This is a prospective study of self-reported eating disorder symptoms (EDS) in youth. The aim was to enhance the detection of EDS and the use of health records in identifying individuals with the highest risk of EDS. 

In the outset, the 8th (n=1036) and 9th (n=855) graders and their school nurses (n=14) in Finnish-language secondary schools in Turku provided data for testing the feasibility of the SCOFF questionnaire in measuring EDS. The 8th grade data were also used to investigate gender differences in EDS. The course of EDS was investigated with a 1-year follow-up of the 9th graders (n=372), and a 4-year follow-up of the initially EDS-positive participants (n=208) and their non-symptomatic counterparts (n=514). 

SCOFF proved to be a feasible screen for all mid-adolescents. With SCOFF, the detection of EDS increased fourfould. In mid-adolescence, girl preponderance in EDS was small and associated with appearance and weight dissatisfaction. At the 1-year follow-up, almost one half of initially EDS-positive adolescents were still symptomatic. Prolonged EDS was associated with anxiety. At the 4-year follow-up, one half of the girls and a significant number of boys with EDS in mid-adolescence were still EDS-positive. The strongest predictor of prolonged EDS was depressiveness in mid-adolescence. On population level, the estimated prevalence of protracted EDS was 6%–11% in females and 0.7%–2% in males in late adolescence. 

The detection of EDS may be enhanced by including SCOFF in health examinations throughout adolescence. To identify individuals most in need for clinical attention, health records provide useful information.