Background: While preliminary evidence supports the criterion validity of the CogState computerized brief battery in mild cognitive impairment (MCI) and Alzheimer disease (AD), definitive validation studies examining a wider range of dementia-related disorders relative to conventional neuropsychological techniques are necessary.
Methods: Participants satisfying clinical consensus criteria for dementia (AD, n = 37; frontotemporal dementia, n = 7; and dementia with Lewy bodies, n = 5), MCI (n = 16), and the healthy controls (n = 22) were administered a battery of brief neuropsychological and select computerized (CogState) cognitive tests. The battery, administered through the University of Michigan Alzheimer's Disease Research Center, included measures of processing speed, attention, working memory, and learning.
Results: CogState and standard neuropsychological task scores were significantly lower for dementia participants than that of the nondementia groups (P < .05), with a single CogState test distinguishing control from MCI participants, but minimal differentiation existing between dementias using the CogState. Correlations were modest between conventional and computerized test scores, covering matching domains and mostly reflecting the multidimensional nature of cognitive paradigms.
Conclusions: Results support the clinical validity of this brief computerized screening battery when used in established dementias, but not to differentiate between various dementias, and suggest that the select CogState battery's effectiveness in identifying MCI from controls was not as strong as identifying specific dementias.