Clinical Manifestations

Area of Science:

• Neurology
• Speech Pathology
• Biomarker Discovery

Background:

• Subtle speech features, such as pauses, are sensitive indicators of Alzheimer's disease (AD) pathology, detectable via PET biomarkers.
• Previous research demonstrated that silent pauses in Cookie Theft (CT) picture descriptions could differentiate cognitively unimpaired (CU) from cognitively impaired individuals, with the latter exhibiting longer speech and pauses.
• This study extends the investigation of automated pause detection methods to the domain of story recall (SR) tasks.

Purpose of the Study:

• To investigate the utility of automated acoustic analysis of silent pauses during story recall (SR) as a potential biomarker for early Alzheimer's disease (AD).
• To compare speech characteristics, specifically pause metrics and duration, between individuals with Mild Cognitive Impairment (MCI) and cognitively unimpaired (CU) controls.
• To assess the correlation of these speech metrics with PET biomarkers (amyloid/tau), neuropsychological test results, and overall cognitive status.

Main Methods:

• Fifty participants (25 MCI, 25 CU), matched for key demographics, were selected from the Wisconsin Registry for Alzheimer's Prevention (WRAP) and Wisconsin Alzheimer's Disease Research Center (WADRC).
• Delayed story recall (SR) audio recordings (Logical Memory for WRAP, Craft Story for WADRC) were processed to extract silent pauses (>80ms) using MATLAB Speech Pause Analysis.
• Statistical analyses included Wilcoxon signed rank tests for group comparisons, Spearman correlations for relationships with clinical/biomarker data, and logistic regression/ROC analyses for diagnostic prediction.

Main Results:

• Cognitively unimpaired (CU) participants showed longer SR duration, while the Mild Cognitive Impairment (MCI) group exhibited longer mean pause duration and a higher percentage of pauses.
• Speech duration in SR was significantly correlated with performance on various neuropsychological tests and with amyloid PET burden.
• Total SR duration emerged as a significant predictor of cognitive status (AUC=0.76), outperforming the Cookie Theft (CT) task in differentiating between cognitive groups.

Conclusions:

• Individuals with MCI demonstrate shorter SR duration and increased pause frequency/duration compared to CU individuals, potentially reflecting episodic memory deficits.
• Shorter speech duration in SR tasks is associated with higher amyloid-beta levels, suggesting a link to AD pathology.
• Automated acoustic analysis of pauses in SR shows promise as a sensitive, non-invasive early marker for AD, meriting further investigation in diverse populations.