You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer
© 2025. This work is published under http://creativecommons.org/licenses/by/4.0/ (the "License"). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.
Abstract
Background
The diagnosis of Mild Cognitive Impairment (MCI) is challenging and relies on accurate cognitive testing. To improve the validity of cognitive test results, demographic normalisation is commonly used to adjust test scores for the effects of factors like age, sex and education. Many studies have developed normative tables/calculators based on these demographic factors for cognitive tests widely used to identify MCI ‐ the Montreal Cognitive Assessment (MoCA) and the Mini‐Mental State Examination (MMSE). Although demographic normalisation aims to improve diagnostic accuracy, its effectiveness has been seldom studied. This is concerning because, while normalisation improves accuracy for variables unrelated to disease, the commonly adjusted demographic factors—age, sex, and education—are themselves risk factors for MCI and Alzheimer's disease.
Method
We hypothesized that demographic normalisation reduces diagnostic accuracy when the demographic variables used (age, sex, education) differ systematically between diagnostic groups due to their association with disease risk. We tested this hypothesis by assessing whether normalisation affects the ability of MoCA and MMSE to distinguish MCI and early dementia from cognitively normal (CN) individuals, using receiver operating characteristic analysis with paired bootstrapping across multiple normalisation methods (bin‐based, linear, quadratic and partial).
Result
Surprisingly, demographic normalisation worsened rather than improved diagnostic accuracy for detecting MCI with both MoCA and MMSE, reducing their effectiveness as diagnostic tools and confirming our hypothesis. Demographic normalisation significantly reduced diagnostic accuracy, with AUC decrements of 0.007–0.015 (2‐tailed p <0.001) in 7 of 8 CN vs MCI conditions, highlighting that disease‐related demographic shifts often negate normalisation benefits and lower the diagnostic performance of MoCA and MMSE. Normalisation failed to consistently improve accuracy for detecting early dementia, with significant decreases in 3 of 8 conditions and a mix of significant and non‐significant increases in the others, indicating inconsistent effects on test performance.
Conclusion
The current findings indicate that normalisation with the commonly used demographic variables (age, sex, and education) has little effect on the accuracy of MoCA & MMSE for identifying MCI and early dementia. Even this little effect often results in a significantly decreased test accuracy, suggesting that demographic normalisation does more harm than good.
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer
Details
1 Harvard University, Madurai, Tamilnadu, India
2 Harvard University School of Enginering and Applied Sciences, Cambridge, MA, USA





