|
Activity Number:
|
176
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Monday, August 4, 2008 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
| Abstract - #301027 |
|
Title:
|
Models for Identifying Risk Factors for Transitions from Intact Cognition to Mild Cognitive Impairment and Dementia
|
|
Author(s):
|
Richard Kryscio*+ and Lei Yu and Suzanne Tyas and David Snowdon
|
|
Companies:
|
University of Kentucky and University of Kentucky and University of Waterloo and University of Kentucky
|
|
Address:
|
230 Sanders-Brown Building, Lexington, KY, 40536,
|
|
Keywords:
|
Alzheimer's disease ; mild cognitive impairment ; transition models ; risk factors ; competing risks
|
|
Abstract:
|
We review strategies for modeling panel data collected on the cognitive status of elderly subjects being followed for the occurrence of dementia. Response at any given wave is a nominal variable representing the states intact cognition, mild cognitive impairment or a deficit in activities of daily living. Interest centers about transitions among these states with eventual absorption into one of two competing states: dementia or death. We compare the relative merits of several modeling strategies that will allow estimation of the effect of covariates on these transitions This includes Markov transition models, and competing risk models. An application to the NUN study data, a longitudinal cohort of 501 members of the School Sisters of Notre Dame aged 75 or older at baseline, will illustrate the results.
|
