Abstract
INTRODUCTION: The brain age gap estimation (BrainAGE) method uses a machine learning model to generate an age estimate from structural magnetic resonance imaging (MRI) scans. The goal was to study the association of brain age with Alzheimer's disease (AD) imaging and plasma biomarkers. METHODS: One hundred twenty-three individuals from the Indiana Memory and Aging Study underwent structural MRI, amyloid and tau positron emission tomography (PET), and plasma sampling. The MRI scans were processed using the software program BrainAgeR to receive a "brain age" estimate. Plasma biomarker concentrations were measured, and partial Pearson correlation models were used to evaluate their relationship with brain age gap (BAG) estimation (BrainAGE = chronological age - MRI estimated brain age). RESULTS: Significant associations between BAG and amyloid and tau levels on PET and in plasma were observed depending on diagnostic categories. DISCUSSION: These findings suggest that BAG is potentially a biomarker of pathology in AD which can be applied to routine brain imaging. HIGHLIGHTS: Novel research that uses an artificial intelligence learning tool to estimate brain age.Findings suggest that brain age gap is associated with plasma and positron emission tomography Alzheimer's disease (AD) biomarkers.Differential relationships are seen in different stages of disease (preclinical vs. clinical).Results could play a role in early AD diagnosis and treatment.