Subscribe to eTOC


Honolulu, HI — Cognitive tests that measure learning and recall can predict which aging patients will go on to develop mild cognitive impairment and which are liable to have Alzheimer disease, researchers reported here at the AAN Annual Meeting.

But until effective treatment is available for dementia, it is too soon to administer the neuropsychological exams to patients, they said.

“What we found is that, if you're down a little in your ability to learn new material or to recall that material, that is an ominous, qualitative predictor of future development of dementia,” said Ronald C. Petersen, MD, Director of the Alzheimer's Disease Research Center at the Mayo Clinic in Rochester, MN.


Dr. Ronald C. Petersen

As potential new therapeutic targets are found, the field is moving toward earlier identification not only of patients with symptoms of dementia, but also of people who are at risk for mild cognitive impairment but have not yet shown any definite signs of cognitive impairment, Dr. Petersen said.

“If we can identify patients who might be at risk – those who are asymptomatic or those who have mild cognitive impairment, the precursor of Alzheimer disease – we can help them before they lose their cognitive function [once drugs are developed],” he said.


To identify factors that would predict conversion to dementia, the researchers turned to the Mayo Alzheimer's Disease Patient Registry, a longitudinal prospective study of aging and dementia funded by the National Institute on Aging (NIA) since 1986.

The subjects, all of whom were cognitively normal on enrollment, were given neuropsychological tests annually. At each visit, physicians held a consensus conference to determine if the patients remained normal or had converted to mild cognitive impairment, Alzheimer disease, or another dementia.

Over the years, the researchers identified 874 subjects who were cognitively normal on enrollment and had at least one follow-up evaluation. The men and women, whose average age was 78.6 years, had an average education level of 13.2 years; almost all were white.

On two tests of overall cognitive function, the Mini-Mental State Exam (MMSE) and the Short Test of Mental Status (STMS), their average baseline scores were 28.1 and 34.0, both of which were in the normal range.

At a median follow-up of 4.8 years, 75 (8.6 percent) of the patients were diagnosed with mild cognitive impairment or Alzheimer disease. Another 50 patients developed other types of dementia, most of which had some element of vascular disease, Dr. Petersen said.

“We took the 75 patients with mild cognitive impairment or dementia and went back and looked at their day-one performance on the tests when they were normal, asking, ‘What is there about this pattern of performance that might predict conversion?’” he said.

About a dozen univariate variables were significant predictors of conversion: age; apolipoprotein E-4 (APOE4) carrier status; and scores on the MMSE, the Global Deterioration Scale (GDS), the STMS, the Dementia Rating Scale, the Auditory Verbal Learning Tests (AVLT) of learning and delayed recalls, and the Free and Cued Selective Reminding Test (FCSRT) of learning, delayed recall, and use of semantic cues.


When these univariate variables were entered into a multivariate Cox regression model, five independent factors emerged as significant: age, APOE4 carrier status, STMS score, AVLT-delayed recall score, and the FCSRT-use of semantic cues, Dr. Petersen said.

When the investigators used any dementia, rather than only mild cognitive impairment or Alzheimer disease, as the crossover event, the same set of univariate predictors were identified. The multivariate Cox regression model showed six variables to be significant: age, APOE4 carrier status, MMSE, GDS, AVLT-delayed recall, and FCSRT-learning.

“While the subjects were clinically normal at the time of entry into the study, certain demographic, genetic, and cognitive factors predicted who would likely progress to mild cognitive impairment, Alzheimer disease, or dementia years later,” Dr. Petersen said.

“The best cognitive predictors of mild cognitive impairment and Alzheimer disease appear to be those measures with an emphasis on learning and recall,” he said. “If you're down a little on these tests, it spoke to a worsening decline. So this gives us an idea that, even though these people look like they are clinically normal, they might have a greater tendency to develop these dementias.”


In a second study, patients with mild cognitive impairment who had the worst baseline memory and executive performance scores on neuropsychological tests were most likely to develop dementia over the next three years, reported Charles DeCarli, MD.

“It is now generally accepted that mild cognitive impairment is a transition phase between normal aging and dementia, particularly Alzheimer disease,” said Dr. DeCarli, Professor of Neurology and Associate Director of the Alzheimer's Disease Center at the University of California-Davis. “But there are still a lot of uncertainties. For example, there are many causes of mild cognitive impairment, including cerebrovascular disease. And not all patients progress to dementia. It is still unclear who will progress and how fast one will progress.”

The NIA-funded prospective study enrolled 52 patients, about three-fourths of whom were men, who had mild cognitive impairment but no signs of dementia.

“Over the next three years, we looked at a variety of factors that we thought could be predictors of dementia, including age, education, gender, baseline cortical gray matter, hippocampus volume, white matter hyperintensities, and memory and executive function on neuropsychological tests,” Dr. DeCarli said.


The study showed that 23 of 52 patients had lacunar infarction on MRI. But 11 of 29 (37.8 percent) patients without lacunes and 6 of 23 (26.1 percent) with lacunes progressed to dementia, “suggesting that evidence of cerebral brain injury on MRI is not a good predictor of dementia,” Dr. DeCarli said.

In a multivariate model that controlled for demographics and MRI variables, small hippocampus volume was significantly associated with increased risk for conversion.

But memory scores proved an even better predictor, Dr. DeCarli said. In a model that included demographic, MRI, and neuropsychological variables, better recall and executive performance were significantly associated with a reduced risk of developing dementia.

For example, every five-point increase in memory performance was associated with a 36-percent reduction in risk for dementia, while a five-point increase in executive function was associated with a 19-percent decreased risk.

“Neuropsychological measures of memory better capture the complex nature of the combined processes of Alzheimer disease and cerebrovascular disease than do the more specific MRI measures in patients with mild cognitive impairment,” Dr. DeCarli said.


Dr. Sanford H. Auerbach

“This suggests that if we had a really sensitive memory measure, we would be able to identify at-risk patients early on.”


A 25-year follow-up study of volunteers in the Framingham Study also showed that memory tests are a sensitive predictor of eventual dementia and Alzheimer disease, reported Sanford H. Auerbach, MD, Professor of Neurology at Boston University School of Medicine.

Scores on frontal deficit tests were not as helpful and APOE4 carrier status appeared to have no value as a predictor of subsequent dementia, he said.

From 1974 to 1976, the substudy enrolled 2,030 volunteers, 59 percent of whom were women. The average age was 67.9 years, with a range of 56 to 88 years. All the subjects were given neuropsychological tests, including three measures of memory and frontal system functions, at the time of enrollment.

At 10 years of follow-up, there were 93 cases of dementia, 62 attributable to Alzheimer disease. By 25 years, 271 patients had developed dementia, 197 of whom had Alzheimer disease.

Although several statistical models were used to analyze the data, “it was clear that mild memory deficits were more predictive than any others of the subsequent development of dementia and Alzheimer disease,” Dr. Auerbach said.

“The results suggest that standard neuropsychological memory tests, designed to detect mild cognitive impairment, may help predict the future risk of developing Alzheimer disease and are more reliable than subjective complaints,” he said.


The findings of all three studies are mutually consistent and supportive, said Dr. Petersen. “Our findings were mainly in a white population, while Dr. DeCarli's were in a more mixed racial group,” he said. “So, if anything, his research makes the findings all the more generalizable.”

Nevertheless, giving learning and recall tests to patients at the current time – before effective treatments are available – would only worry people needlessly, Dr. Petersen said.

But, eventually, he said, “memory tests will be an important tool for the early diagnosis of dementia, just like mammograms for breast cancer or colonoscopy for colon cancer.”