page contents

Approaches to the Reduction of Relapse Rates in Persons with Anorexia Nervosa

August 2001

Contributed by Michele A. Soltis, Penn State College of Medicine


“I am only one; but still I am one. I cannot do everything, but still I can do something.  I will not refuse to do the something I can do.”

                                                                                                            —Helen Keller


With these words, the diligent, sedulous, tireless Helen Keller conveys her belief that persistence, determination, and an unwillingness to settle, surrender, or accept complacency can enable one to overcome immense obstacles, to achieve great goals, or, at the very least, to offer one’s seemingly minute contribution to the realization of a purpose larger than oneself.   By adopting such an outlook, we too can aspire to accomplish those aims which elude us. Whether our focus is on personal strivings or professional endeavors, endurance, perseverance, and steadfastness can assist us in our attempts to actualize our objectives. It is with this mindset that the committees for Healthy People 2010 seek to understand and improve the health and well being of the public at large.


The development of goals and objectives for the advancement of Public Health issues is central to the work of Healthy People. Few such goals are measurable, meaning that the objectives for a given area of Public Health interest are intended to provide “direction for action” in moving towards the fulfillment of a desired, stated end. Areas of interest for which measurable goals are proposed have previously been investigated and evaluated.   As a result, relevant data representative of local, state, or even national populations exists and can serve as a scaffolding to be built upon in future studies. Other objectives are developmental, denoting an array of Public Health topics which are of emerging importance and for which little or no baseline data exists. It is the intent of Healthy People to present such goals and objectives which “provide a vision” for a favored outcome or health status. All Healthy People objectives, whether they be measurable or developmental, must promise value and importance to the Public Health sector.   The objectives should be prevention oriented in nature, should be designed to encourage action, must be supported by scientific evidence, and should afford opportunity for continuity and comparability.


One such area of the Public Health arena considered by Healthy People deals directly with topics encompassed under the umbrella of mental health.   According to Healthy People, mental health cannot be thought of simply as an “absence of mental illness”. Instead, it must be recognized as a condition where successful, prosperous mental functioning prevails. Mental health results in fruitful, prolific activities, flourishing personal relationships, and the devices to adapt to and cope with change, disappointment, grief, and adversity.     Mental health is imperative to an individual’s sense of well being, fulfillment, and interaction with society.

On the flip side, mental disorders connote a faltering in mental health, characterized by derangements in thought processes, behaviors, and moods. Mental disorders may significantly impair one’s ability to function as a productive member of society, and may result in the deterioration of one’s personal relationships and perception of self. The term, mental illness, collectively refers to all of the diagnosable mental disorders.


Unfortunately, an estimated 22% of the U.S. population is plagued by the onset of mental illness in a given year, placing a tremendous burden of disability on the community as a whole. Regrettably, this amount of disability and hardship has gone under recognized, if not unrecognized altogether. The landmark study, Global Burden of Disease, sponsored by the World Health Organization and Harvard University, has demonstrated that mental illness is nearly equivalent to both heart disease and cancer as a cause of incapacity in industrialized nations.   It should be noted that suicide, a leading Public Health concern, is most often attributable to the devastating toll mental illness can take.


Mental illness has been shown to affect people of all ages, genders, races, and socioeconomic groups. Children and adolescents comprise an ever-increasing subset of the population diagnosed with mental disorders. In a given year, one in five children and adolescents between the ages of 9 and 17 are diagnosed with a mental disorder, a behavioral disturbance, or a serious emotional disorder.   An estimated 5% of such diagnosed children and adolescents are severely impaired by their illness. Such mental illness may lead to further difficulties with substance abuse, suicide contemplation and completion, and criminal activity. At the other end of the age continuum, approximately 25% of older Americans suffer with mental disorders such as depression, anxiety, and dementia. These said disorders lie outside of what is considered normal aging.


There is promising news in that a plethora of highly effective treatments have recently become available for the palliation of most mental disorders.   Such said treatments range from myriad pharmacotherapies to various psychotherapies and even to electroconvulsive therapies.     Sadly, recent studies have indicated that only a mere 25% of persons diagnosed with a mental disorder receives medical help in combating their illness. A major measurable objective for the Healthy People 2010 initiative focuses on the need for psychiatric or psychological treatment to be made readily available to inflicted individuals.   Such measures may help to substantially reduce the staggering $80 billion spent per year on disability insurance payments, lost productivity, property loss, crime, and other costs attributed to the detrimental effects of untreated mental illness. In addition to ensuring widespread, accessible treatment for the rehabilitation of those with mental illness, Healthy People strives to eradicate the discrimination and prejudices which plague individuals diagnosed with mental illness.   Such stigmas serve only to present a formidable obstacle to those in dire need of treatment.


It should be noted that as the life expectancy of individuals continues to grow longer, those with untreated mental disorders might live longer, causing the prevalence of mental illness to continue to increase as well.   This ballooning in prevalence will inevitably present daunting challenges in financing, organizing, and delivering much needed mental health care. Taking steps forward today to continue to ensure adequate funding and accessible care will surely make for the ready availability of mental health services tomorrow.


The Healthy People 2010 initiative toils to improve mental health and ensure access to adequate, appropriate, quality mental health care.   In seeking to do so, the initiative has turned its attention to specific issues concerning mental disorders such as schizophrenia, major depression, manic depression, substance abuse and other addictive disorders, panic disorders, obsessive-compulsive disorder, specific phobias, eating disorders, dementia, and post-traumatic stress disorder.   Endeavors to be pursued, as cited by Healthy People, include the creation and implementation of preventive interventions and the promotion of further social and behavioral research.   This research will target the delineation of the possible interweaving of biological, psychological, and environmental processes which may contribute to the manifestation of mental disease. Other quests include the development of better stress management techniques, coping skills, community support programs, and treatment modalities.   Together, such efforts can hopefully help to alleviate the tremendous burden of disease and disability associated with mental illness.



I would now like to shift the focus from the Healthy People 2010 objectives for mental illness in general, to the Healthy People intentions pertaining to one mental disorder in particular. In an effort to curb the devastatingly adverse sequelae of eating disorders, Healthy People has elected to target the evolution of approaches which will hopefully serve to reduce the relapse rates for persons suffering with anorexia nervosa. Currently, little information is available regarding the prediction of clinical course, outcome, and potential for relapse in those patients afflicted with this disease. Attaining such information is complicated by the very complexity of the disorder itself, a dearth of shared terminology and communicative means among those researching this disease, and a paucity of controlled clinical treatment studies and subsequent follow-up investigations. The Healthy People developmental objective aims to cultivate and assess specific interventions that can perhaps prevent relapse after recovery in patients diagnosed with anorexia nervosa.   It is hoped that by decreasing or altogether eliminating the chance of relapse, such afflicted individuals may enjoy a better long-term prognosis and a fuller life.



Anorexia Nervosa is a serious, devastating disease often associated with a chronic course, significant morbidity, and a considerable mortality rate of up to 11%. This eating disorder is characterized by an intense fear of gaining weight, a colossal over-concern with body image and size, and by profound disturbances in eating behaviors aimed at maintaining a low body weight.   Most patients go to extreme lengths to carry out their strange, ritualistic dining patterns in secrecy. Patients with this disease place an excessive amount of importance on body size, allowing their weight to be the sole measure of their personal worth.   Such perceptual derangements lead patients to become dangerously emaciated. Anorexia Nervosa can be diagnosed when the body weight falls 15% or more below the individual’s ideal weight as based on height, age, and gender. According to the Diagnostic and Statistical Manual of Mental Disorders—4th Edition, diagnosis is also contingent on the disturbance of the menstrual cycle in postmenarchal women. The diagnostic criteria state that a postmenarchal woman must be amenorrheic for at least three consecutive menstrual cycles to be considered anorexic.

The prevalence of Anorexia Nervosa is between 0.5 and 1% among all women aged 15-50. Over 90% of all anorexic patients are female. However, an estimated one million men have been diagnosed with the disorder as well.   Anorexia is most frequently found in men who are wrestlers, dancers, or homosexuals. Given that a great deal of emphasis is placed on one’s physical appearance in the gay community, some gay men are led to zealously pursue the thin body ideal. Average age of onset of this disease is age 17, for both sexes.   Onset is rare before puberty or over the age of 40.   This disease has been found to be more common in industrialized societies and in higher socioeconomic classes.

The classic anorexic patient is usually an older adolescent girl who is bright, does well in school, is athletic, usually popular with peers, hails from an upper-middle class family, and is not objectively fat.   She may however be a tad bit overweight, and so she begins a dieting regimen—only to soon slip into the downward spiral characteristic of anorexia nervosa.

The etiologies of this eating disorder remain an enigma, though many theories attempting to speculate on the causes of Anorexia Nervosa abound.   Psychological hypotheses propose that patients may have an immense fear of losing control, may have a poor self-image, may dread the thought of developing adult secondary sexual characteristics and the adult responsibilities that accompany that bodily development, and/or the patient may suffer from the “all or none” way of thinking frequently associated with a borderline personality disorder (e.g. one is either fat or thin, there is no healthy “in-between”).

It has been postulated that many anorexic patients are entwined in suffocating relationships with their families—the self-starvation may be one way in which the patient can assert control over her body, and so seemingly, gain control over her life as well.

Though not specific to eating disorders, past physical and/or sexual abuse may be a risk factor for the development of Anorexia Nervosa.

Social theories hypothesize that societal opinions, which tend to equate a waif-like figure with attractiveness, drive women to develop eating disorders. Such supposed ideologies may in fact be responsible for a small subset of the anorexic patient population—after all, this disease is more common among women who are models, ballerinas, and gymnasts. However, it should be noted that accounts of this disease have been found dating back to periods when societal norms for beauty were very much different—as evidenced by the rotund, robust figures of the women captured in Rembrandt’s Baroque-era paintings.

Familial and genetic data support a heritable basis for this disease.   Family studies reveal an increased incidence in both mood disturbances and eating disorders among first-degree relatives of anorexic patients.   Twin studies show higher concordance rates for monozygotic versus dizygotic twin pairs. Neuroendocrine evidence suggests a biological contribution to the disease, citing aberrations in levels of circulating corticotropin-releasing factor and reduced central nervous system norepinephrine metabolism.

Clinically, anorexic patients do not suffer a loss of appetite; their refusal to eat is in response to their profound fear of gaining weight. Individuals with anorexia nervosa commonly exercise intensely to lose weight and alter their body shape.   Some may primarily restrict food intake as a primary means of weight control (the restricting type of anorexia nervosa), while others employ bingeing and purging to keep the pounds off (the purging type of anorexia nervosa).

Amenorrhea, resulting from decreased secretion of gonadotropin-releasing hormone and subsequent decreased secretion of both leutinizing and follicle-stimulating hormones, is so common that its presence is a diagnostic feature of this disorder. Other common findings include decreased thyroid hormone release, and thus resultant bradycardia, intolerance to cold, and constipation.   The skin may become dry, scaly, and covered in lanugo, or the skin may have an orange-yellowish tint to it due to increased levels of carotene in the blood. Osteoporosis ensues, owing to the decreased levels of estrogen.     Anemia, lymphopenia, hypo-albuminemia, and hypokalemia may occur—the latter contributing to the anorexic patient’s increased susceptibility to cardiac arrhythmia and sudden cardiac death.

Primary management of this patient is directed at the presenting symptoms. All medical complications must be carefully treated and followed.   During starvation, psychotherapy is often of little benefit due to the cognitive impairment that often accompanies the malnourished state.   When patients have stabilized, a therapeutic program, including supervised meals, weight and electrolyte monitoring, psycho-education regarding their illness, nutritional counseling, and individual and family therapy is implemented. Antidepressants are also usually prescribed to treat co-morbid depression.   Clinical recovery does occur in 50-70% of all patients diagnosed with anorexia nervosa who are privy to treatment.

Clinical Recovery is defined as:

  • 1) the achievement and maintenance of a body weight that is at least 85% of what is considered to be the individual’s ideal body weight given the individual’s height, age, and gender
  • 2) the resumption of menstrual periods in women
  • 3) the full development of secondary sexual characteristics in both males and females
  • 4) the cessation of restrictive and/or purging behaviors
  • 5) the termination of maladaptive attitudes regarding dietary habits
  • 6) the demonstration of satisfactory interpersonal relationships

To be considered “recovered”, all of the above criteria must be met and maintained for no less than twelve months.

Studies have estimated that time to clinical recovery may average between 57-79 months, depending on the individual patient.


Studies have suggested that almost 30 to 50% of all patients thought to be successfully treated become ill again within 1 year of achieving clinical recovery. After this first, most difficult post-recovery year, patients may still relapse at the rate of 3% per year. The overall relapse rate is currently about 20%.

Relapse is defined as:

  • 1) the failure of the patient to maintain a body weight that is at least 85% of what is considered to be the individual’s ideal body weight given the individual’s height, age, and gender.
  • 2) the cessation of menstrual periods in women
  • 3) the resumption of restrictive and/or purging behaviors
  • 4) the resumption of maladaptive attitudes regarding dietary habits

Relapse can occur only after the achievement of clinical recovery, as previously defined.

Prognostic predictive factors which may indicate a likelihood for a patient to relapse have been tirelessly sought. Researchers have managed to detect and agree on very few such predictive factors. It has been found that those patients with long duration of illness, those that present for treatment at an extremely low body weight, patients who are of more advanced age at clinical presentation, and those that exhibit purging behaviors are more susceptible to relapse and poorer outcome.   As previously noted, an escalation in the efforts to develop and assay treatment modalities that may help to prevent relapse in these patients is needed. To date, psychiatrists and psychotherapists have been touting a particular kind of psychotherapy, cognitive-behavioral therapy, which may help to reduce relapse rates in anorexics treated with this rehabilitation method.   Other studies have demonstrated the efficacy of certain pharmacotherapies in lowering relapse rates after recovery in these patients.

Let us now take a closer look at these promising, ongoing investigations:

As mentioned, cognitive-behavioral therapy, or CBT, has been said to show promise in preventing relapse in anorexics treated with this psychotherapeutic method.   By combining both cognitive and behavioral therapies, CBT attempts to restructure or reframe the patient’s patterns of perception and situational analyzation in order to modify resultant self-deprecating emotions and destructive behaviors. The cognitive therapy component teaches the patient to recognize that distorted thinking may be responsible for the damaging feelings and emotions which may provoke the patient to indulge in ruinous behaviors.   The behavioral therapy aspect helps to weaken the connections the patient creates between his/her painful emotions and the habitual, injurious behaviors s/he yields to in response. To this end, CBT may employ different behavioral strategies to assist the patient in regaining control. Such strategies include: autogenic, extinction, and relaxation training, biofeedback approaches, hypnosis, graded flooding, and systematic desensitization. CBT is a focused, yet flexible, shorter-term therapy, as it emphasizes the present (with little discussion of one’s past) in an effort to curb the unwanted, oftentimes harmful behaviors. Cognitive-behavioral therapy is emerging as the preferred treatment for conditions other than anorexia nervosa—conditions such as major depression, anxiety disorders, panic attacks, specific phobias, obsessive-compulsive disorders, post-traumatic stress disorder, and substance abuse.   CBT has been said to resemble coaching, mentoring, or tutoring.   It has been successful in encouraging patients to practice valuable coping techniques and emotional management skills.


Another promising approach attempting to reduce the relapse rates in persons with anorexia nervosa involves the use of the anti-depressant medication, fluoxetine. Fluoxetine, or Prozac, is a selective serotonin reuptake inhibitor, primarily prescribed for those individuals suffering from depression.   However, when used in patients recovering from anorexia, the drug has been shown to assist in the prevention of relapse.   Studies performed at the University of Pittsburgh demonstrate that an estimated 62% of patients treated with Prozac for this purpose have maintained their weight at or above 85% of their desired body weight for over one year.   It is important to realize that the Prozac does not treat the eating disorder itself. Investigators have hypothesized that the anti-depressant helps to thwart obsessions, compulsions, depression, and anxiety in these patients, all of which can provoke a relapse.   The Prozac is most effective once the patient has regained weight and is well on her/his way to recovery. It should also be noted that restricting-type anorexics appeared to benefit more from this therapy than restricting-purging type anorexics. This finding raises the possibility that disturbances in serotonin activity may be somewhat responsible for this disorder, perhaps creating a vulnerability for this illness in susceptible individuals.   Further investigations are warranted to fully examine the role that serotonin receptors and other neurochemical activities may play in the pathogenesis of anorexia nervosa.


The question of whether or not eating disorders can be prevented is also an area of ongoing debate and research.   One study evaluated a pilot school-based eating disorder prevention program designed to discourage and reduce dietary restraint among junior-high school age girls. The students’ attitudes and behaviors regarding dietary habits were evaluated both before and after completing the prevention program.   Study results indicate that the students reported an increase in dietary restraint behaviors after completing the prevention program. This may imply that such “prevention programs” may actually be doing more harm than good.

Aside from formal prevention programs, paying careful attention to remarks made about a patient’s or child’s weight, no matter how innocuous the comments may seem, may really make all the difference.

Another proposition regarding an approach to reducing relapse rates in anorexics involves the at-home monitoring of these patients by a Community Psychiatric Nurse after the patient’s discharge from an inpatient or outpatient treatment program. The Community Psychiatric Nurse may offer support and nutritional counseling in maintaining the patient’s weight gain, may effectively discourage the patient from engaging in any restricting and/or purging behaviors, and may reinforce the therapeutic principles implemented during the patient’s CBT sessions.   Studies are currently underway in the United Kingdom to assess the effectiveness and practicality of such an approach.

Other intended research will focus on the efficacy of alternative treatment modalities, such as family therapy sessions, advanced nutritional rehabilitation and counseling, and other various methods of psychotherapy, such as interpersonal therapy, which may help to further modify behaviors and dysfunctional attitudes. Studies examining the use of a sundry of different psychotropic medications in the hopes of reducing the rate of relapse among recovered anorexic patients are also underway.   The necessity of timely diagnosis and effective treatment of co-morbid psychiatric conditions is becoming increasingly apparent as well.

In addition, there is a great need for future studies that possess the following methodological strengths: For instance, although the prevalence of this disease is admittedly small, studies utilizing prospective sampling and data collection may prove to be beneficial in investigating this malady.   Such a prospective study may be less susceptible to recall and report errors and biases. Relative risks, attributable risks, and the numbers needed to treat may demonstrate the statistical significance of such a study. Multiple variable analyses and subgroup analyses may assist in adjusting for potential confounding variables.   Such studies may also be well suited for verifying prior associations in the cause-effect relationship, extrapolating relevant data to similar groups, and even extrapolating beyond the data and to other populations outside the study populace.

Other future research endeavors may find that studying shorter intervals between follow-up periods and using the direct patient interview as a means of data collection may strengthen the said investigation.

Explicit, well-defined inclusion criteria is needed to better characterize the study population and to narrow the group to which the study’s results can be immediately applied or extrapolated.

More appropriate, useful signification of outcome measures is also needed. This includes the demand for universally accepted, standardized definitions for terms such as recovery, remission, relapse, and recurrence.

Larger sample sizes, which can be followed for extended periods of time, may allow a reasonable chance of demonstrating a statistically significant difference between study groups if, in fact, a difference actually does exist in the larger population from which the study samples were drawn. Such larger study samples may posses great statistical power and hence little type II error if in fact, the study’s hypothesis is found to be true.

Lastly, future studies must better qualify the presentation and duration of the index episode, subsequent symptom-free intervals, and later periods of clinical recovery, recurrence, and relapse using longitudinal methods of data evaluation.   Life tables, which allow for the examination of the experience of many groups of individuals over time, may be used.   Examples of such longitudinal life table methods include the Kaplan-Meier and the Cutler-Ederer tables. Regression models, which may describe the association between one dependent variable and one or more independent variables can also be utilized in this capacity.   Such regression techniques may as well prove to be helpful in adjusting for confounding variables.

With our renewed, resolute, unflagging determination and steadfast effort, perhaps answers can be found to the myriad questions posed by this devastating disease. Perhaps this next decade will bring prolonged, permanent recovery and relief to those wasting away at the hands of this practically silent, secretive killer. Perhaps we can help to provide those suffering, emaciated souls with hope for tomorrow and the promise of a new day. We cannot refuse to do the something we can do.



  • 1) Beaumont, P., et al. Treatment of Anorexia Nervosa.   Lancet, 1993, 341: 1635-40.
  • 2) Calvo, S., et al. Between 5 and 9 Years’ Follow-up in the Treatment of Anorexia Nervosa. Psychotherapy & Psychosomatics, 1989, 52:133-9.
  • 3) Carter, J., et al. Primary Prevention of Eating Disorders: Might it Do More Harm Than Good? International Journal of Eating Disorders, 1997, 22: 167-72.
  • 4) Deter, H., and Herzog, W. Anorexia Nervosa in a Long-term Perspective: Results of the Heidelberg-Mannheim Study.   Psychosomatic Medicine, 1994, 56: 20-7.
  • 5) Eckert, E., et al. Ten-year Follow-up of Anorexia Nervosa: Clinical Course and Outcome. Psychological Medicine, 1995, 25: 143-56.
  • 6) Foppiani, L., Luise, L., et al.   Frequency of Recovery from Anorexia Nervosa of a Cohort of Patients Re-Evaluated on a Long-Term Basis Following Intensive Care.   Eating and Weight Disorders, 1998, 3: 90-4.
  • 7) Gillberg, I., et al. Anorexia Nervosa Outcome: A Six-year Controlled Longitudinal Study of 51 Cases Including a Population Cohort. Journal of the American Academy of Child and Adolescent Psychiatry, 1994, 33: 729-39.
  • 8) Herzog, D., et al. Outcome in Anorexia Nervosa and Bulimia Nervosa.   A Review of the Literature. Journal of Nervous and Mental Disease, 1988, 176: 131-43.
  • 9) Herzog, D., et al. Recovery and Relapse in Anorexia and Bulimia Nervosa: A 7.5 Year Follow-up Study. Journal of the American Academy of Child and Adolescent Psychiatry, 1999, 38: 829-37.
  • 10) Isager, T., et al. Death and Relapse in Anorexia Nervosa: Survival Analysis of 151 Cases. Journal of Psychiatric Research, 1985, 19: 515-21.
  • 11) Kaye, W., et al. Double-blind Placebo-controlled Administration of Fluoxetine in Restricting and Restricting-purging Type Anorexics. Biological Psychiatry, 2001, 49: 644-52.
  • 12) Kennedy, S., and Garfinkel, P. Advances in Diagnosis and Treatment of Anorexia Nervosa and Bulimia Nervosa.   Canadian Journal of Psychiatry, 1992, 37: 309-15.
  • 13) Meades, S.   Suggested Community Psychiatric Nursing Interventions with Clients Suffering from Anorexia Nervosa and Bulimia Nervosa.   Journal of Advanced Nursing, 1993, 18: 364-70.
  • 14) Peterson, C., and Mitchell, J. Psychosocial and Pharmacological Treatment of eating Disorders: A Review of Research Findings.   Journal of Clinical Psychology, 1999, 55: 685-97.
  • 15) Pike, K. Long-term Course of Anorexia Nervosa: Response, Relapse, Remission, and Recovery.   Clinical Psychology Review, 1998, 18: 447-75.
  • 16) Strober, M., et al. The Long-term Course of Severe Anorexia Nervosa in Adolescents: Survival Analysis of Recovery, Relapse, and Outcome Predictors Over 10-15 Years in a Prospective Study.   International Journal of Eating Disorders, 1997, 22: 339-60.
  • 17) Touyz, S., and Beumont, P. Anorexia Nervosa: A Follow-up Investigation.   Medical Journal of Australia, 1984, 141: 219-22.
  • 18) S. Department of Health and Human Services. Healthy People 2010. 2nd ed. With Understanding and Improving Health and Objectives for Improving Health. 2 vols. Washington, DC: U.S. Government Printing Office, November 2000.
  • 19) Zipfel, S., et al. Long-term Prognosis in Anorexia Nervosa: Lessons from a 21-year Follow-up Study. Lancet, 2000, 355: 721-22.
Share This