Five-year follow-up of cognitive impairment with no dementia
Tuokko H, Frerichs R, Graham J, Rockwood K, Kristjansson B, Fisk J, Bergman H, et al.;
Commented by , 28 Jul 2003
Aim of the study
To investigate natural history of elderly persons with no cognitive impairment (NCI) versus persons identified with cognitive impairment but no dementia (CIND).
Method
Five-year longitudinal follow-up of all persons without dementia examined during the first phase of the Canadian Study of Health and Aging (CSHA) in 1991, looking at mortality, institutionalization and clinical diagnosis using the DSM III criteria for dementia.
The original sample was population-based across Canada, involving 10,263 persons over age 65, living at home or in institutions. The informant interview was based on the Cambridge Examination for Mental Disorders (CAMDEX) – section H.
Results
From a sample of 883 persons with NCI and 801 persons with CIND, 517 with NCI (59%) and 327 with CIND (41%) were alive and received clinical diagnosis at follow-up.
Persons with CIND were more likely that those with NCI to die (49% vs 30%; odds ratio 2.3; 95% CI 1.9 – 2.8), to be institutionalized (29% vs 14%; odds ratio 2.5; 95% CI 1.9 – 3.3), or to be diagnosed with dementia (47% vs 15%; odds ratio 5.3; 95% CI 3.8 – 7.4).
These higher rates of developing dementia were similar for all sub-categories of CIND, including depression and vascular. Those found to have dementia at follow-up were more likely to have had impaired memory, informant-reported changes in memory, and functional impairment at baseline.
Discussion
These results are similar to findings from other population-based studies, were approximately 15% of older persons with no cognitive impairment progress to dementia over 5 years, versus approximately 50% of older persons with cognitive impairment but not demented.
The closest study in terms of operational definitions is the Kungsholmen Project, were similar proportions of CIND sub-types progressed to dementia and death over 3 years (Palmer et al. Am J Psychiatry 2002; 159; 436-442).
These epidemiological studies are as important as observations made in specialized clinics on a select group of persons with CIND, eg those with amnestic Mild Cognitive Impairment (MCI). As was discussed by Alistair Burns and Michael Zauding in their review on MCI (The Lancet 2002; 360; 1963-1965), the concept of MCI need to be broadened to include not only memory impairment but also other types of cognitive impairments.
Charles DeCarli in a similar review (The Lancet Neurology 2003; 2; 15-21) expanded MCI to include cerebrovascular disease. It is very likely that new definitions of MCI will also include some neuropsychiatric symptoms (Lyketsos et al, JAMA 2002; 288; 1475-1483). Perhaps the term Mild Cognitive Impairment will have to be changed in order to reflect the broad spectrum of prodromal symptoms of dementia.
Another field of study relevant to prodromal stages of dementia is that of biological markers. Two recent articles are worth noting: Frank et al (Neurobiol Aging 2003; 24; 521-536) offer a comprehensive review of individual blood and CSF markers in Alzheimer’s disease, whereas Teunissen et al (Neurobiol Aging 2003; 5822; 1-10, in press) explore the feasibility of combining different serum markers to increase sensitivity and specificity.
It is likely that a combination of clinical and biological (including genetic) markers will be required to reach reliable levels of certainty in the diagnosis of dementia such as AD in its prodromal stages, as new disease-modification therapies emerge.