In utero exposures and the incidence of endometriosis.

OBJECTIVE: To investigate the relation between the fetal environment and endometriosis.
DESIGN: Prospective cohort study.
SETTING: Nurses' Health Study II with 10 years of follow-up. PARTICIPANT(S): Eighty-four thousand, four hundred forty-six women aged 25-42 who had never been diagnosed with endometriosis, infertility, or cancer at baseline in 1989. MAIN OUTCOME MEASURE(S): Incidence of laparoscopically confirmed endometriosis according to birthweight, prematurity, multiple gestation, diethylstilbestrol (DES) exposure, and having been breastfed. RESULT(S): During 566,250 woman-years of follow-up, 1,226 cases of laparoscopically-confirmed endometriosis were reported among women with no past infertility. After adjusting for age, calendar time, parity, race, and body mass index at age 18, we observed a linear increase in the incidence rate with decreasing birthweight (rate ratio [RR] = 1.3 for birthweight <5.5 pounds versus 7.0-8.4 pounds, 95% confidence interval [CI] = 1.0-1.8, P value, test for trend = .01). In addition, women who were born as one of a multiple gestation (i.e., twins or greater number) were at increased risk even after controlling for birthweight (RR = 1.7, CI = 1.2-2.5). The rate of endometriosis was also 80% greater among women exposed to diethylstilbestrol in utero (RR = 1.8, CI = 1.2-2.8). Neither premature delivery nor having been breastfed were associated with the incidence of endometriosis. None of these effect estimates were modified by infertility status at the time of endometriosis diagnosis. CONCLUSION(S): The fetal environment is associated with subsequent laparoscopically confirmed endometriosis in this cohort of US women.