ORCID

Abstract

Objective: To improve early detection of transitional cognitive decline associated with stage 2 preclinical Alzheimer’s disease (AD), this study evaluated three statistical approaches: Reliable Change Index (RCI), Standardised Regression-Based (SRB) models, and Multivariate Base Rate (MVBR) analysis for identifying Alzheimer’s disease related cognitive decline over a 12-month interval. Method: A longitudinal cohort of older adults (N = 91), including healthy controls, individuals with Subjective Cognitive Decline (SCD), and diagnosed AD patients, completed a neuropsychological battery at two time points. Test–retest reliability and practice effects were estimated, followed by calculation of RCI, SRB, and MVBR indices at both 5th and 16th percentile thresholds. Diagnostic accuracy was evaluated using sensitivity, specificity, and ROC analyses. Results: Most cognitive measures demonstrated adequate 12-month retest reliability. The longitudinal assessment of individual cognitive measure, such as the phonemic fluency test (FAS) achieved outstanding diagnostic accuracy for classifying transitional cognitive decline. However, the base rates of cognitive decline when an assessment battery is administered indicates that it is ‘normal’ for healthy individuals to experience a single test result suggestive of cognitive decline. To correct for this, a multivariate base rate approach was used. When cognitive decline was classified using SRB at the 16th percentile, a multivariate approach achieved excellent classification accuracy for identifying cognitive decline associated with Alzheimer’s disease. Normative change data and base rate distributions were generated to facilitate clinical interpretation. Conclusion: The results of this research provide important psychometric information that neuropsychologists can use to accurately classify transitional cognitive decline. While further research is needed to fully validate this approach, the findings of this research have important implications for practice. To support the implementation of this research into practice, a clinician-facing Excel calculator was developed to aid practical implementation in memory clinic settings.

Awarding Institution(s)

University of Plymouth

Supervisor

Jonathan Scott, Donnchadh Murphy, Rupert Noad

Keywords

Alzheimer Disease, Preclinical Alzheimer’s Disease, Longitudinal Cognitive Change, Neuropsychological Tests, Reliable Change Index (RCI), Subjective Cognitive Decline (SCD), Sensitivity and Specificity, Practice Effects, Multivariate Base Rate Analysis (MVBR), Standardised Regression-Based Models (SRB)

Document Type

Thesis

Publication Date

2025

Embargo Period

2026-09-18

Deposit Date

September 2025

Creative Commons License

Creative Commons Attribution-NonCommercial 4.0 International License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License

Download

Share

COinS