Background: Twin studies have found that childhood attention-deficit hyperactivity disorder (ADHD) has a strong genetic component. Estimates of heritability, the extent of non-additive genetic effects and of 'sibling contrast' effects vary between different studies.
Aims: To use multiple informants to assess the extent to which observer effects influence such estimates in an epidemiological sample of twins.
Method: Questionnaire packs were sent to the families and teachers of twins aged 5-16 years in the Bro Taf region of South Wales. The twins were ascertained from community paediatric registers.
Results: Both parent- and teacher- rated data showed a high degree of heritability for ADHD measured as a symptom dimension, but the correlation between the two types of rater was modest. Bivariate analyses suggested that parent and teacher ratings reflect the effects of different genes. Self-report data from twins aged 11-16 years showed no evidence of genetic effects.
Conclusions: Although ADHD is shown to be highly heritable by both parent- and teacher-rated data, the underlying genotypes may be substantially different. This has implications for study designs aiming to find genes that contribute to the disorder.