Psychologists find more sensitive tests for predicting Alzheimer's as well as changes in...( ster. The researchers also gave a Visu...),Health

On the word-pair memory test, people destined to develop Alzheimer's disease didn't do any better when words were related than when they weren't. The authors think these participants may already have lost key knowledge of word attributes that normally help people to more easily remember words by means of their semantic associations. Sometimes, at-risk participants reported a vague sense that one word had something to do with another, but they couldn't say exactly what. The authors suspect they couldn't encode the word pairs at a sufficiently deep level because they'd lost the semantic knowledge that stays intact in normally aging people.

On the word-reading test, word repetition (to measure priming) didn't help high-risk participants to perform better, a sign that they weren't learning implicitly as well as the people who would stay healthy. The authors speculate that because high-risk participants drew less benefit from word repetition, they did not encode the words properly.

These tests remained sensitive to the risk of developing Alzheimer's disease even within a more homogeneous subset of the broader study population, people with mild cognitive impairment. For both the whole and subset study groups, these tests predicted future Alzheimer's diagnosis as much as two years early.

Equally important, the popular Mini Mental Status Exam (MMSE), a test mainly sensitive to episodic memory, was not as good a predictor. Although clinicians use it for quick, easy-to-administer screening, the authors found it to be "less predictive [than the tests sensitive to semantic and implicit memory]. These [MMSE and other purely episodic memory] tasks may only differentiate between pathological and normal aging when dementia has
'"/>

Contact: Pam Willenz
pwillenz@apa.org
American Psychological Association
25-Sep-2005