Abstract P271: Dna Methylation Measures of Aging in Midlife Are Associated with Frailty Components in African American and European American Older Adults: the Atherosclerosis Risk in Communities (ARIC) Study
Steven Nguyen,Weihua Guan,Jan Bressler,Megan Grove,Priya Palta,Rui Xia,Zhiying Wang,Lindsay Fernandez-Rhodes,Anne E. Justice,Yun Li,Eric A. Whitsel,Kari E. North,Myriam Fornage,Eric Boerwinkle,Anna Kucharska-Newton,James S. Pankow,Ellen W. Demerath
DOI: https://doi.org/10.1161/circ.141.suppl_1.p271
IF: 37.8
2020-01-01
Circulation
Abstract:Introduction: Several putative biomarkers of accelerated biological aging have been developed from patterns of DNA methylation in blood and other tissues (DNAm-Age scores). Some of these DNAm-Age scores have been found to be associated with age-related outcomes such as CVD and all-cause mortality, but their associations with frailty are unclear. We hypothesized that higher DNAm-Age scores measured in midlife are associated with higher odds of frailty and its components in late life. Methods: DNAm was measured at visit 2 or 3 in leukocyte DNA from 812 African American (AA) and 400 European American (EA) participants (mean age=54 years in AA, 57 years in EA), and was used to calculate three different DNAm-Age scores (Horvath, Hannum, and PhenoAge). Frailty status was assessed at visit 5, and dichotomized (frail/pre-frail or robust) using the following component criteria: slow walking speed, low grip strength, weight loss, low physical activity, and exhaustion. Logistic regression tested the associations of DNAm-Age scores, as the predictors of frailty status or each frailty component. Model 1 adjusted for visit 2 or 3 age, sex, education, white blood cell count, cell type proportions, and technical variables and model 2 additionally adjusted for BMI, smoking, alcohol, and visit 5 CHD, stroke, hypertension, and T2D. Inverse probability weights accounted for cohort attrition prior to frailty assessment. The Bonferroni-corrected threshold for statistical significance was set to p<8.3E-3 (0.05/6). Results: In African American participants, a 4-year increase in the PhenoAge score in midlife was associated with frailty in late life (OR=1.21, 95% CI=1.09, 1.35, p=3E-4) in model 1, which was attenuated in model 2 (OR=1.13, 95% CI=1.02, 1.26, p=2.4E-2), and with physical activity (model 2 OR=1.32, 95% CI=1.13-1.53, p=4E-4); a 4-year increase in the Hannum score was associated with the exhaustion and grip strength components (OR=1.33, 95% CI=1.13-1.57, p=6E-4; OR=1.38, 95% CI=1.16-1.64, p=3E-4, respectively). In European American participants, the DNAm-Age scores were not associated with frailty however a 4-year increase in both the Horvath and Hannum scores was associated with the weight loss component (model 2 OR=1.56, 95% CI=1.20-2.02, p=9E-4; OR=2.37, 95% CI=1.49-3.76, p=2E-4), respectively. Conclusions: In African American ARIC participants, a higher mid-life PhenoAge score was associated with higher odds of frailty/pre-frailty in late life, consistent with accelerated biological aging, but was not significant after further adjustment. The Horvath, Hannum, and PhenoAge scores were associated with some frailty components in African American and European American ARIC participants, consistent with accelerated biological aging. Following replication, future studies should examine the causal relationship between DNAm-Age, frailty, and frailty components.