Sample heterogeneity could also be driving the lack of strong positive findings in ADHD. For example, different assessment procedures among the three samples (different instruments and different ways of combining information sources) could have increased the phenotypic heterogeneity of cases. We have also demonstrated that persistent ADHD is a more familial form of the disorder and that ADHD symptoms typically persist in 30–60% of childhood cases. 23,24 Focusing on childhood samples may introduce additional noise by including a large subgroup of cases who will remit from ADHD and may have a less “genetic” etiology. Additional genome-wide studies focused on adult ADHD are needed to follow-up on the promising results suggested in Lesch et al. 4 It is also possible that studies of derived phenotypes or of gene-environment interaction might yield evidence for genome-wide significance, although such strategies have not proven effective for the IMAGE GWAS. 6, 19, 25, 26
