P2‐119: Japanese versions equivalent to original English neuropsychological tests in ADNI

Abstract

The number of international research project and that of global clinical trials are increasing in the field of Alzheimer Disease. This situation urges researchers to develop the non-English versions equivalent to original English neuropsychological tests such as MMSE, CDR, Logical Memory, ADAS-COG and so on. In non-English speaking countries, such equivalent versions are prerequisite to collect the Alzheimer Disease, the MCI, and the normal groups equivalent to the groups in the English speaking countries, since neuropsychological tests are employed as the inclusion or the exclusion criteria to classify the population into the AD, the MCI and the normal group. We, Japanese, joined an international research project in US, namely, the Alzheimer Disease Neuroimaging Initiative (ADNI) since 2007 and developed new versions of 6 neuropsychological tests(MMSE,CDR, GDS-S, ADAS-COG, NPI-Q and FAQ). To create the equivalent version, erroneous translation and unnecessary adaptation were avoided. Psycholinguistic data are employed to adapt cultural differences between US and Japan. Some comparisons between the Japanese ADNI data and the US-ANDI data are in progress to clarify whether or not the new Japanese versions are equivalent to the 6 original English tests.In order to clarify whether the original MMSE and MMSE-J are equivalent or not, the data of original English MMSE in the US-ADNI and those of MMSE-J in the Japanese ADNI were analyzed with regression analysis. Excluding subjects with depression, race origin other than white and Japanese, less than 8 years of education and so on, 699 subjects in the US-ADNI and 224 subjects in the Japanese ADNI were available. The dependent variable of the regression model was the score of original English MMSE and that of MMSE-J. To control their effects on the test score, age, years of education and APOE genotype were used as independent variables besides nation. The results showed the insignificant effect of nation and the significant effects of, age, years of education and APOE genotype on the MMSE score. These findings implied that both versions of MMSE were equivalent.