Family-based association studies.

Family-based association methods are useful because they offer improved matching of controls to cases, with the result that they are not susceptible to confounding by population stratification. They also allow analysis of parent-of-origin effects and maternal-fetal interactions. The transmission/disequilibrium test (TDT) is a test of linkage and association that is equivalent to a matched case/control analysis, from which various extensions are possible. A logistic regression formulation leads to modifications for multiallelic markers, haplotypes, and quantitative traits. Some pitfalls are described, for the situations in which one parent is missing, genotyping errors have occurred, and haplotype phase is uncertain. The problem of testing association in general pedigrees is discussed, with particular reference to sib pairs without parents.