Innovative Analytical Methods for DNA Methylation Age

As the US population continues to age in the coming years, the need for biological measures and biomarkers of aging becomes increasingly urgent. Finding and validating biomarkers of aging continues to attract research efforts, but with limited success. Epigenetic modifications are potentially critical to the biological processes that underlie aging (Ben-Avraham et al. 2012). More recently, DNA methylation levels (DNAm) have been identified as useful tools for defining biological age, as Hannum et al. (2013) and Horvath (2013) both combined DNAm at multiple loci to quantify human aging. This “DNA methylation age” predicted all-cause mortality later in life (Marioni et al. 2015). Both Hannum and Horvath applied the ultra-high dimensional (~485K DNA methylation markers) variable selection methodology with elastic net penalty (Zou and Hastie 2005) to derive their epigenetic age indices. However, there are several issues in the original approaches by these researchers. In this grant, we propose more accurate and robust epigenetic age models using ultra-high dimensional DNA methylation markers. We also consider longitudinal DNAm data. We will develop and disseminate a user-friendly statistical software package that will enable researchers to implement these methods with ease. We will apply our methods to two large longitudinal cohort studies: the Coronary Artery Risk Development in Young Adults (CARDIA) and Multi-Ethnic Study of Atherosclerosis (MESA). Our discoveries may illustrate the biological mechanisms underlying traditional and innovative risk factors for mortality and discover more accurate and reliable markers for biological aging. Potential clinical, lifestyle, and pharmaceutical interventions can thus be developed for healthy aging.