The Japanese Version of 26-Item Informant Questionnaire on Cognitive Decline in the Elderly: Development and...

To develop and validate a Japanese version of the 26-item Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE-J) for detecting dementia.

In this cross-sectional study, consecutive outpatients attending a university memory clinic between 2019 and 2021 completed the IQCODE-J through informants. Dementia was diagnosed according to DSM-5 criteria, independent of IQCODE-J scores. Diagnostic accuracy was assessed using receiver operating characteristic analysis and bootstrap internal validation. Concurrent validity was examined by correlations with the Ascertain Dementia 8-item Informant Questionnaire (AD8-J), Clinical Dementia Rating (CDR-J), Mini-Mental State Examination (MMSE-J), and Frontal Assessment Battery (FAB), and by multivariable logistic regression.

Of 166 participants, 93 had non-dementia and 73 had dementia. IQCODE-J scores were higher in the Dementia group (p < 0.001). The apparent optimal cut-off in the development dataset was 4.06. Bootstrap internal validation showed an optimism-corrected area under the curve of 0.74, with sensitivity and specificity of 47.5% and 81.8%, respectively. The bootstrap median cut-off was 4.00. At cut-offs of 3.4-3.6, sensitivity ranged from 87.7% to 78.1%, and specificity from 33.3% to 54.8%. Each 1-point increase in IQCODE-J was associated with higher odds of dementia (adjusted odds ratio 7.55, 95% CI 3.31-18.74). IQCODE-J correlated with the AD8-J, CDR-J, MMSE-J, and FAB (rho = 0.77, 0.45, -0.47, and -0.44, respectively; all p < 0.001).

IQCODE-J is a valid and feasible informant-based tool for assessing cognitive decline in Japanese patients. A cut-off around 4.0 may be practical when specificity is prioritized in referral-based settings.