Related Experiment Video
Updated: Jun 24, 2025

A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
Combining the AD8 and MMSE for community-based dementia screening
Yun-Ru Lu1, Shin-Fang Chang2, Horng-Huei Liou3
1Department of Neurology, China Medical University Hospital Taipei Branch, Taiwan.
Background:
This study aimed to determine whether a cognitive test the Mini-Mental State Examination (MMSE) and the Ascertain Dementia 8 (AD8) instrument applied in combination could improve the accuracy of dementia detection in a community setting.
Methods:
Study participants were recruited from a community-based integrated screening program in Tainan, Taiwan. Participants completed the AD8 and were administered the Chinese version of the MMSE by psychologists. In addition, the presence of dementia was determined by neurologists based on the 2011 National Institute on Aging-Alzheimer's Association guidelines. Logistic regression analysis determined whether the combination of these two tests provided any additional information for dementia detection than either test alone. Receiver operating characteristic (ROC) curve analyses were conducted to explore the performances of different screening modalities in detecting dementia.
Result:
In total, 282 participants with an average age of 69.31 ± 10.27 years were enrolled. The prevalence of dementia among participants aged ≥65 years was 9.29 %. The sensitivity and specificity of the AD8 applied alone for detecting dementia were 64.71 % and 87.89 %, respectively, and of the MMSE applied alone, after adjusting for education level, were 41.18 % and 84.50 %, respectively. Using a cutoff score of 21 for the MMSE resulted in sensitivity of 77.78 % and specificity of 73.58 %. The AD8 and MMSE when combined in parallel yielded 88.89 % sensitivity and 70.16 % specificity. The serial use of the AD8 followed by the MMSE yielded 50 % sensitivity and 93.02 % specificity. Except for when an MMSE cutoff value of 26 was applied, the sensitivity of all examined modalities was poor and specificity was moderate for detecting mild cognitive impairment. ROC curve analysis revealed that the parallel application of the MMSE and AD8 (area under the ROC curve [AUC]: 82.3 % [75.1 %-89.4 %]) resulted in better dementia detection accuracy than the AD8 alone (AUC: 73.3 % [60.7 %-85.9 %]), the MMSE alone (AUC: 77.4 % [67.6 %-87.3 %]), or serial test administration (AUC: 67.6 % [53.4 %-81.8 %]).
Conclusion:
This study successfully demonstrated that the MMSE and AD8 combination for dementia screening could improve detection accuracy in a community setting.
More Related Videos
06:23The 4 Mountains Test: A Short Test of Spatial Memory with High Sensitivity for the Diagnosis of Pre-dementia Alzheimer's Disease
Published on: October 13, 2016
06:58Highlighting and Reducing the Impact of Negative Aging Stereotypes During Older Adults' Cognitive Testing
Published on: January 24, 2020