Analytic Core (Core D)

SUMMARY: CORE D- ANALYTIC CORE. The rigorous integration, analysis and interpretation of a wide array of data from the Cores and Projects is central to the Harvard Aging Brain Study (HABS). The Data Core provides the database and statistical support for these functions. The Core has developed and optimized: (i) the DataCentral database for data entry, quality assurance, storage, backup, security, and query systems, as well as linkage to raw and processed image files, (ii) data distribution for biostatistical analyses to support the project aims, and (iii) blinding to prevent inadvertent bias. The Core also supports, designs, and implements analyses that contribute to the definition and quantification of critical disease pathologies, brain networks, biomarkers, and their interactions, and participates in the selection of subsets of the sample for additional procedures or specific analyses, as needed. We also develop summary measures of brain pathologies and network disruption that can be used in a wide variety of analyses within and across projects, address methods needed for time-to-event and longitudinal analyses across multiple nonlinear processes within the aging brain, and develop models to address questions of causal inference. The Specific Aims of the Data Core are: 1. Data Integrity and Integration, which includes data entry, quality control, and storage; data access and distribution; and data sharing; 2. Statistical Support, Oversight, and Analysis, which involves statistical support for the Projects and Cores including simulation-guided study design and data analysis, and use of modern epidemiologic methods in causal inference to inform model construction, mediation analyses, and adjustment for selection bias and confounding, and 3. Development of Innovative Statistical Models and Methods, which includes joint modeling of PiB, FTP, functional connectivity and cognitive measures; treatment of misaligned biomarker measurements; modeling of censored time-to-event data that adjusts for delayed entry; derivation of efficient prevention trial populations and endpoints; and development of new methods for adjusting for selection bias in the setting of change-in-change analyses. Overall, the overarching responsibilities of the Data Core are to: manage the large and complex array of data that continues to be collected in the project, including neuroimaging data, clinical data, genetic, and biofluid data; integrate all data elements into a single database; and oversee and facilitate the statistical analyses for each Project--and across projects--using optimal analytic methods. Analytic Core support will ensure that the Program Project achieves unbiased and robust results that improve understanding of the course of Alzheimer’s disease and its translation into future efficient clinical trials.