Spelling in a case of logopenic variant of primary progressive aphasia

Abstract

Introduction Dysgraphia is a common feature of Primary Progressive Aphasia (PPA). In particular, different spelling profiles have been reported for the logopenic variant of PPA. Most commonly, there is impairment in lexical spelling and phoneme-to-grapheme conversion (Graham, 2014). The purpose of the current study was to evaluate the spelling performance of a Greek-speaking individual with the clinical phenotype of logopenic variant of PPA. Greek, unlike English, is an orthographically transparent language (Kambanaros & Weekes, 2013), and this is the first study to investigate spelling performance in PPA for Greek. Participant LE is a 67 years old man with 16 years of education. He was diagnosed with mild dementia (MMSE 21/30, CDR=1). Cognitive testing established relatively preserved memory and visuospatial abilities and the presence of an executive functioning deficit. Formal language assessment revealed a mild word finding impairment, mild difficulty repeating sentences, but spared single word comprehension. His speech was characterized as slow, with frequent pauses, hesitations and phonological paraphasias. There was no evidence of a motor speech disorder per se. LE was assigned a PASS Sum-of-Boxes score of 9 (Sapolsky et al., 2010), with writing being disproportionately affected in comparison to the other language domains. Procedures LE’s writing abilities were evaluated using the following experimental tasks: (i) copying of words and pseudowords, (ii) single phoneme to grapheme conversion, (iii) spelling to dictation of 2 letter nonsense words, 3 letter and 4 letter (CVCV) words and (iv) arranging scrabble tiles (4 letter CVCV words). Results Copying accuracy was similar for words and pseudowords (57.89 vs. 63.15%). Single phoneme to grapheme conversion accuracy was 48%. There was a significant effect of word length; he correctly spelled 47.05% of 2-letter vs. 15% of 3-letter and 5% of 4-letter words and copied 100% of 3-letter, 61.53% of 4-letter, 40% of 6-letter and 20% of >8-letter words. There was no significant difference between words of different grammatical class (nouns, verbs and closed class words). Almost all errors were phonologically implausible and included mainly substitutions, omissions, transpositions and mixed errors (e.g. ‘fao’-‘pa’). Furthermore, he made similar errors on pseudoword copying (e.g. ‘tapro’-‘patro’). Conclusions LE’s writing performance may be explained by a combination of two deficits. First, there is evidence of a graphemic buffer disorder based on the error types (substitutions, transpositions, omissions and additions) evident across different written modalities (writing to dictation, copying, arranging tiles) and that the number of errors increases as a function of word length. It is important to note that linguistic factors, such as lexicality, word frequency and grammatical class, do not appear to play a role in spelling accuracy. Second, there is evidence of phonological dysgraphia as LE has difficulty associating sounds with their corresponding graphemes. Increased lexicality effects in reading (95% for words vs 50% for pseudowords), which is the hallmark feature of phonological dyslexia, supports the presence of phonological dysgraphia.