (See Appendix J) to a team contact (not a coach). The contact
person administered the packets in the same manner as described previously. That is, the
informed consent forms were signed and collected before giving out the questionnaires.
The mail-based questionnaire packets came in envelopes as well. Participants were
instructed to place the completed questionnaires in the envelopes, seal them, and return
them to the contact person. The contact person mailed the signed consent forms, the
sealed completed questionnaires, and any unused questionnaires back to the researcher in
self-addressed stamped envelopes that were provided.
Data collection was undertaken during the following two times: once during the
preseason between October 20 and November 1, 2002 and once during the competitive
season between March 7 and March 21, 2003. Collecting the data during the specified
two weeks ensured that all data were collected at similar points during the season. All
data collection procedures were intended to ensure the confidentiality and anonymity of
all participants. Participants were informed that coaches, parents, and teammates would
not have access to their information.
The nonathletes were recruited from physical activity courses at the University of
Florida. Students were given course extra credit for participating. Class instructors
administered consent forms and questionnaire packets to their classes. The packet
included the Demographic Questionnaire For Nonathletes, the Drive for Thinness
Subscale, the Body Dissatisfaction Subscale, the Perfectionism Subscale, and the Social
Physique Anxiety Scale. Data collection procedures were identical to the direct
administration procedures used for athletes, including the assignment of a numeric code.
Participants were asked to provide their e-mail address on the consent form, which was
kept separate from the questionnaires, in order for the researcher to contact them for the
second questionnaire administration. Data for nonathletes were collected during the same
time periods as the athletes. The first time was between October 20 and November 1,
2002 and the second time was between March 7 and March 21, 2003.
Techniques to Increase Response Rate
According to Ransdell (1996) there are four strategies that researchers should use
to increase participant's response rate. These strategies are: 1) to increase perceived
personalization; 2) to gain commitment towards the purpose of the study; 3) to build trust
in the researcher; and 4) to follow up initial contact (Ransdell, 1996). In order to increase
perceived personalization by the participants I followed Ransdell's (1996) suggestions
and used stamps rather than bulk mailing, ink signature on consent forms as opposed to
Xeroxed signature, I included a cover letter to each contact person, and I sent "thank-
you" emails to all athletic contacts and nonathletic participants. In order to gain
participants' commitment to the study I notified coaches of the study before they agreed
to participate, I set cut-off dates for the return of all questionnaires, and I made certain
that the questionnaire was brief. To inspire trust in the researcher, I made it clear that the
study is sponsored by the university by including the IRB approval number on the
consent forms, and I included my advisor's contact information as well as my own.
Finally I followed up with all participants using email reminders as needed.
I also employed specific data collection techniques to improve the honesty of
responses. All participants were assured that even though the study was longitudinal, the
only way they would be identified would be by a number. I also emphasized
confidentiality by informing participants that coaches, parents, trainers, and teammates
would not have access to their responses. In most cases, questionnaires were completed
and then immediately sealed in an envelope to be returned to the researcher.
Prior to examination of the hypotheses, the internal consistency reliability and
descriptive statistics of the study measures were determined.
Drive For Thinness, Body Dissatisfaction, and Perfectionism Subscales
A power analysis indicated that 60 participants per level were needed to achieve a
moderate effect size for the analysis (Cohen, 1992). A 2(group: gymnast, nonathlete) x
2(time: preseason, in-season) repeated measures multivariate analysis of variance (i.e.,
MANOVA) was conducted for the dependent variables Drive For Thinness, Body
Dissatisfaction, and Perfectionism subscales (Garner, 1991). The independent variables
were group (gymnast, nonathlete) and time (preseason, in-season). Prior to conducting
this analysis, the data were examined to ensure they met the appropriate statistical
assumptions. The assumptions for repeated measures MANOVA are: 1) data must be
randomly sampled and independent, 2) each group must be normally distributed, 3)
variance for each dependent variable must be homogeneous, 4) no outliers exist in the
data set, and 5) sphericity (i.e., the variances of differences between any two
measurements within a subject are constant; Grimm & Yarnold, 1995). The first
assumption is an experimental design error, and it was not violated. If violations did
occur for assumptions two, three, four, or five, then appropriate techniques were used to
control for this.
Social Physique Anxiety Scale
A 2(group: gymnast, nonathlete) x 2(time: preseason, in-season) repeated
measures analysis of variance (i.e., ANOVA) was conducted for the Social Physique
Anxiety Scale. The independent variables for this study were group (gymnast, nonathlete)
and time (preseason, in-season). The dependent variable was the social physique anxiety
score. Prior to conducting the analysis, the data were examined to ensure they met the
appropriate statistical assumptions. The assumptions for repeated measures ANOVA are:
1) observations are independent, 2) the population is normally distributed, 3) covariances
are equivalent, 4) variance is homogeneous, and 5) sphericity (i.e., the variances of
differences between any two measurements within a subject are constant; Grimm &
Yarnold, 1995). The first assumption is an experimental design error, and it was not
violated. If violations did occur for assumptions two, three, four, and five, appropriate
techniques to control for this were used.
There were two methods of data collection: direct administration and mail
administration. All of the nonathletes (N= 111) completed the questionnaires via direct
administration. In comparison, 15 (15%) gymnasts completed the questionnaires via
direct administration and 87 (85%) gymnasts completed the questionnaires via mail
In regard to the response rate for Time 1, 155 questionnaires were distributed to
the nonathletes and 104 questionnaires were distributed to the gymnasts. One hundred
and eleven nonathletes and 102 gymnasts completed the Time 1 questionnaire. Thus, the
response rate for the nonathletes was 72% (i.e., 111/155) and the response rate for the
gymnasts was 98% (i.e., 102/104). For Time 2, 17 nonathletes completed the
questionnaire for a response rate of 15% (i.e., 17/111), and 79 gymnasts completed the
Time 2 questionnaire for a response rate of 77% (i.e., 79/102). The 77% response rate for
the gymnasts was largely due to one school of 18 gymnasts failing to return their Time 2
questionnaires. Implications of the low response rate for the nonathletes will be discussed
in Chapter 5. Because of the low response rate of the nonathletes, only the gymnasts who
completed both the Time 1 and Time 2 questionnaires were included in the repeated
measures analysis. Therefore, the second purpose of the study, examining group
differences in disordered eating correlates, could not be analyzed because there were not
enough nonathletes to conduct the analysis with sufficient power (Cohen, 1992).
One-way analysis of variance (ANOVA) on the Time 1 data revealed that the
gymnasts (M= 23.22, SD = 1.75) and nonathletes (M= 22.96, SD = 4.05) did not differ
significantly for BMI [F(1, 210) = .36,p = .55, 2= .002]. The nonathletes (M= 20.06,
SD = 1.50) were significantly older [F(1, 212) = 6.87, p < .01, r2 = .03] than the gymnasts
(M= 19.57, SD = 1.22). For the total sample, the most common race was Caucasian
(77.6%), followed by Asian (6.5%), Hispanic (6.1%), and African-American (5.6%).
Most gymnasts were Caucasian (84.3%) followed by African-American (5.9%), Asian
(3.9%), and Hispanic (2.0%). Most nonathletes were Caucasian (71.4%), followed by
Hispanic (9.8%), Asian (8.9%) and African-American (5.4%). A Pearson Chi-square test
indicated that the two groups did not significantly differ for ethnicity [x2 (5) = 8.71, p =
The most common year in school reported for the total sample was junior
(28.0%), followed by sophomore (25.7%), senior (25.2%), and freshman (20.6%). Most
gymnasts were juniors (30.4%), followed by freshmen (26.5%), sophomores (22.5%),
and seniors (20.6%). Most nonathletes were seniors (29.5%), followed by sophomores
(28.6%), juniors (25.9%), and freshmen (15.2%). A Pearson chi-square test revealed that
the gymnasts and the nonathletes did not differ significantly for year in school [x2 (4) =
7.03,p = .13].
For the sport demographics of the gymnasts, most gymnasts had an athletic
scholarship (73.1%), while 15.4% were walk-ons, 2.6% had an academic scholarship, and
9.0% did not report their scholarship status. For socioeconomic status, most gymnasts
were upper middle class (79.5%), followed by lower middle class (14.1%), and upper
class (3.8%). The highest level of competition before attending college for most gymnasts
was Level 10 (62.8%; the highest competitive level before Elite) followed by Elite
(28.2%), Level 9 (5.1%), and Level 8 (1.3%), while 2.6% did not report their highest
level of competition. During the current season, 29.5% of the gymnasts competed on
vault, 29.5% competed on uneven bars, 33.3% competed on balance beam, 30.8%
competed on floor exercise, and 25.6% competed in the all-around (i.e., all four events).
Nine percent (n = 7) of the gymnasts did not compete due to an injury; therefore these
seven gymnasts were excluded from the repeated measures analysis because they did not
take part in the competitive season.
Reliability of the Study Measures
Reliability of the Eating Disorder Inventory-2 subscales and the Social Physique
Anxiety Scale were established by calculating a coefficient of internal consistency (alpha;
Cronbach, 1951) for each measure for the Time 1 and Time 2 assessments. Alpha
increases as the number of variables increases; therefore, there is no set limit for an
acceptable alpha value. George and Mallery (2001), however, recommended the
guidelines presented in Table 4-1 to interpret the reliability of a measure.
Table 4-1. Guidelines to Interpret the Reliability of a Measure
Alpha Vaules Interpretation
> .9 Excellent
> .8 Good
> .7 Acceptable
> .6 Questionable
> .5 Poor
< .5 Unacceptable
Table 4-2 presents Cronbach's alpha scores for the Eating Disorder Inventory-2
subscales and the Social Physique Anxiety Scale for Time 1 and Time 2 for the gymnasts
who completed both assessments and the total sample. The reliabilities for these
measures were acceptable to excellent (range = 0.71 to 0.92).
Table 4-2. Internal Consistency (Cronbach's Alpha) Estimates for the Study Measures
Gymnasts Total Sample
Study Measures Time 1 Time 2 Time 1 Time 2
(N= 72) (N= 72) (N= 213) (N= 96)
Eating Disorder Inventory-2 Subscales
Drive for Thinness 0.89 0.92 0.90 0.92
Body Dissatisfaction 0.85 0.89 0.84 0.88
Perfectionism 0.71 0.80 0.74 0.79
Social Physique Anxiety Scale 0.91 0.92 0.90 0.92
Group Differences for Participants Who Completed Time 1 Only and Those Who
Completed Time 1 and Time 2
A one-way ANOVA was undertaken to examine the differences for social
physique anxiety between those participants who completed the Time 1 questionnaire
only (n = 117) and those who completed both the Time 1 and Time 2 questionnaires (n =
96). No significant Social Physique Anxiety Scale score group differences were found
[F(1, 211) = 1.28, p = .26, r2 = .01]. A one-way MANOVA was undertaken to examine if
differences for body dissatisfaction, drive for thinness, and perfectionism existed between
those participants who completed the Time 1 questionnaire only (n = 117) and those who
completed both the Time 1 and Time 2 questionnaires (n = 96). The Box's test was not
significant thus, homogeneity of variance-covariance was met [F(6, 261714.1) = .88, p =
.51]. Therefore, Wilks' Lambda test statistic was used to interpret the MANOVA results.
No significant group differences were found [Wilks' Lambda = .99, F(3, 203) = .47, p =
.70, r2 = .01]. Mean and standard deviation scores for participants who completed the
Time 1 only and Time 1 and Time 2 assessments are presented in Table 4-3.
Table 4-3. Mean (M) and Standard Deviation (SD) Scores for Participants Who
Completed Time 1 Only and Those Who Completed Time 1 and Time 2.
Time 1 Only Time 1 and Time 2
Study Measures n = 117 n =96
M (SD) M (SD)
Eating Disorder Inventory-2 Subscales
Body Dissatisfaction 8.54 (6.44) 9.49 (6.42)
Drive for Thinness 5.28 (5.83) 6.09 (5.81)
Perfectionism 6.95 (4.22) 6.85 (4.03)
Social Physique Anxiety 24.20 (7.29) 25.40 (8.04)
Mail Administration and Direct Administration Differences of Gymnasts
Differences between the gymnasts who completed the study via direct
administration (n = 15) and those who completed it via mail (n = 87) were examined on
the study variables (i.e., social physique anxiety, body dissatisfaction, drive for thinness,
and perfectionism) for the Time 1 and Time 2 assessments. A one-way ANOVA revealed
no significant administration differences for Time 1 for social physique anxiety scores
[F(1, 98) =.10, p = .75, r12 < .01] but did reveal a significant administration difference for
Time 2 [F(1,72) = 4.62, p = .04, r2 = .06]. A one-way MANOVA was undertaken to
examine administration differences for body dissatisfaction, drive for thinness, and
perfectionism scores for Time 1. The Box's test was not significant thus, homogeneity of
variance-covariance was met [F(6, 2507.55)= .28, p = .95]. Therefore, Wilks' Lambda
test statistic was used to interpret the MANOVA results. No significant administration
differences were found [Wilks' Lambda= .99, (3, 95)= .34, p = .80, r2 = .01]. A one-
way MANOVA was also undertaken to examine administration differences for body
dissatisfaction, drive for thinness, and perfectionism scores for Time 2. The Box's test
was not significant thus, homogeneity of variance-covariance was met [F(6, 1792.57) =
1.11, p = .35]. Therefore, Wilks' Lambda test statistic was used to interpret the
MANOVA results. No significant administration differences were found [Wilks' Lambda
=.95, (3, 66) = 1.22, p = .31, r2 = .05]. Mean and standard deviation scores for the direct
and mail administration are presented in Table 4-4.
Table 4-4. Mean (M) and Standard Deviation (SD) Scores for Gymnasts Who Completed
the Study Via Mail and Direct Administration.
Time 1 Time 2
Direct Administration Mail-Based Direct Administration Mail-Based
Study Measures n = 14 n = 85 n = 12 n = 61
M (SD) M (SD) M (SD) M (SD)
Eating Disorder Inventory-2 Subscales
Body Dissatisfaction 9.62 (6.60) 8.22 (6.27) 11.18 (8.21) 7.75 (6.80)
Drive for Thinness 5.31 (5.42) 5.35 (5.73) 5.45 (5.05) 5.03 (5.84)
Perfectionism 6.69 (3.73) 7.01 (4.16) 6.36 (3.75) 7.15 (4.57)
Social Physique Anxiety Scale 25.71 (6.37) 25.01 (7.94) 29.25 (6.03) 23.77 (8.39)
Comparison of Time Differences
The primary purpose of this study was to examine if disordered eating correlates
differed for gymnasts between preseason and competitive season. A repeated measures
ANOVA indicated no significant time effect for social physique anxiety [F(1,65) = 1.95,
p = .17, r2= .03]. A repeated MANOVA was conducted to examine the time differences
for body dissatisfaction, drive for thinness, and perfectionism. No significant results were
found [Wilks' Lambda = .93, F(3, 60) = 1.43, p = .24, r2 = .07]. Means and standard
deviations for these study measures are presented in Table 4-5.
Comparison of Group Differences
The secondary purpose of this study was to examine the group differences in
disordered eating correlates of gymnasts and nonathletes. Due to the low response rate of
the nonathletes for Time 2, only Time 1 responses were analyzed.
A one-way ANOVA revealed that the gymnasts and nonathletes did not differ on
social physique anxiety scores [F(1, 211) = .26, p = .61, rl2 < .01]. A one-way MANOVA
was conducted to examine the group differences for body dissatisfaction, drive for
thinness, and perfectionism. The Box's test was not significant thus, homogeneity of
variance-covariance was met [F(6, 300718.0) = .81,p = .56]. Therefore, Wilks' Lambda
test statistic was used to interpret the MANOVA results. No significant group differences
were found on the Eating Disorder Inventory-2 subscales [Wilks' Lambda = .99, F(3,
203) = .65, p = .59, r2 = .01] (See Table 4-5).
Table 4-5. Mean (M) and Standard Deviation (SD) Scores.
Time 1 Time 2 Time 1 Time 2
Study Measures N= 72 n = 66 N= 111 n = 17
M (SD) M (SD) M (SD) M (SD)
Eating Disorder Inventory-2 Subscales
Body Dissatisfaction 9.09 (6.49) 8.18 (7.00) 9.58 (6.55) 11.35 (5.80)
Drive for Thinness 6.03 (5.99) 5.32 (5.94) 6.04 (5.95) 7.29 (6.71)
Perfectionism 6.88(4.12) 7.20(4.54) 7.41 (7.48) 7.37 (4.44)
Social Physique Anxiety Scale 25.37 (8.12) 24.54 (8.29) 24.48 (7.57) 27.94(8.98)
The purpose of this study was to longitudinally examine eating disorder correlates
(e.g., body dissatisfaction, perfectionism, drive for thinness, and social physique anxiety)
of gymnasts and nonathletes over a six-month period. The first hypothesis stated that
gymnasts would have greater disordered eating symptoms during the competitive season
than during the preseason (Dale & Landers, 1999). The results of this study did not
support this hypothesis. The second hypothesis stated that the gymnasts would have
greater disordered eating symptoms than controls during the competitive season, but they
would not differ from controls during the preseason (Dale & Landers; Davis, 1992;
Hausenblas & Carron, 1999; Smolak et al., 2000; Sundgot-Borgen, 1993). Due to the low
response rate of nonathletes, however, this hypothesis could not be completely analyzed,
and only the preseason responses were compared between the athletes and nonathletes. In
support of the null hypothesis, I found that the athletes and nonathletes did not differ
significantly on the Eating Disorder Inventory-2 subscales or the Social Physique
Anxiety Scale. Implications of these results and future research directions are presented
Although Ransdell's (1996) suggestions for maximizing response rate in
questionnaire research were followed (e.g., increase perceived personalization, gain
commitment towards the study, build trust in the capabilities of the researcher, follow-up
initial contact), it was difficult to retain nonathletes from Time 1 to Time 2. The response
rate was 72% at Time 1 and 15% at Time 2 for the nonathletes, and the response rate was
98% at Time 1 and 77% at Time 2 for the gymnasts. It is important to note that the
primary reason for the lower Time 2 response rate for the gymnasts was that one team (n
= 18) failed to return their questionnaires, and the contact person was unresponsive to
two follow up emails.
There are three possible reasons for the low response rate for the nonathletes
compared to the gymnasts. First, I was unable to have a nonathlete contact person for
Time 2. Second, participants were asked to come to my office to complete the Time 2
questionnaire. Finally, I was unable to offer any incentives for Time 2 participants.
Not having a contact person for the nonathletes for Time 2 was problematic. For
Time 1, class instructors served as the contact person and distributed and collected the
questionnaires and then returned them to me. For Time 2, however, data collection took
place during the second semester and the participants were no longer enrolled in the
classes from which they were recruited. Therefore, I did not have a contact person, and I
had to contact each of the participants individually. The students were contacted three
times by email. I sent the first email during the last week of the semester to inform the
students that they would be asked to come to my office to complete the Time 2
questionnaire during the second semester in March. About 40 students replied and were
willing to complete the second questionnaire.
I sent the second email about two weeks before the second data collection would
take place. In this email I reminded the students who I was, what the study was about,
and where they had been recruited. I also gave a schedule of dates and times they could
choose from to sign up to complete the questionnaire. The schedule was a block of two
weeks, and the students could come in any time that I was not either in class or teaching
(about three to four hours per day) between 7:30 a.m. and 5:00 p.m. About 35
participants responded to the email. I sent the third email about one week after the
second. This email contained the same information as the previous one, but I also asked
anyone who had not already responded to please do so because I needed more
participants. Only two people responded.
In comparison, the data collection procedures for the gymnasts were simple, and
they were the same for Time 1 and Time 2, which resulted in a high response rate. The
contact person for each team was responsible for distributing, collecting, and mailing the
questionnaires to me. The teams were free to complete the questionnaires when it was
convenient for them and a prepaid return envelope was included. These procedures had
the following two advantages: a) they minimized the number of people I had to be in
contact with (i.e., one contact person vs. each gymnast), and b) they were convenient for
the athletes by not requiring a lot of time or effort.
A second problem with retaining the nonathletes for the Time 2 data collection
was that the participants were asked to come to my office to complete the questionnaire.
Other data collection options were considered, (i.e., online questionnaire, mail
questionnaire) but based on faculty recommendations, I kept the data collection format as
close to the original as possible so as not to bias the collection procedures. Although I
sent out three email reminders to the participants, most did not respond, and only 17
actually completed the questionnaire.
A third issue was that many of the instructors gave students extra credit for
completing the Time 1 questionnaire, which was not an incentive for Time 2. The extra
credit encouraged the nonathletes to participate in the Time 1 assessment. For Time 2,
however, students were unable to earn extra credit, as they were no longer registered for
the classes from which they were first recruited. I was also unable to offer any incentives
for the participants. This lack of reward, combined with the fact that the participants were
asked to come to my office may have prevented the students from being interested in
participating in the Time 2 data collection. In hindsight, it would have been helpful to
recruit participants from classes that met for an entire year rather than one semester. It
would have also been helpful to be able to offer some incentives for participants (e.g.,
money, prizes, gift certificates). Also, having participant's phone numbers would have
allowed me a more direct, personal way to contact them.
There was a possibility that women with disordered eating problems would
choose not to participate in Time 2 (O'Connor et al., 1995). Therefore, I compared the
body dissatisfaction, drive for thinness, perfectionism, and social physique anxiety scores
of those participants who completed both questionnaires to those who completed only the
first questionnaire. No differences between the two groups were found, thus indicating
that participants chose not to complete the second questionnaire due to a reason other
than body disturbance or disordered eating.
Although group differences between gymnasts and nonathletes over time could
not be compared due to the low response rate of the nonathletes, group differences for
Time 1 were examined. Results indicated that gymnasts and nonathletes did not differ for
any of the study measures (i.e., body dissatisfaction, drive for thinness, perfectionism,
social physique anxiety). These results do not support research that found athletes to be
more at risk for disordered eating than nonathletes (Davis, 1992; Hausenblas & Carron,
1999; Smolak et al., 2000; Sundgot-Borgen, 1993). Additionally, the results of this study
do not support research that found athletes to be less at risk for developing disordered
eating (DiBartolo & Shaffer, 2002; Hausenblas & Symons Downs, 2001; Kirk et al.,
2001; Kurtzman et al., 1989; Petrie, 1996; Snyder & Kivlin, 1975; Wilkins et al., 1991;
Zucker et al., 2001).
The results of this study do support a large body of literature, however, that has
found that athletes and nonathletes are no different for eating disorder risk (Ashley et al.,
1996; Fulkerson et al., 1999; Hausenblas & Carron, 1999; Krane et al., 2001; Taub &
Blinde, 1994; Warren et al., 1990). This study also supports the findings of Harris and
Greco (1990), and Smolak et al. (2000) who found that gymnasts were no different than
adolescent girls and nonathletes for eating disorder symptoms. Most of the research that
found gymnasts to be a high-risk group for eating disorders was done before 1995, when
USA Gymnastics began its proactive approach to dealing with eating disorders
(O'Connor et al., 1995; Petrie, 1993; Rosen & Hough, 1988; USA Gymnastics, 1995).
USA Gymnastics' effort in promoting eating disorder prevention over the past several
years may be effective and resulting in healthier athletes. In contrast to previous research,
participation in the aesthetic sport of gymnastics at the Division I level may not be a risk
factor for disordered eating.
Time Differences for Gymnasts
I compared the differences in disordered eating symptoms of the gymnasts from
preseason to competitive season. Based on Dale and Landers (1999) study of wrestlers, I
hypothesized that the gymnasts would have greater disordered eating symptoms during
the competitive season than during the preseason. The results of this study did not
support this hypothesis. That is, there were no significant time differences for body
dissatisfaction, drive for thinness, perfectionism, or social physique anxiety for the
gymnasts. It is important to note, however, that although these results were not
significant, body dissatisfaction, drive for thinness, and social physique anxiety scores
decreased from preseason to competitive season. These contradictory findings to Dale
and Landers can perhaps be attributed to the differing demands of wrestling and
gymnastics. Wrestlers are conscious of their weight during the competitive season
because weight is directly tied to success. That is, if a wrestler does not make weight for
his weight class, he does not compete (Dale & Landers). Due to the weight classification
system of wrestling, athletes know what they must weight to compete. If they are even
one ounce over their weight class limit, they must lose the weight to compete. This
constant concern with specific weight during the competitive season is likely to be
positively related to disordered eating correlates (Dale & Landers). When the wrestlers
are no longer required to make weight (i.e., postseason, off season), most of them lose
their weight preoccupation (Dale & Landers).
Gymnastics is different from wrestling in that there are no weight specifications
for competitors, but rather there is an implied ideal physique. Gymnasts perceive that
having a similar physique to former and existing champions will increase their chances of
success (USA Gymnastics, 1995). Evidence has shown that smaller, lighter gymnasts
tend to receive higher scores (Claessens, Lefevre, Beunen, & Malina, 1999), and that a
lower BMI is related to better gymnastics performance until the BMI falls into an
unhealthy range (Sherman, Thompson, & Rose, 1996). Gymnasts may try to lose weight
to attain the implied ideal shape or weight desired for competition. There may be no
differences in gymnasts' disordered eating correlates over the course of the season
because gymnasts are either continuously trying to attain the ideal physique and
consistently score high on disordered eating correlates, or they have attained the ideal
physique and are satisfied with their bodies and consistently score low on disordered
eating correlates. The athletic season may not be a factor for gymnasts' disordered eating
correlates, but the attempt to attain the ideal gymnastics physique may be a factor.
Although the findings of this study contradict those of Dale and Landers (1999),
they support the explanation that athletes who have characteristics evident in eating
disorder patients (e.g., perfectionism, drive for thinness, body dissatisfaction) may be
predisposed to develop an eating disorder (Hausenblas & Carron, 1999). If this
explanation is correct, disordered eating correlates would remain the same over time and
the athletic season would have no impact on disordered eating scores. The results of this
study support this explanation, as there were no significant changes in disordered eating
correlates from preseason to competitive season.
There are several study limitations that must be mentioned. First limitation the
response rate of the nonathletes was poor. That is, only 15% of the nonathletes completed
both the Time 1 and Time 2 questionnaires, thus eliminating the control group for the
repeated measures analysis because there was not sufficient power to conduct the analysis
(Cohen, 1992). Not having a control group is problematic because there is no point of
reference to compare the results of the testing group.
A second limitation to this study was that it relied on self-report measures and no
objective measure of disordered eating symptoms was used. Several researchers have
found that women are not always honest when completing disordered eating
questionnaires, and tend to underreport their symptoms (O'Connor et al., 1995; Sundgot-
Borgen, 1993; Wilmore, 1991). To maximize honest responses, participants in this study
were informed that their responses were anonymous, and that their coaches, trainers,
parents, and teammates would not have access to their responses. To help check the
honesty of participant's responses, including a social desirability measure would have
A third limitation was the use of different data collection methods. Direct
administration and mail-based methods were used. Although it is desirable to use the
same method for all data collection, the logistics involved with this study made this
difficult. Because the gymnasts participating in this study were at schools from all over
the country, it was impossible for me to directly administer the questionnaire to them. In
the future, creating an online questionnaire at the onset of the study would be a
convenient and private way to use the same procedure to collect data from a large sample
across a vast geographic region.
This study was one of the first to investigate disordered eating correlates with
athletes over time, and was the first to do so with gymnasts. The results of this study
indicate that further research in this area is needed. First, more longitudinal studies are
needed to investigate how athletes' eating attitudes and behaviors change over time.
Researchers should examine disordered eating correlates in athletes over the course of an
entire year (i.e., preseason, competitive season, postseason, and off-season) to determine
if sport participation plays a role in the development of disordered eating. This would
enable researchers to identify points in the season, if any, when athletes may be
susceptible to developing dangerous eating behaviors.
Second, longitudinal studies investigating childhood and adolescence are
necessary to determine if people with certain intra-individual characteristics (e.g.,
perfectionism, body dissatisfaction) are predisposed to developing eating disorders.
These longitudinal studies should examine the role that an athlete's societal influences,
particularly those of the family and those involved in the athletic environment (e.g.,
coaches, teammates, judges), play in the development of disordered eating over time.
Third, the fact that this study found gymnasts to be different from wrestlers
further supports Hausenblas and Carron's (2002) recommendation that researchers
examine sports individually rather than group different sports together. Future research
should continue to do so with other sports, particularly those proposed to be high-risk
sports (e.g., aesthetic, weight-dependent, endurance; Hausenblas & Carron, 1999). This
study illustrates that the sport-specific demands vary from sport to sport; so combining
sports together does not provide an accurate picture of how the sporting environment
affects athletes' disordered eating correlates. Researchers should also examine individual
sports over time. This would allow researchers to track the development of disordered
eating attitudes and behaviors within the context of the specific sporting environment,
and as athletes progress to higher levels of sport.
Finally, future research involving gymnasts should examine different levels of
gymnastics including gymnasts at both the elite and collegiate levels (Picard, 1999). For
instance, Division I athletes have to earn and keep their athletic scholarships, while
Division II and III athletes do not experience this pressure. This study examined Division
I collegiate gymnasts, but unlike other collegiate athletes, gymnasts are at their peak
around age 16, before they enter college (Sands, Hoffman, & Nattiv, 2002; USA
Gymnastics, 1995). Some collegiate gymnasts consider themselves "washed up" or "over
the hill" by the time the reach the collegiate level (Jarrett, 2001). Therefore, it is
important to examine gymnasts of different levels, particularly because collegiate
gymnastics differs from elite level gymnastics (Jarrett, 2002). The rules, scoring system,
and training regimen of collegiate gymnasts are different from those of an elite level
gymnast. For example, the NCAA only allows collegiate teams to practice 20 hours per
week, while most elite level gymnasts train between 25-45 hours per week (Ryan, 1995).
Also, the main focus for collegiate gymnasts is school, while elite gymnasts focus mainly
on gymnastics (Ryan, 1995).
This study examined the disordered eating correlates in collegiate gymnastics
(proposed to be a high risk group; Petrie, 1993) over the course of a season, and initially
compared them to a group of nonathletes. Contrary to previous research, which found
that women participating in lean, aesthetic, weight dependent, and endurance sports to be
at greater risk for developing disordered eating symptoms than other types of sports
(Hausenblas & Carron, 1999), as well as nonathletes, this study found no difference
between gymnasts and nonathletes. That is, gymnasts and nonathletes were no different
from each other in body dissatisfaction, drive for thinness, perfectionism, or social
physique anxiety (Ashley et al., 1996; Fulkerson et al., 1999; Krane et al., 2001; Taub &
Blinde, 1994; Warren et al., 1990). Also, gymnasts' body dissatisfaction, drive for
thinness, perfectionism, and social physique anxiety did not differ from the preseason to
the competitive season. Participation in Division I gymnastics may not be a risk factor for
disordered eating disordered eating. Due to the serious negative effects of disordered
eating, further research is needed examining eating disorders and athletes using
DEMOGRAPHIC QUESTIONNAIRE FOR ATHLETES
Birthdate (mm/dd/yyyy) and last 4 digits of your SS #
1. Age: Height (ft./inches):
2. Current Weight (pounds): Ideal Weight (What you would like to
3. How many hours per week do you train in the gym?
4. How many hours per week do you train outside the gym? (i.e., weight room,
5. Number of years participating in gymnastics?
6. Indicate your race/ethnicity: (circle one)
Caucasian/ White Asian Hispanic African-American/ Black
American Indian Other
7. Year in school: (circle one)
Freshman Sophomore Junior Senior Grad Other
8. Your family socio-economic status:
Upper class Upper/middle class Lower/middle class Lower class
9. What type of scholarship do you have? (circle all that apply):
Athletic Academic Walk-on Full Partial None Other
10. Please indicate the events on which you usually compete (Circle all that apply):
Vault Uneven Bars Balance Beam Floor Exercise All-Around
11. What was your highest level of competition in gymnastics before becoming a
collegiate gymnast? (i.e. Level 10, elite, national team member, etc.)
12. What was your highest level of competition in gymnastics as a collegiate
gymnast? (i.e. Regionals, NCAA Championships, etc.)
INJURY QUESTIONNAIRE FOR GYMNASTS
Part D Instructions: Please answer the following questions regarding your injury history
during the current gymnastics season.
1. Have you had any injuries this season? (Check one) Yes No
If you answered "Yes" please answer the following questions.
2. Please give a brief description of the injury(s)
3. Did your injury prevent you from participating in team practice for any amount of
time? Yes No
If you answered "Yes" please indicate about how long you were unable to participate.
Days OR Weeks
4. Did your injury(s) prevent you from competing in meets? Yes No
If you answered "Yes" please indicate how many meets you missed due to injury.
5. If you were still able to participate while injured, how did the injury(s) affect your
participation? Please check all that apply.
Had to limit amount of practice time
Did alternative workout (e.g., strength, conditioning, etc.)
Water down level of difficulty in routines
Limited events in practice/competition
DEMOGRAPHIC QUESTIONNAIRE FOR NONATHLETES
Birthdate (mm/dd/yyyy) and last 4 digits of your SS #
1. Age: Height (ft./inches):
2. Current Weight (pounds): Ideal Weight (What you would like to
3. Indicate your race/ethnicity (Check one):
Caucasian/ White Asian Hispanic African-American/ Black
American Indian Other
4. Year in school (Check one):
Freshman Sophomore Junior Senior Grad Other
5. Your family economic status (Check one):
SUpper class Upper/middle class Lower/middle class Lower class
6. Are you a current member of any varsity athletic team? (Check one): Yes No
If you checked "Yes" please describe
7. Are you a current member of a competitive athletic team? (Check one):
If you checked "Yes" please describe:
8. Have you ever been a member of a competitive athletic team? (Check one):
If you checked "Yes" please list the sport and approximate dates of participation for
your most recent athletic team membership.
DRIVE FOR THINNESS SUBSCALE
Instructions: Using the scale provided below, please complete the following questions as
honestly as possible. For each item, decide if the item is true about you.
I eat sweets and carbohydrates without feeling nervous. 1 2
I think about dieting. 1 2 3 4 5 6
I feel extremely guilty after overeating. 1 2 3 4 5 6
I exaggerate or magnify the importance of weight. 1 2 3
If I gain a pound, I worry that I will keep gaining. 1 2 3
I am terrified about gaining weight. 1 2 3 4 5 6
I am preoccupied with the desire to be thinner. 1 2 3 4
3 4 56
BODY DISSATISFACTION SUBSCALE
Instructions: Using the scale provided below, please complete the following questions as
honestly as possible. For each item, decide if the item is true about you.
Never(l) Rarely(2) Sometimes(3) Often(4) Usually(5) Always(6)
1. I think my buttocks are too large. 1 2 3 4 5 6
2. I think that my hips are just the right size. 1 2 3 4 5 6
3. I think that my thighs are just the right size. 1 2 3 4 5 6
4. I think my hips are too big. 1 2 3 4 5 6
5. I think that my stomach is too big. 1 2 3 4 5 6
6. I think that my thighs are too large. 1 2 3 4 5 6
7. I like the shape ofmy buttocks. 1 2 3 4 5 6
8. I think that my stomach is just the right size. 1 2 3 4 5 6
9. I feel satisfied with the shape of my body. 1 2 3 4 5 6
Instructions: Using the scale provided below, please complete the following questions as
honestly as possible. For each item, decide if the item is true about you.
Never(l) Rarely(2) Sometimes(3) Often(4) Usually(5) Always(6)
1. I hate being less than the best at things. 1 2 3 4 5 6
2. Only outstanding performance is good enough in my family. 1 2 3 4 5 6
3. As a child, I tried very hard to avoid disappointing my parents and teachers. 1 2 3 4 5 6
4. I have extremely high goals. 1 2 3 4 5 6
5. My parents have expected excellence of me. 1 2 3 4 5 6
6. I feel that I must do things perfectly or not do them at all. 1 2 3 4 5 6
SOCIAL PHYSIQUE ANXIETY SCALE
Instructions: A number of statements which people have used to describe themselves are
given below. Please read each statement carefully. Indicate the degree to which the
statement is characteristic or true of how you feel right now, at this moment in time using
the scale provided below. Please place your answer in the blank space provided after each
1 = not at all characteristic
2 = slightly characteristic
3 = moderately characteristic
4 = very characteristic
5 = extremely characteristic
1. I wish I wasn't so uptight about my physique/ figure.
2. I am bothered by thoughts that other people are evaluating my weight or muscular
3. Unattractive features of my physique/ figure make me nervous in this social setting.
4. In the presence of others, I feel apprehensive about my physique/ figure.
5. I am comfortable with how fit my body appears to others.
6. It makes me uncomfortable to know others are evaluating my physique/ figure.
7. When it comes to displaying my physique/ figure to others, I am a shy person.
8. I usually feel relaxed when it is obvious that others are looking at my
9. When in exercise clothes, I often feel nervous about the shape of my body.
SCHOOLS CONTACTED FOR PARTICIPATION
School Initial Response Outcome
Arizona State University No Response N/A
Auburn University Yes Data collected
Brigham Young University No Response N/A
George Washington University No Response N/A
Illinois State University Yes Data collected
Iowa State University Yes Unable to follow up
James Madison University Yes Data collected
Lousiana State University No Response N/A
Michigan State University No N/A
North Carolina State University No Response N/A
Penn State University No N/A
Temple University Yes Unable to follow up
Towson University No Response N/A
University of Alabama No N/A
University of Arizona No Response N/A
University of Arkansas No N/A
University of California- Los Angeles No Response N/A
University of Denver Yes Data collected
University of Florida Yes Data collected
University of Georgia Yes Data collected
University of Illinois- Champaign Yes Unable to follow up
University of Illinois- Chicago No Response N/A
University of Iowa No Response N/A
University of Kentucky No Response N/A
University of Maryland Yes Data collected
University of Michigan No N/A
University of North Carolina- Chapel No N/A
University of Rhode Island No Response N/A
University of Utah No Response N/A
University of West Virginia No Response N/A
LETTER TO COACHES
I am a graduate student in the department of Exercise and Sport Sciences at the
University of Florida, working under the supervision of Dr. Heather Hausenblas. As both
a coach and a former gymnast, I have been involved with the sport of gymnastics for
most of my life. Consequently, I have chosen to study the nutritional attitudes and self-
perceptions of gymnasts for my Masters thesis. I am having gymnasts from all over the
country complete surveys for my study, and I was hoping to make arrangements for your
team to participate. Those that wish to participate will be asked to complete two
anonymous surveys over the course of the year. One will be in October, and one will be
in March. They will take approximately 10 minutes to complete and will not take time
away from practice, as the athletes will complete them on their own time. If you choose
to participate, please provide the name and either phone number or e-mail address of a
trainer or manager for the team whom I can use as a contact person to distribute and
collect the surveys.
If you could let me know by Friday, October 4 whether or not you would be willing to
participate, I would greatly appreciate it. Feel free to contact me if you have any
questions. Thank you very much for your help.
To: Volunteers for the Changes in Self-perceptions and Nutritional Attitudes Study
From: Dr. Heather Hausenblas and Jessica Halvorsen
RE: Informed Consent
The purpose of this statement is to summarize the study we are conducting and to explain
what we are asking you to do. Every participant will be assigned a code number and will
not be identified by name, but rather by the code. Your identity will be kept confidential
to the extent provided by law. All data will be treated in strict confidence and will be
locked in a filing cabinet in the Exercise Psychology Laboratory in Room 145, Florida
Dr. Hausenblas is an Assistant Professor in Exercise Psychology and Jessica Halvorsen is
a Master's student in Sport and Exercise Psychology at the University of Florida. We are
interested in examining how the nutritional attitudes and self-perceptions of athletes and
non-athletes change over time. This study will take place twice, and the second time will
take place approximately 4-5 months after the first time. Both times you will be asked to
complete several questionnaires that will take approximately 10 minutes to complete.
These questionnaires will include items regarding body image, nutritional attitudes, and
self-perceptions. If at any time you do not wish to continue with the study, you are free to
discontinue your involvement without consequence. There are no risks involved with
participating in the study. The benefits associated with the study are a better
understanding of how self-perceptions and nutritional attitudes change over time.
Your participation in voluntary, but it is hoped that you will take part in this study.
Without the cooperation of volunteers, projects like this would not be possible. Please
feel free to ask any questions you may have at this time. If you have any additional
questions or concerns during the course of the study, please contact Dr. Hausenblas (392-
0584 ext. 1292) or Jessica Halvorsen (392-0580 ext. 1368). If you have any questions or
concerns about your rights as a research participant, you may contact the University of
Florida Institutional Review Board at Box 112250, University of Florida, Gainesville, FL
32611-2250, or call (352) 392-0433.
If you have no further questions at this time, and if you agree to volunteer to become
involved in this study, please read the following statement and sign your name in the
I have read the procedure described above. I voluntarily agree to participate in the study,
and I have received a copy of this description.
Principal Investigator's Signature:
INSTRUCTIONS FOR CONTACT PERSON
Dear Contact Person,
Thank you for your help with my thesis study! I appreciate your time and
assistance. Here are directions for distributing and collecting the questionnaires. Please
look them over and let me know if you have questions.
Questionnaires need to be completed any time between October 20, 2002 and
November 1, 2002. They take approximately 10 minutes to complete, so the end of
practice may be the most appropriate time to give them out. You will need to hand out the
consent forms, collect them back, hand out the questionnaires, collect them back, and
then mail everything back to me. I have provided pre-addressed prepaid envelopes for
you to use to send everything back to me. Here are more detailed instructions:
1. Distribute the Informed Consent forms. The consent form explains the purpose
and rationale of the study and provides contact information for my advisor, the
University of Florida Institutional Review Board, and me. It is very important that
everyone who completes a questionnaire has signed an informed consent form.
After you pass them out, please reiterate the following points:
a. Participation is voluntary
b. They may leave blank any questions that they do not want to answer
c. They may stop at any time, or turn in a blank questionnaire
d. Their names will only be on the consent forms, which are kept entirely
separate from the questionnaires.
2. Collect the consent forms back and place them into the envelope to be mailed
back to me.
3. Distribute envelopes that contain the questionnaires. Each person will have
her own envelope containing a questionnaire. (The envelopes are addressed to my
advisor and are prepaid, but please do not put the envelopes themselves in the
mail, as there is currently no money on that account.) Please explain the
a. Ask everyone to write their code number in the spaces provided on the
top of their questionnaire. The number is 12 digits long, and it is their birth
date (mm/dd/yyyy) followed by the last 4 digits of their social security
number. This number will be used ONLY to keep the first and second sets
of questionnaires matched up.
b. Please instruct gymnasts to put their completed questionnaire back into
their envelope and seal it.
4. Collect sealed envelopes back and place in envelope to be mailed back to me.
5. Some final thoughts:
a. Please express my gratitude to everyone for their help and participation!
b. If you could, please send me a quick email once you have put the return
envelopes in the mail just to let me know.
c. This process will be repeated again in March.
Thank you again for all your help, and please let me know if you have any
questions or if I can be of further assistance.
LIST OF REFERENCES
Allaz, A.F., Bernstein, M., Rouget, P., Archinard, M., & Morabia, A. (1998). Body
weight preoccupation in middle-age and ageing women: A general population
survey. International Journal ofEating Disorders, 23(3), 287-294.
American Psychological Association. (1994). Diagnostic and statistical manual of mental
disorders. (4th ed.) Washington DC: Author.
Anderson, A.A. (1992). Eating disorders in males: A special case? In K.D. Brownell, J.
Rodin, & J.H. Wilmore (Eds.), Eating, body weight, andperformance in udllete
Disorders of modern society (pp. 172-190). Malvern, PA: Lea & Febiger.
Ashley, C.D., Smith, J.F., Robinson, J.B., & Richardson, M.T. (1996). Disordered eating
in female collegiate athletes and collegiate females in an advanced program of
study: A preliminary investigation. International Journal of Sport Nutrition, 6, 391-
Bane, S., & McAuley, E. (1998). Body image and exercise. In J.L. Duda (Ed.), Advances
in sport and exercise psychology measurement (pp. 311-324). Morgantown, WV:
Fitness Information Technology.
Beals, K.A., & Manore, M.M. (2002). Disorders of the female athlete triad among
collegiate athletes. International Journal of Sport Nutrition, 12, 281-293.
Berry, T.R., & Howe, B.L. (2000). Risk factors for disordered eating in female university
athletes. Journal of Sport Behavior, 23(3), 207-218.
Byely, L., Archibald, A.B., Graber, J., Brooks-Gunn, J. (2000). A prospective study of
familial and social influences on girls' body image and dieting. International
Journal ofEating Disorders, 28, 155-164.
Cash, T.F., & Henry, P.E. (1995). Women's body images: The results of a national survey
in the U.S.A. Sex Roles, 33(1-2), 19-28.
Claessens, A.L., Lefevre, J., Beunen, G., & Malina, R.M. (1999). The contribution of
anthropometric characteristics to performance scores in elite female gymnasts.
Journal of Sports Medicine and Physical Fitness, 39(4), 355-360.
Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155-159.
Costin, C. (1999). The eating disorder sourcebook: A comprehensive guide to the causes,
treatments, and prevention of eating disorders (2nd ed.). Lincolnwood, Illinois:
Cox, L.M., Lantz, C.D., & Mayhew, J.L. (1997). The role of social physique anxiety
another variables in predicting eating behaviors in college students. International
Journal of Sport Nutrition, 7, 310-317.
Cronbach, L.J. (1951). Coefficient alpha and the internal structure of tests.Psychometrika,
Dale, K.S., & Landers, D.M. (1999). Weight control in wrestling: Eating disorders or
disordered eating? Medicine and Science in Sports and Exercise, 1382-1389.
Davis, C. (1992). Body image, dieting behaviors and personality factors: A study ofhigh-
performance athletes. International Journal of Sport Psychology, 23, 179-192.
Davis, C. (1997). Normal and neurotic perfectionism in eating disorders: An
interactivemodel. International Journal ofEating Disorders, 22, 421-426.
Davison, K., Earnest, M., & Birch, L. (2002). Participation in aesthetic sports and girls'
weight concerns at ages 5 and 7 years. International Journal ofEating Disorders,
Demarest, J., & Allen, R. (2000). Body image: Gender, ethnic, and age differences.
Journal of Social Psychology, 140(4), 465-472.
DiBartolo, P.M., Shaffer, C. (2002). A comparison of female college athletes and
nonathletes: Eating disorder symptomatology and psychological well-being.
Journal of Sport and Exercise Psychology, 24(1), 33-41.
Dingemens, A.E., Bruna, M.J., & van Furth, E.F. (2002). Binge eating disorder: A
review. International Journal of Obesity, 26, 299-307.
Fender-Scarr, L. K. (1999). An investigation of athlete, coach, and family variables in
relationship to eating pathology in adolescent female gymnasts. Dissertation
Abstracts International Section A: Humanities & Social Sciences, 59(10-A), 3743.
(UMI No. mtOO10)
Fulkerson, J.A., Keel, P.K., Leon, G.R, & Dorr, T. (1999). Eating disordered behaviors
and personality characteristics of high school athletes and non athletes.
International Journal ofEating Disorders, 26(1), 73-79.
Garfinkle, P.E., & Garner, D.M. (1982). Anorexia nervosa: A multidimensional
perspective. New York: Brunner-Mazel.
Garner, D.M. (1991). Eating Disorder Inventory-2. Professional manual. Odessa, FL:
Psychological Assessment Resources.
Garner, D.M., & Garfinkle, P.E. (1979). The Eating Attitudes Test: An index of the
symptoms of anorexia nervosa. Psychological Medicine, 9, 273-279.
Garner, D.M., Olmstead, M.P., Bohr, Y., & Garfinkle, P.E. (1982). The Eating Attitudes
Test: Psychometric features and clinical correlates. Psychological Medicine, 12,
Garner, D.M., Olmstead, M.P., & Polivy, J. (1983). Development and validation of a
multidimensional eating disorder inventory for anorexia nervosa and bulimia.
International Journal ofEating Disorders, 2, 1534-1545.
Garner, D.M., Rosen, L., & Barry, D. (1998). Eating disorders among athletes: Research
and recommendations. Child and Adolescent Psychiatric Clinics of North America,
George, D. & Mallery, P. (2001). SPSSfor Windows step by step: A simple guide and
reference 1010 update (3rd ed.). Boston, MA: Allyn and Bacon.
Griffin, J. & Harris, M. B. (1996). Coaches' attitudes, knowledge, experiences, and
recommendations regarding weight control. The Sport Psychologist, 10, 180-194.
Grimm, L.G., & Yarnold, P.R. (Eds.). (1995). Reading and understanding multivariate
statistics. Washington, D.C., American Psychological Association.
Haase, A.M., & Prapavessis, H. (1998). Social physique anxiety and eating
attitudes:Moderating effects of body mass and gender. Psychology, Health, &
Medicine, 3(2), 201-210
Haase, A.M., Prapavessis, H., & Owens, R.G. (2002). Perfectionism, social physique
anxiety and disordered eating: A comparison of male and female elite athletes.
Psychology of Sport and Exercise, 3, 209-222.
Halmi, K., Sunday, S., Strober, M., Kaplan, A., Woodside, D.B., Fichter, M., et al.
(2000). Perfectionism in anorexia nervosa: variation by clinical subtype,
obsessionality, and pathological eating behaviors. American Journal ofPsychiatry,
Harris, M.B., & Greco, D. (1990). Weight control and weight concern in competitive
female gymnasts. Journal of Sport and Exercise Psychology, 12(4), 427-433.
Hart, E.A., Leary, M.R., & Rejeski, W.J. (1989). The measurement of social physique
anxiety. Journal of Sport and Exercise Psychology, 11, 94-104.
Hausenblas, H.A. (1998). Eating disorders andadiletew Unpublished doctoral
dissertation, The University of Western Ontario, London, Ontario, Canada.
Hausenblas, H.A., & Carron, A.V. (1999). Eating disorder indices and athletes: An
integration. Journal of Sport and Exercise Psychology, 21, 230-258.
Hausenblas, H.A., & Carron, A.V. (2002). Assessing eating disorder symptoms in sport
groups: A critique with recommendations for future research. International Sports
Journal, winter, 65-74.
Hausenblas, H.A., & Mack, D.E. (1999). Social physique anxiety and eating disorder
correlates among female athletic and nonathletic populations. Journal of Sport
Behavior, 22(4), 502-513.
Hausenblas, H.A., & Symons Downs, D. (2001). Comparison of body image between
athletes and non athletes: A meta-analytic review. Journal ofApplied Sport
Psychology, 13(3), 323-33.
Highet, R. (1989). Athletic amenorrhoea: An update on aetiology, complications, and
management. Sports Medicine, 7(2), 82-108.
James, B., & Collins, D. (1997). Self-presentational sources of competitive stress during
performance. Journal of Sport & Exercise Psychology, 19(1), 17-35.
Jarrett, C. (2001). The road to a college scholarship. USA Gymnastics Magazine, 30(2),
Jarrett, C. (2002). Club v. college gymnastics. Available from USA Gymnastics, Pan
American Plaza, 201 South Capital Avenue, Indianapolis, IN 46225
Johnson, C., Powers, P., & Dick, R. (1999). Athletes and eating disorders: The National
Collegiate Athletic Association study. International Journal ofEating Disorders,
Kirk, G., Singh, K., & Getz, H. (2001). Risk of eating disorders among female college
athletes and nonathletes. Journal of College Counseling, 4(2), 122-132.
Krane, V., Stiles-Shipley, J.A., Waldron, J., & Michalenok, J. (2001). Relationship
among body satisfaction, social physique anxiety, and eating behaviors in female
athletes and exercisers. Journal of Sport Behavior, 24(3), 247-265.
Kurtzman, F.D., Yager, J., Landsverk, J, Weismeier, E., & Bodurka, D.C. (1989). Eating
disorders among selected female student populations at UCLA. Journal of the
American Dietetic Association, 89, 45-53.
Martin, K.A., Rejeski, W.J., Leary, M.R., AcAuley, E., & Bane, S. (1997). Is the social
physique anxiety scale really multidimensional? Conceptual and statistical
arguments for a unidimensional model. Journal of Sport and Exercise Psychology,
Mintz, L.B., O'Holloran, M.S., Mulholland, A.M., & Schneider, P.A. (1997).
Questionnaire for eating disorder diagnosis: Reliability and validity of
operationalizing DSM-IV into a self-report format. Journal of Counseling
Psychology, 44, 63-79.
Nagel, K.L., & Jones, K.H. (1992). Sociological factors in the development of eating
disorders. Adolescence, 27, 107-112.
O'Connor, P.J, Lewis, R.D., & Kirchner, E.M. (1995). Eating disorder symptoms in
female college gymnasts. Medicine and Science in Sports and Exercise, 27, 550-
Otis, C.L., Drinkwater, B., Johnson, M., Loucks, A., & Wilmore. (1997). Position stand:
The female athlete triad. Medicine and Science in Sport and Exercise, 29(5), i-ix.
Petire, T.A. (1996). Differences between male and female college lean sport athletes,
nonlean sport athletes, and nonathletes on behavioral and psychological indices of
eating disorders. Journal ofApplied Sport Psychology, 8, 218-230.
Petrie, T.A. (1993). Disordered eating in female collegiate gymnasts: Prevalence and
personality/ attitudinal correlates. Journal of Sport and Exercise Psychology, 15,
Picard, C. (1999). The level of competition as a factor for the development of eating
disorders in female collegiate athletes. Journal of Youth andAdolescence, 28(5),
Ransdell, L. (1996). Maximizing response rate in questionnaire research. American
Journal of Health Behavior, 20(2), 50-56.
Reel, J.J., & Gill, D.L. (1996). Psychosocial factors related to eating disorders among
high school and college female cheerleaders. The Sport Psychologist, 10(2), 195-
Rock, C.L. (1999). Nutritional and medical assessment and management of eating
disorders. Nutrition in Clinical Care, 2, 332-343.
Rosen, L.W., and Hough, D.O. (1988). Pathogenic weight-control behaviors in female
college gymnasts. Physician and Sportsmedicine, 16, 141-146.
Ryan, J. (1995). Little girls in pretty boxes: The making and breaking of elite gymnasts
and figure skaters. New York, NY: Warner Books Edition.
Sands, W.A., Hofman, M.G., Nattiv, A. (2002). Menstruation, disordered eating
behavior, and stature: A comparison of female gymnasts and their mothers.
International Sports Journal, winter, 1-13.
Scanlan, T.K., Stein, G.L., & Ravizza, K. (1991). An in-depth study of former elite figure
skaters: III. Sources of stress. Journal of Sport & Exercise Psychology, 13(2), 103-
Schlundt, D.G., & Johnson, W.G. (1990). Eating disorders: Assessment and treatment.
Needham Heights, MA: Allyn & Bacon.
Sherman, R.T., Thompson, R.A., & Rose, J.S. (1996). Body Mass Index and
athleticperformance in elite female gymnasts. Journal of Sport Behavior, 19(4),
Smith, M.C., & Thelen, M.H. (1984). Development and validation of a test for bulimia.
Journal of Consulting & Clinical Psychology, 52(5), 863-872.
Smolak, L., Murnen, S.K., & Ruble, A.E. (2000). Female athletes and eating problems: A
meta-analysis. International Journal ofEating Disorders, 27(4), 371-380.
Snyder, E.E. & Kivlin, J.E. (1975). Women athletes and aspects of psychological well-
being and body image. Research Quarterly, 46(2), 191-199.
Stoutjesdyk, D., & Jevne, R. (1993). Eating disorders among high performance athletes.
Journal of Youth andAdolescence, 22, 271-278.
Striegel-Moore, R.H., Silberstein, L.R., & Rodin, J. (1986). Toward an understanding of
risk factors for bulimia. American Psychologist, 41, 246-263.
Strober, M. (1986). Anorexia nervosa: History and psychological concepts. In K.D.
Brownell & J.P. Foreyt (Eds.), Handbook of eating disorders: Physiology,
psychology, and treatment of obesity, anorexia, and bulimia (pp.23 1-246). New
York: Basic Books.
Sundgot-Borgen, J. (1993). Prevalence of eating disorders in elite female athletes.
International Journal of Sport Nutrition, 3, 29-40.
Sundgot-Borgen, J. (1994). Risk and trigger factors for the development of eating
disorders in female elite athletes. Medicine and Science in Sports and Exercise, 26,
Sundgot-Borgen, J., & Larsen, S. (1993). Pathogenic weight-control methods and self-
reported eating disorders in female elite athletes and controls. Scandinavian
Journal of Medicine and Science in Sports, 3, 150-155.
Taub, D.E., & Blinde, E.M. (1994). Disordered eating and weight control among
adolescent female athletes and performance squad members. Journal of Adolescent
Research, 9(4), 483-497.
Terry-Short, L.A., Owens, R.G., Slade, P.D., & Dewey, M.E. (1995). Positive and
negative perfectionism. Personality and individual Differences, 18(5), 663-668.
Thelen, M.H., Farmer, J., Wonderlich, S., & Smith, M. (1991). A revision of the Bulimia
Test: The BULIT-R. Psychological Assessment, 3(1), 119-124.
Thelen, M.H., Mintz, L.B., & Bowman, J.S. (1996). The Bulimia Test-Revised:
Validation with DSM-IV criteria for bulimia nervosa. Psychological Assessment, 8,
Thompson, R. & Sherman, R.T. (1999a). Athletes, athletic performance, and eating
disorders:Healthier alternatives. Journal of Social Issues, 55(2), 317-337.
Thompson, R. & Sherman, R.T. (1999b). "Good athlete" traits and characteristics of
anorexia nervosa: Are they similar? Eating Disorders: The Journal of Treatment
and Prevention, 7(3), 181-190.
Tyrka, A.R., Waldron, I., Graber, J.A., & Brooks-Gunn, J. (2002). Prospective predictors
of the onset of anorexic and bulimic syndromes. International Journal of Eating
Disorders, 32, 282-290.
USA Gymnastics. (1995). Task force on USA Gymnastics response to the female athlete
triad. Available from USA Gymnastics, Pan American Plaza, 201 South Capital
Avenue, Indianapolis, IN 46225.
Walsh, J.M., Wheat, M.E., & Freund, K. (2000). Detection, evaluation, and treatment of
eating disorders: The role of the primary care physician. Journal of General
Internal Medicine, 15, 577-590.
Warren, B.J., Stanton, A.L., & Blessing, D.L. (1990). Disordered eating patterns in
competitive female athletes. International Journal ofEating Disorders, 9(5), 565-
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(2), 129-143.
Williamson, D.A., Anderson, D.A., Jackman, L.P., & Jackson, S.R. (1995). Assessment
of eating disordered thoughts, feelings, and behaviors. Handbook of assessment
methods for eating behavior and weight-related problems: Measures, theory, and
research. Thousand Oaks, CA: Sage.
Williamson, D.A., Netemeyer, R.G., Jackman, L.P., Anderson, D.A., Funsch, C.L., &
Rabalais, J.Y. (1995). Structural equation modeling of risk factors for development
of eating disorder symptoms in female athletes. International Journal of Eating
Disorders, 17(4), 387-393.
Wilmore, J.H. (1991). Eating and weight disorders in the female athlete. International
Journal of Sport Nutrition, 1, 104-117.
Wood, A., Waller, G., & Gowers, S. (1994). Predictors of eating psychopathology in
adolescent girls. European Eating Disorders Review, 2, 6-13.
Zucker, N.L., Womble, L.G., Williamson, D.A., & Perrin, L.A. (2001). Protective factors
for eating disorders in female college athletes. Eating Disorders: The Journal of
Treatment and Prevention, 7(3), 207-218.
Originally from New Jersey, I obtained a B.S. in psychology from James Madison
University and a M.S. in Exercise and Sport Sciences from the University of Florida. My
academic concentration was in sport and exercise psychology and my research interests
include exercise adherence, as well as eating attitudes and behaviors of female athletic
populations. Specifically, I am interested in developing and promoting healthy sport
environments for young female athletes.
I have been an instructor for college-level conditioning, tennis, and gymnastics. I
have also worked as a gymnastics coach for both the developmental and high school