Factor-analytic study of the Anorectic Behavior Observation Scale in Japan: comparisons with the original Belgian study

Psychiatry Res. 2002 Aug 30;111(2-3):241-6. doi: 10.1016/s0165-1781(02)00143-9.

Abstract

The Anorectic Behavior Observation Scale (ABOS) is a questionnaire developed to obtain information from relatives about behaviors and attitudes of patients with eating disorders. The original report of the ABOS revealed three factors. This is the first study to confirm the factor structure by use of confirmatory analyses. Relatives of 102 patients with eating disorders completed the ABOS, and exploratory factor analyses were performed. Confirmatory factor analyses were performed for this Japanese sample, and the fitness of the factor structure was compared between that in the original report proposed by Vandereycken and that used in the present study. A three-factor structure was revealed by the data of Japanese patients, including factor I (eating behavior meeting the criteria of anorexia; concern with weight, foods, and denial), factor II (bulimia; including bingeing, disposing, furtive eating and purging, and talk about thinness and dieting), and factor III (hyperactivity, eating slowly, and chopping food into very small pieces). Confirmatory analyses supported the improvement of fitness in this modified three-factor structure in comparison with the original three-factor model. This Japanese ABOS study revealed three dimensions, which differed from the original subscales in Belgium. Cross-cultural differences were observed, though the ABOS may be useful in Japan as an instrument for assessing eating disorders.

MeSH terms

  • Adolescent
  • Adult
  • Anorexia Nervosa / diagnosis*
  • Anorexia Nervosa / psychology
  • Belgium
  • Bulimia / diagnosis*
  • Bulimia / psychology
  • Cross-Cultural Comparison*
  • Factor Analysis, Statistical
  • Feeding and Eating Disorders / diagnosis*
  • Feeding and Eating Disorders / psychology
  • Female
  • Humans
  • Japan
  • Male
  • Personality Assessment / statistics & numerical data*
  • Psychometrics
  • Reproducibility of Results