Comparison of age and time-to-death in the dedifferentiation of late-life cognitive abilities.

The dedifferentiation hypothesis proposes that specific cognitive abilities become more highly associated with general ability in old age as a result of increasing biological constraints on fluid intelligence. There is limited evidence for the hypothesis, and research has not tended to clearly distinguish age dedifferentiation from ability differentiation and other age-related phenomena. The present study examined age dedifferentiation using a structural equation model that controlled for ability differentiation, along with linear and quadratic effects of age. Time-to-death was examined as an alternative time metric to chronological age, as it may better represent biological constraints. The Canberra Longitudinal Study community-based cohort, consisting of 896 Australian adults aged 70 and over, provided data from 687 decedents who were followed for up to 17 years. Results indicated little support for the age dedifferentiation hypothesis, with only two of seven cognitive tests showing significant age dedifferentiation. The time-to-death metric showed more evidence of dedifferentiation, with four of the seven tests exhibiting dedifferentiation. However, after excluding participants with possible cognitive impairment, all of the dedifferentiation effects were attenuated to nonsignificance. Age dedifferentiation effects may therefore reflect dementia and other mortality-related pathology rather than being an inevitable outcome of advanced age. Alternative developmental theories for cognitive function must better account for the diversity of late-life abilities and pathology.