Sampling procedure, participation rates and representativeness in the Swedish part of the European Youth Heart Study (EYHS).

OBJECTIVE: The European Youth Heart Study (EYHS) is a cross-sectional, school-based population study on risk factors for future cardiovascular disease in children, with an overall participation rate in Sweden of about 50%. To study the representativeness of the participants in the Swedish part of EYHS, a comprehensive non-participant follow-up study was carried out.
DESIGN: A structured multilevel analysis model was developed, addressing each level in the sampling procedure. The income, educational and occupational categories of the geographical regions of the study (level I), school catchment areas (level II) and parents (level III) were compared with official data. Participating and non-participating pupils (level IV) were compared through a questionnaire.
SETTING: Thirty-seven state schools in two regions of Central Sweden (Orebro and southern Stockholm) were visited during the school year 1998/1999.
SUBJECTS: Boys and girls aged 9 and 15 years were randomly sampled through a multiphase sampling procedure.
RESULTS: Data for socio-economic status for levels I and II corresponded well to national and regional official data. At level III, non-manually working parents were slightly over-represented among parents of participating children. At level IV, non-participating subjects corresponded in most respects to participants with a few exceptions--mainly more interest in physical exercise among participants.
CONCLUSIONS: Based on the knowledge from the non-participant study, we do not foresee problems regarding interpretation of the outcomes in the EYHS, despite the low participation rate.