AD is the most common form of dementia, impacting 11% of people over 65 years of age. There currently exist no truly effective treatments or prevention strategies for AD, in part because of the difficulty of identifying pre-symptomatic individuals for clinical trials. The goal of this study is to further characterize and validate a novel prediction model we have generated based on blood levels of circRNAs predictive of diagnosis and risk of AD.