SAN DIEGO—"Those individuals who ultimately develop dementia or Alzheimer's disease presumably pass through a transitional stage," said Ronald Petersen, MD, PhD, at the 52nd Annual Meeting of the American Academy of Neurology. In this transitional phase, he continued, there may be some overlap between normal cognition and mild cognitive impairment, indicating that some of the individuals in the normal part of the cognitive spectrum may be at risk for ultimately developing cognitive decline. But can this cognitive decline be predicted? According to Dr. Petersen and researchers from the Mayo Clinic, Rochester, Minnesota, a combination of measures can separate the effects of normal aging from preclinical dementia. Their research indicated that age, an index of learning, delayed recall, a higher-order delayed recall term, and a global deterioration measure were all accurate indicators of potential cognitive decline, as was a patient's subjective perception of his or her own cognitive ability.

PATIENT PROFILE

To determine which variables in normal subjects may predict their subsequent decline to either mild cognitive impairment or Alzheimer's disease, a study cohort was drawn from the Mayo Clinic Alzheimer's Disease Research Center/Alzheimer's Disease Patient Registry. Of the 500 control subjects enrolled in a longitudinal study of aging and dementia, a subgroup of 54 subjects was identified. These subjects (mean age at study entry, 82; education, 13.2 years; Mini-Mental State Examination [MMSE] score, 28.1) were originally recruited as cognitively normal but subsequently declined to mild cognitive impairment or dementia over a median of 4.5 years of follow-up.

The researchers plotted the patients' conversions from normal to mild cognitive impairment or Alzheimer's disease. "At the outset of the study, all patients had normal cognitive function," said Dr. Petersen.

MEASURES ARE THE VARIABLES

Data collected on these individuals included demographic variables, a variety of cognitive rating scales (verbal intelligence, naming test, fluency) and memory measures (learning and delayed recall assessments from the Free and Cued Selective Reminding Test [FCSRT] and the Auditory Verbal Learning Test [AVLT]). "Out of this set of variables, a subset fell out as being significant as univariate predictors of change over time." Age and scores on the MMSE, Dementia Rating Scale, Boston Naming Test, Global Deterioration Scale (GDS), and learning and delayed recall measures on both memory instruments were all significant in univariate analysis, Dr. Petersen reported. A multivariate prediction model further revealed that age at time of enrollment, GDS score, and performance on the learning measure derived from FCSRT and the delayed recall measure of the AVLT over a 30-minute delay interval were significant predictors of cognitive decline.

According to their scores on the GDS, the patients were put into two groups. The GDS1 subjects were normal, had no complaints, and were doing well. Patients in the GDS2 group were also normal but had a subjective impression about their cognitive impairment. "They felt their memory may not be as good as it used to be," explained Dr. Petersen. "To our surprise," he continued, "this did turn out to be a useful predictor in who might subsequently decline. So apparently these people were telling us something about their subjective impression of their cognitive function."

Two instruments were used for the memory measures: a learning measure from the FCSRT and a delayed recall measure from the AVLT. The tests were conducted separately so as not to interfere with each other.

The learning measure, derived from the FCSRT, was a multi-trial free-recall test using a word list of 16 titles. The words were presented alone and also with a taxonomic category. For example: if the word to be remembered was "sweater" it was also presented as an article of clothing. If the word was "banana" it was also presented as "fruit." If the patient did not free-recall an item, the taxonomic category was given as a cue. "This measure is an index of the subject's ability to benefit from these cues. The data showed that most individuals who were able to benefit from the cues declined less rapidly than those who were unable to benefit from the cues," said Dr. Petersen. And this may also have an implication for the anatomic distribution of the disease, he said.

The same pattern of delayed recall is seen with the under-80 age-group as with the over-80 age-group, "although the delayed recall is certainly more dramatic in the older subjects," according to Dr. Petersen. Interestingly, APOE*E4 carrier status was not significant, nor was a positive family history of dementia. This may relate to the age of the subjects but was not significant overall, he said.

"So, from that we concluded that cognitive variables in normal subjects may be used as predictors of subsequent cognitive decline. An index of learning and an index of delayed recall may be useful; and age and subject impression of cognition can also be used to predict decline further down the road," he said. Identifying those at risk for cognitive decline may be useful in designing primary prevention trials, he suggested.

—Heidi Moore

Suggested Reading
Petersen RC, Smith GE, Waring SC, et al. Mild cognitive impairment: clinical characterization and outcome. Arch Neurol. 1999;56:303-308.
Jack CR Jr, Petersen RC, Xu YC, et al. Prediction of AD with MRI-based hippocampal volume in mild cognitive impairment. Neurology. 1999;52:1397-1403.

Return to table of contents