Share this page | Modelresource's Facebook Group

The Muse Study Results

Executive Summary

Model and Undergraduate Self-Esteem (MUSE) Study


The purpose of the Model and Undergraduate Self-Esteem (MUSE) Study was to compare models' and non-models' eating and exercise behaviour, self-esteem, and thoughts towards themselves and others. Models have, especially of late, been stereotyped in the media as having eating disorders. Their thin physiques, coupled with the tragic deaths of a small number of models from anorexia nervosa, has lead to model bans and guidelines issued around the world pertaining to the weight and ages of models on runways. Despite all of this, little research exists on the modelling population.

Participants and Methods

A total of 339 female participants completed the online questionnaire between December 2006 and March 2007. Non-models consisted of female undergraduate students from the University of Waterloo. Models were recruited through Toronto and Montreal modelling agencies via e-mail. Participation for both groups was voluntary and anonymous. All participants were to be over the age of 18 to satisfy the university's ethics concerns. The survey was conducted in an online format to accommodate the nomadic nature of fashion models.

The questionnaire was composed of previously validated and frequently used assessment tools as well as a small number of questions developed by the researcher and general demographic questions. It is important to note that the study did not seek to clinically diagnose participants, but rather, to “flag” those who might be eating disordered by way of assessing the eating attitudes and behaviours. This was accomplished by administering the Eating Attitudes Test (EAT). Other validated assessment tools contained within the survey included: Rosenberg Self-Esteem Scale (RSE), Appearance Self-Esteem Scale (ASE), Physical Appearance Comparison Scale (PACS), and the Sociocultural Attitudes Towards Appearance Questionnaire (SATAQ).


Body Size

  • Models were significantly taller and weighed significantly less than the comparison sample. The average BMI of models was 17.4 (18.0 controlling for age) vs. 22.7 for the non-models. Eating Behaviour
  • Eating attitudes and disordered eating behaviour did not differ significantly between models and non-models as measured by average EAT scores and EAT subscales.
  • Models and non-models did not differ significantly on whether they had ever: gone on eating binges, used diet pills, laxatives or diuretics, been treated for an eating disorder or recently thought of or attempted suicide.
  • Models were more likely to have indicated “yes” to whether they had ever vomited after eating to control weight or shape (24% of models vs. 12% of non-models). However, when asked the frequency of such behaviour in the past six months, no models came anywhere close to meeting clinical criteria for bulimia nervosa. Although non-models were less likely to indicate they had ever vomited after eating, some did so frequently enough that it is likely they could be diagnosed with bulimia.
  • There was some evidence that models below the age of 25 were more at risk for disordered eating behaviour than models over the age of 25; however, the average scores for both ages of models were well within the “normal” range.
  • No significant differences were found between “editorial/runway models” and commercial/lifestyle” models on eating behaviour.

    Self-esteem, Comparisons, Media Influence, Appearance

  • Models had significantly higher self-esteem than non-models.
  • Models and non-models did not differ on the degree to which they compare their physical appearance to that of others.
  • Models and non-models did not differ on how much they reported being influenced by media images. However, non-models were more likely to be influenced by images of athletes than models were.
  • Younger models (< 25) had lower general and appearance self-esteem than older models. Younger models (< 25) also more frequently compared their physical appearance to that of others.
  • No significant difference was found between “editorial/runway” models and “commercial/lifestyle” models on self-esteem, media influence or comparisons to others.
  • Models agreed significantly more frequently than non-models that, overall, they were happy with their physical appearance and were significantly more comfortable describing themselves as “beautiful”.

    Exercise/Physical Activity

  • Models and non-models did not differ on the frequency or duration of exercise they engaged in.

    Beliefs About Models

  • Non-models were more likely than models to indicate that models cause eating disorders in other women and girls, though both were a majority.
  • Non-models were far more likely to indicate that they felt models were more likely to have disordered eating behaviour than other women and girls

    Models' Beliefs

  • 44% of models disagreed and 31% agreed that the demands to be thin have been unfair, inappropriate or extreme.
  • 74% of models did not have to lose weight to begin modelling and 65% did not have to lose weight to continue modelling.
  • 67% of models did not have to change their eating or exercise habits to begin modelling and 34% did not have to change them to continue modelling.
  • The reasons models reported pursuing modelling included: passion for fashion (22.1%), travel (16.2%), just for fun (10.3%), money (9.1%), fame (4.4%), prestige (4.4%), self-confidence (1.5%), and other (22.1%).
  • 93% of models indicated that overall they felt modelling had a positive effect on them.
  • More models (44.1%) disagreed that demands to be thin have been unfair, inappropriate or extreme than models than agreed (30.9%). 25% remained neutral.

    About the models

  • The average age they began modelling was 16.9 + 3.6 years.
  • Regarding education, 41% were currently attending university or college and 40% had completed a university or college in the past.
  • One-quarter of the models were “in development or new faces”, 63.9% were professional models, 6.9% were former models, and 4.2% were modelling students (exclusively). Regarding type of modelling, 41.7% indicated “commercial/lifestyle”, 51.4% indicated “fashion/editorial/runway”, 5.4% indicated “plus size/full figure”, and 2.8% indicated “specialty/parts”.
  • Models were asked to indicate how long they have been modelling for: 6.9% had modelled for less than six months, 11.1% for six months to one year, 23.6% for one to three years, 11.1% for three to five years, 20.8% for five to 10 years, and 26.4% for 10 years of more.
  • The majority of the models (57%) modelled part-time.
  • All models were represented by at least one Canadian agency, but over half (58%) had experience working outside of Canadian markets in fashion capitals such as New York, Paris and Milan. In relation to this, 38.9% were currently not represented outside of Canada, 27.8% had representation in one to three cities outside of Canada, 22.2% were represented in four to seven, 8.3% were represented in seven to 10, and 2.8% were represented in more than 10 cities.


  • Smoking was significantly more common for models: 18% of models were smokers, compared to just 3% of non-models


    Research evidence from this study suggests that models are not at high risk for disordered eating behaviour, despite being much thinner than the comparison sample. These findings suggest that fashion models are genetically predisposed to be extremely thin and extremely tall in comparison to average female weight and height measurements. It appears that models maintain their thin builds without compromising eating or exercising excessively.

    The bans and guidelines that were issued and the committees that brainstormed them, have thus far failed to involve models in the decision making process, and perhaps even proceeded with their decision making without sufficient research about them. That the average BMI of the models is below the cut-off issued by one such “skinny model ban” (i.e., Spain's) together with the finding that disordered eating behaviour does not differ from a comparison sample whose BMI was considerably higher, suggests that such guidelines should be revised.

    Canadian models are known for their “thin but healthy” look and may be slightly less pressured to be extremely thin than models who call other fashion capitals their home. If this sample of models can be considered healthy, as appears to be the case, then it might be worthwhile for the international industry to consider such an aesthetic a goal for its models.

    It is hoped that the MUSE Study results, in addition to providing models with a voice on this topic, will make a significant contribution to the ongoing debate about model's physiques and help to continue the very important discussion.


    Due to an insufficient number of male models, plus size models, and modelling students participating in the study, analysis of these groups is on hold until a larger sample size can be obtained. The MUSE researchers are currently looking to recruit more participants in these categories – if interested, please contact

    The researcher is also looking to interview models for more in-depth research analysis, if interested, please contact

    Please contact the researchers for limitations about the study.




    News Release

    Executive Summary