Item‐level verbal fluency psycholinguistic metrics predict preclinical Alzheimer's disease
Michael A. Sugarman, Mackenzie Stabile, David G. Clark, Paul J. Nietert, Federico Rodriguez‐Porcel, Justin Bushnell, Andreana BenitezAbstract
INTRODUCTION
Preclinical Alzheimer's disease (pAD) is characterized by subtle disruptions in semantic network functioning. Psycholinguistic metrics from item‐level responses during verbal fluency tasks could reflect these changes in the asymptomatic stage.
METHODS
We analyzed psycholinguistic metrics for Animal and F Fluency (age of acquisition [AoA], semantic neighborhood density [SND], and lexical frequency [LF]), testing whether these metrics differentiated 50 participants with positron emission tomography (PET)‐confirmed pAD (Aβ+) from 99 controls (Aβ‐) and predicted cognitive decline.
RESULTS
Aβ+ participants generated Animal Fluency responses with younger AoA, higher LF, and marginally higher SND than Aβ‐ ( d s = 0.33‐0.44), despite comparable total scores. All three metrics predicted amyloid positivity in models controlling for age and total score. No significant findings were observed for F Fluency, and no metrics predicted cognitive decline.
DISCUSSION
Psycholinguistic metrics for Animal, but not F Fluency responses may reflect changes in semantic network functioning in pAD not captured by the total score.