Linguistic metrics of semantic fluency relate to longitudinal cortical thinning in older adults without dementia: The SMART‐MR study

Background The need persists for sensitive, low-cost, and high-access cognitive markers to complement biomarker information in the earliest stages of Alzheimer’s disease. Sensitive may be derived from at item level neuropsychological tests. We investigated if item-level metrics add beyond total score relationship semantic fluency—generating as many words possible a category given time frame—to cortical thinning across eight years middle-aged older adults without dementia. Method Participants were part SMART-MR cohort (n = 251, age 59.6±8.6, 19.5% women), designed investigate brain changes individuals with arterial disease, which puts them high risk Global thickness was MRI scans two visits approximately apart. Item-level linguistic databases: mean log lexical frequency (how often word occurs our language) acquisition (AoA; what learned) generated on 2-minute (animal) fluency task (Fig 1). calculated residualized scores representing deviation metric is expected based account collinearity. Linear regression models—adjusted age, sex/gender, education, history stroke—analyzed between or determinants follow-up outcome, adjusting baseline level. Result Total alone not related (β -.023[-.001;.000], p .466). However, higher-than-expected -.065[-.054;-.002], .033) lower-than-expected AoA .066[.001;.019], more follow-up. An alternative analytic approach capture change, using difference adjustment, yielded similar results. Conclusion Alternative fluency, opposed traditional score, future neurodegeneration high-risk dementia diagnosis. results seem robust different approaches. Valuable could leveraged accurate identification clinical addition evidence, intervention timely