Predicting Progression to Dementia Using Auditory Verbal Learning Test in Community-Dwelling Older Adults Based On...

Area of Science:

• Gerontology
• Artificial Intelligence
• Cognitive Science

Background:

• Dementia diagnosis is challenging in primary healthcare.
• Early identification of dementia is crucial for timely intervention.

Purpose of the Study:

• Develop and validate machine learning models to predict dementia in older adults.
• Identify key predictive features for dementia development from cognitive assessments.

Main Methods:

• Four machine learning models (logistic regression, decision tree, random forest, gradient-boosted trees) were developed.
• A cohort of 1,162 older adults with normal cognition was analyzed.
• Models were validated on an independent cohort (n=1,370) using wrapper feature selection.

Main Results:

• The random forest model achieved 93% accuracy and an AUC of 0.88 in the primary cohort.
• External validation showed acceptable performance for random forest (AUC=0.81).
• Logistic regression performed better in the validation cohort with an AUC of 0.88.

Conclusions:

• Machine learning provides a viable strategy for dementia prediction in primary care.
• Memory tests and demographic data are key predictors.
• This framework supports cognitive change monitoring and early intervention.