F4-01-02: increasing the impact of the ideas study using brain health registry online data collection

We evaluated feasibility and validity of remotely-collected online data in older adults with mild cognitive impairment (MCI) and dementia. Valid online methods have the potential to facilitate efficient identification of older adults with and at risk for Alzheimer's disease (AD). Imaging Dementia–Evidence for Amyloid Scanning (IDEAS) study participants were invited to join the Brain Health Registry (BHR). Participants completed online Cogstate Brief Battery (CBB); and self- and study partner-report Everyday Cognition scale (self-ECog and SP-ECog) as a measure of subjective decline. BHR and IDEAS study data (β-amyloid PET results and clinical diagnosis) were linked. We compared demographic and cognitive profiles of general BHR older adults and IDEAS participants, and analyzed agreement and associations between online variables and clinically-confirmed diagnosis or β-amyloid (Aβ) PET results. Compared to the general BHR cohort of older adults, IDEAS participants were older with more males and more participants with MCI and AD. Completion rates of CBB were lower for IDEAS versus general BHR (Table 1). For CBB and ECog, general BHR cognitively-unimpaired participants performed better than IDEAS MCI, who performed better than IDEAS AD (p<0.001). In IDEAS participants with linked clinical data (n=575, Table 1), agreement between BHR self-report and clinical data was 97% for Aβ, 60% for MCI diagnosis, and 72% for AD diagnosis. Agreement between BHR SP-report and clinical data was 98% for Aβ, 61% for MCI diagnosis, and 64% for AD diagnosis. Aβ+ participants performed worse on CBB across cognitive domains (p = 0.01 to 0.05), and had worse SP-ECog memory scores (p = 0.003). Clinical diagnosis was associated with CBB scores and SP-report ECog (p<0.001 for all). There were no associations between self-report ECog and any outcome measure.