SYNOPSISThe Eating Attitudes Test (Garner & Garfinkel, 1979) was administered to 578 students (446 females and 132 males) at a College of Technology and to 14 ‘control’ subjects fulfilling diagnostic criteria for anorexia nervosa. Although no male students scored highly, a total of 28 female students (6·3% of the female student sample) scored in the ‘anorexic’ range. These ‘high scorers’, together with a random control group of 28 ‘non-high scorers’, were subsequently Interviewed. Interview revealed that the symptoms of anorexia nervosa were common in the high scoring group but virtually absent in the student control group. It is concluded that a substantial proportion of post-pubertal females (approximately 5%) develop a subclinical form of anorexia nervosa. The implications of these findings are discussed in relation to the aetiology, prevention and treatment of disturbances in eating behaviour.

A number of authors have emphasized the importance of self-esteem in the aetiology of the eating disorders anorexia nervosa and bulimia nervosa. Evidence for such theorizing, however, mainly derives from clinical observations on people being treated for eating disorders. This study is the first prospective study to investigate the role of self-esteem in aetiology prior to the onset of an eating disorder. Self-esteem was measured in 594 schoolgirls aged 11-12 using the Rosenberg Self-Esteem Scale (Rosenberg, 1965). Almost 400 of these girls were successfully followed up at age 15-16 and they completed a questionnaire examining eating and other psychological problems. Results showed that girls with low self-esteem at age 11-12 were at significantly greater risk of developing the more severe signs of eating disorders, as well as other psychological problems, by the age of 15-16. It is argued that more research is needed to replicate and extend these findings. The results also give weight to the case for examining the potential role of self-esteem enhancement in the prevention of eating disorders.

Research suggests that approximately 50 per cent of patients drop out prematurely from services for anorexia nervosa. This research set out to investigate the possible reasons for this high drop‐out rate. The experiences of eight women were explored through written accounts and semi‐structured interviews. The information was analysed using a social constructionist revision of grounded theory. A central theme of control emerged which was affected by interactions with the wider social system. The cumulative battle for control between the anorectic and others appeared to result in the act of dropping out as the treatment approach, timing and context recreated the setting conditions of lost control. The organizational and clinical implications of this model are discussed. Copyright © 2002 John Wiley & Sons, Ltd and Eating Disorders Association.

The study confirms previous evidence of a markedly increased mortality risk for anorexia nervosa, but also suggests that the risk is not confined to this eating disorder and that greater attention needs to be focused on wider psychiatric comorbidity.