PROJECT SUMMARY / ABSTRACT
New submission for PAS-19-391 (Small Research Grant Program for the Next Generation of Researchers in
AD/ADRD Research: Area of Focus Archiving and Leveraging Existing Data Sets for Analyses [R03]).
Recent statistics show that if everyone alive in 2018 who will develop Alzheimer’s disease (AD) had early
and accurate diagnosis, there would be a savings of $7.9 trillion in medical and long-term care costs. Although
the advent of Alzheimer’s disease (AD) biomarkers has revolutionized our understanding of AD pathogenesis
during the preclinical phase, these approaches are often expensive, invasive, inaccessible due to rural location,
cost, or medical contraindications. Additionally, beyond a focus on AD biomarkers alone, it is essential to
emphasize that cognitive difficulties and subsequent functional impairment are the features of the disease that
negatively impact the lives of patients and their families. Emerging evidence suggests that subtle cognitive
changes may develop much earlier than originally described in models of AD and these subtle cognitive changes
add meaningful prognostic value, above and beyond AD biomarkers, in predicting progression to mild cognitive
impairment (MCI) and dementia. The gold standard approach for identifying subtle cognitive decline in preclinical
AD, however, remains unclear. Therefore, we propose to apply four different classification algorithms for subtle
cognitive decline in the open source Alzheimer’s Disease Neuroimaging Initiative (ADNI) dataset. The four
classifications methods include two subjective approaches: self-reported subjective cognitive decline (Self-SCD)
and informant-reported subjective cognitive decline (Inform-SCD); and two objective approaches: a sensitive
neuropsychological individual test-based approach called objectively-defined subtle cognitive decline (Obj-SCD)
and a composite score-based approach using the preclinical Alzheimer’s cognitive composite to identify subtle
cognitive decline (PACC-SCD). The specific aims of this proposal include: 1) Compare AD biomarkers across 4
subtle cognitive decline definitions (Self-SCD, Inform-SCD, Obj-SCD, and PACC-SCD); 2) Determine which
subtle cognitive decline definitions best capture objective cognitive decline in the year preceding the subtle
cognitive decline classification; and 3) Examine longitudinal a) clinical and b) biomarker progression across
subtle cognitive decline definitions to determine the definitions with the best predictive utility. Results of the
proposed aims will likely impact the design of future studies, as having simple, yet reliable, and highly cost-
efficient methods for narrowing the initial pool of participants so that only those at greatest risk require a PET
scan for screening purposes would result in significant cost-savings. Additionally, these results will serve as
critical preliminary data for future grant applications, and be a key step toward improving clinically meaningful
early detection methods of those at risk for future decline, particularly those with limited access to AD biomarker
testing.