Cognitive data harmonization in the ADRC Network and beyond-Past, present, and future.
Paul K Crane, Changye Li, Trevor Cohen, Aaron Seitz, Emma Rhodes, Seo-Eun Choi, Michael Lee, Shubhabrata Mukherjee, Connie Nakano, Samuel Albertson, Ali Asadollahi, Karen Velderrain-Lopez, Jeanne Gallée, Laura A Rabin, Leslie Gaynor
Abstract
Open AccessThe National Alzheimer's Coordinating Center (NACC) Uniform Data Set (UDS) includes extensive cognitive data contributed by Alzheimer's Disease Research Centers (ADRCs) since 2005. The UDS has evolved over time and is on its fourth iteration. In addition to this core dataset, individual ADRCs have administered augmented neuropsychological batteries to research participants that go beyond the UDS. Here we describe ongoing efforts and future plans to optimize use of these data with modern psychometric methods. Modern psychometric methods address challenges from an evolving battery of cognitive tests. To date, most efforts using modern psychometric methods have focused on harmonization and co-calibration of ADRCs' UDS and non-UDS cognitive data, while recent efforts address additional areas such as subjective cognitive impairment. Modern psychometric methods provide a workable framework for anticipated future developments, including digital cognitive testing and analyses of spoken responses. These methods facilitate optimal use of NACC's data riches to further its mission to facilitate collaborative research. HIGHLIGHTS: National Alzheimer's Coordinating Center (NACC) data collection includes an extensive cognitive battery that has changed over time. An ambitious project has harmonized and co-calibrated cognitive domain scores for memory, executive functioning, and language. Scores and their standard errors are available from the NACC. Those scores are co-calibrated with domain scores from many additional studies, facilitating cross-study scientific investigation. Future opportunities include digital data collection, consideration of neuropsychiatric symptoms and subjective cognitive impairment, and other uses of the granular cognitive data.