Evaluating the Clinical Validity of Novel Gene-Environment Risk Profiles

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): This project will develop and evaluate genetic and gene-environment risk profiles for breast cancer and type 2 diabetes using data from the Nurses' Health Study (NHS) and Health Professionals' Follow-Up Study (HPFS), based on validated risk markers from recent genome-wide association studies. Because the NHS and HPFS are cohorts with extensive prospectively-collected data on known, modifiable environmental, lifestyle and anthropometric risk factors, this project will be the first to empirically examine how much the inclusion of genetic factors and gene-environment interactions improves clinical validity relative to standard risk prediction models which only use non-genetic factors. In particular, the project will calculate empirical absolute risks (5-, 10-year and lifetime risks) of breast cancer and type 2 diabetes for individuals with different profiles. These absolute risk estimates are needed to evaluate the risk-benefit balance for "personalized prevention" strategies, such as recommending that women with a particular gene-environment profile start mammography screening at earlier ages. This project will also develop new statistical methods to infer absolute disease risks for genetic and gene-environment risk profiles by combining data on genetic markers and environmental exposures collected in a nested case-control with data on environmental exposures collected on the entire underlying cohort. These methods will be broadly applicable to other cohort studies and other common diseases and traits of public health relevance. RELEVANCE: This project will develop and evaluate genetic and gene-environment risk profiles for breast cancer and type 2 diabetes using data from the Nurses' Health Study and Health Professionals' Follow-Up Study, based on validated risk markers from recent genome-wide association studies. This work will provide estimates of the net benefits of "personalized prevention" strategies based on these profiles. The statistical methods developed for this project will be broadly applicable to other cohort studies and other common diseases and traits of public health relevance.