Background: The prevalence of autism is reported to be on the rise worldwide. Change of diagnostic criteria and a broadening of the concept of autism have been mentioned as contributing factors. Further studies of general populations are needed. The present study assessed the distribution of autistic features in a total population of children 7-9 years of age, and explored the impact of age, gender, informant, and participation bias on symptom report.
Methods: Teacher and parent forms of the Autism Spectrum Screening Questionnaire (ASSQ) were used in the "Bergen Child Study", a total population study of 9430 children aged 7-9 years. Completed teacher forms were returned for 97% and parent forms for 71% of the children. High-scorers were defined according to the ASSQ validation study by Ehlers, Gillberg, and Wing (1999).
Results: The distribution of ASSQ scores was found to be almost continuous. Of the children with both a teacher and a parent form, 2.1% were defined as high-scorers. Children without parent informed consent (i.e., anonymous children) obtained significantly higher teacher scores than those who had questionnaires completed by both parent and teacher. Adjusting prevalence for the anonymous children, the prevalence of high-scorers was 2.7% of the total population. Age did not affect symptom scores. Boys scored higher and parents reported more symptoms, particularly in girls. Agreement between informants was low to moderate.
Conclusions: Autism symptoms are not uncommon in the general population of children. Our findings are consistent with the concept of autism as a spectrum. Non-responders had a higher load of autism symptoms than identified children, indicating that reports on the prevalence of autism in a responder group underestimate true prevalence. Large differences across informants suggested the need to gather information both from families and from schools when screening for autism spectrum disorders.