Long Term Effects of a First Pregnancy on the Hormonal Environment: Estrogens and Androgens*

An early (but not a late) first pregnancy is known to be protective for breast cancer. This effect might be mediated through a long term change in the hormonal environment caused by the early first pregnancy. To investigate the possibility of such a change we carried out a prospective longitudinal study of serum and urinary estrogens and serum androgens in four groups of women, namely early (age, 18–23 yr) first pregnancy (n = 15), early control (n = 20), late (age, 29–40 yr) first pregnancy (n = 9), and late control (n = 20). The pregnancy groups were studied before (initial visit) and 7–19 months after a first pregnancy (return visit). The control groups were similarly studied, but without an intervening pregnancy. The following were measured: serum estrone (E1), 17β-estradiol (E2), estriol (E3), and E1 sulfate; urinary total E1, E2, E3, and glucosiduronates of these three estrogens; and serum testosterone, dehydroepi-androsterone sulfate (DHAS), and dehydroepiandrosterone (DHA). There was no significant change between the initial and return visits in serum E1, E2, E1 sulfate, or any of the urinary estrogens in either pregnancy group or in the corresponding control groups. There was, however, a significant increase in serum E3 between initial and return visits for both pregnancy groups compared with the control values. There was no significant change in serum testosterone. There was a marked significant decrease in both serum DHAS and DHA between initial and return visits in both pregnancy groups compared with the corresponding control group values. There was also a significant increase in the serum E3 to DHA ratio in both pregnancy groups. A cross-sectional study (measuring serum DHAS and DHA only) was then carried out in a series of parous and nulliparous women. The serum DHAS and DHA levels were markedly and significantly lower in parous than in nulliparous women, as expected. There was no significant relationship between serum DHAS or DHA levels and months elapsed (up to 150) since last delivery, indicating that the changes last at least for this period of time. There was no significant relationship between serum DHAS or DHA levels and parity (one to three previous pregnancies), indicating that the changes occur only after a first pregnancy. We conclude that 1) a first pregnancy induces a long term alteration in the hormonal environment in respect to serum E3, which increases significantly, and serum DHAS and DHA, which decrease significantly; 2) these changes occur with both an early and a late first pregnancy; 3) the duration of the effect is at least 19 months after delivery for serum E3 and at least 150 months (12–13 yr) after delivery for serum DHAS and DHA; and 4) the effect occurs only after a first pregnancy and not after succeeding pregnancies. These long term hormonal changes may form part of a protective mechanism for breast cancer, associated with an early first pregnancy. Since the change occurs after both early and late first pregnancy, the effect may be dependent on the duration of exposure rather than to the change itself.