Dementia: Update for the Practitioner

Neuropsychology of Mild Cognitive Impairment,
Alzheimer's Disease, Dementia with Lewy Bodies,
and Frontotemporal Dementia
Penne Sims, Ph.D.

Screening Measures

In 2001 the American Academy of Neurology published practice guidelines for early detection of dementia, and these include recommendations of screening measures for cognitive functioning. In addition to neuropsychological testing, they recommended a number of commonly used tests.


	Mini Mental State Exam
		Orientation
	(5)	What is the year, season, date, month?
	(5)	Where are we: state, country, town or city, hospital, floor?
		Registration
	(3)	Repeat the following: apple, table, penny.
		Attention and Calculation
	(5)	Serial 7s: 93, 86, 79, 72, 65 or spell world backwards.
		Recall
	(3)	What were the three objects: apple, table, penny?
		Language
	(2)	Name a pencil and watch.
	(1)	Repeat the following: No ifs, ands, or buts
	(3)	Three-step command: Take this piece of paper in your right hand, fold it in half, and put it on the floor.
	(1)	Read and obey: Close your eyes.
	(1)	Write a sentence.
	(1)	Copy the design:
	(30)

	The Folstein Mini Mental State Exam is a thirty-item test used to evaluate patients for cognitive problems and dementia. Courtesy of Dr. Karen Bell

The Folstein Mini Mental exam is one of the most common measures and is used when evaluating a patient who has functional complaints. The Folstein is a thirty-item test that takes about five minutes to give. We generally look for scores below 25 out of 30 to be considered impaired, but this can be misleading for many patients depending on age and level of education. For example, a patient who is 70 years old with an eighth-grade education who scores a 24 out of 30 may be less of a concern than someone age 70 with an M.D. or Ph.D. who also scores 24. It is important to utilize the age and education-corrected norms when using the MMSE. On average, patients with dementia decline three points per year on the MMSE.

Many studies have found that people with cognitive impairment perform on neuropsychological tests much like people who are normal or have no complaints at all. Neuropsychological testing using the MMSE is not necessarily as useful in discriminating normal patients from those you think may have MCI.

There are limitations with Folstein MMSE. It is heavily weighted toward verbal items, and those tend to be very simple. Also, if you have monitored a patient fewer than three years, any changes you see in the score may be somewhat misleading because there tends to be a large measurement error. There is a substantial variation in change over a short period of time depending on how the patient is feeling—maybe the weather was bad, maybe the examiner was tired, maybe the patient was tired. Without extended monitoring, you will not know whether a three- or four-point change is as significant as it seems.

Other cognitive domains like distractibility, abstraction or executive functioning, and visual-spatial skills may be limited or completely omitted from the Folstein. There have been some modifications to the test, and one of those is a more comprehensive 57-item measure that we have used at Columbia. One of the more common overall measures of cognitive functioning in use is the Mattis Dementia Rating Scale. It is much more comprehensive and is better than the Folstein at picking up some of the cognitive domains omitted from the MMSE. The Brief Focus Cognitive Assessment includes tests such as clock-drawing, telling time, or making change, but these are limited in their utility because they only assess a very specific cognitive skill.

The structured informant interview is extremely important in terms of providing information on the history and nature of the difficulties and change or decline over time. Neuropsychological testing is a much more comprehensive measurement of cognitive functioning than the tests discussed above. As such, it is important to identify what constitutes a deficit and what signifies a decline in neuropsychological performance. As before, the standard cutoff is 1.5 standard deviations below the mean, and some people require 2 standard deviations to consider a deficit significant. Here the deficit is based on a comparison of the patient's performance with normative data that tend to be stratified by age, education, and sometimes by sex.

As with any test, there are going to be limitations. Sometimes the normative sample is not appropriate for very low- or very high-functioning individuals. In these cases, we prefer to make a within-subject comparison, by which I mean we have to compare the individual to himself or herself. One way to do this is to estimate the individual's premorbid level of functioning by assessing his or her level of education, level of occupational functioning, and/or performance on tasks like vocabulary or reading that can be relatively resistant to aging effects. Alternatively, we can look at two or more consecutive evaluations of the individual to identify more specific changes in cognitive functioning over time.

PREVIOUS | NEXT