August 2009  << Back  

  Can't open the PDF? Click here for help.



Anorexia nervosa treatment from the patient perspective: A metasynthesis of qualitative studies

Cybele Ribeiro Espíndola, MSci

Department of Psychiatry, Federal University of São Paulo/UNIFESP, São Paulo, Brazil

Sergio Luís Blay, MD, PhD

Department of Psychiatry, Federal University of São Paulo/UNIFESP, São Paulo, Brazil

BACKGROUND: Anorexia nervosa is a complex condition that brings great challenges at each stage of treatment. This study aims to organize the body of information available in qualitative studies about the treatment of anorexia nervosa through a systematic literature review and metasynthesis.

METHODS: Searches were carried out on the following databases for the years 1990 to 2005: PubMed, ISI, PsycINFO, EMBASE, LILACS, and SciELO from 1990 to 2005. A meta-ethnographic approach was used to synthesize the data through second-order and third-order interpretations.

RESULTS: The search revealed 3415 studies, of which 16 were selected. Two concepts for second-order interpretation emerged from the process: (1) the process of change (phases of recovery, factors favoring recovery, factors limiting recovery) and (2) perception of the treatment modalities. From the second-order categories, a third-order category was derived—reconciliation, in which personal and environmental acceptance have a central role.

CONCLUSIONS: Recovery from anorexia nervosa, as a very complex process, goes well beyond conventional treatment. Self-acceptance, determination, and spirituality are equally important elements.

KEYWORDS: anorexia nervosa, treatment, systematic review, metasynthesis



Various intervention methods are used for treatment of anorexia nervosa. However, experimental studies testing the efficacy of these methods have shown limited results, even considering short-term follow-up.1-5 Frequently, those who remain in treatment do not adhere to instructions and, when they do adhere, they are at great risk of relapse. In a follow-up study of women with anorexia nervosa who participated in interpersonal psychotherapy, Keel et al6 found a 36% relapse index. According to Treasure and Ward,7 about 50% of patients who participated in multiprofessional treatment achieved full recovery, 20% remained with residual symptoms, and 30% presented a chronic course independent of the treatment used. The mortality varied between 0% and 25%.1 Many individuals with anorexia nervosa display bulimic behavior, and most bulimic individuals have a history of anorexia.2

Guidelines for the treatment of patients with eating disorders have recently been launched in many countries;2,8-10 these guidelines are based on the available scientific evidence. In the case of anorexia nervosa, the guidelines are based on very weak evidence. For example, in a study titled “The only evidence that anything works in adult anorexics,” cognitive behavioral therapy reduced the rate of relapse from 53% to 22%.11 A review of 5512 studies on the same topic found only 6 studies that fulfilled scientific criteria, from which only 2 indicated some effect of treatment.12 In recent years, progress in terms of anorexia nervosa treatment, and particularly outcome, has been modest.1

Most of the studies about women’s recovery in anorexia nervosa have used quantitative methods that do not focus on women’s diverse experiences of recovery. However, a number of studies using qualitative methods have been presented recently. The results of these studies reveal that recovery from anorexia may depend on more than treatment factors. In a review of 23 studies, Bell13 investigated the opinions of patients with eating disorders as to which treatment they considered the most useful. Overall, treatments involving the psychological context of various theoretical and methodological approaches were viewed as very useful, whereas medical interventions focused exclusively on weight were seen as not useful.

From a theoretical point of view, the understanding of the clinical manifestations of anorexia nervosa has been enhanced by psychodynamic and cognitive investigations. The psychodynamic theories were enriched by the work of Bruch, Boris, Skårderud, Farber et al, and others.14-17 Among various considerations, anorexia nervosa is seen as a multidetermined symptom, which may include attempts to create a new identity, cope with counter-attacks to the self, give birth to a true self, devise defense mechanisms to cope with parental conflicts, manage annihilation anxieties, and develop “emotional metaphors” to articulate where emotions materialize in the body. The cognitive theories, including those put forth by Kleinfeld and collaborators,18 are supported by two basic assumptions. The first is that avoiding food is primary for the maintenance of the disorder. The second is that anorexia nervosa would have a positive function in the patient’s life, thus offering a way out of the difficulties faced at different stages of the development of the disorder, in addition to the cognitive distortions that accompany it. In this sense, the disorder to which the patient is attached is constantly being reinforced. Other important factors may include support from family, contact with friends, engagement in activities that help the patient focus on something other than the eating disorders, and experiences that improve self-esteem.19-21 Other cognitive models of anorexia nervosa are proposed by other investigators. Vitousek and Gray22 proposed a combined theoretical model, in which psychodynamic and cognitive elements participate. For that purpose, they combine the treatment of anorexia nervosa with psychodynamic techniques allied to the approach involving false beliefs, nutritional issues, and difficulties with family. Fairburn et al23 have created a “transdiagnostic” model, in which they aim to explain how anorexia is maintained. Their model is based on psychological therapies, particularly cognitive-behavioral therapy, and so suggests areas in which clinicians could provide psychological treatment. The key premise of this model is that all major eating disorders share core types of psychopathology that help maintain the eating disorder behavior. These include clinical perfectionism, chronic low self-esteem, mood intolerance (inability to cope appropriately with certain emotional states), and interpersonal difficulties.

Metasynthesis is a method that involves rigorous examination and interpretation of the findings of a number of studies using qualitative methods.24 It relies on the synthesis of themes and textual quotations from qualitative reports, and the goal is to produce new and integrated interpretations of findings that are more substantial than those resulting from individual investigations.25,26 Synthesizing the findings of qualitative studies on women’s recovery in anorexia has important implications, both for increasing knowledge of the therapeutic process for treating these patients and for the possibility of modifying clinical practice through more advanced understanding.26

Keeping in mind the difficulty of treating these patients and the fact that until now, no reviews on this subject have been published, our goal was to synthesize studies that used qualitative methodology through a systematic review and metasynthesis of how patients understand anorexia nervosa treatment and recovery.


The research encompassed 3 distinct phases: (1) systematic literature review, (2) critical appraisal of the articles, and (3) metasynthesis.

1. Systematic review

Search sources. An exhaustive electronic bibliographic search was carried out using the following databases: PubMed, LILACS, SciELO, ISI, PsycINFO, and EMBASE.

Search strategy for electronic databases. This research used the following key words: eating disorder, anorexia nervosa, AND qualitative research; qualitative study; phenomenology; perspective; perception; experiences; and comprehension—respecting the peculiarities of each database.

Selection of qualitative studies. The selection of manuscripts is not free from debate. As proposed by Dixon-Woods et al25 we opted for a quality inclusion strategy.

INCLUSION CRITERIA. (a) Studies were published in English, Spanish, French, or Portuguese in the 15 years from 1990 to 2005. (b) Articles had to report qualitative research about patient experiences with anorexia nervosa, according to DSM-IV criteria.27 We focused on patients with anorexia nervosa independent of the degree of severity, and included adolescents and adults only. (c) Articles had to report the following methodological structure: original study; clear theoretical framework, and purposeful sample with sample size defined by saturation; analysis based on qualitative methods of data extraction; and results obtained through text-based transcriptions and interpretations.

EXCLUSION CRITERIA. (a) Chapters or books, as well as masters theses or dissertations; (b) studies focused on psychiatric disorders or comorbidities other than eating disorders; (c) investigations assessing children or the elderly; and (d) secondary analyses of previous studies

2. Critical assessment of papers.

The adopted quality criteria were assessed by the standardized form Critical Appraisal Skills Programme (CASP),28 which traces lines for quality appraisal of qualitative research. CASP consists of 10 items that allow classification of articles in categories referring to methodological structure. The studies were classified as categories A or B.

Category A studies have a low risk of bias. They meet ≥9 of the 10 items proposed. Adopted criteria are as follows: (1) the objective is clear and justified; (2) the methodological design is adequate to the objectives; (3) the methodological procedures are presented and discussed; (4) the sample selection is purposeful; (5) data collection is described, including specific instruments and saturation process used; (6) the relationship between researcher and interviewee is considered; (7) ethical cares are considered; (8) the analysis is dense and well-founded; (9) the results are presented and discussed, point to the credibility aspect, and use triangulation; (10) the contributions and implications of the knowledge generated by the research, as well as its limitations, are discussed.

Category B studies meet ≥5 but <9 of the 10 items proposed. Assignment of studies to category B status means that they partially meet adopted criteria, presenting a risk of moderate bias. Case studies and convenience sample belong to this category.

There were no disagreements between the two researchers as to which studies were deemed to fit the inclusion criteria for the review.

3. Metasynthesis

Extraction and data synthesis. Metasynthesis is a method involving induction and interpretation that provides an alternative to traditional synthesis methods by allowing the researcher to understand and transfer ideas, concepts, and metaphors across different studies. Meta-ethnography is one of the most well developed and frequently used methods for synthesizing findings of qualitative studies.29-32 Part of the appeal of meta-ethnography to qualitative researchers is its potential to preserve the properties of primary data and facilitate the identification of themes that run both within and across studies. It is also one of the few methods to provide explicit guidelines for conducting a synthesis. Unlike traditional review work, meta-ethnography aims to derive new insights. The meta-ethnographic method involves selecting relevant empirical studies to be synthesized and then reading them repeatedly and noting key concepts. These key concepts become the raw data for the synthesis. The synthesis is achieved through 3 techniques:

1. Reciprocal translation analysis, which entails examining the key concepts across each study. An attempt is made to translate the concepts into each other based on a comparative approach. Judgments about the ability of the concept of one study to capture concepts of others are based on the attributes of themes themselves, including cogency, economy, and scope. The concept that is considered “most adequate” is chosen.

2. Refutational synthesis, in which the key concepts and themes in each study are identified and contradictions between the reports are characterized. The “refutations” are examined, and an attempt is made to explain them.

3. Lines of argument synthesis, which involves building a general interpretation grounded on the findings of the separate studies (similar to comparative analysis of the Grounded Theory).

Although these translations allow comparisons between different studies, they preserve the structure of relations between concepts. The translation process goes through two stages as suggested by Noblit and Hare.33 The first, called second-order interpretation, is based purely on original results and is the basis for the synthesis itself. The contexts and concepts relevant to each study are registered for a better understanding of interpretations. Text rereadings are done to standardize terminology and incorporate new concepts. The start of the synthesis process translates the findings from an individual study to provide an understanding of how the work interrelates with others. Each new concept is examined through convergent and divergent cases using a process called reciprocal translation.

The second stage is called third-order interpretation. In this stage, interpretation goes beyond the meaning of the original results and interpretations, advancing conceptually and deriving a new reading of the original categories synthesized. As a result, third-order interpretation can constitute a new construction of hypotheses or theories concerning the area of study.

The reading and extraction of categories were carried out by two independent reviewers (C.R.E. and S.L.B.). Categories used in assessment and metasynthesis were obtained through a consensus among appraisers. One of the articles (Woods34) was used as a reference for organizing the comparison process between the different investigations.


The combined search strategies yielded 3415 documents. The abstracts and titles of the citations were read; 2995 studies were rejected, the majority of which did not use qualitative methodology or were note written in the adopted language inclusion criteria. The remaining 420 documents were read in detail. An additional 404 manuscripts were rejected for the following reasons: the study had other objectives; the study did not focus on the theme; the sample was outside the age range (ie, elderly or children); the study was not original; the study was theoretical; the methodology was insufficiently described or inappropriate; clinical descriptions were inadequate; or the study included epidemiological studies on clinical samples or in the community. The final number of investigations was 15. According to the quality criteria applied, 2 studies were classified as category A and 13 as category B. TABLES 1 AND 2 illustrate some of the methodological characteristics of studies on anorexia nervosa and on mixed disorders, anorexia nervosa, and bulimia, respectively.


Studies of anorexia nervosa (N=10)

Author/year Country Participants/treatment modality Method of data collection Study qualitya
Hsu et al (1992)41 United Kingdom 6 patients (age 17 to 44)/outpatient treatment Case study B
Surgenor et al (2003)43 New Zealand 7 women/outpatient and hospital treatment Interviews B
Tozzi et al (2003)19 New Zealand 69 women (average age, 32.3)/outpatient and hospital treatment Interviews, questionnaires B
Tan et al (2003)37 United Kingdom 10 women patients (age 13 to 21) and 9 of their mothers outpatient and hospital treatment Interviews B
Tan et al (2003)45 United Kingdom 10 women patients (age 13 to 21) and 9 of their mothers outpatient and hospital treatment Interviews B
Williams et al (2003)46 United States 28 women (age 18 to 43)/outpatient treatment Interviews B
Chan and Ma (2003)42 China 1 patient (age 25) and family/ outpatient treatment Interviews, videotapes, family therapy sessions B
Colton and Pistrang (2004)34 United Kingdom 19 women (age 12 to 17)/hospital treatment Interviews A
Lamoureux and Bottorff (2005)39 Canada 9 women patients (age 19 to 48)/outpatient and hospital treatment Interviews, grounded theory B
Weaver et al (2005)38 Canada 12 women patients (age 14 to 63)/outpatient treatment Interviews, grounded theory B
a A: low risk of bias; B: moderate risk of bias (per Critical Appraisal Skills Programme criteria28).


Studies with mixed eating disorders/anorexia nervosa and bulimia (N=5)

Author/year Country Participants/treatment modality Method of data collection Study qualitya
Redenbach and Lawler (2003)20 Australia 5 women (older than 18)/outpatient and hospital treatment Interviews A
D’Abundo Chally (2004)35 United States 20 women/outpatient and hospital treatment and as well as some without treatment Interviews, focus group, participant observation B
Woods (2004)36 United States 16 women and 2 men (age 18 to 21)/alternative treatment setting; did not accept treatment Interviews by e-mail B
Cockell et al (2004)40 Canada 32 subjects/outpatient and hospital treatment Interviews B
Keski-Rahkonen and Tozzi (2005)22 United States 32 women and 3 men (age 13 to 53)/alternative treatment setting Internet messages B
a A: low risk of bias; B: moderate risk of bias (per Critical Appraisal Skills Programme criteria28).

Samples from the 15 studies included 306 subjects (298 women, 97%; 8 men, 3%), age 12 to 63. Most studies (10 of 15) examined subjects over age 18. Studies were conducted in the following countries: 4 in England, 4 in the united States, 3 in Canada, 2 in New Zealand, 1 in Australia, and 1 in China. Most of the studies (13 of 15, 86.67%) used semistructured interviews, but other formats included 1 case study, 1 focal group study, and 1 study that used messages on blogs on the Internet. More than half of the studies were published after 2000.

Most patients were seen in clinics and hospitals (n=8), whereas others were seen only in outpatient clinics (n=3), or exclusively in hospitals (n=1). Two studies examined patients in alternative treatment settings,27,28 and no studies were based on commentaries by family members.

Second-Order Interpretation

From readings and the saturation process—the point at which no new information or themes are observed in the data—two second-order themes were identified in the studies: (1) the process of change with treatment (phases of recovery, factors favoring recovery, and factors limiting recovery) and (2) perception of treatment modalities. (See TABLE 3 for second-order concepts and interpretations.)


Synthesis: Second-order interpretation/Process of change

Phases of recovery
Centered identity in anorexia21,34,37-39
Factors favoring recovery
Satisfactory affective relationships19,20,35,36,40,41
Application of cognitive/psychological abilities and nutritional knowledge38,40
Increasing vital space19,40,41
Factors limiting recovery
Fear of change37,42
Lack of motivation20,34
Lack of social support36,40
Professional rigidity43,45
Media influence36,46
Perception of treatment modalities
Pharmacologic treatment19,21,34
Nutritional treatment21,34
Individual psychotherapy19,34,39,41
Group therapy19,20,34

The process of change

PHASES OF RECOVERY. This category includes behavior and cognition constituting the recovery process, which is presented in stages.

  1. Centered identity in anorexia.21,34,37-39 First, the identity of the subject is centered on anorexia nervosa, and subjects believe that the condition is part of their identity. As a result, fighting against anorexia nervosa is a difficult fight against one’s own identity.

  2. Self-knowledge.20,38,39 The self-knowledge process expressed in some commentaries allows patients to find their identity, promoting a better understanding of the self and the construction of new paradigms about themselves and about life.

  3. Self-acceptance.19-21,35,36,38 With the development of self-knowledge, patients with anorexia nervosa may change the dysfunctional view they have of themselves and begin to function in a more integrated manner. Returning to previously held values and establishing expectations are fundamentals of the process of change.

  4. Determination.19-21,35,38 Later, after developing a broader understanding of the experience, the patient adopts determination, ie the movement to have initiative and assume a commitment to the reconstruction of new behavior.

  5. Maintenance.20,38,40 In this category, the patient tries to maintain and consolidate what was conquered before. Inasmuch as new habits are maintained, existence without anorexia nervosa becomes possible.

FACTORS FAVORING RECOVERY. This category identifies factors that facilitate the recovery process. Studies focused on the patient’s own perspective on recovery consider extra-professional help of great importance in recovery. Satisfactory affective relations, increased living space, and spirituality are the most common factors.

  1. Satisfactory affective relations.19,20,35,36,40,41 Satisfactory affective relationships, whether with partners, family members, or friends, are emphasized. Acceptance and understanding are central elements in these relationships.

  2. Application of cognitive/psychological abilities and nutritional knowledge.38,40 Distorted ideas about food, such as caloric value, are prejudicial; in this light, knowledge helps patients in modifying inappropriate food attitudes.

  3. Increasing vital space.19,40,41 Power concentrated in aspects of life other than the body and the disorder also help in recovery. This process includes concrete changes. In this way, actions such as moving to a new house, getting a new job, or taking care of an animal represent useful attitudes and symbolize a new beginning.

  4. Spirituality.19,35,41 Having religious belief and faith were also identified as important aspects in the recovery process.

FACTORS LIMITING RECOVERY. There are various internal and external barriers a person with anorexia nervosa must pass through in the recovery process. Fear of change, ambiguous feelings, and absence of social support as well as professional rigidity and indifference are some of the barriers most often mentioned by patients.

  1. Fear of change.37,42 This fear involves life changes in general, including changes in weight.

  2. Lack of motivation.20,34 Treatment will be successful and recovery possible only if the patient is motivated.

  3. Ambivalence.34,43 “I don’t want to change my body shape; I want to be free of disturbance but not of my body”…”I really want my body reduced as much as possible.”43 These comments reveal ambivalence about treatment and recovery. Patients become anxious about both being free of the disorder as they also desire to remain thin.

  4. Lack of social support.36,40,44 Cockell et al40 and De la Rie et al44 identified lack of social support as a barrier to change.

  5. Professional rigidity.43,45 “Health professionals need to understand that people hide behind anorexia; and the sterner and more rigorous they are, the more we will hide behind it and so it will get stronger.”43 An imposing, indifferent manner does not seem to garner results with these patients.

  6. Media influence.36,46 “I hate TV. It brings all kinds of bad old feelings back to me. I don’t watch anymore.”46 The problem with media influence is how the message is received, interpreted, and understood by individuals with anorexia nervosa.

Perception of treatment modalities. This category investigates patient perceptions of the different types of interventions that constitute treatment for anorexia.

  1. Pharmacologic treatment.19,21,34 Although without much emphasis, the use of medication is considered useful in addressing mood problems and reducing anxiety, but it is not considered important for addressing the central aspects of anorexia, in that it is limited to organic determinants.

  2. Nutritional treatment.21,34 “An eating disorder does not disappear just because you start eating right.”21 This category shows that the treatment of anorexia nervosa is not exclusively a question of weight and eating habits. Nutritional treatment designed to increase weight, which may at times involve the imposition of certain rules, is understood to be unsatisfactory, in that no consideration is given to the psychological aspects of anorexia nervosa, nor does it provide emotional support to the patient.

  3. Individual psychotherapy.19,34,39,41 Individual psychotherapy was considered useful by most patients, as a way of managing the emotional aspects of anorexia nervosa and in promoting motivation and willpower. Psychotherapy is also considered a privileged space where the patient can feel understood and accepted in an unconditional manner.

  4. Group therapy.19,20,34 Individuals’ positions within a group may bring out the ambiguity that is attributed to group therapy, which is sometimes considered useful but also has some caveats. Being together with people who also have anorexia nervosa may be beneficial in the sense of promoting support, but it may also entail a series of negative effects that result in increased stress, eg, patients may compete for thinness. Being together with other participants can also mean learning new bad habits concerning eating disorders, which may make patients feel even worse at times.

Third-Order Interpretations

The second-order categories allow for the formulation of a third-order category: the process of reconciliation with oneself. This meta-category was not present individually in the articles studied but brings together a great deal of information relative to the treatment of anorexia nervosa (see TABLE 4).

The “reconciliation process” includes 3 axes of understanding. The first is characterized by the acceptance of anorexia nervosa as an affliction, in that one of the characteristics of the process is a lack of criticism of the problem. Patients often consider this eating behavior normal and do not perceive it as a behavior problem. They adopt dietary restriction as a way of life, presenting various arguments and intellectualizations on the subject. The second axis considers the person’s need to perceive that they are an object of acceptance by others, especially a family member or a health care professional. This process includes the notion of approximation, and not confrontation, with what the patient is or represents. Finally, self-reconciliation is achieved (see TABLE 4).


Synthesis: Third-order interpretation/ Reconciliation

Derived from the second-order interpretations
  • Recovery phases (centered identity in anorexia, self-knowledge, self-acceptance, determination, maintenance)

  • Factors favoring recovery (satisfactory affective relationships, broadening of vital space)

  • Factors limiting recovery (fear of change, lack of motivation, ambivalence)


Second-Order Interpretation

Second-order analysis resulted in two concepts: the process of change and the perception of treatment modalities. The process of change is complex and long and is made up of various phases. Initially, a person with anorexia nervosa lives the affliction so intensely as to forget about himself or herself. Life becomes anorexia nervosa, as if the person has been swallowed up. In a theoretical study, Buckroyd47 argued that, in the process of change, it is essential to separate anorexia nervosa from the person and his or her individuality and healthy qualities.

Regarding factors that facilitate change, the following were identified as contributing to treatment success: satisfactory affective relationships that provide acceptance and understanding as key elements; broadening life space with involvement in different activities; development of cognitive abilities through psychoeducational work about inadequate eating habits and their nutritional aspects; and, finally, spirituality that provides emotional comfort. In a qualitative study based on interviews, Woods36 affirmed that satisfactory affective relationships based on understanding and acceptance by the other are not just supportive, ie, they do not simply serve as support in a helpful manner but, rather, they serve as true active interventions.

Among factors that limit the process of change are fear of change, absence of social support, and lack of motivation. According to a qualitative study by KeskiRahkonen and Tozzi,21 the value of professional help is conditioned by the patient’s desire to change. If the patient is not willing to change, treatment will not be successful.

Another factor raised by patients that may limit the process of change is ambiguity of feelings, rigidity, and lack of sympathy on the part of health professionals. Gabbard48 argued that establishing a bond with a patient who has anorexia nervosa is a slow and difficult task, in that the therapist/physician may be considered the enemy. In psychotherapy, the therapist must synchronize with the emotional experience of the patient and give meaning to that patient’s experience. This notion was supported by Dare et al5 in a study on the effectiveness of psychological therapies, in which a friendly therapist-client relationship appears as the most powerful element in promoting change.

Regarding the perception of treatment modalities, it should first be emphasized that treatment of anorexia nervosa must be understood in a broad sense, considering the subject as a whole, thus including biological, psychological, and social dimensions. However, although each of these interventions has its role and degree of importance according to the comments analyzed, patients have a preference for psychosocial approaches. Pharmacological approaches can be seen as directed toward treating anorexia nervosa and not the patient, ignoring that patient’s intense emotional needs.

Third-Order Interpretation: Reconciliation

Third-order interpretation revealed a meta-category involving a subject’s self-reconciliation. One of the central questions in this disorder is a conflict or discomfort with the body itself. This is experienced with intense suffering in that it involves the subject’s self-image. The anorexic individual has to experience a process of being accepted by self and others. This broader acceptance—inner and outer—is what facilitates recovery, according to patients. In a study of anorexia nervosa and quality of life, De la Rie et al44 indicated that 93% of the patients interviewed felt that caring and belonging are a central aspect of quality of life.

On the basis of clinical cases on the role of acceptance in this clinical picture, Hycner and Jacob49 consider a primal factor in recovery to be what they call “confirming the other.” Confirming means a conscious effort to see the other as a unique, singular being who has a right to his or her own individuality, respecting it and accepting it in one’s own capacity to make choices and be responsible for them.

This synthesis of qualitative research allows the integration of findings from small studies to be used as a tool to better understand how individuals with anorexia nervosa view treatment. Meta-ethnography can be used to synthesize studies with various qualitative methods—not limited to ethnographic research—to explore a wide range of experience while simultaneously increasing the size and diversity of the total sample.

Clinical consensus has demonstrated that health professionals should not be limited to weight gain as an exclusive objective in the treatment of patients with anorexia nervosa. Difficulties in the treatment field are enhanced by the frequent association of anorexia nervosa with multiple comorbid factors, personality disorders, family organization, genetics, temperament, body image, and mechanisms of defense.50-59 All of these combined factors may be related to the heterogeneity of patients with anorexia nervosa and could have important psychotherapeutic and prognostic value. Our analysis showed that, of the studies we identified, patients had pointed out important treatment experiences that could help the management of patients who have anorexia nervosa.

Overall, it can be observed that the studies examined in this review support the evidence presented by the works of Bruch, Boris, Skårderud, Farber et al, Kleinfeld et al, and Bell.13-18 Hence, we postulate that the difficulty in treating patients with anorexia nervosa relies on the complex psychological mechanisms involved—mainly, the concept of identity, which is so impaired in these cases. This concept can partly explain the lack of therapeutic success frequently reported in experimental studies and systematic meta-analyses. Furthermore, even when a patient recognizes the existence of the eating disorder, that does not eliminate the simultaneous presence of a group of other psychological components that cause patients to minimize the nature of the problem and the risk involved.

These findings can be useful to health professionals who are involved in the treatment of anorexia nervosa. The treatment of anorexia nervosa that exclusively considers the notion of weight can be frustrating for the patient. The exclusion of the psychological elements discussed previously can limit the relationship between the health care professional and the patient. The consequence is obvious and can jeopardize not only the enrollment in any kind of therapeutic process (independent of its therapeutic approach), but can also influence the patient’s involvement with the therapeutic processes. False assumptions about anorexia nervosa—both from patients and doctors—can be at the core of many unsuccessful therapeutic cases. The biggest contribution of the anorexic patient to the investigation of therapeutic processes might be in emphasizing the complex nature of this disorder and the need to account for this complexity in the process of intervention.

This synthesis of qualitative research shows that several main themes were described as important in the treatment process: phases of recovery, factors favoring recovery, factors limiting recovery, perception of treatment modalities, and self-reconciliation. The themes identified in this synthesis confirm other research in anorexia nervosa,59,60 but our synthesis expands on previous work by showing the large spectrum of alternatives faced by patients with anorexia nervosa when attempting to deal with and treat their symptoms. For the health care practitioner, the diagnosis of anorexia nervosa is challenging, as the likelihood of successfully treating such patients is low. However, an awareness of the themes indicated by patients is vital, as taking into account some of these issues may help some patients. Anorexia nervosa is a heterogeneous entity manifested by symptoms leading to clinically significant impairment or distress. There are indications that the outcome may be affected by the following factors: religion and spirituality,61-63 internal motivation to change, recovery as a work in progress, the perceived value of the treatment experience, developing supportive relationships, awareness and tolerance of negative emotion and self-validation,60 awareness of the process of change,59 relevant turning points,19,64 fear of change,65 the therapeutic relationship,66,67 and information.68 Another interesting factor affecting outcome is media information, such as television. For Williams et al,46 the media has two main functions: (1) maintaining the disorder, and (2) helping to maintain anorexia and dictating rigid aesthetic standards.

This study presents some positive aspects that should be emphasized. As far as we know, this is the first metasynthesis study of anorexia nervosa from the patient’s perspective. Second, the results point toward a complex phenomenon in which the question of identity emerges, along with its connection to outcome.

However, the results of this investigation should be viewed in the light of some limitations. First, most of the results come from studies classified as category B according to the criteria for evaluating qualitative methodology, leading to cautious interpretation. Second, almost all of the studies consider the experience of women who have anorexia nervosa. Third, scientific production is concentrated in a few developed countries, which can lead to cultural bias. Fourth, we relied on self-report. However, self-report has been found to be a valid measure in anorexia nervosa and other eating problems.69-71

Further investigations with qualitative methodology are needed. Increased scientific production involving more methodological rigor is needed, as are studies in developing countries and those that include the male population.


This study was supported by FAPESP, Fundação de Amparo à Pesquisa do Estado de São Paulo (The State of São Paulo Research Foundation), Grant 07/50739-1.


The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.


  1. Steinhausen HC. The outcome of anorexia nervosa in the 20th century. Am J Psychiatry. 2002;159:1284–1293.
  2. Wilson GT, Shafran R. Eating disorders guidelines from NICE. Lancet. 2005;365:79–81.
  3. Claudino AM, Hay P, Lima MS, et al. Antidepressants for anorexia nervosa. Cochrane Database Syst Rev. 2006;(1):CD004365.
  4. Bulik CM, Berkman ND, Brownley KA, et al. Anorexia nervosa treatment: a systematic review of randomized controlled trials. Int J Eat Disord. 2007;40(4):310–320.
  5. Dare C, Eisler I, Russell G, et al. Psychological therapies for adults with anorexia nervosa. Br J Psychiatry. 2001;178:216–221.
  6. Keel PK, Dorer DJ, Franko DL, et al. Postremission predictors of relapse in women with eating disorders. Am J Psychiatry. 2005;62:2263–2268.
  7. Treasure JL, Ward A. A practical guide to the use of motivational interviewing in anorexia nervosa. Eur Eat Disord Rev. 1997;5:102–114.
  8. Geller J, Cockell SJ, Zaitsoff S, et al. Predicting behaviour change in anorexia nervosa: A comparison of readiness assessment strategies. Paper presented at: Meeting of the Eating Disorders Research Society; 1999; San Diego, CA.
  9. Miller WR, Rollnick S. Motivational Interviewing: Preparing People for Change. New York, NY: Guilford Press; 1999.
  10. Geller J, Drab D. The Readiness and Motivation Interview: A symptom-specific measure of readiness for change in the eating disorders. European Eat Disord Rev. 1999;7:259–278.
  11. Pike KM, Walsh BT, Vitousek KGT, et al. Cognitive behavior therapy in the posthospitalization treatment of anorexia nervosa. Am J Psychiatry. 2003;160:2046–2049.
  12. Hay P, Bacaltchuk J, Claudino A, et al. Individual psychotherapy in the outpatient treatment of adults with anorexia nervosa. Cochrane Database Syst Rev. 2003;(4):CD003909.
  13. Bell L. What can we learn from consumer studies and qualitative research in the treatment of eating disorders? Eat Weight Disord. 2003;8:181–187.
  14. Bruch H. The changing picture of an illness: anorexia nervosa. In: Attachment and the Therapeutic Process. Madison, CT: International universities Press; 1987:20-30.
  15. Boris HN. The problem of anorexia nervosa. Int J Psychoanal. 1984;65:315–322.
  16. Skårderud F. Eating one’s words. Part I: “Concretised metaphors” and reflective function in anorexia nervosa—an interview study. Eur Eat Disord Rev. 2007;15:163–174.
  17. Farber SK, Jackson CC, Tabin JK, et al. Death and annihilation anxieties in anorexia nervosa, bulimia, and self-mutilation. Psychoanal Psychol. 2007;24:289–305.
  18. Kleinfield EI, Wagner S, Halmi KA. Cognitive-behavioral treatment of anorexia nervosa. Psychiatr Clin North Am. 1996;19:715–737.
  19. *Tozzi F, Sullivan PF, Fear JL, et al. Causes and recovery in anorexia nervosa: the patient’s perspective. Int J Eat Disord. 2003;33:143–154.
  20. *Redenbach J, Lawler J. Recovery from disordered eating: what life histories reveal. Contemp Nurse. 2003;15:148–156.
  21. *Keski-Rahkonen A, Tozzi F. The process of recovery in eating disorder sufferers’ own words: an Internet-based study. Int J Eat Disord. 2005;37(suppl):S80–S86.
  22. Vitousek KM, Gray JA. Psychotherapy of eating disorders. In: Gabbard G, Beck J, Holmes JA, eds. Oxford Textbook of Psychotherapy. Oxford, uK: Oxford university Press; 2005:177-200.
  23. Fairburn CG, Cooper Z, Shafran R. Cognitive behavior therapy for eating disorders: a “transdiagnostic” theory and treatment. Behav Res Ther. 2003;41:509–528.
  24. Jensen LA, Allen MN. Meta-synthesis of qualitative findings. Qual Health Res. 1996;6:553–560.
  25. Dixon-Woods M, Fitzpatrick R, Roberts K. Including qualitative research in systematic reviews: opportunities and problems. J Eval Clin Pract. 2001;7:125–133.
  26. Paterson BL, Thorne SE, Canam C, et al. Meta-Study of Qualitative Health Research: A Practical Guide to Meta-Analysis and Metasynthesis. Thousand Oaks, CA: Sage Publications; 2001.
  27.  American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 4th ed, text revision. Washington, DC: American Psychiatric Association; 2000.
  28.  CASP. Critical Appraisal Skills Programme (CASP)—making sense of evidence. London, UK: Oxford; 2002.
  29. Britten N, Campbell R, Pope C, et al. using meta ethnography to synthesise qualitative research: a worked example. J Health Serv Res Policy. 2002;7:209–215.
  30. Nelson A. A metasynthesis: Mothering other-than-normal children. Qual Health Res. 2002;12:515–530.
  31. Campbell R, Pound P, Pope C, et al. Evaluating meta-ethnography: a synthesis of qualitative research on lay experiences of diabetes and diabetes care. Soc Sci Med. 2003;56:671–684.
  32. Sandelowski M, Barroso J. Towards a metasynthesis of qualitative findings on motherhood in HIV positive women. Res Nurs Health. 2003;26:153–170.
  33. Noblit GW, Hare RD. Meta-ethnography: Synthesizing Qualitative Studies. Newbury Park, CA: Sage Publications; 1988.
  34. *Colton A, Pistrang N. Adolescents’ experiences of inpatient treatment for anorexia nervosa. Eur Eat Disord Rev. 2004;12:307–316.
  35. *D’Abundo M, Chally P. Struggling with recovery: participant perspectives on battling an eating disorder. Qual Health Res. 2004;14:1094–1106.
  36. *Woods S. untreated recovery from eating disorders. Adolescence. 2004;39:361–371.
  37. *Tan J, Hope RA, Stewart A, et al. Control and compulsory treatment in anorexia nervosa: the views of patients and parents. Int J Law Psychiat. 2003;26:627–645.
  38. *Weaver K, Wuest J, Ciliska D. understanding women’s journey of recovering from anorexia nervosa. Qual Health Res. 2005;15:188–206.
  39. *Lamoureux M, Bottorff J. Becoming the real me: recovering from anorexia nervosa. Health Care Women Int. 2005;26:170–188.
  40. *Cockell SJ, Zaitsoff SL, Geller J. Maintaining change following eating disorder treatment. Prof Psychol Res Pr. 2004;35:527–534.
  41. *Hsu LKG, Crisp AH, Callender JS. Recovery in anorexia nervosa—the patient’s perspective. Int J Eat Disord. 1992;11:341–350.
  42. *Chan Z, Ma J. Anorexic body: A qualitative study. Forum Qualitative Sozialforschung/Forum Qualitative Social Research. 2003;4(1):article 1.
  43. *Surgenor LJ, Plumridge EW, Horn J. Knowing one’s self “anorexic”: implications for therapeutic practice. Int J Eat Disord. 2003;33:22–32.
  44. *De la Rie S, Noordenbos G, Donker M, et al. The patient’s view on quality of life and eating disorders. Int J Eat Disord. 2007;40:13–20.
  45. *Tan J, Hope RA, Stewart A. Anorexia nervosa and personal identity: the accounts of patients and their parents. Int J Law Psychiat. 2003;26:533–548.
  46. *Williams MS, Thomsen SR, McCoy JK. Looking for an accurate mirror: a model for the relationship between media use and anorexia. Eat Behav. 2003;4:127–134.
  47. Buckroyd J. Anorexia e bulimia: Esclarecendo suas dúvidas. São Paulo, Brazil: Agora; 2000.
  48. Gabbard G. Psiquiatria psicodinâmica. Porto Alegre, Brazil: Artmed;1998.
  49. Hycner R, Jacob L. Relação e cura em Gestalt-terapia/Tradução de Elisa. Plass e Maria Portella. São Paulo, Brazil: Summus; 1997.
  50. Garfinkel PE, Garner DM. Anorexia Nervosa: A Multidimensional Perspective. New York, NY: BrunnerMazel; 1982.
  51. Garner DM, Garner MV, Rosen LW. Anorexia nervosa “restrictors” who purge: Implications for subtyping anorexia nervosa. Int J Eat Disord. 1993;13:171–185.
  52. Thompson-Brenner H, Eddy KT, Satir AD, et al. Personality subtypes in adolescents with eating disorders: validation of a classification approach. J Child Psychol Psychiatry. 2008;49(2):170–180.
  53. Casper RC, Troiani M. Family functioning in anorexia nervosa differs by subtype. Int J Eat Disord. 2001;30(3):338–342.
  54. Wade TD, Bulik CM, Neale M, et al. Anorexia nervosa and major depression: shared genetic and environmental risk factors. Am J Psychiatry. 2000;157(3):469–471.
  55. Díaz-Marsá M, Carrasco JL, Sáiz J. A study of temperament and personality in anorexia and bulimia nervosa. J Personal Disord. 2000;14(4):352–359.
  56. Santos PCM, Júnior Ny, Pessa RP, et al. Anorexia nervosa e bulimia: aspectos psicopatológicos, demográficos, diagnósticos e clínicos. Rev Bras Psiquiatr. 1988;10:34–41.
  57. Gothelf D, Apter A, Ratzoni G, et al. Defense mechanisms in severe adolescent anorexia nervosa. J Am Acad Child Adolesc Psychiatry. 1995;34(12):1648–1654.
  58. Jiménez-Murcia S, Fernández-Aranda F, Raich RM, et al. Obsessive-compulsive and eating disorders: comparison of clinical and personality features. Psychiatry Clin Neurosci. 2007;61(4):385–391.
  59. Prochaska JO, DiClemente CC. The transtheoretical model of change. In: Norcross JC, Goldfried MR, eds. Handbook of Psychotherapy Integration. New York, NY: Basic Books; 1992:300-334.
  60. Federici A, Kaplan AS. The patient’s account of relapse and recovery in anorexia nervosa: a qualitative study. Eur Eat Disord Rev. 2008;16(1):1–10.
  61. Koenig HG, McCullough M, Larson DB. Handbook of Religion and Health: A Century of Research Reviewed. New York, NY: Oxford university Press; 2001.
  62. Sloan RP, Bagiella E, Powell T. Religion, spirituality, and medicine. Lancet. 1999;353:664–667.
  63. Powell LH, Shahabi L, Thoresen CE. Religion and spirituality. Linkages to physical health. Am Psychol. 2003;58(1):36–52.
  64. Nilsson K, Hägglöf B. Patient perspectives of recovery in adolescent onset anorexia nervosa. Eat Disord. 2006;14(4):305–311.
  65. Rosenvinge JH, Klusmeier AK. Treatment for eating disorders from a patient satisfaction perspective: a Norwegian replication of a British study. Eur Eat Disorders Rev. 2000;8:459–66.
  66. Button E, Warren R. Living with anorexia nervosa: the experience of a cohort of suffers from anorexia nervosa 7.5 years after initial presentation to a specialized eating disorders service. Eur Eat Disord Rev. 2001;9(2):74–96.
  67. Pettersen G, Rosenvinge JH. Improvement and recovery from eating disorders: a patient perspective. Eat Disord. 2002;10:61–71.
  68. Diener E, Fujita J. Resources, personal strivings and subjective well-being: a nomothetic and idiographic approach. J Pers Soc Psychol. 2005;51:1058–1068.
  69. Fairburn CG, Beglin S. Assessment of eating disorder pathology: interview or self-report questionnaire? Int J Eat Disord. 1994;16:363–370.
  70. Fichter MM, Quadflieg N. Comparing self- and expert rating: a self-report screening version (SIAB-S) of the Structured Interview for Anorexic and Bulimic Syndromes for DSM-IV and ICD-10 (SIAB-Ex). Eur Arch Psychiatry Clin Neurosci. 2000;250(4):175–185.
  71. Wolk SL, Loeb KL, Walsh BT. Assessment of patients with anorexia nervosa: interview versus self-report. Int J Eat Disord. 2005;37(2):92–99.

CORRESPONDENCE: Sergio L. Blay, MD, PhD Department of Psychiatry Escola Paulista de Medicina-UNIFESP R. Botucatu, 740 CEP 04023-900 São Paulo SP Brazil. E-MAIL: