Abstract TU99: Changes in DNA methylation-based biological aging predict structural and vascular brain damage in older adults: Age, Gene/Environment Susceptibility - Reykjavik Study

Background: DNA methylation (DNAm)-based biological aging (BA) is linked to poor structural and functional brain integrity, but prior studies often had small samples, cross-sectional designs, or focused narrowly on specific brain health indicators. We examined whether changes in BA are associated with multiple domains of brain health. Methods: Data were drawn from the Age, Gene/Environment Susceptibility–Reykjavik Study (AGES-RS), including participants with DNAm, MRI, and cognitive data (baseline mean age = 75.5 ± 4.8 years; 57.6% women). DNAm was assayed using the Illumina Infinium MethylEPIC array. Brain volumes and infarcts were measured by MRI, and cognition by standard tests. BA pace was estimated with DunedinPACE (DDPACE; >1 = faster aging). Baseline data (2002–2006) included 2,602 individuals, with 2,081 re-examined after ~5 years. Associations between baseline DDPACE and brain volume, infarcts, and cognition at follow-up were analyzed, adjusting for age, sex, education, batch, and blood cell composition. DDPACE and continuous outcomes were standardized. Additionally, mediation analyses tested whether baseline DDPACE mediated associations between midlife lifestyle (e.g., smoking) and health factors (e.g., total cholesterol), their composite score (Life’s Simple 7), and late-life brain outcomes (mean age = 81 ± 4.8 years). Results: Higher baseline DDPACE was inversely associated with white matter (WM, β=-0.14, 95% CI -0.19 – -0.09), gray matter (GM, β=-0.09, -0.14 – -0.05), and total brain volume (TBV, β=-0.12, -0.17– -0.08), and positively with WM lesions (WML, β=0.06, 0.01–0.11) at follow-up, all P<0.05. Associations remained significant after adjusting for baseline brain volumes, except WML. Higher baseline DDPACE also predicted poorer cognition (e.g., processing speed β=-0.10, -0.14 – -0.05, P<0.001) and greater brain infarct risk (OR=1.22, 1.07–1.39, P<0.01). Participants who shifted to faster aging (>1 SD above mean at follow-up) had higher WML (β=0.05, 0.01–0.09, P<0.01). Mediation analyses indicated that DDPACE accounted for 18.9–30.7% of the association between midlife smoking and late-life WM, GM, TBV, and WML, and 15–18% of the LS7 effect on these outcomes (all P < 2.2×10 -16 ). Conclusion: A DNAm-based biomarker of aging predicts brain damage and mediates the association of midlife cardiovascular health-related risk factors with late-life brain health.