Neuroimaging Predictors of Alzheimer's Disease and Cognitive Decline

National Institute On Aging

Investigators

Linked publications, trials & patents

Abstract

Through our research program on early markers of Alzheimers disease (AD), we continue to perform serial magnetic resonance imaging (MRI), including measures of vascular changes, positron emission tomography (PET), and neuropsychological assessments in participants from the Baltimore Longitudinal Study of Aging (BLSA) to investigate the neurobiologic basis of individual differences in memory change over time, as well as predictors of cognitive impairment and resilience. These evaluations allow us to examine changes in brain structure and function which may be early preclinical predictors of cognitive change, impairment, including AD, and resilience or maintenance of cognitive health. We also investigate modifiers of cognitive and brain changes, including sex differences in cognitive and brain aging, genetic, metabolic, and inflammatory risk factors. Longitudinal neuroimaging evaluations of BLSA participants began in 1994 with approximately 160 individuals aged 55-85 at enrollment. In 2009, we expanded MRI assessments to more than 1000 individuals (with greater than 60% receiving 2 or more scans) and cognitive assessments. Our progress over the last year has focused on continued acquisition of MRI scans in BLSA participants, analyses of existing data, and methods development of new approaches to analysis of longitudinal neuroimaging data. In addition, we took advantage of newly available Simoa assays to investigate AD biomarkers and modifiers. In collaboration with Drs. Moghekar, Walker, and colleagues at the University of Gothenburg, we have completed Quanterix Neuro4plex plasma assays for ABeta40, ABeta42, Neurofilament Light, GFAP and ptau181 and ptau231 for all participants receiving PET scans, as well as selected participants in a case-control design, and the majority of individuals with a baseline 3T MRI scan. In a key paper (Bilgel et al 2023), we related cross-sectional and longitudinal plasma biomarkers of AD neuropathology, astrocytic activation, and neuronal injury to brain amyloid burden measured by PET-PiB. We investigated the cross-sectional sensitivity and specificity of these biomarkers against PET amyloid burden (positive or negative), the intraclass correlation coefficients as measures of reliability over time, and longitudinal changes in these biomarkers as a function of PET amyloid status (positive vs. negative). A multivariable model incorporating all biomarkers and demographic variables yielded the highest AUC in predicting presence of brain amyloid. As expected, AB42/AB40 ratio was lower and ptau values were higher in amyloid positive versus negative individuals. Interestingly we found significant declines in the AB42/AB40 ratio in amyloid negative but not positive individuals, suggesting the change in plasma ABeta had already occurred before PET amyloid positivity. Conversely, ptau181 and ptau231 showed significant longitudinal increases in amyloid positive but not negative individuals. We continue to work closely with Dr. Christos Davatzikos and his group at University of Pennsylvania through the iSTAGING consortium and with Dr. Bennett Landman and Derek Archer at Vanderbilt University. Publications over the last year have included identification of functional connectivity patterns in large scale harmonized resting state fMR scans (Zhou et al., 2023), characterizing white matter changes using free water corrected DTI data (Archer et al., 2023) and in relation to clinical staging of disease (Yang et al., 2023). The data generated under this project continue to be used by many intramural and extramural collaborators, involving analysis of skeletal mitochondrial function (Tian et al., 2023a), lipid profiles (Tian et al., 2023b), olfaction (Tian et al., 2023c), multi-sensory loss (Cai et al, 2023; Yesantharao et al, 2023), energetics (Dougherty et al., 2023), and circadian rhythms (Rabinowitz et al., 2022). They also are actively used through our participation in the Preclinical AD Consortium, led by Dr. Marilyn Albert of Johns Hopkins.