Measurement error issues in studying the effect of gene-environment interactions on disease risks

Complex diseases, such as different types of cancers, are influenced by genetic and environmental factors and their interactions. There is overwhelming evidence that the effects of environmental factors on most cancers are modified by individual genetic characteristics. The accuracy of assessing the effects of gene-environment interactions on disease risks depends on how accurately the exposure to environmental factors can be measured or how accurately genetic makeup can be classified or both. Measurement error or misclassification can seriously distort the true effects of gene-environment interaction and produce biased estimates of the effects. Dr. Shahadut Hossain is developing a flexible modeling approach to adjust for biases when some of the quantitative environmental exposures are measured inaccurately. Hossain is also working to extend this methodology so that it can incorporate both exposure measurement errors and gene misclassification. His research involves studies of non-Hodgkin lymphoma, ovarian cancer and prostate cancer conducted with the Cancer Control Research Program at the BC Cancer Agency. Hossain hopes his work will enable the assessment of gene-environment interactions to be done more precisely, contributing to a better understanding of the effects of these interactions and more effective intervention strategies to prevent these diseases.