Disordered eating behavior can range in severity from various types of dieting to extreme weight control behaviors including purging behaviors and extreme self starvation (American Psychiatric Association, 2000). Eating disorders currently effect a small proportion of the population, however maladaptive patterns of eating and food restriction are probably more common. Examining the possible mechanisms that may contribute to the development of disordered eating is important in formulating treatment and prevention strategies to target these unhealthy behaviors. We propose to study the effects of social learning variables on the development of disordered eating behavior.
AIM I: Testing social learning variables as contributing to the development of disordered eating patterns.
Self-esteem is a construct that is associated with numerous positive life outcomes including job satisfaction, improved phsycial health, relationship satisfaction and positive and negative affect (Orth, Robins & Widaman, 2012). In particular, self-esteem has been positively correlated with body satisfaction in women, which may in turn influence disordered eating behaviors (Koyuncu et al., 2010). With the relationship between self esteem and positive mental health outcomes, self esteem may serve as a moderator in the relationship between social learning variables and disordered eating.
AIM II: Testing whether self-esteem is a moderator in the relationship between social learning variables and disordered eating behaviors.
Bandura was the first major theories to examine the concept of Social Learning Theory (Olsen & Hergenhahn, 2009). According to Bandura, learning results from interactions we experience with other people (Olsen & Hergenhahn, 2009). There are four major processes involved in social learning: attentional, retentional, behavioral production and motivational processes (Olsen & Hergenhahn, 2009). According to these four processes someone must attend to a model, retain the information provided by the model, translate the information obtained by the model into performance and have motivation to engage in the behavior that was modeled (Olsen & Hergenhahn, 2009). Bandura further argued reciprocal determinism guides behavior, where the person, their environment and the person’s behavior all interact with one another to produce future behavior (Olsen & Hergenhahn, 2009).
Bandura’s (1962) most recognized study of imitative learning examined children’s imitation of an adult model’s aggressive behaviors. Bandura (1962) argued that observation of a models behavior may elicit imitative responses from the children, without requiring the presence of reinforcement (Bandura, et al., 1962). Bandura examined whether observed behaviors could be generalized to new settings, where the model performing the original observed behavior was absent (Bandura et al., 1962). He found that observing a model’s aggressive behaviors in one setting was correlated with subsequent aggression in another setting where the model was absent (Bandura et al., 1962). Specifically, participants who had observed an aggressive model, compared to participants who observed a non aggressive model, engaged in physical and verbal aggression in another setting (Bandura et al., 1962).
Social Learning Theory could be applied to the study of disordered eating. Many factors contribute to and interact with each other to influence the development and maintenance of eating disorders. Some of these factors relate to Social Learning Theory, as influences of media exposure, family factors and social comparisons, may contribute to disordered eating patterns (Grabet et al., 2008; Stice, 2002; Neumark-Sztainer, 2010; Myers and Crowther, 2009) .
Current Western society emphasizes thinness as a physical ideal for women meet. Garner, Garfinkel, Schwartz and Thompson, (1980) chronicled the evolution of the thin ideal in Western society by examining changes in various popular culture phenomenon. When examining playmate centerfolds over time, these researchers noted the percent of average weight for playmates age and height decreased significantly from 1959-1978. Similar results were found when examining Miss America contestants from 1959-1978. Each year, the average weight of Miss American contestants declined by about .28 pounds, and the average weight of Miss American winners declined by .37 pounds a year. Furthermore, the winners of the Miss America Pageant were found to weight significantly less than the contestants each year. Finally, the authors examined the number of diet articles included in numerous women’s magazines and found the number of diet articles had increased from 1959-1973. Although this article examines changes in popular culture in the 1960’s and 1970’s this trend in an increasingly thin ideal for women is still evident today as disordered eating rates continue to rise and American models and actresses continue to appear extremely thin.
Although the thin ideal is more prominent in American females, studies have also examined the evolution of male physiques. Pope, Olivardia, Gruber and Borowiecki (1998) examined the changing male physique through male action toys in the past 30 years. The researchers found that male action figures, including GI Joe and Star Wars figures, have become more muscular over time. This research may explain the phenomenon where men experience body dissatisfaction related to inadequate muscularity.
These evolving physical ideals, which emphasize thinness for women and muscularity for men, are conveyed through multiple forms of media including television, magazines and the internet. Studies have shown that media exposure can have a negative impact on body satisfaction and eating behaviors. One meta-analysis by Grabe, Ward and Hyde (2008) evaluated the effects of media exposure on female body dissatisfaction, internalization of the thin ideal and eating behaviors in a meta analysis of both experimental and correlational studies. The authors concluded that media exposure has a small to moderate effect on body dissatisfaction (Cohen’s d = -0.28), indicating that increased media exposure has an associated with decreases in body satisfaction. Similarly, media exposure had a moderate negative effect on internalization of the thin ideal (Cohen’s d = -0.39). Finally, media exposure was also moderately negatively associated with eating behaviors (Cohen’s d = -0.30), suggesting media exposure is related to greater eating pathology.
Another meta-analysis by Stice (2002), examined factors that contribute to the development and maintenance of disordered eating. Some elements in his meta analysis may apply to how social learning theory may explain disordered eating. Sociocultural pressures to be thin were associated with an increase in body dissatisfaction, dieting and negative affect. These factors have been shown to be associated with the development of disordered eating. In addition, sociocultural pressures to be thin also predicted binge eating, bulimic symptoms and eating pathology. Stice (2002) also found that internalization of this thin ideal was associated with an increase in body dissatisfaction, dieting and negative affect, and was also associated with the onset of binge eating and bulimic symptoms.
These studies show that factors related to social learning and exposure to others such as media exposure and sociocultural pressures to be thin are related to disordered eating and that increases in both factors that can contribute to eating pathology, such as body dissatisfaction, dieting, bulimic symptoms and eating pathology.
Family factors that contribute to disordered eating, may also be explained in terms of Social Learning Theory. Neumark-Sztainer, Bauer, Friend, Hannan, Story & Berge (2010) examined how family weight talk and dieting can influence body dissatisfaction and disordered eating in a sample of adolescent girls. The researchers found that weight teasing was associated with higher BMI, body dissatisfaction, unhealthy weight control behaviors and binge eating in their sample. Additionally, parent weight talk, or discussion about their own weight, was associated with numerous eating disordered behaviors. Finally, mother’s who dieted was associated with extreme weight control behaviors by daughter.
Modeling of eating pathology and social comparisons to other women’s body image may also impact the development of body dissatisfaction and eating pathology. In the meta-analysis by Stice (2002) modeling of body image disturbance and eating pathology was found to increase onset of binge eating and bulimic symptoms. In another meta-analysis Myers and Crowther (2009) examined how social comparison impacts body dissatisfaction. The researchers found a large effect size for the impact of social comparison on body dissatisfaction (d = 0.77). Therefore, when individuals engaged in social comparison with others, body dissatisfaction was increased. Moderators in this relationship included gender and age, with women and younger individuals, experiencing greater body dissatisfaction when engaged in social comparison. The influence of social comparisons on body dissatisfaction is important as body dissatisfaction is a risk factor associated with dieting and eating pathology (Stice, 2002).
Self-esteem has been associated with an array of positive outcomes. Orth, Robins and Widaman (2012) examined the impact of self-esteem on various important life outcomes including occupational and relationship satisfaction, salary, and positive and negative affect. The researchers found that self-esteem effected affect, depression, relationship satisfaction, job satisfaction and overall health. Cross-lagged regression analysis revealed that self-esteem was most likely a cause of these outcomes rather than an outcome.
Studies of Terror Management Theory suggest that self esteem may possess an anxiety buffering function. Greenberg et al. (1992) experimentally manipulated self-esteem and found that increasing self-esteem was associated with decreases in anxiety when participants were asked to view images of death. In a subsequent study, Greenberg et al. (1992) found that providing positive feedback about participants personalities also resulted in decreases anxiety following exposure to death images. When anticipating a painful shock, participants in the high self-esteem condition had decreased self-report and physiological signs of anxiety (Greenberg et al., 1992).
Self-esteem has also been shown to increase body satisfaction in women. Koyuncu et al. (2010) found a significant, positive association between self-esteem and body image, with higher self-esteem being related to a more positive body image. In this same study, self-esteem was also significantly negatively related to social physique anxiety. Previous research has examined self-esteem as a moderator of the relationship between social comparisons and body image in women. Jones and Buckingham (2005) found participants in the lower self-esteem group, experienced higher levels of body esteem after making downward comparisons to their peers, whereas women with higher self-esteem experienced higher body image after making upward social comparisons. Vohs et al. (2001) examined the relationship between body dissatisfaction, perfectionism and self-esteem on bulimic behavior. They found a three way interaction between self-esteem, perfectionism and body dissatisfaction, where perfectionism and body dissatisfaction combined to increase bulimic symptoms, but only in participants who experienced low self-esteem.
If increased self esteem has been shown to be associated with general positive outcomes, including positive increases in body image, self-esteem may serve as a moderator of the relationship between social learning and disordered eating behaviors.
Disordered eating behaviors contribute to a myriad of mental and physical health problems. Extreme food refusal can result in the development of Anorexia Nervosa, an eating disorder characterized by refusal to maintain normal body weight, fear of gaining weight or becoming fat, and a disturbance in the way one perceives their weight or shape (American Psychiatric Association, 2000). Anorexia Nervosa has been associated with numerous negative health outcomes including hypotension, hypothermia, anemia, cardiovascular difficulties, dental problems, osteoporosis and death, with mortality rates ranging from 4.-0-5.6% (American Psychiatric Association, 2000; Crow et al., 2009; Lock et al., 2010). In addition to physical health problems associated with Anorexia Nervosa, many psychological disorders, including depressive disorders, anxiety disorders, obsessive-compulsive disorder and personality disorders are highly co-morbid with Anorexia Nervosa (Lock et al., 2010).
Bulimia Nervosa is the second diagnostic category in the current DSM-IV-TR for eating disorders and is characterized by eating significantly large amount of food in discrete periods of time, followed by inappropriate compensatory behaviors, in an attempt to lessen weight gain (American Psychiatric Association, 2000). Bulimia Nervosa, although not traditionally believed to be as severe as Anorexia Nervosa, is associated with poor mental and physical health outcomes including increased rates of depressive disorders, anxiety disorders, substance abuse disorders, electrolyte imbalances, metabolic alkalosis, metabolic acidosis (American Psychiatric Association, 2000). Mortality rates for Bulimia Nervosa have been found to be 3.9%, almost as high as some morality rates found in Anorexia Nervosa (Crow et al., 2009).
Although Anorexia Nervosa and Bulimia Nervosa are generally thought to be the most severe types of eating pathology, however subthreshold cases of eating pathology are also associated with poor outcomes and mortality rates for individuals diagnosed with Eating Disorder Not Otherwise Specified, where individuals exhibit maladaptive eating patterns but do not meet full DSM-IV-TR criteria for either Anorexia Nervosa or Bulimia Nervosa, have been found to be as high as 5.2% (Crow, et al., 2009). Furthermore, subthreshold Bulimia Nervosa, Binge Eating Disorder and Purging Disorder were associated with greater impairment and distress (Stice et al., 2009). These subthreshold cases may also progress to disorders that meet diagnostic significance, especially for cases of subthreshold Binge Eating Disorder and Bulimia Nervosa (Stice et al., 2009). This suggests that this population of subthreshold individuals should not be ignored in the prevention and treatment or disordered eating.
The numerous negative mental and physical health outcomes associated with eating disorders and subthreshold cases of disordered eating behavior suggest that understanding variables that can contribute to the development of these disorders is important. Identifying both risk and protective factors for disordered eating can help implement effective prevention and intervention strategies to reduce and prevent disordered eating behaviors.
We predict that correlational analyses will reveal that greater exposure to modeling of disordered eating habits and exposure to thin images will be associated with an increase in disordered eating patterns. Furthermore, we hypothesize regression analyses will show that self-esteem moderates this relationship. Particularly, participants with greater self-esteem will show significantly less disordered eating symptoms, even when exposed to modeling of disordered eating behaviors. Subsequently, lower levels of self-esteem will be associated with even greater patterns of disordered eating after exposure to disordered eating models.
V. Research Design
Participants will be high school aged children who are randomly selected to participate in the study from a sample of 15 randomly selected schools in the city of Chicago. Participation in this study will be voluntary and informed consent documents will be provided to obtain consent from the parents. In addition, assent documents will be provided to all participants in this study to acknowledge their willingness to participate. Participants may choose to be entered to a raffle to win one of four $50 Amazon gift cards for their participation in the study. About 500 participants will be expected to participate.
Demographic questionnaire: A demographic questionnaire would be administered to participants. Information regarding participant age, gender, ethnicity, race, and year in school will be obtained.
Eating Disorder Examination (16th Edition) The Eating Disorder Examination is a clinician administered evaluation attitudes and behavior of eating disorders designed by Cooper and Fariburn (1987). The EDE has diagnostic items used to diagnosis an eating disorders as well as subscales designed to assess for pathology including Restraint, Eating Concerns, Weight Concern, Shape Concern. All examiners will be doctoral level psychology students with significant training in administration and scoring of the measure. The Eating Disorders Examination has been shown to have adequate reliability. Test-retest reliability has been shown to be .70 or higher and inter-rater reliability is around .90 (Rizvi et al., 2000). Clinical doctoral students, who have been adequately trained by professionals in the field, will conduct these interviews.
General Media Habits Questionnaire- Adult Version (GMHQ-A) The General Media Habits Questionnaire – Adult Version was developed by Douglas Gentile and measures various types of media exposure including media violence exposure, prosocial media exposure, aggressive media exposure, and video game exposure (Gentile, 2011). This version has been validated in secondary school aged children through adults (Gentile, 2011).
Social Learning Questionnaire: A social learning questionnaire will be designed to assess participants level of exposure to disordered eating behaviors by parents and peers. Questions will be designed to assess types of disordered eating habits, such as dieting, negative weight talk and weight teasing, which have been associated with disordered eating in previous research (Neumark-Sztainer et al., 2010).
Rosenberg Self Esteem Scale – The Rosenberg Self-Esteem Scale (RSE) is a 10 item, self report measure assessing participants feelings about themselves and their self esteem. This scale was designed by Morris Rosenberg (1965). Item include “I feel that I have a number of good qualities,” “I feel I do not have much to be proud of,” and “I certainly feel useless at times.” The RSE has been found to have good psychometric properties with internal consistency reliabilities of ?? = .80 (Galli et al., 2011). One year test retest reliability has also been strong with r = .77 (McCarthy & Hodge, 1982).
Participants will be female high school aged children, randomly selected to participate in the study from a random sample of 15 schools in the city of Chicago. Participants will be told the study examines body image concerns in high school females. Informed consent will be obtained from parents of the students and assent will be sought from the participants themselves. Participants will be asked to spend about one hour and a half to participate in the study. Participants will begin with completing the demographic questionnaire, social learning questionnaire and XXX. Following completion of the questionnaires, participants will be administered the Eating Disorder Examination by trained clinical psychology doctoral students.
Following completion of the study, participants will be given an opportunity to enter a raffle to win one of four $50 Amazon gift cards for compensation for their time. Participants will then be debriefed by the experimenter and will be given information regarding the actual purpose of the study. Participants who meet criteria for an eating disorder based on their answers to the Eating Disorder Examination will be referred to community resources for eating disorders.