Determinants of Course Patterns of Mild Cognitive Impairment (MCI)

The concept of MCI was developed to gain a predictor for future dementia. Annual conversion rates differ in short- and long term studies and the lack of a linear trend between both suggest that MCI subsumes a heterogenic group of more than one disorder, which may be stable, improve or progress to dementia. To distinguish progressive MCI from stable or remittent MCI, this study investigates indicators predicting the course of MCI and allowing a prognosis of the risk of future dementia. 357 subjects from the AgeCoDe cohort with MCI at baseline were followed up for 36 months by interviews and cognitive testing. Groups of remittent, unstable, stable and progressive MCI were built and analysed by multivariate and CART analysis for possible baseline-predictors for future course. The group of progressive MCI consisted of 22.4% of all subjects with MCI. The following variables were found to predict the conversion to dementia in a statistically significant manner (p<0.01): MCI subtype, Hypercholesterolemia, history of TIA, renal insufficiency, and a poor score of 14.5 points or lower in the CERAD Wordlist learning and recall subtest. The latter alone increases the risk of conversion to dementia from 22.3% to 48.6%. A score of more than 14.5 points decreases the risk of progressive MCI to 10.8%. No clear determinants were found for courses of unstable and stable MCI. In persons with MCI a poor score of 14.5 points or lower in the CERAD Wordlist learning and recall subtest is the best predictor of progression to dementia. For courses of unstable and stable MCI no clear predictors were found.