A - 16 A Proof-of-Concept for the Construct Validity of Cognitive Dispersion: Longitudinal Relationship between Cognitive Dispersion and Preclinical Alzheimer’s Cognitive Composite Scores

Abstract Objective To examine the construct validity of cognitive dispersion (CD) by examining its relationship with the Preclinical Alzheimer’s Cognitive Composite (PACC) scores across multiple visits among cognitively healthy individuals with hypertension and memory concerns/significant family history of dementia. Methods 435 participants (Mage = 68.90, Meducation = 16.29 years) from the rrAD study with neurocognitive data from five visits over 24 months were included. PACC total scores were derived as per literature from the MMSE, Free and Cued Selective Reminding Test, Digit Symbol Substitution Test, and Logical Memory delayed recall. Raw scores were converted to z-scores using the baseline sample mean and standard deviation (SD) and calculating the mean of these z-scores to form the PACC score. CD was calculated as the intraindividual SD of the z-scores of all tests within the PACC battery. Linear regression, with time representing the five time points as a predictor, examined the relationship between cognitive dispersion and PACC scores. Results The model significantly predicted PACC scores (F[2, 2278] = 65.83, p < 0.001, R^2 = 0.05). Time alone did not significantly predict PACC scores (p = 0.41), but CD was a significant predictor when controlling for time (b = −0.51, SE = 0.04, t = −11.47, p < 0.001), such that greater CD was related to lower PACC composite scores. Conclusion The findings suggest an inverse relationship between CD and PACC scores over time, supporting CD as a potential marker of preclinical cognitive change in this sample. Future research may explore how lifestyle interventions influence this relationship while considering demographic factors such as age, sex, and race/ethnicity.