Terminal-Decline Effects for Select Cognitive Tasks after Controlling for Preclinical Dementia

Abstract

In a previous study, the authors found no accelerated decline in close proximity to death for a measure of global cognitive functioning, after excluding persons in a preclinical phase of dementia. However, specific cognitive tasks might be more sensitive to terminal-decline effects. The purpose of this study was to explore possible terminal-decline effects for a range of cognitive tasks after controlling for preclinical dementia. Community-based cohort study. The Kungsholmen district of Stockholm. A total of 585 persons (75+ years) were repeatedly assessed over an 11-year period. Level and change in cognitive performance were compared for three groups: persons in close proximity to death, persons in a preclinical phase of dementia, and persons who remained alive and nondemented throughout the study. Tasks assessing primary and episodic memory, verbal ability, and visuospatial skill. Compared with an analysis where all dead subjects were included in the impending-death group, removing the preclinical dementia cases resulted in markedly attenuated mortality-related effects. However, the impending-death group still declined at a faster rate relative to the comparison group on Digit Span-forward, word recognition, and category fluency. Notably, these were tasks for which the comparison group showed no significant decline. A considerable proportion of the terminal-decline effect is accounted for by the impact of preclinical dementia. However, for tasks that are relatively resistant to age-related change, such effects might be detected independently of preclinical dementia.