To make further progress with the investigation of common variation in ASD, several strategies are being implemented. First, substantial increases in sample size are necessary. This is on the near horizon, with the ongoing activities from groups such as the iPSYCH collaborative likely to bring data from thousands of additional ASD cases to the GWAS effort. Secondly, genetic designs for studies of complex developmental disorders, including ASD and ADHD, have tended to favour a trio-based design and family-based association testing. Although this has provided strength in reducing artefacts resulting from population stratification, recent simulations advise against their use in common and complex polygenic traits, especially where assortative mating may be involved and where the family is known to be multiplex [74]. In such cases, the trio design can underestimate the SNP heritability and the power to observe association.
