Research Using the NACC Database

A Clinical Taxometric Investigation of Alzheimer's Disease and Other Dementias

 
   Lead Investigator:    Brandon Gavett
   Institution      :    Boston
   E-Mail           :    bgavett@uccs.edu
   Proposal ID      :    112


   Publications:


1  Gavett BE
   Dementia has a Categorical, not Dimensional, Latent Structure
   Journal Article  Psychol Aging,27:791-797,2012,PMC3445289

   PUBLISHED





   Proposal Description:
   Despite advances in the clinical identification of Alzheimer's disease and other dementias, there remains considerable overlap between normal aging and dementia.  Some have argued that neurodegenerative diseases represent the extreme end of a continuum of age-related changes, but in general, the conventional school of thought suggests that neurodegenerative diseases are qualitatively distinct from normal aging, both clinically and neuropathologically. Recent statistical advances allow for issues such as this to be addressed empirically. Taxometric procedures were designed to examine the underlying dimensional structure of a construct as measured by clinical variables (e.g., cognitive test performance), with the goal of determining whether the underlying entity is better represented as categorical or continuous. The specific aim of the current study is to apply taxometrics methods to manifest variables collected from aging participants in order to determine whether the clinical picture of dementia is better represented as falling on the extreme end of a continuous aging spectrum or as a discrete clinical entity. Because this is an exploratory study, we do not formally hypothesize whether the data will support a continuous versus categorical structure. However, if the underlying structure is found to be categorical, the taxometric procedures allow for an estimation of the base rate of the taxon(s) and can identify which indicator variables are the most useful at discriminating taxon members from non-members. A similar study was conducted using NACC data by Walters (2010), which suggested that the construct of dementia was distributed dimensionally. However, the Walters (2010) study used a very restrictive conceptualization of dementia (i.e., low cognitive functioning), without considering functional implications or change from a previous ability level. We propose to replicate and extend Walters' findings by focusing more specifically on Alzheimer's disease and by broa