Issues in memory assessment of the elderly

The pattern of memory test performance by elderly individuals demonstrates significant overlap between populations. Consequently, the memory tests themselves poorly differentiate normal aging, dementia, and depression. The additional problem of relatively little adequate normative information for elderly individuals compounds this problem. Thus, inferences regarding the nature of assessed deficits should be based in part on information that extends beyond the traditional quantitative numerical analyses of test performances. For example, depressed individuals tend to make fewer false-positive responses and frequently will respond stating "I don't know." Similarly, if there is evidence for risk of memory impairment, the threshold for inferring assessed memory deficit may be altered. Finally, when independent information is present to confirm the presence of structural change in the brain, the inference of specific memory impairments can be strengthened. One not only needs to look at the entire pattern of neuropsychological deficits, but also use the neuropsychological findings in the context of other clinical findings. Perhaps the greatest utility of neuropsychological memory assessment of the elderly is providing standardized assessments for repeat evaluations. By repeating the assessment, deterioration as a function of a progressive dementia can be assessed. Alternatively, the effect of treatment for depression can be monitored and an improvement in memory test scores would be expected. Throughout, however, the careful clinician will remember that the inferences regarding an individual's level of function are not stronger than the reliability, representativeness, and validity of the instruments employed.