摘要：Male aging and obesity have both been shown to contribute to declines in fertility in men. Recent work in aging has shown consistent epigenetic changes to sperm as a man ages. In fact, our lab has built a tool that utilizes DNA methylation signatures from sperm to effectively predict an individual’s age. Herein, we performed this preliminary cohort study to determine if increased BMI accelerates the epigenetic aging in sperm. A total of 96 participants were divided into four age groups (22–24, 30, 40–41, and > 48 years of age) and additionally parsed into two BMI sub-categories (normal and high/obese). We found no statistically significant epigenetic age acceleration. However, it is important to note that within each age category, high BMI individuals were predicted to be older on average than their actual age (~ 1.4 years), which was not observed in the normal BMI group. To further investigate this, we re-trained a model using only the present data with and without BMI as a feature. We found a modest but non-significant improvement in prediction with BMI [r2 = 0.8814, mean absolute error (MAE) = 3.2913] compared to prediction without BMI (r2 = 0.8739, MAE = 3.3567). Future studies with higher numbers of age-matched individuals are needed to definitively understand the impact of BMI on epigenetic aging in sperm.
其他摘要：Abstract Male aging and obesity have both been shown to contribute to declines in fertility in men. Recent work in aging has shown consistent epigenetic changes to sperm as a man ages. In fact, our lab has built a tool that utilizes DNA methylation signatures from sperm to effectively predict an individual’s age. Herein, we performed this preliminary cohort study to determine if increased BMI accelerates the epigenetic aging in sperm. A total of 96 participants were divided into four age groups (22–24, 30, 40–41, and > 48 years of age) and additionally parsed into two BMI sub-categories (normal and high/obese). We found no statistically significant epigenetic age acceleration. However, it is important to note that within each age category, high BMI individuals were predicted to be older on average than their actual age (~ 1.4 years), which was not observed in the normal BMI group. To further investigate this, we re-trained a model using only the present data with and without BMI as a feature. We found a modest but non-significant improvement in prediction with BMI [r 2 = 0.8814, mean absolute error (MAE) = 3.2913] compared to prediction without BMI (r 2 = 0.8739, MAE = 3.3567). Future studies with higher numbers of age-matched individuals are needed to definitively understand the impact of BMI on epigenetic aging in sperm.
