While there exist over 7,000 living languages worldwide, speech and language studies across neurodegenerative diseases primarily concentrates on fewer than twenty Germanic and Romance languages, with English being the most extensively studied. The applicability of these findings to languages with unique linguistic attributes, such as Chinese languages, remains uncertain. This talk will introduce the Chinese Language Assessment in Alzheimer’s Disease (AD) and Primary Progressive Aphasia (PPA) (CLAP) projects, which examined various linguistically specific features in Chinese speaking PPA and AD patients, with special focus on semantic and phonemic fluency performance. In English, AD patients were noted to exhibit greater semantic fluency (SF) impairment compared to phonemic fluency (PF), attributed to ventral language pathway involvement. However, Chinese languages adopt logographic script where all phonetic units inherently carry semantic meanings. Naturally, there is a lack of clarity on accurately assessing and distinguishing PF from SF in Chinese languages. This research aims to elucidate the roles of various Chinese fluency tests in PPA and AD diagnosis.

This study included seventy-two cognitively normal participants (CN), twenty-nine individuals with mild cognitive impairment (MCI), six AD patient, and sixteen logopenic variant (lv) PPA participants, all of whom are native Cantonese and Mandarin speakers. Participants were assigned three PF and one SF tasks: (1) PF at phoneme level: voice Mandarin/Cantonese sounds with the PinYin consonant of /p/ and /t/, (2) PF at character level: verbally express compound words with the monosyllabic character “心” and “國” (3) PF at syllable level: write characters with the /si/ sound, and (4) SF: produce words that belong to the semantic categories of “four-legged animal” and “vegetable”. ANOVA analysis was employed for group comparison, and Pearson correlations were conducted to explore the association between PF and a semantic association test.

In terms of SF test, participants with MCI, AD and lvPPA generated fewer animal and vegetable words than cognitively normal participants (animal: p< .001; vegetable: p< .001). Compared to CN individuals, participants with AD and lvPPA produced significantly fewer words and characters in PF at syllable and character level (Syllable: M_CN =7.3+/- 2.3, M_MCI =5.5+/-2.5, M_AD=5.0+/-1.5, M_lvPPA =4.1+/-1.7, p< .001) (Character “心”: M_CN=12.1+/- 3.5, M_MCI=10.6+/-4.2, M_AD=5.0+/-1.4, M_lvPPA=5.0+/-2.7, p< .001; “國”: M_CN=8.9+/-2.9, M_MCI=6.8+/-2.2, M_AD=4.5+/-2.1, M_lvPPA=5.2+/-2.6, p<.001). Conversely, no significant differences were noted in PF at phoneme level. (/p/: p=0.108; /t/: p=0.531). PF performance at character level were significantly associated with semantic association test (“心”: r=0.243, p=0.042; “國”: r=0.289 p=0.015), whereas no such association was noted for PF at phoneme level.

The findings suggest that SF, PF at character and syllable level tests are noticeably affected in patients with AD and lvPPA, whereas PF at the phoneme level appears to be less impacted. The distinct variations observed in various phonological fluency tests likely stem from the semantic involvement in Chinese compound word and character generation. This underscores the role of linguistic typology in dementia symptomatology, emphasizing its significance for equitable care in diverse populations.