Biomarkers

Area of Science:

• Neuroscience
• Cognitive Science
• Biomarkers

Background:

• Subjective Cognitive Decline (SCD) is an early stage of cognitive impairment.
• Remote, automated speech-based cognitive assessments offer a novel approach to evaluate cognitive function.
• Validation of these tools against traditional methods is crucial for their clinical adoption.

Purpose of the Study:

• To validate the Mili-platform, a remote, automated speech-based cognitive assessment.
• To compare speech-derived cognitive scores (SB-C) with standard paper-and-pencil cognitive evaluations.
• To assess the utility of speech biomarkers in individuals with SCD.

Main Methods:

• 91 participants with SCD from the B-AARC cohort were enrolled.
• Speech samples were collected via the Mili platform during cognitive tasks (semantic verbal fluency, Auditory Verbal Learning Test).
• Speech-based cognitive scores (SB-C) were generated and correlated with paper-and-pencil cognitive composite scores (executive function, memory, language, global cognition).

Main Results:

• SB-C scores demonstrated moderate correlations with paper-and-pencil cognitive domains.
• The SB-C global cognition score significantly correlated with the PACC (r=0.57, p<0.0001).
• SB-C executive function and memory scores showed significant correlations with corresponding paper-based assessments.

Conclusions:

• Remote, automated speech-based cognitive assessments are a valid and scalable tool for individuals at risk of Alzheimer's Disease (AD).
• Speech features may provide additional, sensitive information not captured by traditional tests.
• Further research is warranted to explore speech biomarkers for predicting cognitive decline and brain pathology.