Revised Criteria for Mild Cognitive Impairment: Validation within a Longitudinal Population Study

Sylvaine Artero, Ronald Petersen, Jacques Touchon, Karen Ritchie
2006 Dementia and Geriatric Cognitive Disorders  
were combined. Conclusion: Incorporating the possibility of change in activity level and alteration of non-mnesic cognitive functions have been found to ameliorate the original algorithm and better define subjects converting to dementia. This definition may be applicable to both clinical and population research. The identification of cases of subclinical cognitive impairment at high risk of evolving towards dementia is important both for early treatment and protection of the individual. The
more » ... ept of mild cognitive impairment (MCI) [1] is now widely used to describe this high-risk group [2] [3] [4] [5] [6] , and numerous large-scale research programmes have now been undertaken with a view to providing treatment. As mild cases of cognitive disorder more commonly present initially to the general practitioner than to the specialist, it is important that diagnostic criteria perform adequately within this setting. Previous attempts to apply MCI criteria in general practice have been disappointing, the algorithm giving variable predictive validity with regard to the identification of dementia converters [7, 8] . A recent consensus conference held in Stockholm [9] provided an opportunity for clinicians and researchers working in the field of MCI to compare data sets and to propose modifications to the existing cri-Abstract Background: Mild cognitive impairment (MCI) refers to the transitional zone between normal ageing and dementia. Current criteria perform poorly within the general population setting. Revisions have been proposed based on results obtained from clinical and epidemiological studies. Objective: To evaluate revised diagnostic criteria for mild cognitive impairment (MCI-R) incorporating changes in activity level and non-mnesic cognitive functioning. Method: MCI-R subjects were recruited from a representative network of general practitioners in the south of France. A computerized neuropsychometric examination was given. At 2 years of follow-up, a diagnosis of dementia was made by a neurologist using DSM-IIIR criteria and without knowledge of the results of the cognitive testing. Rates of conversion to incident dementia were assessed by receiver operating characteristics analysis. Results: The MCI-R prevalence was found to be 16.6% using revised criteria. A significantly better prediction of transition to dementia (AUC = 0.80, sensitivity: 95%, specificity: 66%) was obtained with MCI-R than with the previous MCI criteria (AUC = 0.48, sensitivity: 5%, specificity: 91%). The predictive power was found to increase when MCI subtypes
doi:10.1159/000096287 pmid:17047325 fatcat:2whcsadggnednhi3nsfeha7obi