Alzheimer's Disease Risks and Projections Using Multi-State Models with Biomarkers

点击翻译按钮获取中文摘要

Project Summary As the world's population ages, the numbers of persons with Alzheimer's disease (AD) will significantly increase. Knowledge about the pathogenesis of preclinical AD and biomarkers has grown enormously in the last few years. However, little is known about how to utilize biomarker screening tests in clinical settings to obtain individualized risk predictions, to obtain demographic projections of numbers of persons in preclinical states, and to evaluate the public health impact of potential interventions that delay transitions between preclinical states. Our proposed work will fill a critical void by utilizing multistate modeling to address emerging questions in clinical practice and public health that are raised by the expanding knowledge of AD biomarkers. Aim 1 is to perform a systematic review in order to develop a framework for a multi-state model for progression through the pre-clinical period of Alzheimer's disease that incorporates biomarkers and competing risks of death, and to calibrate the model's transition rates. Aim 2 is to calculate absolute risks of developing clinical AD during one's lifetime based on biomarker screenings that account for competing risks of death and ages at screens. Aim 3 is to develop U.S. population projections of numbers of persons in preclinical and clinical states, and to evaluate the potential impact of future interventions that may decrease preclinical transition rates. Interventions may act at various stages of the natural history of preclinical disease, and our models will be used to evaluate their potential effects on population projections. Our approach extends our previous calculation methods used to project clinical AD to now incorporate biomarker preclinical disease states. Our methods include systematic review, mathematical modeling, and Monte Carlo simulations. The proposed research is significant and innovative because it will synthesize new biomarker evidence about the preclinical period of Alzheimer's disease into a unifying multi-state model that will have application for clinical practice and public health. The research will provide a platform for incorporating new information about biomarkers to enable ongoing updating and refinement of individualized estimates of risks of clinical AD, population projections, and the impact of potential interventions on those projections.