Date of Award
Doctor of Philosophy (PhD)
Preclinical Alzheimer disease (AD) is characterized as the point at which a person is clinically normal but exhibits AD-related neuropathological change and is associated with developing AD-related dementia in the future (Jack et al., 2018). As such, there has been a focus on identifying procedures sensitive to the preclinical stage. However, the current methods used to detect biomarker abnormalities associated with preclinical AD (e.g., lumbar puncture, PET scan, and MRI) are invasive and/or expensive, which limits feasibility for widescale screening. This leaves the need to develop a non-invasive, time-efficient, and cost-effective screening measure to identify those who are at greater risk of preclinical AD. Such a measure could be used to inform decisions regarding when to use more invasive and/or expensive methods.
Preclinical AD has been associated with subtle, but observable, changes in performance on neuropsychological and experimental measures of memory, attention, and spatial navigation (Allison et al., 2016; Balota et al., 2020; Hedden et al., 2013; Langbaum et al., 2014; Levine et al., 2020; Millar et al., 2017). Unfortunately, these tasks can be time-consuming, which limits their feasibility in clinical settings. The goal of this study is to examine self- and informant-reported questionnaires assessing changes in these cognitive domains to identify a questionnaire-based screening measure for preclinical AD that could be easily administered in clinical settings.
This dissertation comprised three independent samples, two recruited from Washington University using the Volunteer for Health (VFH) program and Alzheimer Disease Research Center (ADRC) and one including of preexisting data from the Alzheimer’s Disease Neuroimaging Initiative (ADNI). The project had four specific aims to examine the diagnostic abilities of self- and informant-reported measures of memory, attention, and spatial navigation in preclinical AD. The first aim was to assess the reliability (internal consistency) and validity (confirmatory factor analysis) of the six questionnaires (VFH and ADNI samples). The second aim was to compare the diagnostic accuracy of the six questionnaires using receiver operating characteristic analyses (ADRC and ADNI samples). The third aim was to assess the predictive ability of the questionnaires when controlling for other factors associated with preclinical AD (personality, depressive symptoms, and anxiety symptoms in the ADRC sample and depressive symptoms in the ADNI sample). The final aim was to compare the diagnostic accuracy of the six questionnaires with a previously established measures of cognition (e.g., a self and informant measure of early dementia in the ADRC sample or with neuropsychological composites in the ADNI sample).
All questionnaires demonstrated appropriate reliability and validity. In both samples, self-reported questionnaires were significant predictors of preclinical AD, whereas informant-reported questionnaires were not. Additionally, self-reported attention remained a significant predictor of preclinical AD when controlling for depressive symptoms in the ADNI sample. Of note, the ADRC sample was underpowered based on a priori power analysis. In addition, although significant, the self-reported ADNI questionnaires demonstrated weak diagnostic accuracy in predicting preclinical AD (area under the curve=.564-.592). Given these limitations, it is unclear whether these questionnaires would be appropriate for widespread clinical use.
Although this study was unable to identify a questionnaire that was both diagnostically accurate and highly sensitive to preclinical AD, the results serve to inform the future development of screening tools for early AD pathological change. These results provide an important foundation for the future development of cognitive screening tools in the preclinical stage.
Chair and Committee
Andrew Aschenbrenner, David Balota, Thomas Rodebaugh, Jeffrey Zacks,
Levine, Taylor Fama, "Comparison of self- and informant-reported change in memory, attention, and spatial navigation in predicting preclinical Alzheimer disease" (2023). Arts & Sciences Electronic Theses and Dissertations. 2875.