CogState computerized testing is associated with amyloid and FDG-PET: Implications for secondary preventive trials

Abstract

Inexpensive and non-invasive screening tools of significant brain amyloidosis and neurodegeneration are needed for secondary preventive Alzheimer's disease (AD) trials enrolling cognitively normal (CN) individuals. We examined the association between performance on the CogState computerized neuropsychological battery and MR, FDG-PET, and PiB-PET. We studied 153 Mayo Clinic Study of Aging participants who were cognitively normal (CN) (aged 51–70) and had undergone MRI (n=152), FDG PET (n=120), and/or PiB PET (n=121) and the CogState computerized assessment within a 5 month period. The following CogState tests were administered: Detection (DET) - measuring simple reaction time, Identification (IDN) - measuring choice reaction time, One Card Learning (OCL) - measuring visual learning, One Back test (ONB) - measuring working memory and attention, and the Groton Maze Learning Test (GMLT) - measuring spatial working memory. Linear regression was used to assess the relationship between each test (average reaction time for all tests and also accuracy on OCL and ONB) and continuous measures of PiB PET SUVR, FDG PET metabolism in an “AD signature region”, and hippocampal atrophy, controlling for age. A slower reaction time for DET (p=0.005), IDN (p=0.02), and OCB (p=0.02) was associated with lower FDG PET metabolism. A faster response time (i.e., more moves per second) on the GMLT (p=0.009) was the only test associated with larger hippocampal volumes. While elevated amyloid (PIB SUVR>1.5) was uncommon in this age range, an intriguing finding that needs to be borne out by additional data was that high amyloid trended towards an association with slower reaction times for ONB (p=0.08), OCL (p=0.05), and GMLT (p=0.002). There were no associations between accuracy on the OCL or ONB and neuroimaging measures. These results suggest that the CogState computerized assessment may be a useful indicator of neurodegenerative pathology among young CN individuals. While additional data are needed, especially among persons with higher amyloid burden, there were also trends between slower reaction times on memory-related measures and higher amyloid. CogState is brief (∼15 minutes), can be administered in the home without the need for an in-person visit, and may be a useful screening tool for enhancing clinical trial enrollment of persons harboring preclinical AD biomarkers.