Why Girls Starve Themselves: New Research in Anorexia Nervosa References

The November death of an Israeli fashion model whose weight had dropped below 60 lb was chilling even in a world that prizes rail-thin models as an ideal of feminine chic. Social critics have long blamed the fashion industry's use of such models for inspiring teenagers and young women to engage in extreme dieting. But at the recent Annual Meeting of the California Psychiatric Association, in Huntington Beach, eating disorders expert Walter Kaye, MD, reminded attendees that the causes of anorexia nervosa (AN) relate more to genetics and neurobiology than to size-zero models on catwalks.¹

While a perception persists that AN and bulimia nervosa (BN) are culturally determined disorders, anorexia predates our current culture, said Kaye, professor of psychiatry and director of the eating disorders program at the University of California, San Diego (UCSD). He cited anorexia symptoms described in 1689 as "nervous consumption" by Sir Richard Morton in his work, Phthisiologia, or a Treatise on Consumption.

Despite our societal obsession with dieting and weight, Kaye said, the prevalence of anorexia subtypes is low (0.25% for AN-restricting type, and 0.25% for AN-binge-eating/purging type), which indicates that factors beyond culture are at work.

Symptoms exhibited by patients with AN can be puzzling, Kaye noted. They see themselves as fat but see other people with anorexia as thin. They tend to restrict their eating and have odd food choices, yet they also have an obsessive interest in diet, recipes, and cooking for others. They often exercise compulsively, are anhedonic and ascetic, and find little in life rewarding other than the pursuit of thinness.

Lack of insight on the part of individuals with AN makes it difficult for clinicians to engage them in treatment, Kaye said. They do not see themselves as having problems, do not learn very well from experience, and do not think logically.

Family studies and links to anxiety

Recent research, Kaye said, indicates that biology plays a substantial role in determining an individual's vulnerability to an eating disorder. Family studies show an increased rate of AN, BN, and eating disorders not otherwise specified in first-degree relatives, and the cross-transmission of AN and BN in families.

Twin studies reveal that AN and BN are highly heritable disorders (50% to 80%), so genes are more important than culture or environment, Kaye noted.

"One important new insight over the past decade is that we have come to realize there are traits people have in childhood that put them at risk for eating disorders," Kaye said. "These kids tend to be anxious, perfectionistic, and have an obsessive personality and negative self-evaluation before . . . an eating disorder [develops]."

While malnutrition may exaggerate the behavioral traits, Kaye noted that the traits tend to persist after recovery. He described clinical studies from 3 different continents showing that between 60% and 80% of individuals with AN or BN have an anxiety disorder during their lifetime and that the onset of the anxiety disorder usually precedes the onset of the eating disorder. The most common anxiety disorders comorbid with AN and BN are obsessive-compulsive disorder (41%) and social phobia (20%).²

Hypothesizing about the link between anxiety and AN, Kaye suggested that restricting or binge behavior might be used as a means of reducing dysphoric feelings.

Brain imaging

New brain imaging studies in AN, using functional MRI (fMRI) and positron emission tomography (PET), are enabling researchers to identify and understand the brain circuits that may contribute to such symptoms as altered reward, anxiety, and appetite modulation; altered body image; and obsessiveness.

For example, several lines of evidence support the possibility that disturbances of dopamine function could contribute to alterations of weight, feeding behavior, motor activity, and reward mechanisms in persons with AN. Using PET scans, Frank and colleagues³ studied 10 women at least 1 year after they recovered from AN. Compared with 12 healthy control subjects, they had significantly higher [¹¹C]raclopride binding potential in the basal ganglia's anteroventral striatum irrespective of their age, body mass index, or time since recovery. The researchers also found increased dopamine D₂ and D₃ receptor activity in another part of the basal ganglia called the dorsal caudate, which was related to anorexics' responses to and avoidance of harm. Such differences, the research team concluded, might explain why women with AN often exhibit exaggerated worry and concern about what might happen in the future. In addition, they said, individuals with AN may have a dopamine-related disturbance of reward mechanisms that contributes to their ascetic temperament and a lack of pleasure in eating and other activities.

Another newly released study used the fMRI approach to test the idea that individuals with AN have differential neural activation in primary and secondary taste cortical regions after sucrose and water administration. Individuals with AN often avoid normally pleasurable foods and fail to appropriately respond to hunger.

In that study led by Wagner and Kaye,⁴ the researchers looked at images of the brains of 16 women who had recovered from AN and of 16 control subjects. They measured their brains' reactions to pleasant taste (sucrose) and neutral taste (distilled water). Compared with the control group, individuals who had recovered from AN showed a significantly lower neural activation in the insula, including the primary cortical taste region and ventral and dorsal striatum, to both sucrose and water. In addition, insular neural activity correlated with a pleasantness rating for sucrose in the control group but not in the recovered AN group. The results of the study are the first evidence that individuals with AN process taste in a different way from those without an eating disorder.

Just last month, Wagner and associates⁵ published results from another fMRI study to assess the response of the anteroventral striatum to reward and loss in persons with AN. They used event-related fMRI to examine the blood oxygen level-dependent- signal while participants (13 women who had recovered from AN and 13 healthy comparison women) performed a simple choice and feedback test. In contrast to the healthy controls, the women who had recovered from AN showed greater activation of the dorsal caudate and did not differentiate between reward and punishment in the anteroventral striatum limbic circuits. On the basis of these findings, the researchers concluded that individuals with AN have an impaired ability to identify the emotional significance of a stimulus but have increased traffic in neural circuits that are concerned with planning and consequences.

Treatment approaches

Over time, many people recover from AN and BN, Kaye said, with as many as 50% to 70% getting better in their 20s. Still, 30% of individuals with AN remain chronically ill and 10% die of the disorder, making it the psychiatric disorder with the highest mortality rate. Weight restoration is very important in the short run, to keep patients from starving to death, but many patients continue to have symptoms.

Developing more effective treatments for acute symptoms is frustrating, Kaye added. The first generation of treatment involved weight restoration, using behavior modification, antidepressants and neuroleptics--"without a lot of proof that they were actually successful,"--as well as providing treatment in structured settings. Relapse rates were high.

Goals for the second generation of treatment, Kaye said, include preventing relapse; developing specialized, effective psychotherapies; using controlled trials to identify useful medications; and treating patients without hospitalization.

We are not there yet, he said, because "we have not understood the pathogenesis and physiology of these illnesses. Until we do that, it is hard to come up with effective treatments."

Drugs do not improve weight restoration very much, Kaye said, but SSRIs may prevent relapse in some people after weight restoration.⁶ However, in a recent study⁷ in which outpatients were randomized to receive fluoxetine (Prozac) or placebo for up to 1 year along with cognitive-behavioral therapy (CBT), no benefit for fluoxetine was shown over placebo.⁷

While some modest improvement in persons with AN (weight gain, reduced agitation, less treatment resistance) has been seen in case reports and open trials with atypical antipsychotics, such as olanzapine (Zyprexa), risperidone (Risperdal), or quetiapine (Seroquel), controlled trials are needed. In any case, Kaye has found that individuals with AN often read about the weight gain that can be associated with atypical antipsychotics and then refuse to take them.

Role of Maudsley treatment

Few controlled studies have been done so far on the use of psychotherapy for persons with AN, Kaye said. In one study, CBT was evaluated as a posthospitalization treatment for AN in adults and was found to be more effective than nutritional counseling in improving outcome and preventing relapse.⁸

Family-based therapies appear promising, Kaye said. The family-based Maudsley treatment, developed in England, appears to be effective early in the course of the illness. The treatment focuses on symptoms rather than blame and takes a nonauthoritarian therapeutic stance in which the family is made responsible to re-feed the child.

The NIMH is supporting a 6-center study (including UCSD) of persons aged 12 to 18 years who have AN. The study will compare the Maudsley treatment with psychodynamic family therapy and fluoxetine with placebo. Maudsley family therapy already has become an integral part of the intensive family-based treatment program for AN at UCSD, one of the first of its kind in the country.

Unfortunately, communities often do not have a therapist or program specialized in the treatment of AN, so individuals end up being sent to a residential treatment facility, usually at some distance, making family involvement with treatment difficult. To address this issue, a new treatment program at UCSD has the individual with AN, the siblings, and the parents stay at a nearby residential hotel for about a week. The 40-hour program consists of a comprehensive medical and psychological evaluation of the patient; intensive psychoeducation for the entire family, including a discussion of the symptoms and causes of AN; a 5-session introduction to Maudsley family therapy; other treatments; and discharge planning.

"Before they return to their home states, we do a consultation with the patient's home therapist or physician regarding follow-up care," Kaye said. "We have had some good success with this new approach."

1. Kaye W. Why do they keep doing it? How the brain drives thoughts and behaviors in anorexia nervosa. Presented at: 19th Annual Premier Conference of the California Psychiatric Association; October 5-7, 2007; Huntington Beach, Calif.

2. Kaye WH, Bulik CM, Thornton L, et al. Comorbidity of anxiety disorders with anorexia and bulimia nervosa. Am J Psychiatry. 2004;161:2215-2221.

3. Frank GK, Bailer UF, Henry SE, et al. Increased dopamine D₂/D₃ receptor binding after recovery from anorexia nervosa measured by positron emission tomography and [¹¹c]raclopride. Biol Psychiatry. 2005; 58:908-912.

4. Wagner A, Aizenstein H, Mazurkewicz L, et al. Altered insula response to taste stimuli in individuals recovered from restricting-type anorexia nervosa. Neuropsychopharmacology. 2007 May 9; [Epub ahead of print].

5. Wagner A, Aizenstein H, Venkatraman VK, et al. Altered reward processing in women recovered from anorexia nervosa. Am J Psychiatry. 2007;164:1842-1849.

6. Kaye WH, Nagata T, Weltzin TE, et al. Double-blind placebo-controlled administration of fluoxetine in restricting- and restricting-purging type anorexia nervosa. Biol Psychiatry. 2001;49:644-652.

7. Walsh BT, Kaplan AS, Attia E, et al. Fluoxetine after weight restoration in anorexia nervosa: a randomized controlled trial. JAMA. 2006;295:2605-2612.

8. Pike KM, Walsh BT, Vitousek K, et al. Cognitive behavior therapy in the posthospitalization treatment of anorexia nervosa. Am J Psychiatry. 2003;160:2046-2049. *