Latent Class Analysis of Cognition from the ADC

 
   Lead Investigator:    Denise LaBelle
   Institution      :    Cleveland Clinic - Lou Ruvo Center for Brain Healt
   E-Mail           :    labelld@ccf.org
   Proposal ID      :    804



   Proposal Description:
   Latent class analysis represents a data-driven, person-centered, analytic         method which can be used to detect self-segregating latent subpopulations from    within a heterogeneous overall sample by analyzing patterns of performance on     across variables of interest. The overarching goal of the proposed research is    to apply this method to cognitive indices from the ADC dataset, and               subsequently compare empirically-derived cognitive phenotypes to clinical         diagnoses and post-mortem pathological findings.                                  Objective 1: Identify latent classes of cognitive performance from baseline       data                                                                              Objective 2: Correlate latent cognitive phenotypes at baseline with clinical      diagnosis at baseline and later secondary assessment to determine whether         baseline cognitive status is a useful predictor of clinical diagnosis             Objective 3: Correlate latent cognitive phenotypes at baseline with post-mortem   findings to determine whether baseline cognitive status is a useful predictor     of pathological diagnosis