event
Wednesday 21 Nov 2012: Within-Subject Variance as a Predictor of Health Outcomes
Professor Michael Elliott - Ann Arbor University, USA
Rolle Building Room 213, Plymouth University 15:00-16:00
Growth Mixture Models (GMMs) are used to model heterogeneity in longitudinal trajectories. GMMs assume that each subject's growth curve, characterized by random coefficients in mixed e ffects models, belongs to an underlying latent cluster with a cluster-specifi c mean profi le. Within-subject variability is typically treated as a nuisance and assumed to be non-diff erential. Elliott (2007) extended the idea of modeling random e ffects as fi nite mixtures as in GMMs into the variance structure setting, where underlying `clusters'
of within-subject variabilities were related to the health outcome of interest while the subject-specifi c trajectories were treated entirely as nuisance and modeled by penalized
smoothing splines. We extend these ideas by allowing `heterogeneities' (i.e., clusters) in both the growth curves and within-subject variabilities and develop a method that simultaneously examines the association between the underlying mean growth pro file and the variance clusters with a cross-sectional binary health outcome. We consider an applications to predict onset of senility in a population sample of older adults using memory test scores and to predict severe hot flushes using the hormone levels collected over time for women in menopausal transition.
