Pregnancy is linked to faster epigenetic aging in young women

Abstract

A central prediction of evolutionary theory is that energy invested into reproduction comes at the expense of somatic maintenance and repair, accelerating biological aging. Supporting this prediction are findings that high fertility among women predicts shorter lifespan and poorer health later in life. However, biological aging is thought to begin before age-related health declines, limiting the applicability of morbidity and mortality for studying the aging process earlier in life. Here, we examine the relationship between reproductive history and biological aging in a sample of young (20 to 22yo) men and women from the Cebu Longitudinal Health and Nutrition Survey, located in the Philippines (n = 1,735). We quantify biological aging using six measures, collectively known as epigenetic clocks, reflecting various facets of cellular aging, health, and mortality risk. In a subset of women, we test whether longitudinal changes in gravidity between young and early-middle adulthood (25 to 31yo) are associated with changes in epigenetic aging during that time. Cross-sectionally, gravidity was associated with all six measures of accelerated epigenetic aging in women (n = 825). Furthermore, longitudinal increases in gravidity were linked to accelerated epigenetic aging in two epigenetic clocks (n = 331). In contrast, the number of pregnancies a man reported fathering was not associated with epigenetic aging among same-aged cohort men (n = 910). These effects were robust to socioecological, environmental, and immunological factors, consistent with the hypothesis that pregnancy accelerates biological aging and that these effects can be detected in young women in a high-fertility context.

Keywords: biological aging; costs of reproduction; fatherhood; pregnancy.

Fig. 1.

This figure shows the relationship between the number of pregnancies and cross-sectional measures of first-generation (Horvath and Hannum) and second-generation (PhenoAge and GrimAge) epigenetic clocks, DunedinPACE pace of aging, and a DNAm surrogate measure for leukocyte telomere length (DNAmTL). Higher values for all clocks correspond to accelerated biological aging, except for DNAmTL, where lower values correspond to shorter telomere length and accelerated aging. Four high-leverage data points with DNAm Aging values >3 SD from the mean are excluded from the figure but were included in all analyses using robust regression (n = 825). P-values are inset. Effect sizes and P-values are provided in Table 3.

Fig. 2.

This figure shows the longitudinal association between the change in pregnancy number and change in biological aging measured using first-generation (Horvath and Hannum) and second-generation (PhenoAge and GrimAge) epigenetic clocks, DunedinPACE pace of aging, and a DNAm surrogate measure for leukocyte telomere length (DNAmTL). Higher values for all clocks correspond to accelerated biological aging, except for DNAmTL, where lower values correspond to shorter telomere length and accelerated aging. All women were pregnant at the second time point, but current pregnancies were not included in change scores. Four high-leverage data points with DNAm Aging values >3 SD from the mean are excluded from the figure but were included in all analyses using robust regression (n = 331). P-values are Inset. Effect sizes and P values are provided in Table 4.