RESUMO
Objectives: This study investigated whether employing a phonological or semantic strategy elicited a better performance on a letter fluency task for people with Alzheimer's disease (AD). Methods: Sixty participants with probable AD were extracted from the DementiaBank database. After applying exclusion criteria, 47 participants were included in the final analysis. We used phonological and semantic strategies to analyze participants' responses to the letter fluency task. The phonological strategy analysis was based on the number of switches and the mean cluster size, and the semantic strategy analysis was based on semantic relatedness, which quantified word-similarity change by adapting the concept of persistence length from analyses of DNA and protein structures. We employed Pearson correlation coefficients to determine whether any strategy indexes were significantly related to the number of correct responses and used stepwise multiple regression analyses to determine the best predictor. Results: Participants who relied on phonological strategy performed better on the letter fluency task. The number of correct responses was significantly positively correlated with phonological strategy but significantly negatively correlated with semantic strategy. The number of switches, mean cluster size, and semantic relatedness were all significant predictors, explaining 68.1% of the variance. Conclusion: Our results suggested that individuals with AD who engaged in phonological strategy performed better on the letter fluency task than those who relied on semantic strategy.