PROJECT SUMMARY – Analysis Core (AC) The Analysis Core supports the Duke OAIC effort to understand and optimize reserve and resilience by providing data management and analytic support (Aim 1), providing methodological instruction (Aim 2), and incorporating and developing innovative biostatistical analytic methodologies into the OAIC program (Aim 3). The Analysis Core contains all the expertise needed to provide analytic support to junior and senior faculty across the range of study designs and analytic issues, including biostatisticians with expertise in study design, longitudinal analysis, psychometrics and estimation of latent variables; bioinformaticists with experience in genetic and high dimensional data analysis; and day-to-day monitoring of studies and data management. Data management will use secure web-based methods (REDCap) and methods for managing high dimensional metabolomic, proteomic, and genetic data. Duke OAIC supported studies are constructed and managed so that standardized analytic methods and common measures across studies can be employed. In addition to provision of technical analytic and data management support, the Analysis Core will provide consultation and training support to the faculty of the Duke OAIC (Aim 2). The Core will also pursue methodologic goals of interest to biostatisticians advance statistical science and address analytic issues encountered in current research (Aim 3). In particular, the study of resilience and reserve will require estimation of multi-parameter models and latent classes. A Developmental Project is proposed to develop estimation models and assess statistical performance (false detection rate, stability, power, bias, and validity) of these new classes of models. Working closely with the Molecular Measures and Health and Mobility Measures Cores, the Analysis Core will focus on methods for examining trajectories of change in the biological and clinical variables, develop aggregation techniques for this high dimensional data, establish temporal ordering, assess mediation and moderation pathways, and assess the statistical properties and constancy of the relationships across studies.