Estimation of the population "at risk" for eating disorders in a non-clinical Swedish sample: a repeated measure...

To estimate the late-adolescent population "at risk" for eating disorders (EDs), establish its stability over time using repeated measures, and analyse its social and demographic background factors.

The Eating Disorder Inventory (EDI) and a number of related questions were administered to students in their last year of high school in five Swedish towns in 1993 (n=3855) and 1998 (n=2925).

It was found that 6% of girls were "at risk" for eating disorders with no increase over time. There was an association between this risk and certain socio-demographic factors (living alone, moving away from home early, excessive physical training and experiences of bullying). No differences were found between cities of different size.

A considerable proportion of girls in late adolescence are preoccupied with weight and dieting, and may be considered "at risk" for developing a clinical eating disorder. This proportion seems to be fairly stable over time.