Emerging issues in the genetics of dyslexia: a methodological preview

A review of the classic and recent evidence on the genetics of reading disability (RD) shows encouraging progress, and accumulating evidence of genetic risk factors that operate within families and are separately localizable to more than one chromosomal region. The accelerating pace of these findings, however, suggests the need to consider some methodological issues about the design and interpretation of current and future studies. A major issue is the shape of the distribution of reading ability in the population, and we offer three tests of increasing rigor for determining whether those distributions are categorical, and hence not suitable for analyses that depend on the assumption of a continuous normal distribution. These tests are as follows: a nonnormal preponderance of cases with RD (i.e., the hump in the lower end of the distribution); a difference in the within-group variance-covariance matrices for typical readers compared to those with RD; and a correlation between a neurogenetically relevant criterion and a categorical reading variable that is larger than the correlation between the same criterion and a continuous version of the same reading variable. We emphasize also the importance of interactive relationships between multiple genetic loci, the variations in genotypic range as well as type of affectedness, the need to account for remediation variance, and the importance of lifespan changes in the phenotypes.