Phthalate Exposure and Reproductive Hormone Concentrations in Pregnancy

Abstract

Background

Some phthalate chemicals can affect hormone physiology in utero resulting in adverse reproductive health outcomes in animal models. It is unknown whether these exposures are related to circulating maternal hormone concentrations during pregnancy.

Methods

We used multivariate linear regression to estimate associations between phthalate metabolite concentrations and concurrent serum free and total testosterone and estradiol levels in 180 pregnant women within the Study for Future Families. We also examined associations between prenatal serum hormone concentrations and anogenital outcome in infants. All analyses were adjusted for appropriate confounding variables.

Results

Total testosterone, free testosterone, and estradiol concentrations ranged from 8 to 406 ng/dl, 0.03 to 1.2 ng/dl, and 529 to 40600 pg/ml, respectively. We observed an inverse association between log sum di-2-ethyl hexyl phthalate (DEHP) metabolite concentrations and lower log total testosterone concentrations (−0.15, 95% CI −0.26, −0.04) and log free testosterone (−0.15, 95% CI −0.27, −0.03). This relationship persisted regardless of fetal sex. Similarly, we observed an inverse association between log mono-butyl phthalate (MBP) concentrations and log total and free testosterone concentrations in women carrying male fetuses. Mono-ethyl phthalate (MEP) concentrations were positively associated with log total and free testosterone concentrations in women carrying male fetuses (0.09, 95%CI 0.003, 0.17 and 0.10, 95% CI 0.01, 0.19 respectively). Prenatal hormone concentrations were not significantly associated with infant anogenital outcomes.

Conclusions

Our preliminary data suggest that DEHP metabolite, MBP, and MEP exposures during pregnancy are associated with prenatal sex steroid hormone concentrations, but sex steroid hormone concentrations were not associated with infant reproductive outcomes.

Introduction

Phthalates are synthetic endocrine disrupting chemicals that can affect sex steroid hormone concentrations, signaling, and function during gestation in rodent models (Fisher 2004). Changes in hormone concentrations during this important reproductive programming and developmental window can lead to significant downstream birth defects in offspring within animal models (Gray et al. 2006). In humans, phthalate exposure during pregnancy is related to a number of adverse childhood health outcomes that may by hormone mediated including reduced anogenital distance and changes in sex specific childhood behavior, respectively (Swan et al. 2005, Engel et al. 2009, Engel et al. 2010). Whether phthalate and exposure during pregnancy is related to sex steroid hormone concentrations remains undetermined but may have important implications for fetal development and sex steroid mediated health outcomes.

Exposure to di-2-ethyl hexyl phthalate (DEHP), dibutyl phthalate (DBP), and benzyl butyl phthalate (BBzP) during gestation leads to reductions in testosterone concentrations in both the pregnant dam and in male offspring and subsequent adverse male reproductive abnormalities (Parks et al. 2000, Kavlock et al. 2002, Mylchreest et al. 2002, Foster 2005). In humans, prenatal exposure to DEHP has been negatively associated with free testosterone concentrations in cord blood (Lin et al. 2011). DEHP has been negatively associated with estradiol concentrations in adult female rodents and in vitro in ovarian granulosa cells via decreased aromatase transcription, but this relationship remains to be explored in pregnant women (Laskey & Berman 1993, Davis et al. 1994, Lovekamp-Swan & Davis 2003). Studies examining the impact of diethyl phthalate (DEP) exposure during gestation found reduced testosterone concentrations in offspring in one rodent study, but not with offspring reproductive tract defects (Fujii et al. 2005, Howdeshell et al. 2008).

The maternal, placental, and fetal compartments work in concert to produce circulating reproductive hormones during pregnancy, and the influence of each compartment changes depending on gestational period (Speroff 2005, Strauss JF 2009) Tulchinsky et al. 1972. In early gestation, both the maternal and fetal compartments contribute to circulating hormone concentrations. In later gestation, the fetal adrenal glands provide the precursors for maternal estrogen and androgen production (Speroff 2005) Siiteri et al. 1966. Therefore, circulating maternal hormone concentrations can reflect fetal production, and the influence of the fetal compartment increases as gestation progresses (Speroff 2005, Strauss JF 2009). Total and free testosterone concentrations rise throughout pregnancy but not to the same extent as estrogens, and sources of the increased testosterone are not well characterized but are likely fetal in origin (Bammann et al. 1980, Toriola et al. 2011). The fetus controls estrogen production by producing androgens that are delivered to the placenta and aromatized, and concentrations rise rapidly as pregnancy progresses (Siiteri et al. 1966) (Madden et al. 1978). Results on whether reproductive hormone concentrations differ by fetal sex in the first trimester of pregnancy are conflicting (Klinga et al. 1978, Glass & Klein 1981), but late in gestation, reproductive hormones in the fetal compartment and likely to a great extent, those in the maternal compartment are reflective of fetal production and control (Speroff 2005, Faupel-Badger et al. 2011).

Estradiol and testosterone are important sex steroid hormones that can impact maternal and fetal health (Melmed S 2011). Estrogens are important for many homeostatic functions as well as for normal uteroplacental blood flow, mammary gland development, and fetal adrenal gland function during pregnancy (Speroff 2005, Melmed S 2011). Similarly, testosterone plays a role in several body functions and is important for reproductive function and appropriate male genital tract development during pregnancy (Speroff 2005, Melmed S 2011). The focus of this study is to examine the relationship between phthalate exposure and testosterone and estradiol concentrations in pregnancy as well as the relationship between prenatal hormones and infant reproductive outcomes.

Methods

Study Participants

Pregnant women were originally recruited in the first phase of the Study for Future Families (SFFI), a multicenter pregnancy cohort study, at prenatal clinics in Los Angeles, California, Minneapolis, Minnesota, and Columbia, Missouri, from September 1999 through August 2002. Methods are described in detail elsewhere (Swan et al. 2005). Briefly, couples whose pregnancy was not medically assisted were eligible unless the woman or her partner was < 18 years of age, either partner did not read and speak Spanish or English, or the father was unavailable or unknown. All participants completed a questionnaire and gave blood and urine samples on the same day, and the majority of samples were collected in the second or third trimester of pregnancy (98%). Eligibility criteria for the current analysis included data on prenatal urinary phthalate measurement, prenatal hormone measurement, and complete data on covariates. Anogenital distance (AGD) is defined as the distance from the anus to anterior base of the penis in males and the distance from the anus to the anterior base of the clitoris in females. Anoscrotal distance (ASD) is the distance from the anus to the base of the scrotum, and anofourchette distance (AFD) is the distance from the anus to the base of the posterior fourchette. All measurements were performed in infants born to moms in SFF I according to a standardized protocol described previously (Swan et al 2006). In total, we analyzed data for 180 mothers who met these parameters. SFF obtained human subjects approval, and all participants signed informed consents. The present analyses were deemed exempt for human subjects review, as all data were de-identified.

Serum Hormone Measurements

All hormone measurements were measured at the Endocrine and Metabolic Research Laboratory at Los Angeles Biomedical Research Center at Harbor-UCLA Medical Center using assays that have been fully validated and reported. Serum testosterone was measured by liquid chromatography tandem mass spectrometry (LC-MS/MS) (Shiraishi et al. 2008). Testosterone (>99% pure Sigma Aldrich, St. Louis, MO) was used as a calibration standard. 1, 2 deuterated (D₂)-testosterone (>98% pure, Cambridge Isotope Laboratories, Inc., Andover, MA) was used as the internal standard for T measurements. LC-MS/MS runs were performed with a Shimadzu HPLC system (Columbia, MD) attached to an Applied Biosystems API5000 LC-MS/MS (Foster City, CA) equipped with a TurboIon Spray source. The calibration standards showed a linear response from 1 ng/dl (0.35 nmol/liter) to 2000 ng/dL (69.3 nmol/liter) for testosterone. The within and between run precision was less than 5 % and the recovery of samples spiked with the steroids was between 100 to 113% for testosterone. The lower limit of quantification for testosterone was 2 ng/dl or 0.069 nmol/liter) (Shiraishi S et al. 2008). Free testosterone was measured by equilibrium dialysis using labeled testosterone as described previously (Qoubaitary et al. 2006).

Serum estradiol concentrations for all mothers of boys were measured with a validated LC-MS/MS method (Rothman et al. 2011). Calibration standards and test samples were prepared for LC-MS/MS. First the proteins in the sera were precipitated using acetonitrile. The supernatant was dried and reconstituted with phosphate buffered saline and then extracted with 2ml of diethyl ether. LC-MS/MS were performed with a Shimadzu HPLC system (Columbia, MD) attached to an Applied Biosystems API5000 LC-MS/MS (Foster City, CA) equipped with a TurboIon Spray source. The calibration standards showed a linear response from 1.0pg/ml to 1000pg/ml of estradiol. Analysis of four pools with estradiol concentrations ranging from about 10pg/ml to 500pg/ml showed the within run precision range to be from 2.78% to 3.58%. The between run precision for estradiol (computed over 27 runs) ranged from 3.7% to 9.8%. The recovery of samples spiked with the steroids was between 84.4% to 129% with a mean recovery of 101.8% for E2. The LOQ was 2pg/ml for estradiol. Serum estradiol concentrations for mothers of girls were measured in 2003 in a different laboratory by radioimmunoassay (Pantex, Santa Monica, CA, US). The limit of detection was 4.9 pg/ml and the inter assay coefficient of variation was below 9%. We applied a correction factor to the concentrations of mothers of girls by radioimmunoassay in order to use estradiol concentrations for all mothers in the analysis. The adjustment factor was created based on 156 samples of overlapping estradiol concentrations from 2003 and 2011. The values from 2003 were lower compared to those from 2011 but despite the differing concentrations, the data was highly correlated (0.99). We calculated the estradiol concentrations as [2011 estradiol concentration = 1.33063 × E2(2003) − 0.00001007 [E2(2003)]².

Phthalate Metabolite Measurements

The Division of Laboratory Sciences within the National Center for Environmental Health, Centers for Disease Control, conducted the analyses and had no access to participant data. The analytical method for urinary phthalate metabolites involved the enzymatic deconjugation of the metabolites from their glucuronidated form, followed by concentration of the analytes of interest by solid-phase extraction, separation with high-performance liquid chromatography, and detection by isotope-dilution tandem mass spectrometry (Silva et al. 2004, Kato et al. 2005). This approach allows for the simultaneous quantification in human urine of the following phthalate metabolites reported in this work: MEP, MBP, monomethyl phthalate (MMP), MBzP, mono-isobutyl phthalate (MiBP), mono-2-ethylhexyl phthalate (MEHP), and two oxidative metabolites of DEHP, mono-2-ethyl-5-hydroxyhexyl phthalate (MEHHP), and mono-2-ethyl-5-oxo-hexyl phthalate (MEOHP). Isotopically labeled internal standards and conjugated internal standards were used to increase precision and accuracy of the measurements. Quality control and reagent blank samples were analyzed along with unknown samples to monitor performance of the method. Urinary creatinine was measured at the time of phthalate analysis.

Statistical Analysis

We examined the range of all exposure and outcome variables under consideration including concentration and distribution of maternal serum hormone concentrations and urinary phthalate concentrations. Because the distributions for total testosterone, free testosterone, and estradiol were right skewed, we used log transformations in all statistical analyses. Most phthalate metabolite concentrations were above the LOD, which was between 0.95 – 1.07 μg/L, depending upon the analyte. Concentrations below the LOD were assigned a value equal to the LOD divided by the square root of two (Hornung RW 1990). The percent of samples below the limit of detection (LOD) for each of the phthalate metabolites was <15% except for MMP which was over 50%. We chose not to include MMP in statistical analyses. All phthalate metabolite concentrations were logarithmically transformed to normalize distributions. MEHP, MEOHP, and MEHHP are metabolites of a single parent compound (DEHP). Therefore, we used the molar sum of these metabolites to reflect total DEHP exposure.

We examined the bivariate relationship between covariates chosen a priori and log hormone concentrations. We chose these factors based on past studies examining prenatal hormone concentrations as well as studies examining phthalate exposures. Based on this analysis, we included maternal age, gestational age at blood draw, urinary creatinine as continuous covariates. We included study center, parity, and education as categorical covariates. We used linear regression to explore the associations between log transformed prenatal phthalate metabolite concentrations and concurrent log transformed maternal total testosterone, free testosterone, and estradiol concentrations in individual models.

We fit two models with and without sex-by-phthalate interactions. We reported both sets of results regardless of the significance of the interaction, as others have done (Sagiv et al. 2012). The first model assumes a common slope for boys and girls. The second model allows for separate slopes for boys and girls with the inclusion of a sex by phthalate interaction terms (see Equation 1). We used this model in order to avoid a loss of power from traditional stratification methods.

Sex-Specific Linear Regression model*

$$Y={\beta }_0+\left(\mathit{SEX}\right)\left(\mathit{log}\right(\mathit{MEHP}\left)\right){\beta }_1+(1-\mathit{SEX})\left(\mathit{log}\right(\mathit{MEHP}\left)\right){\beta }_2+\left(\mathit{SEX}\right){\beta }_3+\mathit{COVARIATES}$$

*where sex is coded as 0 = boys, 1 = girls

We used multivariate linear regression to explore the associations between log transformed prenatal hormone concentrations and AGD, ASD, and AFD in male and female infants separately. All analyses were adjusted for weight percentile of infant, age of infant, study center and gestational age at blood draw, and maternal age.

Results

The 180 pregnant women including in the analyses were primarily white (79%) and between ages 20–40 (97%). Women were highly educated (>68% with at least a college degree), and 98% were in their second or third trimester of pregnancy at the time of blood and urine collection (Table 1). Within this group, the highest urinary phthalate concentrations were observed for MEP, and the lowest for MIBP (Table 2). Total testosterone concentrations ranged from 8 to 406 ng/dl, and free testosterone concentrations ranged from 0.03 – 1.2 ng/dl. Estradiol concentrations ranged from 560 to 40,600 pg/ml (log values shown in Figures 1a, 1b, 2a, 2b).

Table 1.

Demographic characteristics (point estimate, 95% CI) for 180 pregnant women

	N (%)
Ethnicity
White	143 (79)
Hispanic	24 (13)
Asian	10 (6)
African-American/Other	3
Maternal Age (years)
<20–30	84 (46)
>30–42	96 (54)
Study Center
MN	83 (86)
CA	44 (24)
MO	53 (29)
Education
Grade School Only/High School Only/Some College	58 (32)
Graduated College or Technical School	64 (36)
Some Graduate Work or Graduate Degree	58 (32)
Parity
Nulliparous	99 (55)
Parous	81 (45)
Gestational Age at Blood Draw (weeks)
0–20	4 (2)
>20–30	68 (38)
>30	108 (60)
Infant Sex in Second Trimester
Male	31 (46)
Female	37 (54)
Infant Sex in Third Trimester
Male	62 (57)
Female	46 (43)

Table 2.

Distribution of Phthalate Metabolite Concentrations in μg/L (N=180)

	Geometric Mean	25%	50%	75%
MBP	16.43	9.20	17.35	54.85
MBzP	9.80	4.50	11.00	38.60
MEP	148.55	47.00	126.40	529.60
MIBP	2.64	1.15	2.70	4.85
SumDEHP metabolites (μM/L)	10.87	5.53	9.99	21.05

figure 1.

figure 2.

In bivariate analysis of participant characteristics and hormone concentrations, increased maternal age and parity were associated with reduced log testosterone and log free testosterone.. Ethnicity was not significantly associated with any hormone concentrations. Having graduated college or technical school was significantly associated with lower testosterone concentrations in pregnant women, but this relationship was not significant in the higher educated women. Increasing gestational age was significantly associated with estradiol concentrations but not testosterone concentrations. In the third trimester of pregnancy, female sex was associated with reduced log total and log free testosterone concentrations and increased log estradiol concentrations (Table 3).

Table 3.

Bivariate relationships between participant characteristics and log hormone concentrations (point estimate, 95% CI) for 180 pregnant women

	Log Total Testosterone	Log Free Testosterone	Log Estradiol
Ethnicity
White	ref	ref	ref
Hispanic	0.15 (−0.26, 0.55)	0.15 (−0.30, 0.60)	0.21 (−0.32, 0.76)
Asian	0.26 (−0.01, 0.53)	0.18 (−0.12, 0.48)	−0.15 (−0.57, 0.28)
African-American/Other	0.29 (−0.43, 1.01)	0.08 (−0.72, 0.88)	0.71 (−0.41, 1.8)
Maternal Age (years)
<20–30	ref	ref	ref
>30–42	−0.31* (−0.49, −0.13)	−0.30* (−0.51, −0.10)	−0.20 (−0.46, 0.05)
Study Center
MN	ref	ref	ref
CA	0.19 (−0.04, 0.42)	0.07 (−0.18, 0.33)	−0.11 (−0.48, 0.25)
MO	−0.04 (−0.26, 0.17)	−0.08 (−0.32, 0.16)	−0.08 (−0.37, 0.20)
Education
Grade School Only/High School Only/Some College	ref	ref	ref
Graduated College or Technical School	−0.32* (−0.54, −0.10)	−0.29* (−0.54, −0.05)	0.09 (−0.23, 0.41)
Some Graduate Work or Graduate Degree	−0.15 (−0.37, 0.08)	−0.10 (−0.35, 0.16)	−0.01 (−0.31, 0.32)
Parity
Nulliparous	ref	ref	ref
Parous	−0.40* (−0.57, −0.22)	−0.36* (−0.56, −0.16)	−0.27* (−0.52, −0.12)
Gestational Age at Blood Draw (weeks)
0–24	ref	ref	ref
>24–41	0.03 (−0.16, 0.22)	0.08 (−0.13, 0.29)	1.10* (0.91, 1.30)
Infant Sex in Second Trimester
Male	ref	ref	ref
Female	0.09 (−0.28, 0.46)	−0.30 (−0.66, 0.06)	0.16 (−0.22, 0.54)
Infant Sex in Third Trimester
Male	ref	ref	ref
Female	−0.23* (−0.44, −0.02)	−0.48* (−0.71, −0.24)	0.04 (−0.15, 0.24)

^*Indicates p value less than 0.05

Within the model that estimated sex specific slopes, we observed a lower log total concentration in relation to a log unit increase in the sum of DEHP metabolites in women with both male (−0.07, 95% CI −0.20, 0.06) and female fetuses (−0.15, −0.26, −0.04) but this result was only statistically significant for women carrying female fetuses (Table 4). We observed similar results with log free testosterone in relation to the sum of DEHP metabolites. We observed some evidence for effect modification for MBP and MBZP in relation to log total and free testosterone concentrations with positive estimates for women with male fetuses and negative estimates for women with female fetuses. We also observed a higher log total testosterone concentration in relation to a log unit increase in MEP (0.09, 95% CI 0.003, 0.17) in women with male fetuses only (Table 4). None of the prenatal log hormone concentrations were associated with AGD, ASD, or AFD in male or female infants (Table 5).

Table 4.

Linear regression analysis with sex specific slopes of phthalate metabolites in relation to log total testosterone, free testosterone, and estradiol concentrations in pregnant women (N=180)

		Women with Male Fetuses (N=94)			Women with Female Fetuses (N=86)
Serum Hormone	Phthalate or BPA Metabolite	Coefficient	95% CI	p value	Coefficient	95% CI	p value
Log Total
Testosterone	Log mbp	0.15	−0.04, 0.33	0.12	−0.20	−0.39, −0.01	0.04
	Log mbzp	0.06	−0.07, 0.19	0.37	−0.13	−0.26, 0.01	0.06
	Log mep	0.09	0.003, 0.17	0.04	0.03	−0.06, 0.11	0.56
	Log mibp	−0.03	−0.18, 0.13	0.75	−0.10	−0.28, 0.07	0.23
	Log sumdehp	−0.07	−0.20, 0.06	0.30	−0.15	−0.26, −0.04	0.01
Log Free
Testosterone	Log mbp	0.13	−0.07, 0.33	0.22	−0.21	−0.42, 0.004	0.05
	Log mbzp	0.07	−0.07, 0.21	0.35	−0.10	−0.25, 0.04	0.16
	Log mep	0.10	0.01, 0.19	0.04	−0.02	−0.10, 0.08	0.75
	Log mibp	−0.03	−0.20, 0.14	0.77	−0.11	−0.30, 0.08	0.26
	Log sumdehp	−0.04	−0.18, 0.10	0.55	−0.15	−0.27, −0.03	0.01
Log Estradiol	Log mbp	0.04	−0.10, 0.18	0.57	−0.002	−0.18, 0.17	0.97
	Log mbzp	−0.03	−0.12, 0.07	0.62	−0.10	−0.23, 0.03	0.14
	Log mep	−0.03	−0.10, 0.04	0.38	0.01	−0.07, 0.10	0.78
	Log mibp	0.003	−0.12, 0.12	0.95	0.03	−0.14, 0.20	0.73
	Log sumdehp	−0.06	−0.17, 0.04	0.21	−0.08	−0.18, 0.01	0.08

Table 5.

Linear regression analysis slopes of anogenital outcomes in infants in relation to log total testosterone, free testosterone, and estradiol concentrations in pregnant women (N=180)

		Male Newborns (N=94)			Female Newborns (N=86)
Anogenital Outcome	Serum Hormone	Coefficient	95% CI	p value	Coefficient	95% CI	p value
AGD	Log Testosterone	−1.50	−4.55, 1.55	0.33	0.33	−1.17, 1.83	0.66
	Log Free Testosterone	−1.16	−3.98, 1.67	0.42	0.30	−1.09, 1.70	0.67
	Log Estradiol	1.98	−1.44, 5.39	0.25	−0.43	−3.54, 2.68	0.78
ASD/AFD	Log Testosterone	−0.18	−3.37, 3.73	0.92	−0.56	−2.44, 1.32	0.56
	Log Free Testosterone	0.04	−3.24, 3.33	0.98	−0.33	−2.08, 1.43	0.71
	Log Estradiol	1.38	−2.60, 5.35	0.49	−0.97	−4.69, 2.75	0.60

Discussion

We report an association between increased prenatal DEHP metabolite exposure and lower testosterone concentrations in women carrying both male and female fetuses, and MBP was also associated with lower testosterone concentrations but only in women carrying female fetuses. We found the opposite relationship with increasing MEP exposure associated with higher testosterone concentrations, and this relationship was strongest in women with male fetuses. We did not find an association between prenatal hormone concentrations and infant anogenital outcomes. These findings are preliminary based on a small sample size but suggest that exposure to some phthalates during pregnancy may affect sex steroid hormone concentrations.

Free testosterone is unbound and is bioavailable to peripheral tissues while total testosterone reflects both bound and free testosterone concentrations (Speroff 2005). The pearson correlation coefficient between the total and free testosterone was > 0.9, and regression results were similar for both which supports the validity of both the hormonal and statistical analyses. The origin of prenatal circulating testosterone concentrations in humans is unclear but is thought to be primarily directed by the fetus as pregnancy progresses via fetal testicular Leydig cell production (Bammann et al. 1980, Kerlan et al. 1994). In women carrying female fetuses, testosterone is likely produced in similar manner to non-pregnant women, through ovarian, adrenal gland, and peripheral production (Speroff 2005). From our results, it may be that DEHP metabolites are both exerting an effect on adrenal as well as Leydig cell production of testosterone because we see reductions in testosterone in women carrying both male and female fetuses. Testosterone plays an important role in sex differentiation during early human gestation. In male fetuses, increasing testosterone concentrations lead to virilization and normal development of male genital structures. In females, estrogen is the predominant sex steroid hormone and increased testosterone concentrations can lead to abnormal virilization (Speroff 2005). In addition, normal estrogen and testosterone play an integral role in an adult female's general endocrine homeostasis and are known to be associated with a variety of health outcomes (Longcope 1986).

In animal studies, DEHP and DBP exposure in pregnant dams leads to lower intra-testicular testosterone concentrations in male offspring via a direct testicular toxic effect on Leydig cells (Parks et al. 2000, Foster 2005, Howdeshell et al. 2008). Therefore, our study may suggest that DEHP could be acting via a direct testicular toxic effect on fetal Leydig cells leading to reductions in circulating maternal testosterone concentrations. It is unclear whether DEHP exposure could also affect ovarian, adrenal, or androstenedione production of testosterone as there are no studies of this mechanism in the literature.

Several studies suggest that DEP is not as potent an endocrine disrupting chemical as DEHP or DBP, but one study did observe reduced testosterone concentrations in male offspring of pregnant dams exposed to DEP at 3000 and 15000ppm during gestation (Gray et al. 2000, Fujii et al. 2005, Howdeshell et al. 2008), and some human epidemiologic studies have found relationships between DEP exposure and adverse endocrine and reproductive health outcomes (Lopez-Carrillo et al. 2010, Tranfo et al. 2011). Results of our study suggest that DEP may be acting via novel mechanism in relation to increased prenatal circulating testosterone concentrations given that we observed an increase in testosterone in women carrying male fetuses. Another possibility is that the DEP finding is due to chance alone.

During gestation, estradiol production occurs primarily in the placenta through aromatization of androgens (Speroff 2005). Benzyl butyl phthalate (BBzP) and DBP have been associated with weak estrogenic activity but other studies have documented an anti-estrogenic effect in peripheral tissues (Jobling et al. 1995, Lee et al. 2004). In two rodent studies, adult female circulating estrogen concentrations were decreased in relation to high dose DEHP exposure (Davis et al. 1994, Hirosawa et al. 2006). Overall, the impact of phthalate exposure on estrogen steroidogenesis is not well understood with conflicting results in the literature. We observed the suggestion of lower estradiol concentrations in relation to increased DEHP exposure (as well as in the sex specific analyses) but this association was not statistically significant.

We examined the relationship between prenatal sex steroid hormones and sex-specific AGD within the SFF dataset but did not observe any statistically significant relationships. However, this analysis was limited in several ways including exposure misclassification as well as lack of precision in outcomes. The timing of serum collection was not appropriate. The genital reproductive programming window is early in pregnancy and the majority of the SFF (>90%) serum samples were taken late in the second trimester or within the third trimester of pregnancy when they are less likely to have impacted genital development. The age of the infants at AGD exam was highly variable (9–36 months), making it more difficult to detect associations if they existed. Future studies would need to assess hormone concentrations in the first trimester of pregnancy when genital structures develop and then examine genital outcomes at birth.

Limitations of our study included spot urine samples to reflect urinary phthalate exposures. Although there can be variability in phthalate concentrations over time, several studies suggest that a spot sample can be representative of a three month period of exposure in adult women (Hoppin et al. 2002). One study of phthalate variability in pregnancy showed that DEHP metabolites decreased as the pregnancy progressed while MEP and MiBP concentrations increased (Braun et al 2012). If these findings are generalizable, urine DEHP metabolite concentrations in our study may have been an underestimate of exposure during the first trimester while MEP concentrations could have been an overestimate. Urine and serum were collected at the same time for each woman. Ideally, we would like to have collected serum after the urine samples as it is not clear whether an effect would be seen concurrently or after hours or days. We do not have multiple measurements within the same women which would allow for longitudinal analysis to examine change in concentrations. We have samples from different time points during pregnancy and attempted to correct for this by adjusting for gestational age at blood draw in our statistical analyses. Regardless, residual confounding by covariates in the analysis could introduce bias in the analysis. We conducted two sensitivity analyses. The first excluded the first trimester samples (N=4), and results were unchanged. We then restricted to women in the third trimester only (N=96) and found that all estimates were in the same direction and of similar magnitude to the primary analysis. We were unable to directly assess fetal hormone concentrations in this study, and methodology to do this (amniotic fluid collection) would not be feasible in the general population.

Conclusion

Exposure to DEHP, MBP, and MEP during pregnancy may be related to changes in circulating prenatal testosterone concentrations. If confirmed, these findings are significant because testosterone is involved in the development of maternal and infant health outcomes including cardiovascular disease, hormone mediated reproductive disease, neurodevelopmental outcomes, and fetal genital tract development. These results are preliminary and should be confirmed in future larger human cohort studies.

List of Abbreviations/Definitions

CDC

Centers for Disease Control

LOD

limit of detection

DEP

diethyl phthalate

DEHP

di-2-ethyl hexyl phthalate

DBP

dibutyl phthalate

BBzP

butylbenzyl phthalate

MEP

monoethyl phthalate

MBP

mono-n-butyl phthalate

MBzP

monobenzyl phthalate

MiBP

monoisobutyl phthalate

MMP

monomethyl phthalate

MEHP

mono-2-ethyl hexyl phthalate

MEOHP

mono-2-ethyl-5-oxo-hexyl phthalate

MEHHP

mono-2-ethyl-5-hydroxyhexyl phthalate

ng/dL

nanograms/decililiter

nmol/liter

nanomoles/liter

pg/ml

picograms/milliliter

μg/L

microgram/L