Related Experiment Video
Updated: Jun 26, 2026

A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
Does a reliable decline in Mini Mental State Examination total score predict dementia? Diagnostic accuracy of two
Anke Hensel1, Tobias Luck, Melanie Luppa
1Department of Psychiatry, Public Health, University of Leipzig, Leipzig, Germany.
Background/Aim:
Norms for change in Mini Mental State Examination (MMSE) total score suggest that only a decline of at least 2-4 points indicates a reliable change. However, it is unknown whether change norms (Reliable Change Indices, RCIs) of the MMSE total score are suitable to predict future dementia.
Methods:
554 elderly individuals aged 75 and over without dementia at the first 2 visits were tested with the MMSE at a maximum of 6 visits with 1.5-year intervals. Two different RCIs for change in MMSE score (first to second visit) were computed - one RCI which corrects for practice and one RCI which corrects for regression to the mean. The main outcome measure was the diagnosis of dementia.
Results:
During the study, 88 persons developed dementia. RCIs were significantly associated with future dementia diagnosis. The best cutoff for raw change in MMSE total score to predict dementia was -1 point (sensitivity = 48%, specificity = 67%, relative risk = 1.6). With the RCI + regression to the mean, the diagnostic accuracy was moderate (sensitivity = 61%, specificity = 72%, relative risk = 3.2).
Conclusion:
A change in MMSE total score is significantly associated with future dementia, but the diagnostic accuracy for dementia prediction is rather low.
Insights
Changes in Mini Mental State Examination (MMSE) scores can predict future dementia. While a decline of -1 point shows association, the diagnostic accuracy for dementia prediction remains low.
Area of Science:
- Gerontology
- Neurology
- Cognitive Science
Background:
- Established norms indicate a 2-4 point MMSE score decline signifies reliable change.
- The predictive value of MMSE change norms for future dementia diagnosis is not well-understood.
Purpose of the Study:
- To evaluate the suitability of MMSE change norms, specifically Reliable Change Indices (RCIs), for predicting dementia in elderly individuals.
- To assess the diagnostic accuracy of MMSE score changes for future dementia prediction.
Main Methods:
- Longitudinal study of 554 individuals aged 75+ without dementia at baseline.
- MMSE administered up to 6 times over 1.5-year intervals.
- Calculated two RCIs for MMSE score change (visit 1 to 2) and correlated with dementia diagnosis.
Main Results:
- 88 participants developed dementia during the study.
- RCIs demonstrated a significant association with subsequent dementia diagnosis.
- A raw MMSE change cutoff of -1 point yielded moderate predictive accuracy (sensitivity 48%, specificity 67%).
- RCI accounting for regression to the mean showed improved accuracy (sensitivity 61%, specificity 72%).
Conclusions:
- Changes in MMSE total score are significantly linked to future dementia risk.
- The diagnostic accuracy of MMSE score changes for predicting dementia is currently limited.
Related Concept Videos
Dementia l: Introduction
Alzheimer Disease l: Introduction
Dementia
The progression of dementia is generally gradual.

