Dementia Insights: The Role of Cognitive Fluctuations and Impaired Episodic Memory in the Diagnosis of Dementia: New Considerations

In this month's “Dementia Insights,” I will bring attention to a few recently published clinical observations that play an important role in evaluating patients with suspected cognitive/ dementia disorders.

The first is the evaluation and role of cognitive fluctuations (CF) in the neurological history, and what they mean in the differential diagnosis of cognitive disorders. Traditionally CF is considered a hallmark in the diagnosis of Lewy Body dementia (LBD). Now it has been shown to occur in the normal population and Alzheimer's disease. Second is that impaired episodic memory (memory storage), which traditionally has been the mainstay in the diagnosis of amnestic mild cognitive impairment and Alzheimer's disease, now is shown to occur in the early diagnosis of the behavioral variant of frontal temporal dementia (bvFTD).

Cognitive Fluctuations
Escandon1 studied the role of cognitive fluctuations and what it means on cognitive testing in normal aging and Alzheimer's disease. Since cognitive fluctuations are a mainstay in the diagnosis of LBD, patients with this diagnosis were excluded from the study. The term cognitive fluctuation is usually defined as spontaneous alterations in cognition, attention, and arousal. As neurologists, when we identify cognitive fluctuation in our suspected dementia patients, we should have already ruled out a seizure disorder, structural brain disorders (tumors, hydrocephalus, etc.) sleep apnea and other sleep disorders, suspected hypoglycemia and side effects, and complications of various medications.

There were 511 participants enrolled in a longterm study of memory and aging at Washington University. They were interviewed by neurologists with the participation and knowledge of a collateral source at the initial visit and follow-up visit. The study used the dementia rating scale (DRS) to determine the presence or absence of dementia. This scale evaluates all aspects of cognitive function and activities of daily living.

This study also included detailed neuropsychological testing. It used the Mayo Clinic Cognitive Fluctuation questionnaire, which was given to the individual that accompanied the patient. That person had to give yes/no answers to four questions. Three or more “yes” answers were required for a diagnosis of cognitive fluctuation. The four questions assessed:

1.) Drowsiness and lethargy all the time or several times a day despite getting enough sleep the night before.

2.) Daytime sleep for two or more hours before 7 p.m.

3.) Times when patient's flow of ideas seem disorganized, unclear or not logical,

4.) Staring into space for long periods of time.

A sleep questionnaire was also used; it asked the informant to rate the level of alertness for the past three months on a scale from sleeping all day to being fully awake. There was also a question asked about acting out dreams to detect the presence of rapid eye movement (REM) sleep behavioral disorder. In this study, there was no correlation or relationship between the REM behavior question and cognitive fluctuations, supporting the fact that the patient population in this study was likely free of unrecognized Lewy Body dementia.

The dementia rating scale scores of the participants in the study were as follows: 295 participants had a CDR of 0 (no dementia), 145 participants had a CDR of 0.5 (mild dementia), and 71 participants had a CDR of 1 (moderate dementia). Cognitive fluctuations were seen in all the CDR levels, but were much greater in the 0.5 and 1.0 CDR levels. In CDR 0, or normals group, 0.7 percent had CF. In the 0.5 CDR level 8.5 percent had CF. And in the CDR 1 group, 18.5 percent had cognitive fluctuations.

The study found no relationship between those with cognitive fluctuations and the answers to the sleep questionnaire. However, there was direct correlation between the alertness of the participant and the level of dementia. The more alert the participant, the less dementia he/she appeared to have.

The results of this study revealed some very useful clinical information:

1.) Cognitive fluctuations can occur in the normal population but are infrequent.

2.) Twelve to 15 percent of patients with CF have Alzheimer's disease. Of note, 70 percent of the remaining cases had dementia, but the cause was not clear at the time of the study.

Long-term follow-up of these cases may reveal even more Alzheimer's disease.

3.) An established history of definite CF in a patient increases the likely diagnosis of dementia by eight fold compared to no CF history.

4.) Neurologists need to take more detailed histories of suspected CF. With an incomplete history, we may be accepting CF as representing memory failure or the idea of “good and bad days.” Using the Mayo Clinic Cognitive Fluctuation questionnaire or something similar takes a very short time, and it can be filled out by the caregiver or informant while the patient is being examined. Remember, the possible diagnosis of dementia with Lewy Bodies includes CF, visual hallucinations, and mild Parkinson's features. But definite CF and visual hallucinations can also be seen in early Alzheimer's disease.

The biological basis for CF is not well understood. Loss of attention control occurs in both LBD and Alzheimer's disease. Ascending cholinergic pathways from pedunculopontine and lateral dorsal tegmental nuclei are involved in arousal. Patients with LBD have fewer of these pathways than normal individuals. This has not been studied in Alzheimer's.

Memory Storage
Traditional teaching tells us that memory storage and episodic memory are usually normal in early behavioral variant of FTD and can be used reliably to differentiate bvFTD from Alzheimer's disease. Hombergel2 reported on a study of patients with a clinical diagnosis of bvFTD, with a threeyear follow-up. All participants met the clinical diagnosis criteria for bvFTD. An informant substantiated personality and social behavioral changes.

There were 59 subjects classified into progressive and non-progressive frontal dementia over a three-year period based on cognitive scores and activities of daily living assessments. Of those, 19 died. Neuropathology was available for 10. Seven had frontal temporal dementia with storage of tau, and three had frontal temporal dementia of ubiquitin storage.

The classification of bvFTD into progressive and non-progressive categories is a new development. ³ Some patients show rapid progression, while others may show little or no progression for 10 years. Rapid progression and slow progression can be separated on MRI and PET scans despite the fact that both forms have equivalent behavioral abnormalities.⁴ Some recent studies have suggested that the rapidly progressive form of bvFTD has orbital medial frontal lobe atrophy on MRI scan and hypo-metabolism on PET scan.⁵ The nonprogressive form did not show MRI or PET abnormalities.

The patient groups did not differ on the CDR score. The most striking finding of the study is that the degree of episodic memory impairment and memory storage in patients with progressive bvFTD was similar to patients with Alzheimer's disease.

Previous research has suggested that patients with bvFTD have been able to perform better on recognition recall and not show the accelerated forgetting typical of Alzheimer's disease in delayed free recall.6 Verbal cueing was shown to improve word completion tasks in bvFTD, suggesting that effective retrieval strategies are the key. Memory storage was considered to be normal. This does not appear to be the case in many patients with progressive bvFTD.

This study suggests that we cannot rely on free recall after a delay to help separate bvFTD from Alzheimer's disease. Impaired episodic memory does not exclude a diagnosis of bvFTD. The best indicator for bvFTD is behavioral disturbance, which can be a problem if a reliable informant is not always available.

This study also showed that orientation scores of time and place are usually preserved in bvFTD as compared to Alzheimer's disease. To support this memory storage finding, some pathologic studies have reported hippocampal atrophy in bvFTD even in the early course of the disease and hypo-metabolism in medial temporal lobe on PET as well. Recall-based tasks have also been shown to place strong functional demands on the prefrontal cortex and could help account for the observed memory problem.

As already mentioned, not all cases of bvFTD progress over three years, and they are not easily distinguishable. Persistent and consistent low scores of memory on delayed recall with a behavioral disturbance suggest the progressive form of bvFTD. Mild memory impairment that is very subtle or patchy in the presence of a behavioral disorder is more in favor of the non-progressive form of bvFTD.

Reconsidering “Diagnostic” Characteristics
These two studies suggest that the presence of cognitive fluctuations and abnormalities in episodic memory and delayed recall can occur in several dementias. They should not be considered diagnostic when comparing Alzheimer's disease, Lewy Body dementia, and bvFTD.

Ronald Devere, MD is Director of the Taste & Smell Disorders Clinic and Alzheimer Disease & Memory Disorders Center in Austin, Texas.

1. Escandon, A., Alhamadi Noor, Galvin, J.E., Effect of Cognitive Fluctuations On Neuropsychological Evaluation in Aging and Dementia, Neurology, Vol. 74, January 19, 2010, pp. 210-217.
2. Hornberger, M., Piguet, O., Graham, A.J., Nestor, P.J., and Hooges, J.R., How Preserved Is Episodic Memory In Behavior Variant Frontal Temporal Dementia? Neurology, Vol. 74, February 9, 2010, pp.472-79.
3. Hornberger, M., Shelly, B.P., Kipps, C.M., Can Progressive and Non Progressive Behavior Variant Frontal Temporal Dementia Be Distinguished On Presentation? JNNP, E Pub. 2009.
4. Hodges, J., Editor, Frontal Temporal Dementia Syndromes, Cambridge: Cambridge University Press, 2007.
5. Kipps, C.M., Hodges, J.R., Fryer, T.D., Nestor, P.J., Combined MR Imaging and PET Brain Imaging In Behavior Variant Frontal Temporal Dementia: Defining The Clinical Type, Brain, Vol. 132, 2009, pp. 2566-78.
6. Perry, R.J., Hodges, J.R., Differentiating Frontal Temporal Variant Of Frontal Temporal Dementia from Alzheimer's Disease, Neurology, Vol. 54, 2000, pp. 2277-84.