A - 131 Patterns of Functional Decline in Alzheimer Biomarker-Positive Vs. Negative Cognitive Impairment

Abstract Objective To contrast longitudinal patterns of functional decline between groups of cognitively impaired individuals with or without biomarker support for Alzheimer’s disease (ad). Method We requested data from the National Alzheimer’s Coordinating Center on individuals who underwent serial evaluation with the Functional Activities Questionnaire (FAQ) and those for whom cerebrospinal fluid biomarkers for ad were available. By cross-referencing the two datasets, we identified 570 subjects, 71 of whom were biomarker positive and 499 biomarker negative. Item-level FAQ item-level scores were then entered as predictor variables into lagged mixed-effects regression models with a single FAQ item as the dependent variable in each model (one model for each of the ten FAQ items). Coefficients for statistically significant predictors of follow-up FAQ items were plotted as arcs on a directed graph, with one vertex for each of the ten FAQ items. We repeated this process for biomarker positive and negative individuals and created a map depicting effect size differences, i.e., biomarker positive minus biomarker-negative regression coefficients (Δ = β + − β-) for statistically significant arcs (α = 0.05). Results Twelve out of 21 arcs that were significant in the biomarker-positive individuals were significant only for this group. The largest effect size difference was STOVE → STOVE (Δ = 0.233). Twenty of 29 arcs that were significant in the biomarker-negative individuals were significant only for this group. The largest effect size difference was GAMES → GAMES (Δ = −0.168). Conclusion ad biomarker status modulated the patterns of longitudinal functional change in this sample.