P2‐313: The Montreal Cognitive Assessment memory index score (MOCA‐MIS) and total MOCA score to help predict MCI conversion to Alzheimer's disease

The MoCA has been shown to distinguish Cognitively Normal Subjects from patients with MCI or AD in a cross sectional study. This study aims to evaluate the usefulness of the MoCA Total Score (MoCA-TS), and MoCA Memory Index Score (MoCA-MIS) to predict MCI to AD conversion in MCI subjects who were followed for up to 5 years. Data was collected from charts of consecutive subjects presenting with cognitive complaints to the CEDRA/Neuro Rive-Sud Memory clinic between 2005 and 2011. MCI diagnosis was made according to Petersen's criteria. ROC curves were analyzed to evaluate the predictive ability of the MoCA-TS, and MoCA-MIS to distinguish MCI Converters (MCI-C) from MCI-Non Converters (MCI-NC). MoCA-MIS is calculated by adding delayed recall score (X 3), category-cued recall (X 2), and multiple choice-cued recall (X 1), with possible scores of 0-15. 168 MCI subjects met study criteria. 116/168 subjects converted to AD (69.04%), and 52 subjects remained MCI on follow-up. Average time to conversion to AD was 18 months (Range 2-68 months), and average follow-up for MCI-NC was 20 months (Range 2-62). Average MCI Converters (MCI-C) and MCI-Non-Converters (MCI-NC) baseline MoCA Total Scores (MoCA-TS) were: 19.34 (SD 3.68), and 22.15 (SD 3.50) respectively (t = 4.64, p< 0.001), and Memory Index Scores (MoCA-MIS) were 6.66 (SD 3.00) and 8.54 (SD 3.18) respectively (t = 3.68, p< 0.001). ROC curve analysis showed that both MoCA-TS and MoCA-MIS are equally predictive of MCI conversion to AD with ROC curve values of 0.71 and 0.67 respectively. Using cut-offs for MoCA-TS (19-20/30) and MoCA-MIS (6-7/15), MCI conversion rates were 53% when both scores were above the cut-off and 91% when both scores were below the cut-off. Logistic regression analysis did not show any significant age or education effects.