This study aimed to test whether prospective memory (PM) was an early cognitive marker of future cognitive decline and incident dementia using longitudinal data spanning 8years from the Sydney Memory and Ageing Study. At baseline, 121 participants aged 72-91years were tested in PM using a validated PM task, Virtual Week, which included time- and event-based tasks presented with varying regularity. Responses were scored "Correct" if completed accurately and "Missed" if the target was not remembered at any time. Measures of cognition were taken at baseline and 2-year intervals over 8years. Dementia diagnoses were made by expert consensus panels using Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition criteria. Linear mixed models and Cox proportional hazards regression models were used to analyze the data, controlling for potential confounds. Both decreased PM accuracy and missed PM responses were associated with rate of cognitive decline measured by Mini-Mental State Examination over 8years and global cognitive decline over 4years. Risk of incident dementia increased with poorer baseline PM ability and missed responses. These effects remained significant after controlling for baseline cognition and were strongest for event-based and regular PM tasks. PM is a sensitive early marker of future cognitive decline and risk of incident dementia. PM tasks supported by spontaneous retrieval (event-based) and those with lower retrospective memory demands (regular tasks) function as particularly sensitive predictors. In other words, deficits in performing less effortful PM tasks best predicted cognitive decline. These findings may encourage clinicians to incorporate PM tasks in clinical assessments.
Read full abstract