Modeling Health Outcomes via Values, Gradients, or Variation of Follicle-Stimulating Hormone in Penn Ovarian Aging Study

• Activity Number: 314
• Type: Contributed
• Date/Time: Tuesday, August 6, 2013 : 8:30 AM to 10:20 AM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Bei Jiang*+ and Michael Elliott and Mary Sammel and Naisyin Wang
• Companies: University of Michigan and University of Michigan and University of Pennsylvania and University of Michigan
• Keywords: joint modeling ; time varying ; short term variability ; long term variability ; gradients

Abstract

Longitudinal predictors of health outcomes are typically summarized as mean profiles to predict distal outcomes. But such methods assume that these summary measures entirely capture the correlation between the dynamics of the longitudinal process and the health risks. We consider a joint modeling method that incorporates information about the mean and variability of the longitudinal predictors to predict outcomes, distinguishing between short-term residual variability and long-term variability captured by the derivative of mean trajectory. To maximize the model flexibility, we allow the potential effects of longitudinal trajectories on the health risks to be time varying. This feature allows further identification of critical time window of higher health risks. We apply this method to investigate the association between the individual profiles of follicle stimulating hormone (FSH) measures for women in transition to menopause and the severity of menopause-related symptoms using data from Penn Ovarian Aging Study.