A Follow-up of the Early Pregnancy Study Cohort

We propose a follow-up study of women who participated in the North Carolina Early Pregnancy Study (EPS) in 1982-1986. Women enrolled in the study after discontinuing birth control and were followed for the occurrence of a pregnancy. Participants completed daily diaries and collected first morning urine specimens which allowed the identification of menstrual cycles. Thus each woman s time to pregnancy was prospectively measured. We intend to trace the women and ask them to complete a questionnaire that includes their recall of their time to pregnancy during the study. This will allow us to address two major questions, one regarding the recall of time-to-pregnancy, and the other regarding the natural length of gestation. In addition, we will investigate associations of bisphenal A (BPA), phthalate metabolites, 2,4- and 2,5-dichlorophenols, Triclosan, benzophenone-3, and methyl-, propyl- and butyl-parabens with reproductive outcomes. This is described in a separate section at the end of this protocol because it does not involve recontacting participants, only assay of stored urine samples for purposes within the goals of the original study.

1. Recall of time-to-pregnancy

Time to pregnancy is a commonly studied estimate of fecundability. This outcome has been used to study the reproductive effects of environmental toxins, cigarette smoking, obesity and numerous other exposures. Prospective studies of time to pregnancy require enrollment of women who are just starting to attempt pregnancy and following them until a pregnancy is conceived. This type of study can be time-consuming and costly. It is also difficult to identify a population of women who are just beginning to attempt pregnancy. A more efficient design is to ask women to recall their time to pregnancy. This design is limited however, as it is unknown how well women can recall their time to pregnancy. Only two studies have examined this question in a population with both prospective and retrospectively recalled time to pregnancy. One of the studies involved women from the Netherlands, the recall period was short (3-20 months), and only part of the woman s time to pregnancy was prospectively observed. The second study had a longer recall period (1-10 years) but included only 43 women. 2. Length of gestation

Gestational length is the time from conception to delivery. The EPS has very good data on the first data point, and we know date of delivery. However, our data on delivery is limited by the fact that we do not know which deliveries were induced and which were the product of natural onset of labor and delivery. We plan to collect this information from the mothers. Such information will not only allow us to consider the natural variability in gestational age (a measure that does not exist at present) but also to estimate associations between events in very early pregnancy and the subsequent duration of pregnancy. The EPS has detailed measurements of the time from ovulation to implantation, the rate of increase in human chorionic gonadotropin and corpus luteum rescue. To our knowledge, these early pregnancy characteristics have not been investigated for associations with length of gestation.