Report Calls for Major Improvements in Eating Disorder Research

A landmark review of the eating disorders literature may well help improve the design of future eating disorders research. The recently released report, Management of Eating Disorders, also analyzes treatment efficacy and outcome for anorexia nervosa (AN), bulimia nervosa (BN), and binge eating disorder (BED). The review was released by the Health and Human Services Agency for Healthcare Research and Quality early in April. (To obtain copies of the study, see the box on page 3.)

A Major Finding: Study Designs Varied Widely

One of the major findings from reviewing the treatment efficacy and outcomes for patients with AN, BN, and BED is that the literature is highly variable. The researchers found that for AN, for example, the literature on medications turned out to be “sparse and inconclusive.” For example, not a single study combining medication with behavioral interventions met inclusion criteria.

According to Cynthia Bulik, PhD, Jordan Distinguished Professor of Eating Disorders in the Department of Psychiatry, Professor of Nutrition, and Director of the Eating Disorders Program at the University of North Carolina at Chapel Hill, and an author of the report, many eating disorders management studies were not well designed. Many studies included too few participants, provided too little information about the participants, and ignored the possible harms associated with of some types of treatment. For example, studies that address the optimal approach to re-nourishing patients or the optimal conditions under which a patient should be stepped down or discharged from inpatient treatment are absent from the literature.

AN: No Effective Medications

No medications currently available are effective for patients with AN. This was not surprising, said Dr. Bulik. She added, “It is a common frustration that there aren’t any medications that are effective in AN, especially in the underweight state.” She noted that most medications that have been tried have been used in the hopes that their side effects (that are usually undesirable for the disorders they are used to treat—such as weight gain in depression or weight gain with atypical antipsychotics)—somehow assist with weight gain in AN. This report is not about specific cases and does not mean that medications may not be effective for an individual. The report simply summarizes the randomized controlled trials that have been published. “However, the trials were far too small to have included meaningful, preplanned subgroup analyses that could indicate whether there are certain groups of patients who may respond better than others,” she said.

Treatment for BN and BED: A More Positive Picture

The case seemed to be different for BN, however. Several medications and behavioral therapies are helpful for patients with BN and BED. The researchers found that several types of medications were helpful for BED patients, at least in the short term. Reflecting the state of the science, fluoxetine is the only FDA-approved drug for any eating disorder (approved for BN), and it does not work across the board. Renewed efforts to determine how to treat individuals who do not respond to fluoxetine and do not respond to cognitive behavioral therapy (CBT) are critical next steps, according to Dr. Bulik.

Several behavioral therapies also helped combat BN and BED. Individual or group CBT and interpersonal psychotherapy were useful for reducing the core symptoms of binge eating and purging and for alleviating the psychological symptoms of the disorder. For BED, CBT reduced the number of binge days or binge episodes. It did not, however, help BED patients lose significant amounts of weight.

Family Therapy Was Effective for Some Patients

Another finding from the study was that most types of traditional family therapy were ineffective for adults with longstanding AN. One form of family therapy, which encourages parents to control their child’s nutrition, appeared to be helpful. Dr. Bulik notes that this form of family therapy seems to be effective for younger patients still living in a family context where parents can still take control of the refeeding process. This becomes less feasible as patients grow older, when the issues become quite different, she said. No studies have yet explored other types of therapy with couples, for example, for adult married or partnered patients. “We know virtually nothing about other types of family therapy that could be tried for older patients,” Dr. Bulik said.

Sex, Race, Ethnicity, Gender, Age Ignored

According to Dr. Bulik, the review revealed a virtual lack of information about treatment needs by sex, race, ethnic group, or age. She added, “It was absolutely striking how little we know about these questions. Moreover, we have little to no information about optimal refeeding strategies for anorexia nervosa. Although we emphasize on a daily basis that the first and most critical steps in treating AN are weight gain and re-nutrition, there were no trials that actually looked at how best to help underweight individuals gain weight effectively.”

Stereotypes limited studies. Older studies were affected by stereotypes about eating disorder patients. Some studies did not even include information about gender, assuming that all patients were white and female.

Most studies were too small. Another area that needs improvement is the size of eating disorder trials. Most eating disorders treatment studies are still small, single-site studies. The average study of AN treatment involves 23 patients, said Dr. Bulik. Compared with other specialty areas, studies with such small numbers of patients would not even be considered valid. Future multi-site trials will improve patient recruitment, help buffer high dropout rates, and improve the generalization of results. Working in partnership with insurance companies to enable such trials in the current reimbursement setting may be critical to success.

Challenges for AN treatment in other settings. As the report noted, most clinical trials for AN, in particular, do not adequately reflect the type of treatment usually delivered in the community. In addition, clinical trials for AN do not address some of the key challenges facing the clinicians in inpatient, partial hospitalization, or residential settings. No clinical studies address the best approach to inpatient weight restoration that can achieve the most lasting gain.

Lack of correlation between types of studies. Another weakness that appeared in the review was the lack of “cross-talk” between the outcomes and the treatment literatures. As the report notes, outcomes literature reveals intriguing problems that persist years after the onset of AN. An example is the presence of autism spectrum disorders reported in a cohort of individuals with AN followed for years in Göteborg, Sweden. Such observations could provide critical information to individuals designing new interventions for AN.

More data on recovery time. Studies that address factors associated with successful outcomes in AN and BN should explore recovery patterns and examine how current diagnostic language captures those trajectories, said Dr. Bulik. An example would be a patient with AN who is evaluated 5 years after the onset of the illness and is given a diagnosis of EDNOS. This pattern fails to acknowledge that the patient is on a recovery trajectory from AN and her symptoms may reflect “residual AN” rather than a different diagnostic entity.

Questions of cost-effectiveness. Another area that needs work is analysis of cost-effectiveness of treatment. Only rarely did studies assess the cost-effectiveness of interventions for AN, BN, or BED, according to the researchers.

There also were gaps in the overall evidence base. The researchers found that the literature on AN, BN, and BED was geographically imbalanced. Although the U.S. has contributed greatly to the literature on BN and BED, it has spent much less time on treatment and outcome of AN. This is one area that could be improved.

Much Work Lies Ahead

Obviously, there is much work to do. Partnerships with industry and universities will enable researchers to study larger groups of patients. Patients face real barriers to insurance, and better education about the seriousness of and cost of treating eating disorders will help. This was one of numerous topics recently addressed at a Congressional briefing and the Lobby Day held by the Eating Disorders Coalition. Senators and members of Congress were urged to support legislation that would lead to mental health parity and ensure coverage for eating disorders treatment.