High and low scorers were determined for one hundred male and female gym users from Leeds are (43 male, 57 female) aged 26 ± 7.93 (mean ± SD) years, height 1.73 ± 0.11m, weight 71.7 ± 13.23 Kg, on 3 Eating Disorder Inventory (EDI) subscales: drive for thinness, bulimia and body dissatisfaction.  These EDI subscale scores were compared with additional questions that assessed the participant’s physical activity behaviour and food avoidance.  Results showed that females scored highly on the 3 EDI subscales whereas males typically scored low (p = £0.05).  Aggregated data revealed that regardless of gender high scorers on each EDI subscale generally participated in more endurance exercise for longer duration’s of time, avoided confectionery and chocolate for reasons of weight loss and attempted dieting more frequently than low scorers.  In conclusion although the present research suggest that the participants were not suffering from a clinically diagnosed eating disorder per se, high scorers repeatedly displayed characteristics associated with less documented eating disorders (Anorexia athletica, atypical & muscle dysmorphia) and avoided certain foods with the purpose of weight loss. 

 

Although there is extensive research and information available regarding eating disorders and physical activity behaviour, there are very few studies that look at the connection between the two (Davis, Fox, Cowles, Hastins & Schwass, 1990; Thompson & Sherman, 1993; Davis, Kennedy, Ravelski & Dionne, 1994; Brewerton, Stellefson, Hibbs, Hodges & Cochrane, 1995: Olson, Williford, Richards & Pugh, 1996; Zmijewski & Howard, 2003).

Previous research on eating disorders has often focused on those who fulfil the diagnostic criteria of either anorexia or bulimia nervosa.  However, it appears that ‘atypical’ eating disorders are often being neglected despite the fact they are just as common as anorexia or bulimia nervosa (Fairburn & Walsh, 2000).

Substantial research does also exist linking elite athletes with eating disorders; unfortunately limited data is offered for those involved in any others forms of regular exercise. Additionally as much of the available research concentrating primarily on females (Andersen, Cohn & Holbrook, 2000) the need for further research is evident.

Overweight and obesity are on the increase (Thompson & Sherman, 1993; Fairburn & Wilson, 1993; Andersen et al, 2000; Fairburn and Bromwell, 2002, Ogden, 2003).  Consequently dieting tends to be a frequent option in today’s society.  The Department of Health (DoH) reported in 2002 that the prevalence of obesity has now reached 21% in both males and females in England (Health Survey for England 2000).  Ogden (2003) stated that over the past few years there has been a dramatic rise in interest of all aspects of diet, from healthy eating through to eating disorders; due to this epidemic disordered eating is becoming more common.

Palmer (1996, cited in the Eating Disorders Association UK (EDA) web-site, 2004) stated that the best estimates in the UK suggest approximately one young woman in a hundred has bulimia nervosa and probably somewhat fewer have anorexia nervosa.  An estimate of the numbers of men in the eating disordered population varies with the criteria adopted and is thought to vary between 1:6 and 1:201. The most commonly used proxy is 10% of all cases of eating disorders will be male. (EDA 2000).  However Anred (2000a) claims that as the study of eating disorders is a relatively new field so statistical data may be limited

The two best known eating disorders anorexia and bulimia nervosa were classified in 1970 and 1979 (for clinical criteria of anorexia and bulimia nervosa see Tables 1 & 2: Appendices A & B).  In 1980 the existence of atypical eating disorders was diagnosed.  Fairburn and Walsh (2000) referred atypical as eating disorders other than anorexia or bulimia (for clinical criteria see Tables 1 and 2: Appendix C).  Most recently two new eating disorders have been distinguished – namely ‘night eating syndrome’ and ‘binge eating disorder’ – however; diagnostic status still remains subject of debate (Garfinkel, 2002).

Prevalence of newly recognised eating disorders is a frequent focus for new research.  Two of the most relevant unofficial eating conditions with respect to the present study are anorexia athletica and muscle dysmorphia (Anred, 2003b).  Anorexia athletica commonly known to researchers as obligatory exercise sees individual’s addicted to exercise, which is often running or performing another intense aerobic routine.  The person feels obliged to pursue an excessive exercise routine, in spite of injury, damaged relationships, and too much time taken off work in service of this activity (Anred, 2003b).  Muscle Dysmorphia is currently conceptualized as a form of body dysmorphic disorder (BDD), but does have similarities to eating disorders as reflected by its previous name ‘reverse anorexia’ (Phillips, 2002).  People with this disorder are obsessed by being too small and underdeveloped.  Both men and women have been seen to suffer from muscle dysmorphia, however, researchers suspect the disorder is more common in males since the culturally defined ideal male is big and strong while the ideal female is small and thin (Anred, 2003b). 

Research overlooking ‘atypical’ eating disorders is a major source of concern for two main reasons.  Firstly, focusing exclusively on anorexia and bulimia nervosa has lead researchers to fail in viewing eating disorders in their entirety.  Secondly, by only providing treatment for those who have been clinically diagnosed with anorexia or bulimia nervosa has left the needs of a large population of patients being ignored (Fairburn & Walsh, 2002).

Several studies (Thompson & Sherman, 1993; Andersen et al, 2000; Ogden, 2003; Andersen & Bulik, 2003; Anred 2003b) have indicated that at least 90% of those with eating disorders are female, however, recently it has been argued that this may be solely due to cases being underreported in males (Thompson & Sherman, 1993).  Andersen and colleagues (2000) believe it would not be surprising if closer to 25-30% of suffers were male, but there are only hints from early scientific data to confirm this figure.  Researchers (Thompson & Sherman 1993; Andersen et al, 2000; Anderson & Bulik, 2003; Anred, 2003a; Engel, 2003; Ogden, 2003) often suggest males with eating disorders may tend to use excessive exercise rather than strict dieting or other weight loss methods such as diuretics, diet pills or laxatives.  As a result, it may be more difficult for a male to recognise that he actually has an eating disorder (Thompson & Sherman, 1993).  Mangweth and colleagues (2001), cited by Anderson and colleagues (2000), stated that the function of excessive exercise in males is primarily to increase weight rather than reduce it (Muscle Dysmorphia).

It is thought that athletes involved in ‘thinness demand’ sports would be at higher risk from developing eating disorders.  For example in several studies (Borgen & Corbin, 1987; Davis and Cowles’ 1989; Thompson & Sherman, 1993, Smolak et al 2000; Phillips, 2002; Engel 2003), these athletes showed greater tendencies towards eating disorders, however, research in this area is inconclusive.  Thompson and Sherman (1993) suggested that the notion that participating in sport and exercise can cause an eating disorder could probably be attributed to two factors.  The first factor results from high levels of physical activity or exercise can cause an eating disorder to develop (Davis et al, 1990).  Davis and colleagues research concluded that regular participation in a fitness program might be fundamentally related to excessive concern with weight and diet.    The second factor involves pressures for the particular sport or by particular coaches (Thompson & Sherman, 1993).

Nevonen and Broberg (2001) found that generally women with eating disordered problems did score significantly higher on Eating Disorder Inventory (EDI) compared to normal controls. A study by Oslon et al. (1996) revealed that a number of aerobic instructors possessed scores on the EDI similar to those who have eating disorders.  However it must be taken into account that existing physiological screening tests that detect eating disorders could be considered biased towards women; hence suggesting another possible reason for the low level of reported eating disorders in males.  Because of this more and more medical researchers are now recognising the need for independent studies for both male and females for every disorder instead of assuming that studies based on one gender are applicable to the other (Andersen, 2000).

Encouragement or demand for reduced weight or body fat has become quite popular in sports circles today (Thompson et al, 1993).  Studies have consistently shown that individuals involved in sport or high levels of physical activity show increased prevalence of body dissatisfaction, a greater pre-occupation with weight and body shape and more disordered patterns of eating compared to the general population (Davis, 1994; Brewerton, 1995; Olson, 1996; Bryne 2002, Bean, 2003; Engel 2003).

Body dissatisfaction (See Fig 1) could be one of the factors leading to dieting.  Most dieters are seen to be women, however research has claimed that the majority of the population attempts to restrain their food intake at some point in their life (Ogden, 2003).

  

Fig 1: Body Dissatisfaction

Source: adapted from Ogden, 2003, Pg. 85

Researchers cited in Fairburn and Wilson (1993) suggested that it seems dieting either contributes to, or at the very least exacerbates binge eating.  A consensus also exists indicating strict dieting and weight loss usually constitutes an early step in the development of both anorexia and bulimia nervosa.  A more controversial angle of research has focused on the extent to which dieting per se raises the chances of developing an eating disorder (Lowe, 2000).  Polivy and Herman (2000) believed that individuals who restrain their food intake would have lower self-esteem than unrestrained eaters would.  Furthermore, the former will tend to report being more anxious and neurotic.  Restrained eaters also have shown to have unrealistic expectations about self improvements following weight loss, and expect eating to decrease negative affect.  In addition Bean (2003) claimed that it is as much about attitude and behaviour towards food as it is about consumption of food.

Fedroff, Polivy & Herman (2003) found that exposure to food cues increased disordered eating especially in restrained eaters.  According to the American Psychiatric Association (APA, 1994) conditions in which cognitive or environmental constraints have been implemented so there is limited access to certain food are main features of anorexia and bulimia nervosa.  Specific food restrictions (e.g. high fat and/or sugar foods) and food deprivation are also central components of dietary treatments for obesity (Raynor & Epstein, 2002).  However, Corwin (2000) claims restriction and/or deprivation may actually contribute to overeating, hence reducing the effectiveness of obesity treatment.     Food restriction and/or deprivation do appear to affect current eating behaviour in humans (Polivy et al, 2003).  For example, men who had reduced intake to about 1500 kcal/day for 24-weeks reported problems with uncontrollable eating increased appetites for sweet food even after weight gain occurred (Polivy et al, 2003).

Many athletes or individuals participating in high levels of physical activity tend to be very careful about what they eat and often experiment with various dietary programmes in order to enhance performance.  However, there is a thin line between paying attention to detail and obsessive eating behaviours (Bean, 2003).  In summary Bean (2003) claims that intense and excessive training programmes combined with food restriction may lead to an obsessive preoccupation with body weight and calorific intake and eventually eating disordered eating.

For the present study it is hypothesised that participants (regardless of gender) who scored highly on any of the 3 subscales of the EDI (drive for thinness, bulimia and body dissatisfaction) will participate in greater amounts of physical activity and/or avoid certain (especially high energy density) foods and/or diet more than participants who score low on any the mentioned EDI subscales.

 

One hundred gym users from Leeds area (43 male, 57 female) aged 26 ± 7.93 (mean ± SD) years, height 1.73 ± 0.11 m, weight 71.7 ± 13.23 Kg (See appendix J)

In total 112 self-report questionnaires were distributed to gym users randomly selected across three health clubs situated in and around Leeds City Centre; 12 questionnaires were discarded from the study due to being incomplete leaving 100 questionnaires in total to be analysed for the final results.  Consent from every participant was obtained prior to the study (See appendix E).

Disordered eating and physical activity behaviour among male and female gym users was assessed via self-report questionnaire (See appendix H).  This included 3 of the 8 subscales from the EDI (Garner et al., 1983) and a set of additional questions regarding individuals physical activity behaviour and food avoidance (See appendix H).

The Eating Disorder Inventory (EDI; Gardner, Olmsted & Polivy 1983) is a 64-item, 6-point forced choice inventory that assesses several behavioural and psychological traits common in eating disorders.  The EDI is especially recommended as a screening instrument to detect at risk populations (Nevonen & Broberg.  2001).  The EDI consists of 8 subscales: drive for thinness, bulimia, body dissatisfaction, ineffectiveness, perfectionism, interpersonal distrust, and interceptive awareness and maturity fears.  For this study, 3 of the 8 subscales were utilised namely drive for thinness, body dissatisfaction and bulimia.  The 3 chosen subscales were utilised as they evaluate attitudes, and behaviours concerned with eating, weight and bodyshape whereas the other 5 subscales primarily relate to general psychological traits.  It has been used successfully in numerous research studies regarding eating disorders (Olson et al, 1996; Iorio, Margiotta, D’Orsi, Bellini & Boschi, 2000; Jones, Bennet, Olmsted, Lawson & Rodin, 2001; Nevonen & Broberg, 2001 Barry, Griol & Masheb, 2003; Engel, Johnson, Powers, Crosby, Wonderlich, Wittrock & Mitchell, 2003).

As with all self-report measures consideration needs to be taken regarding the legitimacy of the responses.   A study performed by Engel et al (2003) found that self report measures resulted in considerable under reporting of disordered eating symptoms in athletes.  Many researchers (Thompson & Sherman, 1993; Fairburn & Wilson, 1993, Andersen al, 2000; Fairburn & Bromwell, 2002; Ogden 2003) have suggested that a considerable amount of suffers of eating disorders (especially athletes) are often in denial.  Answers given on a self-report measure may reflect this accordingly.  Andersen and colleagues (2000) criticise the use of the EDI for another reason.  They suggest such psychological tests for assessing the prevalence of eating disorders are biased towards diagnosing women feeling most men may find it humorous when they are asked typical questions such as “I think my thighs are too large” or “I like the shape of my buttocks”.  These factors have been reviewed in more detail in the Discussion section.

Additional questions utilised in the questionnaire accessed physical activity behaviour and food avoidance.  The closed structure of the additional question was intended to ensure participants did not feel obligated to respond in a certain pattern that may have shaped ideal results.  Initially, gender, age, weight and height were established, followed by 4 closed questions that assessed time, duration, preferred activity and main reasons for attending a gym (weight loss, general fitness, health benefits, enjoyment, weight gain, recommended by other and other).  Finally three closed questions assessed food avoidance (if any), reason for this avoidance and whether any mainstream diets had been tried past or present (See appendix F).  The foods and beverages chosen for assessment of avoidance were based upon categories primarily reduced or removed by recommendation from usual eating patterns when pursuing one of the mainstream ‘fad’ diets.  Food categories given for avoidance were dairy, confectionery, starches, chocolate, alcohol, fizzy drinks, meat and other.

Initially, the questionnaire was piloted to 2 female and 2 male participants; here any necessary changes could be made prior to the official study.  Subsequent changes made to the questionnaire were minor grammatical alterations.   Participants were required to offer consent prior to completion of the questionnaire (See appendix D).

The scores from the 3 subscales of the EDI were calculated via a specifically designed computer program.  Participants regardless of gender* were divided into different sub-groups of high and low scorers based on the cut-off scores for the 3 subscales of the EDI.  The cut-off scores for each of the 3 subscales were based upon the median value. Clinical sample cut-off scores could not be utilised for the present study due to the small sample size.  Median cut off scores for each subscale was as followed: drive for thinness 2, body dissatisfaction 6 and bulimia 0. Chi Square was employed (utilising computer program SPSS 11.0) to compare frequencies of food avoidance and physical activity behaviour inside the two sub-groups (it is expected that those who scored high on EDI will avoid more foods and participate in larger amounts of physical activity than those who scored low).  Statistical significance was limited for the present study; therefore it was often necessary to apply aggregated data.

* It would have been valuable to verify the gender distribution in the 3 subscales for food avoidance and physical activity however, due to the small sample size for the gender, results would have resulted in non significant.

 

Table 1: Gender characteristics of sample according to scoring on the 3 EDI subscales

	Drive for Thinness Low                High (n=53)          (n=47)	Bulimia Low                High (n=59)          (n=41)	Body Dissatisfaction Low                High (n=51)          (n=49)
Gender           Male                     (n=43)	34                         9	35                         8	32                       11
Female                     (n=57)	19                       38	24                       33	19                       38

Gender highlighted significant difference between high and low scorers on all 3 subscales.  Prevalence of female high scorers was significantly higher than that of male (See Figures 2 - 4; Table 1).   As stated in data analysis, the following results will be exposed regardless of gender distribution.

Fig 2: Number of male and females for high (n=47) and low (n=53) scorers on drive for thinness (p=0.000)

Fig 3: Number of male and females for high (n=49) and low (n=51) scorers on body dissatisfaction (p=0.000)

Fig 4: Number of male and females for high (n=41) and low (n=59) scorers on bulimia (p=0.000)

No significance was shown between high and low scorers across any of the 3 subscales for BMI (See appendix J; Table 2).

Table 2: BMI characteristics of sample according to the 3 EDI subscales

	Drive for Thinness Low                High (n=53)          (n=47)	Bulimia Low                High (n=59)           (n=41)	Body Dissatisfaction Low                High (n=51)          (n=49)
BMI classification            Underweight	1	1	1
Normal	30                       31	36                       25	30                       31
Overweight	18                       15	18                       15	18                       15
Obese 1	18                       15	3	2
Obese 2	2	2	1

Number of gym sessions There was no significance between high and low scorers on the 3 subscales for the number of gym session (See appendix J; Table 2).  However, when aggregating the data for 0 to 4 times a week a more than 5 sessions, high scorers on drive for thinness attended more than 5 sessions (X²₍₁₎ =6.4, p = 0.014) while low scorers frequented the gym less times (See Fig 5).

Fig 5: Number of gym sessions per week for high (n=47) and low (n=53) scorers on drive for thinness (p = 0.014)

            Despite the Lack of significance between data it was observed that on each subscale high scorers generally attended the gym more times a week than low scorers (See Fig 6).

Fig 6: Gym attendance for high and low scorers on each of the 3 EDI subscales

Reasons for attending the gym

Aggregated data exposed general fitness as being the main reason given for attending the gym.  However high scorers on each subscale also gave weight loss as one of their main reasons for attending the gym (See Appendix M).

Average duration of gym session

Significance was found between high and low scorers on bulimia (X²₍₁₎ =6.0, p = 0.048).  Low scorers on bulimia participated in longer gym session, while high scorers spend less time (See Fig 7).

Fig 7: Average duration of gym sessions for high (n=41) and low (n=59) scorers on bulimia (P = 0.048)

No significance was found between high and low scorers on the 2 remaining subscales (See appendix J; Table 3).  However observed data indicated high scorer’s duration of gym session is generally longer than that of low scorers (See Fig 8).

Fig 8: Duration of gym sessions for high and low scorers on the 3 subscales of the EDI Most frequent gym activity

There was no significance between high and low scorers on the 3 subscales for the most frequent activity participated in within the gym (See appendix J; Table 4).  However, when aggregating the data for aerobic training and other activities, high scores on body dissatisfaction participated in more aerobic training (X²₍₁₎ =5.9, p = 0.021) while low scorers participated in a range of activities, such as fitness classes and resistance training (See Fig 9). 

Fig 9: Aerobic training compared with all other activities for high (n=49) and low (n=51) scorers on body dissatisfaction (p=0.021) Food avoidance No significance was found between high and low scorers on the 3 subscales with respect to food avoidance (See appendix K).  However, when aggregating the data there were more high scorers on body dissatisfaction than low scorers who avoided confectionery with the purpose of losing weight (X²₍₁₎ =6.1, p = 0.017)(See Fig 10).

Fig 10: All reasons (except weight loss) compared with weight loss for high (n=49) and low (n=51) scorers on body dissatisfaction (p=0.017)

Although no significance was found between high and low scorers on the 3 subscales generally those who scored highly did avoid food more than the low scorers (See Fig 11, 12 & 13).

Fig 11: Food avoided by high (n=41) and low (n=59) for scorers on bulimia. 

Fig 12: Food avoided by high (n=49) and low (n=51) for scorers on body dissatisfaction

Fig 13: Food avoided by high (n=47) and low (n=53) for scorers on drive for thinness

Aggregated data for individuals who had attempting at least one ‘fad’ diet compared to those that had never dieted highlighted high scorers on bulimia subscale had attempted to diet (X²₍₁₎ =6.5,p = 0.011) while low scorers had diet less frequently. There was no significance between high and low scorers on the other 2 subscales, however is can be observed that high scorers had attempted at least one diet more than low scorers on all 3 subscales (See Fig 14; Appendix L).

Fig 14: Prevalence of attempting at least one diet compared to that of none for high and low scorers on all the 3 EDI subscales. High (n-=41) and low (n=59) scorers on bulimia revealed significance (p=0.011)

 

Results indicate that the percentage of females who scored highly on the 3 EDI subscales was significantly higher than the percentage of males (See Figures2 – 4; Table 1).  This finding supports the general belief that females are more susceptible to suffer from disordered eating than males.  Andersen and colleagues (2000) suggested that existing psychological screening tests for detecting eating disorders (Such as the EDI) could be considered biased towards women. An alternative explanation for the low prevalence of eating disorders in males could be that newly recognised eating disorders predominately associated with males  (namely anorexia athletica and muscle dysmorphia) are not currently assessed by any existing psychological screening test.  Men may not necessarily be preoccupied with the desire to be smaller and thinner but actually the desire to be bigger and more defined (Anred 2003b).  The EDI would only assess such responses via the body dissatisfaction subscale.

Studies (Oslon et al, 1996; Iorio et al 2000; Jones et al 2001; Nevonen & Broberg, 2001) utilising the EDI have generally found that females with problems associated with eating disorders did score significantly higher compared to normal controls.  Numerous researchers (Thompson & Sherman 1993; Andersen et al, 2000; Anderson & Bulik, 2003; Anred, 2003a; Engel, 2003; Ogden, 2003) have suggested that males may tend to use excessive exercise sooner than strict dieting or other weight loss methods such as purging, laxatives or diuretics.  As a result detecting males with eating disorders could be increasingly difficult.   To offer support to the above findings it would be necessary to concentrate solely on male’s reasons for pursuing exercise.  However time constraints surrounding the present study made this unfeasible.  Future research would therefore be beneficial.

Researchers (Thompson & Sherman, 1993; Andersen et al, 2000; Ogden, 2003; Andersen & Bulik, 2003; Anred 2003b) have broadly stated that at least 90% of those suffering with eating disorders are female.  However a growing consensus exists that this may be solely attributed to cases being underreported in males.

On each EDI subscale prevalence of attempted diets was elevated in those individuals who scored highly.  Low scorers were shown to have attempted fewer diets in comparison (See Fig 14; Appendix M).  Ogden (2003) feels body dissatisfaction is the main consequence of dieting.  Measurements of body dissatisfaction can be considered in terms of three different perspectives (See Fig 1).  In particular bulimics have often been reported to having heightened levels of body dissatisfaction (Beumont, 2002; Barry 2003; Ogden, 2003).  In addition Polivy and Herman (2002) stated that the literature on the effects of dieting suggest that while society may encourage and congratulate weight loss, dieting has a variety of associations and consequences that imply it is less than desirable.

Aggregated data exposed significance between high and low diet scores for the bulimia subscale.  High scorers had generally attempted at least one diet, whereas low scores had dieted less frequently (See Fig 14).  Ogden (2003) stated that suffers of bulimia nervosa engage in compensatory behaviour as a means to manage any weight caused by binges.  The most common form is self-induced vomiting.  However dieting could be considered as a less severe form of compensatory behaviour for those who have bulimic tendencies.  A more controversial angle could suggest use of dieting is an initial step towards the development of bulimia nervosa.  Lowe (2000) stated that a general consensus exists that dieting and weight loss usually constitutes an early step in the development of both anorexia and bulimia nervosa.  Thompson and Sherman (1993) add that the individual who will develop bulimia nervosa begins with dieting in an effort to lose weight.

Regardless of statistical significance it was observed that for whatever reason high scorers on all 3 subscales generally did avoid certain foods more than the low scorers (See Figures 11 - 13).  Ogden (2003) suggested that due to a dramatic rise in all aspects of diet disordered eating is becoming more common.  In addition, Corwin (2000) stated that restriction and/or deprivation may actually contribute to overeating.  Intention to restrict certain foods in order to lose weight may therefore prove to have an undesired affect.  Lowe (2000) however, stated that one study found that most restrained eaters assessed via Herman and Polivys 10-item Restraint Scale (RS) are not necessarily dieting to lose weight.  Collective data highlighted that for high scorers on body dissatisfaction weight loss was given as the main reason for avoiding confectionery.  Low scorers offered a variety of responses (See Fig 10).  Polivy and Herman (2000) believed that individuals who restrain their food intake would have lower self-esteem and increased body dissatisfaction compared to unrestrained eaters.

Interestingly high scorers on the bulimic subscale actually avoided less confectionery and chocolate than low scorers.  This could be attributed to sweet high-fat foods being the main food source eaten by a bulimic during a binge.  Binge foods such as fatty sweet high-energy foods that patients deny themselves publicly are selected because they are easy to swallow and regurgitate (Beumont, 2002).

Aggregated data revealed significance in relation to gym attendance for high and low scorers on the drive for thinness subscale.  High scorers primarily attended the gym more than five times per week, whilst low scores frequented the gym less often (See Fig 5).  Overall it was observed that generally high scorers for each subscale did attend the gym more frequently compared to low scorers (See Fig 6).  Thompson and Sherman (1993) suggest outside the sport environment, others may criticise an individual for spending too much time exercising – this same individual is much less likely to be criticised in a sport environment that either implicitly or explicitly communicates the necessity of excessive exercise.  Unfortunately it is this acceptance that helps the individual to hide their disorder making it increasingly difficult to identify. Mangweth and colleagues (2001) stated that with respect to males the role of excessive exercise is predominantly to increase weight rather than reduce (Muscle dysmorphia).

High scorers average duration of a gym session for bulimia was statistically less than for the low scorers (See Fig 7).  Many may feel this could be an error in results.  However Beumont (2002) suggested that bulimics would typically use induced vomiting as the process of undoing a binge.  Exercise and other reported means (laxatives, diuretics) generally fall secondary to this.  In addition Thompson and Sherman (1993) stated that the bulimic is usually able to be less compulsive about exercise.  On the remaining 2 subscales the majority of cases highlighted that high scorer’s duration of an average gym session was longer than for low scorers (See Fig 8).  Numerous researchers (Thompson & Sherman, 1993; Fairburn & Wilson, 1993; Fairburn & Bromwell, 2002; Ogden 2003) have stated that individuals suffering with an eating disorder will often use exercise as well as diet as a method of controlling their weight.

Collective data revealed high scorers on body dissatisfaction gave aerobic training as their main activity to engage in at the gym whereas low scorers typically gave a range of activities.  Beumont (2002) claimed that sufferers of eating disorders would exercise deliberately to burn calories to induce weight loss.  For some the activity may be surreptitious while for others it may be strenuous physical exercise usually in the form of aerobic classes or running.  Endurance sports or activities that require endurance training are apt to be attractive to someone who over values the use of exercise or who uses exercise as means of weight loss (Thompson & Sherman, 1993).

Phillips (2002) stated that individuals preoccupied with their body typically focus excessively on body fat, diet and exercise.  In agreement Ogden (2003) suggested that the consequences of body dissatisfaction in women are to diet, exercise and in extreme cases resort in plastic surgery.  Whereas in men typically sees them opting for exercise as a solution to their problem.  It was decided that due to time constraints and the small sample size gender distribution would not be looked at. However, to extent research in the future it would be valuable to focus upon gender distribution.  Stice (2002) additionally stated that exposure to media-potrayed thin ideal images results in increased body dissatisfaction and negative effects (e.g. depression, shame & anger).

When only one reason was given collective data highlighted that general fitness was the main reason for attending the gym.  However, when more than one reason was given the vast majority of high scorers on each subscale gave weight loss (See appendix L).  Thompson and Sherman (1993) suggested that as body weight decreases the motivational value of exercise increases. As previously discussed researchers (Thompson & Sherman 1993; Andersen et al, 2000; Anderson & Bulik, 2003; Anred, 2003a; Engel, 2003; Ogden, 2003) have frequently stated that exercise is often used to induce weight loss.

Clinical cut off points for the 3 EDI subscales could not be utilised as presently none exist for both clinical and control British populations.  Thus, Cut off points for the present study, were based upon the median score.   Only a small sample size was assessed due to time constraints of the study.  Consequently significant results obtained from the sample were limited.  Lack of statistical significance made it necessary to sometimes focus on aggregated data.  A larger sample size could have increased the probability of producing further quantitative results.  Furthermore, self-report measures are often criticised since there is no guarantee that response given will be legitimate.  With reference to eating disorders numerous researchers (Thompson & Sherman, 1993; Fairburn & Wilson, 1993, Andersen al, 2000; Fairburn & Bromwell, 2002; Ogden 2003) have suggested that those suffering will often be in denial.  Refusal to acknowledge one is suffering with disordered eating patterns or excessive exercise may influence response given on a self-report measure.   Reduced validity of results could be a consequence.  A larger sample size and use of clinical cut off points could have produced more conclusive results.

In accordance to literature more women from the present study than men scored females highly on the 3 EDI subscales. Further investigation into gender would have been valuable.  Statistical significance was limited therefore in many instances aggregated data had to be applied.

 

Conclusion

Although the present research suggest that the participants were not suffering from a clinically diagnosed eating disorder per se, high scorers repeatedly displayed characteristics associated with less documented eating disorders (Anorexia athletica, atypical & muscle dysmorphia) and avoided certain foods with the purpose of weight loss.

Andersen, A., Cohn, L. and Holbrook, T. (2000) Making weight – Men’s conflicts with food, weight, shape and appearance. USA:Gurze books.

Anderson, C.B. and Bulik, C.M. (2003) Gender differences in compensatory behaviors, weight and shape salience, and drive for thinness.  Journal of Eating Behaviors. [ONLINE]

Available: http://www.sciencedirect.com. 22/11/03

Anred (2003a) Statistics: How many people have eating disorders?  Accessed on the 21/20/03 from the worldwide web: http://www.anred.com/stats.html.

Anred (2003b) less well known eating disorders and related problems.  Accessed on the 21/10/03 from the worldwide web: http://www.anred.com/deflesser.html.

Barry, D.T., Griol, C.M., and Masheb, R.M. (2003).  Comparison of patients with bulimia nervosa, obese patients with binge eating disorder and non-obese patients with binge eating disorder [Abstract].  J Nerv Ment Dis. (9):589-94.  Accessed on the 9/11/03 from the worldwide web: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=…

Beumont, P.J.V. (2002) Clinical presentation of Anorexia and Bulimia Nervosa.  In Fairburn and Brownell (eds) Eating Disorders and Obesity: a comprehensive handbook. (Pg 162-170) London:Guilford.

Brewerton, T.D., Stellefson, E.J., Hibbs, N., Hodges, E.L and Cochrane, C.E (1995) Comparison of eating disorder in patients with and without compulsive exercising [Abstract].  International Journal of Eating Disorders 17(4):413-6. Accessed on the 16/10/03 from the worldwide web: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=…

Byrne, S.M. (2002).  Sport, occupation and eating disorders.  In Fairburn and Brownell (eds) Eating Disorders and Obesity: a comprehensive handbook. (Pg 256-257) London:Guilford.

Davis, C., Fox, J., Cowles, M., Hastins, P. and Schwass, K. (1990).  The functional role of exercise in the development of weight and diet concerns in women [Abstract].   Journal Psychosom. Res. 34:563-74.  Accessed on the 16/10/03 from the worldwide web: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=…

Davis, C., Kennedy, S.H., Ravelski, E., and Dionne, M. (1994).  The role of physical activity in the development and maintenance of eating disorders [Abstract].  Psychol Med 24(4):957-67.  Accessed on the 16/10/03 from the worldwide web:

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=…

Department of Health (2002), Specialised Services National Definition Set: 35 Specialised morbid obesity services (all ages).  Accessed on the 04/05/04 from the worldwide web: http://www.dh.gov.uk/PolicyAndGuidance/ HealthAndSocialCareTopics/SpecialisedServices
Definition/SpecialisedServicesDefinitionArticle/fs/en?CONTENT_ID=4001826&chk=a017Ao

Eating Disorder Association UK (2000) Some statistics.  Accessed on the 04/05/04 from the worldwide web: http://www.edauk.com/sub_some_statistics.htm

Eating Disorder Inventory (2003) Eating Disorder Inventory.  Accessed on the 28/20/03 from the worldwide web: http://www.cps.nova.edu/~cppphelp/EDI.html

Engel, S.G., Johnson, C., Powers, P.S., Crosby, R.D, Wonderlich, S.A., Wittrock, D.A., and Mitchell, J.E. (2003).  Predictors of disordered eating in a sample of elite Division I college athletes.  Eating Behaviors 4:333-343

 Fairburn, C.G. and Walsh, B.T. (2002) Atypical eating disorders (Eating Disorder Not Otherwise Specified).  In Fairburn and Brownell (eds) Eating Disorders and Obesity: a comprehensive handbook. (Pg 171-177) London:Guilford.

Fairburn, C.G and Wilson, T (1993) Binge eating: nature and assessment. London:Guilford

Fedoroff, I., Polivy, J., and Herman, P. (2003) The specifity of restrained versus un restrained eaters’ responses to food cues: general desire to eat, or craving for cues food? Appetite 41:7-13

Garfinkel, P.E. (2002). Classification and Diagnosis of Eating Disorders.  In Fairburn and Brownell (eds) Eating Disorders and Obesity: a comprehensive handbook. (Pg. 155-161) London:Guilford.

Iorio, D., Margiotta, N., D’Orsi, P., Bellini, O., and Boschi, V. (2000) The Eating Disorder Inventory in evaluation of impaired eating behaviour in subjects request nutritional consultation [Abstract].  Eat Weight Disord 5(4):206-10. Accessed on the 9/11/03 from the worldwide web: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=…

Jones, J.M., Bennett, S., Olmsted, M.P., Lawson, M.L., and Rodin, G. (2001) Disordered eating attitudes and behaviours in teenaged girls: a school based study.  CMAJ 165(5):547-52.

Nevonen, L. and Broberg, A.G. (2001) Validating the Eating Disorder Inventory-2 (EDI-2) in Sweden [Abstract].  Journal of Eating Behaviors. 6:59-67.  Accessed on the 30/10/03 from the worldwide web: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=…

Ogden, J. (2003) The psychology of eating: from healthy to disordered behavior. Oxford:Blackwell.

Olson, M.S., Williford, H.N., Richards, L.A. and Pugh, S. (1996) Self-reports on eating disorders inventory by female aerobics instructors [Abstract].  Percep Mot Skills. 82:1051-8.  Accessed on the 16/10/03 from the worldwide web:

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=…

Phillips, K.A. (2002). Body image and Body Dysmorphic Disorder.  In Fairburn and Brownell (eds) Eating Disorders and Obesity: a comprehensive handbook. (Pg. 113-120) London:Guilford.

Polivy, J., and Herman, C.P. (1993) Etiology of binge eating: psychological mechanisms.  In C.G. Fairburn & G.T Wilson (eds).  Binge eating: Nature and assessment and treatment (Pg. 173-205).  New York: Guilford Press.

Raynor, H.A., & Epstein, L.H. (2002) The relative-reinforcing value of food under differing levels of food deprivation an restriction.  Appetite 40:15-24

Smolak, L., Murnen, S,K., & Ruble, A.E., (2000) Female athletes and eating problems: A meta-analysis.  International Journal of Eating Disorders 27(4): 371-380

Thompson, R.A and Sherman, R.T. (1993) Helping Athletes with Eating Disorders. Human Kinetics:Champaign IL.