Software Program Accurately Assesses Body Image in Patients With Anorexia Nervosa

Caroline Cassels

May 24, 2010

May 24, 2010 (New Orleans, Louisiana) — A new software program can objectively assess body image distortion in patients with anorexia nervosa (AN), including perception of current and desired body size.

New research presented here at the American Psychiatric Association 2010 Annual Meeting shows that the Body Image Software (BIS) was able to provide an accurate objective measure of body image — a core diagnostic feature of the disease — in a group of acutely ill patients with AN.

Jennifer Fier

"Body image distortion is one of the key required symptoms for the diagnosis of anorexia nervosa, but it is very, very hard to measure. Prior to having this particular technology, you really just had to interview the individual [to assess body image perception]," principal investigator Jennifer Hagman, MD, from the University of Colorado Denver, told Medscape Psychiatry.

"It was very subjective, and when you are trying to research anorexia nervosa and assess outcomes of different interventions, to have a core symptom that you can't really quantify makes it very difficult," she added.

Developed by Rick M. Gardner, PhD, also from the University of Colorado Denver, BIS has previously been used to measure perception of body size in a number of populations including obese individuals, physically and sexually abused children, normal children over time, college students, and women in different phases of the menstrual cycle.

However, said coinvestigator Jennifer Fier, MSIV, this is the first study to look at the software's validity in a group of young women with AN.

"We wanted to use this software to assess how much body distortion these patients had and how great their drive for thinness was," said Dr. Hagman.

According to investigators, there are 2 components of body image disturbance. These include perceptual distortion, defined as the inaccurate judgment of one's body size, and attitudinal dissatisfaction with the size or shape of one's body. These 2 components function independent of each other, but most patients with AN experience both.

Perception vs Reality

The study included 66 female teenagers and young adults, with a mean age of 16 years and a mean body mass index of 15.9 kg/m2. The median percentage of ideal body weight at the time of testing was 78.2%.

All patients completed 3 BIS tasks, explained Dr. Hagman. Individuals were required to look at a distorted image of themselves on the computer. The BIS, she added, is extremely user-friendly and takes no more that 20 to 30 minutes to administer.

In the first task, designed to measure body distortion, participants were required to adjust the image either wider or thinner, according to their current perceived body size. In the second task, which measured their drive for thinness, the patients were required to adjust their image to their desired body size.

In the third BIS task, known as the Adjustment Probit Estimation (APE), participants were required to look at 8 blocks of 40 images of themselves and identify each image as either "wider" or "thinner" than their actual size. The BIS program adjusted the images until the subject reached a point of subjective equality, which is their perceived current size, explained Dr. Hagman.

Overall, the researchers found study participants perceived their body size as 10.4% larger than it actually was, and the average desired body size was 6.8% smaller than participants' actual body size.

"Those 2 numbers are really interesting when you look at them in conjunction with each other because there is essentially a 17% difference between what patients think they look like and what they want to look like," said Ms. Fier.

"When subjects adjust the image to their perceived body image, it is of a normal or even slightly overweight person. As a clinician who has been working in this field for 2 decades, it was really striking for me to be able to see the image my patient sees and what they experience. For me this really changed my understanding of what body image distortion is really like for these patients," said Dr. Hagman.

On the APE task, the point of subjective equality was 7.0% larger than participants' actual size, and researchers also found that on average, patients with AN could reliably detect a 2.3% change in body size — a finding that Dr. Hagman said is on par with the general population and confirms that body image distortion in this patient population is not the result of a visual perception problem.

In addition, the researchers found that the APE task results were highly correlated with the Drive for Thinness and Body Dissatisfaction subscales of the Eating Disorders Inventory, as well as the Color-A-Person Test — a finding that confirms the validity of the BIS, said Ms. Fier.

"This study demonstrates that body image distortion is very real, and they truly do experience themselves as much bigger than they really are, and that the drive for thinness is also very real," said Dr. Hagman.

She added that currently there are few treatment effective treatment options for AN, which affects about 0.1% of the population. In addition, it has the highest suicide rate of any mental illness and carries significant and severe morbidity.

Predictor of Treatment Response?

Graham Redgrave, MD, assistant director of the eating disorders program and an assistant professor of psychiatry and behavioral sciences, Johns Hopkins University, Baltimore, Maryland, said the study makes an important contribution to AN research.

"People have had a long-standing interest in body image as part of eating disorders not only in patients with anorexia but also in those with bulimia who also have distorted body image and intense dissatisfaction with their bodies," he told Medscape Psychiatry.

Dr. Redgrave added that during the past decade there have been a number of research groups working to develop different types of tools, including some similar efforts to develop body image software programs.

The fact that patients were able to discriminate very small changes in size does suggest that body image distortion is not a visual processing problem. However, he added, future research that compares healthy controls to patients with AN would strengthen the validity of this result.

He added that if the findings are confirmed in a case-control study, the potential software could have clinical utility as an accurate and quantitative measure that would help predict treatment response.

"For instance," said Dr. Redgrave, "being able to predict who's going to need inpatient hospitalization vs...going to need partial hospital care — which is much cheaper — would be very useful.

"Right now...the treatment of anorexia, along with the treatment of schizophrenia, is probably the most expensive treatment for a mental health condition because the hospital stays are so long.

"It may be that an accurate measure of body image perception could help clinicians more accurately select those who would be able to manage at a lower level of care, and so you could step them down earlier or you could potentially avoid inpatient hospitalization altogether. That said, no clinician is going to make that judgment solely on the basis of a test like this, but if it adds value then that would be great," he said.

The study was supported by the Clinical Translational Research Center and by Training Medical Students to Become Child Psychiatrists. The authors have disclosed no relevant financial relationships.

American Psychiatric Association 2010 Annual Meeting: Abstract NR1-31. Presented May 24, 2010.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.