Clinical Utility of a Reaction-Time Attention Task in the Evaluation of Cognitive Impairment in Elderly with High Educational Disparity.

Abstract

The Clinical Dementia Rating (CDR) scale is commonly used to stage cognitive impairment, despite having educational limitations. In elderly with low education, a previous study has shown that intraindividual variability of reaction time (CV) and commission errors (CE), measured using a culture-free Go/No-Go task, can reliably distinguish early Alzheimer's disease (AD) from mild cognitive impairment (MCI) and healthy controls. We aimed to extend the clinical utility of this culture-free Go/No-Go task in a sample with high educational disparity. One hundred and ten participants with a wide range of years of formal education (0-14 years) were randomly selected from a geriatric unit and divided based on their CDR scores into cognitively unimpaired (CDR = 0), MCI (CDR = 0.5), and early AD (CDR = 1). All underwent a 90-s reaction-time test that measured the variables previously found to predict CDR in low educated elderly. Here we added years of formal education (educational level) to the model. Multivariate analyses compared differences in group means using educational level as confounding factor. A confirmatory discriminant analyses was performed, to assess if CDR scores could be predicted by the two Go/No-Go variables in a sample with high educational disparity. Over all three groups, differences in both CE and CV reached statistical significance (p < 0.05). The discriminant analysis demonstrated that CV and CE discriminated cognitively impaired from cognitively normal elderly. These results remained similar when discriminating MCI from cognitively unimpaired elderly. The Go/No-Go task reliably discriminates elderly with MCI from elderly without cognitive impairment independent of educational disparity.