We analysed data from people with clinical diagnoses of Alzheimer’s disease (n = 18),28 (link) behavioural variant of FTD (bvFTD; n = 16),29 (link) CBS (n = 17),30 (link) PSP (n = 36),31 (link) nfvPPA (n = 26) and svPPA (n = 26),32 (link),33 (link) as well as healthy controls (n = 33). Patients were recruited from specialist memory and movement disorders clinics at Cambridge University Hospitals NHS Trust in observational studies (REC references 07/Q0102/3, 10/H0308/34, 12/EE/0475, 14/LO/2045 and 16/LO/1735).
Verbal fluency tests were administered during the baseline visit. Participants were asked to name as many words as they could that (i) began with the letter ‘P’ (excluding people and place names) and (ii) that belonged to the category of ‘animals’, to assess phonemic/letter and semantic/category fluency, respectively. Words were recorded over 60 s for each task and transcribed by the examiner. Features (e.g. frequency and concreteness) were extracted for bigrams when available; otherwise, the bigram was included as a single entry. Where ratings for pluralized words were unavailable, the word properties for the singular version were extracted. In case of homonyms, when the intended word was unclear, the most frequent word was transcribed then analysed. We indicated ‘NA’ for features for which a rating was not available. Where a feature was not available for a given word (i.e. outside of the psycholinguistic databases), the word was excluded from the PCA. The total word count, excluding errors and repetitions, was calculated.
For each of the words, we obtained ratings for psycholinguistic properties from the MRC Psycholinguistic Database34 and the English Lexicon Project35 (link) as listed inTable 1 .
Verbal fluency tests were administered during the baseline visit. Participants were asked to name as many words as they could that (i) began with the letter ‘P’ (excluding people and place names) and (ii) that belonged to the category of ‘animals’, to assess phonemic/letter and semantic/category fluency, respectively. Words were recorded over 60 s for each task and transcribed by the examiner. Features (e.g. frequency and concreteness) were extracted for bigrams when available; otherwise, the bigram was included as a single entry. Where ratings for pluralized words were unavailable, the word properties for the singular version were extracted. In case of homonyms, when the intended word was unclear, the most frequent word was transcribed then analysed. We indicated ‘NA’ for features for which a rating was not available. Where a feature was not available for a given word (i.e. outside of the psycholinguistic databases), the word was excluded from the PCA. The total word count, excluding errors and repetitions, was calculated.
For each of the words, we obtained ratings for psycholinguistic properties from the MRC Psycholinguistic Database34 and the English Lexicon Project35 (link) as listed in
Free full text: Click here
