Progression of mild cognitive impairment to dementia: a challenge to current thinking
Busse A, Angermeyer MC, Riedel-Heller SG;
Commented by , 24 Nov 2006
Background
Current literature suggests a conversion rate of mild cognitive impairment (MCI) to dementia of 10–15% per year. Better understanding of this process is desirable for primary and secondary prevention of all dementias, particularly Alzheimer’s Disease.
Aim
This study sought to examine the time-dependent evolution from MCI to dementia in a community sample.
Method
Sample - subjects were derived from the Leipzig Longitudinal Study of the Aged. 1500 people aged 75+ were identified by random sampling, and a further 192 living in homes for the elderly were included on a proportional basis. Clinical interviews, including informant interview and neuropsychological assessment were conducted: 17% were found to have DSM-IV dementia and excluded.
The remaining subjects (1045) were then considered for this study, with a further 65 excluded due to Parkinson’s Disease, learning disability or brain cancer: 980 were therefore included at baseline.
Follow up – repeat assessments occurred at 1½ years, 3 years, 4½ years and 6 years after baseline assessment.
Results
Of the 980 without dementia at baseline, 75% were women, with mean age 81.5 years. Mean MMSE score was 27. 9–41% could be classified as having mild cognitive impairment, depending on criteria used.
After baseline examination, a further 117 dropped out. Follow up data was collected on 863 participants for an average of 4.3 years (standard deviation 1.94). 171 (19.8%) developed dementia during this period and 195 died.
Whichever diagnostic cut-off was used for MCI, the proportion of participants who developed dementia was approximately double during the first 18 months of observation, compared to later follow up; approximately 20% converted to dementia in the first 18 months. In total, 60–65% of included subjects had developed dementia at the time of death.
Dr Seymour's comments
The received wisdom in the literature is of a conversion rate of 10–15% per year from MCI to dementia, with a linear progression to dementia over time. The headline finding of this study challenges this assumption of linear progression as the conversion rate was highest in the first 18 months of follow up. A recent Dutch study suggested conversion rates were also dependent on age, with risk of developing dementia from MCI increasing with age.
All studies of MCI are compromised because of the lack of internationally agreed nosology. As the authors acknowledge, long term studies in subjects of this age are also prone to selection bias by attrition/high drop-out rates. Another potential bias is that increasing cognitive impairment, often not labelled "dementia", is common in the weeks and months before death.
MCI is clearly a heterogenous group of conditions: some people improve cognitively, some remain the same, and some go on to develop dementia. The group who improve are likely to have a co-existing medical or psychiatric problems effecting cognition that subsequently improve, whereas the group that go on to develop dementia are likely to be in a prodromal phase of dementia when they are diagnosed with MCI.
It is thus logical that the latter group convert to dementia sooner rather than later.
Needless to say, more research is needed to help clinicians predict which patients with MCI are going to convert to dementia. If deficits are remaining static after 2–3 years, long term follow up may not be justified.