The sensitivity and specificity of a Chinese version of Montreal Cognitive Assessment (MoCA-CV) for Screening mild cognitive impairment (MCI) were examined.

MoCA-CV and MMSE together with a battery of neuropsychological tests were administered to 1056 elders who were recruited from three regions of Beijing (i.e. new town, old town, rural areas).

Twenty-six participants were diagnosed with dementia, 118 were MCI by psychiatrists and neurologists, and 875 were considered to be cognitively normal based on independent neuropsychological tested. Using the recommended cut-off score of 26, the MoCA-CV had an excellent sensitivity of 90.4% to detect MCI from controls, while the specificity was relatively poor (32.0%). Using a lower cut-off score of 22, the MoCA-CV exhibited the optimal balanced sensitivity of 68.7% and specificity of 65.2%. Among all the seven cognitive sub-domains, delayed recall performed the best differentiation of MCI from normal controls. Item analysis showed the internal consistency was relatively low in both naming and sentence repetition tasks which related with the characteristics of Chinese language and culture. Region differences were not significant when such confounding demographic variables as age, gender and education were controlled.

The MoCA-CV is a brief and effective screening tool for detecting MCI. However, presumably due to linguistic and culture issues and the lower education level of Chinese elders, the MoCA-CV did not show better screening capability than MMSE as expected. Some items may need further adaptation to accommodate the language, culture and education characteristics of Chinese participants.