Follow us
P087 – An Array Based Epigenetic Clock for Accurate Age Prediction of East Asian Populations
Aging is a complex and gradual process that leads to progressive changes in the human body over time. This process is influenced by a wide range of genetic, environmental, and lifestyle factors, resulting in substantial inter-individual variability in aging phenotypes. DNA methylation has emerged as a robust biomarker that reflects these biological changes, and methylation levels at CpG sites have been widely used for age estimation. However, the performance of epigenetic clocks may vary across ethnic groups, and large-scale East Asian datasets remain underrepresented in model development. In this study, we developed a blood-based DNA methylation age prediction model tailored to East Asian populations and assessed its applicability. We identified age-associated CpG sites using Illumina’s EPIC v.1 array data (1,008 individuals aged 20-80 from the Gene–Environment Interaction and Phenotype (GENIE) cohort), and constructed a least absolute shrinkage and selection operator (LASSO)-based model consisting of 190 CpG sites to balance accuracy and model complexity. This model demonstrated high prediction accuracy (Pearson’s correlation coefficient = 0.93, mean absolute error = 2.88 years) on an independent dataset of 2,350 individuals aged 40-80 from the Korean Genome and Epidemiology Study (KoGES). Compared with existing epigenetic clocks (e.g., the Horvath clock), our model showed the lowest prediction error across all four East Asian datasets (mean absolute error: 2.47–3.59 years) and maintained low error levels in three non–East Asian datasets (mean absolute error: 1.73–5.72 years). Notably, the model exhibited high accuracy even in saliva, supporting its applicability to commonly encountered forensic samples. Regression analyses showed that metabolic-related biomarkers were significantly associated with extrinsic epigenetic age acceleration, and both extrinsic and intrinsic epigenetic age acceleration were associated with metabolic syndrome status (P < 0.05). In conclusion, the developed epigenetic clock based on a Korean population dataset enables accurate age estimation in East Asian populations, with potential implications for understanding aging-related biological variation.
Authors
- Sang Un Park (Department of Forensic Medicine, Seoul National University College of Medicine, South Korea)
- Eun Kyung Choe (Department of Surgery, Seoul National University Hospital Healthcare System Gangnam Center, South Korea)
- Soong Deok Lee (Department of Forensic Medicine, Seoul National University College of Medicine, South Korea)
- Hwan Young Lee (Department of Forensic Medicine, Seoul National University College of Medicine, South Korea)
On the same topic
