Eating Disorders Review
July/August 2012
Volume 23, Issue 4

Bridging the Gap between Science
and Evidence-Based Practice

At the 2012 International Conference on Eating Disorders, May 3-6, in Austin, TX, one of the most-discussed topics was the need for and challenges of integrating scientific findings into real-world clinical practice.

In his keynote address, Scott O. Lilienfeld, PhD, Professor of Psychology at Emory University, Atlanta, told eating disorders professionals that in today's world, good science is being crowded out by pseudoscience, and that it is often difficult to distinguish one from the other. It is imperative that eating disorders professionals find a balance between science and evidence-based practice, Dr. Lilienfeld told the audience. "Science," he said, "is necessary, not because it's perfect but because it helps safeguard us against bias, naïve realism, and illusory correlation." In the same way, he added, evidence-based practice is needed because it minimizes these biases.

Dr. Lilienfeld works outside the field of eating disorders, but said his outsider's perspective might prove helpful. He said, "Today we live in a very confusing world filled with lots of medical claims, some of which are well supported, but others are not, and it is often hard to distinguish the science from the pseudoscience. We haven't done as good a job as we could have to distinguish good science from pseudoscience."

"The good news is that psychotherapy works," he said, and added,  "we can demonstrate the positive effects of a variety of psychotherapies for mood disorders, anxiety disorders, sexual dysfunction, insomnia, and bulimia nervosa, for example, when compared with the lack of treatment with a plausible "placebo" intervention. However, researchers such as Drs. Judith Banker and Kelly Klump (2010) have reported a substantial science-practice gap in eating disorders treatment, he added.

According to Dr. Lilienfeld, this science-practice gap is caused by a number of factors, including a clash of world views, and by misconceptions about science in general and of evidence-based practice in particular. He added that researchers have shown that most therapists who treat clients with eating disorders do not administer scientifically supported therapies (Mussell et al., 2000), and that most clients with depression and panic attacks do not receive scientifically supported therapies (Kestler et al., 2001). In addition, Dr. Lilienfeld pointed out that 75% of licensed clinical social workers use one or more unsupported therapies, including age regression, psychodrama, and neurolinguistic programming, for example (Pignotti and Thayer, 2009).

Dr. Lilienfeld then explored a number of popular psychological myths and misconceptions. We are not doing a good enough job of educating the public well about psychotherapy, he said, adding that "good psychotherapy has been under-hyped." And, Dr. Lilienfeld added, the science-practice "gap" could be better described as the science-practice "canyon" because it is so broad. Noting that "our clients deserve the best," he urged audience members to work to close the science-practice gap. Part of the problem lies in a conflict between romantic vs. empirical traditions, as outlined by psychiatrist Dr. Paul McHugh, of Johns Hopkins University. Romantics, Dr. Lilienfeld said, believe that questions are best settled by intuition and clinical experience, not research. In contrast, empiricists believe that questions are best settled by research, not intuition or clinical experience. He believes that this split between romantics and empiricists probably underlies much of the science-practice gap and resistance to evidence-based practice.

A number of misconceptions exist about evidence-based practice, Dr. Lilienfeld stressed, including the belief that it stifles creativity and that it requires a cookie-cutter approach to patients. Other misconceptions are that the approach is not helpful because all individuals are unique; or that evidence-based practice isn't needed because "we can judge the effectiveness using our clinical experience and intuition."

Treatment of eating disorders patients is also affected by confirmation bias, Dr. Lilienfeld said. He explained that there is a tendency to seek out "evidence that supports our individual views and thus to deny, dismiss or distort evidence that does not do so." This tendency affects scientists at least as often as nonscientists, he noted (Mahoney, 1977). One historical example of this was widespread adoption of prefrontal lobotomy in the late 1940s; in fact, the scientist who developed it was awarded the Nobel Prize in 1949.

Another bias is naïve realism, or the belief that the world is exactly as we see it. Illusory correlation, or a tendency to perceive correlations when they are absent or to exaggerate the magnitude of actual correlations is another confounding problem. Some examples of this, he said, are shown in the argument about a connection between vaccinations and autism and the belief in the "lunar lunacy effect," or the belief that the moon has powerful effects upon us.

There are ethical dangers from some of these biases, according to Dr. Lilienfeld, and he pointed to the true cost of lost opportunities when clients spend time and money and effort to seek treatments that can't help them. "I worry that these types of approaches have undermined our credibility—psychology and psychiatry are not viewed well," Dr. Lilienfeld said. He then quoted the late physicist Richard Feynman, who defined the essence of science as "bending over backwards to prove ourselves wrong."

Dr. Lilienfeld and some colleagues are currently writing an article on causes of spurious therapeutic effectiveness. No matter how clever or good we are, there is frequently no way to know if the therapy worked, he said. Some causes include regression to the mean, placebo effects, multiple pre-treatment interference, maturation of the patient, and history.

All individuals in the mental health field want to help people, but we disagree about how to do it, he stressed, noting that "Science is our best hope to root out our errors and is at the same time a prescription for humility." Poor clinical care comes from overconfidence and reluctance to root out errors. He told the audience that evidence-based practice is important because it helps us correct and rule out our errors and mistakes within our ‘web of belief.'

Dr. Lilienfeld urged the audience to try to heal the pernicious divide between science and practice by recognizing the proper place for romanticism, which is to think big, to be bold and to listen to clinical intuition, but then to be able to turn off the romanticism and to use a scientific approach to be rigorous and to realize that everyone is prone to error.

Plenary Session II: Implementing Science into Clinical Practice

In a plenary session, "The Good, the Bad, and the Ugly of Integrating Clinical Research and Practice," Dasha Nicholls, MBBS, MRCPsych, MD, FAED, head of the Feeding and Eating Disorders Service at Great Ormond Street Hospital for Children, London, reminded the audience of the many challenges that face clinicians in "real-world practice." Dr. Nicholls said that in her practice, where she treats children and teens with eating disorders and very ill patients with anorexia nervosa, she may not know all the latest research, but her expertise, like many professionals treating eating disorders, is personal and unique.

Clinical work is complicated, she said, and there is much to process and much clinical judgment comes into play. There is a need for mutual respect and ongoing dialog, she said, but she also called for mutual understanding about others' ways of thinking. One of the issues with different ways of thinking has to do with different purposes of thinking, she added. Dr. Nicholls said, "Through multiple lenses a clinician reaches a clinical judgment, because you must make a decision—patients come to see you for that opinion, and can't wait 10 years for research to prove a treatment approach."

Dr. Nicholls said that the role of a researcher is to take a highly focused research question, perhaps from a review of the medical literature, to look for any possible source of bias, including outcome bias, context and personal biases, and to eliminate such biases. The role of the researcher, then, is to be totally "agnostic" and open-minded about the outcome. This approach is the opposite of the role of the clinician, who cares a great deal about the outcome. Thus, it is easy to see that the purposes of the researcher and the clinician are diametrically opposed, which causes a tension between the confidence a clinician needs and the open-minded and agnostic view needed for good research.

Another potential problem a clinician faces is that his or her judgment will often be different from that of other clinicians, and to mitigate this problem, she said, we build in supervision, writing, and communicating with others. She noted that there is ambiguity about the role of the multidisciplinary team in delivery of empirically supported treatments. Another issue is individual competence versus collective competence, she added. Also, some empirically supported treatments can only be tested and delivered in certain health care contexts, for example, possible acute hospitalization followed by day care vs. inpatient treatment.

To explore the challenge of implementing evidence-based therapy into clinical practice, Dr. Nicholls conducted a brief survey of 23 of her colleagues. She said 73% described themselves as pure clinicians, and 27% identified themselves as clinician-researchers. The question she posed was, "How much of your current practice do you estimate has a sound evidence base?" The majority, 68%, said "some," while 32% replied that "most" of their practice had a sound evidence base. Some of the reasons her colleagues gave for not implementing research into their practices were: (1) research did not ask or answer the relevant question; (2) it can be hard to identify the clinical implications of research; and (3) clinicians don't like using treatment manuals and don't have the time to read research papers. Finally (4) some felt that most treatment manuals are out of date or that research did not apply to clinical work.

When Dr. Nicholls qualified the answers further, she found a number of reasons that research was not implemented or prioritized among the colleagues. Some comments included: ‘It's hard to get everyone to agree on protocols—we all have our own ideas and service constraints'; ‘It is fantastically difficult to do good research that answers the difficult questions,' and ‘There is a need to compare a manualized with an non-manualized version of the same therapy—that is, does the manual matter?

Dr. Nicholls concluded that what she and other clinicians need are "frameworks for practice based on research, clinical experience, and the perspectives of patients and caregivers, that facilitate problem-solving at any individual patient level." While some of this is admittedly already available, there are still huge gaps between research and clinicians, she said. She added that there is a long way to go; what is needed, she said, is a perspective on individuals who are doing well in treatment, those who are not doing that well, and better ways of understanding the treatment response. Other factors are: knowing when intervention should be halted, who should deliver treatment, and a better way to evaluate therapist skill.

A Better Means of Accreditation

Craig Johnson, Chief Clinical Officer at the Eating Disorders Recovery Center of Denver and Professor of Psychiatry at the University of Colorado, proposed an action plan that would lead to increased competency among clinicians and adoption and dissemination of evidence-based therapy among eating disorders professionals. Noting that he and Dr. Chris Fairburn suggested such a path several years ago, he expressed disappointment that not much progress had been made since then.

Dr. Johnson proposed a plan of action that would create a mechanism that disseminates evidene-based therapy and that would identify to the public—a crucial point, he said—professionals who have demonstrated competencies in treatment. To accomplish this, he suggested coordination between researchers generating evidence-based therapy and consumers of treatment. He added that those who have the most influence upon what treatment to deliver are those that pay for the treatment.

Dr. Johnson praised the National Institute for Clinical Excellence (NICE) in the United Kingdom, which has identified certain areas of evidence-based treatment in eating disorders. For example, he said, cognitive behavioral therapy for bulimia nervosa has been graded as an A-level treatment; in the B-level category, where there is research but not randomized trials, interpersonal psychotherapy and the SSRIs are included for bulimia nervosa. He predicted that DBT will soon be included at the B-level and family-based therapy will be included at the A-level in the next NICE iteration.

Noting that "We do have treatments to disseminate, but nobody thinks EBTs are being adopted by most clinicians in the US. One of the challenges is showing expertise, and some progress has been made through the AED Credentialing Task Force, and collaborations with the Academy for Eating Disorders (AED) and the National Eating Disorders Association (NEDA). Dr. Johnson singled out professionals at NEDA who he said pushed the professional community to come forward with some criteria to show expertise in the field. He said, "Dr. Mary Tantillo of NEDA deserves a purple heart for all her efforts."

The crux of the problem was not that the groups couldn't agree on professional standards but rather finding a business model to implement the standards, Dr. Johnson commented. While establishing the standards took a year, finding the correct business model took about 7 years, he added. Dr. Johnson also noted that "We are within months of being able to partner with the joint Commission on Accreditation of Rehabilitation Facilities (CARF), an independent, nonprofit organization that accredits programs of health and human services.

Credentialing task forces can help with this challenge. Dr. Johnson said he would like to see the AED take on the task of credentialing individuals; however, he added that most recognize there are liability concerns and we should respect that, he added. He would like to see the organization extend the scope of the credentialing task force to individual practitioners and to continue the joint venture between IAED and the Academy. Dr. Johnson predicted, "If we in the Academy do nothing, we will see a progressive loss of relevancy to the practitioners out there."

 (In the next issue, the dialogue about efforts to close the gap between research and evidence-based practice continues with presentations by G. Terence Wilson, PhD and Ulrike Schmidt, MD, PhD, FAED.)

UPDATE: Low Self-Compassion
and Eating Disorders

Researchers in British Columbia are exploring the possibility that healthy self-compassion can play a helpful role among people with eating disorders. Dr. Josie Geller and her colleagues at the University of British Columbia and St. Paul's Hospital, Vancouver, reported at the recent International Conference on Eating Disorders meeting in Austin, TX, that 80% of participants in the researchers' ongoing study have low self-compassion, and the remaining 20% have scored in the moderate range. As the researchers suspected, global and subscale self-compassion scores (derived from the Self-Compassion Scale and other measures of disordered psychological health, along with demographic information) strongly and significantly correlated with eating disorder symptoms such as drive for thinness and body dissatisfaction. The preliminary results also could be correlated with other intrapersonal characteristics, including perfectionism, impulse regulation, maturity fears, and introspective awareness. Dr. Geller told the audience that these preliminary findings support the role that low self-compassion may play in the development and maintenance of shape and weight concerns, as well as disordered eating and maladaptive coping. Thus far, 70 patients have been enrolled in the study.

Bipolar Disorder and Eating Disorders:
A Comorbid Rollercoaster

Bipolar disorder takes patients on a lifelong rollercoaster ride, from the highs of the manic phase to the lows of the depressive phase. Comorbidities can also obscure and delay the diagnosis. Three recent studies have focused on connections between bipolar disorder and eating disorders.

Aware that the percentage of patients with bipolar disorders who have comorbid eating disorders has been reported ranging from 2% to 15%, researchers at the Universidade Federal da Bahia, Porto Alegre, Brazil, wondered how having an eating disorder might affect a person with bipolar disorder. Dr. Camila Seixas and colleagues enrolled 356 bipolar patients in their study (Rev Bras Pisquitar 2012; 34:1516). All patients had undergone treatment between March 31, 2005 and March 22, 2009 at two university hospitals in Salvador and Porto Alegre, Brazil. These patients then completed a series of clinical and sociodemographic questionnaires. The World Health Organization Scale for quality of life evaluation was used for a subsample of 180 participants.

The researchers report that the prevalence of eating disorders was higher than in the general population, as other studies have reported. The mean age of study participants was 41 years. Among the study participants, 19 (5.3%) had a current of lifetime diagnosis of an eating disorder. Among the 19, 11 (57.9%) had bulimia nervosa (BN) and 8 (42.1%) had anorexia nervosa (AN). The authors found significant differences among bipolar patients with and without eating disorders, and the group with eating disorders contained a larger number of females than in the other group. The only male participant met the diagnostic criteria for BN. The authors also found that the odds of having an eating disorder increased with the number of comorbidities. And, as in previous studies, the authors found that eating disorders occurred more frequently during the depressive phase; both the HAM-D and the WHO Quality of Life Questionnaire were significant predictors of the presence of eating disorders in bipolar patients.

The study had several limitations: the cross-sectional study design precluded an opportunity to examine any temporal relationship between BD and eating disorders, body mass index (BMI) was not measured, and binge eating disorder (BED) was not included in the study because it is not included as a diagnostic category in the DSM-IV. Finally, all participants were adults. The authors suggest that future studies be designed to differentiate bipolar children and adolescents with comorbid eating disorders from the adult population.

Cormorbidity was associated with more severe illness

In an earlier study, Dr. S.L. McElroy and colleagues at the Lindner Center of HOPE, Mason, OH, reported that patients with bipolar disorder not infrequently have comorbid eating disorders, and this comorbidity is associated with an earlier age of onset and more severe course of bipolar illness (J Affect Disord 2011; 128:191). This particularly affects women. Of 875 outpatients with DSM-IV bipolar I or II disorders, 125 (14.3%) met criteria for at least one comorbid lifetime Axis I eating disorder; BED was more common (77 patients) than was BN (n=42) or AN (n=27). Having a lifetime comorbid eating disorder was associated with being female, and a host of other characteristics, including: younger age, earlier age of onset of mood symptoms and of bipolar disorder, history of rapid cycling and suicidal attempts, greater mean BMI, obesity and severe obesity and family history including depression, bipolar disorder, alcoholism, and drug abuse. Lifetime AN was associated with normal weight and a lifetime anxiety disorder, BN was associated with overweight, and lifetime BED was associated with obesity and severe obesity.

Future study will search the bipolar spectrum in eating disorders

Another Brazilian group at the University of San Paolo is conducting a study, ESPECTRA (Occurrence of Bipolar Spectrum Disorders in Eating Disorder Patients) and is currently enrolling patients in a single-site cross-sectional study designed to evaluate the prevalence of bipolar spectrum in a sample of eating disorders patients (BMC Psychiatry 2011; 11:59. Women 18 to 45 years of age will be evaluated. Stay tuned for results.

Exploring the Brain for Clues
to Anorexia Nervosa

Dr. Laura M. Holsen and colleagues at Harvard had a novel idea--using advanced imaging technology to explore whether patients with anorexia nervosa (AN) have disrupted brain circuitry that interferes with normal appetite and satiety. The authors hypothesized that a causal factor for AN might be lower-than-normal levels of food motivation in appetite-controlling sections of the brain.

Dr. Holsen and colleagues used functional magnetic resonance imaging (fMRI) of the brain among women with active AN, a second group of women diagnosed with AN who were weight-restored, and a group of healthy controls (J Psychiatry Neurosci 2012; DOI: 10.1503/jpn. 11-156). Women with restricting-type AN (active AN group; n=12), and weight-restored women with a history of AN-restricting type or binge eating/purging type (weight-restored group; n=10) between the ages of 19 and 28 years were recruited. All met DSM-IV criteria for a diagnosis of AN. Eleven healthy women from the community who had regular menses, no pubertal delay, and who were currently at 90% to 110 % of ideal body weight (IBW) were added as a control group.

After a 12-hr fast, healthy controls and weight-restored women were studied during the follicular phase of their menstrual cycle (days 1-10). The first fMRI session was set for 8 am, and then participants were asked to consume a 400-kcal mixed meal standardized for micro- and macronutrient content over the course of 15 minutes. When the women finished eating, the staff weighed the portion of the uneaten meal to determine caloric intake. A second fMRI scan began at about 9:15 am. Immediately before and after their fMRI scanning sessions, the participants rated their hunger and desire to eat their favorite food, using visual analog scales and scored their current anxiety level using the State-Trait Anxiety Inventory. After each fMRI session, participants rated a selection of the food and nonfood stimuli on a scale ranging from highly unappetizing to highly appetizing. After the post-meal scan, participants also completed several questionnaires such as the Beck Depression Inventory and the Eating Disorder Examination Questionnaire.

Low-activity areas of the brain were identified

The women with active AN had significantly lower percentages of IBW than did the weight-restored and control groups. Women with active disease also had significantly greater eating disorder symptoms and higher levels of trait anxiety than did the weight-restored and control groups. Women in the control group had higher ratings of hunger and desire to eat their favorite foods than did women with active disease and, to a lesser degree, women in the weight-restored group. For women in the control group, high-calorie food was more appalling and appetizing than for the other groups. Women with active disease showed higher levels of state anxiety than did controls and weight-restored women both before and after scanning

The authors found hypoactivity in the hypothalamus, amygdala, and anterior insula in the weight-restored women compared to control women during periods when appetite was greatest; they believe this is compelling evidence that these low-activity areas of the brain may point to long-term physiological traits tied to AN. They added that the functions of these regions of the brain are not limited to appetite and food intake regulation, but are also highly involved in the response to stressful stimuli in healthy controls and have been implicated in anxiety disorder and obsessive-compulsive disorder.

Self-reported hunger and perception of the reward value of food are significantly altered in women with active AN, and debate continues about whether these abnormalities are driven by physiologic or psychologic processes. The authors add that even though weight gain through restoration of healthy eating patterns is the primary goal of treatment for patients with AN, disturbed eating behavior remains even after recovery. Thus, patients are always at risk of relapse. Evidence of possible  state and trait patterns of hypoactivation in food motivation in regions involved in appetite-related behavior underscore the importance of seeking out other potential state and trait characteristics when investigating brain phenotypes in persons with AN.

Clues in the Brain to Patterns of Eating

Differences in brain circuitry may help explain why one person develops anorexia nervosa (AN) and another becomes obese. Using functional magnetic resonance imaging (fMRI), Dr. Guido Frank and co-workers at the University of Colorado, Aurora, found that reward circuits in the brain were sensitized in anorexic women and desensitized in obese women. fMRI measures brain activity by detecting the changes in blood oxygenation and flow that occur in response to neural activity. When an area of the brain is more active, it consumes more oxygen to meet this demand and more blood flows to the active area and can be measured with fMRI scans.

Dr. Frank and colleagues studied 21 patients with restricting-type AN (mean age: 22 years), 19 obese patients (mean age: 27 years), and 23 healthy control women (mean age: 24 years) with fMRI and a reward-conditioning task Neuropsychopharmacology 2012; May 2. doi:10.1038/nnp.2012.51 [Epub ahead of print]). The participants were visually conditioned to associate certain shapes with either a sweet or a non-sweet solution, and then received the taste solution expectantly or unexpectedly.

Activity in the occipitofrontal cortex was found and used to differentiate all three groups. Brain reward circuits were more responsive to food stimuli in the women with AN but less responsive in the obese women. The authors found that during the fMRI sessions, an unexpected sweet-tasting solution led to increased neural activation of reward systems in the patients with AN and to diminished activation in the obese individuals and control group. In animal studies, food restriction and weight loss have been associated with greater dopamine-related reward in the brain. Although the mechanisms for the association is not yet known, the authors feel these brain reward response patterns could become biomarkers of the respective weight state. The findings also suggest that eating behavior is related to brain dopamine pathways involved in addictions.

EDNOS: A Catch-All Diagnosis?

Binge eating followed by compensatory behaviors such as self-induced vomiting and use of laxatives and diuretics provides the most common symptoms of full- and subthreshold eating disorders. Such behaviors, with or without binge eating, form the basis for a diagnosis of eating disorders not otherwise specified (EDNOS).

According to Dr. Dawit Shawel Abebe and colleagues of the Norwegian Social Research organization, the broad category identified as EDNOS has become a catch-all category for patients, and needs to be better defined. That is, future diagnostic categories in the DSM-5 may need to be better defined to reflect combinations of binge eating with compensatory behaviors, whether these include purging or not. According to the researchers, those who purge are at special risk.

Dr. Abebe and colleagues gathered data from a Norwegian national representative longitudinal study, "Young in Norway," which followed individuals from 14 to 34 years of age and was conducted at four time points: 1992, 1994, 1999, and 2005 (BMC Public Health 2012; 12:32). The study was conducted in 67 representative schools in Norway and sampled thousands of students 12 to 20 years of age, in grades 7 to 12. At each of the 4 time points, students who responded to the earlier study were invited to participate once more, and a 67% overall response rate was recorded.

At each point, students completed self-report questionnaires that examined items such as binge eating and compensatory behaviors (Bulimic Investigatory Test, Edinburgh, or BITE), eating problems (the BITE test and the Eating Attitude Test12), satisfaction with appearance, depressive symptoms (Depressive Mood Inventory), anxiety symptoms (Hopkins Symptom Checklist), and measures of general self-worth, alcohol consumption, relationship with parents, self-concept (Rosenberg's Stability of Self scale), social support, and loneliness. Body mass index (BMI, kg/m²) was computed from self-reported measures of height and weight.

Females were at higher risk

The researchers found that subclinical groups of individuals with differing combinations of eating disorder symptoms showed distinct age- and gender-related trends. For example, females were at significantly higher risk for purging behaviors than were males. The prevalence of binge eating and compensatory behaviors in females and compensatory behaviors in males gradually declined with the transition from adolescence to young adulthood. Individuals with subclinical forms of eating disorders differed significantly on measures of general eating and psychosocial problems.

Purging emerged as the most serious type of behavior, compared with compensatory behaviors, binging-compensatory behaviors, and non-purging compensatory behaviors. Individuals who purged, mostly by using self-induced vomiting, had significantly higher levels of appearance dissatisfaction, anxiety and depressive symptoms, drank larger amounts of alcohol, and were more unstable and lonelier than other participants. For both males and females, the prevalence of compensatory behaviors decreased from age 14 to 16 to 23 years and over. However, for binge eating, a significant decrease was found only for females, whose binge eating also declined more markedly over time than did that of males.

The authors concluded that a substantial proportion of teens report binge eating and using inappropriate weight control methods. As they transition to young adulthood, the risk declines somewhat. According to the authors, early preventive interventions could potentially decrease the likelihood that the inappropriate behaviors will progress to full-threshold eating disorders.

Focusing on Family Dynamics
in Treatment for AN

In what is believed to be the first study of its type, French and Canadian researchers found that the addition of family therapy sessions focusing on the dynamics of the family itself rather than upon eating symptoms improved treatment outcome in girls with severe anorexia nervosa (AN) (PLoS One 7(1):e28249.doi:10.1371/journal.pone/0028249).

The severe effects of AN are well known, and its mortality rate of 10% is the highest among psychiatric illnesses, yet research on treatment is scarce and the many multidisciplinary approaches have rarely been evaluated, according to Dr. Nathalie Godart and colleagues. Thus, the authors designed a study to compare two post-hospitalization outpatient treatment programs for adolescents with severe AN. Patients were randomized to treatment as usual or treatment as usual with additional sessions of family therapy. Sixty female AN adolescents 13 to 19 years of age were included in the randomized parallel controlled trial conducted from 2002 to 2004.

Teens in the treatment as usual group (n=30) had ambulatory care tailored to their individual mental and physical needs, including individual consultations, regular interviews involving the parents and, if required, individual psychotherapy sessions. Parents were advised to be supportive but to leave decisions about food to their teens. They were also asked not to directly discuss difficulties they observed with their daughter during or after the meal but to wait to bring up these issues at the time of consultation with the psychiatrist and their daughter.

The family therapy group (n=30) received treatment as usual, with the addition of a program of family therapy that focused not only on current issues but also on unresolved issues from the past as well as expectations of how these might impact the future. The sessions focused on the family dynamic as a whole and did not address eating behaviors directly (these behaviors were dealt with by the referring psychiatrist). The family therapy sessions included the patient, her parents and her siblings if they were older than 6 and living in the home. These 90-minute sessions were held every 3 or 4 weeks and ran for up to 18 months.

At the time of randomization, and then again 18 months later, a series of questionnaires and assessments were administered. In addition, the patient's body mass index (BMI, kg/m²), menstrual status, use of contraceptives, and the number of hospitalizations during the course of follow-up were recorded. Patient outcome was reported as good when weight was greater than the 10th percentile BMI, with regular menstruation; intermediate outcome was weight greater than the 10th percentile BMI but with amenorrhea (defined as absence of menstruation for the past 3 months); and poor outcome was weight lower than the 10th percentile and/or presence of bulimic symptoms..

Additional therapy led to healthy advantages

The researchers found the two groups had certain characteristics in common: pre-study BMIs were much lower than the third percentile BMI, and post-study BMIs were above 17 kg/m^2. A fourth of the participants had been previously hospitalized for treatment of AN; there were comparable comorbid mood and anxiety disorders, and teens in both groups had received comparable psychiatric consultations. The participants in the adjunctive family therapy group attended an average of 11.8 family therapy sessions, and the total number of treatment sessions did not differ between the two treatment groups. And, as a whole, between the beginning of the study and 18 month follow-up, all participants had significant improvement for all parameters.

A higher proportion of teens fell into the good and intermediate outcome categories after treatment with adjunctive family therapy; as a group, these adolescents achieved a "good" or "intermediate" outcome 3.2 times as often as those from the treatment as usual group. The family therapy group also achieved a healthy weight about 3 times as often as those in the treatments as usual group. When each group was considered separately, only the family therapy group showed a significant evolution in higher BMIs and they had amenorrhea significantly less often.

The authors had hypothesized that adjunctive family therapy would reduce the number of relapses, but this did not occur. They noted that this result was similar to those of other follow-up studies of adolescent AN outpatients, where such percentages range from 25% to 30%.

BOOK REVIEW: Fear of Food: A History
of Why We Worry about What We Eat

(By Harvey Levenstein. University of Chicago Press, 2012; 218 pp; $25)

Since the world is full of potentially fearful things and few people have time to fear all of them, we're usually selective. We choose such things as climate change, cancer, Alzheimer's disease, terrorism, animal rights, various types of intolerance (or even, for some people, tolerance), political instability, and so forth. Each fear has its constituents, and in many instances entire industries have been spawned from the efforts of well-intentioned scientists and policy makers, often reinforced by the not-always-good intentions of media hawks and activists who make careers and money from playing on assorted pet fears. Let's call it "fear porn" when it hits the 24/7 media cycle on the airwaves.

Among topics that prominently tap into deep public wells of fear are those related to food. For people apprehensive about potential contamination, poisonings, unhealthy additives, bad fats, food dyes or preservatives causing anything from cancer to atherosclerosis to ADHD to God knows what, "Fear of Food" ranks high on the list of matters to be concerned about.

In his book Fear of Food, Harvey Levenstein, an emeritus professor of history at McMaster University in Ontario, Canada, serves up a delightful review of the origins, motivations, and sustaining factors behind the huge number of warnings that have emerged over the past 120 to 130 years or so concerning the safety of our food supply. We see repeatedly how good scientific observations transmute into overblown pseudoscientific pronouncements that enter the popular consciousness and become foci for concerns: Germophobias? Autointoxicating foods? Good/bad/too little/too many vitamins? And, for those of us working with eating disorders, lipophobias?

How well are we doing today with respect to accurate information/hype ratios concerning natural-organic-processed-raw-foods in relation to actual individual health and longevity? With regard to the "science" in the food sciences, Levenstein notes the applicability of the clever observation (elsewhere known as "Gibson's Law") that "for every PhD there is an equal and opposite PhD." His bottom line is that we're best off being skeptical about all the claims we hear – that we perhaps take this all with a pinch of salt (and, as one wry reviewer remarks, "just a pinch, though – too much could be bad for you"). And, along with food author Michael Pollan, he emphasizes that eating habits, and food claims, do best with moderation.

Originally presented as a series of invited lectures, this readable book would be well suited for a variety of college courses concerning social history, social psychology, and food science. It would also be fun to give to and discuss with your food faddist friends.

— JY

Smoking and Disordered Eating

Patients with eating disorders can be very creative about finding ways to lose weight, including excessive exercise and dietary restriction, not to mention cigarette smoking. A study at Yale recently examined the relationships between cigarette smoking for weight control and eating disorder symptoms in a community sample of adult female smokers.

Marney A. White reported that 107 adult female smokers (drawn from an original sample of 1141 male and female participants) completed a battery of questionnaires, including the Eating Disorder Examination-Questionnaire (EDE-Q) (Comprehensive Psychiatry 2012; 53:403). The key areas measured included weight-control smoking, including smoking to prevent overeating, smoking to undo the effects of overeating, and smoking to feel less hungry. Women who used these measures were then compared (on the EDE-Q) with women who denied using such behaviors and scores and frequency of binge eating and purging.

The results of this community survey showed that a significant proportion of the adult female smokers participating in the study reported smoking cigarettes for weight control, More than half of the participants reported smoking more when dieting (55.1%; n=59) or smoking to suppress hunger (54.2%; n=58). Forty-three percent (n=46) reported smoking more after overeating and 43% (n=46) reported smoking to prevent overeating. More than a fourth of the women endorsed smoking more to undo the effects of overeating than for the stimulation from nicotine; that is, they turned to smoking as a means to control their weight.

The authors suggest that clinicians treating patients with eating disorders monitor and address smoking with these patients because they may be using smoking to compensate for overeating and/or to help them restrict their food intake.

Life Transitions Can Trigger Eating Disorders

Traumatic life events, such as relationship changes, the loss of a loved one, or a sexual assault, can trigger eating disorders, according to the results of a small study of 26 women and 1 man ranging in age from 17 to 64 years (median age: 27 years). As Dr. Jerica M. Berge and colleagues at the University of Minnesota recently reported (J Clin Nurs 2012; 21:1355), even a small change, such as moving to a new home or enrolling in a new school, may trigger anorexia nervosa (AN) or bulimia nervosa (BN).

The patients in the retrospective study had been receiving treatment for an eating disorder for from 10 months to 18 years. Nine had AN, 1 had a combination of BN and AN, 3 had BN, and the other 14 participants had eating disorders that did not meet the diagnostic criteria for a single, specific eating disorder. The participants were studied with semistructured interviews.

Six transitional events

Notably, since this study involved a small number of subjects and used a retrospective design, its findings should be verified with prospective research and larger samples of subjects. Nevertheless, the authors' findings are of interest. Six themes were identified as family life transitional events that preceded the onset of an eating disorder: (1) school transitions, (2) death of a family member, (3) relationship changes, (4) home and job situations, (5) illness/hospitalization, and (6) sexual abuse or incest.

The researchers found that transitional events in the family life cycle, followed by a lack of needed support during the transitions, could precipitate the onset of an eating disorder. This lack of support at times of stress was particularly important in the evolution of the eating disorder, and the authors suggest that future research might lead to interventions to help the family and patient reduce the stress from these transitional periods. Helping parents become aware of stressors and more supportive might be even more important than trying to solve or fix the individual problem, according to Dr. Berge and colleagues.

Q & A: Rectal Prolapse after
Long-term Laxative Use

Q. One of my patients with anorexia nervosa binge/purge type who has always complained of constipation and who in the past used large quantities of laxative tablets was recently diagnosed with rectal prolapse. What sort of condition is that? Is that common in patients with eating disorders? (JH, Philadelphia)

A. While not a common feature of eating disorders, cases of rectal prolapse have certainly been reported in the literature, primarily in patients with anorexia nervosa of the binge/purge type, but also in patients with bulimia nervosa, particularly those who suffer from irritable bowel syndrome (IBS), chronic constipation, and laxative abuse (Dis Colon Rectum 1997; 40:1382; IJED 2011; 44:95).

Rectal prolapse consists of the protrusion of rectal muscle wall beyond the verge (opening) of the anus. Factors contributing to the chronic constipation include poor diet and inadequate fiber, dehydration, and a variety of other constitutional and external causes including the anticholingeric effects of some medications. The increased intra-abdominal pressures associated with vomiting are also thought to possibly contribute to the mechanism of prolapse. In rare instances patients have been known to engage in "rectal purging," i.e., intrarectal digital dilatation to help relieve the constipation, or for the sensation (IJED 2012; 45:456). Little is known about this phenomenon since few clinicians routinely ask patients about this behavior, and shame is likely to prevent many patients from talking to their clinicians about this practice.

Treating rectal prolapse usually requires surgery, but to prevent future recurrences steps are needed to reduce constipation (and rectal digitation in instances where the patient is practicing this behavior).

— JY