[Features of a modified WHO/UCLA AVLT performance in amnestic mild cognitive impairment and mild Alzheimer's disease].

Objective: To compare the features of a modified WHO/UCLA AVLT performance in the cognitive normal, amnestic mild cognitive impairment (aMCI) and mild Alzheimer's disease (mild AD) patients. Method: A total of 105 cases of cognitivenormal (CN), 48 aMCI and 50 mild AD patients were included between 2016 and 2018. All subjects undertook detailed neuropsychological tests and brain MRI/CT scan. Results: The total score of five learning trials in CN, aMCI and AD groups were 53.9±6.9, 34.6±8.3 and 23.7±6.2, respectively (P<0.001). The score of 20-min delay recallwere 12.5±1.6, 4.3±3.0 and 0.6±1.0, respectively (P<0.001) in three groups. The score of cued recall were 13.0±1.4, 7.0±2.4 and 2.6±2.0, respectively (P<0.001). The score of 20-min delay recall had the largest effect sizes between CN and aMCI groups (Cohen'd=3.8, 95%CI,3.3-4.4), and CN and mild AD groups (Cohen'd=8.1, 95%CI 7.1-9.1). Cued recall had the largest effect size between aMCI and mild AD groups (Cohen'd=2.04, 95%CI 1.5-2.5). The scores of learning total score, 20-min delay recall, cued recall and recognition had the strong relationships with the scores of mini-mental state examination (MMSE) and Montreal cognitive assessment (MoCA) , but obtaining moderate relationships with Boston naming test and trail making test (TMT) and weak relationships with digit span and figure copy. Age and education had no relationship with the main indices of this modified AVLT. Conclusions: The modified WHO/UCLA AVLT is still an age and education fair test to assess memory domain function. Qualitative analysis of AVLT profiles may be useful to differentiate the CN, aMCI and mild AD in Chinese sample.