Utility of a shortened Hasegawa Dementia Scale Revised questionnaire to rapidly screen and diagnose Alzheimer's disease.

Aims: The aim of this study was to analyze the sensitivity and specificity of a shortened Hasegawa Dementia Scale Revised (shortened HDS-R) questionnaire and explore its utility for the rapid screening and diagnosis of Alzheimer's disease (AD).

Methods: We included 113 patients over the age of 60 years who visited our hospital from June 2018 to January 2021 including 70 subjects with AD and 43 healthy subjects. AD was diagnosed in accordance with the diagnostic criteria of the Diagnostic and Statistical Manual of Mental Disorders, 4th edition, and the standard HDS-R questionnaire was used as a neuropsychological examination. The shortened HDS-R questionnaire was composed of the first seven subdomains (1 to 7) of the HDS-R questionnaire and excluded subdomains 8 and 9. Magnetic resonance imaging (MRI) was performed to calculate the degree of atrophy of the whole brain, hippocampus, and parahippocampal gyrus.

Results: The cumulative contribution ratio of subdomains 1 to 7 of the HDS-R questionnaire was as high as 94%, indicating that the construct validity of the shortened HDS-R was very good. The correlation coefficient of the total scores of the shortened HDS-R and the HDS-R was 0.96, indicating that the criterion-related validity was also very good. Furthermore, the shortened HDS-R was significantly negatively correlated with the degree of atrophy in the whole brain, hippocampus, and parahippocampal gyrus, indicating that its concurrent validity was very good in relation to imaging parameters. Cronbach's α coefficient of the shortened HDS-R was 0.76, and the correlation coefficient of the item-total correlation analysis was between 0.68 and 0.76, indicating that this questionnaire has high internal consistency and reliability. The total shortened HDS-R score of the normal group (17.0 ± 1.9) was significantly higher than that of the AD group (8.6 ± 3.8), demonstrating that the total shortened HDS-R score can be used to identify healthy individuals and patients with AD. When the cutoff score was 14 of 15, the sensitivity was 92.9% and the specificity was 88.4%. The diagnostic ability of the shortened HDS-R was 91.2%, which indicates that it is similar to the full HDS-R questionnaire as an AD screening tool.

Conclusion: As a neuropsychological examination questionnaire for the screening and diagnosis of AD, the shortened HDS-R had very high validity and reliability. Its sensitivity, specificity, and diagnostic ability were similar to those of the gold standard HDS-R; therefore, it can be considered a concise and useful questionnaire for AD screening and diagnosis in the older population.