The validity and test-retest reliability of the Japanese version of the Mini Mental State Examination （MMSE-J）（Sugishita 2006）was preliminarily evaluated in 2010（Sugishita, Hemmi and JADNI, 2010）based on data from the Japanese Alzheimer’s Disease Neuroimaging Initiative（J-ADNI）.　However, in March, 2014, Drs. Sugishita, M and Asada, T reported a total of 105 cases with confirmed or suspected falsification of data or protocol violation to the Special Investigative Committee of the University of Tokyo. In June 20, 2014, the Committee confirmed that data had been inappropriately modified by inappropriate persons.　In October 2014, 129 cases with confirmed or suspected falsification of data or protocol violation were reported to the Third Party Panel of the University of Tokyo.　The Third Party Panel examined all 129 subjects and found no falsification of data or protocol violations. However, the four written refutations by Dr. M. Sugishita argued against the criticism of the Third Party Panel and it became clear that the Third Party Panel had made errors in judgement（cf. http://www.geocities.jp/shinjitunodentatu/daisannsyaiin.html, Sugishita et al. 2016, http://sinjitunodentatu.web.fc2.com/）.

　　Consequently, Sugishita et al. （2016） reexamined the validity and reliability of the MMSE-J after excluding cases with confirmed or suspected falsification of data or protocol violations.　The paper also indicated that 1） Because the serial 7s is rather easy for the Japanese subject, it is not appropriate to use the backward repetition task instead of the serial 7s. 2） In the previous study（Sugishita et al. 2010）, the medical doctor sometimes knew the MMSE-J score in advance, which influenced his diagnosis of the subject.　Consequently, the reported validity is problematic.　3） The US-ADNI data showed that the conversion rate of the amnestic type of the MCI to the AD was 16.5% one year after the first examination.　However, the J-ADNI data revealed that the conversion rate was extremely high, that is, 29.0%, which suggests that the J-ADNI data is dubious.　The validity and reliability of the MMSE-J should be evaluated not on the J-ADNI data but on a new data.

　　The aim of the present study was to resolve the above-mentioned three problems and complete standardization of the MMSE-J.　In the present study, 1） The problem of the serial 7s and the backward repetition task was examined using Folstein’s original procedure of the Attention and Calculation task of the MMSE （2001）.　Folstein’s original procedure is “If the patient refuses to perform the serial 7s task, ask him to spell “world” forwards and then backwards.”　（Because one can perform neither backward spelling nor forward spelling in Japanese, we employed the backward repetition and forward repetition of a Japanese word）.　2） To estimate the validity, the DSM-5 and the FAST were employed as the external standard. Sensibility and specificity by ROC analysis was employed as index of validity.　3） Instead of the J-ADNI data, 381 subject （Normal 175, MCI 137, Mild AD 69）were newly recruited from the 6 hospital / clinics for standardization of the MMSE-J with Folstein’s original procedure of Attention and Calculation Task. The validity and reliability of the MMSE-J with are estimated on 379 subjects （Normal 175, MCI 137, Mild AD 67）

　　The results showed that in the MCI and the Mild AD groups, gender, age and years of education did not influence on the MMSE-J score.　In the Normal group, the MMSE score was significantly low, when years of education was low or age was high.　Gender did not influence on the MMSE-J score.　The number of the subject who could not perform the serial 7s or refused to perform it was very small, 5 and 2 respectively.

(View PDF for the rest of the abstract.)