|
Abstract:
|
In studies of cognitive aging, it is crucial to distinguish subtypes of longitudinal cognition change while accounting for the effects of given covariates. The patterns of the longitudinal cognition trajectories are typically nonlinear with heterogeneous shapes that do not follow a simple parametric form. Although functional clustering has been a useful nonparametric tool to characterize the subtypes of nonlinear longitudinal trajectories, most existing approaches do not allow controlling for the possible functional effects of covariates in clustering. In this paper, we propose a novel latent class functional mixed-effects model for flexible conditional functional clustering, where the covariates are modeled as having fixed functional effects, and the random curves are assumed to be generated by a mixture of Gaussian processes. Each component of the mixture corresponds to a cluster of trajectories with similar patterns after controlling for the functional effects of the covariates. The new method is applied to the latest data from the Religious Orders Study and Rush Memory and Aging Project; and four novel subtypes of cognitive patterns are identified.
|
