Clinical utility of mild cognitive impairment subtypes and number of impaired cognitive domains at predicting progression to dementia: A 20-year retrospective study.

Objective: To determine the utility of mild cognitive impairment (MCI) subtypes and number of impaired cognitive domains on initial assessment at predicting progression to dementia in a sample of memory clinic patients over a 20-year period.

Methods: A retrospective analysis was conducted of those presenting to a memory clinic with MCI from 1 January 1999 to 31 December 2018 inclusive. Those with MCI were broken down into one of the four subtypes using recommended cut-off scores on the Cambridge Cognitive Assessment (CAMCOG). Binomial logistic regression analysis was used to determine the utility of MCI subtypes and number of impaired cognitive domains as predictors for dementia.

Results: Overall 1188 individuals with MCI diagnosis were identified, with 378 (32%) progressing to dementia, with median [range] time to diagnosis of 2 years [1-8.4]. Six hundred and forty-nine (55%) were identified as amnestic MCI and 539 (45%) as non-amnestic MCI. Amnestic MCI was a significant predictor of progression compared to non-amnestic MCI (OR = 1.85, df = 1, P < .001). Number of cognitive domains impaired was also a significant predictor of progression to dementia (OR = 1.07, df = 1, P = .01) but the single-/multi-domain distinction was not (OR = 1.29, df = 1, P = .36).

Conclusion: This study shows that approximately 32% of those diagnosed with MCI in a memory clinic progressed to dementia, with a median time to progression of 2 years. Those with amnestic MCI are almost twice as likely to progress to dementia than non-amnestic MCI and that therefore this is a useful distinction. However, the utility of the single- and multi-domain MCI distinction is called into question by our findings.