Project Summary
The Alzheimer’s Association’s 2019 Facts and Figures Report reports that while a cognitive assessment is
required for all Medicare Annual Wellness Visit, only 16% of surveyed older adults reported having any form of
memory assessment. While standardized neuropsychological tasks are sensitive to gross memory deficits, they
are not as sensitive to milder deficits associated with healthy aging or early disease stages and they are often
poorly suited for routine testing outside of trained neuropsychologist practitioners. Thus, there is a critical need
for an easy-to administer, reliable, and sensitive measure of hippocampal memory function for clinical use. We
have designed the Mnemonic Similarity Task (MST), a modified object recognition memory task, to provide not
only a traditional measure of object recognition memory, but also a measure of “mnemonic discrimination” that
is highly sensitive hippocampal function by placing strong demands on pattern separation. Thus, the goal of this
proposal is to complete the development and validation of version of the clinical MST (cMST) as a fully
encapsulated tool that would be ready for use in clinical research and for evaluation as a routine clinical tool.
First, we will develop an optimized version of the MST designed for clinical use as a sensitive assay of
hippocampal function that rapidly, but robustly, estimates hippocampal memory performance. The design goals
are that it is: 1) easy to administer in a short amount of time, 2) has clear scoring and interpretation of results, 3)
is sensitive to modest hippocampal dysfunction, 4) demonstrates reliability, and 5) exhibits general utility across
a range of racial/ethnic/age groups. At the core of the MST is the use of highly similar lure items that have a
range of pre-determined “mnemonic similarity” to the original studied item. We have now developed a version of
the MST that uses a continuous recognition memory format and optimizes the distribution of targets, foils, and
lures, along with the mnemonic similarity of those lures, for use in the cMST. Here, we will determine whether
that is an ideal format or whether a Bayesian adaptive version is superior.
Then we will establish normative data on the cMST across the lifespan and in early dementia. In Aim 2.1, we will
evaluate the sensitivity and validity of the cMST to memory decline in healthy aging, comparing it to the standard
research-grade MST and to traditional neuropsychological tests in a large lifespan sample. Repeat testing of
individuals will be used to establish reliability and assess practice effects. We will use these data to create
normative data. In Aim 2.2, we will extend these investigations to clinical populations - specifically to those with
Mild Cognitive Impairment or early AD through the UCI Alzheimer’s Disease Research Center (ADRC) to
determine both viability and normative performance of our measures in these impaired populations and
diagnostic ability of the cMST. By working with our ADRC for testing both healthy and impaired individuals, we
will gain access to a host of other existing biomarkers of hippocampal decline available in this cohort, such as
word recall, CSF tau levels, and hippocampal volume measures via structural MRI.
Finally, we will create a model for large-scale distribution of the cMST via online administration. The ultimate
goal of this proposal is to create an encapsulated version of the cMST that can be adopted for clinical use that
is independent of a research lab. In Aim 3, we will distribute this task via a downloadable link that can be used
on either desktop/laptop devices or on touchscreen tablets to a large and diverse sample. UCI’s large Consent
2 Contact (C2C) database (N=3250) provides an outstanding test case for large-scale data collection outside of
a clinical or lab-based setting and for testing its potential use as a diagnostic screening tool.