A two-stage epidemiological study of abnormal eating attitudes and their prospective risk factors in Polish schoolgirls.

In this longitudinal two-stage screening study of abnormal eating attitudes and behaviours in Polish schoolgirls self-report questionnaires (EAT-26) were completed by 747 schoolgirls aged between 14 and 16 years. On the basis of their EAT-26 scores 167 girls were selected for a clinical interview conducted blind to questionnaire scores. No clinical cases of DSM-III-R anorexia or bulimia nervosa were identified, however, there was a point prevalence of 2.34% for subclinical eating disorder and of 28.6% for dieting. Natural history was investigated, repeating the screening and interviews after 10 months. Both the group mean EAT-26 score and the proportion of subclinical cases remained the same at follow-up however, individual girls showed substantial fluctuations over time. On re-screening 11.5% of the sample had moved EAT category (based on scoring below or above the clinical cut-off point). Follow-up interviews showed that 58% (7) of the original subclinical cases were no longer cases while seven girls were "new' subclinical cases. This second stage of the study also aimed to investigate the risk factors for the development of eating disorders. Several family factors distinguished the dieters from non-dieters at the initial stage. Subjects who began dieting between the initial and follow-up interviews had higher Body Mass Index, felt more overweight and had greater discrepancy between their current and ideal weights than the non-dieters. However, analysis of regression revealed no specific risk factors for development of a subclinical syndrome.