| Abstrakt: |
Background: Lack of practice effects (PE) in common cognitive tests has been associated with increased risk of dementia. Previous studies have found connections between recall of proper names (PN) and amyloid positivity in cognitively unimpaired adults at risk for Alzheimer's Disease (AD; Mueller et al., 2020). For several language‐related tests, including PN, we compared PE modeling options and then examined whether amyloid status moderated PE. Method: Participants with language tasks and C‐Pittsburgh Compound‐B (PiB) amyloid positron emission tomography (PET) were selected from the Wisconsin Registry for Alzheimer's Prevention (WRAP; n = 377). PE operationalizations included: number of language task exposures – 1 ("PEnvis‐1"), the square root of PEnvis‐1 ("PEsqrt") and categorical PE ("PEcat"; reference = 0 (baseline) vs 1, 2, or 3+ follow‐up assessments). Language outcomes variables included letter fluency (CFL), animal naming, and logical memory (LM). From LM, we utilized item‐level data to create immediate and delayed PN recall. We used linear mixed effects models to identify the best fitting PE operationalization for each outcome and ran subsequent models examining amyloid*PE interactions (covariates: age at assessment, gender, and WRAT‐III reading score; random effects for intercept). Result: Sample characteristics are shown in Table 1 overall and by amyloid status. AIC fit statistics were best overall for PEnvis‐1. In subsequent models, the PE*amyloid interaction was significant for LM Total Score (β = ‐0.08, 95% CI = (‐0.15, ‐0.01), p = 0.02) and PN immediate recall (β = ‐0.22, 95% CI = (‐0.38, ‐0.05), p = 0.01). For both outcomes, predicted Z‐scores increased with more practice for amyloid positive participants, but decreased for amyloid negative participants (Table 2; Figure 1). The main effect of PE was significant for CFL (β = 0.12, 95% CI = (0.08, 0.15 p<0.001) and Animal Naming (β = 0.07, 95% CI = (0.03‐0.11), p = 0.002). Conclusion: PEnvis‐1 was generally the best‐fitting PE operationalization. In our relatively young sample (baseline mean age = 54.1; mean years between baseline and most recent LM = 9.12), improvements with practice for animal naming and CFL were not modified by amyloid status. In contrast, those with amyloid negative biomarkers showed practice‐related improvements on LM while those with positive amyloid showed decreased performance with more practice. [ABSTRACT FROM AUTHOR] |
