Statistical methods for tracking clinical course

 
   Lead Investigator:    Alireza Atri
   Institution      :    Bedford VAMC  Massachusetts General Hospital
   E-Mail           :    aatri@partners.org
   Proposal ID      :    209



   Proposal Description:
   Project 2 abstract: obtain sample size estimates. To adapt and apply advanced multivariate longitudinal modeling approaches to classify, compare and assess the determinants and predictive factors, including demographic and clinical patterns (e.g. unique profiles such as gender, history, education, and motor findings), latent constructs (e.g. cognitive reserve), and use of anti-AD medications and supplements, that may potentially impact the trajectory of progression of cognition, function, behavior, overall  severity, and clinical milestones in healthy cognitive aging, cognitive impairment, and dementia, including the Alzheimer???s disease spectrum.  Specific Aim 1: Adaptation and application of advanced multivariate modeling approaches including Latent Variable Models and Cluster Analysis to affect specific aim#1 in Project 1.  Specific Aim 2: Adaptation and application of advanced multivariate modeling approaches including mixed effect regression models, generalized estimating  equations, and latent variable growth curve models to affect specific aim#2 in Project 1.  Specific Aim 3: To develop and provide advanced tools, including software code and programs, to assist NACC UDS researchers to analyze and model missing data, better classify and track performance on clinical tests and measures (including neuropsychological tests), predict progression, and conduct power analysis and obtain sample size estimates.