Genetics accounts for more than half of anorexia liability, UNC-led study concludes

March 05, 2006
A new study led by University of North Carolina at Chapel Hill researchers estimates that 56 percent of the liability for developing anorexia nervosa is determined by genetics.In addition, the study found that the personality trait of “neuroticism” (a tendency to be anxious and depressed) earlier in life is a significant factor associated with development of the eating disorder later.

Anorexia nervosa is a psychiatric illness characterized by an individual’s refusal to maintain a minimally acceptable body weight, intense fear of weight gain and a distorted body image. It occurs primarily among females in adolescence and young adulthood and is associated with the highest mortality rate of any mental disorder.

Dr. Cynthia M. Bulik

Dr. Cynthia M. Bulik

This study is the first published in the medical literature to estimate how much liability for developing anorexia nervosa is due to genetics, and the first to find a statistically significant association between the prospective risk factor of neuroticism and later development of anorexia, said Dr. Cynthia M. Bulik, lead author of the study, published in the March issue of the Archives of General Psychiatry.

“What this study shows is that anorexia nervosa is moderately heritable and may be predicted by the presence of early neuroticism, which reflects proneness to depression and anxiety,” Bulik said. “Fifty-six percent heritability – that’s a fairly large contribution of genes. The remaining liability is due to environmental factors.”

Bulik is the William R. and Jeanne H. Jordan distinguished professor of eating disorders in UNC’s School of Medicine and director of the UNC Eating Disorders Program at UNC Hospitals. She also is a professor of nutrition, a department housed in the schools of public health and medicine, and holds the only endowed professorship in eating disorders nationwide.

The reason she and her co-authors reached these conclusions where previous studies could not, Bulik said, is that their study was based on data obtained from screening a very large sample of twins. Their sample, from the Swedish Twin Registry, consisted of 31,406 individuals born between 1935 and 1958. None of the previous studies had samples nearly as large, Bulik said.

Working with colleagues at the Karolinska Institute in Stockholm, Bulik’s team screened members of the sample for a range of disorders, including anorexia nervosa, using diagnostic criteria from the American Psychiatric Association’s “Diagnostic and Statistical Manual of Mental Disorders – Fourth Edition.” Information from sample members collected in 1972-1973 was used to examine prospective risk factors.

About half of the members of the sample were monozygotic, or identical, twin pairs, who are genetically identical. The other half were dizygotic, or fraternal, twins, who are no more similar genetically than siblings who are not twins.

“In this big population, we compared the group of monozygotic twins with the group of dizygotic twins and asked the question, ‘How often do both twins in a twin pair have the disorder?’ If you find that both members of monozygotic twin pairs have the disorder more frequently than both members of dizygotic twin pairs, that suggests there’s a genetic component to the disorder,” Bulik said.

That’s what the researchers found, she added.

“It was more common for both members of identical twin pairs to have it than for both members of fraternal twin pairs.”

A statistical analysis of the data they collected from the twins in the sample estimated that genetics accounted for 56 percent of the liability for developing anorexia within that population.

Since the Swedish Twin Registry contained data on sample members dating back to 1972-1973, Bulik’s team also was able to determine the factors or features that these people had earlier in life that predicted they were going to develop anorexia nervosa later.

“What we found was that neuroticism, measured in 1973, was the strongest predictor of the development of anorexia nervosa later in life,” Bulik said. Other prospective risk factors that were examined, including a low body mass index and excessive exercise levels, were not found to be predictive.

One of the next steps, Bulik said, is to put these two pieces of information together to figure out exactly what is inherited. “We suggest that there are some basic biological differences between people with anorexia nervosa and everybody else,” Bulik said. “When most of us get hungry, or starved, we get more anxious. But these people’s bodies respond differently. They say that food deprivation makes them feel more calm and more in control, which is one reason they keep doing it.”

Bulik’s UNC co-authors were Drs. Patrick F. Sullivan, a professor in the department of genetics; Federica Tozzi, who was a postdoctoral fellow in the department of psychiatry; and Helena Furberg, a postdoctoral fellow in the department of genetics. Co-authors from the Karolinska Institute were Drs. Paul Lichtenstein, associate professor; and Nancy L. Pedersen, professor and department head, both in the department of medical epidemiology and biostatistics.


Note: To request an interview with Bulik, call Stephanie Crayton at 919-966-2860 or

For further information please contact Ramona DuBose either by phone at 919-966-7467 or by e-mail at