BOSTON—The standard statistical methods for assessing risk factors of patients associated with the time to dementia or estimating the incidence rate of dementia assume that everyone will eventually develop dementia. This assumption has been challenged by researchers at the University of Washington, in a presentation at the 2013 Alzheimer’s Association International Conference.
“We obtained highly statistically significant evidence for the existence of an immune subgroup,” reported Xiao-Hua Andrew Zhou, PhD, Professor in the Department of Biostatistics and a coinvestigator at the National Alzheimer’s Coordinating Center, School of Public Health and Community Medicine, University of Washington in Seattle. “The existence of an immune subgroup implies that statistical methods of future studies would likely need to be adjusted to account for this, rather than using the standard survival analysis methods.”
Results were based on an analysis of participants from the National Alzheimer’s Coordinating Center’s Uniform Data Set (UDS), which included 27,196 persons who were enrolled from 2005 through December 2012. The data were compiled from 29 National Institute on Aging–funded Alzheimer’s Disease Centers and included standardized clinical examinations, testing, diagnostic evaluation, and data collection protocol with annual longitudinal follow-up.
A total of 8,775 persons had been diagnosed as cognitively normal by a clinician when they were entered into the UDS database and served as the study population. After a mean follow-up of 5.5 years, 275 persons developed dementia. The researchers determined an estimated mean immune rate of 86.79% after controlling for age, sex, race, marital status, stroke, Parkinson’s disease, depression, diabetes, and maternal/paternal history of dementia.
Dr. Zhou pointed out several limitations to his group’s findings, including the fact that the UDS is not a representative sample of the US population. “So we focus on the existence of the immune subpopulation rather than estimating the true immune rate,” he commented.
“On average, study subjects are followed for only 5.5 years. The majority of them are right censored. Datasets with longer follow-up period are needed for accurate estimation of immune rate. Further studies are needed to confirm this existence of an immune subgroup and to determine the traits of the immune group,” he said.
—Colby Stong
Editor