Fluency in spontaneous speech predicts individual variance in executive function among seniors

Abstract Background Several recent studies have used automated linguistic analysis of naturalistic speech in a picture‐description task (e.g. Cookie Theft) combined with machine learning approaches to distinguish the those dementia from healthy age‐matched controls. Extension these techniques predict continuous variables related cognitive status offers means track severity over time and evaluate effectiveness interventions. Here we examine relationship variability executive function, ability manage conflicting information speeded performance. Higher function confers protection against clinical manifestation despite underlying neurodegeneration. Method In new sample 76 adults aged 65‐75, measured using an extensive test battery, elicited spoken picture descriptions same individuals. The battery included N‐back, Simon task, verbal fluency, Color Word Interference (CWI or Stroop test), Trails A & B, elevator subtests Test Everyday Attention, Logical Memory. Two description narratives were recorded subjected generating 400 features, grouped into 8 composite measures. Result Exploratory factor revealed one accounting for most variance battery. This exhibited significant correlations four eight measures, strongest weakest: word finding difficulty, local coherence, lexical richness, syntactic complexity, reflecting greater fluency spontaneous descriptive speech. No found repetitiveness, global units, sentiment. Ongoing analyses include characterization language measures larger normative sample. Conclusion Older individuals higher as captured by common laboratory tests, also exhibit more fluent quantified analysis. difficulties may reflect overall slowing processes access long‐term memory. Automated ultimately serve inexpensive repeatable measure older at risk dementia.