VuLogoPsychology Department

Health Psychology Home Page

Papers written by students providing scientific reviews of topics related to health and well being

  HomeWeight LossAlternative Therapy | Supplements | Eating Disorders | Fitness | About this Page |

Sororities: A breeding ground for eating disorders?

Valerie Kuznik






Eating disorders are a serious, under-recognized issue in today’s society, especially for women; approximately 6% of women have some type of eating disorder, and college women in particular have a high frequency of disordered eating (Alexander, 1998).

A possible motivation for this high prevalence of eating disorders is society’s “emphasis on thinness” (Schulken, Pinciaro, Sawyer, Jensen, Hoban, 1997). This emphasis explains why groups that have a higher risk of developing disordered eating are those that “emphasize body weight and shape for professional competence, such as actors, dancers, and athletes” (Schulken et al, 1997) or social groups that “emphasize a thin ideal” (Schulken et al, 1997).

Thanks to movies, books, and gossip, sororities have come to be stereotypically labeled as one of these social groups. As a result, many see sororities today as a breeding ground for eating disorders due to the supposed obsession with appearance and thinness. However, those who investigate further often find that these stereotypes are false, and that sororities can be home to intelligent, well-rounded women who value a person’s achievements and personality over their clothes and physique.

As a result, one must question the stereotype that if a woman is in a sorority, she must have an eating disorder. This brings many questions to the surface: Are eating disorders really more statistically prevalent among women in sororities as compare to those not associated? If this is indeed the case, is it because the actual sorority causes the high frequency due to supposed appearance-obsession, or are women with predispositions to develop eating disorders just more likely to join sororities? And if being in a sorority really makes women more susceptible to eating disorders, what exactly causes this vulnerability?


Method Definitions


            Several questionnaires and surveys are used in studies to determine eating disorder prevalence and the severity of pathology.

The Eating Disorder Inventory (ETI) was created in order to measures symptoms (both behavioral and psychological) of bulimia and anorexia nervosa using eight subscales: drive for thinness, bulimia, body dissatisfaction, ineffectiveness, perfectionism, interpersonal distrust, interoceptive awareness, and maturity fears (Alexander, 1998).

The Eating Attitudes Test (EAT) is a survey with forty questions that assesses anorexia components of behavior and attitude. (Alexander, 1998) The lower the score on the EAT, the more severe of an eating abnormality the participant suffers from.

The Bulimia Test-Revised (BULIT-R) is a questionnaire with 36 items that evaluates bulimia based on criteria for diagnosis set by the DSM-III (Alexander, 1998). The higher the score, the “more severe [the] pathology” (Alexander, 1998) is.

The Body Mass Index (BMI) Silhouettes Survey includes seven silhouettes, each assigned a BMI value. Participants are asked to choose a silhouette that they think is similar to their own, a silhouette that they think matches what a woman should look like, and a silhouette that they would most like to have (Schulken, Pinciaro, Sawyer, Jensen, Hoban, 1997). The results reflect body image and perception.

The Rosenberg Self-Esteem Scale (RSES) is a ten-question survey. It is composed of statements, for example, “At times I think I am no good at all” ( and the participant answers on a scale of “strongly agree” to “strongly disagree.” ( The study measures the participant’s self esteem.

The Center for Epidemiologic Studies Depression Scale is used to determine one’s “depression quotient” ( The survey is comprised of twenty questions that test for depression by providing statements and interpreting responses that indicate how rarely or frequently the stated feeling or action occurred (


Study 1:

“The Prevalence of Eating Disorders and Eating Disordered Behaviors in Sororities”


            In 1998, Laurel A. Alexander conducted a study entitled “The Prevalence of Eating Disorders and Eating Disordered Behaviors in Sororities.” This study’s background information cited two possible sources for higher prevalence of eating disorders in sororities: the first was the age of sorority women, and the second was the fact that many women in sororities are of a higher socioeconomic class (Alexander, 1998). These sources seem to support that sorority women are more susceptible to eating disorders rather than joining a sorority causes eating disorders in its members. Alexander also noted that another source could be membership in the sorority itself, as previous studies have found that women overly concentrated on “creating and maintaining an effective social façade” are at higher risk for developing eating disorders” (Alexander, 1998), which seems contradictory to the previous source suggestion. Thus, from the background information alone, there is no clear conclusion.

            The three groups in the study were a sorority member group, an “activity comparison” group (included members of dance groups or sports teams), and a control group. The measures used to compare the groups were the Eating Disorder Inventory (EDI), the Eating Attitudes Test (EAT), and the Bulimia Test-Revised (Alexander, 1998). The study involved 239 participants, all women, making up the three groups.

            On the EAT, the “activity comparison” (Alexander, 1998) group scored the lowest, followed by the sorority group and then by the control group, scoring the highest. These results indicated that the activity group met the most anorexic components out of all three groups. Similarly, on the bulimia portion of the EDI, the scores from lowest to highest were the activity group, the sorority group, and last the control group. To the contrary, on the Ineffectiveness portion of the EDI, the control group scored the lowest, then the sorority group, and finally the activity group with the highest score. As a lower score relates to “more severe pathology,” it is clear that while the activity group exhibits more bulimic tendencies, the control group has greater “feelings of self-doubt, self-loathing, powerlessness and insecurity” (Alexander, 1998). Finally, on the BULIT-R the activity group scored highest and the control group lowest, with the sorority group in between. As a higher score on this test reflects “more severe pathology” (Alexander, 1998), the activity group was found to meet the most components of bulimia (Alexander, 1998).

While the sorority group tended to score with “more severe pathology” (Alexander, 1998) than the control group on the three questionnaires, it cannot be ignored that the sorority group still scored less severely pathologically than the activity group. This can be translated to generalize that while members of sororities may be more prone to eating disorder behavior and eating disorders themselves than the average college woman, college women participating in sports or dance teams see the highest prevalence of these disorders and behaviors. The only area in which the sorority women seemed more susceptible to eating disorder behavior was in their scores on the ineffectiveness portion of the EDI, meaning that as a whole, the women in sororities on this campus were more insecure than women involved in dance and sports (Alexander, 1998). Alexander draws this conclusion and celebrates the success of the hypothesis (predicting that sorority women would have more self-doubt) without recognizing that the sorority women, though more insecure than the athletes, were still more secure than the average woman on campus. This distinction is incredibly significant in determining whether or not sorority women really are at higher risk for developing an eating disorder than non-sorority members. After recognizing this distinction, it is clear that sorority women are not necessarily at higher risk; in fact, they could be at lower risk due to lower insecurity.

            The study recognized its own limitation in noting the small size of the sample groups. Additionally, the study warned that the data might not be as significant because the Greek life on this particular campus was probably not a good representation of the Greek life on other campuses across the country. This limitation could be compensated for by performing this study on multiple campuses (Alexander, 1998). Additionally limiting was the fact that subjects for the study were recruited from the sororities directly, as the Greek life population on campus was not large enough to guarantee that a random sampling of students would have sufficient Greek representation (Alexander, 1998). As a result of these limitations, the data may be skewed and less accurate.

            Thus, the only conclusion that can be made from this study is that women involved in sports and dance on campus are more likely to have some sort of disordered eating behavior than women in sororities, who are more likely to have some sort of disordered eating behavior than the average college woman involved in neither athletics nor Greek life. This study suggests that while sorority women were more at risk than average campus women, there are other groups with higher risk that must be recognized.


Study 2:

“Sorority Women’s Body Size Perceptions and Their Weight-Related Attitudes and Behaviors”


            A second study, “Sorority Women’s Body Size Perceptions and Their Weight-Related Attitudes and Behaviors,” was conducted by Ellen D. Schulken, Paul J. Pinciaro, Robin G. Sawyer, JoAnne G. Jensen, and Mary T. Hoban in 1997. The study’s background section claims that women in sororities “are sometimes pressured into developing potentially dangerous weight-control behaviors” (Schulken, Pinciaro, Sawyer, Jensen, Hoban, 1997) as “binge eating may be acquired through modeling” (Schulken et al, 1997). Though this modeling behavior may in fact be accurate, they have provided no evidence that the binge behavior being modeled exists in sororities in the first place; a behavior cannot be modeled if it is not present to begin with. Thus, this claim is unsupported and invalid.

            The study cited many previous studies’ findings on the topic of eating disorders and sororities to provide sufficient background information. A study by Meilman et al that was cited found that of all students in a sample (freshmen excluded) who purged, 72.2% were members of sororities (or fraternities). Interestingly, Meilman et al admitted that they had not reached a firm conclusion as to whether disordered eating was a result of membership in Greek life or if instead, those more susceptible to developing eating disorders were just more likely to “go Greek,” a question also looking to be answered in this paper (Schulken et al, 1997). Thus, though this statistic undeniably reflects that the majority of bulimics sampled were in sororities, it does not mean that the supposed “thin ideal” of sororities caused this. Another limiting aspect of this study was that it only surveyed for bulimia and therefore can only be taken for what it represents; the results do not speak to eating disorders in general but instead, just bulimia.

            Another study cited, conducted in 1997, surveyed women who came into a campus’s “primary care clinic.” The study found that 2.5% of women in sororities who were sampled were bulimic, while only 0.8% of athletes sampled were bulimic (Schulken et al, 1997). These results are almost completely irrelevant for various reasons. First, the study only sampled women who entered the clinic, meaning that rather than randomly sampling the entire student population, only women with a health issue that brought them into the clinic were surveyed, which did not create a very legitimate sample group. Second, the percentages recorded referred to how many women in that group (sorority woman/athlete) were bulimic, not how many women who were bulimic fell into that group. This distinction is significant because the percentages are completely dependant upon how many women from that group entered the clinic and most likely do not accurately reflect the frequency of bulimia in each group. However, these statistics do importantly serve to put Meilman et al’s study’s statistics into perspective: while Meilman et al reflected that a majority of bulimics were sorority women, this study reflected that only a very small percent of sorority women were bulimic. This study helps to point out that by no means are all sorority women bulimic, despite the misleading statistic from Meilman et al.

            In the actual study conducted by Schulken et al, the women sampled were at least eighteen years old and in one of twelve sororities “at a large mid-Atlantic coast university” (Schulken et al, 1997), with 627 women participating in all. The measures used to conduct the study were a “demographic questionnaire” (age, grade, height, weight, race), the EDI, and the Body Mass Index (BMI) Silhouettes Survey. Data was collected by three females at sororities’ weekly chapter meetings, with each member present at the meetings asked to fill out the three surveys (Schulken et al, 1997).

            Demographically, Schulken et al determined that 91.9% of the women surveyed were white and that the average age of women surveyed was 19.5.  The EDI scores for the sorority women was compared to five previous studies of surveyed college women, four of which concerning women without sorority affiliation and one study of women in sororities. The results of the EDI reflected that the sorority women in this study had higher scores in the drive for thinness, body dissatisfaction, and bulimia subcategories. On the body silhouette questionnaire, 62.1% of the sorority women chose a figure with a BMI considered to be underweight when asked to select the silhouette that a woman should look like. When asked to select a silhouette that resembled their own physique, 36.1% of sorority women considered to be underweight selected a figure with a higher BMI than their own. To the contrary, of sorority women who were at a normal weight or overweight, only between 1.6-8.3% chose a figure with a higher BMI than their own. Finally, only 19% of all of the sorority women chose a silhouette that “represented an acceptable weight” (Schulken et al, 1997) (meaning its BMI fell in the normal range) when asked to select a silhouette of how they wished they looked, while 81% selected figures with BMI’s considered to be underweight (Schulken et al, 1997).

            The study reasoned that as sorority women in this study scored higher on select BMI subcategories, they may have a higher body preoccupation (dissatisfaction, fear of gaining weight, etc.), and as this is considered “relevant to the development of eating disorders,” these women may constitute “an at-risk population” (Schulken et al, 1997). The results regarding choosing a silhouette of a higher BMI than one’s own is considered to be characteristic of Anorexia. However, although the study claims that “there may be women in this population whose body-size perceptions are distorted” (Schulken et al, 1997) due to the 36.1% of underweight women who chose silhouettes that were larger, this percentage only makes up one-third of all underweight sorority women, not to mention the majority of normal and overweight sorority women (recall these percentages were quite low, at 1.6% and 8.3% respectively). Thus, this assumption seems to make a generalization for the entire sorority population based on a small percentage, and does not even compare these statistics with those of non-Greek college women.  Finally, although it is true that a majority of the women chose an underweight silhouette for their ideal as well as for what they thought a woman should look like, the study offers no comparative data with that found of these sorority women, and as a result, it seems unjustified to conclude that “thinness is an ideal among these sorority women” (Schulken et al, 1997) when there is no data to present this thinness ideal for the general collegiate women population (Schulken et al, 1997). If the general population were to have a substantially similar results, this conclusion would be less valid.

Study 3:

“A Prospective Study of Disordered Eating among Sorority and Nonsorority Women”


            The previous studies did not touch on whether or not being in a sorority actually causes eating disorders, but “A Prospective Study of Disordered Eating among Sorority and Nonsorority Women” by Kelly C. Allison (University of Pennsylvania School of Medicine, Department of Psychiatry) and Crystal L. Park (University of Connecticut, Department of Psychology) examines just that (Allison, Park, 2004) focuses on just that.

            Very significantly, this study cited a study by Atlas & Morier, in which characteristics of those women who rushed sororities were investigated. Atlas & Morier found that these women were prettier, from higher economic classes, more eager to “fit in,” and drank alcohol more often (Allison et al, 2004). These characteristics are all stereotypical of sorority women, so this study proved that the stereotype was true in this case.

            Additionally significant is the evidence of modeling behavior that another study cited suggests. A survey by Crandall found that over a period of time, a group of friends ”who modeled certain eating behaviors…came to accept these behaviors as norms” (Allison et al, 2004). The study found that as the year went on, the binge behavior inside of friend groups became more homogenous (Schulken et al, 1997). This is significant as sororities take the form of “friend groups,” and thus, members might be more likely to behave similarly and come to see the behavior of their sisters as normal, even if this includes disordered eating. However, this modeling would not just be an issue in sororities but also on teams, in clubs, and in “unofficial” social friend groups as well. As a result, this finding does not shed considerable light on eating disorders in sororities when this principle applies to every social group.

            Schulken et al’s “A Prospective Study…” surveyed 102 women throughout three of their four years “at a Midwestern state university” (Schulken et al, 1997).  Data was collected at three time periods: T1, during the fall of their freshman year; T2, during the winter of their sophomore year; and T3, during the winter of their junior year. The following was assessed at each time period: T1, ideal weight; T2, sorority membership; T3, ideal weight. Additionally, at all three time periods, the Drive for Thinness (DT), Bulimia (B), and Body Dissatisfaction portions of the EDI; the Center for Epidemiologic Studies Depression Scale; and the Rosenberg Self-Esteem Scale (RSES) were measured (Schulken et al, 1997).

            Unexpectedly, Allison and Park found that over the three years, women who ended up joining a sorority actually gained more weight than those that did not join. Additionally, over the three years there was no significant difference between sorority and non-sorority women regarding BMI, actual/ideal weight discrepancy, ideal weight, self-esteem and depression. The only significant difference found was a higher drive for thinness (part of the EDI) in the women who joined sororities at the third time period than those who did not (Allison et al, 2004).

            Overall, the study concluded that women who join sororities do not enter with a higher likelihood of engaging in disordered eating than women who do not join. Further, women who joined sororities actually gained more weight than those who did not, contrary to the stereotypical assumption that sorority women are appearance-obsessed and starve themselves. However, this seems to contradict the finding that women who joined sororities had an increased drive for thinness while those who did not had a decrease (Allison et al, 2004). Thus, the results of this study appear somewhat unclear due to this incongruity: the women were supposedly more concerned with thinness, yet gained more weight. Though it is true that some eating disorders might cause weight gain, the majority do not, and thus, an irregularity exists.

            This study is limited in several areas. First, like Schulken et a, this study had a very homogenous sample group (majority white). Additionally, the university used in this study had a high amount of Greek life, with a quarter of all freshmen women joining sororities, so this may not be representative of Greek life for colleges across the country. Finally, as mentioned before, the study’s results are partially contradictory and therefore weaken its credibility.

            However, the study’s findings that women who join sororities do not enter with a higher probability of engaging in disordered eating than those who do not is significant. Thus, the question “are women who are more susceptible simply more attracted to sorority life?” is answered with a supported “no.”




            Based on the results collected from the three previously mentioned studies, some aspects of the relationship between sororities and eating disorders have been clarified, while others have been blurred further.

            As to whether or not sorority women themselves are more likely to engage in disordered eating, it has been proven by Alexander that while sorority women are more disordered in their eating behavior that the average college woman, they are still less likely to be disordered than women who engage in sports. Thus, although sororities are the recipients of negative stereotypes regarding eating disorders, athletes are actually a more accurate group for this.

            Another aspect of sororities and eating disorders is whether sororities actually cause eating disorders or if the women who join are simply more susceptible due to their own personal factors and just so happen to be attracted to sorority life. Schulken et al unjustly concluded that sorority women have poor self-image based on very low statistics, and Allison et al addressed this issue although the results were somewhat muddled. While Allison et al established from their results that women entering sororities are not in fact more susceptible to eating disorders than those who do not join, their other data was not as conclusive. Later data suggested that being in a sorority caused an increased preoccupation with thinness but also a greater weight gain than those not in sororities. It is hard to gather a conclusion from this confusing data, and therefore it was unjustified when the study concluded that due to the preoccupation-with-thinness data, “dieting and weight issues continue to be emphasized within Greek organizations” (Allison et al, 2004). If this were true, the members would not have gained more weight than non-members. Thus, this study is a perfect example that this issue is not cut and dry. As a result, no firm conclusion can be gathered this aspect of sororities and eating disorders.

            Something that none of these studies took into account was the little recognized concept that no two sororities are the same. Even if sororities share the same Greek letters but are organizations at two different universities, there is almost no likelihood of the member groups of those two chapters being similar. Each university has a separate culture, and even within a university, different sororities have different values and even stereotypes. Although sororities are known for being typically shallow, appearance-obsessed, and ditzy, others are known for their sisterhood bond, their academic performance, or their devotion to service. As a result, it would be almost impossible to gather concrete data on this subject matter. Some data would definitely suggest a correlation between sorority membership and disordered eating if the right chapters at the right schools were surveyed. To the contrary, if a sorority with a reputation for being home to outcasts and less popular women was surveyed, their eating disorder statistics would most likely be lower because perhaps that house does not pick members during recruitment based on appearance, but rather based on personality and campus involvement. To conclude, all of the studies discussed were limited in that they failed to recognize the possibility of difference between actual sorority women, and not just that between sorority and non-sorority women. Lack of recognition of this distinction most likely caused the sporadic and inconsistent data.

            As a whole, the only conclusion that can be made based on limitations and contradictive findings in the three studies is that sorority women are not more susceptible to eating disorder behavior as they enter sororities (Allison et al, 2004) and that athletes are actually more prone to eating disorder behavior than sorority members (Alexander, 1998). It is clear that the relationship between sorority membership and eating disorders is not as cut and dry as the media and the general population make it out to be; just because a woman calls herself by her letters, she is in no means necessarily engaging in disordered eating.















Works Cited


Alexander, L.A. (1998). The prevalence of eating disorders and eating disordered behaviors in sororities. College Student Journal, 32, 66–75.

Allison, K. C., & Park, C. L. (2004). A prospective study of disordered

         eating among sorority and nonsorority women. International Journal of

         Eating Disorders, 35(3), 354 –358.

Schulken, E. D., Pinciaro, P. J., Sawyer, R. G., Jensen, J. G., & Hoban,

         M. T. (1997). Sorority women’s body size perceptions and their weight-

         related attitudes and behaviors. Journal of American College Health, 46,

         69 –74.





Psychology Department

The Health Psychology Home Page is produced and maintained by David Schlundt, PhD.


VuLogoVanderbilt Homepage

Return to the Health Psychology Home Page
Send E-mail comments or questions to Dr. Schlundt