Date of Award

8-2011

Degree Type

Report

Degree Name

Master of Science (MS)

Department

Mathematics and Statistics

First Advisor

Dr. Christopher Corcoran

Abstract

The Cache Lifespan Stressors and Alzheimer's Disease (LSAD) study has access to data from the Cache County Study on Memory Health and Aging (CCS) that have been linked to the extensive genealogical and vital records from the Utah Population Database (UPDB). Information about stressful life events experienced by the original 5092 CCS participants has been extracted objectively from the UPDB, without the possibility of recall bias. This information was then statistically analyzed to look for relationships between key stressors and dementia risk. The LSAD study made it possible to examine the correlation between stressors as well as look at patterns or groupings that may exist among the stressors.

For this project, we will apply, compare and contrast a variety of methods to explore the relationship between stressors and dementia status. We first assessed significant associations between stressors and dementia using chisquare tests, logistic regression, and stepwise logistic regression. CART and Random Forests were also used to evaluate which variables are most predictive of dementia status. To explore interrelationships among the stressors, principal components analysis, factor analysis, and cluster analysis were performed. These last methods also obtain dimension reduction of the set of stressor variables to a smaller set of factors or clusters of stressors. For each of the methods, we explored some basic strategies to account for the high rate of missingness among some of the stressors.

Comments

This work made publicly available electronically on August 30, 2011.

Share

COinS