Stricker Learning Span: A Computer Adaptive Word List Memory Test Optimized for Remote Assessment

PROJECT SUMMARY/ABSTRACT Due to COVID-19, remote cognitive assessment has transitioned from an important research aim to an immediate and urgent research and clinical need. This need has underscored the lack of well-validated, sensitive, reliable, and well-normed tests available for remote assessment. Word list memory performance changes early in the preclinical phase of Alzheimer?s disease (AD), up to 20 years prior to onset of mild cognitive impairment (MCI). Although delayed word list recall is viewed as the most sensitive memory measure in studies of MCI and AD dementia, traditional word list memory test paradigms require interactive administration with an examiner. Further, recency effects may influence the sensitivity of delayed word list recall: correct recall of words at the end of the list during immediate recall trials may represent attentional span/working memory for some individuals. When these words are subsequently not recalled at delay, they are considered forgotten, but an alternate view is that the words were never fully encoded during repeated learning trials. In line with this, the central deficit in AD may be one of learning. To help address the critical need for a sensitive and brief remote learning measure, we have developed a new word list learning paradigm. The Stricker Learning Span (SLS) aims to achieve the known sensitivity of word list memory tests to the earliest preclinical AD changes while allowing reliable remote self-administration on a smartphone or alternative device (e.g., PC, tablet) through a web-based platform. This flexible approach is necessary to reach the most users and eliminate potential health disparities. We have transformed the traditional manner of administering a word list memory test in several important ways, resulting in a novel paradigm that takes full advantage of computer- based administration through use of a computer adaptive testing procedure focused on learning, visual presentation of the word list, 4-choice recognition to assess words learned, and a complete data capture approach to allow future studies focused on process-based measures and application of machine learning techniques. This study will leverage the resources of the ongoing population-based Mayo Clinic Study of Aging, which already incorporates in-clinic and remote computerized cognitive assessment. Specifics aims include to (1) determine the acceptability and efficiency of the unsupervised SLS, (2) demonstrate the construct validity of the SLS by examining association with a traditional word list memory measure, (3) determine the criterion validity of the SLS through sensitivity to preclinical AD (cross-sectional effect sizes between cognitively unimpaired amyloid PET positive and negative groups) and associations with continuous AD neuroimaging biomarkers (tau PET, hippocampal volume) in all participants with neuroimaging data, and (4) determine 7.5 month test-retest reliability for the SLS. Sensitive, remotely administered computerized measures will help to identify individuals with preclinical AD for anti-amyloid and anti-tau clinical trials, as well as individuals meeting criteria for treatment once anti-amyloid or anti-tau therapies are approved.