Optimizing Cognitive Assessment in DIAN with Smartphone-based burst testing

点击翻译按钮获取中文摘要

The Ambulatory Research in Cognition (ARC) study will leverage the increasing popularity of smartphones to improve upon standard in-clinic cognitive testing, providing a robust characterization of cognition in participants enrolled in the Dominantly-Inherited Alzheimer Network (DIAN) study. One of the most important and face-valid indicators of Alzheimer’s disease (AD) is a change in cognition, but assessing cognition in-clinic has several drawbacks. First, performance is influenced by day to day fluctuations in stress, fatigue, sleep patterns, and mood. Second, the testing takes place in environments that are fundamentally different from where cognitively demanding tasks are performed in daily life. Finally, by design, cognition is typically assessed very infrequently, usually once per year or every two years (as is the case in DIAN). One solution would be to increase the frequency of in-clinic testing, but this is impractical and burdensome, especially for DIAN participants. DIAN participants come from families that have a ~50% chance of inheriting an extremely rare autosomal dominant causal mutation for AD. Carriers of these mutations become symptomatic between 30 and 50 years of age, decades earlier than the more common sporadic form of AD. Most are still working and many are primary caretakers for their young children and often for affected family members, therefore traveling to sites more frequently for assessments is extraordinarily burdensome. The ARC study will provide a solution to these difficulties by assessing cognition with an iOS and Android app installed on ~248 DIAN participants’ personal smartphones. Every three months, participants will complete extremely brief (<3 minutes each) testing sessions 4x/day over the course of one week. Tests are averaged across the week to provide a score that captures and normalizes natural variability and dramatically increases reliability. Pilot studies show that ARC assessments demonstrate extraordinary reliability, ranging from 0.86 to 0.98. Another advantage is the ability to measure variability within a day, across a week, and across the 3- month intervals of assessments. We hypothesize that ARC assessments will be more sensitive than in-clinic assessments to disease stage and AD biomarkers and that mutation carriers in DIAN will show more variability in cognitive performance than non-carriers, even in the preclinical stages of disease. If successful, the increases in sensitivity and reliability of ARC assessments will provide extraordinary statistical power to characterize cognitive decline in observational studies of AD. In addition, ARC assessments could benefit clinical trials by shortening trial duration and requiring fewer participants.