Predictability of a favorable outcome in anorexia nervosa

Eur Psychiatry. 2005 Mar;20(2):165-72. doi: 10.1016/j.eurpsy.2004.09.006.

Abstract

Objective: In a long-term follow-up of anorexia nervosa (AN) patients, somatic, psychological and social variables at clinical presentation should be investigated using a multilevel approach.

Methods: This study isolated predictors known from the literature over longer time periods and carried out a separate investigation of predictors in a sample of 81 AN patients of the Heidelberg-Mannheim study over a mean period of 12 years (range 9-19 years). Separate hierarchic regression analyses on the basis of the course of the Morgan-Russell categories were calculated for four individually recorded areas: anamnestic, psychological, somatic and social data sets.

Results: Age at the onset of the disease, purging behavior, low serum albumin, high glutamic-oxalo acetic transaminase (GOT) psychopathology (ANSS) and social pathology had the highest predictive value qualities. In survival analysis overall assessment of all six main predictors at clinical presentation could differentiate all patients who recovered from those who remained ill (log-rank test P = 0.019).

Conclusion: A small number of variables were important for detecting a good or poor long-term course of AN. At onset of the disease, it seems necessary to evaluate these psychological, somatic and social predictors.

MeSH terms

  • Adaptation, Psychological
  • Adolescent
  • Adult
  • Albumins / metabolism
  • Anorexia Nervosa / blood
  • Anorexia Nervosa / epidemiology
  • Anorexia Nervosa / therapy*
  • Aspartate Aminotransferases / blood
  • Child
  • Disease Progression
  • Female
  • Follow-Up Studies
  • Health Status
  • Humans
  • Male
  • Mental Disorders / epidemiology
  • Prospective Studies
  • Regression Analysis
  • Retrospective Studies
  • Social Behavior
  • Survival Analysis
  • Treatment Outcome

Substances

  • Albumins
  • Aspartate Aminotransferases