CHICAGO—The best predictor of Alzheimer’s disease dementia may be a verbal memory test, according to psychologists Mark Boulos, BSc, and Konstantine K. Zakzanis, PhD, of the University of Toronto. They presented their research at the 110th Annual Convention of the American Psychological Association.

In a meta-analysis of 31 studies that included 1,144 patients with Alzheimer’s disease and 6,046 healthy controls, the researchers reviewed both neuropsychologic and neuroimaging tests to determine their accuracy in detecting preclinical dementia and Alzheimer’s disease. They also paired a genetic susceptibility to dementia and Alzheimer’s disease with results on both neuropsychologic and neuroanatomic tests in an attempt to identify which tests would prove most accurate in identifying preclinical disease in patients with the apolipoprotein E (APOE) gene.

MAGNITUDE OF MEMORY LOSS

Their findings support the use of the California Verbal Learning Test (long-delay recall, percent retained) as the best predictor of Alzheimer’s dementia, with executive function–type measures and neuroimaging techniques not being good predictors. Other long- and short-delay verbal memory tests were also good predictors of dementia. Although these findings were based on the results of only one study, Mr. Boulos maintained that they are “very encouraging” due to the weight of the effect size, which was large enough to discriminate between nearly all patients who developed Alzheimer’s dementia and those who did not.

In presenting their findings, the investigators noted that a decline in memory, especially in verbal episodic memory, can be observed in healthy elderly as well as in elderly individuals with mild cognitive impairments and that most people with this deficit will not be affected by Alzheimer’s disease. What is the difference between those elderly with normal memory impairments and those with preclinical Alzheimer’s dementia? The answer, according to the researchers, lies in the magnitude of the memory deficit. Because their study was a meta-analysis of effect sizes, the results allowed them to compare the size of the difference between normal older adults with normal memory decline and older adults with actual dementia.

Although memory assessments were also more predictive of the APOE epsilon4 allele than neuroimaging tests, the most predictive test for the APOE epsilon4 allele still appears to be an odor identification test.

MEASURE FOR MEASURE

The researchers also found a significant relationship between results on the Wechsler Adult Intelligence Scale (WAIS) digit span–backward test and percentage of female patients who converted to dementia, suggesting that as women who converted to dementia got older, their performance on the WAIS digit span test declined. Previous work by Dr. Zakzanis found that duration of disease is tied to the magnitude of effect once the disease was diagnosed. Also of note in this study was the high test sensitivity of the conceptualization subscore of the Dementia Rating Scale—a non-memory measure with a test sensitivity greater than 90%.

Changes in the hippocampus were the best volumetric or neuroimaging measure—which Mr. Boulos said “makes sense and is a plausible finding”—but in general volumetric measures were less sensitive to preclinical stages of the dementia than were the neuropsychologic tests. Although neuroimaging was not an important indicator of conversion to dementia, Mr. Boulos recognizes its applications.

The researchers emphasized that their work underscores the importance of early testing so that those predisposed to Alzheimer’s disease can attempt to maintain a higher quality of life. “As an unobtrusive measure with such high sensitivity, it will be important for future researchers to determine the reliability of this test [California Verbal Learning Test] sensitivity in future studies employing the measure,” they reported.

NR

Return to table of contents