Elsevier

Psychiatry Research

Volume 200, Issues 2–3, 30 December 2012, Pages 518-523
Psychiatry Research

A prospective study of predictors of relapse in anorexia nervosa: Implications for relapse prevention

https://doi.org/10.1016/j.psychres.2012.04.037Get rights and content

Abstract

Anorexia nervosa (AN) is a serious psychiatric disorder with a high rate of relapse. The goal of this study was to identify predictors of relapse in adult AN using a prospective, longitudinal design. Participants were 100 AN patients who had successfully completed specialized inpatient/day treatment, were weight-restored to a body mass index (BMI) of at least 20 for a minimum of 2 weeks, and reported less than one binge-purge (BP) episode over the previous 28 days at the end of treatment. Predictor variables included baseline demographic and clinical features, behavioral and psychological changes during treatment, residual psychopathology at post-treatment, and motivation to recover. Results indicated that 41% of participants relapsed during the 1-year follow-up period. The highest risk period for relapse was between 4 and 9 months post-treatment. Predictors of relapse included: the BP subtype of AN, severity of checking behaviors at pre-treatment, decrease in motivation to recover during treatment, and lower motivation to recover at post-treatment. These findings suggest that individuals with the BP subtype of AN are particularly susceptible to relapse. Increasing and maintaining motivation to recover during acute treatment may have an important impact on long-term outcome.

Introduction

Anorexia nervosa (AN) is a serious psychiatric disorder characterized by extreme food restriction, maintenance of an abnormally low body weight, intense fear of weight gain, and body image disturbance (American Psychiatric Association, 1994). In about half of cases, there is also recurrent binge eating and/or purging behavior. The disorder is associated with significant psychiatric comorbidity, serious medical complications, and major impairments in psychosocial functioning. The cumulative mortality rate is approximately 5.6% per decade, with about half of deaths due to cardiac failure and half due to suicide (Sullivan, 1995, Herzog et al., 2000). Nutritional rehabilitation and weight restoration, often involving a period of inpatient or day hospital treatment, usually constitute the first step in the recovery process in severe cases. Cohort studies suggest that most adult patients respond well to hospital-based treatment programs for AN in the short-term (e.g., Olmsted et al., 2010) but that the rate of relapse is high (Carter et al., 2004). Studies have reported a wide range of estimates of the rate of relapse in AN depending upon the definitions of relapse used, the length of follow-up, and the methodologies employed. Reported rates of relapse have ranged from 9% to 65% (e.g., Strober et al., 1997, Carter et al., 2004, Keel et al., 2005, Walsh et al., 2006), with the relapse rates tending to be lower in adolescent samples than among adults. Studies that have examined the timing of relapse have shown that the risk of relapse is highest during the first 4–12 months post-treatment (e.g., Carter et al., 2004, McFarlane et al., 2008).

Identifying variables predictive of relapse in AN that are amenable to change may suggest important targets for both acute and relapse prevention interventions. It may also advance understanding of the psychopathology of AN by shedding light on key maintaining factors. Several studies have found that individuals who present with more severe or chronic eating disorders are more likely to relapse including older age (Deter and Herzog, 1994); a higher number of previous hospitalizations (Treat et al., 2008); higher pre-treatment purge frequency (Deter and Herzog, 1994); and greater dietary restriction (McFarlane et al., 2008). However, baseline variables are not really modifiable, but may identify high risk groups. Several studies have shown that patients who make slower progress during acute treatment are more prone to relapse including a lower rate of weight gain (Castro et al., 2004, Treat et al., 2008) and slower adherence to the prescribed meal plan during day hospital treatment (McFarlane et al., 2008). These variables may reflect lower motivation or greater resistance to change. Other research has identified certain areas of vulnerability following acute treatment that predict subsequent relapse including higher residual weight concern (Carter et al., 2004); greater body image disturbance (Keel et al., 2005); and higher residual weight-based self esteem (McFarlane et al., 2008). Relapse has also been shown to be associated with lower body mass index (BMI) (Kaplan et al., 2009) and lower percent body fat (Mayer et al., 2007) at post-treatment. One study found that patients who reported lower levels of motivation to recover at the end of acute treatment were more prone to relapse (Castro-Fornieles et al., 2007). The aim of the present study was to identify predictors of relapse in AN that might inform the development of more effective treatment and relapse prevention interventions.

Section snippets

Participants

The participants were 100 consecutive patients who met Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) criteria for AN based on Eating Disorder Examination (EDE; Fairburn and Cooper, 1993) interview. All were admitted to the hybrid inpatient/day treatment unit of the Eating Disorders Program at a large general hospital between 2000 and 2008. This program is a specialized hospital-based program operated by a multi-disciplinary team that focuses on nutritional

Attrition and missing data imputation

Maximum likelihood imputation based on EM algorithm was used to impute missing discharge data for those cases whose discharge assessments had missing data. The frequency (percent) of missing data for the discharge variables ranged from 0 (0%) to 18 (18%) cases out of 100.

Rate and timing of relapse

Fig. 1 presents the survival curve and hazard function for the 100 participants. The highest risk of relapse was between 4 and 9 months post-discharge. After that, the survival curve began to level off suggesting less risk of

Discussion

Comparable to previous studies, we found that, at 12 months follow-up, 41% of patients with AN had relapsed and that the highest risk for relapse occurred between 4 and 9 months post-treatment. Among patients who did not relapse, 25% of them maintained a BMI above 19.5 for the entire 1 year follow-up period, and most of these had the restricting subtype of the disorder. The remainder dropped below a BMI of 19.5, and 40% maintained a BMI between 17.5 and 19.5. This indicates that a significant

Acknowledgments

This research was supported by the Canadian Institute for Health Research (grant MOP-86531). The authors are grateful to Dr. Allison Kelly for her statistical advice and her comments on the manuscript.

Declaration of Interest

None.

References (41)

  • J.C. Carter et al.

    The impact of childhood sexual abuse in anorexia nervosa

    Child Abuse & Neglect

    (2006)
  • P. van Oppen et al.

    The structure of obsessive–compulsive symptoms

    Behaviour Research and Therapy

    (1995)
  • American Psychiatric Association

    Diagnostic and statistical manual of mental disorders

    (1994)
  • A.T. Beck et al.

    Manual for the Beck Depression Inventory-II

    (1996)
  • Bewell-Weiss, C.V., Mills, J.S., Westra, H., Carter, J.C., Motivational interviewing as a prelude to intensive eating...
  • C.V. Bewell et al.

    Readiness to change mediates the impact of eating disorder symptomatology on treatment outcome in anorexia nervosa

    International Journal of Eating Disorders

    (2008)
  • J.C. Carter et al.

    Relapse in anorexia nervosa: a survival analysis

    Psychological Medicine: A Journal of Research in Psychiatry and the Allied Sciences

    (2004)
  • J. Castro-Fornieles et al.

    Predictors of weight maintenance after hospital discharge in adolescent anorexia nervosa

    International Journal of Eating Disorders

    (2007)
  • J. Castro et al.

    Predictors of rehospitalization after total weight recovery in adolescents with anorexia nervosa

    International Journal of Eating Disorders

    (2004)
  • J. Cohen

    Statistical Power Analysis for the Behavioral Sciences

    (1988)
  • Z. Cooper et al.

    The validity of the eating disorder examination and its subscales

    British Journal of Psychiatry

    (1989)
  • Z. Cooper et al.

    The eating disorder examination: a semi-structured interview for the assessment of the specific psychopathology of eating disorders

    International Journal of Eating Disorders

    (1987)
  • D.R. Cox et al.

    Analysis of Survival Data

    (1984)
  • L.R. Derogatis et al.

    The brief symptom inventory: an introductory report

    Psychological Medicine: A Journal of Research in Psychiatry and the Allied Sciences

    (1983)
  • H.-C. Deter et al.

    Anorexia nervosa in a long-term perspective: results of the Heidelberg–Mannheim study

    Psychosomatic Medicine

    (1994)
  • C.G. Fairburn et al.

    Assessment of eating disorders: interview or self-report questionnaire?

    International Journal of Eating Disorders

    (1994)
  • C.G. Fairburn et al.
  • A. Favaro et al.

    Purging behaviors, suicide attempts, and psychiatric symptoms in 398 eating disordered subjects

    International Journal of Eating Disorders

    (1996)
  • M.B. First et al.

    Structured Clinical Interview For DSM-IV Axis I Disorders

    (1996)
  • J. Geller et al.

    Readiness to change dietary restriction predicts outcomes in the eating disorders

    Eating Disorders: The Journal of Treatment & Prevention

    (2004)
  • Cited by (130)

    • An integrated review of fear and avoidance learning in anxiety disorders and application to eating disorders

      2022, New Ideas in Psychology
      Citation Excerpt :

      Specifically, even in rigorous trials with expert providers, CBT-E treatments are only effective for approximately one third to one half of individuals (Keel & Mitchell, 1997). Additionally, approximately one-third to one-half of people who do improve from CBT-E treatment will experience at least one ED relapse after discharge (Carter et al., 2012; Keel et al., 2005). These rates of remission and relapse may be even worse when accounting for the lack of specialized ED providers and training and use of non-evidence based treatments in clinical practice (Cooper & Bailey-Straebler, 2015; Haas & Clopton, 2003; Von Ranson et al., 2013).

    • The transdiagnostic role of food craving: An application of substance use models

      2022, Appetite
      Citation Excerpt :

      Eating disorders (EDs) encompass a diverse range of behaviors, including restrictive eating, overeating, and compensatory behaviors (e.g., vomiting; Larson, Neumark-Sztainer, & Story, 2009; Neumark-Sztainer, Wall, Larson, Eisenberg, & Loth, 2011). The heterogeneity of these behaviors, paired with weak treatment outcomes and high relapse rates (Carter et al., 2012; McFarlane, Olmsted, & Trottier, 2008), underscore the fundamental question of whether current eating disorder diagnoses adequately represent the underlying processes that contribute to the development and maintenance of disordered eating behavior patterns. Given the numerous constellations of presenting behaviors and the behavioral and diagnostic crossover commonly observed during treatment (e.g., bulimia nervosa transitions to binge eating disorder: Castellini et al., 2013; Eddy et al., 2008), it is not surprising that EDs are associated with heightened economic burden (Stuhldreher et al., 2012; Weissman & Rosselli, 2017).

    View all citing articles on Scopus
    View full text