Epigenetic Associations with Sociodemographic Factors and Cardiometabolic Profiles Across the Life Course

Cardiovascular disease (CVD) is the leading cause of morbidity and mortality among U.S.adults. Cardiometabolic risk factors contributing to the development of CVD include hypertension,diabetes, dyslipidemia, overweight and obesity, and inflammation, all of which have long-standingracial/ethnic and socioeconomic disparities in the U.S. DNA methylation, an epigeneticmechanism that regulates transcription without altering the DNA sequence, is associated withcardiometabolic risk factors and CVD. A better understanding of the relationships betweenmethylation-based biomarkers (e.g., individual methylation sites (CpGs), epigenetic clocks, andpoly-epigenetic scores (PES)), sociodemographic and behavioral factors, and cardiometabolicprofiles can shed light on advancing strategies in the precision health context. This dissertationexplores epigenetic associations with sociodemographic factors and cardiometabolic profiles inmulti-ancestry U.S. cohorts.In Aim 1, we investigated the associations between five measures of epigenetic ageacceleration and four blood lipid measures in older adults from the Health and Retirement Study(HRS, mean age = 70 years). We also examined whether demographic factors (i.e., age, sex, andeducational attainment) modified these relationships. Greater epigenetic age acceleration wasassociated with lower total cholesterol (TC), high-density lipoprotein-cholesterol (HDL-C), andlow-density lipoprotein-cholesterol (LDL-C), and higher triglycerides (TG) (p<0.05), although theeffect sizes were relatively small. The associations were stronger in younger participants, females,and those with higher educational attainment.
In Aim 2, we constructed trait-specific PESs for eight cardiometabolic risk factors (systolicand diastolic blood pressure (SBP, DBP), body mass index (BMI), C-reactive protein (CRP), HDLC, LDL-C, TG, and fasting glucose), and examined their associations with the correspondingcardiometabolic traits in older U.S. adults from the HRS. We next investigated how demographics(i.e., age, sex, and educational attainment) and health behaviors (i.e., smoking, alcoholconsumption, and physical activity) modified these associations. All PESs were positivelyassociated with their traits (P < 0.05), and most associations remained consistent across race/ethnicgroups. Associations for BMI, HDL-C, and TG were stronger in younger participants, and BMIand HDL-C were also stronger in females. The association for CRP was stronger among thosewith a high school degree. Finally, the association for HDL-C was stronger among currentsmokers.In Aim 3, we utilized the structured life-course modeling approach (SLCMA) to examinethe time-sensitive associations between income-to-poverty ratio (IPR) and epigenome-widemethylation at two time points (ages 9 and 15) in U.S. children and adolescents from the Future ofFamilies and Child Wellbeing Study. A total of 269 and 285 CpGs at ages 9 and 15, respectively,were identified with FDR-q<0.05 and >3% of DNA methylation explained by the selected IPR.Most CpGs were associated with IPR in early childhood (at or before age 5) rather than in laterchildhood (age 9), adolescence (age 15), or accumulated poverty throughout childhood andadolescence. The CpGs identified at both ages were primarily enriched for immune systemfunction.This dissertation provides a deep characterization of the relationships among epigeneticbiomarkers and cardiometabolic risk factors in older adults, identifying subgroups for which thesebiomarkers perform particularly well. It further identifies methylation profiles in children andxviiiadolescents influenced by demographic factors in early life, and points to biological pathways thatmay underly the link between early life adverse events and later life health outcomes. Together,these findings may help to advance strategies for more precise prevention and early interventionefforts for the U.S. population throughout the life course.