One of the things revealed in Spaan's research was that elderly persons who were found to have dementia two years later, were scarcely better at remembering word pairs clearly linked in meaning (for example, pipe - cigar) than word pairs without such a link (for example nail - butter). However, elderly persons who did not have dementia two years later normally benefited from such a link in meaning when remembering word pairs. Spaan concluded that the memory problems of elderly people in an early stage of dementia could be attributed to a disrupted semantic processing.
In addition it transpired that a test for the unconscious (implicit) recall of previously presented words significantly improved the prediction of dementia: elderly persons in the early stage of dementia did not benefit from the repeated presentation of words.
With these data it is possible to distinguish 'normal' age-related memory problems that generally occur in the elderly from memory problems typical of the early stages of dementia. This is particularly important, as the present generation of drugs aimed at inhibiting the course of dementia are only useful if administered in the earliest stages of the disease.
A large group of elderly people who lived (semi-) independently, were subjected by Spaan to a wide range of memory tests developed at the University of Amsterdam. These elderly persons participated in the population study Longitudinal Aging Study Amsterdam from the Free University of Amsterdam. The tests were conducted on two occasions with an interval of two years.
Using the official diagnostic criteria, nobody in the group was found to have dementia during the first assessment. During the repeat assessment two years later, it was established who had developed dementia during the intervening period. Using data from the first assessment, Spaan compared the memory test performance of the elderly persons who later developed dementia with those who did not.
Spaan argues that the current methods used to detect dementia concentrate too much on the so-called episodic memory. That is the conscious memory of personal and specific information related to time and place. The researcher recommends the use of memory tests that in addition to the episodic memory also assume a semantic memory (concerning general knowledge and information, including our vocabulary) and an implicit memory (which includes various unconscious learning processes). In this manner the current test material could be improved in order to recognise dementia at an earlier stage.
Spaan's research was part of the Netherlands Organisation for Scientific Research programme 'Memory processes and dementia'.
For further information please contact Dr Pauline Spaan (Department of Psychonomics, University of Amsterdam), tel. 31-205-256-644, fax 31-206-391-656, e-mail: firstname.lastname@example.org. The doctoral thesis was defended on 30 January 2003. Ms Spaan's supervisors were Prof. J.G.W. Raaijmakers and Prof. C. Jonker.
The research was funded by the Netherlands Organisation for Scientific Research (NWO).