Connected speech profiles in mild cognitive impairment reflect global cognition

Area of Science:

• Neurology
• Linguistics
• Psychology

Background:

• Mild cognitive impairment (MCI) often precedes Alzheimer's disease and is associated with diverse connected speech (CS) changes.
• Existing research on CS in MCI presents heterogeneous findings, suggesting multiple underlying profiles.
• Cognitive heterogeneity in MCI further supports the possibility of distinct CS patterns.

Purpose of the Study:

• To identify distinct connected speech (CS) profiles within the mild cognitive impairment (MCI) population.
• To investigate whether these identified CS profiles correlate with specific cognitive characteristics.

Main Methods:

• Connected speech (CS) characteristics were analyzed from 109 controls and 210 individuals with MCI from the COMPASS-ND study.
• A Two-Step Cluster Analysis was employed to identify distinct CS profiles in the MCI cohort.
• Linguistic and cognitive features of identified CS profiles were compared against each other and against control groups.

Main Results:

• Two distinct CS profiles were identified in individuals with MCI.
• Profile 1: Characterized by reduced syntactic complexity and semantic content, associated with broader cognitive deficits (visuospatial, memory, executive functions).
• Profile 2: Characterized by dysfluencies, longer speech samples, and reduced semantic efficiency, primarily linked to episodic memory impairments.

Conclusions:

• Connected speech (CS) analysis can differentiate between MCI subtypes with varying cognitive severities.
• CS analysis serves as a potential screening tool to identify individuals with MCI who may require further medical or cognitive assessment.
• Identifying specific CS profiles can guide tailored cognitive and psychological interventions for individuals with MCI.