Core B- Biostat Core

点击翻译按钮获取中文摘要

Abstract Overview: The Sedentary Time and Aging Research (STAR) Program Biostatistics Core B will provide unified support for informatics and statistical needs across the three STAR Projects. The core will implement best practices for data collection and management, and apply rigorous statistical methods for data analysis. The Specific Aims are to provide comprehensive informatics support and data management, enhance communication, and to conduct data analysis and develop novel statistical approaches applicable across STAR projects. Informatics: The Biostatistics Core will coordinate and support database development and maintenance for all STAR Projects. The core will undertake extensive data validation, standardized monitoring and reporting across STAR Projects, including protocol compliance, form completion, accrual reporting, automated interim data quality checks, locking and documenting of finalized study data sets. Patient confidentiality will be protected. The STAR informatics system, hosted at the UCSD Supercomputer center, is HIPAA compliant. Statistical Analysis: The Biostatistics Core will provide rigorous unified statistical expertise, implementing best practices to analyze data across the three STAR Projects including intent-to-treat principles, longitudinal data analysis, and missing data methods. In addition, the Biostatistics core will develop and apply novel state- of-the-art statistical methods, pertinent to all three STAR Projects, including methods for modeling multiple longitudinal behavioral, biomarker, and health outcomes. New Methodology: Leveraging objective biomarker and 24-hour accelerometer measures, the Biostatistics Core will integrate data across STAR Projects, and test supplemental aims. We will identify temporal patterns of health behaviors (sedentary behavior, physical activity, sleep) in postmenopausal women, and ascertain if these patterns change after participation in the intervention trials. Also, we will develop algorithms to identify “responders”, i.e., participants with the greatest improvement in biomarker profiles in the intervention trials. We believe these supplemental aims could elucidate if/how health behavior accumulation (e.g., long versus interrupted bouts of sedentary behavior or physical activity) impact health outcomes, and provide prediction rules to identify likely “responders,” which could inform design of future trials and tailor interventions to participant needs to best effect successful behavior change and improve health. Impact. In summary, the STAR Biostatistics Core will provide a sound, robust and reproducible framework for supporting the overall objective of the STAR Projects: to enhance scientific knowledge of and examine the biobehavioral efficacy of interventions to interrupt sedentary behavior in a diverse sample of postmenopausal women. We expect that STAR program findings will ultimately inform public health recommendations for healthy living.