Effects of socio-demographic variables on performance on the Cambridge neuropsychological automated tests for the assessment of dementia and Portuguese norms for older adults living in retirement homes

Abstract

Objective: This study aimed to analyze the effects of age, education, gender, computer experience, institutionalization time, and psychotropic drug use on performance on four tests of the Cambridge Neuropsychological Test Automated Battery (CANTAB) which are recommended for the assessment of dementia (Rapid Visual Information Processing [RVP], Paired Associates Learning [PAL], Spatial Working Memory [SWM], and Reaction Time [RTI]), and to provide norms for Portuguese older persons without neuropsychiatric diagnoses who are living in retirement homes. Method: The normative sample included 128 adults aged 69–96 years who had no neuropsychiatric diagnosis and who had lived in retirement homes for 3–232 months. The CANTAB was administered, at the latest, one week after a screening session that comprised an interview and the administration of pencil-and-paper tests. Results: The simultaneous multiple linear regression models were significant (p < .05) for all tests except the RTI five-choice movement time measure. The total variance explained by the socio-demographic variables was smaller for the CANTAB measures (4–14%) than for the pencil-and-paper tests (10–33%). Significant effects involving age or gender were observed for RVP, PAL, and SWM. A marginally significant computer experience effect was found for the RTI simple movement time measure. We additionally observed significant effects of education, age, gender, and computer experience on several pencil-and-paper tests. Conclusions: Our findings suggest that different socio-demographic variables influence distinct tests and measures of the same test, and that the associations between computer experience and several pencil-and-paper tests may be mediated by possible cognitive skills developed through computer use.