| Abstract [eng] |
With the rapidly ageing population, the prevalence of age-related health disorders is increasing, making the assessment of biological ageing increasingly important. Although ageing is most commonly described in terms of chronological age, this indicator does not reflect individual differences in the ageing process. Epigenetic biomarkers, particularly epigenetic clocks based on DNA methylation, allow the estimation of epigenetic age and the calculation of epigenetic age acceleration (EAA), which reflects the rate of biological ageing and is associated with various age-related diseases and a decline in functional capacity. Sarcopenia and frailty syndrome are common conditions in older adults, associated with an increased risk of disability, hospitalisation, and mortality; however, their underlying biological mechanisms remain insufficiently understood. The aim of this study was to compare four epigenetic clock models (Horvath, Hannum, PhenoAge, and PCGrimAge) between older adults with and without sarcopenia and/or frailty syndrome, and to evaluate their epigenetic age and epigenetic age acceleration. The study included 235 participants, divided into a sarcopenia and/or frailty syndrome group (n = 64) and a control group (n = 171). Participants underwent anthropometric, functional, and psychophysiological assessments, and epigenetic age was calculated based on DNA methylation data. Statistical analysis was performed using parametric and non-parametric methods, correlation and regression analyses, as well as model accuracy assessment using MAE, RMSE, and BIAS indicators. The results showed that epigenetic age values were higher in the sarcopenia and/or frailty syndrome group across all applied models compared to the control group; however, these differences were partly explained by the higher chronological age of the participants in this group. After evaluating epigenetic age acceleration, it was found that EAA values in the control group were close to zero or negative, whereas in the sarcopenia and/ or frailty syndrome group they were positive. Statistically significant differences in EAA between groups were observed using the Horvath, Hannum, and PhenoAge models (p < 0.05), while no significant differences were found using the PCGrimAge model. The largest between-group difference was observed with the PhenoAge model. It was also found that accelerated epigenetic ageing (EAA > 0) was more frequently observed in the sarcopenia and/or frailty syndrome group. All epigenetic clock models showed statistically significant correlations with chronological age, with the strongest correlation observed for the PCGrimAge model. Regression analysis demonstrated that this model had the highest explanatory power in both groups, while the Hannum model showed a regression coefficient closest to the ideal value in the control group. Model error analysis revealed systematic deviations, whereas the PCGrimAge model demonstrated the most stable performance across groups. In conclusion, epigenetic age acceleration is higher in the sarcopenia and/or frailty syndrome group, and accelerated epigenetic ageing is observed more frequently compared to the control group. Different epigenetic clocks reflect these changes inconsistently; therefore, their application should be guided by the specific aim of the study. |
