A Prospective Cohort Study of Multimodal Biomarkers and Progression From Mild Cognitive Impairment to Dementia

This prospective, longitudinal observational cohort study will characterize multidimensional biomarker profiles across the cognitive continuum. Its principal aim is to develop integrated approaches for disease classification and prediction of progression from mild cognitive impairment (MCI) to dementia. The study will enroll 600 participants in three cohorts: approximately 200 cognitively unimpaired participants, 200 participants with mild cognitive impairment, and 200 participants with dementia. The mild cognitive impairment and dementia cohorts may include Alzheimer disease-related and non-Alzheimer disease-related clinical subtypes, without fixed enrollment quotas for these subtypes. No therapeutic intervention will be assigned. Baseline evaluations will include demographic and clinical characterization, comprehensive neuropsychological testing, and blood collection. Participants will also undergo multimodal neuroimaging. Blood-based analyses will assess established and exploratory biomarkers. Established measures will include amyloid beta 42 and phosphorylated tau 217. These biomarkers will be evaluated using chemiluminescent enzyme immunoassay, electrochemiluminescence immunoassay, immunoprecipitation-mass spectrometry, and single-molecule array platforms. Exploratory analyses will examine protein biomarkers, microRNA expression profiles, and lipid-metabolomic markers. Apolipoprotein E genotype will also be determined. Blood collection, neuropsychological testing, and neuroimaging will be repeated at 12 months. Cross-sectional analyses will compare biomarker profiles across the study groups. Longitudinal analyses will characterize changes in cognition, circulating biomarkers, and neuroimaging measures. Associations among blood biomarkers, neuroimaging measures, and clinical phenotypes will be examined using prespecified statistical methods. Multivariable statistical and machine-learning approaches will then be used to identify informative combinations of clinical, cognitive, genetic, blood-based, and neuroimaging features. These features will be incorporated into diagnostic and risk-prediction models for MCI, dementia, and progression from MCI to dementia.