Mild cognitive impairment in the older population: Who is missed and does it matter?
✍ Scribed by Blossom C. M. Stephan; Carol Brayne; Ian G. McKeith; John Bond; Fiona E. Matthews
- Publisher
- John Wiley and Sons
- Year
- 2008
- Tongue
- English
- Weight
- 127 KB
- Volume
- 23
- Category
- Article
- ISSN
- 0885-6230
- DOI
- 10.1002/gps.2013
No coin nor oath required. For personal study only.
✦ Synopsis
Abstract
Objectives
Classifications of mild cognitive impairment (MCI) vary in the precision of the defining criteria. Their value in clinical settings is different from population settings. This difference depending on setting is to be expected, but must be well understood if population screening for dementia and pre‐dementia states is to be considered. Of importance is the impact of missed diagnosis. The magnitude of missed ‘at‐risk’ cases in the application of different MCI criteria in the population is unknown.
Methods
Data were from the Medical Research Council Cognitive Function and Ageing Study, a large population based study of older aged individuals in the UK. Prevalence and two‐year progression to dementia in individuals whose impairment failed to fulfil published criteria for MCI was evaluated.
Results
Prevalence estimates of individuals not classified from current MCI definitions were extremely variable (range 2.5–41.0%). Rates of progression to dementia in these non‐classified groups were also very variable (3.7–30.0%), reflecting heterogeneity in MCI classification requirements.
Conclusions
Narrow definitions of MCI developed for clinical settings when applied in the population result in a large proportion of individuals who progress to dementia being excluded from MCI classifications. More broadly defined criteria would be better for selection of individuals at risk of dementia in population settings, but at the possibility of high false positive rates. While exclusion may be a good thing in the population since most people are presumably ‘normal’, over‐inclusion is more likely to be harmful. Further work needs to investigate the best classification system for application in the population. Copyright © 2008 John Wiley & Sons, Ltd.