Send to

Choose Destination
J Alzheimers Dis. 2011;24(1):125-36. doi: 10.3233/JAD-2010-100450.

Differential cognitive deterioration in dementia: a two year longitudinal study.

Author information

Department of Psychiatry, University of Cambridge School of Clinical Medicine, Addenbrooke's Hospital, Cambridge, UK.


The ability to predict cognitive deterioration in patients with dementia holds valuable potential for clinical trials and early intervention. This study identified cognitive domains deteriorating differentially over time as well as baseline predictors of subsequent cognitive decline in patients referred to a memory clinic. Twenty-six subjects with Alzheimer's disease (AD) and 43 subjects with Subjective Memory Impairment (SMI) were entered into a longitudinal study in which cognitive function was assessed at baseline and at 8-monthly intervals for 2 years, using a range of well-validated measures. Thirty-seven patients with depression and 39 healthy controls were also longitudinally assessed. AD was associated with disproportionate deterioration over time on general measures of cognitive function, multiple measures of mnemonic processing, mental fluency (letter and category), and aspects of motor speed. SMI showed restricted relative cognitive deterioration on general measures of cognitive function, on a subset of memory measures, and on letter but not category fluency. Secondary analysis showed that earliest detectable ADAS-cog and MMSE decline in AD was at 16 months, while several specific neuropsychological indices were sensitive as early as 8 months (graded naming test, semantic naming, and the category/letter fluency tests). In combination, baseline/early changes in cognitive performance, alongside clinical measures, predicted 48% of disease progression over two years in memory impaired patients as a whole. These findings have implications for identifying patients likely to benefit from disease modifying agents, and for designing, powering, enriching, and implementing future clinical trials. Follow-up studies in independent populations are needed to validate predictive algorithms identified.

[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for IOS Press
Loading ...
Support Center