Basic Science and Pathogenesis.

Area of Science:

• Neuroscience
• Cognitive Psychology
• Linguistics

Background:

• Verbal fluency tasks are standard for distinguishing mild cognitive impairment (MCI) from healthy controls (HC).
• MCI patients typically show lower word production efficiency within semantic categories compared to HC.
• Existing methods may be unreliable when HC and MCI patients have similar word output.

Purpose of the Study:

• To investigate the utility of lexical property distribution analysis in detecting MCI, particularly in individuals with high verbal fluency.
• To explore if analyzing word properties within word lists can improve MCI detection.

Main Methods:

• Eighty native Chinese speakers (40 HC, 40 MCI) performed a one-minute verbal fluency task (Animal category).
• Participants were stratified into high and low proficiency subgroups (HC, LC, HP, LP).
• Distribution analysis examined word frequency, contextual diversity, semantic diversity, surprisal, and frequency variation.

Main Results:

• Distribution analysis revealed significant differences between control and patient groups, mainly driven by the low-proficiency (LP) MCI group.
• A subtle difference was observed between low-proficiency controls (LC) and high-proficiency (HP) MCI patients based on individual word properties.
• Frequency variation analysis of the first 15 words clearly differentiated control and patient groups, with LC starting with high-frequency words and HP maintaining a consistent frequency band.

Conclusions:

• Individual word property analysis distinguished only low-proficiency MCI patients.
• Considering frequency variation in word lists enabled differentiation between high-proficiency MCI patients and controls.
• Findings suggest MCI patients may have altered semantic memory structures, but larger studies are needed to validate the positional word frequency approach.