Abstract Background/Introduction DNA methylation (epigenetic) age is a reliable measure of biological age and is associated with frailty. Epigenetic age acceleration (EAA) can independently predict adverse cardiometabolic risks. However, neither has been compared against atherosclerotic cardiovascular disease (ASCVD) risk equations previously. Purpose To compare five commonly used epigenetic clocks against two clinically established ASCVD risk estimators. Methods Five epigenetic clocks, including the original Horvath (2013), Hannum (2013), PhenoAge (2018), GrimAge (2019) and GrimAge2 (2022), were analysed in older adults (age ≥60y) of the UFO cohort. Individuals with a history of heart failure or advanced therapies including implantable devices and heart transplant were excluded. Equal ratios of robust, pre-frail, and frail individuals classified by the FRAIL scale (Table) were selected prior to analysis. The MethylationEPIC v1.0 array (Illumina, USA) was used for analysis of DNA methylation in blood cells. Serum was analysed for lipid and lipoprotein attributes using 1H-NMR spectroscopy (Nightingale Health, Finland). EAA (independent variable) was calculated as the residuals in the correlation between epigenetic and chronological age. Risk estimates from the ACC/AHA ASCVD (2013) pooled cohort equations (PCEs) and the PREVENT (2024) risk equation were used as dependent variables. Multiple imputation was performed for incomplete systolic blood pressure data (Fig.1A). Regression models were adjusted for age, sex and smoking. A p-value 0.05 was considered statistically significant. Results Of 535 older adults, 186 were robust, 180 pre-frail and 169 frail (Table). Increased frailty was associated with older age, female sex, and smaller low-density lipoprotein (LDL) particle size (diameter) (Table 0.022; Fig.1D). Using the 2013 PCEs to predict 10-year ASCVD risk, it was determined that GrimAA and Grim2AA, but not other EAAs, were significantly associated (adjusted p0.01 and p0.001, respectively; Fig.1E); whereas PhenoAA, GrimAA and Grim2AA were associated with the PREVENT risk equation (adjusted p0.01, p0.00001 and p0.0000001, respectively; Fig.1E). After excluding 92 individuals with previous ASCVD events (i.e. ischaemic heart disease, myocardial infarction and/or stroke), GrimAA and Grim2AA remained associated with PREVENT, whereas Grim2AA only was associated with the 2013 PCEs (adjusted p0.05; Fig.1F). Addition of EAA to risk equations did not further improve the C-statistic for prediction of adverse risks. Conclusion(s) Grim2AA is an independent predictor of frailty and ASCVD risk in older adults. The ASCVD risk estimated by Grim2AA is consistent with that of the ACC/AHA PCEs (2013) and the PREVENT (2024) risk equation.
Ng et al. (Sat,) studied this question.