P2–260: Disease progression markers from connected speech
Samrah Ahmed, Celeste de Jager, A.-M. Haigh, Peter Garrard · Alzheimer s & Dementia · 2013
Our group recently described syntactic simplification and impairments in lexico-semantic processing in samples of connected speech samples obtained at the time of diagnosis from patients in whom AD was subsequently confirmed at post mortem. In this study we set out to identify features of connected speech that: i) had been abnormal during the MCI phase of these patients' illness; and ii) showed consistently greater deviation from normal performance with disease progression. Members of the OPTIMA cohort, in whom diagnoses of both MCI and probable AD had been made during longitudinal evaluation, were selected for the study. Transcripts of the Cookie Theft picture description task were subjected to analysis using the Quantitative Production Analysis (QPA) approach. QPA produces a series of measurements and error counts that can be grouped under four broad descriptive domains: syntactic complexity, lexical content, speech production and fluency. Semantic content was quantified by identifying and counting semantic units. Variables on which the majority of cases showed a linear decline between the MCI and early AD stages and were taken as potential markers of progression. These were then applied to transcripts obtained from the same patients at a uniform stage of moderate disease severity. There were significant linear trends over the three stages of the disease (MCI, early and moderate AD) for: syntactic complexity (F(1,8)=12.304, p0.05) or fluency errors (F(1,8)=2.735, p>0.05). The study found that subtle changes in language function typical of AD are evident during the preclinical stage, and that a number of these indices showed a progressive decline through later clinical stages of the disease. Translation of these striking observations into simple and specific markers, of language change in AD could have far-reaching clinical consequences.