This is the first study to document measurable concentrations of BPA in the urine of a Mexican population. Furthermore, the present study suggests, based on a single spot urine sample collected during the third trimester, that women who delivered less than or equal to 37 weeks of gestation and prematurely (< 37 weeks) had higher concentrations of BPA in their urine compared to women delivering after 37 weeks.
Geometric mean urinary BPA concentrations (1.52 μg/L uncorrected, 1.95 μg/g Cr) in this pilot analysis from the ELEMENT study were very similar to concentrations among pregnant women in the Netherlands (1.1 μg/L uncorrected, 1.7 μg/g Cr) [7
], New York City (1.3 μg/L uncorrected) [21
], Cincinnati (2.0 μg/L uncorrected, 2.2 μg/g Cr) [25
], and women in the general US population from NHANES 2003-2004 (2.4 μg/L, 2.8 μg/g Cr) [6
]. The geometric mean concentration in our study was approximately one-half of that reported in a small (n = 10) subset of the Norwegian Mother and Child Cohort Study (4.5 μg/L uncorrected, 5.9 μg/g Cr) [11
To date there has been limited human data on preterm birth, gestational length, and prenatal BPA exposure. In the Children's Environmental Health study, Wolff et al. found no association between BPA exposure during the third trimester and gestational length among 404 pregnant women in New York City [21
]. In addition, a small scale study (N = 40) in Southeastern Michigan that measured BPA in blood of women at the time of delivery also found no difference in gestational length between women with blood BPA concentrations > 5 and ≤ 5 ng/mL[20
]. However, unlike the present study, these studies were not designed to assess preterm birth in relation to BPA exposure. Other small-scale human studies of BPA exposure in pregnancy have reported increases in the risk of spontaneous abortion [26
], sister chromatid exchanges [27
], and abnormal fetal karyotype [28
] with increased BPA concentrations. As far as we are aware, this is the first study to specifically assess the relationship between BPA exposure biomarker concentrations during pregnancy and preterm birth.
Preterm birth is likely a condition with multiple etiologies [29
]. BPA may harm fetal growth and promote early parturition through various mechanisms, as it has been shown to disrupt a variety of biologic functions including steroid hormone synthesis and metabolism [30
], peroxisome proliferation [32
], cytokine networks [18
], genotoxicity [33
], and oxidative stress [36
]. Low doses of BPA also induced apoptosis and increased output of matrix metalloproteinase-9, an enzyme associated with preterm birth, in ovarian granulosa cells in dose-dependent patterns [41
]. Recently, it has been shown that human primary cytotrophoblast cells undergo a dose-dependent increase in TNF-α production and apoptosis with increasing environmentally relevant (0.0002 to 0.2 ug/mL) levels of BPA [18
]. Supporting these initial results, Morck et. al. also demonstrated that similar levels of BPA exposure can induce cell death in a human choriocarcinoma cell line and increase secretion of β-hCG and caspase-3 cleavage in first trimester human chorionic villous explant cultures [17
Our study has several limitations including its small sample size. Gestational length was estimated by date of maternally-recalled last menstrual period which may be an unreliable measure, varying as much as ± 7-21 days, depending on a host of factors including nutrition, physical activity, smoking, alcohol consumption, stress, and inter-pregnancy interval [43
]. Because urinary BPA was only measured at a single time point in this study, the potential for measurement error exists since exposure is likely to be variable over time [47
]. However, unless errors in last menstrual period recall were associated with BPA urinary concentrations and/or measurement error of BPA concentrations was systematic in relation to gestational age, one would expect that, on average, any bias in effect estimates would be toward the null.