P1‐427: A community‐based study of the Montreal Cognitive Assessment (MoCA) in Japan: The Kurihara Project

The Montreal Cognitive Assessment (MoCA) was a new screening instrument for mild cognitive dysfunction. There was a clinic-based study using the MoCA Japanese version (MoCA-J), but no previous community-based studies. The purpose of this study is to examine the validity of the MoCA of old-old people in the community. We obtained the agreement from 229 community dwellers, aged 75+ years in 2010 living in Kurihara, Northern Japan. We utilized Clinical Dementia Rating (CDR), clinical examination, neuropsychological tests and MRI. Of the 229 subjects, there were 98 with a CDR 0 (healthy), 107 with a CDR 0.5 (Mild Cognitive Impairment; MCI), and 24 with a CDR 1+ (dementia). The MCI and dementia groups were significantly older than the healthy group in age (healthy vs. MCI vs. dementia; 78.8 vs. 80.8 vs. 82.6 years (mean); p < 0.001) and lower in educational level (9.5 vs. 8.6 vs. 8.0 years (mean); p < 0.001). We analyzed the differences between the 3 CDR groups in the MoCA, MoCA subscales, and Mini-Mental State Examination (MMSE). There were significant decreases step by step through 3 CDR groups in the MoCA (21.5 vs. 17.6 vs. 12.0; p < 0.001) and MMSE (25.5 vs. 22.9 vs. 16.0; p < 0.001). In MoCA subscales, there were significant decreases step by step through 3 CDR groups in naming, memory (encoding), vigilance, serial 7s, and abstraction (p < 0.05). The MCI and dementia groups had significantly lower scores than the healthy group on alternating Trail Making/visuoconstructional skill and delayed recall (p < 0.05). The dementia group had significantly lower scores than the healthy and MCI groups on orientation (p < 0.05). From the receiver operating characteristic (ROC) curve, the area under the curve (AUC) in the MoCA (healthy vs. MCI+) was 0.78 (p < 0.001), and the cutoff was at 20/21 (sensitivity 0.77, specificity 0.57). These results suggest that the MoCA is an appropriate method of evaluation for detecting MCI subjects and dementia patients in the community. Since the MoCA was affected by older age and lower educational level, the cutoff scores should be re-considered.