CAN COMPUTERIZED NEUROPSYCHOLOGICAL TESTS BETTER DISCRIMINATE SUBJECTIVE COGNITIVE DECLINE IN ELDERLY?

Abstract

Background: Subjects with subjective cognitive decline (SCD) are at higher risk of developing mild cognitive impairment (MCI) and Alzheimer’s dementia (AD). By definition, they perform normally on conventional neuropsychological tests. However, it is unclear whether computerized tests can detect subtle changes in this population. Objective: To compare cognitive performance in conventional and computerized tests of the CANTAB battery in people over 55 years old, divided into three subgroups: Controls, SCD and MCI, according to the NIA-AA 2018 criteria. Methods: We included 64 volunteers: 19 controls, 15 SCD, 30 MCI. Principal Component Analysis (PCA) model was used in both tests and naive bayes classifier were used to distinguish SCD from controls. Results: In conventional tests, variability of 57,17%, differentiating only the MCI. The CANTAB showed a subtle difference in dispersion between SCD and controls, with a variability of 30,12%. Cognitive functions with greater differentiation: episodic visual memory and new learning with variability of 72,65%, visual perception and immediate visual memory 51,95% variability, with similar results between the SCD and MCI groups. Attention and psychomotor speed 23,89%. Sustained attention and psychomotor speed with 71,3%. Adjusted for demographic variables, 52,63% of the SCD were classified as MCI in the computerized test, while the conventional one did not change. Conclusion: Computerized tests seem to be more sensitive in differentiating SCDs from controls, resembling the MCI group.