Introduction: Semantic memory is impaired in mild cognitive impairment (MCI). Twomain hypotheses about this finding are debated and refer to the degradation of stored knowledgeversus the impairment of semantic access mechanisms. The aim of our study is to evaluate semanticimpairment in MCI versus healthy subjects (HS) by an experiment evaluating semantic priming.
Methods: We enrolled 27 MCI and 20 HS. MCI group were divided, according to follow up, intoconverters-MCI and non converters-MCI. The semantic task consisted of 108 pairs of words, 54 ofwhich were semantically associated. Stimuli were presented 250 or 900 ms later the appearance ofthe target in a randomized manner. Data were analyzed using factorial ANOVA.
Results: Both HSand MCI answered more quickly for word than for non-word at both stimulus onset asynchrony(SOA) intervals. At 250 ms, both MCI and HS experienced a shorter time of response for relatedwordthan for unrelated words (priming effect), while only the converters-MCI subgroup lost thepriming effect. Further, we observed a rather larger Cohen's d effect size in non converters-MCIthan in converters-MCI.
Conclusion: Our data, and in particular the absence of a semantic primingeffect in converters-MCI, could reflect the impairment of semantic knowledge rather than theaccessibility of semantic stores in MCI individuals that progress to dementia.
Keywords: Alzheimer disease; mild cognitive impairment; semantic priming.