Combined healthy lifestyle score and risk of epigenetic aging: a discordant monozygotic twin study

Abstract

We investigated whether lifestyle influences epigenetic aging in 143 monozygotic twin pairs discordant for the combined healthy lifestyle score. Twins were scored for four lifestyle factors as unhealthy or healthy; non-smoker, moderate drinker, adequate fruit and vegetable intake, and sufficient physical activity. The combined healthy lifestyle score was calculated for each participant by summing the binary score for each factor. Individual and co-twin analyses were used to assess the relationship between single or combined lifestyle scores, along with DNA methylation age acceleration (AA) calculated using Horvath's and Li's epigenetic clocks, focusing on AA and intrinsic epigenetic age acceleration (IEAA) measures. Compared with the twins that scored no or one healthy lifestyle point, those who scored four healthy lifestyle points had lower Li_IEAA with similar results observed in the co-twin analysis. No significant relationships were found in analyses based on Horvath's clock, although the direction of correlations was consistent with that determined using Li's clock. Smoking and drinking did not significantly affect DNA methylation AA; however, physical activity and intake of vegetables and fruits did, although the influence varied depending on the epigenetic clock. Our findings suggest that a healthy lifestyle may be an important way to delay aging and prevent age-related diseases.

Keywords: Chinese; DNA methylation age; epigenetic aging; healthy lifestyle; twin.

Figure 1

Distribution of two kinds of predicted DNAm age and chronological age by healthy lifestyle score. Data were shown as mean±standard error.

Figure 2

Association between DNAm age acceleration and combined healthy lifestyle score. Figure 2.1 presented the result generated by mixed effect model and the result of co-twin analysis were shown in Figure 2.2. AA:age acceleration; IEAA: intrinsic epigenetic age acceleration, Li and Horvath present two algorithms for calculating mage respectively.