A1080
Title: Modelling biomarker variability in joint analysis of longitudinal and time-to-event data
Authors: Christiana Charalambous - University of Manchester (United Kingdom) [presenting]
Abstract: Visit-to-visit variability of a biomarker has been recognized as an important driver in predicting related diseases. However, existing variability measures are criticized for being entangled with random variability caused by measurement error or for being unreliable. A new measure is proposed to describe the biological variability of a biomarker by capturing the fluctuation of individual trajectories behind longitudinal measurements. Given a mixed-effects model for longitudinal data with the mean function over time specified by cubic splines, the proposed variability measure can be mathematically expressed as a quadratic form of random effects. For the time-to-event data, a Cox model is used, which incorporates the defined variability and the current level of the underlying longitudinal trajectory as covariates. The proposed joint models are further extended to incorporate the weighted cumulative effects of both biomarker level and variability on the survival hazard. Simulation studies are conducted to reveal the advantages of the proposed methods over the two-stage method, as well as a simpler joint modelling approach that does not take biomarker variability into account. Finally, the models are applied to investigate the effect of systolic blood pressure variability on cardiovascular events in the Medical Research Council elderly trial.
