The Link between Brain age and Alzheimer’s Disease development using different neuroimaging modalities

Background Brain age (BA) provides a proxy of general brain health through the prediction chronological (CA) from neuroimaging data healthy individuals. During aging, adult undergoes changes both on morphological and metabolic level. Here, we compare link between BA predicted morphology (structural MRI) or metabolism (18F-FDG-PET) cognitive function, as well pathophysiological measures Alzheimer’s disease (AD) in cognitively normal individuals (CN) patients with mild impairment (MCI). Method Several machine learning algorithms were trained to predict 256 matched MRI 18F-FDG-PET scans CN Disease Neuroimaging Initiative. Trained then used test sample (n = 123) MCI 621). Partial spearman correlations brain-predicted difference (BA – CA; BPAD) composite scores function (memory: ADNI-MEM; executive function: ADNI-EF), AD (amyloid (cerebrospinal fluid (CSF) PET) tau pathology (CSF p-tau)) subsequently calculated these samples correcting for sex. Result 18F-FDG-PET- MRI-predicted showed mean absolute error 2.77 1.81 years CN, respectively. ADNI-MEM declined increasing 18F-FDG-PET-BPAD, while ADNI-EF MRI-BPAD CN. Higher BPAD was significantly associated worse modalities, stronger associations found function. 18F-FDG-PET-, but not correlated CSF amyloid load In MCI, higher modalities more p-tau pathology, only an increase pathology. Associations generally stronger. Conclusion can be reliably basis MRI. Amyloid-neuropathology, predisposing AD, initial memory decline is related features advanced ahead clinical onset disease. Onset tau-related neurodegeneration objective strongly signals