Our study provides little support for association with additional candidate genes for ADHD previously implicated via meta-analysis; 17, 20 only 6 of the 42 candidates examined demonstrated nominal gene-wide statistical significance, and we failed to replicate any of the positive associations reported with individual SNPs examined and subjected to meta-analysis. This is not surprising considering the number of trios examined and that any associated genes are likely to exert a small or modest increase in risk (i.e. odds ratios <1.5). The sample size required to detect such an association (OR=1.5, p=5E-08, MAF=0.25) is at least 1,188 trios. However, many tens of thousands of subjects will likely be need to uncover true risk loci of smaller effect. For example, GWAS with samples of 29,136, 34,433 and 71,225 subjects were needed to identify replicated genetic loci for blood pressure. 21, 22
