A Joint Model for the Analysis of Multiscale Time to Events and Longitudinal Binary Data

• Activity Number: 224
• Type: Topic Contributed
• Date/Time: Monday, August 2, 2010 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section
• Author(s): Rajeshwari Sundaram*+ and Alexander Mclain and Paul S. Albert and Germaine M. Buck Louis and Kirsten J. Lum
• Companies: Eunice Kennedy Shriver National Institute of Child Health and Human Development and Eunice Kennedy Shriver National Institute of Child Health and Human Development and Eunice Kennedy Shriver National Institute of Child Health and Human Development and Eunice Kennedy Shriver National Institute of Child Health and Human Development and The Johns Hopkins University
• Keywords: Survival ; Joint modeling ; Fecundity ; Likelihood

Abstract

Joint modeling of longitudinal and survival data is of considerable importance in many biomedical studies. This talk is motivated by a prospective pregnancy study, where interest lies in understanding the associations between longitudinally measured lifestyle factors, intercourse behavior, menstrual cycle length and time to pregnancy. These data have a unique feature, namely, the longitudinal processes are measured on a daily time-scale within each menstrual cycle and time to event is measured on the cycle level. Furthermore, menstrual cycle lengths vary between and within women and may share common risk factors with the longitudinal as well as the cycle level time to event. We propose a joint model for the longitudinal process and the two multiscale time to events, with bivariate latent variable capturing dependencies. We will illustrate our methods using a prospective pregnancy study.