Comparison of latent cognitive reserve markers and a latent marker score in Alzheimer's disease. (31st December 2021)

• Record Type: Journal Article
• Authors: Ersoezlue, Ersin; Rauchmann, Boris‐Stephan; Perneczky, Robert
• Is Part Of: Alzheimer's & dementia. Volume 17(2021)Supplement 5
• Page Start: n/a
• Page End: n/a
• Publication Date: 2021-12-31
• Subjects: Alzheimer's disease -- Periodicals; Alzheimer Disease -- Periodicals; Dementia -- Periodicals; Démence; Maladie d'Alzheimer; Périodique électronique (Descripteur de forme); Ressource Internet (Descripteur de forme); 616.83
• DOI: 10.1002/alz.058604
• Languages: English
• ISSNs: 1552-5260

Abstract

Abstract: Background: The cognitive reserve (CR) hypothesis explains inter‐individual differences in the association between observed cognitive function and estimated neurodegenerative burden. Latent CR marker approach were suggested as a promising method to estimate cognitive reserve. We aimed to compare different CR marker approaches and test a clinical latent CR marker score. Method: Amyloid‐β (Aβ)‐negative and cognitively normal participants were included as healthy controls (HC) and Aβ‐positive and clinically impaired participants were included as AD spectrum (ADS) from the AD Neuroimaging Initiative (ADNI) study as two separate CSF data‐only discovery and a PET data‐only replication cohorts. A latent CR marker factor (CRM‐Factor) was obtained as the latent global cognitive domain score (gCDS) through a principal component analysis remaining after accounting for neuropathological burden and demographic variables. A residual CR marker (CRM‐Residual) was calculated as residualized gCDS in a multilinear regression analysis. We then estimated a CR marker score (CRM‐Score) as predictions from a multiple regression model with CRM‐Factor as dependent variable and demographics, total Mini‐Mental State Examination score and binarized biomarker status as predictors. Markers have been derived at baseline. We tested the moderation of each CR marker on the individual trajectories of memory decline in separate linear mixed models. Result: Higher years of education were predicted by the CR markers when tested separately. CRM‐Residual levels were strongly associated with CRM‐Factor, while CRM‐Residual was moderately associated with CRM‐Factor and CRM‐Score. In linear mixed models, patients in ADS with higher CR revealed a slower memory decline, mainly in preclinical AD. Conclusion: Our results support the operational value and effectivity of the latent CR marker approach. We defined a novel latent CR Marker Score that may be operationalized in independent cohorts without conducting a data‐driven analysis. In line with previous studies, latent CR markers estimated differences in cognitive trajectories predominantly in the MCI stage. Our marker score approach can offer a standardized clinical scoring for simple, cohort‐independent and effective estimation the individual CR in clinical and epidemiological studies, making a better individual risk assessment and estimating therapeutic effects to strengthen the CR.