In the current research, investigators utilized the knowledge that subtle cognitive changes may precede disease onset by as much as 7 to 10 years to hypothesize that changes in cognitive performance indicative of high risk for disease might be found on neuropsychological testing well before the first clinically apparent symptoms are seen.

The American team evaluated neuropsychological test data from 40 people with average age 75 years who were enrolled in a long-term study on Alzheimer's disease. Mark Jacobson, Ph.D., and colleagues compared the cognitive performance of two groups: the preclinical group, 20 people who were diagnosed with possible Alzheimer's disease within a year or two of testing and the control group, 20 people who remained symptom-free for several years after testing.

The researchers examined the comparative performance of the two groups, all of whom were clinically normal at the time of testing, on a variety of neuropsychological tests. The most revealing data emerged from tests in which people had to name common items from picture and those in which they had to use blocks to copy printed block designs (visuoconstruction). These skills tests have been proved useful in following Alzheimer's disease because they reflect performance of cerebral regions that are among the earliest to deteriorate.

The data supported the team's hypothesis that testing might reveal cognitive performance deficits not yet apparent clinically. The preclinical group demonstrated significantly greater discrepancies between naming ability and visuoconstruction ability than did the control group. In addition, the preclinical group had a significantly higher-than-expected proportion of these asymmetric profiles than were found in a large normative group: The preclinical subjects clearly had discrepancies in verbal-versus-visuoconstruction performance that were greater in magnitude and more common than those seen in control subjects.

The authors suggest that this kind of simple, noninvasive testing could be of particular value for populations known to be at risk because of family history of genotype. The key, according to lead author Jacobson, is to test and track performance abilities over time, looking not only for change in any one test but in the level of one test relative to other skills tests.

"Our analyses," he continues, "show that although Alzheimer's disease is typically marked by a global cognitive decline, some individuals in a preclinical stage may show a subtle change in one area of cognition prior to this global decline. As a result, in addition to looking at cognitive deficits, we might also look for large differences in one cognitive domain relative to another -- especially if this difference represents a change from earlier patterns of functioning."

The authors note that genotype has consistently been shown to be a more accurate predictor of risk than asymmetry in testing or brain imaging. However, they assert that genetic testing may be more indicative of incidence rate than timing of onset of disease. They hope that research will better define cognitive asymmetry and raise the predictive power of cognitive profiling such that false positive results are minimized.

Jacobson and his colleagues are now studying elderly people on a longitudinal basis to see whether relative differences in cognition between preclinical and control, non-disease groups are always present or represent actual changes in cognition before clinical onset of disease. They also hope that research will better define cognitive asymmetry and raise the predictive power of cognitive profiling in order to minimize false positive test results.