Assessing effects of disease genes and gene-environment interactions: the case-spouse design and the counterfactual-control analysis

Abstract

Background: Assessing joint genetic and environmental contributions to disease risk is the central issue in many genetic epidemiological studies. To characterise the effects of a gene, the case-control study may suffer from the problem of population stratification bias. For a late onset disease, recruiting control subjects into case-parents and case-sibling studies may be difficult.

Methods: Two novel approaches to analysing case-spouse data are introduced: the 1:1 case-counterfactual-control analysis (genotype swapping between the case and their spouse) and the 1:5 case-counterfactual-controls analysis (allele swapping).

Results: Both can be implemented using statistical packages that allow matched analysis (the conditional logistic regression) to yield valid estimates of the genotype relative risk, the gene-environment interaction parameter, the gene-sex interaction parameter, and the gene-environment-sex three factor interaction parameter (if desired), if certain assumptions are fulfilled.

Conclusion: Because of the ease in recruiting subjects, and in collecting and analysing data, this approach makes a convenient tool for gene characterisation.