Case-control studies augmented by responses and covariates from family members allow investigators to study the association of genetics and environment with the response. In addition, such case-control family data allow investigators to directly relate within-family differences in covariates to differences in the response. However, existing approaches for case-control family data parametrize covariate effects in terms of the marginal probability of response, the same covariate effects that one estimates from standard case-control studies.
This talk focuses on the estimation of within-family covariate effects and presents a profile-likelihood approach thatapplies generally to settings where one has a fully specified model for the vector of responses in a family and particularly to family-specific models such as binary mixed-effects models. We illustrate our approach using data from a case-control family study of brain cancer and consider the role of conditional likelihood methods.
|