The effect of smoking on age estimation by DNA methylation in semen samples

The determination of biological age is gaining importance in forensic identification and other legal contexts. DNA methylation and, in particular, alterations in CpG dinucleotide sites form the basis of biological age estimation models. In this study, the effect of long-term smoking on biological age estimation was examined by using the DNA methylation-based age estimation panel for human semen samples developed by Er et al (2024) (1). The study included 70 male participants ages between 19 and 58 who smoked at least 20 cigarettes per day and 70 control subjects who had never smoked. The DNA samples obtained from the semen samples were analyzed using bisulfite conversion and the SNaPshot® minisequencing technique. The analysis revealed a negative correlation between the cg23488376 locus and age (β=–0.353; p<0.01), while a weak positive correlation was observed between the NOX4 locus and age (β=0.336; p<0.01). These relationships were more pronounced in smokers (cg23488376 for β=–0.550; NOX4 for β=0.469; p<0.01), indicating that smoking had a significant effect on DNA methylation profiles. The error values related to model performance were calculated as MAE=6.94 years, RMSE=10.40 years for the smoking group and MAE=8.02 years, RMSE=10.37 years for the non-smoking group. These values were found to be higher than the error rates reported by Er et al. (2024) (MAE=4.1 years, RMSE=5.0 years). This difference is thought to stem solely from the examination of the smoking variable and the exclusion of lifestyle differences. It has been determined that the model generally predicts a higher age than chronological age in individuals who smoke (p=0.014). The results revealed that the DNA methylation-based age prediction model increased the differences between estimated age and chronological age by accelerating aging in smokers. These findings emphasize the importance of considering smoking in biological age prediction studies and the necessity of large samples and multidimensional lifestyle data in future studies. Keywords: DNA Methylation, age prediction, epigenetics, semen, smoking.