Biomarker Development for Use in Regression Calibration

Abstract

It is challenging to alleviate systematic measurement error in self-reported data when studying the associations between dietary intakes and chronic disease risk. The regression calibration method has been used for this purpose when an objectively measured biomarker that satisfies a classical measurement error assumption is available. The requirement for the biomarkers needs to be quite strong and very few dietary intake biomarkers as such have been developed. Feeding studies provide opportunities to develop such potential biomarkers using regression methods with a much larger variety of dietary variables. However, the measurement error for the resulting biomarkers will be of Berkson type and these biomarkers are not suitable to the existing regression calibration method. Ignoring the violation of the classical measurement error assumption can lead to severe biases in disease association estimates. In this project, we propose three ways to obtain consistent estimates of such associations under rare disease assumption. The asymptotics of the proposed estimators is derived. Theoretical and numerical analyses were performed to compare these estimators. Estimation procedures are applied to the Women’s Health Initiative (WHI) data to re-examine the associations between dietary intakes and cardiovascular diseases.