CoChat
Sign Up
Cited by 111Eating Disorders and BehaviorsNutrition, Health and Food BehaviorChild Nutrition and Feeding Issues

Factors associated with recovery from anorexia nervosa

Open access
Zerwas, Stephanie, Lund, Brian C., Von Holle, Ann, Thornton, Laura M., Berrettini, Wade H., Brandt, Harry, Crawford, Steven, Fichter, Manfred M. +10 more·2013·Journal of Psychiatric Research
DOI 10.1016/j.jpsychires.2013.02.011PMC PMC3682792
Open this paper in CoChat — free account

Abstract

Previous studies of prognostic factors of anorexia nervosa (AN) course and recovery have followed clinical populations after treatment discharge. This retrospective study examined the association between prognostic factors--eating disorder features, personality traits, and psychiatric comorbidity--and likelihood of recovery in a large sample of women with AN participating in a multi-site genetic study. The study included 680 women with AN. Recovery was defined as the offset of AN symptoms if the participant experienced at least one year without any eating disorder symptoms of low weight, dieting, binge eating, and inappropriate compensatory behaviors. Participants completed a structured interview about eating disorders features, psychiatric comorbidity, and self-report measures of personality. Survival analysis was applied to model time to recovery from AN. Cox regression models were used to fit associations between predictors and the probability of recovery. In the final model, likelihood of recovery was significantly predicted by the following prognostic factors: vomiting, impulsivity, and trait anxiety. Self-induced vomiting and greater trait anxiety were negative prognostic factors and predicted lower likelihood of recovery. Greater impulsivity was a positive prognostic factor and predicted greater likelihood of recovery. There was a significant interaction between impulsivity and time; the association between impulsivity and likelihood of recovery decreased as duration of AN increased. The anxiolytic function of some AN behaviors may impede recovery for individuals with greater trait anxiety.

Integrity signals

Registry

Crossref · 10.1016/j.jpsychires.2013.02.011

Cites

62 of 62 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 36 cited sources checked · 0 supported · 0 flagged

  • Findingown data

    In Model 4, greater impulsivity at onset was associated with a higher probability of recovery (hazard ratio = 1.07; 7% higher hazard).

    “Participants had a significantly higher probability of recovery when they also endorsed more impulsivity at the onset of eating disorder (hazard ratio = 1.07; 7% higher hazard)”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Temperament and personality traits are stable and not substantially affected by the emotional and cognitive stress of extreme dietary restriction and malnutrition.
  • Findingown data

    In the analysis sample of 680 women with anorexia nervosa, 123 participants (18.1%) met the study's recovery criterion of at least 12 months without eating disorder symptoms.

    “Of the 680 women with AN, 123 (18.1%) met the study criterion for recovery, defined as 12 months symptom free.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Participants did not have a period of recovery (12 months without symptoms of anorexia nervosa) and relapse back to anorexia nervosa prior to the recovery event captured by the measures.
  • Findingown data

    In Model 1, endorsing vomiting was associated with a lower probability of recovery from anorexia nervosa (hazard ratio = 0.37; 63% lower hazard).

    “In Model 1, participants had a lower probability of recovery (hazard ratio = 0.37; 63% lower hazard) if they endorsed vomiting.”
    Quote in the paperSubstantial claim
  • Findingown data

    In Model 2, greater impulsivity at anorexia nervosa onset was associated with a higher probability of recovery (hazard ratio = 1.05; 5% higher hazard).

    “At eating disorder onset, participants had a significantly higher probability of recovery when they endorsed more impulsivity (hazard ratio = 1.05; 5% higher hazard).”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Temperament and personality traits are stable and not substantially affected by the emotional and cognitive stress of extreme dietary restriction and malnutrition.
  • Findingown data

    In Model 1, no eating disorder features other than vomiting were significant prognostic factors for recovery.

    “There were no other significant prognostic factors.”
    Quote in the paperNot a substantial claim
  • May combine two claims. 12
  • Classified as probably not a substantial claim. 21

  • May combine two claims.

    Findingown data

    In the discussion, age of onset, laxative abuse, fasting, excessive exercise, novelty seeking, harm avoidance, reward dependence, persistence, concern over mistakes, and comorbid OCD, MDD, alcohol and substance abuse or dependence, PTSD, and borderline personality disorder were not significant predictors of recovery.

    “Finally, age of onset, laxative abuse, fasting, excessive exercise, novelty seeking, harm avoidance, reward dependence, persistence, concern over mistakes, and the presence of comorbid OCD, MDD, alcohol and substance abuse or dependence, PTSD, and borderline personality disorder were not significant predictors.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Participants were able to report the temporality of their psychiatric diagnoses accurately.

  • May combine two claims.

    Findingcitation

    The percentage of participants in recovery from anorexia nervosa in the current study was 18.4%, which was lower than estimates of 20-50% in other studies.

    “Similarly the percentage of participants in recovery from AN in the current study was unusually low (18.4%) compared with estimates of 20–50% in other studies (Herzog et al., 1999b, Strober et al., 1997, Clausen, 2004).”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites Herzog et al., 1999b

      Not checked in the free run.

    • cites Strober et al., 1997 source

      Not checked in the free run.

    • cites Clausen, 2004 source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Among participants classified as recovered, the mean time since the last eating disorder symptom was 5.7 years (SD = 5.6).

    “Of the participants who were recovered, the mean number of years since the last eating disorder symptom was 5.7 (SD = 5.6).”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductnone given

    Three models were pre-specified, with covariates grouped by domain: eating disorder features, personality traits, and psychiatric comorbidity.

    “Three models were pre-specified, with covariates grouped by domain: eating disorder features (Model 1), personality traits (Model 2), and psychiatric comorbidity (Model 3).”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Study design and conductcitation

    All p-values in the final models 1-4 were adjusted by model according to the Benjamini-Hochberg method of false discovery rate control.

    “All p-values in the final models 1–4 were adjusted by model according to the Benjamini-Hochberg method of false discovery rate (FDR) control.(Benjamini and Hochberg, 1995)”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Benjamini and Hochberg, 1995 source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    Only the presence of compensatory behaviors was recorded, not their duration; therefore, duration of these behaviors could not be included as prognostic factors.

    “Only the presence of these behaviors was recorded not duration; thus, we were not able to include the duration of these compensatory behaviors as prognostic factors.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Interpretationreasoning

    Increased severity from combined psychological disturbance could interfere with treatment and decrease the likelihood of recovery.

    “Increased severity from combined psychological disturbance could interfere with treatment and decrease the likelihood of recovery.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S1 (asserted).

    Interpretationreasoning

    The contradictory findings across studies on impulsivity may be due to differences in how impulsivity was assessed.

    “We hypothesize that these contradictory findings across studies may be due to differences in impulsivity assessment.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationcitation

    The overall time to recovery in this sample was 8-9 years, which was longer than the median 6.5 years for full recovery reported in previous research.

    “However, the overall time to recovery was somewhat longer in this sample (8–9 years) compared with previous research, which reported the median length of illness to be 6.5 (79 months) years for full recovery (Strober et al., 1997).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Strober et al., 1997 source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Interpretationreasoning

    Impulsivity in the present study was measured with the Barratt Impulsivity Scale, which may capture different psychological constructs than Fichter et al.'s measure.

    “Impulsivity in the present study was measured with the BIS, which may capture different psychological constructs.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    The Barratt Impulsivity Scale has been found to assess the degree of impulse control in women with eating disorders.

    “The BIS measures cognitive, motor, and non-planning impulsiveness and has been found to assess the degree of impulse control in women with eating disorders (Bulik et al., 1997b).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Bulik et al., 1997b

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    The definition of fasting as 24 hours without eating may have led to a more conservative interpretation than a less stringent criterion such as missing two consecutive meals or eight waking hours without eating.

    “Also, the definition of fasting as 24 hours without eating may have led to a more conservative interpretation than a less stringent criterion such as missing two consecutive meals or eight waking hours without eating.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Study design and conductcitation

    Interviewers at each site completed a training program, and all eating disorder diagnoses were confirmed by the principal investigator at each site after reviewing the SIAB; the project coordinator of the data core independently reviewed every participant's SIAB interview to confirm diagnoses.

    “Briefly, interviewers at each site completed a training program and all eating disorder diagnoses were confirmed by the principal investigator at each site after reviewing the SIAB. Second, the project coordinator of the data core reviewed every participant’s SIAB interview independently to confirm diagnoses (Kaye et al., 2000).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Kaye et al., 2000 source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Premisecitation

    No consensus definition of anorexia nervosa recovery exists.

    “Because no consensus definition of AN recovery exists (Bardone-Cone et al., 2010, Hoek, 2006, Pike, 1998)”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites Bardone-Cone et al., 2010 source

      Not checked in the free run.

    • cites Hoek, 2006 source

      Not checked in the free run.

    • cites Pike, 1998 source

      Not checked in the free run.

  • Classified as probably not a substantial claim. May combine two claims.

    Interpretationreasoning

    Impulsivity as measured by the Barratt Impulsivity Scale may be associated with recovery from anorexia nervosa because it tempers the rigidity and intractability often associated with the disorder and could encourage experimenting with healthier eating behaviors.

    “Impulsivity as measured by the BIS may be associated with recovery from AN because it tempers the rigidity and intractability often associated with AN and could encourage experimenting with healthier eating behaviors”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Core features of anorexia nervosa-food restriction and exercise-are hypothesized to have an anxiolytic effect.

    “In addition, core features of AN—food restriction and exercise—are hypothesized to have an anxiolytic effect (Dellava et al., 2010, Kaye, 2008)”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Dellava et al., 2010 source

      Not checked in the free run.

    • cites Kaye, 2008 source

      Not checked in the free run.

  • Classified as probably not a substantial claim. May combine two claims.

    Study design and conductreasoning

    The prognostic factors examined were limited to fewer than ten covariates per model and to conditions previously associated with anorexia nervosa recovery, in order to reduce overfitting.

    “We limited the prognostic factors examined in these analyses to groups less than ten covariates per model and to conditions that had been found to be associated with AN recovery in previous studies in an effort to reduce overfitting.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductreasoning

    Because recruitment source and treatment seeking were not directly measured, the study was unable to assess treatment heterogeneity and examine treatment experience as a prognostic factor.

    “However, because some of the participants in the present study were also recruited from specialty centers yet recruitment source and treatment seeking were not directly measured, we were unable to assess treatment heterogeneity and examine treatment experience as a prognostic factor.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductreasoning

    Axis I disorders were coded as present only if their onset was before the participant's last anorexia nervosa symptom, to avoid estimating associations between post-recovery Axis I disorders and recovery likelihood.

    “In all cases, Axis I disorders were only coded as present if onset was before the participant experienced her last AN symptom to prohibit estimation of associations between Axis I disorders occurring after recovery and the likelihood of recovery.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Prior research reported a linear relationship between trait anxiety scale responses and remission, with a one-unit increase on the trait anxiety scale reducing the chance of remission from anorexia nervosa by a factor of 1.15.

    “Others have reported a linear relationship between responses to the trait anxiety scale from the STAI and remission; an increase of one unit on the trait anxiety scale reduced the chance of remission from AN by a factor of 1.15 (Yackobovitch-Gavan et al., 2008).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Yackobovitch-Gavan et al., 2008

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    The present definition of recovery did not include assessment of psychosocial functioning because none was available.

    “However, the present definition of recovery did not include assessment of psychosocial functioning because none was available.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductnone given

    Each model was evaluated for fit through residual analysis and a test for interaction with time for each covariate.

    “Each model was evaluated for fit through residual analysis and a test for interaction with time for each covariate.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductown data

    No variance inflation factor exceeded 2.1, which was not deemed a strong indicator of multicollinearity.

    “No VIF exceeded 2.1, a value not deemed a strong indicator of multicollinearity.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning

    These traits likely represent a pathway to anorexia nervosa in addition to decreasing the likelihood of recovery from the disorder.

    “these traits likely represent a pathway to AN in addition to decreasing the likelihood of recovery from the disorder.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Interpretationreasoning

    Fichter et al.'s measure of impulsivity might more appropriately index sensation-seeking behaviors.

    “whereas Fichter et al.’s (2006) measure might more appropriately index sensation-seeking behaviors.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductcitation

    The interquartile range for the interval between disorder onset and assessment interview was 4 to 14 years, which could lead to incomplete or biased reporting.

    “Third, the interquartile range for the interval between the onset of the disorder and the assessment interview ranged from 4 to 14 years, which is a significant length of time and could lead to incomplete or biased reporting.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationcitation

    The negative association between vomiting and recovery may be related to higher levels of psychological disturbance in individuals with anorexia nervosa and purging.

    “Why vomiting is a negative prognostic factor is unclear, but may be related to the higher levels of psychological disturbance seen in individuals with AN and purging.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Keel et al., 2004 source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    The concern over mistakes scale of the Frost Multidimensional Perfectionism Scale has been found to be uniquely elevated in anorexia nervosa compared to other psychiatric disorders.

    “The concern over mistakes scale measures the tendency to associate mistakes with a loss of respect and failure (Frost et al., 1990) and has been found to be uniquely elevated in AN compared to other psychiatric disorders (Bulik et al., 2003).”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Frost et al., 1990 source

      Not checked in the free run.

    • cites Bulik et al., 2003 source

      Not checked in the free run.

28 of 28

  • The authors state this as an assumption.

    Participants were able to report the temporality of their psychiatric diagnoses accurately.

    “These analyses also assume that participants were able to report temporality of their psychiatric diagnoses accurately.”

  • The authors state this as an assumption.

    Participants did not have a period of recovery (12 months without symptoms of anorexia nervosa) and relapse back to anorexia nervosa prior to the recovery event captured by the measures.

    “our analyses assume that participants did not have a period of recovery (12 months without symptoms of AN) and relapse back to AN prior to the recovery event captured by our measures.”

  • The paper relies on this without support.

    Temperament and personality traits are stable and not substantially affected by the emotional and cognitive stress of extreme dietary restriction and malnutrition.

    “although a convincing theoretical argument could be made that temperament and personality are stable, it is possible that women with AN might respond differently to these assessments when experiencing the emotional and cognitive stress of extreme dietary restriction and malnutrition than they would when fully weight-restored during recovery”
3 of 3

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Anxiety is a transdiagnostic factor involved in the development and maintenance of anorexia nervosa and is predictive for its course and treatment outcomes.

    “A transdiagnostic factor that is involved in the development, maintenance of, and is predictive for the course and treatment outcomes of anorexia nervosa is anxiety [10–20].”

    From Anxiety matters: a pilot lab study into food, weight, and virtual body exposure in anorexia nervosa (2024) · cites this paper as [14]

    The Results (Model 4: Combined Model; page unavailable) show that greater trait anxiety predicted a lower likelihood of recovery from anorexia nervosa. The Discussion suggests that anxiety-related traits may precede onset, and the paper proposes that the anxiolytic function of some anorexia nervosa behaviors may impede recovery. It does not establish anxiety as a transdiagnostic factor involved in development and maintenance or test its prediction of treatment outcomes specifically. (LLM, conf=0.85)

    Experimental — accuracy not yet measured

  • Trait anxiety predicts a lower likelihood of anorexia nervosa recovery.

    “and predicts a lower likelihood of anorexia nervosa recovery,3”

    From Neural correlates of a mindfulness-based intervention in anorexia nervosa (2023) · cites this paper as [3]

Cites

Flagged first, then by citations

Cited by

Sorted by citations

Related

Authors
Zerwas, S., Lund, B. C., Von Holle, A., Thornton, L. M., Berrettini, W. H., Brandt, H., Crawford, S., Fichter, M. M., Halmi, K. A., Johnson, C., Kaplan, A. S., La Via, M., Mitchell, J., Rotondo, A., Strober, M., Woodside, D. B., Kaye, W. H., & Bulik, C. M.
Affiliations
  • University of North Carolina at Chapel Hill (US)
  • University of Iowa (US)
  • University of Pennsylvania (US)
  • Sheppard Pratt Health System (US)
  • Ludwig-Maximilians-Universität München (DE)
  • Weill Cornell Medicine (US)
  • University Health Network (CA)
  • University of North Dakota (US)
  • University of Pisa (IT)
  • University of California, Los Angeles (US)
  • University of Toronto (CA)
  • University of California San Diego (US)
Corresponding author
Zerwas, Stephanie
ORCID
Venue
Journal of Psychiatric Research
Publisher
Elsevier BV
Published
2013-07
Volume
47
Issue
7
Pages
972-979
ISSN
0022-3956 (ISSN-L)
DOI
10.1016/j.jpsychires.2013.02.011
PMC
PMC3682792
OpenAlex
W2019515004
Cited by
111
References
62
Citations by year
2026: 4 · 2024: 14 · 2023: 7 · 2022: 5 · 2021: 14 · 2020: 8 · 2019: 7 · 2018: 10 · 2017: 11 · 2016: 7 · 2015: 14 · 2014: 6 · 2013: 4
FWCI
11.48
Type
journal article
Language
en
Open access
Licence
Locations
  • Journal of Psychiatric Research (publishedVersion)
  • PubMed (publishedVersion)
  • Europe PMC (PubMed Central) (acceptedVersion)
  • PubMed Central (acceptedVersion)
Funders
National Institutes of Health (MH66117, UL1RR025747, K12-HD01441)
Topics
  • Social Sciences › Psychology › Clinical Psychology › Eating Disorders and Behaviors
  • Health Sciences › Nursing › Nutrition and Dietetics › Nutrition, Health and Food Behavior
  • Health Sciences › Medicine › Psychiatry and Mental health › Child Nutrition and Feeding Issues
Keywords
anorexia nervosa, recovery, prognostic factors, self-induced vomiting, impulsivity, trait anxiety, psychiatric comorbidity, personality traits, disease duration, women
MeSH terms
Adolescent, Adult, Aged, Anorexia Nervosa, Mental Disorders, Female, Humans, International Cooperation, Middle Aged, Personality, Prognosis, Retrospective Studies, Time Factors, Comorbidity, Survival Analysis, Recovery of Function, Young Adult

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-09-28.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-09-28.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review