The transition from genetic linkage analyses to association studies (Risch and Merikangas, 1996; Li et al., 2005) presents a situation when affected sibs or affected pedigree members are often used as case samples in a case-control association study (Bourgain, 2005; Epstein et al., 2005; Moore et al., 2005; Biedermann et al., 2006; Klei and Roeder, 2007; Köhler et al., 2007; Yoo et al., 2007; Visscher et al., 2008; Knight et al., 2009). Since the correlation structure between sibs or relatives is given, it is not necessary to use techniques such as the generalized estimating equation as has been carried out in (Silverberg et al., 2003). Instead, variance of correlated samples can be calculated (Slager and Schaid, 2001) and its effect on the test statistic can be determined. The method discussed in (Slager and Schaid, 2001) is however only applied to the Armitage trend test.
