Association of language markers with future cognitive impairment and presence of limbic predominant age related TDP‐43 encephalopathy

INTRODUCTION: Accessible biomarkers for prediction of cognitive impairment and diagnosis of limbic predominant age related TDP-43 encephalopathy neuropathologic change (LATE-NC) are needed. METHODS: Written descriptions of the cookie-theft picture were produced by 134 participants of the Leisure World Cohort Study who subsequently joined the 90+ Study and underwent regular assessments and autopsy. We examined the relationships between linguistic markers and future cognitive impairment and the presence of neuropathologic changes. RESULTS: Mean age at writing was 85 and there was an average interval of 8 years to the development of cognitive impairment. Future cognitive impairment was associated with less grammatical units, fewer clauses, and fewer pictorial themes. LATE-NC was the only neuropathologic change associated with language markers, including higher proportions of content words, less complete units, and more closed class word errors. DISCUSSION: Our results suggest that linguistic markers obtained long before the development of cognitive impairment might serve as biomarkers for future cognitive impairment and LATE-NC. Highlights: We analyzed the writing samples of 134 participants in the Leisure World Cohort Study who later joined the 90+ Study and had longitudinal assessments and came to autopsy. We found that features extracted from these writing samples differed between participants who died with cognitive impairment and those who died with normal cognition. Of note, participants who developed cognitive impairment were all cognitively normal at the time of writing of the samples and developed cognitive impairment on average 8 years later. We found a different set of features were related to the presence of limbic predominant age related TDP-43 encephalopathy (LATE) at post mortem. No features were related to either Alzheimer's disease neuropathology or Lewy body (LB) pathology.