Clinical Manifestations.

Area of Science:

• Neuroscience
• Gerontology
• Clinical Psychology

Background:

• Functional impairment (FI) is crucial for diagnosing and monitoring neurodegenerative dementia (ND).
• Traditional assessments often focus only on activities of daily living (ADLs), neglecting other functional dimensions.
• This study evaluates a multidimensional approach to capture a broader functional phenotype.

Purpose of the Study:

• To assess the utility of the Chilean multidimensional functional assessment (Ch-MFA) in differentiating Alzheimer's disease (AD) and frontotemporal dementia (FTD).
• To evaluate the Ch-MFA's ability to predict dementia severity using the Clinical Dementia Rating (CDR).
• To compare the performance of a multidimensional assessment against unidimensional approaches.

Main Methods:

• Recruited 308 participants (209 controls, 44 FTD, 55 AD) in Peru.
• Utilized comprehensive functional assessments including ADLs, social behavior, mobility, and everyday cognition.
• Employed Random Forest models with cross-validation for predictive analysis and feature selection.

Main Results:

• The Ch-MFA achieved an average f1-score of 0.81 for differentiating dementia subtypes, outperforming unidimensional assessments.
• For severity prediction, Ch-MFA showed an average f1-score of 0.65 for severities 0.5, 1, and 2.
• A subset of 4 Ch-MFA questions achieved an f1-score of 0.88 for differentiating severities 1 and 2.

Conclusions:

• The Ch-MFA demonstrates superior performance compared to unidimensional assessments in differentiating dementia types and severity.
• Different functional dimensions are critical for distinguishing dementia subtypes versus predicting their severity.
• A multidimensional functional phenotype better reflects the specific characteristics of dementia subtypes.