Comparing epigenetic, proteomic, and MRI aging clocks in sporadic and autosomal-dominant Alzheimer disease

/ABSTRACT While older age is the greatest known risk factor for Alzheimer disease (AD), it is not causative. Increased risk of AD is likely driven in part by cumulative, progressive physiological processes across biological levels, termed biological aging. Biological age can be estimated through machine learning-based “clock” models using rich, multivariate physiological signatures, including DNA methylation (epigenetic age), protein concentration (proteomic age), and brain structure (brain-predicted age). By characterizing biological age at the epigenetic, proteomic, and endophenotypic layers, these clocks may detect complementary signatures of biological aging, which could identify unique processes of AD progression or resilience. The proposed K01 research study will rigorously compare state-of-the-art biological age clocks across epigenetic, proteomic, and neuroimaging modalities using available, deeply-phenotyped, longitudinal human AD datasets. The study will test whether epigenetic, proteomic, and brain-predicted age clocks represent complementary signatures of biological aging, disease progression, and resilience in sporadic late onset AD (sLOAD), which is the most common form of AD (Aim 1), and in autosomal dominant AD (ADAD), in which the accelerated aging hypothesis can be tested with minimal age-related confounding factors (Aim 2). The final aim will test whether biological age estimates from these clocks differ between sLOAD and ADAD (Aim 3). As protein expression inherently links genetic and phenotypic layers, we predict that proteomic age will mediate the association between epigenetic and brain-predicted age clocks in both sLOAD and ADAD. If these multimodal clocks capture complementary signals of AD risk and resilience, we predict that each clock will explain unique related variance in both sLOAD and ADAD. If pathological severity is greater in ADAD than sLOAD, we predict that age clocks will exhibit greater sensitivity to ADAD than sLOAD. This study will address an important knowledge gap on whether multimodal age clocks provide additive measures of AD staging or resilience. Future independent research projects may explore multimodal biological age longitudinally, in relation to additional pathological, risk, and resilience factors, or using more advanced modeling approaches. This K01 award will support the candidate’s transition towards an independent career with a unique, transdisciplinary research program in aging, AD, and AD-related dementias (ADRD). A carefully-tailored training plan will build upon the candidate’s established expertise in cognitive psychology, neuroimaging, and machine learning with additional expertise in the analysis and interpretation of epigenetic (Goal 1) and proteomic datasets (Goal 2), frameworks of resilience and protective factors in AD (Goal 3), and development of the professional skills required to lead and support an independent research lab (Goal 4). These training goals will be supported by individualized mentorship from a team of domain experts, didactic coursework, practical skill-building, and completion of the research aims. Project Number: 1K01AG090753-01 | Fiscal Year: 2025 | NIH Institute/Center: National Institute on Aging (NIA) | Principal Investigator: Peter Millar | Institution: WASHINGTON UNIVERSITY, SAINT LOUIS, MO | Award Amount: $128,373 | Activity Code: K01 | Study Section: Career Development For Early Career Investigators Study Section [AGCD-2] View on NIH RePORTER: https://reporter.nih.gov/project-details/11033122