Key points are not available for this paper at this time.
Background As language deficits have been found to be a strong predictor of conversion from MCI (Mild Cognitive Impairment) to dementia, connected speech analysis provides sensitive measures of cognitive decline through micro-linguistic features.Aims This study investigated specific linguistic measures in connected speech of MCI patients and healthy controls (HC), examining differences in micro-linguistic features and their correlation with cognitive function.Methods & procedures We analyzed language samples from 40 MCI patients and 22 healthy controls from the Delaware English Protocol Corpus of Dementia Bank. Participants completed five language tasks including picture descriptions, story retelling, and procedural narratives. An independent t-test was performed to compare the linguistic measures between the MCI group and the HC group. Correlation analysis showed highly positive relationships were excluded and the remaining variables were then used as predictors in the regression analysis. Stepwise multiple linear regression analysis to examine the influence of linguistic feature on cognitive function (measured by Montreal Cognitive Assessment scores).Outcomes & results MCI patients demonstrated significantly reduced lexical semantic measures and morpho-syntactic measures. Correlation analysis showed highly positive relationships between MLU in morphemes and Verb Utt. Stepwise multiple linear regression identified MLU in morphemes as a significant predictor of cognitive status (B = 0.79, F (1, 60) = 16.26, p < 0.01).Conclusion MCI patients exhibit distinct patterns of linguistic impairment characterized by reduced lexical diversity, simplified syntactic structures, and decreased propositional density. MLU in morphemes emerges as a particularly valuable linguistic marker for cognitive assessment and may sensitively reflect cognitive variation, which have auxiliary value to distinguish and predict MCI.
Chen et al. (Sun,) studied this question.