|
Abstract:
|
Promoting positive lifestyle behaviors to attenuate lifetime risk of non-communicable disease is of great interest to public health researchers. However, due to gaps/biases in participant-reporting and unobserved heterogeneity in lifestyle behaviors across various at-risk population subgroups, statistical modeling to assess these measurements is challenging. Biomarkers of chronic disease may provide a proximal measure of energy balance, but the lack of non-invasive and inexpensive biomarkers limits their routine use in population studies. Moreover, the measurement of biomarkers is often subject to left-censoring due to detection limits, which leads to various statistical challenges. We propose a new method that constructs a quantile-specific weighted index of multiple behavioral components to address the aforementioned challenges. Under the censored quantile regression framework, the proposed method provides greater flexibility to characterize risk-specific behavioral patterns of dietary intake and physical activity in relation to longitudinal biomarker levels that are left-censored. The performance of the proposed method is evaluated through simulations and real data applications.
|
