Combining the transmission disequilibrium test and case-control methodology using generalized logistic regression

To study the role of genetic factors in the etiology, susceptibility, or severity of disease, several methods are available. In a transmission disequilibrium test, genotypes of cases are compared to those of their parents to explore whether a specific allele, or marker, at a locus of interest appears to be transmitted in excess of what is expected on the basis of Mendelian inheritance. Such apparent excess transmission indicates that cases are being selected for that allele, thereby providing evidence that this allele is a risk factor for disease. In case-control studies, genotypes of cases are compared to those of controls from the same population to identify whether a specific allele is associated with disease. If so, either the allele at this locus or one in linkage disequilibrium with it may be causally related to the etiology of the disease. Here, we discuss the problem of combining a transmission disequilibrium test and a case-control comparison, in order to integrate all available information, and thereby increase statistical power. As the same cases are used in both approaches, the two results are not independent. However, parents of cases can be independently compared to controls. Both the issue of testing for a genetic effect and the estimation of relative risks under the multiplicative model using generalized logistic regression are discussed.